The Oxford Handbook of Eating Disorders

OX F O R D L I B R A RY O F P S YC H O LO G Y

Area Editors

Clinical Psychology
David H. Barlow

Cognitive Neuroscience
Kevin N. Ochsner and Stephen M. Kosslyn

Cognitive Psychology
Daniel Reisberg

Counseling Psychology
Elizabeth M. Altmaier and Jo-​Ida C. Hansen

Developmental Psychology
Philip David Zelazo

Health Psychology
Howard S. Friedman

History of Psychology
David B. Baker

Methods and Measurement

Todd D. Little

Neuropsychology
Kenneth M. Adams

Organizational Psychology
Steve W. J. Kozlowski

Personality and Social Psychology

Kay Deaux and Mark Snyder
 OXFORD LIBRARY OF PSYCHOLOGY

The Oxford Handbook

of Eating Disorders
Second Edition
Edited by
W. Stewart Agras
Athena Robinson

1
1
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.

Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.

© Oxford University Press 2018

First Edition published in 2010

Second Edition published in 2018

All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.

You must not circulate this work in any other form

and you must impose this same condition on any acquirer.

Library of Congress Cataloging-​in-​Publication Data

Names: Agras, W. Stewart, editor. | Robinson, Athena Hagler, editor.
Title: The Oxford handbook of eating disorders / edited by W. Stewart Agras and Athena Robinson.
Description: Second edition. | Oxford ; New York : Oxford University Press, [2018] |
Series: Oxford library of psychology | Includes bibliographical references and index.
Identifiers: LCCN 2017044310 (print) | LCCN 2017044448 (ebook) |
ISBN 9780190621018 (updf ) | ISBN 9780190662721 (epub) | ISBN 9780190620998
Subjects: LCSH: Eating disorders.
Classification: LCC RC552.E18 (ebook) | LCC RC552.E18 O97 2018 (print) |
DDC 616.85/26—dc23
LC record available at https://lccn.loc.gov/2017044310

9 8 7 6 5 4 3 2 1
Printed by Sheridan Books, Inc., United States of America
S H O RT CO N T E N T S

About the Editors  vii

Contributors ix

Contents xiii

Chapters 1–524

Index 525

v
A B O U T T H E   E D I TO R S

W. Stewart Agras
W. Stewart Agras, MD, is professor emeritus in the Department of Psychiatry
and Behavioral Sciences at Stanford University. He was editor of the Journal of
Applied Behavior Analysis and the Annals of Behavioral Medicine and president
of the Association for Behavioral and Cognitive Therapies and the Society of
Behavioral Medicine. He has been working in the field of eating disorders for the
past 30 years, focusing on the treatment of anorexia nervosa, bulimia nervosa,
and binge eating disorder, and continues an active research program at Stanford.

Athena Robinson
Athena Robinson, PhD, is a clinical associate professor in the Department of
Psychiatry and Behavioral Sciences at Stanford University’s School of Medicine.
Her core areas of programmatic research include treatment outcome and imple-
mentation of evidence based treatments for eating disorders. She is an attending
faculty member in Stanford’s Eating Disorders Clinic and Dialectical Behavior
Therapy Program, and is the clinic’s liaison with Stanford campus, including
Student Health Services and Athletics. She teaches and supervises graduate stu-
dents and postdoctoral fellows.

vii
 CO N T R I B U TO R S

W. Stewart Agras Eunice Y. Chen

Department of Psychiatry and Behavioral Department of Psychology
Sciences Temple University
Stanford University School of Medicine Philadelphia, PA
Stanford, CA Ross Crosby
Kelly C. Allison Neuropsychiatric Research Institute
Department of Psychiatry Department of Clinical Neuroscience
Perelman School of Medicine at the University of North Dakota School of
University of Pennsylvania Medicine and Health Sciences
Philadelphia, PA Fargo, ND
Drew A. Anderson Scott J. Crow
Department of Psychology Department of Psychiatry
SUNY at Albany University of Minnesota School of
Albany, NY Medicine
Eileen Anderson-​Fye Minneapolis, MN
Department of Bioethics Cortney Dable
Case Western Reserve University Department of Psychology and
Cleveland, OH Neuroscience
Courtney Arena Duke University
Department of Psychology and Durham, NC
Neuroscience Antonios Dakanalis
Duke University Department of Brain and Behavioral
Durham, NC Sciences
Cara Bohon University of Pavia
Department of Psychiatry and Behavioral Pavia, Italy
Sciences Alison M. Darcy
Stanford University School of Medicine Department of Psychiatry and Behavioral
Stanford, CA Sciences
Cynthia M. Bulik Stanford University School of Medicine
Department of Psychiatry Stanford, CA
School of Medicine Martina de Zwann
University of North Carolina Clinic for Psychosomatics and
Chapel Hill, NC Psychotherapy
Natasha L. Burke Hannover Medical School
Department of Medical and Clinical Hannover, Germany
Psychology Joseph Donahue
Uniformed Services University of Health Department of Psychology
Sciences SUNY at Albany
Bethesda, MD Albany, NY

ix
Valerie J. Douglas Sasha Gorrell
Neuropsychiatric Research Institute Department of Psychology
Department of Clinical Neuroscience SUNY at Albany
University of North Dakota School of Albany, NY
Medicine and Health Sciences Anna I. Guerdjikova
Fargo, ND Lindner Center of HOPE, Mason, OH
Alice V. Ely Department of Psychiatry & Behavioral
Department of Psychiatry Neuroscience
University of California, San Diego University of Cincinnati College of
La Jolla, CA Medicine
Lauren E. Ehrlich Cincinnati, OH
Department of Psychology José Gutiérrez-​Maldonado
SUNY at Albany Department of Clinical Psychology and
Albany, NY Psychobiology
Marta Ferrer-​García University of Barcelona
Department of Clinical Psychology and Spain
Psychobiology Katherine A. Halmi
University of Barcelona Department of Psychiatry
Spain Weill Cornell Medical College
Eike Fittig Cornell University
Institut fur Klinische Psychologie und White Plains, NY
Psychotherapie Amy Harrison
Technische University Dresden University College London
Dresden, Germany London, UK
Ellen E. Fitzsimmons-​Craft Caroline E. Haut
Department of Psychiatry Department of Psychiatry
Washington University School of University of Minnesota
Medicine Minneapolis, MN
St. Louis, MO Anja Hilbert
E. Leigh Gibson Department of Medical Psychology and
Clinical and Health Psychology Medical Sociology
Research Center University of Leipzig Medical Center
Department of Psychology Leipzig, Germany
University of Roehampton Jasmine Hill
London, UK Department of Psychology and
Neha J. Goel Neuroscience
Department of Psychiatry and Duke University
Behavioral Sciences Durham, NC
Stanford University School of Medicine Jill L. Holm-​Denoma
Stanford, CA Department of Psychology
Kathryn H. Gordon University of Denver
Neuropsychiatric Research Institute Denver, CO
Department of Clinical Neuroscience Caroline Hubble
University of North Dakota School of Department of Psychology and
Medicine and Health Sciences Neuroscience
Fargo, ND Duke University
Durham, NC

x Contributors
Kristian Hütter James E. Mitchell
Institut fur Klinische Psychologie und Neuropsychiatric Research Institute
Psychotherapie Fargo, ND
Technische University Dresden Nicole Mori
Dresden, Germany Lindner Center of HOPE, Mason, OH
Corinna Jacobi Department of Psychiatry & Behavioral
Institut fur Klinische Psychologie und Neuroscience
Psychotherapie University of Cincinnati College of
Technische University Dresden Medicine
Dresden, Germany Cincinnati, OH
Anna M. Karam Lisa Opitz
Department of Psychiatry Department of Medical Psychology and
Washington University School of Medical Sociology
Medicine University of Leipzig Medical Center
St. Louis, MO Leipzig, Germany
Walter H. Kaye Molly Orcutt
Department of Psychiatry Neuropsychiatric Research Institute
University of California, San Diego Fargo, ND
La Jolla, CA Carol B. Peterson
Paul E. Keck Jr. Department of Psychiatry
Department of Psychiatry University of Minnesota
University of Cincinnati College of Minneapolis, MN
Medicine Emily M. Pisetsky
Cincinnati, OH Department of Psychiatry
Pamela K. Keel University of Minnesota
Department of Psychology Minneapolis, MN
Florida State University Renne Rienecke
Tallahassee, FL Department of Pediatrics and
Daniel Le Grange Department of Psychiatry and
Department of Psychiatry Behavioral Sciences
University of California, San Francisco Medical University of South Carolina
San Francisco, CA Charleston, SC
Jennifer D. Lundgren Department of Psychiatry
Department of Psychology University of Michigan
University of Missouri, Kansas City Ann Arbor, MI
Kansas City, MO Giuseppe Riva
Annika P. C. Lutz Centro Studi e Ricerche di Psicologia
Universite du Luxembourg della Comunicazione
Luxembourg Università Cattolica del Sacro Cuore
Susan L. McElroy Milan, Italy
Lindner Center of HOPE, Mason, OH Athena Robinson
Department of Psychiatry & Behavioral Department of Psychiatry and Behavioral
Neuroscience Sciences
University of Cincinnati College of Stanford University School of Medicine
Medicine Stanford, CA
Cincinnati, OH
Philip S. Mehler
University of Colorado
Denver, CO

Contributors xi
Shiri Sadeh-​Sharvit Tracy D. Wade
Department of Psychiatry and Behavioral Department of Psychology
Sciences Flinders University
Stanford University School of Medicine Adelaide, South Australia
Stanford, CA Denise E. Wilfley
Debra L. Safer Department of Psychiatry
Department of Psychiatry and Behavioral Washington University School of
Sciences Medicine
Stanford University School of Medicine St. Louis, MO
Stanford, CA G. Terence Wilson
Heather Shaw Graduate School of Applied and
Oregon Research Institute Professional Psychology
Eugene, OR Rutgers University
Emilie Sohl Piscataway, NJ
Department of Psychology and Stephen A. Wonderlich
Neuroscience Neuropsychiatric Research Institute
Duke University Department of Clinical Neuroscience
Durham, NC University of North Dakota School of
Kristine Steffen Medicine and Health Sciences
Neuropsychiatric Research Institute Fargo, ND
Fargo, ND Angelina Yiu
Eric Stice Department of Psychology
Oregon Research Institute Temple University
Eugene, OR Philadelphia, PA
Marian Tanofsky-​Kraff Jee Yoon
Department of Medical and Clinical Department of Psychology and
Psychology Neuroscience
Uniformed Services University of the Duke University
Health Sciences Durham, NC
Bethesda, MD Nancy Zucker
C. Barr Taylor Department of Psychiatry and Behavioral
Department of Psychiatry and Behavioral Sciences
Sciences Duke University
Stanford University School of Medicine Durham, NC
Stanford, CA
Claus Vögele
Institute for Health and Behaviour
University of Luxembourg
Luxembourg

xii Contributors
CONTENTS

Introduction 1
W. Stewart Agras and Athena Robinson

Part One  • Phenomenology and Epidemiology 

1. The Classification of Eating Disorders  9
Kathryn H. Gordon, Jill M. Holm-​Denoma, Valerie J. Douglas, Ross Crosby,
and Stephen A. Wonderlich
2. Research Domain Criteria: The Impact of RDoC on the
Conceptualization of Eating Disorders  24
Cara Bohon
3. Epidemiology and Course of Eating Disorders  34
Pamela K. Keel

Part Two  • Approaches to Understanding the Eating

Disorders 
4. Appetitive Regulation in Anorexia Nervosa and Bulimia Nervosa  47
Alice V. Ely and Walter H. Kaye
5. Genetic Influences on Eating Disorders  80
Tracy D. Wade and Cynthia M. Bulik
6. Psychosocial Risk Factors for Eating Disorders  106
Corinna Jacobi, Kristian Hütter, and Eike Fittig
7. Dieting and the Eating Disorders  126
Eric Stice and Heather Shaw
8. Mood, Emotions, and Eating Disorders  155
Claus Vögele, Annika P. C. Lutz, and E. Leigh Gibson
9. Cultural Influences on Body Image and Eating Disorders  187
Eileen P. Anderson-Fye

Part Three  • Assessment and Comorbidities of the Eating

Disorders 
10. Psychological Assessment of the Eating Disorders  211
Drew A. Anderson, Joseph Donahue, Lauren E. Ehrlich, and Sasha Gorrell
11. Medical Complications of Anorexia Nervosa and Bulimia Nervosa  222
Philip S. Mehler
12. Psychological Comorbidities of Eating Disorders  229
Katherine A. Halmi

xiii
 Part Four  • Prevention and Treatment 
13. Prevention: Current Status and Underlying Theory  247
C. Barr Taylor, Ellen E. Fitzsimmons-​Craft, and Neha J. Goel
14. Cognitive-Behavioral Therapy for Eating Disorders  271
G. Terence Wilson
15. Interpersonal Psychotherapy for the Treatment of Eating Disorders  287
Natasha L. Burke, Anna M. Karam, Marian Tanofsky-​Kraff,
and Denise E. Wilfley
16. Family Therapy for Eating Disorders  319
Daniel Le Grange and Renne Rienecke
17. Dialectical Behavior Therapy and Emotion-Focused Therapies
for Eating Disorders  334
Eunice Y. Chen, Angelina Yiu, and Debra L. Safer
18. Self-​Help and Stepped Care Treatments for Eating Disorders  351
Carol B. Peterson, Emily M. Pisetsky, and Caroline E. Haut
19. Pharmacotherapy for Eating Disorders  359
Susan L. McElroy, Anna I. Guerdjikova, Nicole Mori, and Paul E. Keck Jr.
20. Cognitive Remediation Therapy for Eating Disorders  395
Amy Harrison
21. Costs and Cost-​Effectiveness in Eating Disorders  410
Scott J. Crow

Part Five  • Emerging Topics 

22. Selective Eating: Normative Developmental Phase or Clinical
Condition?  419
Nancy Zucker, Courtney Arena, Cortney Dable, Jasmine Hill,
Caroline Hubble, Emilie Sohl, and Jee Yoon
23. Emerging Syndromes  438
Kelly C. Allison and Jennifer D. Lundgren
24. Eating Disorders and Problematic Eating Behaviors
After Bariatric Surgery  458
Molly Orcutt, Kristine Steffen, and James E. Mitchell
25. Virtual Reality: Applications to Eating Disorders  470
José Gutiérrez-​Maldonado, Marta Ferrer-​García, Antonios Dakanalis, and
Giuseppe Riva
26. Mobile Device Applications for the Assessment and Treatment of
Eating Disorders  492
Alison M. Darcy and Shiri Sadeh-​Sharvit
27. Internet-​Based Interventions for Eating Disorders  505
Anja Hilbert, Lisa Opitz, and Martina de Zwann
Afterword  520
W. Stewart Agras and Athena Robinson

Index  525

xiv Contents
 Introduction

W. Stewart Agras and Athena Robinson

Abstract
This chapter provides a brief introduction to and overview of the contents of the Handbook. Several
issues are highlighted, including changes since the previous edition of this volume, namely, the revised
Diagnostic and Statistical Manual (DSM-​5); the research domain criteria (RDoC), and recent technological
innovations such as Internet treatment and the use of virtual reality related to eating disorders. Chapters
on selective eating, bariatric surgery, and cognitive remediation have also been added. Themes carried
forward from the previous edition of the Handbook are presented in updated chapters reviewing
etiological, maintenance, assessment, comorbidity, medical complications, and pharmacotherapy, as well
as evidence-​based prevention and treatment considerations.
Key Words:  classification, diagnosis, eating disorder, history, overview, treatment, technology

Introduction extends into adult life, eating disorders in relation to

This book is divided into five sections: phenom- bariatric surgery and the use of technology in treat-
enology and epidemiology of the eating disorders; ment and assessment.
approaches to understanding the disorders; assess-
ment and comorbidities of the disorders; preven- History of the Eating Disorders
tion and treatment; and emerging topics. The first Whether the eating disorders have historical con-
section deals with classification and epidemiology tinuity has been much debated (Habermas, 2005;
of the disorders, together with a consideration of Keel & Klump, 2003). Unfortunately, the histori-
the effect of the impact of the new research domain cal record does not provide sufficiently detailed case
criteria (RDoC) promoted by the National Institute descriptions to enable certain diagnosis. It is clear
of Mental Health, on the conceptualization of that self-​starvation and self-​induced vomiting, com-
the eating disorders. The second section describes bined with asceticism and religious preoccupations
research basic to understanding the eating disorders apparently driving these symptoms, were present in
including biological, psychosocial risk factors, diet- medieval times (Bynum, 1987; Harrison, 2003), as
ing, and mood in the genesis of eating disorders. well as cases of binge eating, often on strange foods,
The third section describes assessment of the eating which is probably why they were recorded. Opinion
disorders, medical and psychological comorbidi- is divided as to whether such individuals would
ties, and medical management. The fourth section meet present-​day diagnostic criteria for an eating
deals with prevention and treatment of the eating disorder or whether true eating disorder syndromes
disorders including psychotherapeutic and psycho- emerged only in the 19th and 20th centuries when
pharmacologic approaches and a consideration of detailed case histories became available (Habermas,
the cost-​ effectiveness of existing treatments. The 2005). Given the biological underpinnings of the
final section contains chapters on emerging topics eating disorders, for example, heritability, two
such as selective eating, a childhood problem that explanations come to mind: First, eating disorders

1
such as anorexia nervosa (AN) and bulimia nervosa disorder, the criteria for the diagnoses of BN and
(BN) may have been present throughout the centu- BED were loosened, and amenorrhea was removed
ries but the historical record is insufficient to fully as a criterion for AN. The RDoC are aimed at cor-
confirm this possibility. Second, cultural conditions recting the problem that research to date has failed
changed at some point, interacting with the genetic to produce enough knowledge about psychopatho-
component, to produce full-​fledged eating disor- logic processes useful for the prevention and treat-
der syndromes. One possible cultural change is the ment of mental disorders. One reason for this is that
focus on weight and shape together with attempts the diagnoses emanating from clinician consensus
to alter these features that became increasingly com- described in the DSM may not represent actual
mon in young women from the mid-​19th century entities, hence approaches to understanding these
onward (Habermas, 2005). “disorders” have little chance of discovering useful
Hence, there are good descriptions of AN psychopathologic processes. The RDoC approach
beginning in the mid-​19th century (Gull, 1874) aims to uncouple DSM diagnoses from research
although BN was first described in detail much later questions and to consider domains that may stretch
(Russell, 1979) and BED, a provisional diagnosis in across the disorders classified in the DSM. This
DSM-​IV, only became a full disorder in DSM-​5. radical approach has spurred a great deal of discus-
Moreover, the impetus for research in BN was the sion and some dissent. The third change is in the
increase in cases seen in North American clinics in use of technologic advances that have led to the
the mid-​1970s. development of Internet-​based treatment for the
Systematic study of the eating disorders began eating disorders together with mobile applications
in the last third of the 20th century, although AN (apps) as aids to therapy allowing for improved self-​
had been described in the 19th century and vari- monitoring. Other advances such as the use of vir-
ous treatments for that disorder were tried, none tual reality for assessment and treatment of some
of them particularly successful, during the next aspects of the eating disorders are also examined.
100 years. Since the 1970s research into the eating Each of these changes is likely to alter our per-
disorders has grown exponentially. The first issue ception and understanding of the eating disorders
of the International Journal of Eating Disorders, the and their treatment in the years to come.
premier journal in the field, appeared in the fall of
1981, encouraging further research and other jour- The Eating Disorders: Boundary Problems
nals including Eating Behaviors; Eating: The Journal One problem in classifying the eating disorders
of Treatment and Prevention; Eating and Weight is that the disorders tend to merge over time. For
Disorders; Journal of Eating Disorders; and the example, it is not uncommon for patients with AN
European Eating Disorders Review have followed. As to begin to binge eat and purge, thus meeting criteria
the field is now maturing, the purpose of this vol- for BN when they no longer meet weight criteria for
ume is to update the state of treatment and research. AN. Indeed, about 25% of participants with BN in
The relatively recent recognition of the eating treatment trials had been diagnosed with AN in the
disorders means that research has lagged behind past (Agras, Walsh, Fairburn, Wilson, & Kraemer,
that of more established fields such as depression 2000; Fairburn & Cooper, 2011). Such individuals
and anxiety disorders. For example, research on tend to have worse treatment outcomes than those
treatment of BN only began in the late 1970s with who have not had past AN, suggesting that the
both pharmacologic and psychotherapeutic studies psychopathologic processes active in AN continue
(Fairburn, 1981; Pope & Hudson, 1982; Schneider to affect outcome. To a lesser extent, there is cross-
& Agras, 1985; Wermuth, Davis, Hollister, & over between BN and BED. When there is a shift
Stunkard, 1977). between syndromes, the question arises: Should the
diagnosis change or should it remain in the previous
Recent Changes diagnostic grouping? Although there is considerable
Three important events have occurred between controversy over this point, it would seem sensible
the editions of this volume. First, DSM-​ 5 was to preserve the original diagnosis rather than assum-
published. Second, the research domain crite- ing, as a diagnostic change does, that there has been
ria (RDoC) were defined. Third, the use of tech- recovery from one syndrome and development of a
nology in the assessment and treatment of eating new one. More problematic again is the fact that the
disorders has increased. In DSM-​5 “Binge Eating residual grouping EDNOS, from DSM-​IV, was the
Disorder” was moved from a provisional to a full most common ED diagnosis (Fairburn & Cooper,

2 Introduction
2011; Vo, Accurso, Goldschmidt, & Le Grange, Risk Factors and Prevention of Eating
2016). This group was largely composed of subclini- Disorders
cal variants of AN, BN, and BED, often with mixed Risk factors can be ascertained, usually after
symptoms together with more tentatively identified preliminary studies finding associations either ret-
entities such as (self-​induced) vomiting disorder rospectively or concurrently between a disorder and
and night eating syndrome. Loosening the binge particular variables, in two main ways. First a risk
eating (and purging) criteria and the introduction factor can be identified from prospective studies.
of the category Other Specified Feeding and Eating Second, a causal risk factor can be identified experi-
Disorder (OSFED) in DSM-​5 is expected to reduce mentally by altering the strength of the risk factor
the number of EDNOS-​type cases that would have and ascertaining the effect of such alteration on the
been seen in DSM-​IV (Machado, Goncalves, & occurrence of the disorder or an important compo-
Hoek, 2012). nent of the disorder. The most difficult ED to study
A further boundary problem is the relationship is AN, because the incidence of this disorder is
of the eating disorders to overweight and obesity. relatively low, requiring very large-​scale prospective
Here, the boundary between BED and obesity is population studies to identify sufficient cases and
the most complex because a substantial proportion the putative risk factors. However, our knowledge
of those with BED are also overweight or obese. of risk factors for BN and to a lesser extent for BED
A  family study helped to clarify the relationship has developed mainly by means of prospective stud-
between these two disorders (Hudson et al., 2006). ies with a few experimental studies aimed at identi-
The authors found an aggregation of BED within fying causal risk factors.
families, probably due to interacting genetic and Knowledge of risk factors is crucial to the devel-
environmental variables. In addition, relatives of opment of effective prevention programs. Among
those with BED had a markedly higher prevalence the factors that form the basis for a number of pre-
of severe obesity than relatives of those without vention studies in adolescents and young women are
BED. These findings suggest that BED is a familial an elevated perceived pressure to be thin emanating
disorder caused by factors distinct from those that from family, peers, and the media; internalization
cause obesity, and that these BED-​specific fam- of the thin-​ideal espoused for women by Western
ily factors also increase the risk of severe obesity. culture; and elevated body mass index and body dis-
Hence obesity may be conceptualized as an entity satisfaction coupled with dieting. These risk factors
separate from BED although BED is a risk factor have predicted eating pathology in a number of pro-
for the development of obesity, especially severe spective studies. Prevention studies now constitute
obesity. a promising research field with some notable suc-
cesses. Importantly, many prevention programs can
Family and Genetic Studies make use of media and the Internet to deliver the
Family and twin studies suggest that the eating intervention, thus reducing cost and providing easy
disorders are heritable, with familial and environ- access to the programs.
mental factors specific to individuals (nonshared Interestingly, there is mixed evidence that diet-
environment) interacting with genes to produce ing is a risk factor for BN and BED despite the
disorders. The estimated contributions of genetic fact that it is universally regarded as a risk factor.
and nonshared environmental variables differ Specifically, while some prospective studies show
considerably from study to study, hence the rela- increases in binge eating and bulimic symptoms
tive contribution of genes and environment to post dietary restraint, some experimental studies
the eating disorders is unclear. Moreover, epigen- show that dietary restriction can reduce binge eat-
etic factors (the influence of environment on gene ing and bulimic symptoms. It may be that a third
expression) provide a pathway for the effects of variable elicited by the assessment of dieting may
early life stressors. Whether or not genetic studies be a risk factor although it is unclear what that fac-
will provide useful leads for treatment is debatable, tor might be.
given the complexity of eating and its disorders.
Following the path that other psychiatric disorders Treatment of the Eating Disorders
pioneered, it is now recognized that the acquisition The relatively low prevalence of AN combined
and analysis of large well-​specified samples, includ- with the reluctance of many patients with the dis-
ing eating disorders and eating disorder symptoms order to seek or follow through with treatment
is needed. makes treatment research for this disorder difficult.

Agras, Robinson 3
Many of the controlled studies that have been com- follow-​up, and has lower dropout rates than CBT.
pleted have sample sizes too small to allow con- However, neither CBT nor IPT has much effect on
clusions about the effectiveness of treatment to be weight, an important issue because the majority of
made. Hence, at this time there are no first-​line patients with BED are overweight. Individuals who
evidence-​ based pharmacological or psychothera- stop binge eating and who maintain abstinence
peutic treatments available for persistent AN. This from binge eating during follow-​up will lose about
is disappointing, given the fact that of all the eating 5  kg. Here, medications such as the antiepileptic
disorders AN has the longest history, even in mod- drug topiramate and similar compounds may be
ern times. The most promising treatment at this useful because such medications have larger effects
time is a family-​based approach for adolescents first on weight than does CBT or IPT and also reduce
developed at the Maudsley Hospital in London, binge eating (Brownley, et  al. 2016). Recently the
UK (Agras, et  al., 2014; Lock, Agras, Bryson, & FDA approved lisdexamphetamine (LDX) for
Kraemer, 2005). This treatment is now supported the treatment of BED, again showing effective-
by a number of controlled studies demonstrating ness in reducing binge eating and weight. Further
that for adolescents it is more effective than indi- research combining medication and psychother-
vidual therapy and a generic family therapy, hence it apy is needed. Hence, CBT, IPT, and antidepres-
can be regarded as a first-​line evidence-​based treat- sants, antiepileptics, and LDX can be regarded as
ment. Moreover, early treatment may reduce the evidence-​based treatments for BED, with CBT and
number of persistent cases. IPT as first-​line treatments. More recently a large-​
The situation is somewhat better for BN with a scale study compared IPT, behavioral weight loss
number of well-​designed studies comparing vari- treatment (BWL), and guided self-​help (CBTgsh)
ous treatments. Although only fluoxetine is FDA for BED (Wilson, Wilfley, Agras, & Bryson, 2010).
approved for use in BN, most antidepressants have At the end of treatment there were no differences
been shown to be effective in reducing binge eat- among the three groups in reducing binge eating.
ing and purging (Hay, 2013; McElroy, Guerdjikova, However, the BWL group lost more weight than the
Mori, & Keck, 2015). However, cognitive-​ other two groups. At 1-​year follow-​up there were no
behavioral therapy (CBT) appears to be more effec- differences between groups on binge eating reduc-
tive than medication in comparative studies (Agras tion, weight losses, or psychopathology, but at the
et al., 1992; Mitchell et al., 1990). Similarly, CBT 2-​year follow-​up both IPT and CBTgsh were supe-
is more effective than interpersonal therapy (IPT) rior to BWL in reducing binge eating. The authors
at the end of treatment, but not at follow-​up (Agras concluded that CBTgsh may be useful as a first step
et al., 2000) with IPT apparently acting more slowly in the treatment of BED, with IPT or CBT being
than CBT. Hence, CBT can be recommended as a used for those who do not improve with guided
first-​line evidence-​based therapy for BN with IPT self-​help. Indeed, the recent release of the National
or medication as secondary choices. More recently, Institute of Clinical Excellence (NICE) guidelines
guided self-​ help treatments based on CBT have for eating disorder treatment suggest use of guided
been found effective in both adults and adolescents self-​help in the initial treatment of BN and BED
and may form the basis for a cost-​effective first step (NICE, 2017).
in the treatment of BN followed by CBT if needed.
Despite these developments, only about 25% to Technology: Assessment and Treatment
35% of patients with BN who are treated with CBT The advent of Internet-​ based and mobile
will recover. Hence, the search for more effective applications in the last few years holds promise
treatments or combinations of treatments for BN for extending the reach of therapists and bringing
needs to continue. treatment to areas where no evidence-​based care is
Considerable progress has been made in devel- available. In a recent review (Agras, Fitzsimmons-​
oping evidence-​based treatments for BED because Craft, & Wilfley, 2017) of technology-​ based
effective treatments for BN have been adapted assessment and treatment, the authors concluded,
for this condition. Both CBT and IPT have been “There is not a strong enough evidence-​base to
shown to be effective for BED in well designed support widespread usage of Internet treatments
studies, with more than 60% of individuals recover- in the clinic. The extant studies provide a signal
ing both at the end of treatment and at follow-​up that effectiveness studies involving comparisons
Wilfley et al., 2002). Interestingly, IPT is as effec- with known effective treatments are feasible.”
tive as CBT both at the end of treatment and at (p.  34). Most studies did not take advantage of

4 Introduction
the Internet to personalize treatment. Hence, Habermas, T. (2005). On the uses of history in psychia-
treatment over the Internet is in the early stage of try:  Diagnostic implications for anorexia nervosa.
International Journal of Eating Disorders, 38, 167–​182.
development. Moreover some aspects of Internet Harrison, K. (2003). Saint Therese of Lisieux. London,
treatment raise ethical concerns. For example, UK: Weidenfeld & Nicholson.
Internet assessment and treatment without a ther- Hay, P. A systematic review of evidence for psychological treat-
apist may not identify important safety concerns ments in eating disorders. (2013). International Journal of
such as low weight, suicidal ideation, and elec- Eating Disorders, 40, 321–​336.
Hudson, J. I., Lalonde, J. K., Berry, J. M., Pindyck, L. J., Bulik,
trolyte abnormalities, or identify newly emerging C. M., Crow, S.,  . . .  Pope, H. G. Jr. (2006). Binge-​eating
psychopathology during treatment. Moreover, it disorder as a distinct familial phenotype in obese individuals.
is questionable whether a patient whose identity Archives of General Psychiatry, 63, 313–​319.
is unknown should be engaged in treatment. In Keel, P. K., & Klump, K. L. (2003). Are eating disorders culture-​
the United States, state licensing regulations vary bound syndromes? Implications for conceptualizing their
etiology. Psychological Bulletin, 129, 747–​769.
concerning treatment by out-​of-​state providers. Lock, J. Agras, W. S., Bryson, S., & Kraemer, H. C. (2005). A
Mobile applications will face these issues once comparison of short and long-​term family therapy for ado-
they cross the line from assessment to treatment. lescent anorexia nervosa. Journal of the American Academy of
Overall, this book delineates the considerable Child & Adolescent Psychiatry, 44, 632–​639.
progress made in understanding and treating the Machado, P. P., Goncalves, S., Hoek, H. W. (2012). DSM-​
5 reduces the proportion of EDNOS cases:  Evidence
eating disorders while drawing attention to the vari- from community samples. International Journal of Eating
ous gaps in our knowledge with suggestions as to Disorders, 46, 60–​65.
how to address them. McElroy, S. L., Guerdjikova, A. I., Mori, N., Keck, P. E.
(2015). Psychopharmacologic treatment of eating dis-
orders:  Emerging findings. Current Psychiatry Reports,
References 17, 35–​42.
Agras, W. S., Rossiter, E. M., Arnow, B., Schneider, J. A., Mitchell, J. E., Pyle, R. L., Eckert, E. D., Hatsukami, D.,
Telch, C. F., Raeburn, S. D.,   .  .  .  Koran L. M. (1992). Pomeroy, C., & Zimmerman, R. (1990). A comparison
Pharmacologic and cognitive-​behavioral treatment for buli- study of antidepressants and structured intensive group psy-
mia nervosa: A controlled comparison. American Journal of chotherapy in the treatment of bulimia nervosa. Archives of
Psychiatry, 149, 82–​87. General Psychiatry, 47, 149–​157.
Agras, W. S., Walsh, B. T., Fairburn, C. G., Wilson, G. T., & National Institute for Health and Care Excellence (2017).
Kraemer, H. C. (2000). A multicenter comparison of Eating disorders:  recognition and treatment. NICE guide-
cognitive-​behavioral therapy and interpersonal psychother- line (NG69).
apy for bulimia nervosa. Archives of General Psychiatry, 57, Pope, H. G., & Hudson, J. I. (1982). Treatment of bulimia with
459–​466. antidepressants. Psychopharmacology, 78, 167–​179.
Agras, W. S., Lock, J., Brandt, H., Bryson, S. W., Dodge, E., Russell, G. F. M. (1979). Bulimia nervosa: An ominous variant
Halmi, K. A.,  . . .  Wilfley, D. (2014). JAMA Psychiatry, 71, of anorexia nervosa. Psychological Medicine, 9, 429–​448.
1279–​1286. Schneider, J. A., & Agras, W. S. (1985). A cognitive-​behavioral
Agras, W. S., Fitzsimmons-​Craft, E. E., Wilfley, D. E. (2017). group treatment of bulimia. British Journal of Psychiatry,
The evolution of cognitive-​behavioral therapy for eating dis- 146, 66–​69.
orders. Behaviour Research and Therapy, 88, 26–​36. Wermuth, B. M., Davis, K. L., Hollister, L. E., & Stunkard, A. J.
Brownley, K. A., Berkman, N. D., Peat, C. M., Lohr, K. N., (1977). Phenytoin treatment of the binge-​eating syndrome.
Cullen, K. E. & Bulik, C. M. (2016). Binge eating disorder American Journal of Psychiatry, 134, 1249–​1253.
in adults:  A  systematic review and meta-​analysis. Annals of Wilfley, D. H., Welch, R. R., Stein, R. I., Spurrell, E. B., Cohen,
Internal Medicine, 165, 409–​420. L. R., Saelens, B. E.,  . . .  Matt, G. E. (2002). A randomized
Bynum, C. (1987). Holy feast and holy fast:  The religious sig- comparison of group cognitive-​behavioral therapy and group
nificance of food to medieval women. Berkeley: University of interpersonal psychotherapy for the treatment of overweight
California. individuals with binge eating disorder. Archives of General
Fairburn, C. G. (1981). A cognitive-​behavioural approach in Psychiatry, 59, 713–​721.
the management of bulimia. Psychological Medicine, 11, Wilson, G. T., Wilfley, D. E., Agras, W. S., & Bryson, S. (2010).
707–​711. Psychological treatments of binge eating disorder. Archives of
Fairburn, C. G., & Cooper, Z. (2011). Eating disorders, General Psychiatry, 67, 94–​101.
DSM-​5 and clinical reality. British Journal of Psychiatry, Vo, M., Accurso, E. C., Goldschmidt, A. B., & Le Grange, D.
198, 8–​10. (2016). The Impact of DSM-​5 on eating disorder diag­
Gull, W. W. (1874). Anorexia nervosa. Transactions of the Clinical noses. International Journal of Eating Disorders, 50, 578–​581.
Society London, 7, 22–​28. doi:10.1002/​eat.22628

Agras, Robinson 5
 PART 
1
Phenomenology
and Epidemiology
 CH A PT E R


The Classification of Eating Disorders
1
Kathryn H. Gordon, Jill M. Holm-​Denoma, Valerie J. Douglas, Ross Crosby,
and Stephen A. Wonderlich

Abstract
The purpose of this chapter is to elucidate the key issues regarding the classification of eating disorders.
To this end, a review of nosological research in the area of eating disorders is presented, with a
particular focus on empirically based techniques such as taxometric analysis, latent class analysis, and
factor mixture modeling. This is followed by a section outlining areas of overlap between the current
Diagnostic and Statistical Manual of Mental Disorders–​Fifth Edition (DSM-​5) eating disorder categories
and their symptoms. Next, eating disorder classification models that are alternatives to the DSM-​5 are
described and critically examined in light of available empirical data. Finally, areas of controversy and
considerations for change in next version of the DSM (i.e., the applicability of DSM criteria to minority
groups, children, and males; the question of whether clinical categories should be differentiated from
research categories) are discussed.
Key Words:  classification, diagnostic model, eating disorder, latent class analysis, nosology, taxometrics

Introduction Mendeleev’s periodic table of elements for the field

Taxonomy (the science of classification) is often of chemistry:  theoretically driven, parsimonious,
undervalued as a glorified form of filing—​with each and instrumental for scientific growth.
species in its folder, like a stamp in its prescribed Within the disciplines of psychiatry and psy-
place in an album; but taxonomy is a fundamen- chology, one of the most commonly used classifi-
tal and dynamic science, dedicated to exploring cation systems for mental disorders is the Diagnostic
the causes of relationships and similarities among and Statistical Manual of Mental Disorders—​Fifth
organisms. Classifications are theories about the Edition (DSM-​5; American Psychiatric Association
basis of natural order, not dull catalogues compiled [APA], 2013). The broad objectives of this chapter
only to avoid chaos (Gould, 1989, p. 98). are (1) to examine how well the DSM-​5 eating dis-
Sound classification systems are the cornerstone order (ED) classification system performs in terms
of scientific progress. The universal language of clas- of the qualities that Gould and others (e.g., Waller
sification systems allows scientists to communicate & Meehl, 1998) have articulated and (2) to illumi-
in an efficient, standardized manner about variables nate pathways for improvement on the current sys-
in their discipline. This function of classification is tem. The chapter begins with a brief description of
extremely important, and makes it possible to con- the DSM’s current system for classifying EDs and a
struct a collective knowledge base on which scien- review of research on nosological issues. Next, the
tists can build and advance the field. However, as overlap between diagnostic entities is reviewed, and
Stephen Jay Gould articulates in the quote above, alternative diagnostic models for EDs are presented.
classification systems, at their best, are not merely Finally, the chapter concludes with suggestions for
arbitrary organization systems. Rather, they are like change and future directions.

9
The Current DSM Classification of Eating average, a minimum of once per week for 3 months.
Disorders The binge eating episodes must be associated with
The DSM-​5 was updated to include two main at least three of the following characteristics: eating
types of disorder characterized by disturbed eating abnormally fast, eating until uncomfortably full,
behaviors: feeding disorders and EDs. There are five eating large amounts of food when not experiencing
categories of EDs:  anorexia nervosa (AN), bulimia physical hunger, being embarrassed by the amount
nervosa (BN), binge eating disorder (BED), other of food consumed and subsequently feeling the
specified feeding or eating disorder (OSFED), and need to eat alone, and feeling guilty, disgusted with
unspecified feeding or eating disorder (UFED). oneself, or depressed afterward. In addition, there
Anorexia nervosa is diagnosed when an individual is significant distress concerning binge eating and
does not consume enough caloric energy to main- the individuals must not regularly use inappropri-
tain a minimal healthy body weight, exhibits intense ate compensatory behaviors. Binge eating disorder
fear about weight gain and/​or fatness despite being is considered mild in severity when there are 1–​3
underweight, and has distorted perceptions related binge eating episodes per week, moderate for 4–​7
to weight and shape (e.g., does not recognize one is episodes, severe when there are 8–​13 episodes per
severely underweight, experiences undue influence week, and extreme if there are 14 or more binge eat-
of shape or weight on mood and self-​evaluation. ing episodes per week (APA, 2013).
Anorexia nervosa is divided into two subtypes:  a Next, there are two related but separate diag-
binge eating/​purging type for those who engage in nostic categories, OSFED and UFED. The first,
binge eating (i.e., an episode wherein one expe- OSFED, is reserved for individuals deemed to have
riences a sense of loss of control while eating and a clinically significant (i.e., distressing and impair-
consumes an objectively large quantity of food) and/​ ing) eating or feeding disorder that does not meet
or purging (i.e., self-​induced vomiting or laxative or full criteria for AN, BN, BED, or any of the feeding
diuretic use) and a restricting type for individuals who disorders. Examples of presentations that can fall
do not regularly engage in binge eating or purging under this category are atypical AN, low frequency/​
behavior (APA, 2013). In addition to categorical dis- duration BN, low frequency/​duration BED, purg-
tinctions, DSM-​5 includes a dimensional measure ing disorder, or night eating syndrome. The second,
of severity for AN based on body mass index (BMI; UFED, is also reserved for individuals deemed to
≥17  kg/​m2  =  mild, 16–​16.99  kg/​m2  =  moderate, have a clinically significant eating or feeding disor-
15–​15.99 kg/​m2 = severe, <15 kg/​m2 = extreme). der that does not meet criteria for the other feed-
Bulimia nervosa is diagnosed when an individual ing or eating disorders; however, this diagnosis is
regularly (i.e., at least once a week on average for a given when the clinician chooses to not specify the
3-​month period) engages in binge eating (i.e., eat- presentation or there is insufficient information to
ing an objectively large amount of food within any make a more specific diagnosis (APA, 2013).
2-​hour period while perceiving a lack of control) Finally, there are two feeding disorders that are
and inappropriate compensatory behaviors (e.g., most relevant to the adult population:  pica and
excessive exercising, purging, fasting). Bulimia ner- rumination disorder. Pica is diagnosed in an indi-
vosa can be diagnosed only if the individual does vidual if, for at least a month-​long period, the indi-
not meet criteria for AN (otherwise an AN diag- vidual persistently consumes substances that are
nosis trumps the BN diagnosis), and BN diagno- nonnutritive and not foods. This eating of nonfood
ses also require that the individual’s self-​evaluation substances must be incongruent with the individu-
is unduly influenced by weight and shape (APA, al’s development level such as seen in adults without
2013). Bulimia nervosa also has a dimensional certain mental disorders (e.g., intellectual develop-
severity specifier determined by the frequency of mental disorder, autism spectrum disorder, schizo-
inappropriate compensatory behaviors per week (1–​ phrenia) and must also not be in the confines of a
3 episodes = mild, 4–​7 episodes = moderate, 8–​13 cultural or social normative practices. Rumination
episodes = severe, 14 or more episodes = extreme). disorder is diagnosed in an individual if, for at least
The DSM-​5 reduced the frequency of disordered a month-​ long period, the individual repeatedly
eating behavior required for BN and eliminated the regurgitates their food and this regurgitation is not
nonpurging/​purging subtype classification. attributed to a gastrointestinal or medical condi-
For DSM-​5, BED went from the appendix to tion. The regurgitation must not exclusively occur
an official eating disorder diagnosis; it is diagnosed during an episode of AN, BN, or BED. The indi-
when an individual experiences binge eating, on vidual may proceed to rechew, reswallow, or eject

10 Cl assification
the regurgitated food. Individuals do not regurgi- samples is hampered by the low percentage of
tate the food with nausea or disgust (APA, 2013). included participants that exhibit concerning ED
To our knowledge, there are not classification stud- symptoms (i.e., having low numbers of participants
ies specifically examining rumination disorder or with ED symptomatology decreases the likelihood
pica in adults, and therefore, these disorders are not of identifying meaningful ED clusters). Thus LPA/​
discussed further in this chapter. LCA research using clinical samples tends to iden-
tify several clusters that exhibit differing types of
Review of Research on Nosological Issues ED pathology, often including groups that resemble
The way in which eating disorders are defined, AN, BN, BED, and mixed ED presentations (e.g.,
classified, and distinguished have important impli- Deschartres et al., 2011; Eddy et al., 2009). In addi-
cations for both the scientific and clinical eating dis- tion, a series of LCA/​LPA studies have suggested
order communities. The eating disorder diagnoses that there may be a meaningful distinction between
currently defined in the DSM-​5 were based jointly people with AN-​like presentations who endorse an
on clinical wisdom and empirical evidence available intense fear of fatness and people who have a very
at the time of publication; however, the validity of low weight but do not endorse strong weight/​fatness
these diagnoses and the criteria used to define DSM concerns (e.g., Crow et al., 2012; Keel et al., 2004;
diagnoses have not always received empirical sup- Wildes, Forbush, & Markon, 2013), and between
port (see Wonderlich, Joiner, Keel, Williamson, & those with BN-​like presentations who engage in
Crosby, 2007). Statistical approaches to the classifi- multiple purging methods versus those who solely
cation of eating disorders, such as latent class analy- vomit (e.g., Crow et  al., 2012; Eddy et  al., 2009;
sis (LCA; Lazarsfeld & Henry, 1968), taxometric Keel et al., 2004).
analysis (TA; Waller & Meehl, 1998), and factor One of the advantages of LCA is its ability to
mixture modeling (FMM; Lubke & Muthen, 2005) determine the optimal number and composition
provide empirically based alternatives to the DSM of homogeneous groups; however, a noted weak-
that may have greater scientific validity and clinical ness of LCA is the tendency for it occasionally to
utility. These empirical approaches are considered in produce spurious classes that actually represent
the text that follows. different points along a continuum of severity
Latent class analysis is designed to identify (Uebersax, 1999). Taxometric analyses were spe-
homogeneous subsets of cases (i.e., latent classes) cifically designed to address the question of whether
using observed signs and symptoms based on the two apparently separate classes represent categori-
principle of conditional independence (Lazarsfeld cally distinct entities (taxa) or superficially differ-
& Henry, 1968). Specifically, classes are created in ent manifestations of a single underlying condition
LCA in such a way that, within each class, the signs (Waldman & Lilienfeld, 2001). Like LCA, TA,
and symptoms are statistically independent (i.e., examines the associations between observed mea-
uncorrelated). Thus, LCA is a statistical model-​ sures; however, unlike LCA, TA specifies a priori the
based approach that provides objective criteria for number of groups (i.e., two). In TA, indicators are
the determination of the optimal number of clus- selected to serve as proxies of group membership.
ters, and provides a probability-​based method for If group membership is based on a dimensional
assigning individuals to classes. Also, LCA requires distinction along an underlying continuum, then
the use of dichotomous indicator variables. A vari- the strength of associations between any two indi-
ant of LCA, latent profile analysis (LPA), allows the cators of group membership will remain constant
use of categorical, ordinal, and continuous variables. across the range of a third indicator. In contrast, if
Well over a dozen studies have used LCA or the distinction between groups is truly qualitative,
LPA to classify individuals with eating disorder then the strength of associations will be lowest when
symptoms. Two to six clusters tend to emerge from examined separately in each distinct group and
these analyses (see Keel, Brown, Holland, & Bodell, highest when examined in a mixed sample contain-
2012, for a more detailed review of LCA/​LPA stud- ing individuals from each group (Gordon, Holm-​
ies). Results from nonclinical or community-​based Denoma, Smith, Fink, & Joiner, 2007).
samples consistently suggest that healthy “nor- Two fairly recent studies have used groups
mals” are distinct from people with disordered eat- identified in LPA to conduct subsequent TA. In
ing symptoms (e.g., Pinheiro, Bulik, Sullivan, & one study, Keel et  al. (2011) first identified five
Machado, 2008; Wade, Crosby, & Martin, 2006); clusters (healthy normals, BN, mixed ED, AN,
however, LCA/​ LPA research using community and BED) using LPA. Next, they conducted a

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 11

series of TAs comparing pairs of identified groups In recent years, researchers have begun using
to one another. Results showed a categorical dis- FMM to evaluate the latent structure of con-
tinction between the healthy normal group and structs. Factor mixture modeling is able to ana-
each of the other four ED groups; however, TA lyze continuous and categorical observed variables
conducted between the various ED groups often in a single model and can also include covariate
resulted in ambiguous results, likely due to rela- effects in different parts of the model. In addition,
tively small sample sizes. Similarly, Thomas et al. unlike TA, FMM can evaluate the possibility of
(2015) used a combined LPA and TA approach in more than two categories using statistical tests that
a sample of individuals from Hong Kong. Four evaluate competing models’ goodness of fit (Lubke
groups emerged in LPA (BN, low-​weight non-​fat-​ & Muthen, 2005). To our knowledge, only one
phobia, classic AN, and BED). Due to the group’s FMM study of eating disorders have been pub-
small number of members, the BED group was lished to date. Keel et al. (2013) examined various
unable to be used in the TA; however, other pair- bulimic syndromes (e.g., AN binge-​purge subtype,
wise TA revealed categorical distinctions between BN, BED, purging disorder, and EDNOS cases
the two low-​weight classes, and between BN and with BN-​like symptoms) to determine whether
the low-​weight non-​fat-​phobia group. The cat- they are best represented as categorically distinct
egorical distinction between the non-​fat-​phobic diagnostic groups, a single group with dimen-
low-​weight group and the classic AN group is sions of severity, or a combination of groups and
largely consistent with the findings of Keel et al. dimensions. Results suggested that a hybrid model
(2011), who also identified a taxonic relationship comprised both categories and a dimension most
between a fat-​phobic low-​weight class and a non-​ accurately reflected the latent structure of bulimic
weight-​phobic (albeit normal-​weight) group. syndromes. Three latent classes were observed (one
Other researcher teams have conducted recent with serious EDs characterized by binging, self-​
TA of eating disorders without using groups iden- induced vomiting, and high levels of fat phobia;
tified a priori through LCA/​ LPA. Results from another group with subthreshold EDs character-
early taxometric analyses of eating disorders (e.g., ized by binging, nonpurging compensatory behav-
Gleaves, Lowe, Green, et al., 2000; Gleaves, Lowe, iors, and low levels of fat phobia; and a final group
Snow, et al., 2000; Williamson et al., 2002; Tylka & with minimal eating disorder symptoms), whereas
Subich, 2003) tended to support the idea of a single the identified dimension that cut across categories
“bulimic taxon” that is categorically distinct from reflected symptom severity.
normality; however, the conclusions from those
early studies have been criticized as methodological Overlap Between Diagnostic Entities
advances, including the development of the com- The previous section consisted of a review of
parison curve fit index (CCFI; Ruscio, 2007), have research on nosological issues and EDs, while the
evolved. current section discusses areas of overlap between
More recently, in both a nonclinical sam- the diagnostic categories. At a basic level, there is
ple (Holm-​Denoma, Richey, Joiner, 2010) and some overlap in the actual diagnostic criteria of
an inpatient ED sample (Olatunji et  al., 2012), AN and BN. Specifically, both diagnoses include a
TAs were conducted using the Eating Disorders criterion reflecting “self-​evaluation is unduly influ-
Inventory subscales as indicators. In the com- enced by body shape and weight” (APA, 2013).
munity sample, body dissatisfaction, bulimia, and This criterion is not currently included for the
drive for thinness were all shown to have a dimen- diagnosis of BED. However, Grilo et  al. (2008)
sional latent structure. In the inpatient sample, argue for its use as a diagnostic specifier for BED,
similar results were obtained, such that categori- based on findings that BED participants with clini-
cal distinctions were not revealed between diag- cal levels of overvaluation of weight/​shape reported
nostic groups (i.e., AN-​R, AN-​BP, and BN) when significantly higher levels of eating pathology and
body dissatisfaction, bulimia, drive for thinness, depression symptoms than BED participants with
and BMI were used as indicators. In addition, a subclinical levels of overvaluation of weight and
recent quantitative review of the latent structure shape. It is interesting to note that these group dif-
of various types of psychopathology concluded ferences (based on levels of overvaluation of weight
that there is little evidence that EDs are categori- and shape) emerged despite a lack of differences in
cally distinct from normalcy (Haslam, Holland, & BMI and binge eating frequency between the two
Kuppens, 2012). BED groups.

12 Cl assification
Overlap Between DSM-​Based Eating possibly BED) on the other. Specifically, individuals
Disorder Diagnoses on External Validators with AN tend to be high in constraint whereas those
There is substantial evidence that meaningful with BN tend to be more impulsive (Cassin & von
differences between the DSM diagnostic classes do Ransom, 2005).
not exist on a host of external validators. For exam- In sum, ED diagnoses, as defined in DSM-​5,
ple, a variety of studies have indicated that although appear to lack discriminant validity on a number
longitudinal stability within an ED diagnosis is of variables. Nonetheless, there are some important
more common than diagnostic crossover, a substan- clinical validators that have been supported in com-
tial minority of people with an ED diagnosis dem- parisons of DSM-​5 classes.
onstrate diagnostic crossover longitudinally (e.g.,
Agras, Crow, Mitchell, Halmi, & Bryson, 2009; Alternative Diagnostic Models
Eddy et al., 2008; Fichter & Quadflieg, 2007). In In response to the aforementioned concerns
addition, family history studies have failed to find about the existing DSM ED classification system,
that eating disorder diagnoses “breed true,” as there some researchers have proposed alternative diagnos-
are frequent cross-​transmissions of diagnoses in fam- tic systems. Two of the proposed alternative models
ily members of both AN and BN probands (Strober, are described in this section. In addition, a criti-
Freeman, Lamper, Diamond, & Kaye, 2000) and cal review of each proposed classification system is
high levels of genetic correlations between AN and provided.
BN (Bulik et al., 2010). Both individuals with AN
and BN have also been shown to have abnormalities Three-​Dimensional Model
in the serotonergic and dopaminergic systems (e.g., After examining the results of a series of taxo-
Frank et al., 2002; Frank et al., 2005). metric studies (Gleaves, Lowe, Snow, et  al., 2000;
Moreover, all of the major eating disorder diagno- Williamson et  al., 2002), Williamson, Gleaves,
ses are characterized by a frequently shared comor- and Stewart (2005) proposed the empirically based
bidity profile; AN, BN, and BED all show high levels three-​dimensional model (TDM) of eating disor-
of mood disorders, substance use disorders, anxiety ders. The TDM hypothesizes that three main fac-
disorders, and personality disorders (Hudson, Hiripi, tors underlie all disordered eating presentations.
Pope, & Kessler, 2007; Johnson, Spitzer, & Williams, The first factor, binge eating, is considered to be
2001; Wonderlich & Mitchell, 1997). Similarly, eat- taxonic (i.e., individuals either engage in binge eat-
ing disorder classes are not distinguishable in terms ing or they do not), whereas the other two factors,
of impaired interpersonal functioning (Gonzales, drive for thinness and fear of fatness/​inappropriate
2001; Hartmann, Zeeck, & Barrett, 2010; Hsu compensatory behaviors, are viewed as dimensional.
et al., 2002) and share many personality correlates Williamson et al. (2005) have conceptualized how
(e.g., for AN and BN: perfectionism, negative emo- each of the existing DSM EDs and obesity would
tionality, Cassin & von Ranson, 2005). fill the TDM’s space when all three factors are
However, it is important to recognize that the simultaneously considered (Figure 1.1).
current DSM-​based eating disorder diagnoses pro- The diagnostic system proposed by the TDM
vide some degree of discriminant validity on clinical differs from that of the DSM in several important
outcome measures. It appears as though AN is dis- ways. First, it posits that eating behaviors with a
tinguished from the other eating disorders in terms binge eating component (e.g., BN, BED, and the
of higher overall mortality rates (Arcelus, Mitchell, binge eating/​purging type of AN) are qualitatively
Wales, & Nielsen, 2011), higher suicide rates (Preti, distinct from those without a binge eating compo-
Rocchi, Sisti, Camboni, & Miotto, 2011), and nent (e.g., the restricting type of AN, obesity, and
lower likelihood of remission or recovery (Fichter & normal eating). Second, individuals with the binge
Quadflieg, 2007; Herzog et al., 1993). It is unclear eating/​purging type of AN are posited to be qualita-
whether people with different ED diagnoses show tively similar to those with both purging and non-
differential response to treatment because no study purging types of BN, but to differ quantitatively in
has been conducted that exposed all three ED diag- their relatively higher levels of drive for thinness.
noses to the same treatment agent; however, the lit- Third, according to this model, BED is qualitatively
erature implies that AN is more treatment resistant similar to other disorders that have a binge eating
than BN and BED (Fichter & Quadflieg, 2007). In component, but quantitatively different from them
addition, some personality differences have been because it is defined by relatively lower levels of fear
observed between AN on one hand and BN (and of fatness and drive for thinness. Finally, only one

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 13

variant of AN exists according to the TDM:  AN, whereas those with relatively lower levels of both fac-
restricting type. tors would be considered obese or normal. Within
One could imagine that a classification system both of the diagnostic categories proposed by the
stemming from the TDM would have two primary TDM (i.e., EDs accompanied by binge eating, and
ED categories:  EDs accompanied by binge eat- EDs without a binge eating component), diagnos-
ing and EDs without a binge eating component. tic labels would refer to individuals who differ with
Within the binge eating category, individuals would regard to severity of drive for thinness and fear of
be plotted on the dimensions of drive for thinness fatness. Thus, within the two diagnostic categories,
and fear of fatness. As described in the preceding no qualitatively distinct groups would exist.
text, those who had the highest levels of drive for The TDM has implications not only for the clas-
thinness would have a syndrome that is like the sification of eating disorders, but also for the eti-
DSM-​5’s AN, binge eating/​ purging diagnosis, ology, assessment, and treatment of them. With
whereas those with relatively lower levels of drive regard to etiology, some assume that taxonic results
for thinness would have syndromes that were sim- signify a genetic or biological basis to a disorder.
ilar to BED and BN. In addition, those with syn- Although taxonic entities may have a higher like-
dromes like the DSM-​5’s BED would have lower lihood of a genetic etiology than dimensional enti-
levels of fear of fatness than BN. Within the non-​ ties (Meehl, 1992), it is conceivable that they may
binge-​eating category, individuals with high levels also result from sociocultural factors (e.g., an envi-
of fear of fatness and drive for thinness would be ronmental mold taxon; Catell, 1946). Similarly,
considered to have a restrictive anorexic syndrome, dimensional entities such as AN, restricting type,

AN-BP
High
BN-P BN-NP

BED
Binge Eating

AN-R

High

Obese
DFT

Normal
Low
High
Low
Low
FF/CB

Figure 1.1  Three-​dimensional model of eating disorders. BN-​P = bulimia nervosa, purging type; BN-​NP = bulimia nervosa,
nonpurging type; AN-​R = anorexia nervosa, restricting type; AN-​BP = anorexia binge/​purge type; BED = binge eating disorder;
FF/​CB = fear of fatness/​concern with body size and shape.  Source: Williamson, D. A., Gleaves, D. H., & Stewart, T. M. [2005]. Categorical
versus dimensional models of eating disorders: An examination of the evidence. International Journal of Eating Disorders, 37, 1–​10. © 2005 Wiley
Periodicals, Inc. Reprinted with permission of John Wiley & Sons, Inc.

14 Cl assification
may be impacted by biological factors (e.g., amount that many distinctive clinical features (i.e., caloric
of serotonin released in the brain) and/​or may result restriction, binge eating, body checking) cut across
from an interaction of biological factors (e.g., geno- ED diagnostic categories. As evidence of their view-
type of the 5-​HT2a receptor gene) and environmen- point, Fairburn cite the fact that some individu-
tal factors (e.g., internalization of the thin ideal). als first diagnosed with a given ED cross over into
With regard to assessment, Gangestad and another ED category over time.
Snyder (1985) reported that mismatching the struc- Given the commonalities across ED diagnoses,
ture (i.e., taxonic vs. dimensional) of a latent vari- Fairburn et  al. (2008) suggests that one’s specific
able with a given assessment modality could result ED diagnosis should not determine the treatment
in distortion of results. For taxonic disorders, a short modality. Rather, cognitive-​behavioral therapy (CBT)
measure that concentrates on dichotomously classi- should be used with all ED patients, and the spe-
fying individuals based on a best cut can minimize cific techniques of CBT that are primarily employed
misclassifications (Meehl, 1992). In contrast, mea- should be based on the psychopathological features
sures of continuous constructs must include items and maintaining mechanisms of each patient. For
that adequately assess all aspects of the latent entity instance, a patient presenting with high levels of clini-
and discriminate across the entire breadth of the cal perfectionism should be treated with a CBT-​based
dimension (Meehl, 1992; Ruscio & Ruscio, 2002). approach that targets the perfectionism, regardless of
Finally, with regard to treatment, Williamson et al. the DSM-​5 defined ED for which she or he meets
(2005) suggest that for disorders that are dimen- criteria. In sum, Fairburn believe that regardless of a
sional in nature (e.g., AN, restricting type), progress patient’s diagnostic status as defined by the DSM, a
during treatment may be relatively slow and difficult similar treatment approach (CBT) should be used.
to detect at any given time. Treatment that moves Although the transdiagnostic approach holds
someone “down” the continuum a small number some intuitive appeal, it is somewhat undermined
of intermediate levels may result in change that is by data that suggest genuine differences between
almost unperceivable in the short term. In contrast, AN and BN exist (e.g., mortality rates; Franko &
researchers have predicted a disorder that is taxonic Keel, 2006). Perhaps the model still offers an inter-
in nature will likely respond to treatment in an esting approach to conceptualizing the dimensions
all-​or-​nothing fashion (Strube, 1989; Williamson that may cut across all ED diagnoses. For instance,
et  al., 2005). In this case, they have hypothesized Fairburn (2008) hypothesizes several characteristics
that change may be difficult to initiate, but once (i.e., maintaining mechanisms) that cut across ED
initiated, should be noticeable and complete. It is diagnoses such as clinical perfectionism, low self-​
important to note that these are hypotheses that esteem, mood lability intolerance, and interpersonal
have yet to be empirically examined. difficulties. Future TA and/​or FMM could exam-
As discussed by Wonderlich et al. (2007), one of ine these maintaining mechanisms to determine
the limitations of the TDM is that it relies on the whether they are indicators of dimensions and/​or
results of taxometric studies that have examined the taxa relevant to eating pathology.
nature of existing DSM ED categories. Therefore,
the model’s generalizability to EDs not currently Which Classification Model Is Best?
included in the DSM-​ 5 (e.g., feeding disorders, As described in the preceding text, the DSM-​
such as pica and rumination disorder) is compro- 5’s current classification system is partially a result
mised. Further, it does not consider certain variables of arbitrary, rather than empirically based, deci-
that are integral to any classification system (e.g., sions. The two alternative models described in
what role comorbidity plays). this section address some of the shortcomings of
the DSM model. For instance, the TDM is based
Transdiagnostic Model on empirical findings, whereas the transdiagnos-
The transdiagnostic approach, first outlined by tic model addresses some DSM validity concerns
Fairburn and his colleagues (Fairburn, Cooper, (e.g., rates of crossover between diagnoses might
& Shafran, 2008; Fairburn, Cooper, & Shafran, indicate an invalid boundary between DSM cat-
2003), proposes that all EDs fall into a single diag- egories, and the transdiagnostic model minimizes
nostic category called “eating disorder.” Its premise the use of the arbitrary diagnostic boundaries).
is that the core psychopathology of all EDs is the However, there are limitations to each of the alter-
overvaluation of control over body shape, weight, native models. For example, the TDM has focused
and eating. Moreover, Fairburn et al. (2008) believe exclusively on DSM-​defined eating disorders.

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 15

 Given the three models discussed in this chapter emotionally dysregulated; compulsive/​emotionally
(i.e., DSM-​5, TDM, and the transdiagnostic model) constricted, and normative (e.g., Wonderlich et al.,
and an infinite array of other possible models, how 2005). Similar results have been obtained in ado-
can one discern which classification system is best? lescent samples (Thompson-​Brenner, Eddy, Satir,
First, it is important to emphasize the importance of Boisseau, & Westen, 2008). The personality clus-
considering multiple variables simultaneously when ters differ from one another with regard to poten-
evaluating a classification system’s validity. Kendler tial etiologic variables such as rates of childhood
(1990) suggests several validators need to be rigor- abuse (Steiger, Israël, Gauvin, Kin, & Young, 2003;
ously evaluated: treatment response and utility, clin- Westen & Harden-​Fischer, 2001), family history
ical course, etiological factors, biological and genetic (Holliday, Landau, Collier, & Treasure, 2006), and
variables, and demographic measures. Second, it is genetic vulnerabilities (Steiger et  al., 2003). Their
currently unclear whether either of the newly pro- eating disorder history and comorbidity profiles are
posed systems demonstrates incremental validity also different (e.g., Tozzi et  al., 2005; Wonderlich
or clinical utility when compared to the DSM-​5’s et al., 2005), as well as their treatment history and
system. Future researchers should consider testing emotional and behavioral ratings (Wonderlich
alternative classification systems against one another et al., 2007). These findings suggest that personality
to determine empirically which one demonstrates variation within a given DSM diagnostic construct
maximal validity, and the field should prioritize the (e.g., AN, restricting type) may be associated with
use of empirically supported diagnostic systems. meaningful conceptual and clinical differences.
Moreover, many empirically derived eating disor- Accordingly, a maximally effective treatment may
der classification systems have been shown to have look different when delivered to an individual with
superior external validity than the DSM’s classifi- AN who has significant emotion dysregulation diffi-
cation system. For instance, LPA studies generated culties in comparison to an individual with AN who
a six-​class system in which three classes (i.e., a class does not have significant emotion dysregulation dif-
that resembles BN, a class that resembles BED, and ficulties (Racine & Wildes, 2015). Some research-
a class of low-​weight individuals who denied a fear ers have described modifications to treatments that
of fatness) had heightened standardized mortal- account for personality variations (Fairburn et  al.,
ity ratios; the six-​class empirically based classifica- 2008; Wonderlich, Mitchell, Peterson, & Crow,
tion resulted in better prediction of mortality risk 2001), but more research is needed.
than the existing classifications in the DSM (Crow
et al., 2012). Conclusion
In conclusion, the way that EDs are classified
Intradiagnostic Heterogeneity by the DSM has a significant impact on empirical
It is well known that there is heterogeneity in research and treatment development. Therefore,
clinical presentation within each ED category it is of utmost importance that future versions of
(Wonderlich et al., 2007), but there seems to be a the DSM be guided by empirical research. In the
limited number of personality patterns when you first section of this chapter, we reviewed nosological
look across all DSM-​based ED categories (Johnson research on EDs. Research using LPA/​LCA tends
& Connors, 1987; Westen & Harden-​Fischer, 2001; to find clusters of eating pathology that are similar
Wonderlich & Mitchell, 1997). Early on, most to AN, BN, BED, and mixed ED presentations. In
researchers focused on how multi-​impulsivity and addition, these types of studies suggest that there
other features often associated with borderline per- may be subgroups of AN (fat phobic versus low
sonality disorder might be related to eating pathol- fear of fat) and BN (single purging methods ver-
ogy (Fichter, Quadflieg, & Rief, 1994; Lacey, 1993; sus individuals who use multiple purging methods.
Steiger & Stotland, 1996; Wonderlich & Swift, Finally, the most recent TA suggest that EDs are not
1990). More recently, however, researchers have categorically distinct from each other or normality
applied statistical approaches such as cluster analysis and may occur on a continuum. However, more
and latent profile analysis to identify multiple per- future research is needed to investigate these find-
sonality profiles that reliably coincide with eating ings further.
disturbances. Next, we reviewed diagnostic overlap among
Empirical studies have consistently identified the eating disorders, and outlined a general lack of
at least three personality clusters that are common differences on certain external validators (e.g., fam-
among adults with eating disorders:  impulsive/​ ily history, comorbid conditions), and significant

16 Cl assification
differences on other validators (e.g., mortality rates criteria were valid, subsequent changes to the clin-
and likelihood of remission or recovery; with AN ical criteria would be warranted. This type of pro-
generally faring worse than BN). Next, we outlined cedure may prevent researchers from limiting their
and evaluated alternative ways to classify eating studies to DSM-​defined disorders, while simultane-
disorders with the TDM and the transdiagnostic ously ensuring that only empirically supported and
model (which suggests one “eating disorder” diag- validated changes eventually occur in the clinically
nosis and emphasizes the lack of differences between based DSM system.
categories). In the final section of the chapter, we Alternatively, the research community may
will address the challenges that confront the classi- adopt a new perspective on empirical studies, such
fication of eating disorders, such as applicability to that it may discourage investigators to adhere rig-
minority groups, children, males, and the question idly to DSM conventions in an effort to expedite
of whether or not clinical categories should be dif- the process of discovery. A  paradigm shift of this
ferentiated from research categories. nature would require researchers to provide clear
rationales for examining alternative sets of symp-
Future Directions and Controversies toms, funding organizations to value the impor-
The following section is focused topics related to tance of scrutinizing existing diagnostic standards,
the future of classification for eating disorders. and the peer-​review system to embrace studies that
examine novel diagnostic concepts. At least one
Because DSM Is Aimed Primarily funding organization, the National Institute of
at Clinicians, Should Different Mental Health, appears to be embracing this notion
Classification Criteria Be Used for Research? through their Research Domain Criteria (RDoC;
Although the DSM-​ 5 clearly states that its National Institute of Mental Health [NIMH],
highest priority is to provide a helpful guide that 2016). NIMH (2016) defines RDoC as “a research
informs clinical practice (APA, 2013, p.  xli), the framework for new ways of studying mental disor-
disorders and associated symptoms described in the ders. It integrates many levels of information (e.g.,
DSM are often used in research endeavors as well. from genomics to self-​report) to better understand
Researchers tend to adopt a rigid DSM definition basic dimensions of functioning underlying the full
of a given mental disorder when conducting empiri- range of human behavior from normal to abnor-
cal studies, and this practice has likely hindered the mal.” Therefore, we may see more empirical data on
progression of knowledge about the disorder’s eti- alternative classification systems for eating disorders
ology and treatment (Grilo, Devlin, Cachelin, & as a result.
Yanovski, 1997; Wilfley, Bishop, Wilson, & Agras,
2007). Specifically, when researchers study a pre- What Cultural Issues Should Be Considered
defined condition, they do not allow themselves to for Future Versions of the DSM?
contest currently accepted conventions; therefore, The majority of research on the classification
the evolution of new valid diagnostic criteria, and of eating disorders is based on White adolescent/​
associated etiological and treatment implications, is young adult females from Western, industrialized
disrupted. nations, despite the fact that EDs are not restricted
In acknowledgment of this problem, ED to this demographic group. The samples used in
experts have encouraged clinicians and researchers studies influence the description, definition, and
alike to stop reifying the DSM (Grilo et al., 1997; classification of EDs. Therefore, a lack of diversity
Kupfer, First, & Regier, 2002). Making modifica- in study samples could limit the generalizability and
tions to the current diagnostic system and common utility of ED classifications used to characterize EDs
research approaches may facilitate this request. One in people from non-​Western cultures and from eth-
such modification may be to introduce a split-​ nic/​racial minority groups.
classification system in which research and clinical Specifically, there are concerns about the DSM-​5
criteria for a given disorder are not identical. For AN criteria set. Fear of weight gain as a criterion
instance, researchers may specify research criteria has been criticized on the grounds that it may not
in need of empirical examination (e.g., frequency be present in individuals displaying apparent AN,
of binge eating that is clinically significant; person- though this is a subject of debate (Habermas, 1989;
ality dimensions that have clinical utility to eating Katzman & Lee, 1997). Attempts to examine evi-
pathology). If a sufficient body of empirical evi- dence of AN across cultures have been marked by
dence accumulated and suggested that the research debates concerning the definition of the illness.

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 17

While there is universal agreement that AN rep- 2000; Gowen, Hayward, Killen, Robinson, & Taylor,
resents a disorder marked by self-​starvation, some 1999; Lester & Petrie, 1995; Pumariega, Gustavson,
experts (Habermas, 1989) have argued that an Gustavson, & Motes, 1994). In contrast, several stud-
excessive fear of weight gain is a necessary moti- ies have failed to find a significant association between
vating force behind food refusal, while others acculturation to Western values and disordered eating
(Katzman & Lee, 1997; Lee, 1995) have argued among Asian women (Gowen et al., 1999; Haudeck,
that it is not a core feature of AN. Cases of AN-​like Rorty, & Henker, 1999; Jackson, Keel, & Lee, 2006).
syndromes have been described in countries all over Instead, results suggest that cultural factors that con-
the world (e.g., South Africa, Nigeria, Zimbabwe, tribute to eating disorders may be native to some
Egypt, United Arab Emirates, Iran, China, Japan, Asian cultures (Jackson et  al., 2006). For example,
Korea, Russia, India, Pakistan, and Malaysia; Keel the virtues of fasting to the point of emaciation are
& Klump, 2003). Food refusal and emaciation are included in the Daoist text Sandong zhunang (Rieger,
reported for all cases and the syndrome is found pre- Touyz, Swain, & Beumont, 2001). Thus, Asian women
dominantly in adolescent and young adult women, may be more likely to suffer from an ED character-
but the presence of an excessive fear of weight gain ized by fasting and very low weight in the absence
as a motivating factor is not universal. In contrast, of weight phobia. Native American groups appear to
BN appears to exist mostly within the Western con- be at increased risk for obesity and increased BMI.
text (Keel & Klump, 2003). Feelings of being overweight have been associated
Another challenge for cross-​cultural research is with the use of purging among Chippewa and Native
the focus on DSM disorders, making it impossible Alaskan females (Rosen et  al., 1988; Story et  al.,
to comment on the possible presence of other EDs 1994). Thus, AN may be underrepresented among
(e.g., purging disorder) outside of Western culture. Native American girls and women, whereas purging
Because binge episodes require large stores of readily disorder may be more common.
edible food, BED, like BN, may be limited to places The failure of a diagnostic system to adequately
where food is abundant and easily obtained. In con- capture disorders of eating that are experienced by
trast, both purging disorder and subjective BED members of ethnic/​racial minority groups may have
may be relatively more common in non-​Western particularly pernicious effects for these individuals.
regions because, like AN, neither requires large Indeed, women from ethnic/​racial minority groups
quantities of food. are more likely to have undiagnosed and untreated
In addition to intercultural differences, research EDs compared with White women (Cachelin,
also supports intracultural differences with regard to Rebeck, Veisel, & Striegel-​Moore, 2001).
EDs. For example, Black American women appear to Given the fact that cultures change over time, it
be less likely to experience body dissatisfaction, dis- is important to recognize that eating disorder rates
ordered eating, and diagnosable EDs compared with and symptom presentations may accordingly change
White American women (Abrams, Allen, & Gray, over time. For instance, BN is currently observed
1993; Rhea, 1999; Striegel-​Moore et  al., 2003). In only in societies that have plentiful access to food;
particular, Black women appear to be protected more however, BN rates may increase in certain societies
from AN and BN than from BED (Smith, Marcus, over time if members of those societies encounter an
Lewis, Fitzgibbon, & Schreiner, 1998; Striegel-​Moore increase in the access they have to large quantities of
et al., 2003), suggesting that Black women may not food. As another example, researchers have demon-
be protected from eating disorders per se, but more strated the exposure to Western media increases the
protected from AN and BN than White women. In rates of ED symptoms among women from cultures
contrast, no specific protection from the develop- that are typically shielded from Western media expo-
ment of body image disturbance or disordered eat- sure (e.g., Eddy, Hennessey, & Thompson-​Brenner,
ing has been observed in Hispanic (Fitzgibbon et al., 2007). Therefore, it is reasonable to predict that as a
1998; le Grange, Stone, & Brownell, 1998; Lester & society’s exposure to traditional Western thin-​ideals
Petrie, 1995; Robinson et al., 1996), Asian (Barnett, increases, the observed rates of eating disorders may
Keel, & Conoscenti, 2001; Robinson et al., 1996) or increase.
Native American samples (Rosen et al., 1988; Smith
& Krejci, 1991; Story et al., 1994). Should There Be Separate Diagnostic
A strong traditional cultural identity appears to Criteria for Men and Women?
protect Black women from certain disordered eating Most research suggests that EDs in men closely
attitudes and behaviors (Chamorrow & Flores-​Ortiz, resemble EDs in women (Keel, Klump, Leon, &

18 Cl assification
Fulkerson, 1998; Leon, Fulkerson, Perry, Keel, & Further, youths with EDs often have difficulty
Klump, 1999; Minnich, Gordon, Holm-​Denoma, articulating or conceptualizing that a large portion
& Troop-​Gordon, 2014). Factors that appear to be of their self-​esteem is related to their body image.
more relevant for men, such as involvement in sports Acknowledging and conveying both fear of fatness
that require low weight (Hausenblas & Carron, and feelings of body dissatisfaction are processes
1999) and sexual orientation (Carlat, Camargo, & that require cognitive and emotional sophistication,
Herzog, 1997; Russell & Keel, 2002), may increase and it is well known that cognitive and emotional
the salience of weight control in a group that is oth- capacities continue developing well into adoles-
erwise less concerned about being thin. cence (Boyer, 2006). Accordingly, youths may not
There has been speculation that men may be at be able to convey verbally their cognitively based
risk for “reverse anorexia” (Pope, Katz, & Hudson, ED symptoms as effectively as adults.
1993) in addition to more traditionally recognized Owing to the difficulty many youths have with
EDs. Instead of viewing the body as much larger recognizing and reporting subjective emotional
than it really is, men with reverse AN are charac- experiences and thought patterns, some research-
terized by viewing their bodies as “puny” despite ers have suggested that focusing on overt behav-
their efforts and success at body-​ building (Pope ioral symptoms may be preferential with working
et al., 1993). This distorted perception contributes with young patients. For instance, Marcus and
to more extreme efforts (excessive exercise, high Kalarchian (2003) suggested that behaviors such as
protein diets, anabolic steroid use) to increase lean hiding food or secretive eating may be good indi-
muscle mass and overall body size. Because this con- cators of binge eating patterns in children, and
dition involves altered eating patterns, the use of Bryant-​Waugh and Lask (1995) suggested describ-
extreme weight control behaviors, and body image ing a childhood ED as one in which “there is an
disturbance, some have argued that it represents excessive preoccupation with weight or shape, and/​
an eating disorder (e.g., Pope et  al., 1993), albeit or food intake” (p. 191), at least in part because food
one rarely seen in women. However, this diagnostic intake can be observed by others. The DSM-​5 may
category is best represented in DSM-​5 currently as address this issue in some cases with the addition of
body dysmorphic disorder, with a muscle dysmor- avoidant/​restrictive food intake disorder, a diagnosis
phia specifier. Because extant research suggests that based on feeding problems that cannot be explained
men with EDs present similarly to females with by other EDs, such as AN.
EDs, there does not appear to be sufficient evidence Other researchers have suggested that the DSM
to justify separate DSM criteria at this time. authors consider adding modifiers to existing ED
symptoms for children and adolescents (Workgroup
Should There Be Separate Diagnostic for Classification of Eating Disorders in Children
Criteria for Children and Early Adolescents? and Adolescents [WCEDCA], 2007). This practice
Although EDs were initially considered Disorders has been useful for other disorders in the DSM-​5
Usually Diagnosed in Infancy, Childhood, and such as major depressive disorder (e.g., irritability
Adolescence by the DSM (van Son et  al., 2006) can be used in place of sad mood for children) and
and typically onset during adolescence (Hoek & obsessive-​compulsive disorder (e.g., children are not
Hoeken, 2003), recent editions of the DSM have required to express the reasons for their compul-
removed EDs from the section focusing on child- sions; APA, 2013). The WCEDCA proposed spe-
hood disorders. Despite this fact, many researchers cific criteria alterations for children and adolescents
and clinicians have raised concerns about whether with EDs such as coding behavioral observations
the DSM’s ED diagnostic criteria are as valid for by others (i.e., parents report their child engages in
youngsters as they are for later adolescents and behaviors that indicate he/​she has an intense fear
adults. of fatness) rather than self-​reports of the affected
A potential problem with applying the DSM-​ youngster (i.e., instead of requiring the child to
5 ED criteria to youths has to do with the devel- directly state that he/​she is terrified of becoming fat
opmentally acquired ability to observe and have even though he/​she is underweight).
insight into one’s own thoughts and feelings. For It is possible that the transdiagnostic model pro-
instance, studies have highlighted the frequent fail- posed by, Fairburn et  al. (2008) may improve the
ure of youths to endorse fear of weight gain despite validity of ED diagnoses for all who suffer from eat-
the presence of behaviors that clearly contribute ing pathology, including youths. In this model, a
to harmful weight loss (Fairburn et  al., 2003). single diagnosis of ED would be given to all affected

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 19

individuals; however, qualifiers relevant to specific Boyer, T. (2006). The development of risk-​ taking:  A  multi-​
manifestations and/​or clinically meaningful variables perspective review. Developmental Review, 26, 291–​345.
Bryant-​Waugh, R., & Lask, B. (1995). Eating disorders in chil-
(e.g., those related to developmental differences in dren. Journal of Child Psychology and Psychiatry, 36, 191–​202.
ED presentations) would be possible. Others have Bulik, C. M., Thornton, L. M., Root, T. L., Pisetsky, E. M.,
suggested that a diagnostic approach be developed Lichtenstein, P., & Pedersen, N. L. (2010). Understanding
in which youths who present with eating pathology the relation between anorexia nervosa and bulimia nervosa
are required to meet some, but not all, diagnostic in a Swedish national twin sample. Biological Psychiatry,
67, 71–​77.
criteria to qualify for an ED diagnosis. For instance, Cachelin, F. M., Rebeck, R., Veisel, C., & Striegel-​Moore, R.
Hebebrand, Casper, Treasure, and Schweiger (2004) H. (2001). Barriers to treatment for eating disorders among
described a model in which a broad array of symp- ethnically diverse women. International Journal of Eating
toms would be assessed (e.g., criteria could be related Disorders, 30, 269–​278.
to age, gender-​specific presentations, and medical Carlat, D. J., Camargo, C. A., Jr., & Herzog, D. B. (1997).
Eating disorders in males: A report on 135 patients. American
conditions such as bradycardia, hypothermia, hypo- Journal of Psychiatry, 154, 1127–​1132.
tension). Not all candidates for an ED diagnosis Cassin, S. E., & von Ranson, K. M. (2005). Personality and eat-
would be evaluated on all criteria, but rather clini- ing disorders: A decade in review. Clinical Psychology Review,
cians would focus on evaluating the criteria that are 25, 895–​991.
relevant given the individual’s demographic and Catell, R. B. (1946). Description and measurement of personality.
New York, NY: World Book.
symptomatic presentation. The proposed models of Chamorrow, R., & Flores-​Ortiz, Y. (2000). Acculturation and
Fairburn and Hebebrand et al. should be empirically disordered eating patterns among Mexican American women.
evaluated further before conclusions are reached International Journal of Eating Disorders, 28, 125–​129.
regarding their utility in the diagnosis of EDs expe- Crow, S. J., Swanson, S. A., Peterson, C. B., Crosby, R. D.,
rienced by youths. Wonderlich, S. A., & Mitchell, J. E. (2012). Latent class
analysis of eating disorders:  Relationship to mortality.
Journal of Abnormal Psychology, 121, 225–​231.
Summary Deschartres, A., Huas, C., Godart, N., Pousset, M., Pham,
In sum, the developers of the next DSM are A., Divac, S. M.,   .  .  .  Falissard, B. (2011). Outcomes of
charged with the task of creating a classification empirical eating disorder phenotypes in a clinical female
system that facilitates clinical work and scientific sample: Results from a latent class analysis. Psychopathology,
44, 12–​20.
research, but also, as Gould (1989, p. 98) suggests Eddy, K. T., Crosby, R. D., Keel, P. K., le Grange, D., Powers,
in the opening quote, reflects “theories about the P., Mitchell, J. E.,  . . .  Wonderlich, S. A. (2009). Empirical
basis of natural order.” There is much to learn about identification and validation of eating disorder phenotypes
the generalizability of the DSM ED categories with in a multisite clinical sample. Journal of Nervous and Mental
regard to gender, developmental period, and diverse Disease, 197, 41–​49.
Eddy, K. T., Dorer, D. J., Franko, D. L., Tahilani, K., Thompson-​
cultural groups. The accumulation of more empiri- Brenner, H., & Herzog, D. B. (2008). Diagnostic crossover
cal data in these areas may lead to a greater under- in anorexia nervosa and bulimia nervosa:  Implications for
standing of the nature of EDs and to the most valid DSM-​5. American Journal of Psychiatry, 165, 245–​250.
form of classification. Eddy, K. T., Hennessey, M., & Thompson-​Brenner, H. (2007).
Eating pathology in East African women: The role of media
exposure and globalization. Journal of Nervous and Mental
References Disease, 195, 196–​202.
Abrams, K. K., Allen, L. R., & Gray, J. J. (1993). Disordered eating
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive
attitudes and behaviors, psychological adjustment, and ethnic
behaviour therapy for eating disorders:  A  “transdiagnostic”
identity: A comparison of black and white female college stu-
theory and treatment. Behaviour research and therapy, 41,
dents. International Journal of Eating Disorders, 14, 49–​57.
509–​528.
Agras, W. S., Crow, S., Mitchell, J. E., Halmi, K. A., & Bryson,
Fairburn, C. G., Cooper, Z., & Shafran, R. (2008). Cognitive
S. (2009). A 4-​year prospective study of eating disorder
behaviour therapy for eating disorders:  A  “transdiagnostic”
NOS compared with full eating disorder syndromes.
theory and treatment. Behaviour Research and Therapy, 41,
International Journal of Eating Disorder, 42, 565–​570.
509–​517.
American Psychiatric Association (APA). (2013). Diagnostic and
Fichter, M. M., & Quadflieg, N. (2007). Long-​term stability
statistical manual of mental disorders (5th ed.). Washington,
of eating disorder diagnoses. International Journal of Eating
DC: Author.
Disorders, Special Issue on Diagnosis and Classification, 40
Arcelus, J., Mitchell, A. J., Wales, J., & Nielsen, S. (2011).
(Supplement), S61–​S66.
Mortality rates in patients with anorexia nervosa and other
Fichter, M. M., Quadflieg, N., & Rief, W. (1994). Course
eating disorders:  A  meta-​analysis of 36 studies. Archives of
of multi-​ impulsive bulimia. Psychological Medicine, 24,
General Psychiatry, 68, 724–​733.
591–​604.
Barnett, H. L., Keel, P. K., & Conoscenti, L. M. (2001). Body
Fitzgibbon, M. L., Spring, B., Avellone, M. E., Blackman, L. R.,
type preferences in Asian and Caucasian college students. Sex
Pingitore, R., & Stolley, M. R. (1998). Correlates of binge
Roles, 45, 867–​878.

20 Cl assification
 eating in Hispanic, black, and white women. International Hausenblas, H. A., & Carron, A. V. (1999). Eating disorder indi-
Journal of Eating Disorders, 24, 43–​52. ces and athletes: An integration. Journal of Sport and Exercise
Frank, G. K, Bailer, U. F., Henry, S. E., Drevets, W., Meltzer, C. Psychology, 21, 230–​258.
C., Price, J. C.,  . . .  Kaye, W. H. (2005). Increased dopa- Hebebrand, J., Casper, R., Treasure, J. L., & Schweiger, U.
mine D2/​D3 receptor binding after recovery from anorexia (2004). The need to revise the diagnostic criteria for anorexia
nervosa measured by positron emission tomography and nervosa. Journal of Neural Transmission, 111, 827–​840.
[11c]raclopride. Biological Psychiatry, 58, 908–​981. Herzog, D. B., Sacks, N. R., Keller, M. B., Lavori, P. W., von
Frank, G. K., Kaye, W. H., Meltzer, C. C., Price, J. C., Greer, Ranson, K. B., & Gray, H. M. (1993). Patterns and predic-
P., McConaha, C., & Skovira, K. (2002). Reduced 5-​HT2A tors of recovery in anorexia nervosa and bulimia nervosa.
receptor binding after recovery from anorexia nervosa. Journal of the American Academy of Child and Adolescent
Biological Psychiatry, 52, 896–​906. Psychiatry, 32, 835–​842.
Franko, D. L., & Keel, P. K. (2006). Suicidality in eating dis- Hoek, H., & Hoeken, D. (2003). Review of prevalence and
orders:  Occurrence, correlates, and clinical implications. incidence of eating disorders. International Journal of Eating
Clinical Psychology Review, 26, 769–​782. Disorders, 34, 383–​396.
Gangestad, S., & Snyder, M. (1985). “To carve nature at its Holliday, J., Landau, S., Collier, D., & Treasure, J. (2006). Do
joints”:  On the existence of discrete classes in personality. illness characteristics and familial risk differ between women
Psychological Review, 92, 317–​349. with anorexia nervosa grouped on the basis of personality
Gleaves, D. H., Lowe, M. R., Green, B. A., Cororve, M. B., pathology? Psychological Medicine, 36, 529–​538.
& Williams, T. L. (2000). Do anorexia and bulimia ner- Holm-​Denoma, J. M., Richey, J. A., Joiner, T. E., Jr. (2010).
vosa occur on a continuum? A taxometric analysis. Behavior The latent structure of dietary restraint, body dissatisfac-
Therapy, 31, 195–​219. tion, and drive for thinness: A series of taxometric analyses.
Gleaves, D. H., Lowe, M. R., Snow, A. C., Green, B. A., & Psychological Assessment, 22, 788–​792.
Murphy-​Eberenz, K. P. (2000). Continuity and discontinu- Hsu, L. K. G., Mulliken, B., McDonagh, B., Das, S. K., Rand,
ity models of bulimia nervosa:  A  taxometric investigation. W., Fairburn, C. G.,  . . .  Roberts, S. (2002). Binge eating
Journal of Abnormal Psychology, 109, 56–​68. disorder in extreme obesity. International Journal of Obesity &
Gonzales, E. G. (2001). Evaluation of self-​concept, body sat- Related Metabolic Disorders, 26, 1398–​1403.
isfaction, and social skills in anorexia and bulimia nervosa. Hudson, J. I., Hiripi, E., Pope, H. G., & Kessler, R. C.
Clinica y Salud, 12, 239–​304. (2007). The prevalence and correlates of eating disorders
Gordon, K. H., Holm-​Denoma, J. M., Smith, A. R., Fink, E. L., in the National Comorbidity Survey Replication. Biological
& Joiner, T. E., Jr. (2007). Taxometric analysis: Introduction Psychiatry, 61, 348–​358.
and overview. International Journal of Eating Disorders, 40, Jackson, S. C., Keel, P. K., & Lee, Y. H. (2006). Trans-​cultural
S35–​S39. comparison of disordered eating in Korean women.
Gould, S. J. (1989). Wonderful life:  The Burgess shale and the International Journal of Eating Disorders, 39, 498–​502.
nature of history. New York, NY: Norton. Johnson, C., & Connors, M. E. (1987). The etiology and treatment
Gowen, L. K., Hayward, C., Killen, J. D., Robinson, T. N, of bulimia nervosa:  A  biopsychosocial perspective. New  York,
& Taylor, C. B. (1999). Acculturation and eating disor- NY: Basic Books.
der symptoms in adolescent girls. Journal of Research on Johnson, J. G., Spitzer, R. L., & Williams, B. W. (2001). Health
Adolescence, 9, 67–​83. problems, impairment, and illnesses associated with bulimia
Grilo, C. M., Devlin, M. J., Cachelin, F. M., & Yanovski, S. nervosa and binge eating disorder among primary care and
Z. (1997). Report of the National Institutes of Health obstetric gynecology patients. Psychological Medicine, 31,
(NIH):  Workshop on the development of research pri- 1455–​1466.
orities in eating disorders. Psychopharmacology Bulletin, 33, Katzman, M. A., & Lee, S. (1997). Beyond body image:  The
321–​333. integration of feminist and transcultural theories in the
Grilo, C. M., Hrabosky, J. I., White, M. A., Allison, K. C., understanding of self starvation. International Journal of
Stunkard, A. J., & Masheb, R. M. (2008). Overvaluation of Eating Disorders, 22, 385–​394.
shape and weight in binge eating disorder and overweight Keel, P. K., Crosby, R. D., Hildebrandt, T. B., Haedt-​Matt, A.
controls:  Refinement of a diagnostic construct. Journal of A., & Gravener, J. A. (2013). Evaluating new severity dimen-
Abnormal Psychology, 117, 414–​419. sions in the DSM-​5 for bulimic syndromes using mixture
Habermas, T. (1989). The psychiatric history of anorexia nervosa modeling. International Journal of Eating Disorders, 46,
and bulimia nervosa:  Weight concerns and bulimic symp- 108–​118.
toms in early case reports. International Journal of Eating Keel, P. K., Brown, T. A, Holland, L. A., & Bodell, L. P. (2012).
Disorders, 8, 259–​273. Empirical classification of eating disorders. Annual Review of
Hartmann, A., Zeeck, A., & Barrett, M. S. (2010). Interpersonal Clinical Psychology, 8, 381–​404.
problems in eating disorders. International Journal of Eating Keel, P. K., Holm-​Denoma, J., Crosby, R. D., Haedt-​Matt, A.
Disorders, 43, 619–​627. A., Gravener, J. A., & Joiner, T. E. (2011). Latent struc-
Haslam, N., Holland, E., & Kuppens, P. (2012). Categories ver- ture of bulimic syndromes:  An empirical approach utiliz-
sus dimensions in personality and psychopathology: A quan- ing latent profile analyses and taxometric analyses. In R.
titative review of taxometric research. Psychological Medicine, H. Striegel-​Moore, S. A. Wonderlich, B. T. Walsh, & J. E.
42, 903–​920. Mitchell (Eds.), Developing an evidence-​based classification of
Haudeck, C., Rorty, M., & Henker, B. (1999). The role of eth- eating disorders: Scientific findings for DSM-​5 (pp. 145–​164).
nicity and parental bonding in the eating and weight con- Washington, DC: American Psychiatric Association.
cerns of Asian-​ American and Caucasian college women. Keel, P. K., Fichter, M., Quadflieg, N., Bulik, C. M., Baxter,
International Journal of Eating Disorders, 25, 425–​433. M. G., Thornton, L.,  . . .  Kaye, W. H. (2004). Application

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 21

 of a latent class analysis to empirically define eating dis- Preti, A., Rocchi, M. B., Sisti, D., Camboni, M. V., & Miotto, P.
order phenotypes. Archives of General Psychiatry, 61, (2011). A comprehensive meta-​analysis of the risk of suicide
192–​200. in eating disorders. Acta Psychiatrica Scandinavia, 124, 6–​17.
Keel, P. K., & Klump, K. L. (2003). Are eating disorders culture-​ Pumariega, A. J., Gustavson, C. R., Gustavson, J. C., & Motes, P.
bound syndromes? Implications for conceptualizing their S. (1994). Eating attitudes in African-​American women: The
genetic bases. Psychological Bulletin, 129, 747–​769. Essence Eating Disorders Survey. Eating Disorders, 2, 5–​16.
Keel, P. K., Klump, K. L., Leon, G. R., & Fulkerson, J. A. Racine, S. E., & Wildes, J. E. (2015). Emotion dysregulation and
(1998). Disordered eating in adolescent males from a school-​ anorexia nervosa:  An exploration of the role of childhood
based sample. International Journal of Eating Disorders, 23, abuse. International Journal of Eating Disorders, 48, 55–​58.
125–​132. Rhea, D. J. (1999). Eating disorder behaviors of ethnically
Kendler, K. S. (1990). Toward a scientific psychiatric nosol- diverse urban female adolescent athletes and non-​athletes.
ogy: Strengths and limitations. Archives of General Psychiatry, Journal of Adolescence, 22, 379–​388.
47, 969–​973. Rieger, E., Touyz, S. W., Swain, T., & Beumont, P. J. (2001).
Kupfer, D., First, M. B., & Regier, D. (2002). A research Cross-​cultural research on anorexia nervosa:  Assumptions
agenda for DSM-​V. Washington, DC: American Psychiatric regarding the role of body weight. International Journal of
Association. Eating Disorders, 29, 205–​215.
Lacey, J. H. (1993). Self-​damaging and addictive behavior in Robinson, T. N., Killen, J. D., Litt, I. F., Hammer, L. D., Wilson,
bulimia nervosa: A cachement area study. British Journal of D. M., Haydel, K. F.,  . . .  Taylor, C. B. (1996). Ethnicity
Psychiatry, 163, 190–​194. and body dissatisfaction:  Are Hispanic and Asian girls at
Lazarsfeld, P. F., & Henry, N. W. (1968). Latent structure analysis. increased risk for eating disorders? Journal of Adolescent
Boston, MA: Houghton Mifflin. Health, 19, 384–​393.
le Grange, D., Stone, A. A., & Brownell, K. D. (1998). Eating dis- Rosen, L., Shafer, C. L., Dummer, G. M., Cross, L. K., Deuman,
turbances in white and minority female dieters. International G. W., & Malmberg, S. R. (1988). Prevalence of patho-
Journal of Eating Disorders, 24, 395–​403. genic weight-​ control behaviors among Native American
Lee, S. (1995). Self-​starvation in context:  Towards a culturally women and girls. International Journal of Eating Disorders,
sensitive understanding of anorexia nervosa. Social Science 7, 807–​811.
Medicine, 41, 25–​36. Ruscio, J., & Ruscio, A. M. (2002). A structure-​based approach
Leon, G. R., Fulkerson, J. A., Perry, C. L., Keel, P. K., & Klump, to psychological assessment: Matching measurement models
K. L. (1999). Three to four year prospective evaluation of to latent structure. Assessment, 9, 4–​16.
personality and behavioral risk factors for later disordered Ruscio, J., Ruscio, A. M., & Meron, M. (2007). Applying the
eating in adolescent girls and boys. Journal of Youth and bootstrap to taxometric analysis: Generating empirical sam-
Adolescence, 28, 181–​196. pling distributions to help interpret results. Multivariate
Lester, R., & Petrie, T. A. (1995). Personality and physical corre- Behavioral Research, 42, 349–​386.
lates of bulimic symptomatology among Mexican American Russell, C. J., & Keel, P. K. (2002). Homosexuality as a specific
female college students. Journal of Counseling Psychology, 42, risk factor for eating disorders in men. International Journal
199–​203. of Eating Disorders, 31, 300–​306.
Lubke, G. H., & Muthen, B. (2005). Investigating population Smith, J. E., & Krejci, J. (1991). Minorities join the major-
heterogeneity with factor mixture models. Psychological ity: Eating disturbances among Hispanic and Native American
Methods, 10, 21–​39. youth. International Journal of Eating Disorders, 10, 179–​186.
Marcus, M., & Kalarchian, M. (2003). Binge eating in children Smith, D. E., Marcus, M. D., Lewis, C., Fitzgibbon, M., &
and adolescents. International Journal of Eating Disorders, 34 Schreiner, P. (1998). Prevalence of binge eating disorder,
(Supplement), S47–​S57. obesity, and depression in a biracial cohort of young adults.
Meehl, P. E. (1992). Factors and taxa, traits and types, differences Annals of Behavioral Medicine, 20, 227–​232.
of degree and differences in kind. Journal of Personality, 60, Steiger, H., Israël, M., Gauvin, L., Kin, N., & Young, S. N.
117–​174. (2003). Implications of compulsive and impulsive traits for
Minnich, A. M., Gordon, K. H., Holm-​Denoma, J. M., & serotonin status in women with bulimia nervosa. Psychiatry
Troop-​Gordon, W. (2014). A test of an interactive model of Research, 120, 219–​229.
binge eating among undergraduate men. Eating Behaviors, Steiger, H., & Stotland, S. (1996). Prospective study of
15, 625–​631. outcome in bulimics as a function of Axis II comor-
National Institute of Mental Health. (2016, November 5). bidity:  Long-​ term responses on eating and psychiatric
Research Domain Criteria (RDoc) Retrieved from https://​ symptoms. International Journal of Eating Disorders, 20,
www.nimh.nih.gov/​research-​priorities/​rdoc/​index.shtml 149–​162.
Olatunji, B. O., Broman-​Fulks, J. J., Ciesielski, B. G., Zawilinski, Story, M., Hauck, F. R., Broussard, B. A., White, L. L., Resnick,
L. L., Shewmaker, S., & Wall, D. (2012). A taxometric inves- M. D., & Blum, R. W. (1994). Weight perceptions and
tigation of the latent structure of eating disorders. Psychiatry weight control practices in American Indian and Alaska
Research, 197, 97–​102. Native adolescents. A  national survey. Archives of Pediatrics
Pinheiro, A. P., Bulik, C. M., Sullivan, P. F., & Machado, P. P. and Adolescent Medicine, 148, 567–​571.
(2008). An empirical study of the typology of bulimic symp- Striegel-​Moore, R. H., Dohm, F. A., Kraemer, H. C., Taylor, C.
toms in young Portuguese women. International Journal of B., Daniels, S., Crawford, P. B., & Schreiber, G. B. (2003).
Eating Disorders, 41, 251–​258. Eating disorders in White and Black women. American
Pope, H. G., Jr., Katz, D. L., & Hudson, J. I. (1993). Anorexia Journal of Psychiatry, 160, 1326–​1331.
nervosa and “reverse anorexia” among 108 male bodybuild- Strober, M., Freeman, R., Lamper, C., Diamond, J., & Kaye,
ers. Comprehensive Psychiatry, 34, 406–​409. W. (2000). Controlled family study of anorexia nervosa and

22 Cl assification
 bulimia nervosa:  Evidence of shared liability and transmis- Westen, D., & Harden-​Fischer, J. (2001). Personality profiles
sion of partial syndromes. American Journal of Psychiatry, in eating disorders: Rethinking the distinction between Axis
157, 393–​401. I and Axis II. American Journal of Psychiatry, 158, 547–​562.
Strube, M. J. (1989). Evidence for the type in Type A  behav- Wildes, J. E., Forbush, K. T., & Markon, K. E. (2013).
ior:  A  taxometric analysis. Journal of Personality and Social Characteristics and stability of empirically derived anorexia
Psychology, 56, 972–​987. nervosa subtypes: Towards the identification of homogenous
Thomas, J. H., Eddy, K. T., Ruscio, J., Ng, K. L., Casale, low-​weight eating disorder phenotypes. Journal of Abnormal
K. E., Becker, A. E., & Lee, S. (2015). Do recognizable Psychology, 122, 1031–​1041.
lifetime eating disorder phenotypes naturally occur in a Wilfley, D. E., Bishop, M. E., Wilson, G. T., & Agras, W. S.
culturally Asian population? A combined latent profile and (2007). Classification of eating disorders:  Toward DSM-​V.
taxometric approach. European Eating Disorders Review, 23, International Journal of Eating Disorders, 40, S123–​S129.
199–​209. Williamson, D. A., Gleaves, D. H., & Stewart, T. M. (2005).
Thompson-​Brenner, H., Eddy, K. T., Satir, D. A., Boisseau, Categorical versus dimensional models of eating disor-
C. L., & Westen, D. (2008). Personality subtypes in ado- ders: An examination of the evidence. International Journal
lescents with eating disorders: Validation of a classification of Eating Disorders, 37, 1–​10.
approach. Journal of Child Psychology and Psychiatry, 49, Williamson, D. A., Womble, L. G., Smeets, M. A. M., Netemeyer,
170–​180. R. G., Thaw, M., Kutlesic, V.,  . . .  Gleaves, D. H. (2002).
Tozzi, F., Thornton, L. M., Klump, K. L., Fichter, M. M., The latent structure of eating disorder symptoms:  A  factor
Halmi, K. A., Kaplan, A. S.,   .  .  .  Kaye W. H. (2005). analytic and taxometric investigation. American Journal of
Symptom fluctuation in eating disorders:  Correlates of Psychiatry, 159, 412–​418.
diagnostic crossover. American Journal of Psychiatry, 162, Wonderlich, S. A., Crosby, R. D., Joiner, T. E.  J., Peterson,
732–​740. C. B., Bardone-​Cone, A. M., Klein, M. H.,   .  .  .  Vrshek,
Tylka, T. L., & Subich, L. M. (2003). Revisiting the latent S. (2005). Personality subtyping and bulimia ner-
structure of eating disorders: Taxometric analyses with non-​ vosa: Psychopathological and genetic correlates. Psychological
behavioral indicators. Journal of Counseling Psychology, 50, Medicine, 35, 649–​657.
276–​286. Wonderlich, S. A., Joiner, T. E., Jr., Keel, P. K., Williamson,
Uebersax, J. S. (1999). Probit latent class analysis with dichoto- D. A., & Crosby, R. D. (2007). Eating disorder diag-
mous or ordered category measures:  Conditional indepen- noses:  Empirical approaches to classification. American
dence/​dependence models. Applied Psychological Measures, Psychologist, 62, 167–​180.
23, 283–​297. Wonderlich, S. A., & Mitchell, J. E. (1997). Eating disorders
van Son, G., van Hoeken, D., Aad, I., Bartelds, A., van Furth, and comorbidity: Empirical, conceptual, and clinical impli-
E., & Hoek, E. (2006). Time trends in the incidence of cations. Psychopharmacology Bulletin, 33, 381–​390.
eating disorders:  A  primary care study in the Netherlands. Wonderlich, S. A., Mitchell, J. E., Peterson, C. B., & Crow, S.
International Journal of Eating Disorders, 39, 565–​569. (2001). Integrative cognitive therapy for bulimic behav-
Wade, T. D., Crosby, R. D., & Martin, N. G. (2006). Use of ior. In R. Striegel-​Moore & L. Smolak (Eds.), Eating dis-
latent profile analysis to identify eating disorder pheno- orders:  Innovations and directions for research and practice
types in an adult Australian twin cohort. Archives of General (pp. 173–​195). New  York, NY:  American Psychological
Psychiatry, 63, 1377–​1384. Association.
Waldman, I. D., & Lilienfeld, S. O. (2001). Applications of taxo- Wonderlich, S. A., & Swift, W. J. (1990). Borderline versus other
metric methods to problems of comorbidity:  Perspectives personality disorders in the eating disorders. International
and challenges. Clinical Psychology:  Science and Practice, 8, Journal of Eating Disorders, 9, 629–​638.
520–​527. Workgroup for Classification of Eating Disorders in Children
Waller, N. G., & Meehl, P. E. (1998). Multivariate taxometric and Adolescents. (2007). Classification of child and ado-
procedures:  Distinguishing types from continua. Newberry lescent eating disturbances. International Journal of Eating
Park, CA: Sage. Disorders, 40 (Supplement), S117–​S122.

Gordon, Holm-Denoma, Dougl as, Crosby, Wonderlich 23

 C H A PT E R


Research Domain Criteria: The Impact of
2 RDoC on the Conceptualization
of Eating Disorders

Cara Bohon

Abstract
A primary goal of the research domain criteria (RDoC) project from the National Institute of Mental
Health in the United States is to better characterize and understand the pathology and etiology of
mental illness by examining constructs with biological underpinnings and their effects on psychiatric
symptoms. This endeavor shows promise in helping to better conceptualize dysfunction in the field of
eating disorders, where there appears to be great heterogeneity within diagnostic groups. Treatments
designed for a particular diagnosis may result in improved remission rates if they instead target underlying
mechanisms of eating disorder symptoms. This system is not without challenge and limitations, however.
This chapter includes a brief review of relevant literature on the proposed RDoC functional domains
in eating disorders and discussion of the benefits and costs of this type of approach in improving
patient care.
Key Words:  RDoC, research domain criteria, eating disorder, classification, research, dysfunction

Introduction (Insel & Cuthbert, 2009). This project emphasizes

Prior and current editions of the Diagnostic the study of observable behavior and neurobiology
Statistical Manual of Mental Disorders (DSM; across a broad spectrum of functioning, from nor-
APA, 2013) have relied on presentation of symp- mal to abnormal. It highlights brain and behavior
toms to determine a diagnosis without consider- relationships over typical symptom clusters used to
ation of pathology or etiology. Thus, diagnoses classify illness. It anchors classification in models of
have tended to be descriptive rather than explan- neural circuitry. As Cuthbert (2014) notes, “RDoC
atory. When research into the biology that could was founded not out of hubris, but rather from
underlie these diagnoses became more prevalent, a sense of humility that we know so little about
it often failed to identify consistent underpin- our disorders as currently conceived that it is nec-
nings to these diagnoses due in part to exten- essary to start over again with a new approach.”
sive biological heterogeneity within a diagnosis. Although it is not intended to be used in clini-
Additionally, although there was improvement cal settings yet, the long-​term goal of the RDoC
in the most recent (fifth) edition of the DSM, project appears to be an improvement in how we
extensive comorbidity among diagnoses and over- identify dysfunction in our patients and develop
specification of categories still remain. In light of more effective treatment approaches. Given the
these shortcomings of the current classification emphasis on this new approach within a primary
system, the National Institute of Mental Health funding agency for research in eating disorders, it
in the United States developed a new initiative, is important to consider the impact of RDoC on
the research domain criteria (RDoC) project our conceptualization of these illnesses.

24
Current Classification of Eating Disorders to treatment failure due to this heterogeneity. An
The creation of the DSM-​ 5 (APA, 2013) intervention that is effective for one patient with a
addressed a number of concerns from the prior particular presentation of symptoms or underlying
diagnostic manual. One of the primary goals was mechanism may not be for another with a slightly
to reduce the number of patients in the “catch-​all” different etiology. Further, many treatment studies
category of eating disorder not otherwise specified, use samples of patients with limited comorbidities
which had previously encompassed the largest per- or constrained symptom presentations (Strober,
centage of patients diagnosed with an eating dis- 2014), which hinders generalizability of the effects.
order (Keel, Brown, Holland, & Bodell, 2012). To Thus, treatments that perform well in tightly con-
do this, the diagnostic criteria for anorexia nervosa trolled trials may fail when applied to more complex
and bulimia nervosa were relaxed. The amenorrhea presentations.
criterion was removed from the diagnosis of ano-
rexia nervosa, and the definition of “low weight” Research Domain Criteria and Eating
was given more flexibility for clinical judgment. Disorders
The minimum frequency of binge eating and purg- Assessing patients in terms of dimensions of
ing episodes in bulimia nervosa was reduced from function that are linked to underlying biology could
twice weekly to once weekly on average over the address some of these challenges. The RDoC project
prior 3 months. Binge eating disorder was officially is currently structured as a matrix with five domains
adopted into the list of diagnostic categories, fur- and eight units of analysis, although the matrix is
ther reducing the not otherwise specified group. not intended to be fixed or comprehensive. Indeed,
Avoidant restrictive food intake disorder was also there are continuing discussions to evaluate and edit
included in the DSM-​5 to address the prevalence the matrix. The domains represent proposed func-
of patients who presented with eating disturbance tional dimensions based on translational research
and food restriction but did not meet criteria for on genes, neural circuits, and behavior and include
other feeding and eating disorders, often lacking a negative valence systems, positive valence systems,
desire to control one’s shape or weight or disturbed cognitive systems, social processes, and arousal and
body image. The DSM-​5 also includes a group of regulatory systems. The RDoC project is currently
diagnoses officially coded as “other specified feeding considering adding a sixth domain: motor systems.
and eating disorders” including atypical anorexia Each of these domains is broken down into con-
nervosa, bulimia nervosa of low frequency and/​or structs and subconstructs. For instance, the nega-
limited duration, binge eating disorder of low fre- tive valence domain is subdivided into acute threat
quency and/​or limited duration, purging disorder, (“fear”), potential threat (“anxiety”), sustained
and night eating syndrome. And finally, the DSM-​5 threat, loss, and frustrative nonreward. Each domain
includes a category of “unspecified feeding or eating is evaluated across eight units of analysis, which
disorder” given when symptoms of an eating disor- include genes, molecules, cells, circuits, physiology,
der that cause significant distress or impairment do behavior, self-​report, and paradigms. A  patient’s
not meet full criteria for any other disorder in the functioning in a particular domain can be assessed
diagnostic class. Thus, the updated diagnostic man- at each of these levels. Describing function in these
ual includes more diagnoses in attempt to increase domains could help develop precision medicine—​
the specificity of the diagnosis. more specific and individualized treatment plans.
Despite the extensive improvements made to the One potential benefit of an RDoC approach
DSM-​5 to improve diagnosis and reduce the prev- in eating disorders is in understanding diagnostic
alence of the eating disorder not otherwise speci- crossover. Although stability of diagnosis is more
fied category, there are still patients who require common in almost all eating disorders, there is a
the unspecified diagnosis, as well as heterogeneity percentage of patients who cross over from one
within diagnostic categories (Wildes & Marcus, diagnosis to another (see Keel et  al., 2012 for a
2015). This heterogeneity exists both for severity of review). In binge eating disorder, crossover to a
symptom presentation within a diagnosis and poten- diagnosis of bulimia nervosa is more likely than sta-
tially underlying mechanisms. This is important bility in the diagnosis. The crossover between eating
because treatments for eating disorders have been disorder diagnoses could suggest the existence of
lagging and do not result in symptom remission underlying mechanisms that are driving the various
for the majority of patients (Hay, 2013). Perhaps eating disorder behaviors. For instance, binge eating
using diagnoses to guide treatment decisions leads and purging may both reflect an underlying struggle

Bohon 25
with cognitive control, one of the constructs within of how various RDoC domains are implicated in
the cognitive systems domain of RDoC. Thus, eating disorders, with the potential for improved
individuals who are engaging in binge eating and/​ treatment.
or purging may move between various eating dis-
order diagnoses that include those behaviors, but Negative Valence Systems and Eating
they may not necessarily require a different treat- Disorders
ment approach when the diagnosis changes. Indeed, One construct under the negative valence sys-
rather than develop separate treatment protocols for tems domain of RDoC that appears relevant to
binge eating disorder, bulimia nervosa, and purg- patients with eating disorders is sustained threat.
ing disorder, perhaps a treatment approach that This is defined as an aversive emotional state due
addresses the underlying deficit in cognitive control to prolonged exposure to stimuli, states, or condi-
would be more tailored for particular patients strug- tions. These stimuli can be internal or external and
gling with these behaviors. may be actual or anticipated, but the effects caused
In fact, modular treatment protocols and pre- by the prolonged exposure persist in the absence of
cision medicine could be a primary benefit of an the threat. Using the RDoC matrix, we can iden-
RDoC approach to conceptualizing eating disor- tify patterns of behavior and dysregulated circuit
ders. One example of a modular treatment protocol function present in some eating disorders con-
was developed by Bruce Chorpita and colleagues gruent with sustained threat. At the circuit level,
to treat anxiety disorders in children (Chorpita, there is evidence of disrupted activity in the habit
Taylor, Francis, Moffitt, & Austin, 2004). This pro- system, including the dorsal striatum, in patients
tocol includes individual treatment modules that with anorexia nervosa while selecting food choices
can be assembled according to individual needs of (Foerde, Steinglass, Shohamy, & Walsh, 2015),
the patient. It includes decision flowcharts to guide as well as evidence of hyperactivity in the amyg-
module selection and sequencing. The protocol is dala among restrictive patients in response to food
now referred to as MATCH (Modular Approach to images versus nonfood images (Joos et  al., 2011).
Therapy for Children with Anxiety, Depression, or When selecting food choices, avoidance of high-​fat
Conduct Problems) and has outperformed standard foods was common (Foerde et al., 2015), and anx-
evidence-​based treatments for children (Weisz et al., ious arousal was present while viewing food images
2012). Applying this to an RDoC model of psycho- (Joos et al., 2011). Avoidance and anxious arousal
pathology, modules could be developed to address are two of the behaviors associated with sustained
deficits in various domains and be selected accord- threat, and harm avoidance has indeed been found
ing to an individual’s need, as opposed to the use of in studies of personality traits in anorexia nervosa
a single approach to treat everyone within a diag- (Lilenfeld, Wonderlich, Riso, Crosby, & Mitchell,
nostic category. Researchers have recently taken this 2006). There is also evidence of attentional bias to
approach in the treatment of adolescent depression angry-​threat faces in anorexia nervosa, which is con-
(Henje Blom et al., 2014), creating treatment mod- gruent with sustained threat (Harrison, Tchanturia,
ules that address domains of function from RDoC & Treasure, 2010). Further evidence that disrup-
that can be used as needed for particular patients. tions in sustained threat are present in eating dis-
How might this look for eating disorders? One orders is the presence of heightened sensitivity to
example that has begun to accrue evidence in ano- punishment across eating disorders (Harrison,
rexia nervosa is cognitive remediation therapy O’Brien, Lopez, & Treasure, 2010). Although the
(CRT) (Tchanturia, Davies, & Campbell, 2007), evidence for problems with sustained threat in eat-
which was initially developed to address difficulty ing disorders comes from cross-​sectional studies, it
with cognition in schizophrenia (Wykes & van der may be that this unresolved anxious arousal under-
Gaag, 2001). Its use for anorexia nervosa is based lies some aspects of disordered eating, particularly
on evidence of abnormalities in cognitive systems food restriction. It will be important to investigate
(one of the RDoC domains) that may underlie the sustained threat across eating disorder diagnoses,
illness, such as difficulties in cognitive flexibility but sensitivity to sustained threat may lead individ-
and central coherence. Looking beyond this exam- uals to restrict food intake in order to avoid certain
ple, we can consider other abnormalities in RDoC threats, such as peer rejection. The food restriction
domains present in eating disorders as potential then leads to weight loss and severe malnutrition in
treatment targets. Although not a comprehensive some or to eventual loss of control eating and binge
review, the next sections highlight some examples eating in others. There may be important individual

26 Research Domain Criteria

differences in the influence and importance of sus- had even more diminished ventral striatum response
tained threat in the etiology of disordered eating, to anticipatory reward processing to a monetary
and these differences could influence treatment reward (Balodis et al., 2013, 2014). Neuroimaging
selection, with a focus on sustained threat in some, studies are not all consistent, however, and there
but not all patients. Further, the domains presented is some evidence of reduced brain response to
in the RDoC matrix do not function in isola- food images (Brooks et  al., 2011) and height-
tion. Thus, an individual’s functioning in another ened response to taste (Oberndorfer et  al., 2013)
domain may impact the particular symptom pre- in bulimia nervosa and reduced brain response to
sentation and development. Thus, it is important taste in anorexia nervosa (Oberndorfer et al., 2013).
to consider functioning across multiple domains Although inconsistent, many of the findings from
to evaluate the influence on symptom presentation neuroimaging studies of anticipatory reward are
and development. congruent with studies of self-​report measures of
reward sensitivity showing greater sensitivity in eat-
Positive Valence Systems and Eating ing disorder patients (Glashouwer, Bloot, Veenstra,
Disorders Franken, & de Jong, 2014; Harrison, O’Brien,
Two constructs within the positive valence sys- et al., 2010; Jappe et al., 2011), although some self-​
tems in RDoC are initial and sustained responsive- report measures in anorexia nervosa show reduced
ness to reward. Research in eating disorders has not reward sensitivity (Harrison, O’Brien, et al., 2010).
clearly separated these two constructs, so this review As in the negative valence domain, inconsistencies
will consider the evidence broadly for responsiveness among findings may be related to grouping indi-
to reward. Reward sensitivity is vital to the drive to viduals based on DSM diagnoses, and measuring
eat (Shin, Zheng, & Berthoud, 2009). At the circuit dysfunction in the positive valence domain may
level, we know that reward sensitivity in the brain is help address heterogeneities within diagnoses to
enhanced when the body is in a caloric deficit, lead- improve treatments. Further, functioning in the
ing people to seek and identify high-​calorie foods other domains may impact the measurement of this
to eat to return to homeostasis (Shin et al., 2009). one. In line with this, one study found that negative
In healthy adults, fasting results in increased brain affect related to brain response in the striatum to
response to images of high-​calorie food images in anticipation of a palatable taste in bulimia nervosa
reward-​related regions of brain including the ventral (Bohon & Stice, 2012), was potentially explained
striatum, amygdala, anterior insula, and orbitofron- by connectivity between the amygdala (implicated
tal cortex (OFC) (Goldstone et al., 2009). Thus, we in negative valence systems) and the putamen and
would expect that patients engaging in food restric- insula (implicated in taste reward).
tion would show enhanced reward response to
high-​calorie foods, as well, when they are engaging Cognitive Systems and Eating Disorders
in food restriction. But a number of studies look- Two constructs within the cognitive systems
ing at patients who recovered from anorexia nervosa domain that appear relevant to eating disorders
and are no longer malnourished have still found are those of visual perception and inhibition/​sup-
greater reward sensitivity and response in brain pression. One symptom of anorexia nervosa is a
regions to high-​calorie food images, receipt of mon- disturbance in body image such that individuals
etary reward, and sometimes to pleasurable tastes believe they are fat despite their low weight. This
(Cowdrey, Park, Harmer, & McCabe, 2011; Frank may reflect a true dysfunction in visual percep-
et al., 2012; Wagner et al., 2007). Patients engaging tion. A  recent review of visual processing in ano-
in binge eating, diagnosed with bulimia nervosa or rexia nervosa found an overattention to detail and
binge eating disorder, tend to show greater response reduced focus on global features compared with
in reward-​related regions of the brain to food images healthy controls (Madsen, Bohon, & Feusner,
and anticipation of taste (Karhunen et  al., 2000; 2013). There is evidence of similar abnormal vis-
Schienle, Schäfer, Hermann, & Vaitl, 2009), but ual processing in bulimia nervosa, as well, with per-
diminished brain response to a sweet taste (Bohon formance on neuropsychological measures showing
& Stice, 2011; Frank, Reynolds, Shott, & O’Reilly, weak central coherence, the ability to view things
2011; Frank et  al., 2006). There is also some evi- globally (Lopez, Tchanturia, Stahl, & Treasure,
dence of diminished brain response to monetary 2008a, 2008b, 2009), although the deficit appears
rewards in binge eating disorder and that individu- to be more consistently found in anorexia nervosa
als who continue to binge eat following treatment (Roberts, Tchanturia, & Treasure, 2013). If we

Bohon 27
consider this visual processing difference and relate diagnosis, found that patients engaging in binge
it to perceptual errors of one’s own body, then it eating and purging activated brain regions involved
makes sense that an overattention to detail or more in inhibitory control to a greater extent during a
local visual processing could highlight small “flaws” go/​no-​go task compared with both controls and
in one’s appearance, resulting in disturbed body restricting type anorexia nervosa (Lock, Garrett,
image. Perhaps treatment focused on addressing Beenhakker, & Reiss, 2011). There were no differ-
and/​or acknowledging these visual disturbances can ences in performance on the task, however, which
improve body image and in turn reduce eating dis- again raises the concern of interpreting neural find-
order symptoms. ings to suggest either greater efficiency or disengage-
Despite colloquial beliefs that individuals with ment. Indeed, another study of cognitive control in
restricting-​type anorexia nervosa are more inhib- adolescents with bulimia nervosa found no differ-
ited and those with binge/​purge-​type anorexia ner- ence in performance on a task, but reduced activa-
vosa, bulimia nervosa, or binge eating disorder are tion of frontostriatal regions in the bulimia nervosa
more impulsive, data do not consistently reflect this group compared with controls (Marsh et al., 2011).
pattern (Oberndorfer, Kaye, Simmons, Strigo, & Currently RDoC does not address conflicting data
Matthews, 2011; Wu, Hartmann, Skunde, Herzog, from various units of analysis within a domain, as
& Friederich, 2013). By self-​report, patients with the goal seems to be consensus across these mea-
restricting-​type anorexia nervosa were less impul- sures. Further confusion arises in the application
sive than healthy controls and those with binge/​ of RDoC constructs, such as inhibition, when
purge-​type anorexia nervosa and bulimia nervosa, context affects the measurement of the construct.
but this was not supported by performance on a A recent meta-​analysis found that type of stimulus
stop signal task where participants had to inhibit a impacted effect sizes in studies of inhibitory control
motor response (Claes, Nederkoorn, Vandereycken, in bulimic-​type eating disorders (Wu, Hartmann,
Guerrieri, & Vertommen, 2006). Another study et al., 2013), such that smaller effect sizes were pres-
found no behavioral difference in performance on ent for general stimuli than disease salient stimuli,
a stop signal task, but did find less medial prefron- like food or body images. When the brain is already
tal cortex activation during hard trials in patients taxed because of the arousing and/​ or emotional
recovered from anorexia nervosa compared with content of a stimulus, cognitive systems may not
controls (Oberndorfer et  al., 2011). It is unclear function as well.
whether reduced activation of a brain region indi-
cates more efficient processing or a disengagement Social Processes and Eating Disorders
of neurocircuitry, and thus without a corresponding A recent meta-​analysis reviewed studies evaluat-
difference in performance on the task, it is difficult ing various constructs within the RDoC domain of
to interpret the findings. A  study of patients with social processes (Caglar-​Nazali et  al., 2014). The
bulimia nervosa and binge eating disorder found largest effects in patients with eating disorders were
greater reaction times during a stop signal task found for impaired facial communication, negative
among patients with bulimia nervosa compared self-​evaluation, poor understanding of mental states,
with controls but no difference between patients and sensitivity to social dominance. Specifically,
with binge eating disorder and controls (Wu, Giel, there is more evidence of impaired facial com-
et al., 2013). It is important to note that the study munication in anorexia nervosa, with fewer facial
used separate samples of healthy controls that were expressions produced by patients than controls,
BMI-​matched to each patient group. Because those but also more errors in recognizing facial expres-
with binge eating disorder had greater body mass, sions in others (Caglar-​Nazali et al., 2014). Studies
the matched controls may have struggled with gen- of the understanding of mental states tend to use
eral (nonbinge) overeating, which is also associated tasks such as Reading the Mind in the Eyes, and
with difficulties in inhibitory control (Nederkoorn, poorer understanding of others is more consistently
Smulders, Havermans, Roefs, & Jansen, 2006). poor in anorexia nervosa. Negative self-​evaluation
These studies continue to use diagnostic categories and a sense of social inferiority were more pervasive
and between-​group comparisons to evaluate func- across eating disorder diagnoses, but were usually
tion in inhibition and other RDoC domains/​con- measured via self-​report questionnaires of submis-
structs, which does not allow for investigation of sive behavior or shame (Caglar-​Nazali et al., 2014).
heterogeneity within diagnoses. One study looking Neuroimaging studies have also found impairments
at binge eating and purging behaviors, rather than in theory of mind and social cognition (McAdams

28 Research Domain Criteria

& Krawczyk, 2011; Schulte-​Rüther, Mainz, Fink, disorder category of the DSM-​5, is likely impacted
Herpertz-​Dahlmann, & Konrad, 2012) and self-​ by dysfunction in these constructs. Indeed, patients
evaluation (McAdams & Krawczyk, 2012) in with night eating syndrome have more awakenings
patients with or recovered from anorexia nervosa. during the night than healthy controls (O’Reardon
Patients with bulimia nervosa also showed neural et al., 2004), and show differences in the timing and
differences in social cognition and self-​knowledge, amplitude of circadian markers involved in the regu-
but to a lesser extent than patients with anorexia lation of eating (Goel et al., 2009).
nervosa (McAdams & Krawczyk, 2013). It is
unclear how current emotional state might impact Challenges of the Research Domain
these social processes. Criteria Approach
Despite the promise for developing individual-
Arousal and Regulatory Systems and ized treatments and better conceptualizing the eti-
Eating Disorders ology of psychopathology, RDoC has a number of
Arousal systems are not well defined within the limitations. First, the matrix created for the proj-
RDoC matrix, but are assessed with measures of ect consists of a specified set of domains. Although
emotional reactivity, sensory reactivity, and startle this is a working list and is not comprehensive,
eye-​blink tasks. There is some evidence of a discord- as it currently stands, researchers are encouraged
ance between physiological arousal and self-​reported to work within the presented matrix. A  group of
emotional arousal in adolescents with anorexia experts created this set of domains based on exist-
nervosa, with higher levels of anxiety in response ing literature, but there may be other domains or a
to stress, but no difference in heart rate or cortisol better distinction between domains not present in
compared with controls (Zonnevylle-​Bender et  al., the matrix. Thus, these domains may not represent
2005). Thus, arousal is another area where differ- the true set of functions that would be important to
ent measures and units of analysis may not agree. address to make advances in psychopathology. For
Another study noted that a desire to binge follow- example, emotion regulation may impact multiple
ing a stressor was not accounted for by differences domains, but may not function differently within
in physiologic reactivity (Cattanach, Malley, & each domain. Individuals may have a deficit in emo-
Rodin, 1988). Women with bulimia nervosa show tion regulation that impacts functioning in multi-
blunted systolic blood pressure, heart rate, and epi- ple domains, and addressing the regulatory deficit
nephrine response to stress, but elevated cortisol and would be the true underlying concern. Note that
greater self-​reported distress (Koo-​Loeb, Pedersen, & the arousal and regulatory systems domain in the
Girdler, 1998). There is some evidence of enhanced current RDoC matrix does not include affective or
cortisol awakening response (the increase in cortisol emotion regulation, instead focusing on arousal,
after awakening) in anorexia nervosa, which reflects circadian rhythms, and sleep-​wakefulness.
dysfunction in arousal (Monteleone et  al., 2011; Another challenge in using RDoC to better
Oskis, Loveday, Hucklebridge, Thorn, & Clow, understand psychopathology, including eating dis-
2012), but not in bulimia nervosa (Monteleone, orders, is in the complexity of interactions between
Scognamiglio, Monteleone, Perillo, & Maj, 2014). domains and with environment and development.
Interestingly, sensitivity to reward was positively cor- Although Cuthbert (2014) notes that the working
related with cortisol awakening response in patients group intended for environmental and developmen-
engaging in binge eating and purging, but negatively tal factors to be implicated at each domain and unit
in healthy controls, suggesting differential relations of analysis, they are not explicitly recognized. For a
between the arousal domain of function and positive field such as eating disorders, it is surprising to have
valence systems in individuals (Monteleone et  al., a set of domains intended to conceptualize a patient
2014). Given the inconsistent findings and results presentation that does not include these factors,
from various measures, some domains, such as arou- given what we know about the age of onset of many
sal, may not be as useful. eating disorder symptoms and the influence of envi-
It may be that another construct within this ronment and experience. Even researchers outside
domain, such as circadian rhythms or sleep-​ of the eating disorder field find this approach to lack
wakefulness, may be more relevant. Although there is cohesion (Srivastava, 2013).
little data in the more commonly known eating dis- Because of the potential interactions between
order diagnoses, night eating syndrome, one of the domains, it is unlikely that deficits would need to
diagnoses in the other specified feeding and eating be addressed individually. In other words, rather

Bohon 29
than requiring a treatment module for each area of Further, their sensitivity to the reward value of food,
dysfunction, a primary area may be identified for but not other rewards, seems to be enhanced, par-
a patient that would impact other dysfunctions. ticularly during negative mood states. The patient
For example, if reward sensitivity in some patients is also surrounded by pressure to be thin and was
who binge eat is enhanced by negative affect, then teased for being overweight as a child. This seemed
addressing the negative affect may normalize reward to place an emphasis and increased focus on shape
sensitivity in those patients. and weight, as well as on “healthy eating.” When
her mood is good, however, she is able to manage
Conclusions and Alternatives her thoughts and feelings about her body image and
Existing research on RDoC domains in eat- eats an appropriate balance of nutrients without
ing disorders has tended to examine areas of dys- engaging in binge eating. This difference in behavior
function within diagnostic groups, which relates during different mood states, as well as the differen-
dysfunction to clusters of symptoms. In order to tial reward sensitivity, may highlight the importance
connect dysfunction in these domains to specific of improving emotion regulation and resolving the
symptoms or maladaptive behaviors, research must dysfunction in the negative valence systems first,
cut across diagnoses. A small number of studies have and then moving to lingering concerns about body
done this with binge eating and purging behaviors image and overvaluation of shape and weight.
in both anorexia nervosa and bulimia nervosa (e.g. Further, it may highlight a need for addressing cog-
Lock et al., 2011), but more research is needed to nitive control after inducing a negative mood, so
better understand how eating disorder symptoms that the patient has practice in the skill in the con-
and behaviors relate to underlying dysfunction. text in which she tends struggle. Contrast this with
Despite this gap in knowledge, there is some evi- a patient whose binge eating is not related to mood,
dence suggesting that disruptions or abnormalities and instead seems to be more connected to reward
in RDoC domains are present in patients with eat- sensitivity and loss of control around food cues, as
ing disorders and may be worth investigating fur- well as further enhanced reward sensitivity when in
ther to determine whether change in these domains a food restricted state. This patient may not benefit
of function results in symptom change. Most of from emotion regulation, but instead in improving
our knowledge to date is based on cross-​sectional cognitive control and developing skills in regulat-
data of patients currently ill or recovered, so it is ing positive valence systems (reducing reward sensa-
unclear at this point whether targeting these areas tion) or adapting the environment to accommodate
would impact symptom presentation. Importantly, this reward sensitivity. In other words, if the patient
evidence of interactions between domains of func- cannot reduce the reward sensitivity, reducing the
tion and regulatory processes that affect multiple likelihood of being exposed to certain food cues that
domains will be important to take into account as will trigger binge eating, as well as improving cogni-
this research progresses. tive control, may help her refrain from binge eating.
It may be that future psychiatric evaluations will This individualized approach to patient care, how-
focus less on diagnoses, but instead provide more ever, requires extensive research to create effective
descriptive accounts of an individual’s functioning intervention modules and understand how to create
and history, including the presence of maladaptive these personalized approaches.
behaviors. Treatments would be designed modularly
to address the particular constellation of dysfunc- Future Directions
tion and environmental/​ developmental factors—​ Despite the challenges noted, there is likely great
highlighting primary areas that may require initial benefit in thinking outside of diagnostic categories
intervention. Continued evaluation throughout to approach treatment and meet patients’ needs.
treatment will be important since interactions RDoC-​type research studies have already begun,
between these areas may result in improvement of primarily focused on existing RDoC domains and
subsequent areas without requiring treatment mod- eating disorder symptoms, rather than diagnoses.
ules to directly address them. For example, a patient Tanofsky-​ Kraff and colleagues (Tanofsky-​ Kraff,
may present with complaints about binge eating Engel, Yanovski, Pine, & Nelson, 2013) have pro-
and purging. Through evaluation and testing, it is posed a research design to examine acute threat
clear that their binge eating is driven primarily by (negative valence system) to predict loss of control
loss of control (deficit in cognitive control) during eating in children, which could have implications on
periods of negative affect (negative valence systems). the development of binge eating. This design keeps

30 Research Domain Criteria

its focus away from diagnoses and instead on behav- Journal of Eating Disorders, 44, 585–​595. https://​doi.org/​
ior and the functional domain. Future research will 10.1002/​eat.20869
Bohon, C., & Stice, E. (2012). Negative affect and neural
have to determine how to manage incongruence response to palatable food intake in bulimia nervosa. Appetite,
between units of analysis, as it is unclear how tightly 58, 964–​970. https://​doi.org/​10.1016/​j.appet.2012.02.051
genes, circuits, and behaviors are connected within Brooks, S. J., O’Daly, O. G., Uher, R., Friederich, H.-​ C.,
domains of function. Giampietro, V., Brammer, M.,  . . .  Campbell, I. C. (2011).
As previously described, another approach that Differential neural responses to food images in women
with bulimia versus anorexia nervosa. PloS One, 6, e22259.
may provide greater clinical impact would be the https://​doi.org/​10.1371/​journal.pone.0022259
examination of treatment targets across diagnoses Caglar-​
Nazali, H. P., Corfield, F., Cardi, V., Ambwani, S.,
with the goal of developing modular treatment pro- Leppanen, J., Olabintan, O.,   .  .  .  Treasure, J. (2014).
tocols that capitalize on discoveries in other fields A systematic review and meta-​ analysis of “systems for
and are individualized to the patient to improve social processes” in eating disorders. Neuroscience and
Biobehavioral Reviews, 42, 55–​92. https://​doi.org/​10.1016/​
treatment outcome. This is complex, as it relies on j.neubiorev.2013.12.002
the interactions between many neural and behav- Cattanach, L., Malley, R., & Rodin, J. (1988). Psychologic and
ioral processes. If there are superordinate processes, physiologic reactivity to stressors in eating disordered indi-
like emotion regulation, however, which impact the viduals. Psychosomatic Medicine, 50, 591–​599.
functioning of many other domains and behaviors, Chorpita, B. F., Taylor, A. A., Francis, S. E., Moffitt, C., &
Austin, A. A. (2004). Efficacy of modular cognitive behavior
this may be a fruitful area of study. Indeed, due to therapy for childhood anxiety disorders. Behavior Therapy,
the length of time required to move from research 35, 263–​287. https://​doi.org/​10.1016/​S0005-​789480039-​X
to practice, it is important that our focus not linger Claes, L., Nederkoorn, C., Vandereycken, W., Guerrieri, R., &
entirely on possible underlying mechanisms, but Vertommen, H. (2006). Impulsiveness and lack of inhibitory
also on the treatments with the most potential for control in eating disorders. Eating Behaviors, 7, 196–​203.
https://​doi.org/​10.1016/​j.eatbeh.2006.05.001
impact now. Although RDoC provides encourage- Cowdrey, F. A., Park, R. J., Harmer, C. J., & McCabe, C. (2011).
ment to think outside of the (diagnostic) box, those Increased neural processing of rewarding and aversive food
suffering from eating disorders cannot wait for the stimuli in recovered anorexia nervosa. Biological Psychiatry, 70,
completion of basic biological research; there must 736–​743. https://​doi.org/​10.1016/​j.biopsych.2011.05.028
be simultaneous research on treatments to alleviate Cuthbert, B. N. (2014). The RDoC framework:  Facilitating
transition from ICD/​ DSM to dimensional approaches
known symptoms (Lewis-​Fernández et  al., 2016). that integrate neuroscience and psychopathology. World
Additionally, the predefined matrix set up by the Psychiatry, 13, 28–​35. https://​doi.org/​10.1002/​wps.20087
RDoC project provides another set of constraints, Foerde, K., Steinglass, J. E., Shohamy, D., & Walsh, B. T. (2015).
which may be just as limiting as diagnoses. If we Neural mechanisms supporting maladaptive food choices
instead think broadly about individuals as having in anorexia nervosa. Nature Neuroscience, 18, 1571–​1573.
https://​doi.org/​10.1038/​nn.4136
unique sets of developmental, environmental, and Frank, G. K. W., Reynolds, J. R., Shott, M. E., Jappe, L., Yang,
biological backgrounds, we may be able to better T. T., Tregellas, J. R., & O’Reilly, R. C. (2012). Anorexia
tailor our interventions to meet their needs. nervosa and obesity are associated with opposite brain reward
response. Neuropsychopharmacology, 37, 2031–​2046. https://​
doi.org/​10.1038/​npp.2012.51
References Frank, G. K. W., Reynolds, J. R., Shott, M. E., & O’Reilly, R.
American Psychiatric Association. (2013). Diagnostic and C. (2011). Altered temporal difference learning in bulimia
Statistical Manual of Mental Disorders (5th ed.). Washington, nervosa. Biological Psychiatry, 70, 728–​735. https://​doi.org/​
DC: Author. 10.1016/​j.biopsych.2011.05.011
Balodis, I. M., Grilo, C. M., Kober, H., Worhunsky, P. D., Frank, G. K., Wagner, A., Achenbach, S., McConaha, C.,
White, M. A., Stevens, M. C.,  . . .  Potenza, M. N. (2014). Skovira, K., Aizenstein, H.,   .  .  .  Kaye, W. H. (2006).
A pilot study linking reduced fronto-​Striatal recruitment Altered brain activity in women recovered from bulimic-​
during reward processing to persistent bingeing following type eating disorders after a glucose challenge: A pilot study.
treatment for binge-​ eating disorder. International Journal International Journal of Eating Disorders, 39, 76–​79. https://​
of Eating Disorders, 47, 376–​384. https://​doi.org/​10.1002/​ doi.org/​10.1002/​eat.20210
eat.22204 Glashouwer, K. A., Bloot, L., Veenstra, E. M., Franken, I. H. A.,
Balodis, I. M., Kober, H., Worhunsky, P. D., White, M. A., & de Jong, P. J. (2014). Heightened sensitivity to punish-
Stevens, M. C., Pearlson, G. D.,  . . .  Potenza, M. N. (2013). ment and reward in anorexia nervosa. Appetite, 75, 97–​102.
Monetary reward processing in obese individuals with and https://​doi.org/​10.1016/​j.appet.2013.12.019
without binge eating disorder. Biological Psychiatry, 73, 877–​ Goel, N., Stunkard, A. J., Rogers, N. L., Dongen, H. P. A. V.,
886. https://​doi.org/​10.1016/​j.biopsych.2013.01.014 Allison, K. C., O’Reardon, J. P.,  . . .  Dinges, D. F. (2009).
Bohon, C., & Stice, E. (2011). Reward abnormalities among Circadian rhythm profiles in women with night eating syn-
women with full and subthreshold bulimia nervosa: A func- drome. Journal of Biological Rhythms, 24, 85–​94. https://​doi.
tional magnetic resonance imaging study. International org/​10.1177/​0748730408328914

Bohon 31
Goldstone, A. P., Prechtl de Hernandez, C. G., Beaver, J. D., Lopez, C. A., Tchanturia, K., Stahl, D., & Treasure, J. (2008a).
Muhammed, K., Croese, C., Bell, G.,  . . .  Bell, J. D. (2009). Central coherence in eating disorders:  A  systematic review.
Fasting biases brain reward systems towards high-​ calorie Psychological Medicine, 38, 1393–​1404. https://​doi.org/​
foods. European Journal of Neuroscience, 30, 1625–​1635. 10.1017/​S0033291708003486
https://​doi.org/​10.1111/​j.1460-​9568.2009.06949.x Lopez, C. A., Tchanturia, K., Stahl, D., & Treasure, J. (2008b).
Harrison, A., O’Brien, N., Lopez, C., & Treasure, J. (2010). Central coherence in women with bulimia nervosa.
Sensitivity to reward and punishment in eating disor- International Journal of Eating Disorders, 41, 340–​347.
ders. Psychiatry Research, 177(1–​2), 1–​11. https://​doi.org/​ https://​doi.org/​10.1002/​eat.20511
10.1016/​j.psychres.2009.06.010 Lopez, C. A., Tchanturia, K., Stahl, D., & Treasure, J. (2009).
Harrison, A., Tchanturia, K., & Treasure, J. (2010). Attentional Weak central coherence in eating disorders: A step towards
bias, emotion recognition, and emotion regulation in looking for an endophenotype of eating disorders. Journal
anorexia:  State or trait? Biological Psychiatry, 68, 755–​761. of Clinical and Experimental Neuropsychology, 31, 117–​125.
https://​doi.org/​10.1016/​j.biopsych.2010.04.037 https://​doi.org/​10.1080/​13803390802036092
Hay, P. (2013). A systematic review of evidence for psychologi- Madsen, S. K., Bohon, C., & Feusner, J. D. (2013). Visual
cal treatments in eating disorders: 2005–​2012. International processing in anorexia nervosa and body dysmorphic disor-
Journal of Eating Disorders, 46, 462–​469. https://​doi.org/​ der:  similarities, differences, and future research directions.
10.1002/​eat.22103 Journal of Psychiatric Research, 47, 1483–​1491. https://​doi.
Henje Blom, E., Duncan, L. G., Ho, T. C., Connolly, C. G., org/​10.1016/​j.jpsychires.2013.06.003
LeWinn, K. Z., Chesney, M.,  . . .  Yang, T. T. (2014). The Marsh, R., Horga, G., Wang, Z., Wang, P., Klahr, K. W., Berner,
development of an RDoC-​based treatment program for ado- L. A.,  . . .  Peterson, B. S. (2011). An fMRI study of self-​
lescent depression: “Training for Awareness, Resilience, and regulatory control and conflict resolution in adolescents with
Action” (TARA). Frontiers in Human Neuroscience, 8, 630. bulimia nervosa. American Journal of Psychiatry, 168, 1210–​
https://​doi.org/​10.3389/​fnhum.2014.00630 1220. https://​doi.org/​10.1176/​appi.ajp.2011.11010094
Insel, T. R., & Cuthbert, B. N. (2009). Endophenotypes: Bridging McAdams, C. J., & Krawczyk, D. C. (2011). Impaired neu-
genomic complexity and disorder heterogeneity. Biological ral processing of social attribution in anorexia nervosa.
Psychiatry, 66, 988–​989. https://​doi.org/​10.1016/​ Psychiatry Research:  Neuroimaging, 194, 54–​63. https://​doi.
j.biopsych.2009.10.008 org/​10.1016/​j.pscychresns.2011.06.016
Jappe, L. M., Frank, G. K. W., Shott, M. E., Rollin, M. D. H., McAdams, C. J., & Krawczyk, D. C. (2012). Who am I? How do
Pryor, T., Hagman, J. O.,  . . .  Davis, E. (2011). Heightened I look? Neural differences in self-​identity in anorexia nervosa.
sensitivity to reward and punishment in anorexia nervosa. Social Cognitive and Affective Neuroscience, nss093. https://​
International Journal of Eating Disorders, 44, 317–​324. doi.org/​10.1093/​scan/​nss093
https://​doi.org/​10.1002/​eat.20815 McAdams, C. J., & Krawczyk, D. C. (2013). Neural responses
Joos, A. A.  B., Saum, B., van Elst, L. T., Perlov, E., Glauche, during social and self-​knowledge tasks in bulimia nervosa.
V., Hartmann, A.,   .  .  .  Zeeck, A. (2011). Amygdala Frontiers in Psychiatry, 4, 103. https://​doi.org/​10.3389/​
hyperreactivity in restrictive anorexia nervosa. Psychiatry fpsyt.2013.00103
Research:  Neuroimaging, 191, 189–​195. https://​doi.org/​ Monteleone, P., Scognamiglio, P., Monteleone, A. M., Mastromo,
10.1016/​j.pscychresns.2010.11.008 D., Jr, Steardo, L., Serino, I., & Maj, M. (2011). Abnormal
Karhunen, L. J., Vanninen, E. J., Kuikka, J. T., Lappalainen, diurnal patterns of salivary α-​amylase and cortisol secretion
R. I., Tiihonen, J., & Uusitupa, M. I.  J. (2000). Regional in acute patients with anorexia nervosa. World Journal of
cerebral blood flow during exposure to food in obese binge Biological Psychiatry, 12, 455–​461. https://​doi.org/​10.3109/​
eating women. Psychiatry Research: Neuroimaging, 99, 29–​42. 15622975.2011.590226
https://​doi.org/​10.1016/​S0925-​492700053-​6 Monteleone, P., Scognamiglio, P., Monteleone, A. M., Perillo, D.,
Keel, P. K., Brown, T. A., Holland, L. A., & Bodell, L. P. (2012). & Maj, M. (2014). Cortisol awakening response in patients
Empirical classification of eating disorders. Annual Review with anorexia nervosa or bulimia nervosa:  Relationships
of Clinical Psychology, 8, 381–​404. https://​doi.org/​10.1146/​ to sensitivity to reward and sensitivity to punishment.
annurev-​clinpsy-​032511-​143111 Psychological Medicine, 44, 2653–​2660. https://​doi.org/​
Koo-​Loeb, J. H., Pedersen, C., & Girdler, S. S. (1998). Blunted 10.1017/​S0033291714000270
cardiovascular and catecholamine stress reactivity in women Nederkoorn, C., Smulders, F. T. Y., Havermans, R. C., Roefs, A.,
with bulimia nervosa. Psychiatry Research, 80, 13–​27. & Jansen, A. (2006). Impulsivity in obese women. Appetite,
Lewis-​Fernández, R., Rotheram-​ Borus, M. J., Betts, V. T., 47, 253–​256. https://​doi.org/​10.1016/​j.appet.2006.05.008
Greenman, L., Essock, S. M., Escobar, J. I.,   .  .  .  Iversen, Oberndorfer, T. A., Frank, G. K. W., Simmons, A. N., Wagner,
P. (2016). Rethinking funding priorities in mental health A., McCurdy, D., Fudge, J. L.,  . . .  Kaye, W. H. (2013).
research. British Journal of Psychiatry, 208, 507–​509. https://​ Altered insula response to sweet taste processing after recov-
doi.org/​10.1192/​bjp.bp.115.179895 ery from anorexia and bulimia nervosa. American Journal of
Lilenfeld, L. R.  R., Wonderlich, S., Riso, L. P., Crosby, R., & Psychiatry, 170, 1143–​1151. https://​doi.org/​10.1176/​appi.
Mitchell, J. (2006). Eating disorders and personality: A meth- ajp.2013.11111745
odological and empirical review. Clinical Psychology Review, Oberndorfer, T. A., Kaye, W. H., Simmons, A. N., Strigo, I. A.,
26, 299–​320. https://​doi.org/​10.1016/​j.cpr.2005.10.003 & Matthews, S. C. (2011). Demand-​specific alteration of
Lock, J., Garrett, A., Beenhakker, J., & Reiss, A. L. (2011). medial prefrontal cortex response during an inhibition task
Aberrant brain activation during a response inhibition task in recovered anorexic women. International Journal of Eating
in adolescent eating disorder subtypes. American Journal Disorders, 44, 1–​8. https://​doi.org/​10.1002/​eat.20750
of Psychiatry, 168, 55–​64. https://​doi.org/​10.1176/​appi. O’Reardon, J. P., Ringel, B. L., Dinges, D. F., Allison, K. C.,
ajp.2010.10010056 Rogers, N. L., Martino, N. S., & Stunkard, A. J. (2004).

32 Research Domain Criteria

 Circadian eating and sleeping patterns in the night eating nervosa: preliminary findings. Annals of General Psychiatry, 6,
syndrome. Obesity Research, 12, 1789–​1796. https://​doi.org/​ 14. https://​doi.org/​10.1186/​1744-​859X-​6-​14
10.1038/​oby.2004.222 Wagner, A., Aizenstein, H., Venkatraman, V. K., Fudge, J., May,
Oskis, A., Loveday, C., Hucklebridge, F., Thorn, L., & Clow, A. J. C., Mazurkewicz, L.,  . . .  Kaye, W. H. (2007). Altered
(2012). Diurnal patterns of salivary cortisol and DHEA in reward processing in women recovered from anorexia ner-
adolescent anorexia nervosa. Stress, 15, 601–​607. https://​doi. vosa. American Journal of Psychiatry, 164, 1842–​1849.
org/​10.3109/​10253890.2012.661493 https://​doi.org/​10.1176/​appi.ajp.2007.07040575
Roberts, M. E., Tchanturia, K., & Treasure, J. L. (2013). Is atten- Weisz, J. R., Chorpita, B. F., Palinkas, L. A., Schoenwald, S.
tion to detail a similarly strong candidate endophenotype K., Miranda, J., Bearman, S. K.,  . . .  Research Network on
for anorexia nervosa and bulimia nervosa? World Journal of Youth Mental Health. (2012). Testing standard and modular
Biological Psychiatry, 14, 452–​463. https://​doi.org/​10.3109/​ designs for psychotherapy treating depression, anxiety, and
15622975.2011.639804 conduct problems in youth: A randomized effectiveness trial.
Schienle, A., Schäfer, A., Hermann, A., & Vaitl, D. (2009). Archives of General Psychiatry, 69, 274–​282. https://​doi.org/​
Binge-​eating disorder:  Reward sensitivity and brain activa- 10.1001/​archgenpsychiatry.2011.147
tion to images of food. Biological Psychiatry, 65, 654–​661. Wildes, J. E., & Marcus, M. D. (2015). Application of the
https://​doi.org/​10.1016/​j.biopsych.2008.09.028 research domain criteria (RDoC) framework to eat-
Schulte-​Rüther, M., Mainz, V., Fink, G. R., Herpertz-​ ing disorders:  Emerging concepts and research. Current
Dahlmann, B., & Konrad, K. (2012). Theory of mind Psychiatry Reports, 17, 30. https://​doi.org/​10.1007/​
and the brain in anorexia nervosa:  Relation to treatment s11920-​015-​0572-​2
outcome. Journal of the American Academy of Child and Wu, M., Giel, K. E., Skunde, M., Schag, K., Rudofsky, G.,
Adolescent Psychiatry, 51, 832–​841.e11. https://​doi.org/​ de Zwaan, M.,  . . .  Friederich, H.-​C. (2013). Inhibitory
10.1016/​j.jaac.2012.06.007 control and decision making under risk in bulimia ner-
Shin, A. C., Zheng, H., & Berthoud, H.-​ R. (2009). An vosa and binge-​eating disorder. International Journal of
expanded view of energy homeostasis: Neural integration of Eating Disorders, 46, 721–​728. https://​doi.org/​10.1002/​
metabolic, cognitive, and emotional drives to eat. Physiology eat.22143
and Behavior, 97, 572–​580. https://​doi.org/​10.1016/​ Wu, M., Hartmann, M., Skunde, M., Herzog, W., &
j.physbeh.2009.02.010 Friederich, H.-​C. (2013). Inhibitory control in bulimic-​
Srivastava, S. (2013, May 24). Where is RDoC headed? type eating disorders:  A  systematic review and meta-​
A  look at the eating disorders FOA. Retrieved from analysis. PLOS ONE, 8, e83412. https://​doi.org/​10.1371/​
h t t p s : / / ​ h a r d s c i . w o r d p r e s s . c o m / ​ 2 0 1 3 / ​ 0 5 / ​ 2 4 /​ journal.pone.0083412
where-​is-​rdoc-​headed-​a-​look-​at-​the-​eating-​disorders-​foa/​ Wykes, T., & van der Gaag, M. (2001). Is it time to develop a
Strober, M. (2014). Proposition:  Family based treat- new cognitive therapy for psychosis—​cognitive remediation
ment is overvalued Position:  proposer. Advances in therapy (CRT)? Clinical Psychology Review, 21, 1227–​1256.
Eating Disorders, 2, 264–​284. https://​doi.org/​10.1080/​ https://​doi.org/​10.1016/​S0272-​735800104-​0
21662630.2014.898390 Zonnevylle-​Bender, M. J.  S., van Goozen, S. H.  M., Cohen-​
Tanofsky-​Kraff, M., Engel, S., Yanovski, J. A., Pine, D. S., & Kettenis, P. T., Jansen, L. M. C., van Elburg, A., & Engeland,
Nelson, E. E. (2013). Pediatric disinhibited eating: Toward H.  van. (2005). Adolescent anorexia nervosa patients have
a research domain criteria framework. International Journal a discrepancy between neurophysiological responses and
of Eating Disorders, 46, 451–​455. https://​doi.org/​10.1002/​ self-​
reported emotional arousal to psychosocial stress.
eat.22101 Psychiatry Research, 135, 45–​52. https://​doi.org/​10.1016/​
Tchanturia, K., Davies, H., & Campbell, I. C. (2007). j.psychres.2004.11.006
Cognitive remediation therapy for patients with anorexia

Bohon 33
 C H A PT E R


Epidemiology and Course of
3 Eating Disorders

Pamela K. Keel

Abstract
The epidemiology of eating disorders holds important clues for understanding factors that may
contribute to their etiology. In addition, epidemiological findings speak to the public health significance
of these deleterious syndromes. Information on course and outcome is important for clinicians to
understand the prognosis associated with different disorders of eating and for treatment planning. This
chapter reviews information on the epidemiology and course of anorexia nervosa, bulimia nervosa, and
binge eating disorder. In addition new information is reviewed for two forms of other specified eating or
feeding disorder, purging disorder and night eating syndrome, which were introduced in the Diagnostic
and Statistical Manual for Mental Disorders, fifth edition.
Key Words:  anorexia nervosa, binge eating disorder, bulimia nervosa, course, epidemiology, mortality,
purging disorder, night eating syndrome

Introduction medical literature, considerable information has

Eating disorders represent a significant source been amassed on the epidemiology and course of
of psychiatric morbidity among late adolescent this syndrome.
and young adult women. Basic to the understand-
ing of any mental disorder is careful description of Epidemiology
who is affected by the disorder and disorder course Although AN is the first condition that comes
and outcome. This chapter reviews information on to mind when most people think of eating disor-
the epidemiology and course of the eating disor- ders and represents a high proportion of patients
ders included in the category of eating and feeding encountered in inpatient settings (Dalle Grave &
disorders:  anorexia nervosa (AN), bulimia ner- Calugi, 2007), AN may be less common than the
vosa (BN), binge eating disorder (BED), and two other eating disorders (i.e., BN, BED, and PD) in
forms of other specified feeding or eating disorder the general population. According to the Diagnostic
(OSFED), and purging disorder (PD) and night and Statistical Manual of Mental Disorders (5th ed.;
eating syndrome (NES). DSM-​5; American Psychiatric Association [APA],
2013), the 12-​month prevalence of AN is 0.4% in
Anorexia Nervosa women and approximately one tenth (or 0.04%)
The term “anorexia nervosa” was first introduced that in men. This means that approximately 1 in
into the psychiatric nomenclature in 1874 by Sir 200 women and 1 in 2,000 men will experience
William Gull to describe four adolescent female AN during a 12-​month time frame. Of note, in a
patients who experienced significant weight loss population of 300 million with balanced gender
that could not be related to any medical condition representation, this does not mean that 600,000
(Gull, 1874). Given the long history of AN in the women (150,000,000 × 0.004) and 60,000 men

34
(150,000,000 × 0.0004) have AN, because not all (2003) reported that the increase was evident until
individuals are within the period of risk for the ill- the 1970s, and then incidence rates demonstrated
ness. Age at onset for AN is typically early to late relative stability over time. Two epidemiologi-
adolescence, and, as discussed in the text that fol- cal studies support the stability of AN incidence
lows, a proportion of patients with AN eventually rates from 1994 to 2000 in the United Kingdom
achieves recovery such that prevalence is lower in (Currin, Schmidt, Treasure, & Jick, 2005) and from
older age groups. Thus, the point prevalence, or per- 1985–​1989 to 1995–​1999 to 2005–​2009 in the
centage of individuals who currently meet criteria Netherlands (van Son, van Hoeken, Bartelds, van
for AN, is much lower than 12-​month prevalence Furth, & Hoek, 2006; Smink, van Hoeken, Donker,
and varies by age group and population studied. For Susser, Oldehinkel, & Hoek, 2016). However, evi-
example, Favaro, Ferrara, and Santanastaso (2003) dence suggests increasing rates in the highest risk
reported a lifetime prevalence of AN of 2.0% and a group of females ages 15 to 19 years, with incidence
point prevalence of 0.3% in women in a metropol- nearly doubling from 56.4 to 109.2 per 100,000
itan area of Italy. person years in the Netherlands (van Son et  al.,
Four population-​based epidemiological studies 2006). These results echo findings from the United
have reported lifetime prevalence estimates for AN States (Lucas, Crowson, O’Fallon, & Melton,
in women, representing the percentage of women 1999) in which incidence rates demonstrated a
who have suffered from AN at some point during linear increase from 1935 to 1989 in females aged
their lifetime, including a lifetime prevalence of 15 to 24  years. Javaras et  al. (2015) found signif-
0.9% in the United States (Hudson, Hiripi, Pope, icant increases in AN incidence in females and
& Kessler, 2007), 1.9% in Australia (Wade, Bergin, males from 2001 to 2009 based on the healthcare
Tiggemann, Bulik, & Fairburn, 2006), 2.0% in Italy registry in Sweden as did Steinhausen and Jensen
(Favaro et al., 2003), and 2.2% in Finland (Keski-​ (2015) in Denmark’s nationwide psychiatric reg-
Rahkonen et  al., 2007). Some of the higher esti- istry from 1995 to 2010. Importantly, these rates
mates may reflect detection of cases that are missed would be influenced by improved detection of cases
when ascertainment is based on clinical referral due to greater awareness of eating disorders and
(Keski-​Rahkonen et al., 2007). Indeed, 1-​year prev- their diagnostic features, and trends observed for
alence estimates based on mental healthcare records AN were observed for other psychiatric conditions
are lower than estimates derived from primary care, (Steinhausen & Jensen, 2015). Thus, popular media
which, in turn, are lower than those obtained from portrayals of the “epidemic” of AN have some basis
the community (Hoek, 2006). In addition, results in truth in that rates of the illness have increased
may reflect increasing prevalence of the syndrome over time, and this trend has been particularly true
across cohorts over time. Although the term “rate” for adolescent and young adult females. However,
is often used in conjunction with prevalence esti- sex-​and age-​adjusted incidence rates in the overall
mates, prevalence estimates are typically unsuitable1 population are relatively low, ranging from 1.2 per
for determining time trends (or rates of change) for 100,000 person-​years for severe AN (Milos et  al.,
the presence of eating disorders in a population. 2004) to 12.6 per 100,000 person-​years for more
Prevalence estimates combine new and old cases in broadly defined AN (Steinhausen & Jensen, 2015).
a single value and, thus, reflect both the number of
new cases in a population as well as the chronicity Cross-​Cultural Patterns
of an illness. The studies reviewed in the preceding text
Incidence rates reflect the number of new cases have come from Western cultures because most
(“incident cases”) within a unit of population and research on eating disorders has been conducted
over a unit of time (e.g., cases per 100,000 persons in the United States, Canada, Australia, and west-
per year or per 100,000 person years). As such, these ern European countries. Many have asserted the
rates allow examination of changes over time that possibility that AN may represent a culture-​bound
can address whether or not a condition has become syndrome. However, a review of cross-​ cultural
more common in a population. A  meta-​analysis data revealed evidence of AN throughout the non-​
(Keel & Klump, 2003) and systematic review (Hoek Western world, including cases in individuals with
& van Hoeken, 2003) found significant increases in no evident exposure to Western culture (Keel &
AN incidence during the 20th century. Keel and Klump, 2003). Of note, several cases that emerge
Klump (2003) reported a modest effect size associ- in a non-​Western context and in the absence of
ated with these changes, and Hoek and van Hoeken exposure to Western ideals are characterized by

Keel 35
absence of a fear of gaining weight or becoming fat. ratio (SMR) of approximately 10.0—​or a 10-​fold
However, these features were not described by Gull increase in risk of premature death (Keel et  al.,
(1874) when the syndrome was first introduced 2003; Löwe et al., 2001). Primary causes of death
and may reflect culturally meaningful explanations include starvation and suicide (Nielsen et al., 1998).
for a syndrome for which the true causes remain Keel et al. (2003) reported an SMR due to suicide in
unknown. Although AN has been found in non-​ AN of 56.9 (95% confidence interval [CI] = 15.3–​
Western cultures, the impact of culture on disorder 145.7). A recent meta-​analysis of studies examining
expression is evident in the lower estimates found mortality rates in AN estimated that SMR was 5.86
in these countries. For example, Kolar, Rodriguez, across studies (Arcelus, Mitchell, Wales, & Nielsen,
Chams, and Hoek (2016) reported that the point 2011). One study found that elevated death in AN
prevalence of AN was 0.1% in Latin America coun- was observed during the first 10 years of follow-​up
tries, representing a lower estimate than found in and then decreased (Franko et al., 2013).
other Western countries. Similarly, van Hoeken, One in five patients with AN commits suicide
Burns, and Hoek (2016) found 0% point preva- (Arcelus et al., 2011). A recent meta-​analysis of sui-
lence of AN in countries of Africa. cide in AN provided an SMR of 31 (Preti, Rocchi,
Sisti, Camboni, & Miotto, 2011), meaning that
Course individuals with AN were 31 times more likely to
Anorexia nervosa is associated with a variable die by suicide than were other individuals matched
course and outcome (APA, 2013). In treatment-​ on gender, age, and socioeconomic status. In a large
seeking samples, a minority of patients achieve AN cohort (n  =  432), 17% reported at least one
remission early in the course of illness (i.e., within suicide attempt (Bulik et al., 2008). These findings
1  year) and sustain recovery throughout life, are consistent with results from a review suggesting
whereas many patients struggle with their illness that between 3% and 20% of AN patients endorse
for more than a decade. Across studies examining suicide attempts (Franko & Keel, 2006). The vari-
patients 10 to 20  years after diagnosis, just under ability in proportions of attempts may reflect the
half of patients achieved full recovery, another third composition of samples, given that suicide attempts
remained symptomatic but demonstrated some are more likely to occur in individuals with the
improvement, and 20% remained chronically ill binge-​purge subtype of AN compared to those with
(Steinhausen, 2002). The worst prognosis has been the restricting subtype (Bulik et  al., 2008; Franko
observed in individuals diagnosed at older ages and & Keel, 2006). An examination of methods used
who encounter longer delays between onset of illness by AN patients who completed suicide indicated
and initiation of treatment (Steinhausen, 2002). the use of highly lethal methods that would be fatal
Among patients demonstrating “improvement,” regardless of the attempter’s physical health (Holm-​
some develop binge-​ purge symptoms and expe- Denoma et al., 2008).
rience weight gain resulting in a shift in diagnosis Clinical correlates of suicide attempts included
from AN to BN. Although these individuals may be purging behaviors, depression, substance abuse,
considered “improved” because they no longer meet and history of childhood physical or sexual abuse
full criteria for AN, they are not recovered. Instead, (Franko & Keel, 2006). Prospective predictors of
they have crossed over or migrated from one diag- fatal outcome due to any causes (suicide and non-
nosis to another. At 12-​year follow-​up, Fichter and suicide) included poor psychosocial functioning,
Quadflieg (2007) reported that 9% of patients ini- longer duration of follow-​up, and severity of alcohol
tially diagnosed with AN crossed over to a diagnosis use disorders (Keel et  al., 2003). Thus, comorbid
of BN. Similar issues emerge for those whose illness substance use disorders are associated with increased
changes over time to meet partial criteria for AN. suicide attempts, and severity of alcohol use disorder
Such individuals may be viewed as partially recov- predicts increased risk of premature death in AN.
ered (Herzog et  al., 1999) or they may be viewed Importantly, these course data have come pre-
as having an EDNOS (Fichter & Quadflieg, 2007). dominantly from clinic-​ based samples and may
At 12-​year follow-​up, Fichter and Quadflieg (2007) reflect more dire outcomes than would be observed
reported that 15.7% of patients initially diagnosed for the full population of individuals diagnosed
with AN crossed over to a diagnosis of EDNOS. with AN. Supporting this possibility, course data
Mortality has been observed in approximately 1 from community-​ based epidemiological studies
in 20 patients (Steinhausen, 2002; Sullivan, 1995) suggest that a majority of individuals with lifetime
across studies, reflecting a standardized mortality diagnoses of AN do not currently meet criteria for

36 Epidemiology and Course
the illness (Favaro et al., 2003; Hudson et al., 2007; In contrast to these broad patterns, some recent
Keski-​Rahkonen et al., 2007; Wade et al., 2006). studies have suggested the possibility of declin-
ing rates of BN in the population. Currin et  al.
Bulimia Nervosa (2005) reported an initial increase in BN incidence
The term “bulimia nervosa” was introduced to in women ages 10 to 39  years from less than 25
the medical literature in 1979 by Dr. Gerald Russell per 100,000 person-​years in 1988 to greater than
to describe a series of 30 patients who were binge eat- 50 per 100,000 person-​years in 1996, this was fol-
ing and purging at normal weight (Russell, 1979). lowed by a 38.9% decline in incidence by 2000.
In his seminal article, Russell (1979) described the The decline was largely attributable to changes in
illness as an “ominous variant of anorexia nervosa,” incidence rates for women ages 20 to 39. Incidence
and noted a history of full or partial AN in a number in this group decreased significantly from 56.7 per
of his patients. Given the relatively recent delinea- 100,000 person-​years in 1993 to 28.6 per 100,000
tion of BN from AN, fewer studies have examined person-​years in 2000. In contrast, BN incidence
the epidemiology and course of this syndrome. Yet, rates were comparatively stable in women ages 10
the studies that have emerged suggest that BN dif- to 19 over this period, ranging from 41.0 to 35.8
fers from AN on both epidemiology and outcome. per 100,000 person-​ years from 1993 to 2000.
Smink et al. (2016) found a significant decrease in
Epidemiology BN incidence from 8.6 per 100,000 in 1985–​1989
According to the DSM-​5 (APA, 2013), the 12-​ to 3.2 per 100,000 person years in 2005–​2009 in
month prevalence of BN is 1% to 1.5% in ado- contrast to stability of AN incidence over this same
lescent and young adult women. Consistent with period of observation. Finally, Keel, Heatherton,
this range, recent population-​based studies have Dorer, Joiner, and Zalta (2006) reported a signifi-
reported that BN affects 1.5% of US women cant decline in the point prevalence of BN across
(Hudson et  al., 2007), 2.9% of Australian three college cohorts assessed in 1982, 1992, and
women (Wade et  al., 2006), and 4.6% of Italian 2002, though this was largely explained by particu-
women (Favaro et  al., 2003). Similar to findings larly high rates in the 1982 cohort, as no significant
for AN, the gender ratio of BN is approximately 10 difference was observed between the 1992 and 2002
women for every 1 man, according to the DSM-​5 cohorts.
(APA, 2013). However, this appears to reflect ratios These trends contrast with those reported for
observed from treatment-​ seeking samples. The population-​based cohorts assessed in Australia in
population-​based National Comorbidity Survey—​ 1995 and 2005 (Hay, Mond, Buttner, & Darby,
Replication reported a lifetime prevalence of 1.5% 2008) in which bulimic behaviors, such as binge
for BN in women and 0.5% in men (Hudson eating and purging, increased twofold over time.
et al., 2007). The National Comorbidity Survey—​ Of note, this study did not examine changes in
Replication Adolescent Supplement reported a BN point prevalence over time because baseline
lifetime prevalence of 1.3% for BN in girls 0.5% assessments did not include all items necessary for
for boys (Swanson, Crow, le Grange, Wendsen, & a diagnosis. Taken together, data suggest that BN
Merikangas, 2011). Thus, in the general popula- incidence and prevalence is highly variable within
tion, most of whom never seek treatment for an populations over time. Such patterns are consistent
eating disorder, the gender ratio appears to be with the conclusion that BN is a culturally bound
closer to 3:1. Onset typically occurs in late adoles- syndrome (Keel & Klump, 2003) and more likely to
cence to young adulthood (APA, 2000). Thus, peak fluctuate in relation to immediate cultural factors.
age at onset overlaps somewhat with AN but tends
to fall in a slightly older age range for BN. Cross-​Cultural Patterns
In a meta-​analysis of studies reporting changes As with the review of epidemiological studies for
in BN incidence over time, Keel and Klump (2003) AN, the aforementioned studies on BN have come
detected a significant increase in BN incidence over from Western cultures. A  review of cross-​cultural
the 20th century that was associated with a large evidence for BN (Keel & Klump, 2003) did not
effect size. Consistent with these findings, preva- find any cases of BN in the absence of exposure to
lence estimates across birth cohorts in a large epi- Western influences. Instead, evidence suggested that
demiological sample (Hudson et al., 2007) indicate bulimic symptoms emerged in non-​Western cul-
higher prevalence of BN in younger cohorts com- tures following exposure to Western ideals (Becker,
pared with older cohorts. Burwell, Gilman, Herzog, & Hamburg, 2002). In a

Keel 37
recent meta-​analysis, Kolar et al. (2016) reported a less lethal methods. This pattern has contributed to
1.16% point prevalence for BN in Latin American speculation that suicide attempts may serve a differ-
countries. Similarly, van Hoeken, Burns, and Hoek ent purpose in BN compared to AN. In AN, many
(2016) reported a point prevalence of 0.87% for suicide attempts may be motivated by the desire
BN in African countries. Thus, the emergence of to end life whereas suicide attempts in BN may
BN appears to be closely linked to a combination be motivated by a desire to regulate negative affect
of access to large quantities of edible food, mod- (Franko & Keel, 2006).
ern plumbing to facilitate purging in private, and a
modern idealization of thinness to make purging a Binge Eating Disorder
culturally meaningful response to binge eating (Keel Although the term “binge eating disorder” was
& Klump, 2003). first described by Dr.  Albert Stunkard in 1959
(Stunkard, 1959), it was not included as an offi-
Course cial diagnostic category in the DSM until publica-
Bulimia nervosa is associated with a more favor- tion of the fifth edition in 2013. Data on changing
able course and outcome compared to AN (APA, incidence rates, long-​term course, and outcome for
2013). Across studies examining patients 5 or more BED are limited relative to those available for AN
years after diagnosis, approximately 70% achieve and BN given the relatively recent inclusion of BED
recovery, another 20% remain symptomatic but in the DSM.
demonstrate some improvement, and 10% remain
chronically ill (Fichter & Quadflieg, 2007; Herzog Epidemiology
et al., 1999; Keel & Mitchell, 1997; Keel, Mitchell, The recent replication of the National
Miller, Davis, & Crow, 1999). Longer follow-​up Comorbidity Survey suggested that the lifetime
studies suggest that approximately 65%–​70% of prevalence of BED was 3.5% among women and
BN patients achieve remission, while 10% follow 2.0% among men in the United States (Hudson
a chronic course (Keel, 2016). However, as dura- et al., 2007). Lifetime prevalence for BED was 2.3%
tion of follow-​up increases, women become more in adolescent girls and 0.8% in adolescent boys
likely to suffer from an OSFED rather than meet- (Swanson et  al., 2011). Although BED is signifi-
ing full criteria for BN (Keel & Brown, 2010; Keel, cantly more common in women than in men, gen-
Mitchell, Miller, Davis, & Crow, 2000). der discrepancies appear less robust than reported
In a direct comparison of standardized mortal- for AN or BN. Mean and median age at onset for
ity ratios between AN and BN in a single study BED were 25.4 and 21 years, respectively (Hudson
using the same methods of ascertainment, Keel et  al., 2007), suggesting that BED may be associ-
et al. (2003) found that BN is less likely to result ated with a slightly older age of onset compared to
in premature death compared to AN. This obser- AN and BN. In female cohorts, Wade et al. (2006)
vation was confirmed by a recent meta-​analysis, in reported a lifetime prevalence of 2.9% for BED in
which the 95% confidence interval for the SMR for Australia, and Favaro et al. (2003) reported a life-
BN did not overlap with that for AN (Arcelus et al., time prevalence of 0.6% for BED in Italy. Favaro
2011). Overall, BN is associated with a significant et al. (2003) further reported that the current preva-
increased risk of premature death from all causes lence of BED was 0.1%. In contrast, Hay and col-
(SMR = 1.93) (Arcelus et al., 2011) and from sui- leagues (2008) reported that the current prevalence
cide (SMR = 7.5) (Preti et al., 2011). of BED in a population-​based sample of Australian
Although completed suicide appears to be more men and women was 2.3%, with 67% of cases
common in AN than in BN, the prevalence of occurring in women. Thus, prevalence estimates for
suicide attempts has not differed across narrowly BED vary considerably across studies.
defined eating disorder diagnoses, occurring in Incidence studies have not yet been published
approximately 12% of individuals with eating disor- for BED, making it difficult to draw conclusions
ders in a Swedish Twin Registry (Pisetsky, Thornton, regarding time trends for the number of individuals
Lichtenstein, Pedersen, & Bulik, 2013). In a study affected with the syndrome in a given population
of 295 women with bulimic symptoms, drug over- over time. However, Hudson et al. (2007) reported
dose was the most common method employed for that risk of BED diagnosis increased over succes-
attempting suicide, at 79% (Corcos et  al., 2002). sive birth cohorts from 1944 to 1985. This suggests
Thus, the discrepancy between suicide attempts and that BED has become more common in the United
suicide completions in BN may reflect reliance on States over recent decades. Similarly, in Australia,

38 Epidemiology and Course
Mitchison, Hay, Slewa-​Younan, and Mond (2012) 13.3% and crossover to BN was 10%. Collapsing
reported a significant increase in the prevalence of across eating disorder categories, these results sug-
binge eating from 1998 to 2008; however, this was gest that 20% of BED patients met criteria for an
not restricted to its presence in the syndrome of eating disorder at 6-​year follow-​up, whereas 30%
BED versus BN or OSFED. met criteria for an eating disorder at 12-​year follow-​
up. Overall, results suggest affected individuals
Cross-​Cultural Patterns may have extended periods in which they are free
Kolar et  al. (2016) found a lifetime prevalence of BED symptoms yet may be at risk for a resur-
of 3.5% for BED in Latin American countries in gence of symptoms over an extended period of their
their meta-​analysis. In addition, cases of BED or lives. This conclusion is consistent with retrospec-
subthreshold BED have been reported in China, tive reports of long illness duration (e.g., more than
Malaysia, India, Jordan, Iran, Egypt, and Tunisia 10 years) in epidemiological studies (Striegel-​Moore
(Keel, 2016), covering all five non-​Western regions & Franko, 2008).
originally reviewed by Keel and Klump (2003). Only two deaths have been reported among
These reports are recent, ranging from 2009 to patients with BED, both in the same cohort
2016, suggesting that cases may have emerged fol- (Ficther, Quadflieg, & Gnutzmann, 1998; Fichter,
lowing exposure to Western influences that impact Quadflieg, & Hedlund, 2008), yielding a crude
emergence of BN. mortality ratio of 3% and an SMR of 2.29 (which
was not statistically significant owing to the small
Course sample size). A meta-​analysis of suicide risk in eat-
Given that BED was introduced as a provisional ing disorders found no reports of suicide in patients
diagnostic category in the DSM-​IV (APA, 1994), it with BED (Preti et  al., 2011). In contrast to an
is not surprising that few long-​term follow-​up stud- absence of reported deaths by suicide at long-​term
ies have been conducted on the illness. Fairburn, follow-​up in BED, suicide attempts do not differ in
Cooper, Doll, Norman, and O’Connor (2000) prevalence between BED and other eating disorders
described the natural course of BED over a 5-​ (Pisetsky et al., 2013).
year follow-​up period in a relatively small (n = 48)
community-​based sample. Results from this study Other Specified Feeding or Eating Disorder
suggested that 82% of individuals were substan- Despite inclusion of BED as a full-​threshold eat-
tially improved or recovered, 4% continued to meet ing disorder and expansion of diagnostic boundar-
full criteria for BED, and no deaths were observed. ies for AN and BN in the DSM-​5 (APA, 2013),
Fairburn et al. (2000) further reported no crossover community-​based two-​stage epidemiological stud-
from BED to AN and that 3% of individual with ies have suggested that the most common eating
BED met criteria for BN during follow-​up. Fichter disorder diagnosis is OSFED (Keel, Brown, Holm-​
and Quadflieg (2007) reported somewhat less favor- Denoma, & Bodell, 2011; Machado, Gonçalves,
able outcome at 6-​year follow-​up in their sample & Hoek, 2013). To address the inability to derive
of 60 inpatients with BED; 78.3% were recovered, meaningful information on epidemiology, course,
6.7% continued to meet full criteria for BED, and or outcome from the hodgepodge of syndromes
1.7% had died. Similar to findings from Fairburn previously inhabiting the DSM-​IV eating disorder
et  al. (2000), no crossover from BED to AN was not otherwise specified category (Wilfley, Bishop,
observed; however, 8.3% met criteria for BN at Wilson, & Agras, 2007), the DSM-​ 5 Eating
follow-​up and 5.0% met criteria for an EDNOS Disorders Workgroup introduced named condi-
(Fichter & Quadflieg, 2007). tions among the OSFED. The following sections
Comparison of data from 6-​versus 12-​ year review epidemiology and course for two specified
follow-​up of the BED cohort described by Fichter forms of OSFED, purging disorder (PD) and night
and Quadflied (2007) yields some counterintuitive eating syndrome (NES).
findings. Although recovery rates often improve
as duration of follow-​ up increases, this was not Purging Disorder
observed for BED. Instead, at 12-​year follow-​up, Purging disorder involves recurrent purging
66.7% of patients were recovered, 6.7% continued in the absence of objectively large binge episodes.
to meet full criteria for BED, and 3.3% had died. This feature is accompanied by undue influence of
No cases of crossover to AN were observed at 12-​ weight or shape on self-​evaluation or an intense fear
year follow-​up; however, crossover to EDNOS was of gaining weight or becoming fat (APA, 2013).

Keel 39
Despite the very recent description of PD in the had died over follow-​up, representing a significantly
literature, several epidemiological studies have increased risk of premature death (SMR = 3.9). In
reported on the frequency of this condition, and a comparison of mortality across eating disorders,
some limited data are available on course (Keel & Koch et al. (2013) found a significantly greater risk
Striegel-​Moore, 2009). of death in PD than in BN and no significant differ-
ence between purging disorder and AN.
Epidemiology A prospective community-​ based 6-​ month
Lifetime prevalence estimates for PD in follow-​up of 23 women with PD (Keel, Haedt, &
women have ranged from 1.1% in Italy (Favaro Edler, 2005) found that 13% achieved partial or
et  al., 2003) to 5.3% in Australia (Wade et  al., full remission. Based on combined results from two
2006). Of interest, in both studies, estimates of school-​based longitudinal studies (Allen, Byrne,
PD excluded individuals with lifetime diagnoses Oddy, & Crosby, 2013; Stice, Marti, Shaw, &
of AN, BN, or BED, thus controlling for cross- Jaconis, 2009), approximately 75% of individuals
over. Purging disorder is likely to be more com- with PD achieve remission over follow-​up of inter-
mon than suggested by these studies. Of further mediate length (less than 5 years). However, stud-
interest, both Favaro et al. (2003) and Wade et al. ies examining illness status more than 5 years after
(2006) reported lifetime prevalence estimates for presentation suggest lower recovery rates, with less
PD that were nearly twice as large as estimates than half of individuals with PD achieving remis-
for BED. sion (Allen et al., 2013; Koch, Quadflieg, & Fichter,
Current prevalence for PD has been reported 2013). The data of Koch et al. (2013) indicated that
in 0.5% of Australian women (Hay et al., 2008) PD has a more favorable outcome than AN.
and 0.6% of Canadian women (Gauvin, Steiger,
& Brodeur, 2009) in two-​stage population-​based Night Eating Syndrome
studies. Studies of college-​ based samples in the Night eating syndrome (NES) is characterized
United States suggest that between 0.6% (Haedt by recurrent nocturnal eating that is excessive or
& Keel, 2010) and 0.8% of women (Crowther, interferes with sleep and contributes to distress or
Armey, Luce, Dalton, & Leahey, 2008) have PD. impairment in the individual (APA, 2013). Studies
Consistency of results for point prevalence esti- of NES prevalence have been restricted to single
mates (ranging only from 0.5% to 0.8%) across cohorts, and there are no studies of NES incidence
populations, assessment methods, and syndrome over time. In addition, there are no data regarding
definitions is striking and may reflect the extent course or outcome for NES.
to which questions regarding the use of purging
methods (e.g., self-​induced vomiting, laxative, or Epidemiology
diuretics) to influence weight or shape are associ- In a population-​based sample of Black and White
ated with higher reliability in various methods of women, prevalence of NES was 1.6% (Striegel-​
assessment. Age at onset for PD has been reported Moore, Dohm, Hook, Schreiber, Crawford, &
in late adolescence (Favaro et  al., 2003), and PD Daniels, 2005). In a multicenter study of 2,266
is significantly more common in women than men obese adults assessed before undergoing bariatric
(Haedt & Keel, 2010). surgery, NES was diagnosed in 17.7% of patients
Two studies have examined point prevalence (Mitchell et al., 2015). In a sample of 395 families
of PD in successive cohorts of college students in Canada in which one parent was obese, NES was
(Crowther et  al., 2008; Haedt & Keel, 2010) and found to affect 0.0% of children, 0.5% of mothers,
reported no significant changes over time. Of note, and 0.3% of fathers (Lundgren et al., 2012). In a
1982 was the earliest observation of PD point prev- multisite study of 845 patients diagnosed with type
alence across these studies. Thus, it is possible that 2 diabetes, 3.8% were diagnosed with NES (Allison
the syndrome became increasingly common during et  al., 2007). Differences in sample ascertainment
the 20th century in the period leading up to when are likely to be responsible for this very large range
estimates were first described. of prevalence estimates. Thus, more research is
needed to understand who suffers from NES. One
Course challenge for this research is that NES criteria differ
Koch, Quadflieg, and Fichter (2013) conducted qualitatively from those for defined eating disorders,
a 9-​year follow-​up of 225 patients diagnosed with making data from most epidemiological studies of
PD between 1999 and 2005. Overall, 5% of patients eating disorders irrelevant.

40 Epidemiology and Course
Conclusion when reliable information is obtained on course and
Evidence suggests that AN, BN, and BED have outcome.
affected increasing proportions of the population over
the course of the 20th century. The most dramatic Future Directions
changes have been observed for syndromes charac- More studies are needed to understand the epi-
terized by binge eating and may reflect increased demiology, course, and outcome associated with
urbanization and increased access to large quanti- OSFED such as PD and NES. For these to be
ties of readily edible food (Keel & Klump, 2003). informative, it will be crucial for investigators to
Conversely, the rise in AN incidence may reflect separate different forms of OSFED in providing
the increased idealization of thinness in modern, these descriptive data. Questions that present future
Western culture (Keel & Klump, 2003). However, directions for the field are presented here:
the thin ideal is neither sufficient nor necessary for 1.  Is there a valid threshold for distinguishing
the emergence of a syndrome characterized by self-​ AN from atypical AN based on epidemiological
starvation that predominately affects adolescent and patterns, long-​term course, or outcome?
young adult females (Keel & Klump, 2003). Less 2.  Is there a valid threshold for distinguishing
is known about time trends for the incidence or BN from subthreshold BN based on epidemiologi-
prevalence of PD. Although purging as a behavior cal patterns, long-​term course, or outcome?
demonstrated an increase in one study (Hay et al., 3.  In both community and clinic-​based popu-
2008), two other studies reported no significant lations, has the incidence of BED, PD, or NES
change in PD point prevalence over limited period changed over time?
of observation toward the end of the 20th century 4.  Is BED, PD, or NES a culture-​ bound
(Crowther et al., 2008; Haedt & Keel, 2010). syndrome?
Data on course and outcome suggest that AN, 5.  What are the long-​term course and outcomes
BN, and BED can be placed on a continuum of associated with BED, PD, and NES?
illness severity, with AN residing at the end with
greatest chronicity and risk of premature death Note
and BED residing at the end with lowest chronic- 1. Exceptions to this occur when point prevalence estimates are
taken from independent cohorts at different times from a
ity. However, these conclusions must be tempered
given population (e.g., Keel, Heatherton, Dorer, Joiner, &
by awareness of limited longitudinal data available Zalta, 2006).
for BN and BED. The addition of new prospec-
tive longitudinal studies with longer durations of References
follow-​up may have a particularly profound impact Allen, K. L., Byrne, S. M., Oddy, W. H., Crosby, R. D. (2013).
on our understanding of BED course and outcome, DSM-​ IV-​TR and DSM-​ 5 eating disorders in adoles-
given that comparison of 6-​and 12-​year BED out- cents:  Prevalence, stability, and psychosocial correlates in a
population-​based sample of male and female adolescents.
come suggested less favorable course as duration of
International Journal of Eating Disorders, 122, 720–​732.
follow-​up increased (Fichter & Quadflied, 2007). In Allison, K. C., Crow, S. J., Reeves, R. R., West, D. S., Foreyt, J.
addition, given the significant association between P., Dilillo, V. G.,  . . .  Stunkard, A. J. (2007). Binge eating
BED and obesity (Hudson et  al., 2007), it seems disorder and night eating syndrome in adults with type 2
likely that increased risk of premature death will be diabetes. Obesity, 15, 1287–​1293.
American Psychiatric Association (APA). (1994). Diagnostic and
observed in BED as studies with larger sample sizes
statistical manual of mental disorders (4th ed.). Washington,
and longer follow-​up durations are conducted. DC: Author.
Examining information available on the epide- American Psychiatric Association (APA). (2000). Diagnostic
miology and course of different eating disorders and statistical manual of mental disorders (4th ed., text rev.).
reveals an inverse association between the preva- Washington, DC: Author.
American Psychiatric Association (APA). (2013). Diagnostic and
lence of a syndrome and how much is known about
statistical manual of mental disorders (5th ed.). Washington,
the syndrome’s course and outcome. Thus, we DC: Author.
know the least about the most common syndromes. Arcelus, J., Mitchell, A. J., Wales, J., & Nielsen, S. (2011).
Although much of may be addressed by revisions Mortality rates in patients with anorexia nervosa and other
in the DSM-​5, which included more named and eating disorders:  A  meta-​analysis of 36 studies. Archives of
General Psychiatry, 68, 724–​731.
characterized disorders, such as the promotion
Becker, A. E., Burwell, R. A., Gilman, S. E., Herzog, D. B., &
of BED to an official diagnosis and PD and NES Hamburg, P. (2002). Eating behaviours and attitudes fol-
named as characterized forms of OSFED, there will lowing prolonged exposure to television among ethnic Fijian
be an unavoidable lag between these changes and adolescent girls. British Journal of Psychiatry, 180, 509–​514.

Keel 41
Bulik, C. M., Thornton, L., Pinheiro, A. P., Plotnicov, K., follow-​up study. Journal of the American Academy of Child
Klump, K. L., Brandt, H.,  . . .  Kaye, W. H. (2008). Suicide and Adolescent Psychiatry, 38, 829–​837.
attempts in anorexia nervosa. Psychosomatic Medicine, 70, Hoek, H. W. (2006). Incidence, prevalence and mortality
378–​383. of anorexia nervosa and other eating disorders. Current
Corcos, M., Taieb, O., Benoit-​Lamy, S., Paterniti, S., Jeammet, Opinions in Psychiatry, 19, 389–​394.
P., & Flament, M. F. (2002). Suicide attempts in women Hoek, H. W., & van Hoeken, D. (2003). Review of the preva-
with bulimia nervosa:  Frequency and characteristics. Acta lence and incidence of eating disorders. International Journal
Psychiatrica Scandinavica, 106, 381–​386. of Eating Disorders, 34, 383–​396.
Crowther, J. H., Armey, M., Luce, K. H., Dalton, G. R., & Holm-​Denoma, J. M., Witte, T. K., Gordon, K. H., Herzog, D.
Leahey, T. (2008). The point prevalence of bulimic disorders B., Franko, D. L., Fichter, M.,  . . .  Joiner, T. E., Jr. (2008).
from 1990 to 2004. International Journal of Eating Disorders, Deaths by suicide among individuals with anorexia as arbi-
41, 491–​497. ters between competing explanations of the anorexia-​suicide
Currin, L., Schmidt, U., Treasure, J., & Jick, H. (2005). Time link. Journal of Affective Disorders, 107, 231–​236.
trends in eating disorder incidence. British Journal of Hudson, J. I., Hiripi, E., Pope, H. G., Jr., & Kessler, R. C.
Psychiatry, 186, 132–​135. (2007). The prevalence and correlates of eating disorders
Dalle Grave, R., & Calugi, S. (2007). Eating disorder not other- in the National Comorbidity Survey Replication. Biological
wise specified in an inpatient unit: The impact of altering the Psychiatry, 61, 348–​358.
DSM-​IV criteria for anorexia and bulimia nervosa. European Javaras, K. N., Runfola, C. D., Thornton, L. M., Agerbo, E.,
Eating Disorder Review, 15, 340–​349. Birgegård, A., Norring, C.,  . . .  Bulik, C. M. (2015). Sex-​
Fairburn, C. G., Cooper, Z., Doll, H. A., Norman, P., & and age-​ specific incidence of healthcare-​ register-​
recorded
O’Connor, M. (2000). The natural course of bulimia ner- eating disorders in the complete Swedish 1979–​2001
vosa and binge eating disorder in young women. Archives of birth cohort. International Journal of Eating Disorders, 48,
General Psychiatry, 57, 659–​665. 1017–​1081.
Favaro, A., Ferrara, S., & Santanastaso, P. (2003). The spectrum Keel, P. K. (2016). Eating disorders (2nd ed.). New  York,
of eating disorders in young women: A prevalence study in NY: Oxford University Press.
a general population sample. Psychosomatic Medicine, 65, Keel, P. K., & Brown, T. A. (2010). Update on course and out-
701–​708. come in eating disorders. International Journal of Eating
Fichter, M. M., & Quadflieg, N. (2007). Long-​term stability Disorders, 43, 195–​204.
of eating disorder diagnoses. International Journal of Eating Keel, P. K., Brown, T. A., Holm-​Denoma, J., & Bodell, L. P.
Disorders, 40, S61–​S66. (2011). Comparison of DSM-​IV versus proposed DSM-​5
Ficther, M. M., Quadflieg, N., & Gnutzmann, A. (1998). Binge diagnostic criteria for eating disorders: Reduction of eating
eating disorder:  Treatment outcome over a 6-​year course. disorder not otherwise specified and validity. International
Journal of Psychosomatic Research, 44, 385–​405. Journal of Eating Disorders, 44, 553–​560.
Fichter, M. M., Quadflieg, N., & Hedlund, S. (2008). Long-​ Keel, P. K., Dorer, D. J., Eddy, K. T., Franko, D., Charatan, D.
term course of binge eating disorder and bulimia ner- L., & Herzog, D. B. (2003). Predictors of mortality in eating
vosa:  Relevance for nosology and diagnostic criteria. disorders. Archives of General Psychiatry, 60, 179–​183.
International Journal of Eating Disorders, 41, 577–​586. Keel, P. K., Haedt, A., & Edler, C. (2005). Purging disorder: An
Franko, D. L., & Keel, P. K. (2006). Suicidality in eating dis- ominous variant of bulimia nervosa? International Journal of
orders:  Occurrence, correlates, and clinical implications. Eating Disorders, 38, 191–​199.
Clinical Psychology Review, 26, 769–​782. Keel, P. K., Heatherton, T. F., Dorer, D. J., Joiner, T. E., & Zalta,
Franko, D. L., Keshaviah, A., Eddy, K. T., Krishna, M., Davis, A. K. (2006). Point prevalence of bulimia nervosa in 1982,
M. C., Keel, P. K., & Herzog, D. B. (2013). A longitudinal 1992, and 2002. Psychological Medicine, 36, 119–​127.
investigation of mortality in anorexia nervosa and bulimia Keel, P. K., & Klump, K. L. (2003). Are eating disorders culture-​
nervosa. American Journal of Psychiatry, 170, 917–​925. bound syndromes? Implications for conceptualizing their
Gauvin, L., Steiger, H., & Brodeur, J. M. (2009). Eating-​disorder etiology. Psychological Bulletin, 129, 747–​769.
symptoms and syndromes in a sample of urban-​dwelling Keel, P. K., & Mitchell, J. E. (1997). Outcome in bulimia ner-
Canadian women:  Contributions toward a population vosa. American Journal of Psychiatry, 154, 313–​321.
health perspective. International Journal of Eating Disorders, Keel, P. K., Mitchell, J. E., Miller, K. B., Davis, T. L., & Crow, S.
42, 531–​536. J. (1999). Long-​term outcome of bulimia nervosa. Archives
Gull, W. W. (1874). Anorexia nervosa (apepsia hysterica, of General Psychiatry, 56, 63–​69.
anorexia hysterica). Transactions of the Clinical Society of Keel, P. K., Mitchell, J. E., Miller, K. B., Davis, T. L., & Crow, S.
London, 7, 22–​28. J. (2000). Predictive validity of bulimia nervosa as a diagnos-
Haedt, A. A., & Keel, P. K. (2010). Comparing definitions of tic category. American Journal of Psychiatry, 157, 136–​138.
purging disorder on point prevalence and associations with Keel, P. K., & Striegel-​Moore, R. H. (2009). The validity and
external validators. International Journal of Eating Disorders, clinical utility of purging disorder. International Journal of
43, 433–​439. Eating Disorders, 42, 706–​719.
Hay, P. J., Mond, J., Buttner, P., & Darby, A. (2008). Eating dis- Keski-​Rahkonen, A., Hoek, H. W., Susser, E. S., Linna, M.
order behaviors are increasing: Findings from two sequential S., Sihvola, E., Raevuori, A.,   .  .  .  Rissanen, A. (2007).
community surveys in South Australia. PLoS ONE, 3, e1541. Epidemiology and course of anorexia nervosa in the commu-
Herzog, D. B., Dorer, D. J., Keel, P. K., Selwyn, S. E., Ekeblad, nity. American Journal of Psychiatry, 164, 1259–​1265.
E. R., Flores, A. T.,  . . .  Keller, M. B (1999). Recovery and Koch, S., Quadflieg, N., & Fichter, M. (2013). Purging dis-
relapse in anorexia nervosa and bulimia nervosa: A 7.5 year order:  A  comparison to established eating disorders with

42 Epidemiology and Course
 purging behaviour. European Eating Disorders Review, 21, eating disorders in Dutch primary care: Decreasing incidence
265–​275. of bulimia nervosa but not of anorexia nervosa. Psychological
Kolar, D. R., Rodriguez, D. L., Chams, M. M., & Hoek, H. Medicine, 46, 1189–​1196.
W. (2016). Epidemiology of eating disorders in Latin Steinhausen, H. C. (2002). The outcome of anorexia nervosa in
America:  A  systematic review and meta-​ analysis. Current the 20th century. International Journal of Eating Disorders,
Opinion in Psychiatry, 29, 363–​371. 159, 1284–​1293.
Löwe, B., Zipfel, S., Buchholz, C. Dupont, Y., Reas, D. L., & Steinhausen, H. C., & Jensen, C. M. (2015). Time trends in life-
Herzog, W. (2001). Long-​term outcome of anorexia ner- time incidence rates of first-​time diagnosed anorexia nervosa
vosa in a prospective 21  year follow-​up study. Psychological and bulimia nervosa across 16 years in a Danish nationwide
Medicine, 31, 881–​890. psychiatric registry study. International Journal of Eating
Lucas, A. R., Crowson, C. S., O’Fallon, W. M., & Melton, L. J. Disorders, 48, 845–​850.
(1999). The ups and downs of anorexia nervosa. International Stice, E., Marti, C. N., Shaw, H., & Jaconis, M. (2009). An 8-​
Journal of Eating Disorders, 26, 397–​405. year longitudinal study of the natural history of threshold,
Lundgren, J. D., Drapeau, V., Allison, K. C., Gallant, A. R., subthreshold, and partial eating disorders from a community
Tremblay, A., Lambert, M. A.,  . . .  Stundard, A. J. (2012). sample of adolescents. Journal of Abnormal Psychology, 118,
Prevalence and familial patterns of night eating in the Québec 587–​597.
Adipose and Lifestyle Investigation in Youth (QUALITY) Striegel-​Moore, R. H., Dohm, F. A., Hook, J. M., Schreiber, G.
study. Obesity, 20, 1598–​1603. B., Crawford, P. B., & Daniels, S. R. (2005). Night eating
Machado, P. P., Gonçalves, S., & Hoek, H. W. (2013). DSM-​ syndrome in young adult women: Prevalence and correlates.
5 reduces the proportion of EDNOS cases:  Evidence International Journal of Eating Disorders, 37, 200–​206.
from community samples. International Journal of Eating Striegel-​Moore, R. H., & Franko, D. L. (2008). Should binge
Disorders, 46, 60–​65. eating disorder be included in the DSM-​V? A critical review
Milos, G., Spindler, A., Schnyder, U., Martz, J., Hoek, H. W., of the state of the evidence. Annual Review of Clinical
& Willi, J. (2004). Incidence of severe anorexia nervosa in Psychology, 4, 305–​324.
Switzerland: 40 years of development. International Journal Stunkard, A. J. (1959). Eating patterns and obesity. Psychiatric
of Eating Disorders, 35, 250–​258. Quarterly, 33, 284–​295.
Mitchell, J. E., King, W. C., Courcoulas, A., Dakin, G., Elder, Sullivan, P. F. (1995). Mortality in anorexia nervosa. American
K., Engel, S.,  . . .  Wolfe, B. (2015). Eating behavior and eat- Journal of Psychiatry, 152, 1073–​1074.
ing disorders in adults before bariatric surgery. International Swanson, S. A., Crow, S. J., le Grange, D., Wendsen, J., &
Journal of Eating Disorders, 48, 215–​222. Merikangas, K. R. (2011). Prevalence and correlates of
Mitchison, D., Hay, P., Slewa-​Younan, S., & Mond, J. (2012). eating disorders in adolescents. Results from the National
Time trends in population prevalence of eating disorder Comorbidity Survey Replication Adolescent Supplement.
behaviors and their relationship to quality of life. PLoS One, Archives of General Psychiatry, 68, 714–​723.
7, e48450. van Hoeken, D., Burns, J. K., & Hoek, H. W. (2016).
Nielsen, S., Moller-​Madsen, S., Isager, T., Jorgensen, J., Pagsberg, Epidemiology of eating disorders in Africa. Current Opinion
K., & Theander, S. (1998). Standardized mortality in eating in Psychiatry, 29, 372–​377.
disorders: A quantitative summary of previously published and van Son, G. E., van Hoeken, D., Bartelds, A. I., van Furth,
new evidence. Journal of Psychosomatic Research, 44, 413–​434. E. F., & Hoek, H. W. (2006). Time trends in the inci-
Pisetsky, E. M., Thornton, L. M., Lichtenstein, P., Pedersen, dence of eating disorders:  A  primary care study in the
N. L., & Bulik C. M. (2013). Suicide attempts in women Netherlands. International Journal of Eating Disorders, 39,
with eating disorders. Journal of Abnormal Psychology, 122, 565–​569.
1042–​1056. Wade, T. D., Bergin, J. L., Tiggemann, M., Bulik, C. M., &
Preti, A., Rocchi, M. B., Sisti, D, Camboni, M. V., & Miotto, P. Fairburn, C. G. (2006). Prevalence and long-​term course
(2011). A comprehensive meta-​analysis of the risk of suicide of lifetime eating disorders in an adult Australian twin
in eating disorders. Acta Psychiatrica Scandinavica, 124, 6–​17. cohort. Australian and New Zealand Journal of Psychiatry, 40,
Russell, G. F. M. (1979). Bulimia nervosa: An ominous variant 121–​128.
of anorexia nervosa. Psychological Medicine, 9, 429–​448. Wilfley, D. E., Bishop, M. E., Wilson, G. T., & Agras, W. S.
Smink, F. R., van Hoeken, D., Donker, G. A., Susser, E. S., (2007). Classification of eating disorders:  Toward DSM-​V.
Oldehinkel, A. J., & Hoek, H. W. (2016). Three decades of International Journal of Eating Disorders, 40, S123–​S129.

Keel 43
 PART 
2
Approaches
to Understanding
the Eating Disorders
 CH A PT E R


Appetitive Regulation in Anorexia Nervosa
4 and Bulimia Nervosa

Alice V. Ely and Walter H. Kaye

Abstract
Anorexia and bulimia nervosa are complex disorders with dysregulated appetitive behaviors. The
underlying causes of disturbed patterns of eating are unknown, but a growing body of research suggests
that aberrant functioning of brain or peripheral systems may be responsible. Neuroimaging technologies,
such as positron emission tomography (PET) and functional MRI (fMRI), can be used to explore whether
there are perturbations of the monoamine systems, the neurocircuitry of gustatory processing in eating
disorders, and their relationship to metabolic homeostatic states. Together, PET and fMRI data suggest
that individuals with eating disorders have disturbance of taste-​and reward-​processing regions of the
brain, which may contribute to eating disorder symptoms.
Key Words:  eating disorder, anorexia nervosa, bulimia nervosa, positron emission tomography, functional
magnetic resonance imaging, dopamine, serotonin, anterior insula, orbitofrontal cortex

Introduction function; (2)  neuropeptides involved in hypotha-

Eating disorders (EDs) are characterized by puz- lamic and lower brain center function; and (3) lim-
zling symptoms, and the ED field lags behind other bic and cortical brain circuits that contribute to
psychiatric disorders in identifying responsible brain appetite. We believe that the weight of evidence
regions and circuits. Consequently, our under- supports the involvement of higher order circuits,
standing of the pathophysiology of EDs is limited. as they show persistent altered function after recov-
Although anorexia nervosa (AN) and bulimia ner- ery, and because they code for rewarding and emo-
vosa (BN) are characterized (APA, 2000) as EDs, it tionality properties of food, homeostatic needs,
remains unknown as to whether there is a primary and cognitive modulation (Elman et  al., 2006;
disturbance of appetitive function. The regulation Hinton et al., 2004; Kelley, 2004; Saper, Chou, &
of appetite and feeding are complex phenomena, Elmquist, 2002).
integrating peripheral signals (GI tract, adipose tis-
sue, hormonal secretion, etc.), hypothalamic factors State Versus Trait Characteristics
(neuropeptides), cortical and subcortical processes During their disorders, individuals with AN
(reward, emotionality, cognition), and external and BN have widespread and severe alterations of
influences (Elman, Borsook, & Lukas, 2006; Rolls, brain and peripheral organ function; however, it
1997; Schwartz, Woods, Porte, Seeley, & Baskin, is unclear whether these changes are the cause or
2000). It is possible that a disturbance could occur the consequence of maladaptive appetitive behav-
anywhere in this axis in AN and BN. iors such as extreme restriction and the binge-​purge
A number of subsystems have been investigated cycle. Thus, to understand the etiology and course
in AN and BN:  (1)  peripheral hormones such as of illness of AN and BN, it is useful to divide the
cholecystokinin (CCK) or peripheral autonomic neurobiologic alterations into two categories. First,

47
state-​based alterations secondary to malnutrition that childhood temperament and personality traits
or ongoing symptoms may sustain the illness, and can create a vulnerability for developing AN and BN
perhaps accelerate the establishment of chronic and during adolescence (Kaye et  al., 2013; Wierenga,
enduring disturbance. Second, premorbid, geneti- Ely, et al., 2014). No agreed on definition of recov-
cally determined trait alterations may contribute to ery from AN presently exists, but many studies
an initial vulnerability to develop ED. emphasize a definition that comprises an absence
Starvation and emaciation have profound effects of psychological and behavioral symptoms and sta-
on the function of the brain and other organ sys- ble and healthy body weight for months or years,
tems, lead to neurochemical disturbances that with stable nutrition, the relative absence of dietary
could exaggerate premorbid traits (Pollice, Kaye, abnormalities, and normal menstruation (Bardone-​
Greeno, & Weltzin, 1997), and add other symp- Cone et al., 2010). Although the process of recovery
toms that maintain or accelerate the disease proc- in AN and BN is poorly understood and, in most
ess. For example, AN patients have a reduced brain cases, protracted, approximately 50% to 70% of
volume (see Ellison & Fong, 1998); altered metab- affected individuals will eventually have complete
olism in frontal, cingulate, temporal, and parietal or moderate resolution of the illness, though this
regions (Kaye, Wagner, Frank, & Bailer, 2006); and might not occur until early to mid 20s (Steinhausen,
a regression to prepubertal gonadal function (Boyar 2002; Strober, Freeman, & Morrell, 1997; Wagner,
et al., 1974) that contribute to autonomic and hor- Barbarich, et  al., 2006). Studies have described
monal disturbance as well as mood and cognitive temperament and character traits that persist after
symptoms (Jimerson & Wolfe, 2006), which could long-​term recovery from AN, such as negative emo-
sustain ED behaviors. The fact that such distur- tionality, harm avoidance and perfectionism, desire
bances tend to normalize after weight restoration for thinness, and mild dietary preoccupation, which
suggests that these alterations are secondary to the are similar to those described in childhood in people
disorder and not an underlying cause of AN. The who will go on to develop AN and BN (Anderluh,
difficulty in distinguishing alterations due to state Tchanturia, Rabe-​Hesketh, & Treasure, 2003;
and trait characteristics in ED patients has been Casper, 1990; Lilenfeld, Wonderlich, Riso, Crosby,
a major confound in research into this disorder. & Mitchell, 2006; Srinivasagam et al., 1995; Stice,
Premorbid studies are not practical given the young 2002; Strober, 1980; Wagner, Barbarich, et  al.,
age of potential participants, the rarity of the dis- 2006). Studies (Bulik et  al., 2007) suggest these
order, and the need to follow them for many years. traits are heritable, elevated in unaffected family
An alternative strategy is to study individuals who members, and independent of body weight, provid-
have recovered from AN and BN, thus avoiding the ing further evidence that they confer liability to the
confounding influence of malnutrition and weight development of AN.
loss on biological systems. In summary, there is strong evidence support-
Large-​scale community-​based twin studies have ing a genetic influence in AN and BN, as well as
shown that 50% to 80% of the variance in AN and the persistence after recovery of personality and
BN (Berrettini, 2000; Bulik et  al., 2006; Kendler character traits that may predispose an individual
et al., 1991) can be accounted for by genetic factors. to developing these disorders. These data support
The genetic vulnerability to eating disorders may be the likelihood that such persistent symptoms are
expressed as a more diffuse phenotype of continu- not just “scars” caused by chronic malnutrition, but
ous behavioral traits as suggested by evidence of a may reflect underlying traits that contribute to the
significant heritability of disordered eating attitudes, pathogenesis of this disorder.
weight preoccupation, dissatisfaction with weight
and shape, dietary restraint, binge eating and self-​ Studies of Altered Feeding Behaviors
induced vomiting (Ando, Ichimaru, Konjiki, Shoji, The data on appetite regulation in EDs is small,
& Komaki, 2007; Ando et  al., 2010; Bulik et  al., despite the prominent nature of these symptoms, but
2005; Klump, McGue, & Iacono, 2000; Rutherford, growing. Laboratory studies support clinical obser-
McGuffin, Katz, & Murray, 1993; Wade, Martin, & vations that AN individuals dislike high-​fat foods
Tiggemann, 1998), and familiality of subthreshold (Drewnowski, Pierce, & Halmi, 1988; Fernstrom,
forms of eating disorders (Lilenfeld et  al., 1998; Weltzin, Neuberger, Srinivasagam, & Kaye, 1994)
Strober, Freeman, Lampert, Diamond, & Kaye, and BN individuals tend to binge on sweet and
2000; for review, see Trace, Baker, Peñas-​Lledó, & high-​fat foods (Kaye et  al., 1992; Weltzin, Hsu,
Bulik, 2013). Considerable evidence has suggested Pollice, & Kaye, 1991). Notably, these patterns of

48 Appetitive Regul ation

behavior do not appear to change following weight BN (Halmi, Sunday, Puglisi, & Marchi, 1989).
regain. Individuals with AN tend to report elevated Furthermore, negative mood states may precipi-
anxiety in anticipation of eating, even when emaci- tate a binge (Berg et al., 2013; Hilbert & Tuschen-​
ated (Kissileff et al., 2016), and reduced dysphoric Caffier, 2007; Smyth et al., 2007; Waters, Hill, &
mood with food refusal (Kaye, Bulik, Thornton, Waller, 2001) and overeating may relieve dysphoria
Barbarich, & Masters, 2004; Steinglass et al., 2010). and anxiety (Abraham & Beaumont, 1982; Johnson
Ecological momentary assessment findings suggest & Larson, 1982; Kaye, Gwirtsman, George, Weiss,
that certain trajectories of anxiety during the day & Jimerson, 1986), though a meta-​analysis of eco-
are associated with greater fasting and restriction logical momentary assessment data suggests the
in AN (Lavender et  al., 2014). In BN, behavioral relationship between affect and binge eating may be
studies suggest that metabolic state may contribute more complex (Haedt-​Matt & Keel, 2011). These
to binge eating, but the mechanisms are not well alterations in appetite regulation may be explained
understood. For example, some studies show that by alterations in the neurobiological response to
fasting or dietary restriction may increase the likeli- food reward when hungry and fed in the insula and
hood of future binge eating (Stice, Davis, Miller, & orbitofrontal cortex (described later, see Table 4.1).
Marti, 2008; Stice, Presnell, Groesz, & Shaw, 2005; Taken together, these studies support the possibil-
Telch & Agras, 1996; Zunker et al., 2011), suggest- ity of an altered response to palatable foods and a
ing that restricting intake and prolonging hunger dysphoria-​reducing aspect to pathological eating.
states may confer vulnerability for binge eating.
Conversely, other findings indicate that although Neuropeptide and Neuroendocrine
individuals with BN frequently fast between binge/​ Alterations
purge episodes (American Psychiatric Association, The past decade has witnessed accelerating basic
2013), hunger does not necessarily trigger binge research on the role of neuropeptides in the regu-
eating. In fact, lower dietary intake has been asso- lation of feeding behavior. Indirect evidence from
ciated with a reduction in binge eating (Stice, clinical studies of obesity or animal models of ED
Martinez, Presnell, & Groesz, 2006), and a meta-​ suggests the possibility that altered regulation of
analysis of ecological momentary assessment studies neuropeptides may contribute to abnormal eat-
tracking the temporal relationship of hunger and ing patterns. The mechanisms for controlling food
binge eating determined that binge episodes were intake comprise a complicated interplay between
preceded by lower hunger ratings than were typical peripheral systems (including gustatory stimula-
eating episodes (i.e., participants were less hungry tion, gastrointestinal peptide secretion, and vagal
before binge eating than before meals; Haedt-​Matt afferent nerve responses) and central nervous sys-
& Keel, 2011). Subjective ratings of hunger and tem (CNS) neuropeptides and/​ or monoamines.
fullness indicate that perception of metabolic Thus, studies in animals show that neuropeptides
state may also be impaired in individuals ill with such as CCK, the endogenous opioids such as

Table  4.1  Studies Showing That Food Deprivation Activates (Up Arrow) the  Insula and OFC
Regions When Compared to Satiety
Author Year Technique Duration fasting Insula OFC

Small et al. 2001 PET 4.5 hr ↑

Tataranni et al. 1999 PET 36 hr ↑ ↑

Morris & Dolan 2001 PET 16 hr ↑ ↑

Haase et al. 2009 fMRI 12 hr ↑ ↑

Kringelbach et al. 2003 fMRI 6 hr ↑ ↑

Stice et al. 2013 fMRI ~7 hr ↑

Uher, Treasure, Heining, 2006 fMRI 24 hr ↑

Brammer, & Campbell

 Ely, Kaye 49
beta-​endorphin, and neuropeptide-​Y, regulate the to be at least in part a consequence of hypersecre-
rate of eating, the duration selection (Sam, Troke, tion of endogenous corticotropin-​releasing hormone
Tan, & Bewick, 2012; Suzuki, Simpson, Minnion, (CRH) (Gold et al., 1986; Kaye, Gwirtsman, et al.,
Shillito, & Bloom, 2010) and size of meals, as well 1987; Licinio, Wong, & Gold, 1996; Monteleone
as macronutrient (Morley & Blundell, 1988; Sam et al., 2016; Monteleone et al., 2011; Walsh et al.,
et  al., 2012; Saper et  al., 2002; Schwartz et  al., 1987). Plasma and cerebrospinal fluid (CSF) meas-
2000; Suzuki et al., 2010). In addition to regulat- ures return toward normal with weight restoration,
ing eating behavior, a number of CNS neuropep- it appears likely that activation of the hypothalamic-​
tides contribute to the regulation of neuroendocrine pituitary-​adrenal (HPA) axis is precipitated by the
pathways. Clinical studies suggest that CNS neu- disorder and not vice versa. The observation of
ropeptide alterations may lead to dysregulation of increased CRH activity is of great theoretical inter-
gonadal hormone, cortisol, thyroid hormone, and est in anorexia, since intracerebroventricular CRH
growth hormone secretion in ED (Jimerson & administration in experimental animals produces
Wolfe, 2004; P. Monteleone & Maj, 2013; Stoving, many of the physiologic and behavioral changes
Hangaard, Hansen-​Nord, & Hagen, 1999). associated with AN, including markedly decreased
Research suggests that most of the alterations to eating behavior (Glowa & Gold, 1991). Expression
neuroendocrine and neuropeptide systems appar- of CRH in BN is less clear. The CRH triggers secre-
ent during symptomatic episodes of AN and BN tion of adrenocorticotropic hormone (ACTH); CSF
tend to normalize after recovery. This observation levels of ACTH in ill BN were reported to be simi-
suggests that most of these disturbances are conse- lar to controls, while ACTH levels decreased when
quences of malnutrition, weight loss, and/​or altered BN participants did not engage in binge/​ purge
meal patterns, rather than an underlying cause of behaviors (Gwirtsman et  al., 1989). Given a posi-
initially developing ED. However, understanding tive correlation between ACTH and cortisol levels,
of these neuropeptide disturbances may elucidate these data support the hypothesis that binge/​purge
why recovery from AN or BN can be so difficult. behaviors may relieve dysphoric mood states in BN.
In AN, neuroendocrine and neuropeptide distur- Furthermore, in rats, binge eating prone animals
bance secondary to malnutrition may perpetuate have demonstrated hyporeactivity of the HPA axis,
the food refusal and drive for thinness. Symptoms and higher CRF expression in the bed nucleus of the
such as intolerance of feelings of fullness, obses- stria terminalis, which points to a possible mechan-
sions, anhedonia and dysphoric mood, may be ism underlying stress-​induced binge eating (Calvez,
exaggerated by these neuropeptide alterations and de Ávila, Guèvremont, & Timofeeva, 2016).
thus maintain the disorder. In BN, neuroendocrine
and neuropeptide dysfunction due to chronic bing- Opioid Peptides
ing and purging, in addition to caloric restriction, Studies in laboratory animals raise the possibility
may alter hunger and satiety sensations and ability that altered endogenous opioid activity might con-
to regulate eating behavior. Additionally, mutual tribute to pathological feeding behavior in eating
interactions between neuropeptide, neuroendocrine disorders; opioid agonists generally increase, and
and neurotransmitter pathways may contribute to opioid antagonists decrease, food intake (Morley
the constellation of psychiatric comorbidity often et  al., 1985). State-​related reductions in concen-
observed in these disorders. Even after weight gain trations of CSF beta-​endorphin and related opiate
and normalized eating patterns, many individuals concentrations have been found in both under-
who have recovered from AN or BN have physi- weight AN and ill BN participants (Brewerton,
ological, behavioral and psychological symptoms Lydiard, Laraia, Shook, & Ballenger, 1992; Kaye,
that persist for extended periods of time. Menstrual Berrettini, et  al., 1987; Lesem, Berrettini, Kaye,
cycle dysregulation, for example, may persist for & Jimerson, 1991; Tortorella et al., 2014). In con-
some months after weight restoration. The follow- trast, using the T-​lymphocyte as a model system,
ing sections provide a brief overview of studies of Brambilla et al. (Brambilla, Brunetta, Peirone, et al.,
neuropeptides in AN and BN. 1995) found elevated beta-​endorphin levels in AN,
although the levels were normal in BN (Brambilla,
Hypothalamic-​Pituitary-​Adrenal Brunetta, Draisci, et  al., 1995). If beta-​endorphin
(HPA) Axis activity is a facilitator of feeding behavior, then
When underweight, patients with AN have reduced CSF concentrations could reflect decreased
increased plasma cortisol secretion that is thought central activity of this system, which then maintains

50 Appetitive Regul ation

or facilitates inhibition of feeding behavior in the intake. Release of CCK is thought to be one means
eating disorders. of transmitting satiety signals to the brain by way of
vagal afferents (Gibbs, Young, & Smith, 1973). In
Neuropeptide-​Y and Peptide YY parallel to its role in satiety in rodents, exogenously
These peptides are of considerable theoreti- administered CCK reduces food intake in humans
cal interest, since they are among the most potent (Kissileff, Pi-​Sunyer, Thornton, & Smith, 1981).
endogenous stimulants of feeding behavior within The preponderance of data suggests that patients
the CNS (Kalra, Dube, Sahu, Phelps, & Kalra, 1991; with BN, in comparison to controls, have dimin-
Morley et al., 1985; Schwartz et al., 2000). Peptide ished release of CCK following ingestion of a stan-
YY (PYY) is more potent than neuropeptide-​ Y dardized test-​meal (Devlin et  al., 1997; Geracioti
(NPY) in stimulating food intake; both are selec- & Liddle, 1988; Keel, Wolfe, Liddle, De Young, &
tive for carbohydrate rich foods. In AN women, Jimerson, 2007; Phillipp, Pirke, Kellner, & Krieg,
baseline plasma PYY levels have been reported to be 1991; Pirke, Kellner, Friess, Krieg, & Fichter, 1994),
significantly lower (Germain et al., 2010; Germain though a recent study showed similar CCK levels in
et al., 2007), higher (Misra et al., 2006), and similar BN patients as in healthy controls (Hannon-​Engel,
(Kaye, Berrettini, et al., 1990; Stock et al., 2005) to & Filin, Wolfe, 2013). Measurements of basal CCK
healthy participants, while plasma PYY response to values in blood lymphocytes and in CSF also appear
food intake has been shown to be delayed (Stock to be decreased in patients with BN (Brambilla,
et al., 2005) and not restored with weight restora- Brunetta, Draisci, et al., 1995; Lydiard et al., 1993).
tion (Nakahara et al., 2007). Underweight AN have It has been suggested that the diminished CCK
been shown to have elevations of CSF NPY (Kaye, response to a meal may play a role in diminished
Berrettini, et al., 1990), and while it does not nor- postingestive satiety observed in BN (LaChaussee,
malize with weight restoration in the short term, Kissileff, Walsh, & Hadigan, 1992). The CCK
it does appear to do so in the long term (Gendall, response in BN was found to return toward normal
Kaye, Altemus, McConaha, & La Via, 1999). More following treatment (Geracioti & Liddle, 1988).
recently, it has been reported that the plasma con- Studies of CCK in AN have yielded less con-
centration of NPY was lower in AN than in con- sistent findings. Some studies have found eleva-
trols, while BN participants had elevated NPY tions in basal levels of plasma CCK (Phillipp et al.,
levels (Baranowska, Radzikowska, Wasilewska-​ 1991; Tamai et al., 1993), as well as increased pep-
Dziubinska, Roguski, & Borowiec, 2000). Clearly, tide release following a test meal (Harty, Pearson,
elevated NPY does not result in increased feeding Solomon, & McGuigan, 1991; Phillipp et al., 1991)
in underweight AN; however, it is possible that or elevated during the ill state and normalizing with
increased NPY activity underlies the obsessive and recovery (Cuntz et  al., 2013). One study found
paradoxical interest in dietary intake and food prep- blunting of CCK response to an oral glucose load
aration. On the other hand, CSF levels of NPY and normalized in AN after partial restoration of body
PYY have been reported to be normal in women weight (Tamai et  al., 1993). Other studies have
with BN when measured while participants were found that measures of CCK function in AN were
acutely ill. Although levels of PYY increased above similar to or lower than control values (Baranowska
normal when participants were reassessed after et  al., 2000; Brambilla, Brunetta, Peirone, et  al.,
1 month of abstinence from binging and vomiting, 1995; Geracioti et  al., 1992; Pirke et  al., 1994).
levels of the peptides were similar to control values Further studies are needed to evaluate the relation-
in long-​term recovered individuals (Gendall et al., ship between altered CCK regulation and other
1999). Postprandial PYY response in BN has also indices of abnormal gastric function in symptom-
been shown to interact with a decreased response atic BN and AN (Geliebter et al., 1992).
of ghrelin (Kojima et al., 2005; Monteleone et al.,
2005). A  blunted response of ghrelin and PYY to Leptin
food ingestion suggests failure to attenuate the drive Leptin, the protein product of the ob gene, is
to eat with satiation, which may contribute to over- secreted predominantly by adipose tissue cells.
eating or binge eating behavior. In the hypothalamus, leptin interacts with NPY,
serotonin, and the melanocortins to decrease food
Cholecystokinin intake, thus regulating body fat stores (Friedman
Cholecystokinin (CCK) is a peptide secreted & Halaas, 1998; Zhang et  al., 1994; Zigman &
by the gastrointestinal system in response to food Elmquist, 2003). In rodent models, defects in the

 Ely, Kaye 51
leptin coding sequence resulting in leptin defi- As reviewed (Monteleone et al., 2004), patients
ciency, or impaired leptin receptor function, are with BN, in comparison to carefully matched con-
associated with obesity. In humans, serum and trols, have significantly decreased leptin concentra-
CSF concentrations of leptin are positively corre- tions in serum samples obtained after overnight
lated with fat mass across body mass index (BMI), fast (Baranowska et  al., 2001; Brewerton, Lesem,
including obesity (Considine et al., 1996; Schwartz, Kennedy, & Garvey, 2000; Frederich et  al., 2002;
Peskind, Raskind, Boyko, & Porte, 1996). While Jimerson, Mantzoros, Wolfe, & Metzger, 2000;
rare genetic deficiencies in leptin production have Monteleone, Di Lieto, Tortorella, Longobardi, &
been associated with familial obesity (Farooqi et al., Maj, 2000). Initial findings suggest that serum lep-
2001), obesity in humans is not thought to be a tin levels remain decreased in individuals who have
result of leptin deficiency per se. achieved sustained recovery from BN, compared
Underweight patients with AN have consistently to controls with matched percent body fat. This
been found to have significantly reduced serum lep- finding may be related to evidence for a persistent
tin concentrations in comparison to normal weight decrease in activity in the hypothalamic-​pituitary-​
controls (Baranowska, Wolinska-​Witort, Wasilewska-​ thyroid axis in long-​term recovered BN individuals
Dziubinska, Roguski, & Chmielowska, 2001; Eckert (Wolfe et al., 2000). These alterations could further
et  al., 1998; Grinspoon et  al., 1996; Hebebrand be associated with decreased metabolic rate and a
et al., 1995; Mantzoros, Flier, Lesem, Brewerton, & propensity toward weight gain, contributing to pre-
Jimerson, 1997; Monteleone, DiLieto, Castaldo, & occupation with body weight.
Maj, 2004; Tortorella et  al., 2014). Based on stud-
ies in laboratory animals, it has been suggested that Ghrelin
low leptin levels may contribute to amenorrhea and Ghrelin is a gut-​related peptide that strongly
other hormonal changes in the disorder (Ahima & stimulates increased weight gain and feeding in
Osei, 2004; Ahima, Saper, Flier, & Elmquist, 2000; rats when injected intracerebroventricularly. When
Holtkamp et  al., 2003; Mantzoros et  al., 1997). administered to healthy human volunteers, ghrelin
In healthy volunteers, even modest reductions in results in increased hunger and food intake (Wren
energy intake result in substantial decreases in cir- et al., 2001). In addition, it has been reported that
culating leptin levels (Wolfe, Jimerson, Orlova, & fasting plasma ghrelin concentrations in humans are
Mantzoros, 2004). Although the reduction in fasting negatively correlated with BMI (Shiiya et al., 2002;
serum leptin levels in AN is correlated with reduc- Tanaka et al., 2002), percentage body fat, and fast-
tion in body mass index, there has been some discus- ing leptin and insulin concentrations (Tschop
sion of the possibility that leptin levels in AN may be et  al., 2001). A  number of studies (Jimerson &
higher than expected based on the extent of weight Wolfe, 2006; Monteleone, Serritella, Martiadis,
loss (Frederich, Hu, Raymond, & Pomeroy, 2002; Scognamiglio, & Maj, 2008; Sedlackova et al., 2011)
Jimerson, 2002). Mantzoros et  al. (1997) reported have shown elevation in circulating ghrelin levels in
elevated ratio of CSF to serum leptin in AN com- AN, with a return to normal levels as patients regain
pared to controls, suggesting that the proportional weight (Nakahara et  al., 2007). However, some
decrease in leptin levels with weight loss is greater argue that ghrelin is unlikely to contribute to the
in serum than in CSF. A  longitudinal investigation array of AN symptoms (Stock et al., 2005). Further
during refeeding in AN patients has shown that CSF research is needed to explore the possible existence
leptin concentrations will reach normal values before of ghrelin resistance in cachectic states related to the
full weight restoration, possibly as a consequence of eating disorders.
the relatively rapid and disproportionate accumu- Studies comparing fasting plasma ghrelin con-
lation of fat during weight restoration (Mantzoros centrations in patients with BN and healthy con-
et  al., 1997). This premature normalization of lep- trols have yielded variable results (Kojima et  al.,
tin concentration could contribute to difficulty in 2005; Monteleone et al., 2005; Monteleone et al.,
achieving and sustaining a normal weight in anorexia 2003). Ghrelin may also differentiate between sub-
nervosa. Further research is needed to assess whether types of BN (Germain et  al., 2010), as the above
serum leptin levels at the time of discharge are predic- studies did not separate purging and nonpurging
tive of posthospital clinical course (Holtkamp et al., variants of the disorder, but this is still in conten-
2004; Lob et al., 2003). Plasma and CSF leptin levels tion (Troisi et al., 2005). It is notable, however, that
appear to be similar to control values in long-​term the postprandial decrease in ghrelin levels appears
recovered AN participants (Gendall et al., 1999). to be blunted in patients with BN (Kojima et  al.,

52 Appetitive Regul ation

2005; Monteleone et  al., 2005), consistent with Monoamine Systems
extant research demonstrating diminished satiety There is an abundance of evidence that indi-
responses in the disorder. viduals with AN and BN have disturbances of
monoamine function in the ill state. While less well
Summary of CNS Neuropeptide Alterations studied, monoamine disturbances appear to persist
Individuals with AN tend to have hypersecre- after recovery. Also, PET imaging with selective
tion of CRH and exaggerated HPA function, lower neurotransmitter radioligands has resulted in a
NPY plasma concentration, lower CSF leptin con- technology permitting new insights into regional
centration, and elevated ghrelin levels compared to binding and specificity of 5-​HT and dopamine
healthy controls. However, there has been less agree- neurotransmission in vivo in humans and their
ment in the literature concerning other peptides, relationship to behaviors. Studying neuroreceptor
including opioid, PYY, and CCK expression in AN. networks is of particular importance since results
In contrast, BN has been consistently described by may contribute to developing new pharmaceutical
elevated NPY levels, decreased PPY and CCK fol- interventions.
lowing a meal, and decreased serum leptin levels.
Low CCK release recovers after treatment for BN, Dopamine
but serum leptin levels appears to remain low even The role of dopamine (DA) as a reward pre-
after recovery. Like AN, the literature is unclear as diction signal is well established (Schultz, 1998;
to alterations of CSF beta-​endorphin and related Schultz, Tremblay, & Hollerman, 1998), but rela-
opiate concentrations in BN. A recent meta-​analysis tively little work has been done assessing neuro-
summarized baseline and postprandial levels of PYY, chemical alterations involved in feeding. One study,
ghrelin, and CCK in AN and BN (Prince, Brooks, using PET and the DA D2/​D3 radioligand [11C]
Stahl, & Treasure, 2009), and these alterations are raclopride, found lower [11C]raclopride binding in
described in Table 4.2. the dorsal putamen and caudate nucleus in the sated
It is likely that many of the starvation-​driven state compared to hungry, which may mean that
endocrine and metabolic change are compensa- DA is released during or after the feeding process.
tory and attempt to conserve energy or stimulate In addition, the level meal pleasantness experienced
hunger and feeding (orexigenic) (Schwartz et  al., correlated negatively with [11C]raclopride bind-
2000). For example, AN patients have increased ing, suggesting a positive correlation of pleasant-
orexigenic signals from neuropeptides such as NPY ness rating with DA surge (Small, Jones-​Gotman,
and leptin (see (Inui, 2001). However, starvation & Dagher, 2003). This is surprising, since in many
in AN patients also stimulates feeding-​inhibitory studies pleasantness or hedonic experiences were
(anorexigenic) signaling by increasing levels of in relation to DA activity in the ventral striatum
anorexigenic peptides, such as CRH, CCK, and including the nucleus accumbens.
pancreatic polypeptide, which may contribute
to the perpetuation of symptoms. It is important Dopamine Receptor Binding is Altered
to note that alterations of these and other neuro- in Eating Disorders
peptide systems affected in AN patients, such as The DA system may be of special interest to ED
beta-​endorphin, are also likely to result in altered given its role in motivation, reward, preferences, and
mood, cognitive function, impulse control, and reinforcement (Cannon & Bseikri, 2004). Several
autonomic and hormonal systems (Jimerson & lines of evidence suggest AN individuals have altered
Wolfe, 2006), raising the likelihood they contrib- DA metabolism, including reduced cerebrospinal
ute to the behavioral symptoms associated with the fluid DA metabolites in ill and recovered (REC) AN
ill state. For example, intracerebroventricular CRH (Kaye, Ebert, Raleigh, & Lake, 1984; Kaye, Frank, &
administration in experimental animals produces McConaha, 1999), altered frequency of functional
many of the physiological and behavioral changes polymorphisms of DA D2 receptor genes (Bergen
associated with AN, including hypothalamic hypo- et  al., 2005) including the A1 allele implicated in
gonadism, altered emotionality, decreased sexual hedonic value of rewards (Cohen, Young, Baek,
activity, hyperactivity, and decreased feeding behav- Kessler, & Ranganath, 2005), and feeding (Nisoli
ior (Kaye, Gwirtsman, et al., 1987). Thus, it can be et al., 2007). Impaired visual discrimination learn-
argued that some secondary peptide changes sustain ing (Lawrence, 2003), a task thought to reflect DA
AN behaviors by driving a desire for more dieting signaling function, has also been found in AN, as
and weight loss. well as a generalized failure to activate the appetitive

 Ely, Kaye 53
Table 4.2  Summary of Neuroendocrine/​Neuropeptide Alterations in AN and BN. Increased Compared to Controls
(Up Arrow); Decreased Compared to Controls (Down Arrow); no Difference from Controls (=)
Measurement References (First Author, Year) Group Summary

Cortisol (Monteleone, 2016; Monteleone, 2011)—​CAR Ill AN ↑

(Gold, 1986; Kaye, 1987; Licinio, 1996; Walsh, 1987)—​CRH

(Gwirtsman, 1989)—​ACTH Ill BN =

REC BN ↓

Opioid (Brambilla, 1995) Ill AN ↑

(Lesem, 1991) =

(Kaye, 1987) ↓

(Brewerton, 1992) Ill BN ↓

NPY (Sedlackova, 2011; Kaye, 1990; Baranowska, 2000) Ill AN ↑

(Baranowska, 2000) ↓

(Sedlackova, 2012; Baranowska, 2000) Ill BN ↑

PYY, basal (Misra, 2006, Nakahara, 2007) Ill AN ↑

(Germain, 2007; Germain, 2010) ↓

(Baranowska, 2000; Kaye, 1990; Otto, 2007) =

(Kojima, 2005; Monteleone, 2005) Ill BN =

PYY, postprandial (Nakahara, 2007) Ill AN ↑

(Otto, 2007, Stock, 2005) =

(Kokima, 2005; Monteleone, 2005) Ill BN ↓

CCK, basal (Cuntz, 2013; Phillipp, 1991; Tamai, 1993) Ill AN ↑

(Geracioti, 1992, Pirke, 1994) Ill AN =

(Cuntz, 2013; Geracioti, 1992, Pirke, 1994) REC AN =

(Brambilla, 1995; Lydiard, 1993; Pirke, 1994) Ill BN ↓

(Hannon-​Engel, 2013; Phillipp, 1991) =

(Geracioti, 1988) REC BN =

CCK, (Harty, 1991; Phillipp, 1991) Ill AN ↑

post-​prandial

(Geracioti, 1992, Pirke, 1994) =

(Baranowska, 2000; Brambilla, 1995) ↓

(Geracioti, 1992, Pirke, 1994) REC AN =

(Devlin, 1997; Geracioti, 1988; Keel, 2007; Phillipp, 1991; Pirke, Ill BN ↓
1994)

Leptin (Baranowska, 2001; Eckert, 1998; Grinspoon, 1996; Hebebrand, Ill AN ↓

1995; Mantzoros, 1997; Monteleone, 2004; Monteleone, 2008)

54 Appetitive Regul ation

Table 4.2 Continued
Measurement References (First Author, Year) Group Summary

(Hebebrand, 1995) AN, weight ↑

(Eckert, 1998) restored =

(Gendall, 1999) REC AN =

(Baranowska, 2001; Brewerton, 2000; Frederich, 2002; Ill BN ↓

Jimerson, 2000; Monteleone, 2000; Wolfe, 2000)

(Wolfe, 2000) REC BN ↓

Ghrelin (Monteleone, 2008; Nakahara, 2007; Sedlackova, 2011) Ill AN ↑

(Stock, 2005) =

(Kojima, 2005) Ill BN ↑

(Monteleone, 2005) =

(Monteleone, 2003) ↓

motivational system in a startle task (Friederich Though ill BN participants do not appear to be
et al., 2006) and impaired set shifting (Tchanturia associated with abnormal CSF HVA, several studies
et al., 2004), which, in part, is related to ventral stri- have shown a significant reduction of CSF HVA in
atal DA function (Goto & Grace, 2005). Further, BN patients with high binge frequency (Jimerson,
REC AN had increased binding of D2/​D3 receptors Lesem, Kaye, & Brewerton, 1992; Kaye, Ballenger,
in the anterior ventral striatum (Frank et al., 2005), et  al., 1990). Dopamine metabolism in the CNS
a region that contributes to optimal responses to may explain differences in symptoms between AN,
reward stimuli (Delgado, Nystrom, Fissel, Noll, & AN-​BN, and BN participants. Our group found
Fiez, 2000; Montague, Hyman, & Cohen, 2004; (Kaye et al., 1999) that REC AN participants had
Schultz, 2004). In addition, REC AN showed significantly reduced concentrations of CSF HVA,
positive correlations between DA D2/​D3 binding compared to REC BN women or those with histo-
in the dorsal caudate/​dorsal putamen and anxiety ries of both disorders. Dopamine neuronal function
measures. In summary, striatal DA dysfunction in has been associated with motor activity (Kaye et al.,
AN could contribute to altered reward and affect, 1999), and reward (Blum et  al., 1995; Salamone,
decision-​making, and executive control, as well as 1996). Individuals with AN have stereotyped and
stereotypic motor activity (Yin & Knowlton, 2006) hyperactive motor behavior, anhedonic and restric-
and decreased food ingestion (Halford, Cooper, & tive personalities, and reduced novelty seeking.
Dovey, 2004). Recent studies (Bailer et  al., 2013;
Broft et al., 2012; Frank et al., 2005) suggest that Serotonin
alterations in striatal DA function may contrib- The serotonin (5-​ HT) autoreceptor 5-​ HT1A is
ute to AN behaviors. AN shows increased DA located presynaptically on somatodendritic cell bod-
binding to D2/​D3 receptors in the striatum, and ies in the raphe nucleus, where it functions to decrease
[11C]raclopride, a selective D2 antagonist, demon- 5-​HT neurotransmission (Staley, Malison, & Innis,
strated binding potential in caudate associated with 1998). High densities of postsynaptic 5-​HT1A exist
harm avoidance and trait anxiety. A  recent multi- in the hippocampus, septum, amygdala, and entorhi-
modal study (Bailer et al., 2016) shows that BOLD nal and frontal cortex, where they serve to mediate the
response in the caudate in a monetary choice task effects of released 5-​HT. Postsynaptic 5-​HT2A recep-
is associated with greater D2/​D3 binding, as well tors, which are in high densities in the cerebral cortex
as greater anxiety and harm avoidance in REC AN. and other regions of rodents and humans (Burnet,
This suggests that D2/​D3 binding in AN may be Eastwood, & Harrison, 1997; Saudou & Hen, 1994),
linked with an exaggerated cognitive response to are of interest as they have been implicated in the
feedback and anxious avoidant temperament. modulation of feeding and mood, as well as selective

 Ely, Kaye 55
serotonin-​reuptake inhibitor (SSRI) response (Bailer 5-​HT2A receptors BP values in one study (Bailer
et al., 2004; Bonhomme & Esposito, 1998; De Vry et  al., 2007) and reduced BP in another study
& Schreiber, 2000; Simansky, 1996; Stockmeier, (Audenaert et  al., 2003) in the left frontal, bilat-
1997). Serotonin pathways also play an important eral parietal, and occipital cortex. After recovery,
role in postprandial satiety. Treatments that increase restricting-​type AN individuals (Frank et al., 2002)
intrasynaptic 5-​HT, or directly activate 5-​HT recep- had reduced 5-​HT2A receptors BP in mesial tempo-
tors, tend to reduce food consumption, whereas ral and parietal cortical areas as well as in subgenual
interventions that dampen 5-​HT neurotransmission and pregenual ACC. Similarly, REC bulimic-​type
or block receptor activation reportedly increase food AN (Bailer et  al., 2004) women had reduced 5-​
consumption and promote weight gain (Blundell, HT2A receptor BP in left subgenual ACC, left pari-
1984; Leibowitz & Shor-​Posner, 1986). Moreover, etal, and right occipital cortex. Additionally, REC
CNS 5-​HT pathways have been implicated in the BN women only had reduced [18F]altanserin BP rel-
modulation of mood, impulse regulation and behav- ative to controls in the orbital frontal region (Kaye
ioral constraint, and obsessionality, and they affect et  al., 2001). Serotonin transporter function has
a variety of neuroendocrine systems. Serotonin has recently been shown (Bailer et al., 2013) to interact
been postulated to contribute to temperament or with striatal DA D2/​D3 receptor binding in a sam-
personality traits such as harm avoidance (Cloninger, ple of REC AN and BN, while control participants
1987) or behavioral inhibition (Soubrie, 1986) or did not show the same relationship.
to categorical dimensions such as obsessive com- Studies have also found that AN and BN have
pulsive disorder (Barr, Goodman, Price, McDougle, alterations in 5-​ HT metabolism. When under-
& Charney, 1992), anxiety and fear (Charney, weight, individuals with AN have a significant
Woods, Krystal, & Heninger, 1990), or depression reduction in basal concentrations of the serotonin
(Grahame-​Smith, 1992), as well as satiety for food metabolite 5-​hydroxyindolacetic acid (5-​HIAA) in
consumption. However, the contribution of partic- the CSF compared to healthy controls, as well as
ular components of 5-​HT neuronal systems (i.e., blunted plasma prolactin response to drugs with 5-​
different pathways or receptors) to specific human HT activity and reduced 3H-​imipramine binding.
behaviors remains uncertain. Together, these findings suggest reduced seroto-
nergic activity, although this may arise secondarily
Serotonin Receptor Binding Is Altered from reductions in dietary supplies of the 5-​HT
in Eating Disorders synthesizing amino acid tryptophan. In contrast,
There has been considerable interest in the role CSF concentrations of 5-​HIAA are reported to be
that 5-​HT may play in AN and BN (Brewerton, elevated in long-​term weight-​recovered AN indi-
1995; Jimerson, Lesem, Kaye, Hegg, & Brewerton, viduals. While opposite patterns of serotonergic
1990; Kaye, Gendall, & Strober, 1998; Kaye, activity in acutely ill and long-​term REC AN indi-
Gwirtsman, George, & Ebert, 1991; Steiger et al., viduals may seem counterintuitive, dieting behavior
2004; Treasure & Campbell, 1994). Bailer (Bailer has been shown to lower plasma tryptophan levels
et  al., 2007) reported that ill AN individuals had in otherwise healthy women (Anderson, Parry-​
a 50% to 70% increase in 5-​HT1A receptor bind- Billings, Newsholme, Fairburn, & Cowen, 1990).
ing potential (BP) in subgenual, mesial temporal, Therefore, resumption of normal eating in individ-
orbital frontal, and raphe brain regions, as well as in uals with AN may unmask intrinsic abnormalities
prefrontal, lateral temporal, anterior cingulate cor- in serotonergic systems that mediate behavioral or
tex (ACC), and parietal regions. Increased 5-​HT1A temperamental risk factors that confer vulnerability
postsynaptic activity has also been reported in ill BN to developing AN in the first place.
participants (Tiihonen et  al., 2004). In REC AN Considerable evidence also exists for a dysregu-
binge-​purge type and REC BN (Bailer et al., 2005; lation of serotonergic processes in BN. Examples
Kaye, unpublished data), participants had a signif- include blunted prolactin response to the 5-​ HT
icant 20% to 40% increase in 5-​HT1A receptor BP receptor agonists m-​ chlorophenylpiperazine (m-​
in these same regions compared to control women CPP), 5-​hydroxytrytophan, and DL-​fenfluramine,
(CW) (Bailer et al., 2005). In contrast, REC RAN and enhanced migraine-​like headache response to
women showed no difference in 5-​HT1A receptor m-​CPP challenge. Acute perturbation of serotoner-
BP compared to controls (Bailer et al., 2005). gic tone by dietary depletion of tryptophan has also
With regard to 5-​ HT2A receptor binding, ill been linked to increased food intake and mood irri-
AN participants have been found to have normal tability in individuals with BN compared to healthy

56 Appetitive Regul ation

controls. And, like AN, women with long-​ term D3 receptors in mesial temporal and other lim-
recovery from BN have been shown to have elevated bic regions. Bailer (Bailer et  al., 2004) found that
concentrations of 5-​HIAA in the CSF as well as REC AN-​BN participants showed a positive rela-
increased platelet binding of paroxetine. tionship between [18F]altanserin BP in the left sub-
genual ACC and mesial temporal cortex and harm
Correlations of PET Data with Anxiety and avoidance. For ill AN participants, [18F]altanserin
Harm Avoidance BP was positively related to harm avoidance in the
Clinical and epidemiological studies have con- suprapragenual ACC, frontal, and parietal regions.
sistently shown that one or more anxiety disorders The 5-​HT2A receptor binding and harm avoidance
occur in the majority of people with AN or BN were shown to be negatively correlated in the fron-
(Godart, Flament, Perdereau, & Jeammet, 2002; tal cortex in healthy participants (Moresco et  al.,
Kaye et al., 2004; Kendler et al., 1995; Walters & 2002) and in the prefrontal cortex in patients that
Kendler, 1995). Silberg and Bulik (2005), using attempted suicide (van Heeringen et  al., 2003).
twins, found a unique genetic effect that influences Recent research (Bailer et al., 2013) in a combined
liability to early anxiety and ED symptoms. When a sample of REC AN and BN has shown an inter-
lifetime anxiety disorder is present, the anxiety most action of serotonin transporter function and DA
commonly occurs first in childhood, preceding the receptor binding that significantly predicted harm
onset of AN or BN (Bulik, Sullivan, Fear, & Joyce, avoidance, supporting the possibility that 5-​ HT
1997; Deep, Nagy, Weltzin, Rao, & Kaye, 1995; and DA interactions contribute to HA behaviors
Godart, Flament, Lecrubier, & Jeammet, 2000). in EDs.
Anxiety and harm avoidance remain elevated after There is extensive literature associating the
recovery from AN, AN-​ BN, and BN (Wagner, serotonergic systems and fundamental aspects of
Barbarich, et  al., 2006), even if individuals never behavioral inhibition (Geyer, 1996; Soubrie, 1986).
had a lifetime anxiety disorder diagnosis (Kaye Reduced CSF 5-​HIAA levels are associated with
et al., 2004). Finally, anxiety (Spielberger, Gorsuch, increased impulsivity and aggression in humans
& Lushene, 1970) and Harm Avoidance from the and nonhuman primates, whereas increased CSF
Cloninger Temperament and Character Inventory 5-​HIAA levels are related to behavioral inhibition
(TCI) (Cloninger, Przybeck, Svrakic, & Wetzel, (Fairbanks, Melega, Jorgensen, Kaplan, & McGuire,
1994) have been a robust signal in genetic studies 2001; Westergaard et al., 2003). Thus, it is of inter-
(Bacanu et al., 2005). est that recovered AN and BN women had elevated
Studies in animals and humans implicate the 5-​ CSF 5-​HIAA concentrations. Behaviors found after
HT1A receptor in anxiety (File, Kenny, & Cheeta, recovery from AN and BN, such as anxiety and
2000; Gross et  al., 2002) and depression and/​or perfectionism, tend to be consistent with behav-
suicide (Mann, 1999). Reduced 5-​ HT1A recep- iors displayed by people with high 5-​HIAA levels.
tor BP has been found in ill (Drevets et al., 1999; Reduced CSF 5-​HIAA levels appear to be related to
Sargent et al., 2000) and REC (Bhagwagar, Rabiner, behavioral undercontrol, whereas increased CSF 5-​
Sargent, Grasby, & Cowen, 2004) depressed partici- HIAA concentrations may be related to behavioral
pants, as well as in a primate model for depression overcontrol. Together, these studies contribute to
(Shively et  al., 2006). Parsey (Parsey et  al., 2005) a growing literature suggesting that CSF 5-​HIAA
found no difference in carbonyl-​[11C]WAY100635 concentrations may correlate with a spectrum of
BP in major depressive disorder, although a sub- behavior.
group of never medicated participants had elevated The possibility of serotonergic dysfunction as a
carbonyl-​[11C]WAY100635 BP. Recent studies have common risk factor for BN and AN may seem puz-
found reduced [11C]WAY100635 BP in social pho- zling given well-​recognized differences in behavior
bia (Lanzenberger et al., 2007) and panic disorder in these disorders. However, both disorders respond
(Neumeister et  al., 2004). These findings suggest to 5-​HT-​specific medications, and both disorders
ED and affective disorders share disturbances of have high levels of harm avoidance (see Klump,
common neuronal pathways but are etiologically Bulik, et al., 2000), a personality trait hypothesized
different. to be related to increased 5-​HT activity. Further,
The PET imaging studies in ill and REC AN extant research suggests a familial aggregation of a
and BN participants described above have found range of EDs in relatives of probands with either BN
significant correlations between harm avoidance or AN, and these two disorders are highly comor-
and binding for the 5-​HT1A, 5-​HT2A, and DA D2/​ bid in twin studies. These data raise the possibility

 Ely, Kaye 57
that a disturbance of 5-​HT activity may be a shared Brain Imaging Studies
etiologic vulnerability for the expression of a clus- Brain imaging studies in AN and BN can be
ter of symptoms that are common to both AN and divided in several categories. First, there has been
BN. The development of ED subgroups may be due a substantial literature using computerized tomog-
to factors independent of this shared vulnerabil- raphy, and more recently MRI, that seeks to deter-
ity. For example, the risk for obsessive-​compulsive mine whether structural alterations in the brain
personality disorder is elevated only in restricting-​ are evident in individuals with ED. Second, more
type AN and in their families, and shows a shared recent studies have used functional magnetic reso-
transmission with restrictor-​ type AN (Lilenfeld nance imaging (fMRI) or other technologies to
et al., 1998). This suggests that an additional vul- assess blood flow responses to some stimuli, such
nerability for behavioral overcontrol and rigid and as pictures of food or tastes of food. Third are imag-
inflexible mood states, combined with vulnerability ing studies, such as PET, that employ a radioligand.
for an ED, may result in restricting AN rather than These studies, which may use the glucose analog
another ED. fluorodeoxyglucose (FDG) to study glucose metab-
The premorbid onset and the persistence of anxi- olism, or a ligand that is specific for a serotonin
ety and harm avoidance symptoms after recovery receptor (such as the DA and 5-​HT receptors dis-
suggest these are traits that contribute to the patho- cussed above), provide information that is specific
genesis of AN and BN. The PET imaging data sug- for the system being studied.
gest that such behaviors are related to disturbances of In general, findings from functional and radioli-
5-​HT and DA neurotransmitter function in limbic gand studies have been relatively consistent in that
and executive pathways. This technology holds the most studies have positive findings within frontal,
promise of greater understanding of the complex- cingulate, temporal, and/​or parietal regions. Thus,
ity of neuronal systems in human behavior. For it can be stated that ED individuals, both when ill
example, postsynaptic 5-​ HT1A receptors (Celada, and after recovery, have alterations in brain activity
Puig, Casanovas, Guillazo, & Artigas, 2001; Richer, compared to matched controls. While the literature
Hen, & Blier, 2002; Sibille, Pavlides, Benke, & Toth, is still in its infancy, studies have begun to identify
2000; Szabo & Blier, 2001) have “downstream” regions that may be particularly impacted, though
effects and interactions with other neuronal systems, more sophisticated research in circuitry and behav-
such as norepinephrine, glutamate, and gamma-​ ioral correlates has yet to be conclusive. Sample
aminobutyric acid. Enhanced 5-​ HT1A activity in sizes have been small, and imaging technologies
AN and BN may cause or reflect an altered balance and methods vary widely. Studies to date indicate
between these neuronal systems. Moreover, 5-​HT1A gross alterations of brain function. Because brain
receptors interact with other 5-​HT receptors such pathways are highly complex, however, the neu-
as 5-​HT2A (Martin, Kaplan, & Weir, 1997; Szabo roanatomy of AN and BN have only begun to be
& Blier 2001). The 5-​HT1A post-​synaptic receptors characterized.
mediate locus coeruleus firing through 5-​HT trans- It should be noted that there has been substan-
mission at 5-​HT2A receptors (Szabo et  al., 2001). tial progress in understanding how brain cortical
Theoretically, increased 5-​HT1A and reduced 5-​HT2A regions modulate higher order functions related to
postsynaptic receptor activity in AN could result in appetitive behaviors in healthy humans. Thus, this
increased noradrenergic neuron firing (Szabo et al., section begins first with a review of the neurocir-
2001). Moreover, postsynaptic 5-​ HT1A receptors cuitry involved in appetite regulation, then with an
hyperpolarize and 5-​HT2A receptors depolarize layer overview of fMRI studies of taste in healthy con-
V pyramidal neurons (Martin-​Ruiz et  al., 2001). trols, as they provide a potential baseline that can be
In AN, synergistic effects of these receptors, which used to determine whether individuals with AN and
are colocalized on pyramidal neurons, may reduce BN have some alteration in brain pathways devoted
pyramidal neuronal excitability. to the modulation of feeding.
Taken together, these PET-​ radioligand stud-
ies confirm that altered 5-​HT neuronal pathway Neurocircuitry of Appetite Regulation
activity persists after recovery from AN and BN Many studies using fMRI to test appetitive regu-
and support the possibility that these psychobio- lation, including those in REC EDs from our group
logical alterations might contribute to traits such as and others (Frank et al., 2008; Frank et al., 2006;
increased anxiety, which may contribute to a vul- Wagner, Aizenstein, et  al., 2006; Wagner et  al.,
nerability to develop an eating disorder. 2007), use a pump apparatus to deliver repeat, blind

58 Appetitive Regul ation

gustatory stimuli, such as sucrose (Frank et  al., decisions. In animals, DA modulates the influ-
2003). Although a gustatory task employing sucrose ence of limbic inputs on striatal activity (Goto &
stimuli does not test the complexity of food choices Grace, 2005; Montague et al., 2004; Schultz, 2004;
(Small, 2006), it does activate appetitive pathways Yin & Knowlton, 2006) and mediates the “bind-
of interest. ing” of hedonic evaluation of stimuli to objects or
Sweet taste perception is peripherally medi- acts (“wanting” response) (Berridge & Robinson,
ated by tongue receptors (Chandraskekar, Hoon, 1998). It has been postulated that dorsal striatum is
Ryba, & Zuker, 2006) through cranial nerves, the engaged by real or perceived stimulus-​response out-
nucleus tractus solitarius, and thalamic ventropos- comes with DA projections modulating this behav-
terior medial nucleus, to the primary gustatory ior (O’Doherty et  al., 2004; Tricomi, Delgado, &
cortex, that in humans comprise the frontal oper- Fiez, 2004).
culum and the anterior insula (AI) (Faurion et al., The AI and associated gustatory cortex respond
1999; Ogawa, 1994; Schoenfeld et al., 2004; Scott, not only to the taste and physical properties of food
Yaxley, Sienkiewicz, & Rolls, 1986; Yaxley, Rolls, & but also to its rewarding properties (O’Doherty,
Sienkiewicz, 1990). The AI is contiguous with the Kringelbach, Rolls, Hornak, & Andrews, 2001;
posterior orbitofrontal cortex (OFC) at the oper- Schultz, Tremblay, & Hollerman, 2000; Small,
culum and has bidirectional connections to the 2002; Small, Zatorre, Dagher, Evans, & Jones-​
amygdala, nucleus accumbens (Reynolds & Zahm, Gotman, 2001). Some studies argue that the AI
2005), OFC (Ongur & Price, 2000), medial pre- provide a representation of food in the mouth that
frontal cortex (mPFC) and ACC (Carmichael & is independent of hunger, and thus of reward value
Price, 1996). The AI is centrally placed to receive (Rolls, 2005), whereas the OFC computes the
information about the salience (both appetitive and hedonic value of food (Kringelbach, O’Doherty,
aversive) and relative value of the stimulus environ- Rolls, & Andrews, 2003; O’Doherty et  al., 2000;
ment and integrate this information with the effect Rolls, 2005). Other studies (Small et  al., 2001)
that these stimuli may have on the body. Projections suggest that the AI and OFC have overlapping
from the primary taste cortex reach the central representations of sensory and reward/​ affective
nucleus of the amygdala and from there, the lateral processing of taste. The OFC is associated with
hypothalamus and midbrain dopaminergic regions flexible responses to changing stimuli (Izquierdo,
(Simon, De Araujo, Gutierrez, & Nicolelis, 2006). Cammarota, Medina, & Bevilaqua, 2004; Kazama
The primary taste cortex also projects heavily to the & Bachevalier, 2006) such as the incentive value,
striatum (Kelley, 2004), which receives inputs from for example, whether one is hungry (Critchley &
brain regions involved in reward, incentive learning, Rolls, 1996; Gottfried, O’Doherty, & Dolan, 2003;
and emotional regulation, including the ACC, the Hikosaka & Watanabe, 2000). Of note, the OFC is
ventromedial PFC, the OFC, and AI (Carmichael highly dependent on 5-​HT innervation for flexible
& Price, 1996; Chikama, McFarland, Amaral, & reversal learning (Clarke, Walker, Dalley, Robbins,
Haber, 1997; Fudge, Breibart, Danish, & Pannoni, & Roberts, 2007) so that 5-​HT abnormalities in
2005; Fudge, Breitbart, & McClain, 2004; Haber, ED may contribute to the disturbed inhibitory
Kunishio, Mizobuhi, & Lynd-​Balta, 1995; Haber, control (inability to incorporate changing incentive
Kim, Mailly, & Calzavara, 2006). Human neuro- value of stimuli).
imaging studies show that a highly interconnected Interoceptive processing is also fundamentally
network of brain areas including OFC, mPFC, important for appetite regulation. Information
amygdala, striatum and DA midbrain is involved in about interoceptive state is relayed from the AI to
reward processing of both primary (i.e., pleasura- the ACC, which, as part of the central executive
ble tastes) (Berns, McClure, Pagnoni, & Montague, system, can generate an error signal that is criti-
2001; McClure, Berns, & Montague, 2003) cal for conflict monitoring and the allocation of
and secondary (i.e., money) reinforcers (Breiter, attentional resources (Carter, Botvinick, & Cohan,
Aharon, Kahneman, Dale, & Shizgal, 2001; 1999). Thus, interoception involves monitoring
Delgado et al., 2000; Gehring & Willoughby, 2002; the sensations that are important for the integrity
Montague et  al., 2004; O’Doherty, 2004). These of the internal body state and connecting to sys-
regions code stimulus-​reward value, maintain rep- tems that are important for allocating attention,
resentations of predicted future reward and future evaluating context, and planning actions (Paulus
behavioral choice, and may play a role in inte- & Stein, 2006). In addition, interoceptive sensa-
grating and evaluating reward prediction to guide tions are often associated with intense affective and

 Ely, Kaye 59
motivational components (Paulus & Stein, 2006), cognitive reflection of the taste experience, perhaps
and the evaluative component of the signal is highly on a tertiary level. In a review of their own and
dependent on the homeostatic state of the individ- studies of other investigators, this group proposes
ual. These regions therefore play an important and that OFC activity may be separated into anterior-​
interconnected role in determining homeostatic posterior and medial-​ lateral compartments with
appetitive needs when hungry or satiated. task-​specific responsiveness (Kringelbach et  al.,
2004). With the notion that the OFC is involved
Brain Imaging Studies of Normal Feeding in the evaluation of the reward value that a stimu-
Behavior in Healthy Individuals lus has, medial areas may then respond to reinforc-
Noninvasive brain imaging tools have stimu- ing stimuli and lateral areas to aversive stimuli. This
lated new insights into how cortical brain regions may also be the case for olfactory stimuli (Gottfried,
are involved in the regulation of food intake in O’Doherty, & Dolan, 2002). Anterior areas are
humans and primates. We review recent literature hypothesized to be more involved in abstract stim-
on feeding-​related physiology in healthy controls uli such as monetary reward, as opposed to more
followed by a discussion of brain imaging studies primitive experiences including taste or pain.
in AN and BN. This system of appetite and hunger, Another recent study commented on the interac-
food appetence, ingestion of food, and subsequent tion between the OFC and amygdala (Arana et al.,
subjective experience, is very complex, and study 2003). Using images of menus of varying incentive
methodologies and results are not homogeneous. values, it was found that the amygdala responded
However, those studies may help us guide research relative to the appeal of different aspects presented
on the pathophysiology of EDs and delineate bio- on the menu, whereas the medial OFC response
logical traits. Little work has been done in under- was activated when having to make choices between
standing whether individuals with AN or BN have menus in relation to individual difficulty during
alterations of feeding-​related brain pathways and that task. The lateral OFC was activated when the
how phenomena, such as food-​related anxiety, play preferred menu could not be chosen, and this could
a role in the neurophysiology of EDs. be consistent with some form of aversive experience
Recent fMRI studies have explored the neural as suggested above.
response to taste and smell. Such studies in con- Studies in healthy individuals suggest testa-
trol individuals may help design similar taste and ble hypotheses for understanding AN and BN.
smell studies in the ED population (Cerf-​Ducastel Individuals with restricting type AN may have
& Murphy, 2004; De Araujo, Rolls, Kringelbach, higher lateral OFC activation to aversive and per-
McGlone, & Phillips, 2003). Research from haps anxiety-​provoking food stimuli such as fat
Edmund Rolls and colleagues has provided new (Drewnowski, Halmi, Pierce, Gibbs, & Smith,
understanding of how the brain processes taste. As 1987), but higher medial OFC activity in response
described previously, the insula cortex shows the to exercise, weight loss–​promoting activity, or
primary response to taste recognition, and the OFC particularly “safe” foods. We also propose early
shows secondary response to various taste stimuli. sensory specific satiety in AN, such as acceler-
For example, pleasant gustatory stimulation acti- ated reduction of OFC activity to food stimuli,
vates the OFC. When the same taste stimulus is which would be consistent with early meal ter-
presented repeatedly, resulting in sensory specific mination. In contrast, BN individuals may have
satiety, this activation declines. A  recent study delayed medial OFC activation reduction (e.g.,
from Stice and colleagues (Stice, Burger, & Yokum, delayed sensory-​specific satiety) for initially lower
2013) has shown that activation in response to food habituation, reflecting binge eating vulnerability.
stimuli in the OFC, in addition to the striatum and Since BN individuals may like sweeter stimuli
anterior cingulate cortex, showed a positive corre- than controls (Drewnowski, Bellisle, Aimez, &
lation with the length of time one has been calori- Remy, 1987), they may have increased medial
cally deprived, suggesting that longer-​term negative OFC response but may also have increased ACC
energy balance increases sensitivity to the reward or amygdala activation reflecting heightened anx-
value of food. iety after binge eating episodes. It is important to
The dorsolateral prefrontal cortex (DLPFC) note that humans have substantial variability of
has also been implicated in gustatory processing brain response to taste stimuli, which may limit
(Kringelbach, de Araujo, & Rolls, 2004). This area the power and interpretability of study results
is a cognitive processing center and may suggest (Schoenfeld et al., 2004).

60 Appetitive Regul ation

 Recent brain imaging studies in controls using the incentive and emotional value of taste stimulus.
pictures of food and similar food-​related stimuli A  recent study (Haase, Cerf-​Ducastel, & Murphy,
confirmed that pictures of food activate primary 2009) investigating brain response to different
and secondary taste centers as well as other regions types of taste stimuli (including sucrose, saccharin,
(Wang et  al., 2004). However, high calorie foods caffeine, sodium chloride, and citric acid) in both
may stimulate more medial and dorsolateral pre- hungry and fed states. When sated, the insula and
frontal areas, whereas low calorie food images may OFC demonstrated significantly greater brain acti-
activate medial OFC and temporal regions (Killgore vation in response to all stimuli than the hypothal-
et al., 2003). Brain imaging studies using pictures amus, amygdala, or hippocampus, regions involved
of food potentially test motivational or anticipa- in eating behavior, emotion, and memory, respec-
tory states and possibly the desire to approach food. tively. Sucrose elicited greater activation than the
The hunger state in fact seems to activate amygdala other tastes, and patterns of response differed across
and temporal areas more than the satiation state stimuli depending on hunger state. Together, these
(LaBar et  al., 2001), although gender differences functional taste studies in healthy controls offer
seem to exist. Of further interest, heightened anx- important new leads for the study of altered eating
iety in AN when considering consequences of food physiology in AN and BN.
intake, such as possible weight gain, may interact
with brain response in ACC or amygdala, poten- Functional and Task Activation Studies
tially reflecting such altered processing of hunger in Eating Disorders
and a desire to eat. A number of studies have used fMRI and single
Gender differences in healthy participants are photon emission computed tomography (SPECT)
evident in response to a liquid meal during hunger to investigate appetite regulation in AN and BN. In
or satiation (Del Parigi et al., 2002). In particular, general, sample sizes have been small, and studies
healthy women had higher activity in occipital and have used a range of methods and many different
parietal sensory association cortex and in the dorso- brain regions. Moreover, the resolution of SPECT is
lateral prefrontal cortex, but men had greater acti- poor so that specific regions cannot be clearly iden-
vation in the ventromedial prefrontal cortex when tified. While many of the studies have had positive
sated. Given that AN and BN primarily occur in findings, replication studies have only recently been
women, this contrast is of interest. This study did published. The field is in its infancy but is growing
not report on other behavioral/​emotional param- rapidly.
eters, but it is possible that there are also gender
differences in terms of response to food items and PET and SPECT: Regional Cerebral
taste. However, this is speculative and needs to be Blood Flow
further tested. Women may have greater cognitive Nozoe measured regional cerebral blood flow
activation to food stimuli, as well as sensorimotor (rCBF) using SPECT and detected a significant
cortex activation in response to a liquid meal after increase in response to food intake in the left infe-
satiation. Investigation of the cognitive impact ver- rior frontal cortex in AN compared to controls
sus a more basic physiologic processing in women (Nozoe et al., 1993). In a later study, Nozoe (Nozoe
versus men, and then in ED participants may et  al., 1995) reported that individuals with AN
suggest gender-​based neural differences that may did not have cortical laterality or activated state
maintain symptoms. Such studies may shed light in any cortical area before eating. However, there
on understanding why women are at greater risk of was increased activity in frontal, occipital, parietal
developing EDs. and temporal regions after eating. In contrast, BN
In summary, the study of normal brain activ- individuals showed the highest cortical activity in
ity in relation to food intake has identified cortical the left temporal region and the bilateral inferior
pathways related to the physiology of food intake, frontal regions before eating compared to controls
such as the insula. In particular, the medial OFC but less cortical activity in response to food intake.
may be somewhat specific for pleasant, and the Naruo et  al. (2000) reported that food imagina-
lateral OFC specific for aversive taste experiences. tion assessed by SPECT resulted in greater activa-
In addition, the prefrontal cortex may be involved tion in inferior, superior, prefrontal, and parietal
in cognitive processing of the food ingestion. In regions of the right brain in binge/​purge type AN
this network of activation, the amygdala was con- in comparison to restricting type AN and healthy
firmed to play a role and may respond relative to controls.

 Ely, Kaye 61
 Gordon et al. (2001) used PET to show a high with the insula appears to be altered in AN (Frank,
calorie food stimulus resulted in elevated rCBF in Shott, Riederer, & Pryor, 2016; Kim, Ku, Lee, Lee,
occipital temporal regions in individuals with AN & Jung, 2012; Moody et  al., 2015; Shott, Pryor,
compared to controls. Ellison, Foong, et al. (1998) Yang, & Frank, 2016). Alternatively, adults with
used fMRI and found that individuals with AN, BN have increased insula activation (Brooks et al.,
when viewing pictures of high calorie drinks, had 2011; Schienle, Schafer, Hermann, & Vaitl, 2009;
increased signal changes in the left insula, ante- Weygandt, Schaefer, Schienle, & Haynes, 2012)
rior ACC, and left amygdala/​hippocampal region (for review, see Garcia-​Garcia et al., 2013) implicat-
that were possibly anxiety related. Activation in ing both cognitive and reward circuitry.
the amygdala/​hippocampus region was also nega-
tively correlated to subscales of the Eating Disorder FMRI: Tastes of Food
Inventory-​2 (Shirao, Okamoto, Okada, Okamoto, Studies that manipulate the relationship of food
& Yamawaki, 2003) in healthy volunteers when cues to food receipt are able to examine response to
viewing unpleasant words concerning body image anticipation and consumption of food reward, and
relative to neutral words. In addition, a decrease of may elucidate primary reward processing in ED.
blood flow was detected in the ACC using SPECT A  growing body of research has identified neural
in restricting type AN compared to binge/​purge underpinnings of anticipation and receipt of food
type AN or controls by Naruo et al. (2000). reward in AN (Kaye et  al., 2013), suggesting ill
and REC AN adults have elevated activation with
FMRI: Images of Food anticipation of food cues in the anterior insula, stri-
Neuroimaging research using fMRI have also atum, and frontal regions (Cowdrey, Park, Harmer,
investigated brain response to food pictures and & McCabe, 2011; Frank et al., 2012; Oberndorfer,
tastes. Visual processing of images of food is thought Frank, et al., 2013; Oberndorfer, Simmons, et al.,
to activate anticipatory responses that determine 2013) and reduced activation to tastes of pal-
future feeding behavior (Stice, Spoor, Ng, & Zald, atable foods in the anterior insula and striatum
2009). Studies of visual food cues have demon- (Oberndorfer, Frank, et  al., 2013; Vocks et  al.,
strated notable differences in activation in par- 2011; Wagner et al., 2008). Specifically, our group
ticipants with ED. Adults with AN show reduced has found that REC AN adults have increased insula
neural response to food images in the insula, lat- response to cues (Oberndorfer, Simmons, et  al.,
eral PFC and parietal lobe and increased medial 2013) and reduced insula response when tasting
PFC activation (Gizewski et  al., 2010; Holsen sucrose (Oberndorfer, Frank, et al., 2013; Wagner
et al., 2012; Uher et al., 2003), while connectivity et  al., 2008) (See Figure 4.1). In comparison to

1.2

1.0

0.8
% Signal Change from T01

0.6

0.4

0.2

0.0
2 4 6 8 10 12 14 16 18 20
−0.2

−0.4 Time-Point (Seconds)

Figure 4.1  Coronal view of left anterior insula ROI (x = −41, y = 5, z = 5). Time course of BOLD signal as a mean of all 16 REC
AN (▴) and 16 CW (▪) for taste-​related (sucrose and water) response in the left insula. 

62 Appetitive Regul ation

control women, neural circuits associated with Munte, & Rotte, 2006). One study (Holsen et al.,
reward learning were more active in ill AN adults in 2012) has examined the effects of hunger and sati-
the anterior ventral striatum, insula, and orbitofron- ety in response to food images in both remitted and
tal cortex when engaging in an associative learning ill AN. Both AN groups showed hypoactivation in
task of conditioned visual stimuli and sucrose taste food motivation regions involved in the assessment
(Frank et al., 2012), suggesting enhanced activation of the reward value of food, suggesting these are ill-
to unexpected taste stimuli in reward circuitry in ness state-​independent alterations. AN may fail to
AN. Further, REC AN have increased activity in become appropriately hungry when starved, and are
the ventral striatum in response to both sights and thus able to become emaciated.
flavors of pleasant stimuli (chocolate) and increased Only one study has investigated the impact
activity in the insula and posterior dorsal caudate of hunger and fullness in response to taste in BN
to that of aversive stimuli when compared to con- (Ely, 2017). Our group found that, contrary to the
trols (Cowdrey et al., 2011). These findings suggest pattern exhibited by control participants in which
that an elevated response to predictive cues may be activation in the putamen, amygdala, and insula is
a way to manage distress resulting from aversive elevated in hunger and reduced with satiety, partici-
experience, similar to anticipatory sensitivity seen pants remitted from BN do not show a modulation
in highly anxious individuals (Simmons, Strigo, of response with physiological state. This failure to
Matthews, Paulus, & Stein, 2006). Individuals with devalue reward resulted in elevated activation in the
AN appear to have exaggerated anxiety response to amygdala as compared to controls when full, sug-
anticipated food cues, and diminished reward valu- gesting motivation to eat that does not attenuate
ation response to actual taste. with satiety. In summary, functional neuroimaging
The opposite pattern appears to be evident in studies that showed ill AN individuals pictures of
BN. Ill BN individuals show reduced response to food found altered activity in the insula and OFC
anticipatory food cues in the insula and striatum as well as in mesial temporal, parietal, and the
(Bohon & Stice, 2011; Frank, Reynolds, Shott, ACC regions as compared to CW (Ellison, Foong,
& O’Reilly, 2011) and increased response to taste et al., 1998; Gordon et al., 2001; Naruo et al.,
receipt in the insula and striatum (Oberndorfer 2000; Nozoe et al., 1993; Nozoe et al., 1995; Santel
et  al., 2013; Radeloff et  al., 2014). Further, BN et al., 2006; Uher et al., 2004). Studies using
is linked with decreased activation to unexpected SPECT, PET-​O15, or fMRI found that when ill
receipt of taste rewards as well as omission of AN tasted or were exposed to food images, they
expected stimuli (Frank et al., 2011). Reduced brain had activated temporal regions, and in some stud-
response to a DA-​related reward learning task in the ies, increased anxiety (Ellison, Foong, et al., 1998;
insula, ventral putamen, amygdala, and orbitofron- Gordon et al., 2001; Naruo et al., 2000; Nozoe
tal cortex was also seen in these ill BN participants. et al., 1993). Findings in AN and BN suggest that
Taken together, these results suggest BN may have an ED alters processes of reward and gustatory sen-
abnormalities in reward responding to food cues sation when tasting palatable foods (Small, 2009).
that contrast with AN, that may increase vulner- While the neural activation of those with AN may
ability to disinhibited eating. mimic a continuous state of satiety, dampening
interoceptive and reward responses to food, brain
FMRI Studies of Hunger and Satiety in Eating circuits may be hyperresponsive in BN, manifesting
Disorders as a chronic state of subjective deprivation.
It is also important to assess the impact of hun-
ger and satiety on response to food stimuli in ED. FMRI: Interoception in Eating Disorders
Appetitive dysregulation in AN and BN is poorly These findings raise the question of whether AN
understood. Appetite regulation is a complex proc- have an AI disturbance specifically related to gusta-
ess that involves the integration of a wide variety of tory modulation or a more generalized disturbance
signals such as energy needs in the body, hedonic related to the integration of interoceptive stimuli.
attraction to palatable foods, and long-​term cogni- Interoception, or the perception, processing, and
tive concerns about weight. AN is related to reduced integration of sensory information about bod-
activation in the parietal lobe, dorsal ACC, anterior ily states, has long been thought to be critical for
insula, and amygdala in response to food images self-​awareness as it provides the link between cog-
both when hungry and satiated (Gizewski et  al., nitive and affective processes and the current body
2010; Holsen et al., 2012; Santel, Baving, Krauel, state (Craig, 2002; Paulus & Stein, 2006). Limited

 Ely, Kaye 63
ability to recognize hunger, diminished insight and findings in REC AN (Bergen et  al., 2005; Frank
motivation to change, and altered central coher- et  al., 2005; Kaye et  al., 1999), we performed an
ence, a measure of local versus global cognitive proc- fMRI study (Wagner et  al., 2007) demonstrating
essing (Lopez et al., 2008), are all characteristic of that REC AN women did not differ positive and
AN and could be related to disturbed AI function. negative feedback in the anteroventral striatum, but
Interoception has been hypothesized to drive body did show exaggerated caudate-​dorsal striatum acti-
image distortion and the ability to severely restrict vation in response to both wins and losses as well as
intake (Fassino, Piero, Gramaglia, & Abbate-​Daga, greater response in the DLPFC and parietal cortex
2004; Kaye, Fudge, & Paulus, 2009; Lilenfeld et al., (see Figure 4.2). Further, trait anxiety in the REC
2006; Nunn, Frampton, & Lask, 2012; Pollatos AN group was associated with percent signal change
et al., 2008; Zucker et al., 2013). Ill and REC AN to wins or losses in the left caudate. Altered proc-
describe overresponsiveness to sensation (Brand-​ essing of nonfood rewards in AN was further sup-
Gothelf et  al., 2016), and studies have reported ported by a recent study of ability to delay future
altered interoceptive disturbances in AN in multi- rewards (Wierenga et al., 2015) or delay discount-
ple domains, including the absence of satiety aver- ing. Behaviorally, women with AN have been shown
sion to sucrose (Garfinkel, Moldofsky, & Garner, to have lower delay discounting than healthy indi-
1979; Garfinkel, Moldofsky, Garner, Stancer, & viduals (Decker, Figner, & Steinglass, 2015), in that
Coscina, 1978), heartbeat detection (Eshkevari, they do not devalue monetary reward the longer
Rieger, Musiat, & Treasure, 2014; Khalsa, Craske, they must wait for it to the same extent as controls.
Li, Vangala, Strober, & Feusner, 2015), breathing Wierenga demonstrated that women remitted from
(Khalsa et al., 2015), and physical discomfort (de AN failed to show an elevation of activation of
Zwaan, Biener, Bach, Wiesnagrotzki, & Stacher, reward valuation circuitry when hungry in response
1996; Lautenbacher, Pauls, Strian, & Pirke, Krieg, to monetary reward, but had enhanced response in
1990; Nakai & Koh, 2001; Strigo et al., 2013), and inhibitory control brain regions across metabolic
these disturbances persisted after normalization of state. These findings that underresponse to reward-
weight or failure to rate food as positive when hun- ing stimuli extends beyond food in AN and may
gry (Kerr et  al., 2016; Santel et  al., 2006). Some be due to fundamental dysfunction in valuation of
studies have shown evidence of altered interocep- salient stimuli. Individuals with AN may have both
tion in BN by self-​report (Fassino et  al., 2004; an impaired ability to identify the emotional signif-
Thiels & Patel, 2008) and reduced accuracy for icance of a stimulus and an enhanced ability to plan
bodily signals in REC BN (Klabunde, Acheson, or foresee consequences. Because of AVS pathway
Boutelle, Matthews, & Kaye, 2013). Alternatively, dysregulation, REC AN may focus on long-​term
Eshkevari and colleagues (Eshkevari et  al., 2014) consequences rather than an immediate response to
found no significant differences in accuracy on the salient stimuli. In fact, AN individuals tend to have
heartbeat detection task. A  recent study (Pollatos an enhanced ability to pay attention to detail or use
& Georgiou 2016) found BN did not differ from a logical/​analytic approach, but exhibit worse per-
healthy participants on interoceptive accuracy but formance for global strategies in the here and now
did show reduced interoceptive awareness. These (Lopez et al., 2007; Strupp, Weingartner, Kaye, &
constructs were not correlated among controls, but Gwirtsman, 1986). In particular, the most anxious
in BN found greater interoceptive accuracy was AN individuals may respond in an overly “cogni-
linked to blunted interoceptive awareness, suggest- tive” manner to both negative and positive stimuli.
ing aberrant interoceptive signal processing in BN. Anxiety may interfere with reward valuation and
may have impaired ability to identify emotional sig-
FMRI: Response to Reward Is Altered nificance of the stimuli (Phillips, Drevets, Rauch,
in Eating Disorders & Lane, 2003). This may provide important new
Individuals with AN have long been noted understanding of why it is so difficult to motivate
to be anhedonic and ascetic, able to sustain self-​ AN individuals to engage in treatment since they
denial of food as well as most comforts and plea- may not be able to appreciate rewarding stimuli
sures in life (Frank et al., 2005). They also tend to (Halmi et al., 2005).
be highly harm avoidant and overconcerned about While AN is associated with inhibition, anx-
consequences. This temperament persists, in a more iety, and inflexibility (Harrison, O’Brien, Lopez,
modest form, after recovery (Klump et  al., 2004; & Treasure, 2010; Klump et  al., 2004; Wagner
Wagner, Barbarich, et  al., 2006). Given the DA et al., 2007), individuals with BN tend to be more

64 Appetitive Regul ation

 (a)
0.30

0.25

0.20
Percent Signal Change

0.15

0.10

0.05

0.00
0 2 4 6 8 10 12 14 16
−0.05

−0.10

−0.15 Time (Seconds)

(b)
0.30

0.25

0.20
Percent Signal Change

0.15

0.10

0.05

0.00
0 2 4 6 8 10 12 14 16
−0.05

−0.10

−0.15 Time (Seconds)

Figure 4.2  Time course of BOLD signal as mean percent signal change (from first scan per trial) for loss (dashed line) and win (solid
line) conditions, for REC AN (▲ gray) and CW (■ black) corresponding to (a) left caudate (x = −12, y = 15, z = 7) and (b) left ventral
striatum (x = −10, y = 6, z = −5). 

impulsive, pleasure-​and sensation-​ seeking, and et al., 2010), it also failed to distinguish responses
unconcerned with future consequences (Cassin to positive and negative feedback. Therefore, REC
& von Ranson, 2005). Using the same monetary BN may have executive dysfunction in the integra-
choice task (Delgado et  al., 2000), as described tion of important stimuli with their consequences,
previously in REC AN (Wagner et al., 2007), our reflected in impulsive behavior.
group (Wagner et al., 2010) showed that REC BN
also failed to differentiate feedback valence (wins FMRI: Response to Inhibition Is Altered
and losses) in ventral-​striatal regions in compar- in Eating Disorders
ison to controls, suggesting difficulty in discrim- FMRI studies suggest that individuals with AN
inating positive and negative feedback, perhaps have exaggerated higher-​order inhibitory cognitive
related to increased sensitivity to both reward and control, while those with BN have impaired inhibi-
punishment. While the magnitude of dorsal cau- tion. Recent fMRI research in AN have focused on
date response was normal in REC BN (Wagner error monitoring, set-​shifting, delay discounting,

 Ely, Kaye 65
and behavioral inhibition to assess neural substrates activation of dorsal ACC when making errors
of cognitive control. Overall, results show elevated than when responding correctly. Interestingly,
dorsolateral cognitive circuitry activation linked to BN patients with more severe symptoms were less
impaired set-​shifting and increased cognitive con- accurate in their responses. In a study of ill binge
trol, but reduced activation in medial and lateral eating/​purging adolescents, Lock and colleagues
prefrontal regions during error monitoring and showed this group had increased DLPFC activa-
motor inhibitory control. For example, ill AN adults tion compared to an ill restricting-​type AN group
showed greater activation of dorsolateral frontopari- during No-​Go (i.e., successfully inhibited) trials of
etal networks during a set-​shifting task, potentially the Go/​No-​Go task (Lock, Garrett, Beenhakker, &
due to excessive effortful monitoring of cognitive Reiss, 2011) supporting a role of dysfunctional dor-
control (Zastrow et al., 2009). As described above, sal cognitive neurocircuitry in BN. Together, these
we recently showed that women remitted from AN data suggest that impaired frontostriatal activation
also had elevated cognitive circuitry activation dur- might underlie increased impulsivity and decreased
ing delay discounting (Wierenga, Ely, et al., 2014), capacity for self-​regulatory behaviors, which in turn
suggesting that individuals with AN may rely on could predispose individuals to engage in binge/​
top-​down assessment to compensate for impaired purge behaviors.
motivational processing when evaluating choices or These fMRI studies show altered activity of dor-
making decisions. Conversely, (Pieters et al., 2007) sal cognitive circuitry in both AN and BN, con-
ill AN adults showed blunted dorsal ACC activation sistent with the contention that AN is associated
during errors of commission on a flanker task and with enhanced cognitive control and individuals
reduced cingulate function in relation to executive with BN have reduced ability to self-​regulate. In
function (Ferro et  al., 2005). Similarly, decreased particular, the ability to inhibit the motivational
ventrolateral prefrontal cortex response during set drive to eat may be supported by neural processes
shifting error feedback trials of the Wisconsin Card of exaggerated self-​control (e.g., exaggerated dor-
Sorting Test has been shown in ill AN, suggesting sal cognitive circuit function), whereas lowered
altered response to errors when shifting cognitive set inhibition and cognitive control impulses may be
(Sato et  al., 2013). Using a motor inhibition stop due to blunted function of dorsal cognitive circuit
signal task, we have also demonstrated that both function in combination with exaggerated ventral-​
ill AN adolescents (Wierenga, Bischoff-​ Grethe, striatal reward processing that may increase one’s
et  al., 2014) and REC AN women (Oberndorfer, vulnerability to overeat.
Kaye, Simmons, Strigo, & Matthews, 2011) have
decreased task-​related activation in the middle fron- Future Directions for Eating Disorder
tal gyrus on hard trials (i.e., with greater inhibitory Research
demands). Together, this research raises the possibil- Data to date raise the possibility that individu-
ity that individuals with AN require fewer inhibi- als with ED may have disturbances of circuits
tory resources to maintain behavioral performance, that modulate emotionality and reward as well
reduced activation can accompany more experience as cognition. In turn, such alterations may affect
with a task, potentially suggesting neural efficiency. salient stimuli, such as food, and more complex
In sum, findings show that AN have altered infor- behaviors, such as impulse control. The core traits
mation processing, with deficient processing in cir- that may increase vulnerability for developing an
cuitry concerned with planning and consequences ED are only beginning to be elucidated (Pearson,
and elevated efficiency in regions linked with con- Zapolski, & Smith, 2015), but how these traits are
flict monitoring and motor inhibition. encoded in neuronal circuits remains unknown.
In BN, evidence suggests individuals may There may be a primary disturbance of appetite
have reduced inhibition or greater dysregulation regulation, or disturbed appetite could be second-
due to a failure to appropriately engage fron- ary to altered reward, anxiety, obsessionality, or
tostriatal circuits. For example, during correct other phenomena. Thus, it is important to con-
responses to incongruent trials of the Simon Spatial struct experiments that test behaviors that basic
Incompatibility task, (Marsh et  al., 2011; Marsh science suggests might be encoded in neurocircuits.
et  al., 2009) ill BN participants show reduced For example, the circuitry underlying taste proc-
response in the ACC, inferiolateral PFC, inferior essing is better understood than those implicated
frontal gyrus, and lenticular and caudate nuclei. in response to visual food cues. As a result, investi-
They also demonstrated inappropriately greater gating taste response in ED may more definitively

66 Appetitive Regul ation

highlight aberrant activation. More specifically, that may be useful for stimulating new therapeutic
exploring the effects of hunger and satiety on interventions.
sensory-​hedonic and homeostatic response of the
insula and related brain regions will contribute Summary
to better understanding of the pathophysiology Individuals with AN are able to restrict their
underlying restricted eating in AN, and overeat- food intake everyday and maintain a low weight for
ing in BN. Neurodevelopment may also play a many years. How are individuals with AN able to
role in risk for ED. The development of orbital maintain a chronic diet and become emaciated when
and dorsolateral prefrontal regions of the cortex most people struggle to lose a few pounds? Likewise,
occurs concurrently and after the onset of puberty individuals with BN may lose self-​regulatory control
(Huttenlocher & Dabholkar, 1997) and may chal- by engaging in binge eating and purging behaviors.
lenge vulnerable systems. Increased functional Why do these behaviors relieve dysphoric mood
capacity of these cortical areas may be a substrate states in BN when most people find such behaviors
for excessive worry, perfectionism, and strategiz- aversive?
ing seen in these individuals. Taking the presumed Converging receptor-​ binding and functional
pathological processes and involved brain areas brain imaging data point to a primary disturbance
together, top-​ down amplification of anticipa- of appetite regulation within taste-​and reward-​
tory signal for satiety may result in overrestrictive processing regions of the brain and suggest that this
feeding behavior in AN. Computationally, these may be driving eating disorder symptoms. The AI
processes would involve generating an anticipa- in particular makes critical contributions toward
tory body state model, which describes the neu- determining the hedonic and sensory tone of food
ral circuitry in which a representation of the choices and interoceptive awareness. We hypothesize
physiological condition of the body is instanti- that the AI may play a pivotal role in ED: A failure
ated. Alternatively, uneven development of fron- of the AI to respond appropriately to hunger due
tostriatal regions may play a role in vulnerability to altered interoceptive homeostatic mechanisms,
for BN (Berner & Marsh, 2014). Neurocircuitry perhaps involving disturbed sensory-​hedonic tone,
responsible for reward motivation, including ven- could contribute to the ability of AN individuals to
tral striatum, shows stronger signaling in adoles- restrict food intake and become emaciated, while
cence suggesting earlier maturation (Galvan et al., dysregulation of appetitive signals in this region may
2006; Geier, Terwilliger, Teslovich, Velanova, & result in the extremes of restriction and binge eating
Luna, 2010; Somerville, Hare, & Casey, 2011; seen in BN.
Van Leijenhorst et al., 2010). Exaggerated activa- Numerous functional studies also implicate
tion of reward circuitry in adolescents is evident altered activation of the dorsal and ventral stria-
in fMRI studies of both anticipation and receipt tum, ACC, and OFC, as contributing to distur-
of monetary reward (Ernst et  al., 2005; Galvan bances of feeding behaviors and appetite regulation.
et  al., 2006; Geier et  al., 2010; Van Leijenhorst Behavioral correlations provide further evidence
et  al., 2010). However, ability to delay gratifica- supporting a role for these brain regions in regulat-
tion only improves gradually, raising the possibil- ing taste processing, feeding behavior, and reward
ity that (Steinberg et  al., 2008) Ventral striatum response to gustatory stimuli. Only recently have
influence may bias typical behavior in adolescence noninvasive neuroimaging tools such as PET,
toward greater impulsivity and risk-​taking (Casey, SPECT, and fMRI been available to explore the
Getz, & Galvan, 2008; Casey, Jones, & Hare, neuropathology underlying EDs, and they will
2008; Ernst, Pine, & Hardin, 2006; Somerville likely be instrumental in identifying new therapeu-
et al., 2011; Somerville & Casey, 2010; Steinberg tic targets.
et al., 2008). Berner and Marsh (2014) have theo-
rized that frontostriatal dysfunction characteristic Glossary of Technical Terms
of BN may arise in adolescence due to even ear-
lier maturation of striatal regions contributing to [18F]altanserin: A PET radioligand specific for
reward learning, and delayed development of self-​ 5-​HT2A receptors.
regulatory control. Blood Oxygen Level Dependent Signal: Main
In summary, it is important to generate overarch- outcome measure of fMRI. Takes advantage of
ing hypotheses about complex systems dysfunction differing magnetic properties between oxygenated
in EDs that can be tested against new findings, and and deoxygenated blood and the observation that

 Ely, Kaye 67
metabolically active brain regions receive an influx SPECT—​Single Photon Emission Computed
of oxygenated blood. Tomography
Body Mass Index: Standardized comparison of
height and weight (Kg/​m2). References
Abraham, S., & Beaumont, P. (1982). How patients describe
Binding Potential: Combined measure of bulimia or binge eating. Psychological Medicine, 12,
neurotransmitter receptor availability and affinity 625–​635.
to radioligands. Ahima, R., & Osei, S. (2004). Leptin signaling. Physiology &
Behavior, 81, 223–​241.
Carbonyl-​[11C]WAY100635: A PET radioligand Ahima, R., Saper, C., Flier, J., & Elmquist, J. (2000). Leptin
specific for 5-​HT1A receptors. regulation of neuroendocrine systems. Frontiers in
Neuroendocrinology, 21, 263–​307.
Cortisol: Hormone produced by the adrenal gland.
American Psychiatric Association (APA). (2000). Diagnostic and
Increases blood pressure and blood sugar, and is statistical manual of mental disorders (4th ed., DSM:VI-​TR).
associated with increased stress. Washington, DC.
Fluorodeoxyglucose: A PET radioligand used to American Psychiatric Association (APA). (2013). Diagnostic
and statistical manual of mental disorders (5th ed.; DSM-​V).
study glucose metabolism. Washington, DC: Author.
PET-​O15: PET technique utilizing radiolabeled Anderluh, M. B., Tchanturia, K., Rabe-​Hesketh, S., & Treasure,
oxygen, which measures blood flow throughout J. (2003). Childhood obsessive-​ compulsive personality
traits in adult women with eating disorders:  Defining a
the brain.
broader eating disorder phenotype. The American Journal of
Radioligand: Radioactive substance that binds Psychiatry, 160, 242–​247.
with specificity to particular biological molecules. Anderson, I. M., Parry-​Billings, M., Newsholme, E. A., Fairburn,
Commonly used in PET and SPECT. C. G., & Cowen, P. J. (1990). Dieting reduces plasma trypto-
phan and alters brain 5-​HT function in women. Psychological
Regional Cerebral Blood Flow: Outcome measure Medicine, 20, 785–​791.
of PET-​O15. Ando, T., Ichimaru, Y., Konjiki, F., Shoji, M., & Komaki, G.
(2007). Variations in the preproghrelin gene correlate with
higher body mass index, fat mass, and body dissatisfaction
Abbreviations in young Japanese women. The American Journal of Clinical
5-​HIAA—​5-​Hydroxyindolacetic Acid Nutrition, 86, 25–​32.
Ando, T., Komaki, G., Nishimura, H., Naruo, T., Okabe, K.,
5-​HT—​Serotonin
Kawai, K.,  . . .  Japanese Genetic Research Group for Eating
ACC—​Anterior Cingulate Cortex Disorders. (2010). A ghrelin gene variant may predict cross-
AI—​Anterior Insula over rate from restricting-​type anorexia nervosa to other phe-
AVS—​Anteroventral Striatum notypes of eating disorders: A retrospective survival analysis.
BED—​Binge Eating Disorder Psychiatric Genetics, 20, 153–​159.
Arana, F., Parkinson, J. A., Hinton, E., Holland, A., Owen, A., &
BMI—​Body Mass Index
Roberts, A. (2003). Dissociable contributions of the human
BOLD—​Blood Oxygen Level Dependent Signal amygdala and orbitofrontal cortex to incentive motivation
BP—​Binding Potential and goal selection. Journal of Neuroscience, 23, 9632–​9838.
CCK—​Cholecystokinin Audenaert, K., Van Laere, K., Dumont, F., Vervaet, M., Goethals,
CNS—​Central Nervous System I., Slegers, G.,   .  .  .  Dierckx, R. A. (2003). Decreased 5-​
HT2a receptor binding in patients with anorexia nervosa.
CRH—​Corticotropin-​Releasing Hormone
Journal of Nuclear Medicine, 44, 163–​169.
CSF—​Cerebrospinal Fluid Bacanu, S., Bulik, C., Klump, K., Fichter, M., Halmi, K., Keel,
CW—​Control Women P.,  . . .  Devlin, B. (2005). Linkage analysis of anorexia and
DA—​Dopamine bulimia nervosa cohorts using selected behavioral pheno-
DLPFC—​Dorsolateral Prefrontal Cortex types as quantitative traits or covariates. American Journal of
Medical Genetics B, Neuropsychiatric Genetics, 139, 61–​68.
FDG—​Fluorodeoxyglucose
Bailer, U., Frank, G., Price, J., Meltzer, C., Becker, C., Mathis,
fMRI—​Functional Magnetic Resonance Imaging C.,  . . .  Kaye, W. H. (2013). Interaction between serotonin
HPA—​Hypothalamic-​Pituitary-​Adrenal Axis transporter and dopamine D2/​D3 receptor radioligand meas­
m-​CPP—​m-​Chlorophenylpiperazine ures is associated with harm avoidant symptoms in anorexia
mPFC—​Medial Prefrontal Cortex and bulimia nervosa. Psychiatry Research: Neuroimaging, 211,
160–​168.
NPY—​Neuropeptide-​Y
Bailer, U., Price, J., Meltzer, C., Wagner, A., Mathis, C., Gamst,
OFC—​Orbitofrontal Cortex A.,   .  .  .  Kaye, W. H. (2016). Dopaminergic activity and
PET—​Positron Emission Tomography altered reward modulation in anorexia nervosa-​ insight
PYY—​Peptide YY from multimodal imaging. International Journal of Eating
rCBF—​Regional Cerebral Blood Flow Disorders. [Epub ahead of print]

68 Appetitive Regul ation

Bailer, U. F., Frank, G., Henry, S., Price, J., Meltzer, C., Mathis, Blundell, J. E. (1984). Serotonin and appetite. Neuropharmacology,
C.,  . . .  Kaye, W. H. (2007). Exaggerated 5-​HT1A but nor- 23(12B), 1537–​1551.
mal 5-​HT2A receptor activity in individuals ill with anorexia Bohon, C., & Stice, E. (2011). Reward abnormalities among
nervosa. Biological Psychiatry, 61, 1090–​1099. women with full and subthreshold bulimia nervosa: a func-
Bailer, U. F., Frank, G. K., Henry, S. E., Price, J. C., Meltzer, tional magnetic resonance imaging study. International
C. C., Weissfeld, L.,  . . .  Kaye, W. H. (2005). Altered brain Journal of Eating Disorders, 44, 585–​595.
serotonin 5-​HT1A receptor binding after recovery from Bonhomme, N., & Esposito, E. (1998). Involvement of sero-
anorexia nervosa measured by positron emission tomogra- tonin and dopamine in the mechanism of action of
phy and [11C]WAY100635. Archives of General Psychiatry, 62, novel antidepressant drugs:  a review. Journal of Clinical
1032–​1041. Psychopharmacology, 18, 447–​454.
Bailer, U. F., Price, J. C., Meltzer, C. C., Mathis, C. A., Frank, G. Boyar, R. K., J, Finkelstein, J., Kapen, S., Weiner, H., Weitzman,
K., Weissfeld, L.,  . . .  Kaye, W. H. (2004). Altered 5-​HT2A E., & Hellman, L. (1974). Anorexia nervosa. Immaturity of
receptor binding after recovery from bulimia-​type anorexia the 24-​hour luteinizing hormone secretory pattern. The New
nervosa: Relationships to harm avoidance and drive for thin- England Journal of Medicine, 291, 861–​865.
ness. Neuropsychopharmacology, 29, 1143–​1155. Brambilla, F., Brunetta, M., Draisci, A., Peirone, A., Perna,
Baranowska, B., Radzikowska, M., Wasilewska-​Dziubinska, E., G., Sacerdote, P.,  . . .  Panerai, A. E. (1995). T-​lymphocyte
Roguski, K., & Borowiec, M. (2000). Disturbed release of cholecystokinin-​8 and beta-​endorphin in eating disorders: II.
gastrointestinal peptides in anorexia nervosa and in obesity. Bulimia nervosa. Psychiatry Research, 59, 51–​56.
Diabetes, Obesity and Metabolism, 2, 99–​103. Brambilla, F., Brunetta, M., Peirone, A., Perna, G., Sacerdote,
Baranowska, B., Wolinska-​Witort, E., Wasilewska-​Dziubinska, P., Manfredi, B.,  . . .  Panerai, A. E. (1995). T-​lymphocyte
E., Roguski, K., & Chmielowska, M. (2001). Plasma leptin, cholecystokinin-​ 8 and beta-​ endorphin concentrations in
neuropeptide Y (NPY) and galanin concentrations in buli- eating disorders:  I.  Anorexia nervosa. Psychiatry Research,
mia nervosa and in anorexia nervosa. Neuroendocrinology 59, 43–​50.
Letters, 22, 356–​358. Brand-​Gothelf, A., Parush, S., Eitan, Y., Admoni, S., Gur, E.,
Bardone-​Cone, A., Harney, M., Maldonado, C., Lawson, M., & Stein, D. (2016). Sensory modulation disorder symp-
Robinson, D., Smith, R., & Tosh, A. (2010). Defining toms in anorexia nervosa and bulimia nervosa: A pilot study.
recovery from an eating disorder:  Conceptualization, vali- International Journal of Eating Disorders, 49, 59–​68.
dation, and examination of psychosocial functioning and Breiter, H. C., Aharon, I., Kahneman, D., Dale, A., & Shizgal,
psychiatric comorbidity. Behaviour Research and Therapy, P. (2001). Functional imaging of neural responses to expec-
48, 194–​202. tancy and experience of monetary gains and losses. Neuron,
Barr, L. C., Goodman, W. K., Price, L. H., McDougle, C. J., & 30, 619–​639.
Charney, D. S. (1992). The serotonin hypothesis of obsessive Brewerton, T. D. (1995). Toward a unified theory of sero-
compulsive disorder:  implications of pharmacologic chal- tonin dysregulation in eating and related disorders.
lenge studies. Journal of Clinical Psychiatry, 53 Suppl, 17–​28. Psychoneuroendocrinology, 20, 561–​590.
Berg, K., Crosby, R., Cao, L., Peterson, C., Engel, S. G., Brewerton, T. D., Lesem, M. D., Kennedy, A., & Garvey, W. T.
Mitchell, J. E.,  . . .  Wonderlich, S. A. (2013). Facets of nega- (2000, October 25). Reduced plasma leptin concentration
tive affect prior to and following binge-​only, purge-​only, and in bulimia nervosa. Psychoneuroendocrinology, 25, 649–​658.
binge/​purge events in women with bulimia nervosa. Journal Brewerton, T. D., Lydiard, R. B., Laraia, M. T., Shook, J. E., &
of Abnormal Psychology, 122, 111–​118. Ballenger, J. C. (1992). CSF beta-​endorphin and dynorphin
Bergen, A., Yeager, M., Welch, R., Haque, K., Ganjei, J. K., in bulimia nervosa. The American Journal of Psychiatry, 149,
van den Bree, M. B.,  . . .  Kaye, W. H. (2005). Association 1086–​1090.
of multiple DRD2 polymorphisms with anorexia nervosa. Broft, A., Shingleton, R., Kaufman, J., Liu, F., Kumar, D.,
Neuropsychopharmacology, 30, 1703–​1710. Slifstein, M.,  . . .  Walsh, B. T. (2012). Striatal dopamine in
Berner, L., & Marsh, R. (2014). Frontostriatal circuits and the bulimia nervosa: A pet imaging study. International Journal
development of bulimia nervosa. Frontiers Behavioural Brain of Eating Disorders, 45, 648–​656.
Research, 17, 395. Brooks, S., O’Daly, O., Uher, R., Frederich, H.-​C., Giampietro,
Berns, G., McClure, S., Pagnoni, G., & Montague, P. (2001). V., Brammer, M.,  . . .  Campbell, I. C. (2011). Differential
Predictability modulates human brain response to reward. neural responses to food images in women with bulimia ver-
Journal of Neuroscience, 21, 2793–​2798. sus anorexia nervosa. Plos One, 6, 1–​8.
Berrettini, W. H. (2000). Genetic linkage studies of bipolar dis- Bulik, C., Bacanu, S., Klump, K., Fichter, M., Halmi, K., Keel,
orders. Neurology, Psychiatry and Brain Research, 8, 139–​146. P.,  . . .  Devlin, B. (2005). Selection of eating-​disorder phe-
Berridge, K., & Robinson, T. (1998). What is the role of dopa- notypes for linkage analysis. American Journal of Medical
mine in reward: hedonic impact, reward learning, or incen- Genetics B, Neuropsychiatric Genetics, 139, 81–​87.
tive salience? Brain Res, 28, 309–​369. Bulik, C., Hebebrand, J., Keski-​ Rahkonen, A., Klump, K.,
Bhagwagar, Z., Rabiner, E., Sargent, P., Grasby, P., & Cowen, Reichborn-​Kjennerud, K. S., Mazzeo, S., & Wade, T. D.
P. (2004). Persistent reduction in brain serotonin1A receptor (2007). Genetic epidemiology, endophenotypes, and eat-
binding in recovered depressed men measured by positron ing disorder classification. International Journal of Eating
emission tomography with [11C]WAY-​100635. Molecular Disorders, S52–​S60.
Psychiatry, 9, 386–​392. Bulik, C. M., Sullivan, P. F., Fear, J. L., & Joyce, P. R. (1997).
Blum, K., Sheridan, P. J., Wood, R. C., Braverman, E. R., Chen, Eating disorders and antecedent anxiety disorders:  A  con-
T. J., & Comings, D. E. (1995). Dopamine D2 receptor gene trolled study. Acta Psychiatrica Scandinavica, 96, 101–​107.
variants: association and linkage studies in impulsive-​addictive-​ Bulik, C., Sullivan, P. F., Tozzi, F., Furberg, H., Lichtenstein,
compulsive behaviour. Pharmacogenetics, 5, 121–​141. P., & Pedersen, N. L. (2006). Prevalence, heritability and

 Ely, Kaye 69
 prospective risk factors for anorexia nervosa. Archives of Considine, R., Sinha, M., Heiman, M., Kriauciunas, A., Stephens,
General Psychiatry, 63, 305–​312. T., Nyce, M.,  . . .  Caro, J. F. (1996). Serum immunoreactive-​
Burnet, P. W., Eastwood, S. L., & Harrison, P. J. (1997). leptin concentrations in normal-​weight and obese humans.
[3H]WAY-​100635 for 5-​HT1A receptor autoradiography New England Journal of Medicine, 334, 292–​295.
in human brain:  A  comparison with [3H]8-​OH-​DPAT Cowdrey, F., Park, R., Harmer, C., & McCabe, C. (2011).
and demonstration of increased binding in the frontal Increased neural processing of rewarding and aversive food
cortex in schizophrenia. Neurochemistry International, 30, stimuli in recovered anorexia nervosa. Biological Psychiatry,
565–​574. 70, 736–​743.
Calvez, J., de Ávila, C., Guèvremont, G., & Timofeeva, E. Craig, A. D. (2002). How do you feel? Interoception: The sense
(2016). Stress differentially regulates brain expression of of the physiological condition of the body. Nature Reviews
corticotropin-​releasing factor in binge-​like eating prone and Neuroscience, 3, 655–​666.
resistant female rats. Appetite, 107, 585–​595. Critchley, H., & Rolls, E. (1996). Hunger and satiety modify the
Cannon, C., & Bseikri, M. (2004). Is dopamine required for responses of olfactory and visual neurons in the primate orbi-
natural reward? Physiology & Behavior, 81, 741–​748. tofrontal cortex. Journal of Neurophysiology, 75, 1673–​1686.
Carmichael, S., & Price, J. (1996). Connectional networks Cuntz, U., Enck, P., Frühauf, E., Lehnert, P., Riepl, R., & Fichter,
within the orbital and medial prefrontal cortex of macaque M. (2013). Cholecystokinin revisited: CCK and the hunger
monkeys. Journal of Comparative Neurology, 371, 179–​207. trap in anorexia nervosa. Plos One, 8, 354457.
Carter, C., Botvinick, M., & Cohan, J. (1999). The contribution De Araujo, I., Rolls, E., Kringelbach, M., McGlone, F., &
of the anterior cingulate cortex to executive processes in cog- Phillips, N. (2003). Taste-​ olfactory convergence, and
nition. Reviews in the Neurosciences, 10, 49–​57. the representation of the pleasantness of flavour, in
Casey, B., Getz, S., & Galvan, A. (2008). The adolescent brain. the human brain. European Journal of Neuroscience, 18,
Developmental Review, 28, 62–​77. 2059–​2068.
Casey, B., Jones, R., & Hare, T. (2008). The adolescent brain. Decker, J., Figner, B., & Steinglass, J. (2015). On weight and
Ann NY Acad Sci, 1124, 111–​126. waiting: Delay discounting in anorexia nervosa pretreatment
Casper, R. C. (1990). Personality features of women with good and post-​treatment. Biological Psychiatry, 78, 606–​614.
outcome from restricting anorexia nervosa. Psychosomatic Deep, A. L., Nagy, L. M., Weltzin, T. E., Rao, R., & Kaye, W. H.
Medicine, 52, 156–​170. (1995). Premorbid onset of psychopathology in long-​term
Cassin, S., & von Ranson, K. (2005). Personality and eating disor- recovered anorexia nervosa. International Journal of Eating
ders: A decade in review. Clinical Psychology Review, 25, 895–​916. Disorders, 17, 291–​297.
Celada, P., Puig, M. V., Casanovas, J. M., Guillazo, G., & Artigas, Del Parigi, A., Chen, K., Gautier, J., Salbe, A., Pratley, R.,
F. (2001). Control of dorsal raphe serotonergic neurons by Ravussin, E.,   .  .  .  Tataranni P. A. (2002). Sex differences
the medial prefrontal cortex: Involvement of serotonin-​1A, in the human brain’s response to hunger and satiation. The
GABAA, and glutamate receptors. Journal of Neuroscience, 21, American Journal of Clinical Nutrition, 75, 1017–​1022.
9917–​9929. Delgado, M., Nystrom, L., Fissel, C., Noll, D., & Fiez, J. (2000).
Cerf-​Ducastel, B., & Murphy, C. (2004). Validation of a stimu- Tracking the hemodynamic responses to reward and punishment
lation protocol suited to the investigation of odor-​taste inter- in the striatum. Journal of Neurophysiology, 84, 3072–​3077.
actions with fMRI. Physiology & Behavior, 81, 389–​396. Devlin, M. J., Walsh, B. T., Guss, J. L., Kissileff, H. R., Liddle,
Chandraskekar, J., Hoon, M., Ryba, N., & Zuker, C. (2006). R. A., & Petkova, E. (1997). Postprandial cholecystokinin
The receptors and cells for mammalian taste. Nature, 444, release and gastric emptying in patients with bulimia nervosa.
288–​294. The American Journal of Clinical Nutrition, 65, 114–​120.
Charney, D. S., Woods, S. W., Krystal, J. H., & Heninger, G. De Vry, J., & Schreiber, R. (2000). Effects of selected sero-
R. (1990). Serotonin function and human anxiety disorders. tonin 5-​ HT and 5-​ HT receptor agonists on feeding
Annals of the New York Academy of Sciences, 600, 558–​572. behavior:  Possible mechanisms of action. Neuroscience &
Chikama, M., McFarland, N., Amaral, D., & Haber, S. (1997). Biobehavioral Reviews, 24, 341–​353.
Insular cortical projections to functional regions of the stria- de Zwaan, M., Biener, D., Bach, M., Wiesnagrotzki, S., &
tum correlate with cortical cytoarchitectonic organization in Stacher, G. (1996). Pain sensitivity, alexithymia, and depres-
the primate. Journal of Neuroscience, 17, 9686–​9705. sion in patients with eating disorders:  Are they related?
Clarke, H. F., Walker, S. C., Dalley, J. W., Robbins, T. W., & Journal of Psychosomatic Research, 41, 65–​70.
Roberts, A. C. (2007). Cognitive inflexibility after prefrontal Drevets, W. C., Frank, E., Price, J. C., Kupfer, D. J., Holt, D.,
serotonin depletion is behaviorally and neurochemically spe- Greer, P. J.,  . . .  Mathis, C. (1999). PET imaging of serotonin
cific. Cerebral Cortex, 17, 18–​27. 1A receptor binding in depression. Biological Psychiatry, 46,
Cloninger, C. R. (1987). A systematic method for clinical 1375–​1387.
description and classification of personality variants: A pro- Drewnowski, A., Bellisle, F., Aimez, P., & Remy, B. (1987). Taste
posal. Archives of General Psychiatry, 44, 573–​588. and bulimia. Physiology & Behavior, 41, 621–​626.
Cloninger, C. R., Przybeck, T. R., Svrakic, D. M., & Wetzel, Drewnowski, A., Halmi, K. A., Pierce, B., Gibbs, J., & Smith, G.
R. D. (1994). The Temperament and Character Inventory P. (1987). Taste and eating disorders. The American Journal of
(TCI):  A  guide to its development and use (pp.  19–​28). St. Clinical Nutrition, 46, 442–​450.
Louis, MO:  Center for Psychobiology of Personality, Drewnowski, A., Pierce, B., & Halmi, K. (1988). Fat aversion in
Washington University. eating disorders. Appetite, 10, 119–​131.
Cohen, M., Young, J., Baek, J., Kessler, C., & Ranganath, C. Eckert, E. D., Pomeroy, C., Raymond, N., Kohler, P. F.,
(2005). Individual differences in extraversion and dopamine Thuras, P., & Bowers, C. Y. (1998). Leptin in anorexia ner-
genetics predict neural reward responses. Brain Research. vosa. Journal of Clinical Endocrinology & Metabolism, 83,
Cognitive Brain Research, 25, 851–​861. 791–​795.

70 Appetitive Regul ation

Ellison, A. R., & Fong, J. (1998). Neuroimaging in eating disor- Frank, G., Kaye, W. H., Carter, C., Brooks, S., May, C., Fissel,
ders. In H. W. Hoek, J. L. Treasure & M. A. Katzman (Eds.), K., & Stenger, V. A. (2003). The evaluation of brain activ-
Neurobiology in the treatment of eating disorders (pp. 255–​ ity in response to taste stimuli—​A pilot study and method
269). Chichester, UK: Wiley. for central taste activation as assessed by event related fMRI.
Ellison, Z., Foong, J., Howard, R., Bullmore, E., Williams, S., Journal of Neuroscience Methods, 131(1–​2), 99–​105.
& Treasure, J. (1998). Functional anatomy of calorie fear in Frank, G., Oberndorfer, T., Simmons, A., Paulus, M., Fudge,
anorexia nervosa. Lancet, 352, 1192. J., Yang, T., & Kaye, W. H. (2008). Sucrose activates
Elman, I., Borsook, D., & Lukas, S. (2006). Food intake and reward human taste pathways differently from artificial sweetener.
mechanisms in patients with schizophrenia:  Implications Neuroimage, 39, 1559–​1569.
for metabolic disturbances and treatment with second-​ Frank, G., Reynolds, J., Shott, M., Jappe, L., Yang, T., Tregellas,
generation antipsychotic agents. Neuropsychopharmacology, J., & O’Reilly, R. C. (2012). Anorexia nervosa and obe-
31, 2091–​2120. sity are associated with opposite brain reward response.
Ely, A. V., Wierenga, C. E., Bischoff-Grethe, A., Bailer, U. F., Neuropsychopharmacology, 37, 2031–​2046.
Berner, L. A., Fudge, J. L., . . . Kaye, W. H. (2017). Response Frank, G., Reynolds, J., Shott, M., & O’Reilly, R. (2011).
in taste circuitry is not modulated by hunger and satiety in Altered temporal difference learning in bulimia nervosa.
women remitted from bulimia nervosa. Journal of Abnormal Biological Psychiatry, 70, 728–​735.
Psychology, 126(5), 519. Frank, G., Shott, M., Riederer, J., & Pryor, T. (2016). Altered
Ernst, M., Nelson, E., Jazbec, S., McClure, E., Monk, C., structural and effective connectivity in anorexia and bulimia
Leibenluft, E.,   .  .  .  Pine, D. S. (2005). Amygdala and nervosa in circuits that regulate energy and reward homeosta-
nucleus accumbens in responses to receipt and omis- sis. Translational Psychiatry, 6, e932.
sion of gains in adults and adolescents. Neuroimage, 25, Frank, G., Wagner, A., Brooks-​ Achenbach, S., McConaha,
1279–​1291. C., Skovira, K., Aizenstein, H.,  . . .  Kaye, W. H. (2006).
Ernst, M. P., Pine, D. S., & Hardin, M. (2006). Triadic model Altered brain activity in women recovered from bulimic
of the neurobiology of motivated behavior in adolescence. type eating disorders after a glucose challenge: A pilot study.
Psychological Medicine, 36, 299–​312. International Journal of Eating Disorders, 39, 76–​79.
Eshkevari, E., Rieger, E., Musiat, P., & Treasure, J. (2014). An Frank, G. K., Kaye, W. H., Meltzer, C. C., Price, J. C., Greer,
investigation of interoceptive sensitivity in eating disorders P., McConaha, C., & Skovira, K. (2002). Reduced 5-​HT2A
using a heartbeat detection task and a self‐report measure. receptor binding after recovery from anorexia nervosa.
European Eating Disorders Review, 22, 383–​388. Biological Psychiatry, 52, 896–​906.
Fairbanks, L., Melega, W., Jorgensen, M., Kaplan, J., & McGuire, Frederich, R., Hu, S., Raymond, N., & Pomeroy, C.
M. (2001). Social impulsivity inversely associated with CSF (2002). Leptin in anorexia nervosa and bulimia ner-
5-​HIAA and fluoxetine exposure in vervet monkeys. Journal vosa: Importance of assay technique and method of inter-
of Neuropsychopharmacology, 24, 370–​378. pretation. Journal of Laboratory and Clinical Medicine,
Farooqi, I. S., Keogh, J. M., Kamath, S., Jones, S., Gibson, W. 139, 72–​79.
T., Trussell, R.,  . . .  O’Rahilly, S. (2001). Partial leptin defi- Friederich, H. C., Kumari, V., Uher, R., Riga, M., Schmidt, U.,
ciency and human adiposity. Nature, 414(6859), 34–​35. Campbell, I. C.,  . . .  Treasure, J. (2006). Differential moti-
Fassino, S., Piero, A., Gramaglia, C., & Abbate-​Daga, G. (2004). vational responses to food and pleasurable cues in anorexia
Clinical, psychopathological and personality correlates of and bulimia nervosa: A startle reflex paradigm. Psychological
interoceptive awareness in anorexia nervosa, bulimia nervosa Medicine, 36, 1327–​1335.
and obesity. Psychopathology, 37, 168–​174. Friedman, J., & Halaas, J. (1998). Leptin and the regulation of
Faurion, A., Cerf, B., Van De Moortele, P. F., Lobel, E., Mac body weight in mammals. Nature, 395, 763–​770.
Leod, P., & Le Bihan, D. (1999). Human taste cortical Fudge, J., Breibart, M., Danish, M., & Pannoni, V. (2005).
areas studied with functional magnetic resonance imag- Insular and gustatory inputs to the caudal ventral stria-
ing: Evidence of functional lateralization related to handed- tum in primates. Journal of Comparative Neurology, 490,
ness. Neuroscience Letters, 277, 189–​192. 101–​118.
Fernstrom, M. H., Weltzin, T. E., Neuberger, S., Srinivasagam, Fudge, J., Breitbart, M., & McClain, C. (2004). Amygdaloid
N., & Kaye, W. H. (1994). Twenty-​four-​hour food intake in inputs define a caudal component of the ventral striatum in
patients with anorexia nervosa and in healthy control sub- primates. Journal of Comparative Neurology, 476, 330–​347.
jects. Biological Psychiatry, 36, 696–​702. Galvan, A., Hare, T., Parra, C., Penn, J., Voss, H., Glover, G., &
Ferro, A., Brugnolo, A., De Leo, C., Dessi, B., Girtler, N., Casey, B. J. (2006). Earlier development of the accumbens rel-
Morbelli, S.,   .  .  .  Rodriguez, G. (2005). Stroop inter- ative to orbitofrontal cortex might underlie risk-​taking behav-
ference task and single-​ photon emission tomography in ior in adolescents. Journal of Neuroscience, 26, 6885–​6892.
anorexia:  A  preliminary report. International Journal of Garcia-​Garcia, I., Narberhaus, A., Marques-​Iturria, I., Garolera,
Eating Disorders, 38, 323–​329. M., Radoi, A., Sergura, B.,  . . .  Jurado, M. A. (2013). Neural
File, S. E., Kenny, P. J., & Cheeta, S. (2000). The role of the responses to visual food cues: Insights from functional magnetic
dorsal hippocampal serotonergic and cholinergic systems in resonance imaging. European Eating Disorders Review, 21, 89–​98.
the modulation of anxiety. Pharmacology Biochemistry and Garfinkel, P., Moldofsky, H., & Garner, D. M. (1979). The
Behavior, 66, 65–​72. stability of perceptual disturbances in anorexia nervosa.
Frank, G., Bailer, U. F., Henry, S., Drevets, W., Meltzer, C. C., Psychological Medicine, 9, 703–​708.
Price, J. C.,  . . .  Kaye, W. H. (2005). Increased dopamine Garfinkel, P., Moldofsky, H., Garner, D. M., Stancer, H.
D2/​D3 receptor binding after recovery from anorexia ner- C., & Coscina, D. (1978). Body awareness in anorexia
vosa measured by positron emission tomography and [11C] nervosa:  Disturbances in “body image” and “satiety.”
raclopride. Biological Psychiatry, 58, 908–​912. Psychosomatic Medicine, 40, 487–​498.

 Ely, Kaye 71
Gehring, W., & Willoughby, A. (2002). The medial frontal cor- with positron emission tomography. Journal of Pediatrics,
tex and the rapid processing of monetary gains and losses. 139, 51–​57.
Science, 295, 2279–​2282. Goto, Y., & Grace, A. (2005). Dopaminergic modulation of lim-
Geier, C., Terwilliger, R., Teslovich, T., Velanova, K., & Luna, bic and cortical drive of nucleus accumbens in goal-​directed
B. (2010). Immaturities in reward processing and its influ- behavior. Nature Neuroscience, 386, 14–​17.
ence on inhibitory control in adolescence. Cerebral Cortex, Gottfried, J. A., O’Doherty, J., & Dolan, R. J. (2002). Appetite
20, 1613–​1629. and aversive olfactory learning in humans studied using
Geliebter, A., Melton, P. M., McCray, R. S., Gallagher, D. R., event-​related functional magnetic resonance imaging.
Gage, D., & Hashim, S. A. (1992). Gastric capacity, gas- Journal of Neuroscience, 15, 10829–​10837.
tric emptying, and test-​meal intake in normal and bulimic Gottfried, J., O’Doherty, J., & Dolan, R. (2003). Encoding pre-
women. American Journal of Clinical Nutrition, 56, 656–​661. dictive reward value in human amygdala and orbitofrontal
Gendall, K., Kaye, W., Altemus, M., McConaha, C., & La Via, cortex. Science, 301, 1104–​1107.
M. (1999). Leptin, neuropeptide Y, and peptide YY in long-​ Grahame-​Smith, D. G. (1992). Serotonin in affective disorders.
term recovered eating disorder patients. Biological Psychiatry, International Clinical Psychopharmacology, 6 (Suppl 4), 5–​13.
46, 292–​299. Grinspoon, S., Gulick, T., Askari, H., Landt, M., Lee, K.,
Geracioti, T. D., Jr., & Liddle, R. A. (1988). Impaired chole- Anderson, E.,   .  .  .  Klibanski, A. (1996). Serum leptin
cystokinin secretion in bulimia nervosa. The New England levels in women with anorexia nervosa. Journal of Clinical
Journal of Medicine, 319, 683–​688. Endocrinology & Metabolism, 81, 3861–​3863.
Geracioti, T. D., Jr., Liddle, R. A., Altemus, M., Demitrack, M. Gross, C., Zhuang, X., Stark, K., Ramboz, S., Oosting, R.,
A., & Gold, P. W. (1992). Regulation of appetite and chole- Kirby, L.,  . . .  Hen, R. (2002). Serotonin1A receptor acts dur-
cystokinin secretion in anorexia nervosa. American Journal of ing development to establish normal anxiety-​like behaviour
Psychiatry, 149, 958–​961. in the adult. Nature, 416, 396–​400.
Germain, N., Galusca, B., Grouselle, D., Frere, D., Billard, S., Gwirtsman, H. E., Kaye, W. H., George, D. T., Jimerson, D. C.,
& Epelbaum, J. (2010). Ghrelin and obestatin circadian lev- Ebert, M. H., & Gold, P. W. (1989). Central and peripheral
els differentiate bingeing-​purging from restrictive anorexia ACTH and cortisol levels in anorexia nervosa and bulimia.
nervosa. Journal of Clinical Endocrinology & Metabolism, 95, Archives of General Psychiatry, 46, 61–​69.
357–​3062. Haase, L., Cerf-​Ducastel, B., & Murphy, C. (2009). Cortical
Germain, N., Galusca, B., Le Roux, C., Bossu, C., Ghatei, M., activation in response to pure taste stimuli during the
Lang, F.,   .  .  .  Estour, B. (2007). Constitutional thinness physiological states of hunger and satiety. Neuroimage, 44,
and lean anorexia nervosa display opposite concentrations of 1008–​1021.
peptide YY, glucagon-​line peptide 1, ghrelin, and leptin. The Haber, S., Kunishio, K., Mizobuhi, M., & Lynd-​Balta, E. (1995).
American Journal of Clinical Nutrition, 85, 957–​971. The orbital and medial prefrontal circuit through the primate
Geyer, M. A. (1996). Serotonergic functions in arousal and basal ganglia. Journal of Neuroscience, 15, 4851–​4867.
motor activity. Behavioural Brain Research, 73, 31. Haber, S. N., Kim, K., Mailly, P., & Calzavara, R. (2006).
Gibbs, J., Young, R. C., & Smith, G. P. (1973). Cholecystokinin Reward-​related cortical inputs define a large striatal region in
decreases food intake in rats. Journal of Comparative and primates that interface with associative cortical connections,
Physiological Psychology, 84, 488–​495. providing a substrate for incentive-​based learning. Journal of
Gizewski, E., Rosenberger, C., de Greiff, A., Moll, A., Senf, Neuroscience, 26, 8368–​8376.
W., Wanke, I.,  . . .  Herpertz, S. (2010). Influence of sati- Haedt-​Matt, A., & Keel, P. (2011). Revisiting the affect regula-
ety and subjective valence rating on cerebral activation pat- tion model of binge eating: A meta-​analysis of studies using
terns in response to visual stimulation with high-​ calorie ecological momentary assessment. Psychological Bulletin,
stimuli among restrictive anorectic and control women. 137, 660–​681.
Neuropsychobiology, 62, 182–​192. Halford, J., Cooper, G., & Dovey, T. (2004). The pharmacology of
Glowa, J., & Gold, P. (1991). Corticotropin releasing hormone human appetite expression. Current Drug Targets, 5, 221–​240.
produces profound anorexigenic effects in the rhesus mon- Halmi, K., Agras, W. S., Crow, S., Mitchell, J., Wilson, G.,
key. Neuropeptides, 18, 55–​61. Bryson, S., & Kraemer, H. C. (2005). Predictors of treatment
Godart, N. T., Flament, M. F., Lecrubier, Y., & Jeammet, P. acceptance and completion in anorexia nervosa. Archives of
(2000). Anxiety disorders in anorexia nervosa and bulimia General Psychiatry, 62, 776–​781.
nervosa:  Co-​ morbidity and chronology of appearance. Halmi, K. A., Sunday, S., Puglisi, A., & Marchi, P. (1989).
European Psychiatry, 15, 38–​45. Hunger and satiety in anorexia and bulimia nervosa. Annals
Godart, N. T., Flament, M. F., Perdereau, F., & Jeammet, P. of the New York Academy of Sciences, 575, 431–​444.
(2002). Comorbidity between eating disorders and anxiety Hannon-​Engel, S., & Filin, E. E., Wolfe, B. E. (2013). CCK
disorders: a review. International Journal of Eating Disorders, response in bulimia nervosa and following remission.
32, 253–​270. Physiology & Behavior, 122, 56–​61.
Gold, P. W., Gwirtsman, H., Avgerinos, P. C., Nieman, L. Harrison, A., O’Brien, N., Lopez, C., & Treasure, J. (2010).
K., Gallucci, W. T., Kaye, W.,   .  .  .  Chrousos, G. (1986). Sensitivity to reward and punishment in eating disorders.
Abnormal hypothalamic-​ pituitary-​
adrenal function in Psychiatry Research, 177(1–​2), 1–​11.
anorexia nervosa:  Pathophysiologic mechanisms in under- Harty, R. F., Pearson, P. H., Solomon, T. E., & McGuigan,
weight and weight-​ corrected patients. The New England J. E. (1991). Cholecystokinin, vasoactive intestinal peptide
Journal of Medicine, 314, 1335–​1342. and peptide histidine methionine responses to feeding in
Gordon, C. M., Dougherty, D. D., Fischman, A. J., Emans, S. anorexia nervosa. Regulatory Peptides, 36, 141–​150.
J., Grace, E., Lamm, R.,  . . .  Rauch, S. L. (2001). Neural Hebebrand, J., van der Heyden, J., Devos, R., Kopp, W.,
substrates of anorexia nervosa: A behavioral challenge study Herpertz, S., Remschmidt, H., & Herzog, W. (1995). Plasma

72 Appetitive Regul ation

 concentrations of obese protein in anorexia nervosa. Lancet, food. Proceedings of the National Academy of Sciences of the
346, 1624–​1625. United States of America, 88, 10931–​10935.
Hikosaka, K., & Watanabe, M. (2000). Delay activity of orbital Kaye, W. H., Ballenger, J. C., Lydiard, R. B., Stuart, G. W.,
and lateral prefrontal neurons of the monkey varying with Laraia, M. T., O’Neil, P.,  . . .  Hsu, G. (1990). CSF mono-
different rewards. Cerebral Cortex, 10, 263–​271. amine levels in normal-​weight bulimia: Evidence for abnor-
Hilbert, A., & Tuschen-​Caffier, B. (2007). Maintenance of binge mal noradrenergic activity. American Journal of Psychiatry,
eating through negative mood:  A  naturalistic comparison 147, 225–​229.
of binge eating disorder and bulimia nervosa. International Kaye, W. H., Berrettini, W. H., Gwirtsman, H. E., Chretien, M.,
Journal of Eating Disorders, 40, 521–​530. Gold, P. W., George, D. T.,  . . .  Ebert, M. H. (1987). Reduced
Hinton, E., Parkinson, J. A., Holland, A., Arana, F., Roberts, cerebrospinal fluid levels of immunoreactive pro-​opiomela-
A., & Owen, A. (2004). Neural contributions to the moti- nocortin related peptides (including beta-​ endorphin) in
vational control of appetite in humans. European Journal of anorexia nervosa. Life Sciences, 41, 2147–​2155.
Neuroscience, 20, 1411–​1418. Kaye, W. H., Berrettini, W., Gwirtsman, H., & George, D. T.
Holsen, L., Lawson, E., Blum, K., Ko, E., Makris, N., Fazeli, (1990). Altered cerebrospinal fluid neuropeptide Y and pep-
P.,  . . .  Goldstein, J. M. (2012). Food motivation circuitry tide YY immunoreactivity in anorexia and bulimia nervosa.
hypoactivation related to hedonic and nonhedonic aspects Archives of General Psychiatry, 47, 548–​556.
of hunger and satiety in women with active anorexia nervosa Kaye, W. H., Bulik, C., Thornton, L., Barbarich, N., & Masters,
and weight-​restored women with anorexia nervosa. Journal of K. (2004). Comorbidity of anxiety disorders with anorexia
Psychiatry & Neuroscience, 37, 322–​332. and bulimia nervosa. The American Journal of Psychiatry, 161,
Holtkamp, K., Hebebrand, J., Mika, C., Heer, M., Heussen, 2215–​2221.
N., & Herpertz-​Dahlmann, B. (2004). High serum leptin Kaye, W. H., Ebert, M. H., Raleigh, M., & Lake, R. (1984).
levels subsequent to weight gain predict renewed weight loss Abnormalities in CNS monoamine metabolism in anorexia
in patients with anorexia nervosa. Psychoneuroendocrinology, nervosa. Archives of General Psychiatry, 41, 350–​355.
29, 791–​797. Kaye, W. H., Frank, G. K., & McConaha, C. (1999). Altered
Holtkamp, K., Mika, C., Grzella, I., Heer, M., Pak, H., dopamine activity after recovery from restricting-​ type
Hebebrand, J., & Herpertz-​ Dahlmann, B. (2003). anorexia nervosa. Neuropsychopharmacology, 21, 503–​506.
Reproductive function during weight gain in anorexia ner- Kaye, W. H., Frank, G. K., Meltzer, C. C., Price, J. C.,
vosa:  Leptin represents a metabolic gate to gonadotropin McConaha, C. W., Crossan, P. J.,  . . .  Rhodes, L. (2001).
secretion. Journal of Neural Transmission, 110, 427–​435. Altered serotonin 2A receptor activity in women who
Huttenlocher, P., & Dabholkar, A. (1997). Regional differ- have recovered from bulimia nervosa. American Journal of
ences in synaptogenesis in human cerebral cortex. Journal of Psychiatry, 158, 1152–​1155.
Comparative Neurology, 387, 167–​178. Kaye, W. H., Fudge, J., & Paulus, M. (2009). New insight into
Inui, A. (2001). Eating behavior in anorexia nervosa: An excess symptoms and neurocircuit function of anorexia nervosa.
of both orexigenic and anorexigenic signalling? Molecular Nature Reviews Neuroscience, 10, 573–​584.
Psychiatry, 6, 620–​624. Kaye, W. H., Gendall, K., & Strober, M. (1998). Serotonin
Izquierdo, I., Cammarota, M., Medina, J., & Bevilaqua, L. neuronal function and selective serotonin reuptake inhibi-
(2004). Pharmacological findings on the biochemical bases tor treatment in anorexia and bulimia nervosa. Biological
of memory processes:  A  general view. Neural Plasticity, 11, Psychiatry, 44, 825–​838.
159–​189. Kaye, W. H., Gwirtsman, H. E., George, D. T., & Ebert, M. H.
Jimerson, D., & Wolfe, B. (2006). Psychobiology of eating dis- (1991). Altered serotonin activity in anorexia nervosa after
orders. In S. Wonderlich, J. E. Mitchell, M. de Zwaan, & H. long-​term weight restoration: Does elevated cerebrospinal fluid
Steiger (Ed.), Annual review of eating disorders: Part 2–​2006 5-​hydroxyindoleacetic acid level correlate with rigid and obses-
(pp. 1–​15). Oxford, UK: Radcliffe Publishing. sive behavior? Archives of General Psychiatry, 48, 556–​562.
Jimerson, D., & Wolfe, B. E. (2004). Neuropeptides in eating Kaye, W. H., Gwirtsman, H. E., George, D. T., Ebert, M.
disorders. CNS Spectrums, 9, 516–​522. H., Jimerson, D. C., Tomai, T. P.,   .  .  .  Gold, P. W.
Jimerson, D. C. (2002). Leptin and the neurobiology of eat- (1987). Elevated cerebrospinal fluid levels of immunore-
ing disorders. Journal of Laboratory and Clinical Medicine, active corticotropin-​ releasing hormone in anorexia ner-
139, 70–​71. vosa:  Relation to state of nutrition, adrenal function, and
Jimerson, D. C., Lesem, M. D., Kaye, W. H., & Brewerton, T. D. intensity of depression. Journal of Clinical Endocrinology &
(1992). Low serotonin and dopamine metabolite concentra- Metabolism, 64, 203–​208.
tions in cerebrospinal fluid from bulimic patients with frequent Kaye, W. H., Gwirtsman, H. E., George, D. T., Weiss, S. R., &
binge episodes. Archives of General Psychiatry, 49, 132–​138. Jimerson, D. C. (1986). Relationship of mood alterations to
Jimerson, D. C., Lesem, M. D., Kaye, W. H., Hegg, A. P., & bingeing behaviour in bulimia. British Journal of Psychiatry,
Brewerton, T. D. (1990). Eating disorders and depression:  is 149, 479–​485.
there a serotonin connection? Biological Psychiatry, 28, 443–​454. Kaye, W. H., Weltzin, T. E., McKee, M., McConaha, C., Hansen,
Jimerson, D. C., Mantzoros, C., Wolfe, B. E., & Metzger, E. D. D., & Hsu, L. K. (1992). Laboratory assessment of feeding
(2000). Decreased serum leptin in bulimia nervosa. Journal behavior in bulimia nervosa and healthy women:  Methods
of Clinical Endocrinology & Metabolism, 85, 4511–​4514. for developing a human-​feeding laboratory. The American
Johnson, C., & Larson, R. (1982). Bulimia: An analysis of mood Journal of Clinical Nutrition, 55, 372–​380.
and behavior. Psychosomatic Medicine, 44, 341–​351. Kazama, A., & Bachevalier, J. (2006). Selective aspiration of neu-
Kalra, S. P., Dube, M. G., Sahu, A., Phelps, C. P., & Kalra, P. rotoxic lesions of the orbitofrontal areas 11 and 13 spared
S. (1991). Neuropeptide Y secretion increases in the para- monkeys’ performance on the object reversal discrimination
ventricular nucleus in association with increased appetite for task. Society for Neuroscience Abstracts, 32, 670–​675.

 Ely, Kaye 73
Keel, P., Wolfe, B. E., Liddle, R., De Young, K., & Jimerson, Kringelbach, M. L., de Araujo, I. E. T., & Rolls, E. T. (2004).
D. (2007). Clinical features and physiological response to a Taste-​related activity in the human dorsolateral prefrontal
test meal in purging disorder and bulimia nervosa. Archives cortex. Neuroimage, 21, 781–​788.
of General Psychiatry, 64, 1058–​1066. Kringelbach, M. L., O’Doherty, J., Rolls, E., & Andrews, C.
Kelley, A. E. (2004). Ventral striatal control of appetite motiva- (2003). Activation of the human orbitofrontal cortex to a
tion: Role in ingestive behavior and reward-​related learning. liquid food stimulus is correlated with its subjective pleasant-
Neuroscience & Biobehavioral Reviews, 27, 765–​776. ness. Cerebral Cortex, 13, 1064–​1071.
Kendler, K. S., MacLean, C., Neale, M., Kessler, R., Heath, A., & LaBar, K., Gitelman, D., Parrish, T., Kim, Y., Nobre, A., &
Eaves, L. (1991). The genetic epidemiology of bulimia ner- Mesulam, M. (2001). Hunger selectively modulates corti-
vosa. The American Journal of Psychiatry, 148, 1627–​1637. colimbic activation to food stimuli in humans. Behavioural
Kendler, K. S., Walters, E. E., Neale, M. C., Kessler, R. C., Brain Research, 115, 493–​500.
Heath, A. C., & Eaves, L. J. (1995). The structure of the LaChaussee, J. L., Kissileff, H., Walsh, B., & Hadigan, C. (1992).
genetic and environmental risk factors for six major psychiat- The single-​item meal as a measure of binge-​eating behavior
ric disorders in women: Phobia, generalized anxiety disorder, in patients with bulimia nervosa. Physiology & Behavior, 51,
panic disorder, bulimia, major depression, and alcoholism. 593–​600.
Archives of General Psychiatry, 52, 374–​383. Lanzenberger, R., Mitterhauser, M., Spindelegger, C., Wadsak,
Kerr, K., Moseman, S., Avery, J., Bodurka, J., Zucker, N., & Kyle W., Klein, N., Mien, L.,  . . .  Tauscher, J. (2007). Reduced
Simmons, W. (2016). Altered insula activity during visceral serotonin-​ 1A receptor binding in social anxiety disorder.
interoception in weight-​restored patients with anorexia ner- Biological Psychiatry, 61, 1081–​1089.
vosa Neuropsychopharmacology, 41, 521–​528. Lautenbacher, S., Pauls, A., Strian, F., & Pirke, K. M., Krieg, J
Khalsa, S., Craske, M., Li, W., Vangala, S., Strober, M., & C. (1990). Pain perception in patients with eating disorders.
Feusner, J. (2015). Altered interoceptive awareness in Psychosomatic Medicine, 52, 673–​682.
anorexia nervosa: Effects of meal anticipation, consumption Lavender, J. M., Wonderlich, S. A., Peterson, C., Crosby, R.,
and bodily arousal. International Journal of Eating Disorders, Engel, S., Mitchell, J.,  . . .  Berg, K. C. (2014). Dimensions
48, 889–​897. of emotion dysregulation in bulimia nervosa. European
Killgore, W., Young, A., Femia, L., Bogorodzki, P., Rogowska, Eating Disorders Review, 22, 212–​216.
J., & Yurgelun-​Todd, D. (2003). Cortical and limbic acti- Lawrence, A. (2003). Impaired visual discrimination learning in
vation during viewing of high-​versus low-​ calorie foods. anorexia nervosa. Appetite, 20, 85–​89.
Neuroimage, 19, 1381–​1394. Leibowitz, S. F., & Shor-​Posner, G. (1986). Brain serotonin and
Kim, K., Ku, J., Lee, J., Lee, H., & Jung, Y. (2012). Functional eating behavior. Appetite, 7(Suppl), 1–​14.
and effective connectivity of anterior insula in anorexia Lesem, M. D., Berrettini, W., Kaye, W. H., & Jimerson, D. C.
nervosa and bulimia nervosa. Neuroscience Letters, 521, (1991). Measurement of CSF dynorphin A  1-​8 immuno-
152–​157. reactivity in anorexia nervosa and normal-​weight bulimia.
Kissileff, H., Brunstrom, J., Tesser, R., Bellace, D., Berthod, S., Biological Psychiatry, 29, 244–​252.
Thornton, J., & Halmi, K. (2016). Computerized measure- Licinio, J., Wong, M. L., & Gold, P. W. (1996). The
ment of anticipated anxiety from eating increasing portions hypothalamic-​ pituitary-​adrenal axis in anorexia nervosa.
of food in adolescents with and without anorexia ner- Psychiatry Research, 62, 75–​83.
vosa: Pilot studies. Appetite, 97, 160–​168. Lilenfeld, L., Wonderlich, S., Riso, L. P., Crosby, R., & Mitchell,
Kissileff, H., Pi-​Sunyer, F., Thornton, J., & Smith, G. (1981). J. (2006). Eating disorders and personality: A methodologi-
C-​terminal octapeptide of cholecystokinin decreases food cal and empirical review. Clinical Psychology Review, 26,
intake in man. The American Journal of Clinical Nutrition, 299–​320.
34, 154–​160. Lilenfeld, L. R., Kaye, W. H., Greeno, C. G., Merikangas, K.
Klabunde, M., Acheson, D. Boutelle, K. N., Matthews, S., & R., Plotnicov, K., Pollice, C.,  . . .  Nagy, L. (1998). A con-
Kaye, W. H. (2013). Interoceptive sensitivity deficits in trolled family study of anorexia nervosa and bulimia ner-
women recovered from bulimia nervosa. Eating Behaviors, vosa: Psychiatric disorders in first-​degree relatives and effects
14, 488–​492. of proband comorbidity. Archives of General Psychiatry, 55,
Klump, K., Strober, M., Johnson, C., Thornton, L., Bulik, C., 603–​610.
Devlin, B.,   .  .  .  Kaye, W. H. (2004). Personality charac- Lob, S., Pickel, J., Bidlingmaier, M., Schaaf, L., Backmund,
teristics of women before and after recovery from an eating H., Gerlinghoff, M., & Stalla, G. K. (2003). Serum leptin
disorder. Psychological Medicine, 34, 1407–​1418. monitoring in anorectic patients during refeeding therapy.
Klump, K. L., Bulik, C. M., Pollice, C., Halmi, K. A., Fichter, Experimental and Clinical Endocrinology & Diabetes, 111,
M. M., Berrettini, W. H.,   .  .  .  Kaye, W. H. (2000). 278–​282.
Temperament and character in women with anorexia ner- Lock, J., Garrett, A., Beenhakker, J., & Reiss, A. (2011).
vosa. Journal of Nervous and Mental Disease, 188, 559–​567. Aberrant brain activation during a response inhibition task
Klump, K. L., McGue, M., & Iacono, W. G. (2000). Age differ- in adolescent eating disorder subtypes. The American Journal
ences in genetic and environmental influences on eating atti- of Psychiatry, 168, 55–​64.
tudes and behaviors in preadolescent and adolescent female Lopez, C., Tchanturia, K., Stahl, D., Booth, R., Holliday, J., &
twins. Journal of Abnormal Psychology, 109, 239–​251. Treasure, J. (2007). An examination of central coherence in
Kojima, S., Nakahara, T., Nagai, N., Muranaga, T., Tanaka, M., women with anorexia nervosa. International Journal of Eating
Yasuhara, D.,   .  .  .  Naruo, T. (2005). Altered ghrelin and Disorders. [Epub ahead of print]
peptide YY responses to meals in bulimia nervosa. Clinical Lopez, C., Tchanturia, K., Stahl, D., Booth, R., Holliday, J., &
Endocrinology, 62, 74–​78. Treasure, J. (2008). An examination of the concept of central

74 Appetitive Regul ation

 coherence in women with anorexia nervosa. International ingestion in bulimia nervosa:  Implications for binge-​eating
Journal of Eating Disorders, 41, 143–​152. and compensatory behaviours. Psychological Medicine, 33,
Lydiard, R. B., Brewerton, T. D., Fossey, M. D., Laraia, M. T., 1387–​1394.
Stuart, G., Beinfeld, M. C.,  . . .  Ballenger, J. C. (1993). CSF Monteleone, P., Martiadis, V., Rigamonti, A., Fabrazzo, M.,
cholecystokinin octapeptide in patients with bulimia ner- Giordani, C., Muller, E.,  . . .  Maj, M. (2005). Investigation
vosa and in normal comparison subjects. American Journal of of peptide YY and ghrelin responses to a test meal in bulimia
Psychiatry, 150, 1099–​1101. nervosa. Biol Psychiatry, 57, 926–​931.
Mann, J. J. (1999). Role of the serotonergic system in the Monteleone, P., Scognamiglio, A. M., D, Mastromo, L.,
pathogenesis of major depression and suicidal behavior. Seteardo, L., Jr., Serino, I., & Maj, M. (2011). Abnormal
Neuropsychopharmacology, 21(2 Suppl), 99S–​105S. diurnal patterns of salivary α-​amylase and cortisol secretion
Mantzoros, C., Flier, J. S., Lesem, M. D., Brewerton, T. D., in acute patients with anorexia nervosa. World Journal of
& Jimerson, D. C. (1997). Cerebrospinal fluid leptin in Biological Psychiatry, 12, 455–​461.
anorexia nervosa:  Correlation with nutritional status and Monteleone, P., Serritella, C., Martiadis, V., Scognamiglio, P.,
potential role in resistance to weight gain. Journal of Clinical & Maj, M. (2008). Plasma obestatin, ghrelin, and ghrelin/​
Endocrinology & Metabolism, 82, 1845–​1851. obestatin ratio are increased in underweight patients with
Marsh, R., Horga, G., Wang, Z., Wang, P., Klahr, K., Berner, anorexia nervosa but not in symptomatic patients with buli-
L.,   .  .  .  Peterson, B. S. (2011). An FMRI study of self-​ mia nervosa. Journal of Clinical Endocrinology & Metabolism,
regulatory control and conflict resolution in adolescents with 93, 4418–​4421.
bulimia nervosa. The American Journal of Psychiatry, 168, Moody, T., Sasaki, M., Bohon, C., Strober, M., Bookheimer, S.,
1210–​1220. Sheen, C., & Feusner, J. D. (2015). Functional connectivity for
Marsh, R., Steinglass, J., Gerber, A., Graziano O’Leary, K., Wang, face processing in individuals with body dysmorphic disorder
Z., Murphy, D.,  . . .  Peterson, B. S. (2009). Deficient activ- and anorexia nervosa. Psychological Medicine, 45, 3491–​3503.
ity in the neural systems that mediate self-​regulatory control Moresco, F. M., Dieci, M., Vita, A., Messa, C., Gobbo, C., Galli,
in bulimia nervosa. Archives of General Psychiatry, 66, 51–​63. L.,  . . .  Fazio, F. (2002). In vivo serotonin 5HT2A receptor
Martin, E. R., Kaplan, N. L., & Weir, B. S. (1997). Tests for binding and personality traits in healthy subjects: A positron
linkage and association in nuclear families. American Journal emission tomography study. Neuroimage, 17, 1470–​1478.
of Human Genetics, 61, 439–​448. Morley, J. E., & Blundell, J. E. (1988). The neurobiological basis
Martin-​Ruiz, R., Puig, M. V., Celada, P., Shapiro, D. A., Roth, of eating disorders: Some formulations. Biological Psychiatry,
B. L., Mengod, G.,  . . .  Artigas, F. (2001). Control of sero- 23, 53–​78.
tonergic function in medial prefrontal cortex by serotonin-​ Morley, J. E., Levine, A. S., Gosnell, B. A., Mitchell, J. E.,
2A receptors through a glutamate-​dependent mechanism. Krahn, D. D., & Nizielski, S. E. (1985). Peptides and feed-
Journal of Neuroscience, 21, 9856–​9866. ing. Peptides, 6, 181–​192.
McClure, S., Berns, G., & Montague, P. (2003). Temporal pre- Morris, J. S., & Dolan, R. J. (2001). Involvement of human amyg-
diction errors in a passive learning task activate human stria- dala and orbitofrontal cortex in hunger-​enhanced memory
tum. Neuron, 38, 339–​346. for food stimuli. Journal of Neuroscience, 21(14), 5304–​5310.
Misra, M., Miller, K., Tsai, P., Gallagher, K., Lin, A., Lee, Nakahara, T., Kojima, S., Tanaka, M., Yasuhara, D., Harada, T.,
N.,   .  .  .  Klibanski, A. (2006). Elevated peptide YY levels Sagiyama, K.,  . . .  Inui, A. (2007). Incomplete restoration
in adolescent girls with anorexia nervosa. Journal of Clinical of the secretion of ghrelin and PYY compared to insulin after
Endocrinology & Metabolism, 91, 1027–​1033. food ingestion following weight gain in anorexia nervosa.
Montague, R., Hyman, S., & Cohen, J. (2004). Computational Journal of Psychiatric Research, 41, 814–​820.
roles for dopamine in behavioural control. Nature, 431, Nakai, Y., & Koh, T. (2001). Perception of hunger to insulin‐
760–​767. induced hypoglycemia in anorexia nervosa. International
Monteleone, A., Monteleone, P., Serino, I., Amodio, R., Journal of Eating Disorders, 29, 354–​357.
Monaco, F., & Maj, M. (2016). Underweight subjects with Naruo, T., Nakabeppu, Y., Sagiyama, K., Munemoto, T., Homan,
anorexia nervosa have an enhanced salivary cortisol response N., Deguchi, D.,   .  .  .  Nozoe, S. (2000). Characteristic
not seen in weight restored subjects with anorexia nervosa. regional cerebral blood flow patterns in anorexia nervosa
Psychoneuroendocrinology, 70, 118–​121. patients with binge/​purge behavior. The American Journal of
Monteleone, P., DiLieto, A., Castaldo, E., & Maj, M. (2004). Psychiatry, 157, 1520–​1522.
Leptin functioning in eating disorders. CNS Spectrums, 9, Neumeister, A., Brain, E., Nugent, A., Carson, R., Bonne, O.,
523–​529. Lucnekbaugh, D.,  . . .  Drevets, W. C. (2004). Reduced sero-
Monteleone, P., Di Lieto, A., Tortorella, A., Longobardi, N., tinin type 1A receptor binding in panic disorder. Journal of
& Maj, M. (2000). Circulating leptin in patients with Neuroscience, 24, 589–​591.
anorexia nervosa, bulimia nervosa or binge-​ eating disor- Nisoli, E., Brunani, A., Borgomainerio, E., Tonello, C., Dioni, L.,
der:  Relationship to body weight, eating patterns, psycho- Briscini, L.,  . . .  Carruba, M. O. (2007). D2 dopamine recep-
pathology and endocrine changes. Psychiatry Research, 94, tor (DRD2) gene Taq1A polymorphism and the eating-​related
121–​129. psychological traits in eating disorders (anorexia nervosa and
Monteleone, P., & Maj, M. (2013). Dysfunctions of leptin, bulimia) and obesity. Eating and Weight Disorders, 12, 91–​96.
ghrelin, BDNF and endocannabinoids in eating disor- Nozoe, S., Naruo, T., Nakabeppu, Y., Soejima, Y., Nakajo, M., &
ders:  Beyond the homeostatic control of food intake. Tanaka, H. (1993). Changes in regional cerebral blood flow
Psychoneuroendocrinology, 38, 213–​333. in patients with anorexia nervosa detected through single
Monteleone, P., Martiadis, V., Fabrazzo, M., Serritella, C., photon emission tomography imaging. Biological Psychiatry,
& Maj, M. (2003). Ghrelin and leptin responses to food 34, 578–​580.

 Ely, Kaye 75
Nozoe, S., Naruo, T., Yonekura, R., Nakabeppu, Y., Soejima, Y., and perfectionism in anorexia nervosa. Brain and Cognition,
Nagai, N.,  . . .  Tanaka, H. (1995). Comparison of regional 63, 42–​50.
cerebral blood flow in patients with eating disorders. Brain Pirke, K. M., Kellner, M. B., Friess, E., Krieg, J. C., & Fichter,
Research Bulletin, 36, 251–​255. M. M. (1994). Satiety and cholecystokinin. International
Nunn, K., Frampton, I., & Lask, B. (2012). Anorexia ner- Journal of Eating Disorders, 15, 63–​69.
vosa:  A  noradrenergic dysregulation hypothesis. Medical Pollatos, O., & Georgiou, E. (2016). Normal interoceptive accu-
Hypotheses, 78, 580–​584. racy in women with bulimia nervosa. Psychiatry Research,
Oberndorfer, T., Frank, G., Simmons, A. N., Wagner, A., 240, 328–​332.
McCurdy, D., Fudge, J. L.,  . . .  Kaye, W. H. (2013). Altered Pollatos, O., Kurz, A.-​L., Albrecht, J., Schreder, T., Kleemann,
insula response to sweet taste processing after recovery from A., Schopf, V.,  . . .  Schandry, R. (2008). Reduced percep-
anorexia and bulimia nervosa. The American Journal of tion of bodily signals in anorexia nervosa. Eating Behaviors,
Psychiatry, 214, 132–​141. 9, 381–​388.
Oberndorfer, T., Kaye, W., Simmons, A., Strigo, I., & Matthews, Pollice, C., Kaye, W. H., Greeno, C. G., & Weltzin, T. E. (1997).
S. (2011). Demand-​specific alteration of medial prefrontal Relationship of depression, anxiety, and obsessionality to
cortex response during an inhibition task in recovered anorexic state of illness in anorexia nervosa. International Journal of
women. International Journal of Eating Disorders, 44, 1–​8. Eating Disorders, 21, 367–​376.
Oberndorfer, T., Simmons, A., McCurdy, D., Strigo, I., Prince, A., Brooks, S. J., Stahl, D., & Treasure, J. (2009).
Matthews, S., Yang, T.,  . . .  Kaye, W. (2013). Greater ante- Systematic review and meta-​analysis of the baseline con-
rior insula activation during anticipation of food images in centrations and physiologic responses of gut hormones to
women recovered from anorexia nervosa versus controls. food in eating disorders. The American Journal of Clinical
Psychiatry Research, 214, 132–​141. Nutrition, 89, 755–​765.
O’Doherty, J. (2004). Reward representations and reward-​related Radeloff, D., Willmann, K., Otto, L., Lindner, M., Putnam,
learning in the human brain:  Insights from neuroimaging. K., van Leeuwen, S.,  . . .  Wagner, A. (2014). High-​fat taste
Current Opinion in Neurobiology, 14, 769–​776. challenge reveals altered striatal response in women recov-
O’Doherty, J., Dayan, P., Schultz, J., Deichmann, R., Friston, ered from bulimia nervosa: A pilot study. The World Journal
K. J., & Dolan, R. J. (2004). Dissociable roles of ventral and of Biological Psychiatry, 15, 307–​316. [Epub ahead of print]
dorsal striatum in instrumental conditioning. Science, 304, Reynolds, S., & Zahm, D. (2005). Specificity in the projections
452–​454. of prefrontal and insular cortex to ventral striatopallidum
O’Doherty, J., Kringelbach, M. L., Rolls, E. T., Hornak, J., and the extended amygdala. Journal of Neuroscience, 25,
& Andrews, C. (2001). Abstract reward and punishment 11757–​11767.
representations in the human orbitofrontal cortex. Nature Richer, M., Hen, R., & Blier, P. (2002). Modification of sero-
Neuroscience, 4, 95–​102. tonin neuron properties in mice lacking 5-​HT1A receptors.
O’Doherty, J., Rolls, E. T., Francis, S., Bowtell, R., McGlone, European Journal of Pharmacology, 435, 195–​203.
F., Kobal, G.,  . . .  Ahne, G. (2000). Sensory-​specific satiety-​ Rolls, E. T. (1997). Taste and olfactory processing in the brain
related olfactory activation of the human orbitofrontal cor- and its relation to the control of eating. Critical Review in
tex. Neuroreport, 11, 893–​897. Neurobiology, 11, 263–​287.
Ogawa, H. (1994). Gustatory cortex of primates: Anatomy and Rolls, E. T. (2005). Taste, olfactory, and food texture processing
physiology. Neuroscience Research, 20, 1–​13. in the brain, and the control of food intake. Physiology &
Ongur, D., & Price, J. L. (2000). Organization of networks Behavior, 85, 45–​56.
within the orbital and medial prefrontal cortex of rats, mon- Rutherford, J., McGuffin, P., Katz, R. J., & Murray, R. M.
keys, and humans. Cerebral Cortex, 10, 206–​219. (1993). Genetic influences on eating attitudes in a normal
Otto B., Cuntz U., Otto C., Heldwein W., Riepl R. L., Tschöp female twin population. Psychological Medicine, 23, 425–​436.
M. H. (2007). Peptide YY release in anorectic patients after Salamone, J. D. (1996). The behavioral neurochemistry of
liquid meal, Appetite, 48(3), 301–​304. motivation:  Methodological and conceptual issues in stud-
Parsey, R. V., Oquendo, M. A., Ogden, R. T., Olvet, D., ies of the dynamic activity of nucleus accumbens dopamine.
Simpson, N., Huang, Y.,  . . .  Mann, J. J. (2005). Altered Journal of Neuroscience Methods, 64, 137–​149.
serotonin 1A binding in major depression:  A  [carbonyl-​C-​ Sam, A., Troke, C., Tan, T., & Bewick, G. (2012). The role of the
11]WAY100635 positron emission tomography study. Biol gut/​brain axis in modulating food intake. Neuropharmacology,
Psychiatry, 59, 106–​113. 63, 46–​56.
Paulus, M., & Stein, M. B. (2006). An insular view of anxiety. Santel, S., Baving, L., Krauel, K., Munte, T., & Rotte, M.
Biological Psychiatry, 60, 383–​387. (2006). Hunger and satiety in anorexia nervosa: fMRI dur-
Pearson, C. M., Zapolski, T. C., & Smith, G. T. (2015). A longi- ing cognitive processing of food pictures. Brain Res, 1114,
tudinal test of impulsivity and depression pathways to early 138–​148.
binge eating onset. International Journal of Eating Disorders, Saper, C. B., Chou, T. C., & Elmquist, J. K. (2002). The need
48, 230–​237. to feed: Homeostatic and hedonic control of eating. Neuron,
Phillipp, E., Pirke, K. M., Kellner, M. B., & Krieg, J. C. (1991). 36, 199–​211.
Disturbed cholecystokinin secretion in patients with eating Sargent, P. A., Kjaer, K. H., Bench, C. J., Rabiner, E. A., Messa,
disorders. Life Sciences, 48, 2443–​2450. C., Meyer, J.,  . . .  Cowen, P. J. (2000). Brain serotonin1A
Phillips, M., Drevets, W., Rauch, S. L., & Lane, R. (2003). receptor binding measured by positron emission tomography
Neurobiology of emotion perception I:  The neural basis of with [11C]WAY-​100635: Effects of depression and antidepres-
normal emotion perception Biological Psychiatry, 54, 504–​514. sant treatment. Archives of General Psychiatry, 57, 174–​180.
Pieters, G., de Bruijn, E., Maas, Y., Hulstijn, W., Vandereycken, Sato, Y., Saito, N., Utsumi, A., Aizawa, E., Shoji, T., Izumiyama,
W., Pueskens, J., & Sabbe, B. (2007). Action monitoring M.,   .  .  .  Fukudo, S. (2013). Neural basis of impaired

76 Appetitive Regul ation

 cognitive flexibility in patients with anorexia nervosa. Plos Silberg, J., & Bulik, C. (2005). Developmental association
One, 8, e61108. between eating disorders symptoms and symptoms of
Saudou, F., & Hen, R. (1994). 5-​Hydroxytryptamine receptor depression and anxiety in juvenile twin girls. Journal of Child
subtypes in vertebrates and invertebrates. Neurochemistry Psychology and Psychiatry, 46, 1317–​1326.
International, 25, 503–​532. Simansky, K. J. (1996). Serotonergic control of the organiza-
Schienle, A., Schafer, A., Hermann, A., & Vaitl, D. (2009). tion of feeding and satiety. Behavioural Brain Research,
Binge-​eating disorder:  Reward sensitivity and brain activa- 73(1–​2), 37–​42.
tion to images of food. Biological Psychiatry, 65, 654–​661. Simmons, A., Strigo, I., Matthews, S., Paulus, M., & Stein, M.
Schoenfeld, M., Neuer, G., Tempelmann, C., Schussler, K., (2006). Anticipation of aversive visual stimuli is associated
Noesselt, T., Hopf, J.,  . . .  Heinze, H. J. (2004). Functional with increased insula activation in anxiety-​prone subjects.
magnetic resonance tomography correlates of taste percep- Biological Psychiatry, 60, 402–​409.
tion in the human primary taste cortex. Neuroscience, 127, Simon, S., De Araujo, I., Gutierrez, R., & Nicolelis, M. (2006).
347–​353. The neural mechanisms of gustation: A distributed process-
Schultz, W. (1998). Predictive reward signal of dopamine neu- ing code. Nature Reviews Neuroscience, 7, 890–​901.
rons. Journal of Neurophysiology, 80, 1–​27. Small, D. (2002). Toward an understanding of the brain sub-
Schultz, W. (2004). Neural coding of basic reward terms of strates of reward in humans. Neuron, 22, 668–​671.
animal learning theory, game theory, microeconomics and Small, D. (2006). Central gustatory processing in humans.
behavioural ecology. Current Opinion in Neurobiology, 14, Advances in Oto-​Rhino-​Laryngology, 63, 191–​220.
139–​147. Small, D. (2009). Individual differences in the neurophysiology
Schultz, W., Tremblay, L., & Hollerman, J. (1998). Reward of reward and the obesity epidemic. International Journal of
prediction in primate basal ganglia and frontal cortex. Obesity, 33(Suppl 2), S44–​S48.
Neuropharmacology, 37, 421–​429. Small, D., Jones-​Gotman, M., & Dagher, A. (2003). Feeding-​
Schultz, W., Tremblay, L., & Hollerman, J. R. (2000). Reward induced dopamine release in dorsal striatum correlates with
processing in primate orbitofrontal cortex and basal ganglia. meal pleasantness ratings in healthy human volunteers.
Cerebral Cortex, 10, 272–​284. Neuroimage, 19, 1709–​1715.
Schwartz, M. W., Peskind, E., Raskind, M., Boyko, E. J., Small, D., Zatorre, R., Dagher, A., Evans, A., & Jones-​Gotman,
& Porte, D., Jr. (1996). Cerebrospinal fluid leptin lev- M. (2001). Changes in brain activity related to eating choco-
els: Relationship to plasma levels and to adiposity in humans. late: from pleasure to aversion. Brain, 124, 1720–​1733.
Nature Medicine, 2, 589–​593. Smyth, J., Wonderlich, S., Heron, K., Sliwinski, M., Crosby,
Schwartz, M. W., Woods, S. C., Porte, D., Jr., Seeley, R. J., & R., Mitchell, J., & Engel, S. G. (2007). Daily and momen-
Baskin, D. G. (2000). Central nervous system control of tary mood and stress are associated with binge eating and
food intake. Nature, 404, 661–​671. vomiting in bulimia nervosa patients in the natural envi-
Scott, T. R., Yaxley, S., Sienkiewicz, Z., & Rolls, E. (1986). ronment. Journal of Consulting and Clinical Psychology, 75,
Gustatory responses in the frontal opercular cortex of the alert 629–​638.
cynomolgus monkey. Journal of Neurophysiology, 56, 876–​890. Somerville, L., Hare, T., & Casey, B. (2011). Frontostriatal
Sedlackova, D., Kopeckova, J., Papezova, H., Vybiral, S., maturation predicts cognitive control failure to appetitive
Kvasnickova, H., Hill, M., & Nedvídková, J. (2011). cues in adolescents. Journal of Cognitive Neuroscience, 23,
Changes of plasma obestatin, ghrelin and NPY in 2123–​2134.
anorexia and bulimia nervosa patients before and after a Somerville, L. H., & Casey, B. H. (2010). Developmental neu-
high-​carbohydrate breakfast. Physiological Research, 60, robiology of cognitive control and motivational systems.
165–​173. Current Opinion in Neurobiology, 20, 236–​241.
Shiiya, T., Nakazato, M., Mizuta, M., Date, Y., Mondal, M. S., Soubrie, P. (1986). Reconciling the role of central serotonin neu-
Tanaka, M.,  . . .  Matsukura, S. (2002). Plasma ghrelin lev- rons in human and animal behavior. Behavioral and Brain
els in lean and obese humans and the effect of glucose on Sciences, 9, 319–​335.
ghrelin secretion. Journal of Endocrinology and Metabolism, Spielberger, C., Gorsuch, R., & Lushene, R. (1970). STAI
87, 240–​244. Manual for the State Trait Anxiety Inventory. Palo Alto,
Shirao, N., Okamoto, Y., Okada, G., Okamoto, Y., & Yamawaki, CA: Consulting Psychologists Press.
S. (2003). Temporomesial activation in young females asso- Srinivasagam, N. M., Kaye, W. H., Plotnicov, K. H., Greeno,
ciated with unpleasant words concerning body image. C., Weltzin, T. E., & Rao, R. (1995). Persistent perfection-
Neuropsychobiology, 48, 136–​142. ism, symmetry, and exactness after long-​term recovery from
Shively, C., Friedman, D., Gage, H., Bounds, M., Brown-​ anorexia nervosa. The American Journal of Psychiatry, 152,
Proctor, C., Blair, J.,  . . .  Buchheimer, N. (2006). Behavioral 1630–​1634.
depression and positron emission tomography-​determined Staley, J., Malison, R., & Innis, R. (1998). Imaging of the sero-
serotonin 1A receptor binding potential in cynomolgus tonergic system: Interactions of neuroanatomical and func-
monkeys. Archives of General Psychiatry, 63, 396–​403. tional abnormalities of depression. Biological Psychiatry, 44,
Shott, M., Pryor, T., Yang, T., & Frank, G. (2016). Greater insula 534–​549.
white matter fiber connectivity in women recovered from Steiger, H., Gauvin, L., Israel, M., Kin, N. M., Young, S. N., &
anorexia nervosa. Neuropsychopharmacology, 41, 498–​507. Roussin, J. (2004). Serotonin function, personality-​trait vari-
Sibille, E., Pavlides, C., Benke, D., & Toth, M. (2000). Genetic ations and childhood abuse in women with bulimia-​spectrum
inactivation of the Serotonin(1A) receptor in mice results in eating disorders. Journal of Clinical Psychiatry, 65, 830–​837.
downregulation of major GABA(A) receptor alpha subunits, Steinberg, L., Albert, D., Cauffman, E., Banich, M., Graham, S.,
reduction of GABA(A) receptor binding, and benzodiazepine-​ & Woolard, J. (2008). Age differences in sensation seeking and
resistant anxiety. Journal of Neuroscience, 20, 2758–​2765. impulsivity as indexed by behavior and self-​report: Evidence

 Ely, Kaye 77
 for a dual systems model. Developmental Psychology, 44, in automatic information processing. Neuropsychobiology,
1764–​1778. 15, 89–​94.
Steinglass, J., Sysko, R., Mayer, L., Berner, L., Schebendach, J., Suzuki, K., Simpson, K., Minnion, J., Shillito, J., & Bloom, S.
Wang, Y.,  . . .  Walsh, B. T. (2010). Pre-​meal anxiety and (2010). The role of gut hormones and the hypothalamus in
food intake in anorexia nervosa. Appetite, 55, 214–​218. appetite regulation. Endocrine Journal, 57, 359–​372.
Steinhausen, H. C. (2002). The outcome of anorexia nervosa in Szabo, S. T., & Blier, P. (2001). Serotonin (1A) receptor ligands
the 20th century. The American Journal of Psychiatry, 159, act on norepinephrine neuron firing through excitatory
1284–​1293. amino acid and GABA(A) receptors: A microiontophoretic
Stice, E. (2002). Risk and maintenance factors for eating pathol- study in the rat locus coeruleus. Synapse, 42, 203–​212.
ogy:  A  meta-​analytic review. Psychopharmacology Bulletin, Tamai, H., Takemura, J., Kobayashi, N., Matsubayashi, S.,
128, 825–​848. Matsukura, S., & Nakagawa, T. (1993). Changes in plasma
Stice, E., Burger, K., & Yokum, S. (2013). Caloric deprivation cholecystokinin concentrations after oral glucose toler-
increases responsivity of attention and reward brain regions ance test in anorexia nervosa before and after therapy.
to intake, anticipated intake, and images of palatable foods. Metabolism: Clinical and Experimental, 42, 581–​584.
Neuroimage, 67, 322–​330. Tanaka, M., Naruo, T., Muranaga, T., Yasuhara, D., Shiiya,
Stice, E., Davis, K., Miller, N., & Marti, C. (2008). Fasting T., Nakazato, M.,   .  .  .  Nozoe, S. (2002). Increased fast-
increases risk for onset of binge eating and bulimic pathol- ing plasma ghrelin levels in patients with bulimia nervosa.
ogy:  A  5-​ year prospective study. Journal of Abnormal European Journal of Endocrinology, 146, R1–​R3.
Psychology, 117, 941–​946. Tataranni, P. A., Gautier, J. F., Chen, K., Uecker, A., Bandy, D.,
Stice, E., Martinez, E., Presnell, K., & Groesz, L. (2006). Salbe, A. D.,   .  .  .  Ravussin, E. (1999). Neuroanatomical
Relation of successful dietary restriction to change in bulimic correlates of hunger and satiation in humans using positron
symptoms:  A  prospective study of adolescent girls. Health emission tomography. Proceedings of the National Academy of
Psychology, 25, 274–​281. Sciences, 96(8), 4569–​4574.
Stice, E., Presnell, K., Groesz, L., & Shaw, H. (2005). Effects of Tchanturia, K., Morris, R. G., Anderluh, M. B., Collier, D. A.,
a weight maintenance diet on bulimic symptoms in adoles- Nikolaou, V., & Treasure, J. (2004). Set shifting in anorexia
cent girls: An experimental test of the dietary restraint theory. nervosa: An examination before and after weight gain, in full
Health Psychology, 24, 402–​412. recovery and relationship to childhood and adult OCPD
Stice, E., Spoor, S., Ng, J., & Zald, D. (2009). Relation of obesity traits. Journal of Psychiatric Research, 38, 545–​552.
to consummatory and anticipatory food reward. Physiology & Telch, C. F., & Agras, W. S. (1996). The effects of short-​term
Behavior, 97, 551–​560. food deprivation on caloric intake in eating-​disordered sub-
Stock, S., Leichner, P., Wong, A., Ghatei, M., Kieffer, T., Bloom, ject. Appetite, 26, 221–​234.
S., & Chanoine, J. P. (2005). Ghrelin, peptide YY, glucose-​ Thiels, C., & Patel, J. (2008). Survey of disordered eating and
dependent insulinotropic polypeptide, and hunger responses behaviour in children and adolescents. Zeitschrift für Kinder-​
to a mixed meal in anorexic, obese, and control female ado- und Jugendpsychiatrie und Psychotherapie, 36, 265–​274.
lescents. Journal of Clinical Endocrinology & Metabolism, 90, Tiihonen, J., Keski-​Rahkonen, A., Lopponen, M., Muhonen,
2161–​2168. M., Kajander, J., Allonen, T.,  . . .  Rissanen, A. (2004). Brain
Stockmeier, C. A. (1997). Neurobiology of serotonin in depres- serotonin 1A receptor binding in bulimia nervosa. Biological
sion and suicide. Annals of the New York Academy of Sciences, Psychiatry, 55, 871–​873.
836, 220–​232. Tortorella, A., Brambilla, F., Fabrazzo, M., Volpe, U.,
Stoving, R. K., Hangaard, J., Hansen-​Nord, M., & Hagen, C. Monteleone, A., Mastromo, D., & Monteleone, P. (2014).
(1999). A review of endocrine changes in anorexia nervosa. Central and peripheral peptides regulating eating behaviour
Journal of Psychiatric Research, 33, 139–​152. and energy homeostasis in anorexia nervosa and bulimia ner-
Strigo, I., Matthews, S., Simmons, A., Oberndorfer, T., Klabunde, vosa: A literature review. European Eating Disorders Review,
M., Reinhardt, L., & Kaye, W. H. (2013). Altered insula acti- 22, 307–​320.
vation during pain anticipation in individuals recovered from Trace, S., Baker, J., Peñas-​Lledó, E., & Bulik, C. (2013). The
anorexia nervosa:  Evidence of interoceptive dysregulation. genetics of eating disorders. Annual Review of Clinical
International Journal of Eating Disorders, 46, 23–​33. Psychology, 9, 589–​620.
Strober, M. (1980). Personality and symptomatological features Treasure, J., & Campbell, I. (1994). The case for biology in
in young, nonchronic anorexia nervosa patients. Journal of the aetiology of anorexia nervosa. Psychological Medicine,
Psychosomatic Research, 24, 353–​359. 24, 3–​8.
Strober, M., Freeman, R., Lampert, C., Diamond, J., & Kaye, Tricomi, E. M., Delgado, M. R., & Fiez, J. A. (2004). Modulation
W. (2000). Controlled family study of anorexia nervosa and of caudate activity by action contingency. Neuron, 41,
bulimia nervosa:  Evidence of shared liability and transmis- 281–​292.
sion of partial syndromes. The American Journal of Psychiatry, Troisi, A., Di Lorenzo, G., Lega, I., Tesauro, M., Bertoli, A., Leo,
157, 393–​401. R.,  . . .  Siracusano, A. (2005). Plasma ghrelin in anorexia,
Strober, M., Freeman, R., & Morrell, W. (1997). The long-​term bulimia, and binge-​eating disorder:  Relations with eating
course of severe anorexia nervosa in adolescents:  Survival patterns and circulating concentrations of cortisol and thy-
analysis of recovery, relapse, and outcome predictors over roid hormones. Neuroendocrine, 81, 259–​266.
10–​15 years in a prospective study. International Journal of Tschop, M., Wawarta, R., Reiepl, R., Friedrich, S., Bidlingmaier,
Eating Disorders, 22, 339–​360. M., Landgraf, R., & Folwaczny, C. (2001). Post-​prandial
Strupp, B. J., Weingartner, H., Kaye, W., & Gwirtsman, H. (1986). decrease of circulating human ghrelin levels. Journal of
Cognitive processing in anorexia nervosa:  A  disturbance Endocrinological Investigation, 24, RC19–​RC21.

78 Appetitive Regul ation

Uher, R., Brammer, M., Murphy, T., Campbell, I., Ng, V., emotional state? Behaviour Research and Therapy, 39,
Williams, S., & Treasure, J. (2003). Recovery and chronicity 877–​886.
in anorexia nervosa: Brain activity associated with differential Weltzin, T. E., Hsu, L. K., Pollice, C., & Kaye, W. H. (1991).
outcomes. Biological Psychiatry, 54, 934–​942. Feeding patterns in bulimia nervosa. Biological Psychiatry, 30,
Uher, R., Murphy, T., Brammer, M., Dalgleish, T., Phillips, M., 1093–​1110.
Ng, V.,   .  .  .  Treasure, J. (2004). Medial prefrontal cortex Westergaard, G., Suomi, S., Chavanne, T., Houser, L.,
activity associated with symptom provocation in eating dis- Hurley, A., Cleveland, A.,   .  .  .  Higley, J. D. (2003).
orders. The American Journal of Psychiatry, 161, 1238–​1246. Physiological correlates of aggression and impulsivity in
Uher, R., Treasure, J., Heining, M., Brammer, M. J., & free-​
ranging female primates. Neuropsychopharmacology,
Campbell, I. C. (2006). Cerebral processing of food-​related 28, 1045–​1055.
stimuli:  effects of fasting and gender. Behavioural Brain Weygandt, M., Schaefer, A., Schienle, A., & Haynes, J. (2012).
Research, 169(1), 111–​119. Diagnosing different binge-​eating disorders based on reward-​
van Heeringen, C., Audenaert, K., Van Laere, K., Dumont, related brain activation patterns. Human Brain Mapping, 33,
F., Slegers, G., Mertens, J.,   .  .  .  Dierckx, R. A. (2003). 2135–​2146.
Prefrontal 5-​HT2a receptor binding index, hopelessness and Wierenga, C., Bischoff-​ Grethe, A., Melrose, A., Grenesko-​
personality characteristics in attempted suicide. Journal of Stevens, E., Irvine, Z., Wagner, A.,  . . .  Kaye, W. H. (2014).
Affective Disorders, 74, 149–​158. Altered BOLD response during inhibitory and error pro-
Van Leijenhorst, L., Zanolie, K., Van Meel, C., Westenberg, cessing in adolescents with anorexi anervosa. Plos One, 9,
P., Rombouts, S., & Crone, E. (2010). What motivates the 392017.
adolescent? Brain regions mediating reward sensitivity across Wierenga, C., Bischoff-​ Grethe, A., Melrose, A., Irvine, Z.,
adolescence. Cerebral Cortex, 20, 61–​69. Torres, L., Bailer, U.,  . . .  Kaye, W. H. (2015). Hunger does
Vocks, S., Herpertz, S., Rosenberger, C., Senf, W., & Gizewski, not motivate reward in women remitted from anorexia ner-
E. (2011). Effects of gustatory stimulation on brain activ- vosa. Biological Psychiatry, 77, 642–​652.
ity during hunger and satiety in females with restricting-​ Wierenga, C., Ely, A., Bischoff-​ Grethe, A., Bailer, U.,
type anorexia nervosa: An fMRI study. Journal of Psychiatric Simmons, A., & Kaye, W. H. (2014). Are extremes of
Research, 45, 395–​403. consumption in eating disorders related to an altered bal-
Wade, T., Martin, N. G., & Tiggemann, M. (1998). Genetic and ance between reward and inhibition? Frontiers Behavioural
environmental risk factors for the weight and shape concerns Brain Research, 9, 410.
characteristic of bulimia nervosa. Psychological Medicine, 28, Wolfe, B., Jimerson, D., Orlova, C., & Mantzoros, C. (2004).
761–​771. Effect of dieting on plasma leptin, soluble leptin receptor,
Wagner, A., Aizeinstein, H., Venkatraman, V., Bischoff-​Grethe, adiponectin and resistin levels in healthy volunteers. Clinical
A., Fudge, J., May, J.,   .  .  .  Kaye, W. H. (2010). Altered Endocrinology, 61, 332–​338.
striatal response to reward in bulimia nervosa after recovery. Wolfe, B., Metzger, E., Levine, J., Finkelstein, D., Cooper, T., &
International Journal of Eating Disorders, 43, 289–​294. Jimerson, D. (2000). Serotonin function following remission
Wagner, A., Aizenstein, H., Frank, G. K., Figurski, J., May, J. from bulimia nervosa. Neuropsychopharmacology, 22, 257–​263.
C., Putnam, K.,  . . .  Kaye, W. H. (2006). Neural correlates Wren, A., Seal, L., Cohen, M., Byrnes, A., Front, G., Murphy,
of habituation to taste stimuli in healthy women. Psychiatry K.,   .  .  .  Bloom, S. R. (2001). Ghrelin enhances appetite
Research, 147, 57–​67. and increases foot intake in humans. Journal of Clinical
Wagner, A., Aizenstein, H., Frank, G. K., Figurski, J., May, J. C., Endocrinology & Metabolism, 86, 5992–​5995.
Putnam, K.,  . . .  Kaye, W. H. (2008). Altered insula response Yaxley, S., Rolls, E., & Sienkiewicz, Z. (1990). Gustatory
to a taste stimulus in individuals recovered from restricting-​type responses of single neurons in the insula of the macaque
anorexia nervosa. Neuropsychopharmacology, 33, 513–​523. monkey. Journal of Neurophysiology, 63, 689–​700.
Wagner, A., Aizenstein, H., Venkatraman, M., Fudge, J., May, J., Yin, H., & Knowlton, B. (2006). The role of the basal ganglia in
Mazurkewicz, L.,  . . .  Kaye, W. H. (2007). Altered reward habit formation. Nature Reviews Neuroscience, 7, 464–​476.
processing in women recovered from anorexia nervosa. The Zastrow, A., Kaiser, S. S., C, Walthe, S., Herzog, W., Tchanturia,
American Journal of Psychiatry, 164, 1842–​1849. K.,   .  .  .  Friederich, H. C. (2009). Neural correlates of
Wagner, A., Barbarich, N., Frank, G., Bailer, U., Wonderlich, impaired cognitive-​ behavioral flexibility in anorexia ner-
S. A., Crosby, R. D.,  . . .  Kaye, W. H. (2006). Personality vosa. The American Journal of Psychiatry, 166, 608–​616.
traits after recovery from eating disorders: Do subtypes dif- Zhang, Y., Proenca, R., Maffei, M., Barone, M., Leopold, L., &
fer? International Journal of Eating Disorders, 39, 276–​284. Friedman, J. (1994). Positional cloning of the mouse obese gene
Walsh, B. T., Roose, S. P., Katz, J. L., Dyrenfurth, I., Wright, L., and its human homologue. Nature Genetics, 372, 425–​432.
Vande Wiele, R., & Glassman, A. H. (1987). Hypothalamic-​ Zigman, J., & Elmquist, J. (2003). Minireview: From anorexia
pituitary-​adrenal-​cortical activity in anorexia nervosa and to obesity—​ the yin and yang of body weight control.
bulimia. Psychoneuroendocrinology, 12, 131–​140. Endocrinology, 144, 3749–​3756.
Walters, E. E., & Kendler, K. S. (1995). Anorexia nervosa and Zucker, N., Merwin, R., Bulik, C., Moskovich, A., Wildes,
anorexic-​like syndromes in a population-​based female twin J., & Groh, J. (2013). Subjective experience of sensation
sample. American Journal of Psychiatry, 152, 64–​71. in anorexia nervosa. Behaviour Research and Therapy, 51,
Wang, G., Volkow, N. D., Telang, F., Jayne, M., Ma, J., Rao, M.,  . . .  256–​265.
Fowler, J. S. (2004). Exposure to appetitive food stimuli mark- Zunker, C., Peterson, C., Crosby, R., Cao, L., Engel, S., Mitchell,
edly activates the human brain. Neuroimage, 21, 1790–​1797. J., & Wonderlich, S. A. (2011). Ecological momentary assess-
Waters, A., Hill, A., & Waller, G. (2001). Bulimics’ responses ment of bulimia nervosa: Does dietary restriction predict binge
to food cravings:  Is binge-​eating a product of hunger or eating? Behavioural Brain Research and Therapy, 49, 714–​717.

 Ely, Kaye 79
 C H A PT E R


Genetic Influences on Eating Disorders
5
Tracy D. Wade and Cynthia M. Bulik

Abstract
The current chapter reviews our progress in understanding how genes influence eating disorders by
addressing the following areas: (1) how recognition of genetic influences on eating disorders emerged;
(2) the complexities of gene environment interplay; (3) what twin studies can tell us about gene
environment interplay, and (4) the current state of molecular genetic studies. It is concluded that both
genes and nonshared environment play a critical role in the explanatory framework for the etiology
of eating disorders. Shared environment is likely to contribute to the development of cognition and
attitudes that may initiate disordered eating practices. Researchers are on the cusp of identifying specific
genes that are implicated, and explication of the manner in which genes and the environment work
together to increase risk for eating disorders hinges on the collection of larger samples.
Key Words:  heritability, gene × environment interaction, twin study, molecular genetic study, nonshared
environment

Glossary of Technical Terms Endophenotype

Allele Endophenotypes are measurable compo-
An allele is a viable DNA (deoxyribonucleic nents unseen by the unaided eye along the
acid) coding that occupies a given locus (position) pathway between disease and distal genotype.
on a chromosome. Usually alleles are sequences that Endophenotypes are defined as being (1) heritable,
code for a gene. Humans have two alleles for each (2) cosegregating with a disease trait in the general
trait. If the same allele is present twice, the person population, (3)  state-​independent (i.e., manifest
is homozygous for this characteristic. If, however, in the individual whether or not illness is active),
one chromosome contains one allele and the other (4)  found in nonaffected family members at a
chromosome a contrasting allele, the individual is higher rate than in the general population.
heterozygous. In conditions of heterozygosity, the
phenotype (appearance) may be determined by one Epigenesis
allele and not the other. The allele that determines Epigenesis involves the influence of the envi-
the phenotype is said to be dominantly expressed; ronment on the expression of the genetic code.
it shows dominance over the other allele. The Information transmitted to subsequent generations
expression of the other allele is described as being that does not involve changes in the sequences of
recessive. DNA as gene expression can be modulated with-
out actually changing the DNA but by methylation
Chromosome (which usually silences genes) or any of several other
A structure that contains DNA and resides in the possible ways of changing the ease with which the
nucleus of cells. environment can activate a specific gene.

80
Genome 7-​to 12-​fold increase in the prevalence of AN or
All of the DNA of an organism for one member BN in relatives of eating disordered probands com-
of each chromosome pair. The human genome con- pared to the families of controls (Klump, Kaye, &
tains about 3 billion DNA base pairs. Strober, 2001).
The question that these findings pose is whether
Genotype this increased familial risk relates to the impact of
The genetic constitution of an individual, or the the environment that is shared in families (which
combination of alleles at a particular locus. does not necessarily indicate family environment
but any sources of the environment that may be
Linkage experienced equally by family members), to genetic
Close proximity of loci on a chromosome. influences, or a to combination of both. Evidence
from the last 25 years of research has supported a
Location of Genes substantial genetic contribution to the development
There are 23 pairs of human chromosomes. At of eating disorders. This finding is broadly consis-
some point in each chromosome there is a cen- tent with the emerging research across a number
tromere, a region without genes, where the chro- of major psychiatric disorders and psychopathol-
mosome is attached to its new copy when cells ogy, including emotional and behavioral distur-
reproduce. The short arm of the chromosome above bances (Bouchard & McGue, 2003; Kendler, 2013;
the centromere is called p and the long arm below Cross-​Disorder Group of the Psychiatric Genomics
the centromere is called q. The location of genes is Consortium, 2013).
described in relation to the bands. For example, a The acceptance of genetic influences on behav-
gene at 4p16 means the short arm of chromosome 4 ioral disturbances has been called “one of the
at a particular band, number 6 in region 1. most dramatic shifts in the modern history of the
behavioral sciences” (Plomin, 2000). However,
Phenotype understanding the way in which genetics influence
Characteristics of an individual that result the development of eating disorders is a complex
from interactions between the genotype and the task. Hundreds and perhaps thousands of genes are
environment. hypothesized to exert their influence on the cen-
tral nervous system and possibly metabolism; to
Polymorphisms interact with and be modified by environmental
New DNA differences occur when mutations variables; and potentially to be mediated via psy-
occur when copying DNA—​ these mutations chological variables (Hewitt, 1997). Twin studies
result in different alleles responsible for variations are considered to be a vital and essential companion
in the phenotype. These different alleles are called to molecular genetic investigations (Lyons & Bar,
polymorphisms. 2001; Neiderhiser, 2001; Kendler, 2013), which
together allow us to develop a more sophisticated
Serotonin understanding of the complex interplay between
Also known as 5-​hydroxytryptamine, or 5-​HT, different influences that lead to the emergence and
serotonin is a monoamine neurotransmitter synthe- maintenance of eating disorders.
sized in serotonergic neurons in the central nervous This complexity challenges our ability to dis-
system. seminate the science that now unequivocally
implicates genetic factors in eating disorders
Overview aetiology. Many have been unwilling to relin-
It has been known for some time that eating quish antiquated hypotheses that implicate par-
disorders “run in families,” with an early review of enting. For example, a recent review of the area
published family and twin studies of anorexia ner- concluded, “The decoding of the human genome
vosa (AN) and bulimia nervosa (BN) concluding has made possible research on the genetics of
that four of the five studies indicated elevated risk anorexia. This has failed to confirm the expected
in the family members of affected individuals, or relationship. A  number of leaders in the field of
probands (Spelt & Meyer, 1995). It was estimated family therapy advocate the idea that family rela-
that there was a two-​to threefold risk of developing tionships do not cause anorexia. However, the
an eating disorder in females if a first-​degree rela- failure of genetic explanations draws attention
tive was affected. A subsequent review suggested a back to the possible role of family factors” (Dring,

Wade, Bulik 81
2015, p.  79). A  survey of health professionals, and colleagues (1995) found that reports of fam-
the general public, and female university students ily dysfunction were significantly worse from BN
about the causes of BN showed four underlying probands than from their respective family mem-
dimensions:  (1)  sociocultural pressure, (2)  diet- bers, whose reports of family functioning were
ing and eating practices, (3) family dynamics, and comparable to population norms. As treatment
(4)  psychological vulnerability (Dryer, Tyson, & progressed and the influence of BN decreased,
Kiernan, 2013; Dryer, Uesaka, Manalo, & Tyson, family member reports of family dysfunction
2015). While the university students endorsed remained relatively unchanged, but the reports of
sociocultural pressure more strongly and both lay the probands became more favourable. This find-
groups placed greater importance on dieting and ing is consistent with longitudinal research that
eating practices, the professional groups provided examined the direction of associations between
stronger endorsement for family dynamics. The parent–​adolescent relationships and adoles-
field of psychiatry has seen complete shifts away cent girls’ unhealthy eating, and found a direct
from parent-​blaming toward acceptance of the neu- effect of unhealthy eating on parent–​adolescent
robiological underpinnings of schizophrenia and relationships with no direct effect in the oppo-
autism; our intention is to assist a similar transition site direction (Archibald, Linver, Graber, &
in eating disorders. Brooks-​Gunn, 2002).
Herein, we provide an accessible review of A review of the 15 longitudinal studies of eat-
how genes influence eating disorders balanced by ing disorders meeting requirements for method-
the limitations of our current knowledge and the ological robustness (Jacobi, Hayward, de Zwaan,
direction for future research. In order to address Kraemer, & Agras, 2004) included only five that
these aims, the following issues are examined in incorporated family environment measures; only
turn: (1) how recognition of genetic influences on two of these studies found family variables to
eating disorders emerged; (2) the complexities of be risk factors for eating disorder development.
gene–​environment interplay; (3) what twin stud- Specifically, this included low levels of social sup-
ies can tell us about gene–​environment interplay, port from families (Ghaderi, 2003) and abusive
and (4)  a review of the status of contemporary parental relationships (Johnston, Cohen, Kasen,
molecular genetic studies. & Brook, 2002). An updated review (Jacobi &
Fittig, 2010) confirmed that family environment
How Recognition of Genetic Influence can only be viewed as a retrospective correlate
on Eating Disorders Emerged rather than as a true risk factor for the develop-
Understanding of the role of the family has ment of an eating disorder. This stance is reflected
evolved over the past several decades. in evidence-​based family therapy approaches for
eating disorders, which recognize that the family
Lack of Evidence to Support Causal Role reorganizes themselves around the eating disorder
of Family Environment in such a way that prevents them from being able
There has long been a general consensus among to use their normal adaptive mechanisms to deal
eating disorder researchers that AN and BN are “dis- with change (Eisler, 2005).
orders in which biological, familial, and sociocultural
factors play important etiological roles” (Johnson & Consistent Evidence for a Substantial
Flach, 1985). The focus of the earliest etiological Genetic Contribution to Eating Disorders
research was delineating the specific family struc- Over the last 25  years, a respectable body
ture that produced an eating disorder as opposed of twin studies revealed a substantial and repli-
to other psychopathology such as depression or cated genetic contribution to both AN and BN
anxiety. This interest was fueled by cross-​sectional (Bulik, Sullivan, Wade, & Kendler, 2000; Bulik
research that consistently showed that as the sever- et  al., 2006; Fairburn & Harrison, 2003; Collier
ity of eating pathology increased so too did the & Treasure, 2004; Yilmaz, Hardaway, & Bulik,
proband reports of family dysfunction (Wisotsky 2015). Across the different twin studies examining
et al., 2003). disordered eating, it has been estimated that genetic
However, the limitations of cross-​ sectional influences account for 52% of the estimated median
associations required that the meaning of the variance contributing to eating disorder aetiology
findings be investigated longitudinally. Woodside (Culbert, Racine, & Klump, 2015).

82 Genetic Influences
The Complexities of Gene–​Environment been associated with increased BMI (Locke et  al.,
Interplay 2015). Intriguingly, the majority of these genes are
To fully understand the etiology of eating dis- expressed in the brain, suggesting strongly that cen-
orders, we must consider several complex ways in tral regulation of human BMI occurs (Willer et al.,
which genes and environment can interact. 2009; Locke et al., 2015). Whether AN represents
the opposite extreme of the BMI continuum has been
Numerous Genes Acting on Different raised in the literature (Hebebrand & Remschmidt
Pathways 1995), and several research groups have suggested
There is no one gene that “causes” eating disor- that shared genetic factors may contribute to both
ders. The relevant genes are likely to be associated extremes of BMI dysregulation (Hebebrand &
with multiple pathways, many of which we are not Remschmidt 1995; Hinney, Friedel, Remschmidt, &
currently aware of. Like any complex trait, eating Hebebrand, 2004; Pinheiro, Sullivan, Bacaltchuck,
disorders are understood to be influenced by many Prado-​Lima, & Bulik, 2006; Sulek, Lacinova,
genes and many specific environmental factors, Dolinkova, & Haluzik, 2007; Scherag, Dina, et al.,
where genetic factors are seen to operate in a proba- 2010; Boraska et al., 2014). Using a technique called
bilistic fashion like risk factors rather than predeter- LD-​score regression (LDSR) (Bulik-​Sullivan, Loh,
mined programming (Plomin, 2000). Early research et al., 2015), a significant negative genetic correla-
in eating disorders has focused on the serotonin tion between AN and obesity was revealed (and a
pathway (involved in weight regulation and eating similar genetic correlation with BMI), suggesting
behavior), the catecholamine pathway (including that the same genetic factors that influence normal
the neurotransmitters dopamine, norepinephrine, variation in BMI also influence extreme dysregu-
and epinephrine), and the pathway involved in neu- lated BMI in AN (Bulik-​Sullivan, Finucane, et al.,
ropeptide and feeding regulation (Slof-​Op ‘t Landt 2015). Identifying the implicated genes is more
et al., 2005). Early molecular genetic work using an challenging, however, as 89 SNPs with genomewide
approach called candidate gene association studies, significance for BMI variation and obesity (Fall &
was driven by biological hypotheses and involved Ingelsson, 2014) and 15 SNPs related to extreme
considerable guesswork. For example, as it has long obesity (Fall & Ingelsson, 2014) were not associ-
been known that serotonin is involved with both ated with AN (Boraska et al., 2014). In a subsequent
appetite and mood (Wurtman, 1993), researchers study, however, some alleles predisposing to AN also
targeted single genes in the serotonin pathway to see predisposed to lower BMI (Hinney et al., 2017).
if they differed between individuals with and without This exciting line of research has launched addi-
eating disorders (see Yilmaz et al., 2015, for a review). tional inquiries addressing whether AN should
Keeping in mind that different types of eating dis- be considered not only to be a psychiatric phe-
orders exist, each of which is defined by a number notype, but also a metabolic one. Important next
of different diagnostic criteria, including weight, dys- steps in addition to further clarifying which genes
regulated eating behaviors, and cognitive symptoms, are involved in both obesity and AN risk, is to
and that increased vulnerability to any one of these understand the mechanisms whereby they exert
criteria can arise via a number of different pathways, their effect to help us further understand the path-
it is now clear that focusing on one or a few genes ways that contribute to dysregulated BMI in both
could never capture the complexity of the genetic directions.
effect. Numerous genes acting on different pathways
will be differentially involved in increasing a person’s Genetics of Behavioral, Temperamental,
risk for developing an eating disorder. and Personality Factors Associated
with Eating Disorders
Genetics of Body Mass Index In addition to exploring what we know about
One example of this complex process is illus- the genetics of BMI dysregulation and component
trated by the genetic factors that impact on body symptoms of eating disorders, multiple pathways of
mass index (BMI). Twin, family, and adoption action are also suggested by genetic influences on
studies unequivocally implicate genetic factors in behaviors and temperaments associated with eating
BMI (Maes, Neale, & Eaves, 1997; Hinney, Vogel, disorders that are thought to be important in deter-
& Hebebrand, 2010). At the time of writing, 97 mining their onset and course, but are not currently
genomewide significant (p ≤ 5 × 10−08) loci have included as part of any diagnostic schemes. These

Wade, Bulik 83
continuous measures could result in useful indices emotionality (e.g., ra  =  .51, Singh & Waldman,
of vulnerability that may not necessarily indicate 2010), and also for executive function (Anokhin,
the presence of a clinical syndrome but may be pre- Golosheykin, Grant, & Heath, 2010). As reviewed
dictive of increased risk for developing a disorder below, there is a significant and substantial contri-
(Hewitt, 1997) and also represent quantitative risk bution of both heritability and nonshared environ-
indices that may increase reliable diagnosis (Kendler, ment to phenotypes assessing various dimensions of
Neale, Kessler, Heath, & Eaves, 1993), and result- weight concern.
ing in more focused and powerful molecular genetic
investigations. For example, the importance of these Gene–​Environment Correlations and
types of measures in helping focus work in linkage Interactions
analyses of AN has been shown where incorporation One of the best ways to understand the inter-
of two behavioral covariates, drive for thinness and play between genes and the environment is the
obsessionality, was able to reveal several regions of identification of gene–​environment correlations
interest (Devlin et al., 2002). and genotype–​environment interactions (Rutter
Using this endophenotype approach (expressions & Silberg, 2002). To date, the majority of this
of biological markers for a phenotype that are asso- work in eating disorders has been using variance-​
ciated with illness in the general population that are components approaches in twin models or single-​
a stable and state-​independent characteristic, and candidate genes coupled with environmental
found in nonaffected family members at a higher factors. Ultimately, although very large sample sizes
rate than in the general population), four types of are required, true genotype–​environment interplay
constructs have been proposed as being worthy of will be explored using genomewide data coupled
further investigation (Bulik et  al., 2007). These with environmental variables.
include increased physical activity, dimensions of Genotype–​environment correlations describe
temperament (obsessionality, impulsivity, perfec- the extent to which individuals are exposed to cer-
tionism, and negative emotionality), impaired set tain environments as a function of their genetic vul-
shifting (executive functioning, responsible for nerabilities. Three types of genotype–​environment
the supervision of such cognitive processes as set- correlations have been hypothesized to exist (Scarr
ting goals, planning, and organizing), and dimen- & McCartney, 1983) and previously discussed in
sions reflecting weight concern (including drive for terms of their relationships to eating disorders (e.g.,
thinness). Strober, 1991; Wonderlich, 1992):  passive, evoca-
Each of these phenotypes has been found to tive, and active, each of which is described below.
be significantly heritable, including physical activ- Passive genotype–​environment correlation
ity (54%, 95% CI  =  45%–​62%) (Mustelin et  al., occurs in reared-​together biological relatives, where
2012) and temperament. Genetic factors have been parents provide their children with both genes and
shown to exert a substantial influence on the persis- an environment that is conducive to the develop-
tence of obsessive compulsive behavior over child- ment of certain traits that occur independently of
hood, explaining 59%–​80% of the stability (Krebs, the offspring’s characteristics. For example, parents
Waszczuk, Zavos, Bolton, & Eley, 2015); the genes who are highly concerned about themselves and
significantly associated with obsessive compulsive their children becoming overweight could trans-
symptoms are expressed in the brain and involved mit to their children both the genes for these traits
in the development and control of immune sys- and an environment that is conducive to the trait’s
tem functions and regulation of gene expression of development such as frequent weighing of the child,
muscle-​specific genes (den Braber et al., 2016). One modeling dieting behaviors to the child, and impos-
study has found the heritability of impulsivity to be ing restrictive eating patterns on the child. We
33% (95% CI:  30%–​36%) in males and females know that mothers of the 4-​year-​old children with
(Hemiri, Kuja-​ Halkola, Larsson, & Jayaram-​ feeding problems have a significantly higher rate of
Lindstrom, 2016), and having a significant overlap past or current eating disorders (odds ratio = 11.1,
with genetic risk factors for alcohol dependence 95% confidence interval [CI]: 1.4–​91.8) compared
(ra = 0.40). Heritability has been found to contrib- to the other mothers who have children with other
ute across the different measures of perfectionism, disturbances such as anxiety or mothers whose
ranging from 22% to 43% of the variance (Wade children have no disturbances (Whelan & Cooper,
& Bulik, 2007; Iranzo-​Tatay et al., 2015). There is a 2000). Videotaped interactions suggest that the
moderate additive genetic contribution to negative child’s disturbed eating may not only be caused by

84 Genetic Influences
genetic action (i.e., inheriting the mother’s risk for thus have less scrutiny with respect to their dietary
disordered eating) but also through environmen- intake, less exposure to modeling of healthy eating,
tal action, whereby mothers with eating disorders as well as less experience of eating in a social realm as
use more verbal control in meal and play times opposed to simply a focus on food, then this could
than mothers with postnatal depression or controls reinforce an overvalued importance of controlling
(Stein et  al., 2001). At mealtimes these mothers, dietary intake which could lead to the development
by virtue of their own challenges with eating dis- of unhealthy weight management practices.
orders, are more likely to express negative emotion The second mechanism that describes the
and be less facilitating and more intrusive (Stein, interplay between genes and the environment is
Woolley, Cooper, & Fairburn, 1994; Stein et  al., genotype–​environment interactions that occur
2001; Waugh & Bulik, 1999). when the effect of the environment depends on the
By contrast, both evocative and active genotype–​ genotype, such that individuals will be varyingly
environment correlations are dependent on an susceptible to the influence of high-​risk environ-
individual’s characteristics and/​ or behaviors. An ments proportional to their degree of genetic risk.
evocative genotype–​environment correlation occurs Conversely, the importance of the environment in
when genetically influenced characteristics in the manipulating genetic expression is immense, given
child evoke an environmental response that reflects findings from animal models that early environ-
the genetic trait. For example, an adolescent with mental manipulation can change gene expression
the genetically influenced trait of overweight might (Meaney et  al., 1985). An example of the impor-
experience excessive teasing about body weight and tance of environmental action in humans can be
shape from peers (an environmental response), found in tobacco use when two historical cohorts
and this teasing may result in dysphoric mood in Sweden were compared (Kendler, Thornton, &
that triggers binge eating. An evocative genotype–​ Pedersen, 2000), one where smoking was rare for
environment correlation is suggested by findings women and one in which smoking had become
that people with AN are characterized by obsession- widespread throughout the community. The heri-
ality, rigidity, low impulsivity, fear of uncertainty, tability for tobacco smoking for women increased
and avoidance of novel situations (Cassin & von from 0% to 63%, while heritability for men stayed
Ranson, 2005) and that increased paternal con- relatively stable at around 63%.
trol is uniquely associated with AN in MZ twins Genotype–​environment interactions may to
discordant for AN compared to MZ twins discor- some extent explain why the Western sociocul-
dant for BN or major depression (Wade, Gillespie, tural environment and its emphasis on thinness for
& Martin, 2007). This could suggest that a harm-​ women and muscularity for men impacts adversely
avoidant and timid child evokes a response of over- on some people but not others. In an experimen-
protection from concerned fathers, which in turn tal study, only vulnerable adolescent girls, defined
leads the child to become more timid and uncertain as those with elevated levels of body dissatisfaction
of their autonomy and ability to act independently and perceived pressure to be thin, were adversely
and thus they start to develop an overvalued impor- affected by exposure to a 15-​month subscription of
tance of control over eating and weight in order to a fashion magazine compared to nonvulnerable girls
compensate. in terms of their negative affect (Stice, Spangler, &
Active genotype–​environment correlations refer Agras, 2001). In other words, vulnerability toward
to the situation where individuals actively select or body dissatisfaction can be exacerbated by an envi-
create environments that are correlated with their ronment that further reinforces the dominance of
genetic propensities. An example of this type of cor- the thin-​ideal. Further candidates to examine as
relation would be the selection of weight conscious environmental moderators of genetic risk are indi-
friends by an adolescent who also places impor- cated by research that shows an increase in nega-
tance on physical appearance. Another example is tive life events predicts onset of eating disorders
suggested by a longitudinal study that found that (McKnight Investigators, 2003) and that initial
regular family meals (≥ five meals a week) were asso- social support can protect vulnerable girls against
ciated with lower prevalence of extreme weight con- the impact of fashion magazines (Stice et al., 2001).
trol behaviors in adolescents (Neumark-​ Sztainer, Therefore, in the context of the unhelpful environ-
Eisenberg, Fulkerson, Story, & Larson, 2008). If ment of the thin-​ideal, vulnerability to body dissat-
adolescents who are genetically vulnerable to devel- isfaction could be increased by negative life events
oping eating problems select to eat on their own and or decreased by adequate social support.

Wade, Bulik 85
What Twin Studies can Tell Us About arising from Mendel’s classical experiments, refers
Gene–​Environment Interplay to the interaction between alleles at the same locus.
To date, the majority of the work in understand- In twin studies, dominance is inferred when the
ing gene environment interplay in eating disorders concordance between MZ twins is greater than
has been using variance-​ components approaches twice that of DZ twins.
in twin models or single-​candidate genes coupled Twin studies are flexible tools for modeling
with environmental factors. Ultimately, although gene–​environment interplay that can provide
very large sample sizes are required, true genotype–​ descriptive models tracing the complex causal path-
environment interactions will be explored using way from genotype to phenotype, with a particular
genomewide data coupled with environmental focus on putative environmental risk factors. There
variables. are some indications that there is a different preva-
lence of phenotypes in twins compared to single-
Why Twin Studies? tons, but the direction of impact is unclear. In one
Twin studies can compare trait similarity between study, twins reported less disordered eating than
identical or monozygotic (MZ) twins (who share singletons (Munn-​ Chernoff, von Ranson, et  al.,
100% of their nonsegregating alleles) and noniden- 2013), but another study found that twin status
tical or dizygotic (DZ) twins (who share, on aver- and lower gestational age independently predicted
age, only 50%, just like nontwin siblings), where AN in a Swedish population study, with a 1.5-​fold
structural equation modeling approaches are used difference in AN rates observed between twins and
to examine the relative contribution of three latent singletons (Goodman, Heshmati, Malki, & Koupil,
factors: additive genetic, shared environmental, and 2014). It is unknown whether there is an impact
nonshared environmental effects (Plomin, Defries, of any such differences on generalizing findings of
& McClearn, 1990). Additive genetic effects are twin studies to nontwins.
those genetic factors that “add” rather than inter- Specifically, twin studies can be used to: (1) esti-
act across genes, roughly indicated by a concor- mate heritability of a trait or behavior, (2) investigate
dance between MZ twins for the disorder that is the shared latent risk factors between two or more
approximately twice as high as that of DZ twins. different phenotypes, (3)  inform our understand-
Shared environmental influences are inferred when ing of gene–​environment interplay in the form of
MZ and DZ twin correlations are approximately genotype–​environment correlations and genotype–​
equal, as these factors are common to cotwins grow- environment interactions, and (4) examine develop-
ing up in the same family and therefore contribute mental changes to genetic and environmental risk
to their behavioral similarity (Plomin et al., 1990). factors. The findings to date associated with each
Nonshared environmental influences are unique to area are summarized below.
each cotwin and make them different from each
other. The nonshared environment can be either Heritability Estimates
objective, an actual experience or event that is not Traditionally twin studies have been used to
shared by siblings, or effective, where the same event develop estimates of heritability of phenotypes.
can be experienced uniquely by each family member, However, when it comes to developing accurate
depending on a number of factors such as age and estimates of heritability, twin studies can only
temperament, thus producing differential outcomes develop such estimates commensurate with the
(Turkheimer & Waldron, 2000). It is also worth reliability of the measured trait (Foley, Neale, &
noting that nonshared environmental influences Kendler, 1998), which goes some way to address
can also include a variety of prenatal events (Martin, the observation that “the findings of twin studies [of
Boomsma, & Machin, 1997), that result in MZ eating disorders] are inconsistent  . . .  with [wide]
twins, although carrying identical DNA sequences, estimates for the heritability of liability” (Fairburn,
to exhibit numerous epigenetic differences (Steiger Cowen, & Harrison, 1999, p. 349). More accurate
& Thaler, 2016) that is, there is no structural change measurement of phenotypes is likely to increase esti-
to the DNA, but gene expression is modulated by mates of heritability as measurement error, which
the environment in the womb. There is a fourth contributes to nonshared environmental variance,
source of potential variance, known as genetic dom- will be decreased. There is some disagreement as
inance (Martin, Eaves, Kearsey, & Davies, 1978), to whether interviews or self-​report questionnaires
which is rarely indicated as accounting for variance provide greater reliability in the measurement of
in phenotypes. “Dominance,” a term originally phenotypes (Kendler et  al., 1993; Burt, 2009).

86 Genetic Influences
Lower estimates of heritability are found when et  al., 2016) exist respectively. Results are broadly
using diagnostic interviews as opposed to self-​report consistent with findings for other disorders, in that
questionnaires (Burt, 2009). Use of multiple inter- there is a substantial impact of both genes and non-
view occasions and coassessment of lifetime comor- shared environment. The latter study also found
bidity can improve reliability (Foley et  al., 1998; that the same genetic factors accounted for both
Rice, Rochberg, Endicott, Lavori, Miller, 1992), as initiation and progression of self-​induced vomit-
can the use of valid interviews of eating disorders ing for weight and shape control, but there were
that can address the difficulties inherent in making some unique environmental factors contributing to
a lifetime diagnosis with a sometimes sporadic and the progression to regular purging. This finding is
egosyntonic disorder (Swanson, Brown, Crosby, similar to an examination of latent risk factors for
& Keel, 2014). The tandem use of categorical and DSM-​5 threshold eating disorders and other speci-
dimensional assessments can also enhance the relia- fied feeding and eating disorders in adolescent girls,
bility of our estimates (Kessler, 2002; Kendler et al., which showed the two diagnostic groups to share a
1993; Hewitt, 1997, Fairburn et al., 2007). common genetic basis; however, largely nonoverlap-
It is within the context of the above issues that ping unique environmental influences contributed
the twin studies of the heritability of eating and eat- to the two groups (Fairweather-​Schmidt & Wade,
ing disorders should be interpreted. These studies 2014). This suggests that environmental, but not
are summarized across three different tables, and genetic, risk factors may explain progression to more
include only nonascertained twin studies, which are frequent disordered eating behaviors, and highlights
expected to give less inflated heritability estimates the significance of environmental factors in the eti-
as opposed to studies where twins have been ascer- ology of eating disorders (Peterson et al., 2016).
tained from treatment clinics. With more heterogeneous disorders it may not
In Table 5.1, twin studies of eating disorders be sensible to try to estimate heritability of the entire
are summarized. Eight studies of AN across four disorder, but rather to examine its component parts,
different countries are included, albeit the Danish which may also yield more fruitful linkage and
Twin Registry used a single self-​report question to association studies (Slof-​Op’t Landt et  al., 2005).
assess lifetime history of AN (Kortegaard, Hoerder, Seventeen studies that have examined specific diag-
Joergensen, Gillberg, & Kyvik, 2001). Six of the nostic criteria of eating disorders are summarized in
studies found significant (i.e., confidence intervals Table 5.2. On the whole, the investigation of com-
did not include zero) contributions of heritability, ponent parts does yield clearer results with respect
ranging from 32% to 76% of the variance, and to heritability, showing a significant contribution for
seven studies found significant contributions of all behaviors across all studies. Studies of overeat-
the nonshared environment, ranging from 24% to ing (binge episodes not necessarily involving a loss
68%. None found a significant contribution of the of control) are included as one analysis of the data
shared environment. showed that results were “statistically indistinguisha-
Seven twin studies of BN are summarized in ble” from a phenotype that included (1) loss of con-
Table 5.1, all from the Virginia Twin Registry trol and (2)  an amount of food that others would
albeit one study (Rowe et al., 2002) uses a different consider unusual (Bulik, Sullivan, & Kendler, 1998).
(and younger) sample than the other four studies. Slightly less consistent are results for the cogni-
Previously BN has been described as a heteroge- tive component of eating disorders, importance of
neous disorder (Fairburn, 1991), where diagnos- weight and shape, described as the “core psychopa-
tic features vary in severity, and the course of the thology” of eating disorders (Cooper & Fairburn,
disorder varies widely between patients. Therefore, 1993). A  Norwegian twin study using a single
it is no surprise that estimates for this phenotype self-​report item for males and females (Reichborn-​
are more variable, with four studies showing a sig- Kjennerud et al., 2004) found no heritability impli-
nificant contribution of heritability (28% to 83%), cated in the etiology of this phenotype. However
one study showing a significant contribution of the three studies found a significant contribution
shared environment (37%), and three studies show- of heritability to the phenotype (24% to 49%)
ing significant contributions of the nonshared envi- in young adults (Wade & Bulik, 2007; Spanos,
ronment (35% to 45%). Burt, & Klump, 2010; Mazzeo et  al., 2010). It is
Two studies of binge-​eating disorder (Mitchell of interest to note that the full model tested uni-
et al., 2010; Javaras et al., 2008) and purging dis- variately with the adolescents also suggested the
order (Munn-​Chernoff, Keel et al., 2015; Peterson substantial presence of shared as well as nonshared

Wade, Bulik 87
Table 5.1  Non-​ascertained twin studies of eating disorders
Study Model fitting results1 Population2: How was phenotype
defined?
A (95% CI) C (95% CI) E (95% CI)

Anorexia Nervosa

Wade, Bulik, Neale, .58 (.33–​.84) –​ .42 (.16–​.68) VTR (mean age, 29.3+7.7 years): SCID
& Kendler, 20003 (first wave), met all DSM-​III-​R criteria
except for (i) amenorrhea OR (ii) feeling
fat when emaciated. Bivariate analysis
with major depression

Klump, Miller, Keel, .76 (.35–​.95) –​ .24 (.05–​.65) MTFS (mean age,
McGue, & Iacono, 17.5+0.5 years): SCID, met DSM-​IV
2001 criteria for (i) < 85% IBW and met all
but one criteria for AN OR (ii) <90%
IBW AND at least one cognitive
symptom of AN AND scored above
mean on EDI for all twins in (i)

Kortegaard, Hoerder, .48 (.27–​.65) –​ .52 (NR) Danish Twin Register (aged 11–​41 years):
Joergensen, Gillberg, SRQ “Have you ever had AN?”
& Kyvik, 2001

Bulik, et al., 2006 .56 (0–​.87) .05 (0–​.64) .38 (.13–​.84) Swedish Twin Registry (aged 49–​
58 years): SCID, narrow (all DSM-​IV
criteria)

.31 (0–​.62) 0 (0–​.44) .68 (.37–​1.0) SCID, broad (all DSM-​IV criteria with
exception of amenorrhea)

Mazzeo et al., 2009 .22 (0–​.50) .14 (0–​.44) .64 (.49–​.79) Norwegian Institute of Public Health
Twin Panel (aged 19 to 36 years): CIDI
(DSM-​IV)

Dellava, Kendler, & .32 (.01–​.70) –​ .68 (.30–​.99) Virginia Adult Twin Study of Psychiatric
Neale, 2011 and Substance Use Disorder, Interview
questions mapping on to DSM-​III-​R

Cederlof et al., 2015 .38 (.20–​.54) –​ .62 (.46–​.80) Swedish Twin study of Adults: Genes and
Environment (aged 20–​47 years): SCID
(DSM-​IV), bivariate analysis with
obsessive compulsive disorder

Thornton, Welch, .38 (.08–​.53) 0 (0–​.24) .62 (.47–​.79) Swedish Twin study of Adults: Genes and
Munn-​Chernoff, Environment (aged 20–​47 years): SCID
Lichtenstein, & (DSM-​IV), multivariate analysis with
Bulik, 2016 depression and suicide attempts

Bulimia Nervosa

Kendler, et al., 1991 VTR (mean age, 30.1+7.6 years): SCID

(first wave)

.55 (0–​.77) –​ .45 (.23–​.77) narrow -​met all DSM-​III-​R criteria

.52 (NR) –​ .48 (NR) broad -​met “most but not all” DSM-​III-​
R criteria
Table 5.1 Continued
Study Model fitting results1 Population2: How was phenotype
defined?
A (95% CI) C (95% CI) E (95% CI)

Walters, et al., 1992 .50 (NR) –​ .50 (NR) VTR: SCID (first wave) broad DSM-​III-​
R. Analysis with major depression.

Kendler, et al., 1995 .28 (.07–​.62) .37 (.10–​.59) .35 (.19–​.49) VTR: SCID (first wave) broad DSM-​
III-​R. Multivariate analysis with 5 other
psychiatric disorders.

Bulik, Sullivan, & .83 (.49–​1.0) 0 (0–​.30) .17 (0–​.36) VTR: SCID (first and third waves),
Kendler, 1998 met all DSM-​III-​R criteria except binge
eating frequency. Bivariate analysis

Rowe, et al., 2002 .54 (.44–​.62) –​ .46 (NR) VTSABD (aged 8–​17 years): Child
and Adolescent Psychiatric Assessment,
number of DSM-​III-​R criteria (excluding
weight and shape importance).

Baker, Mazzeo, & .31 (NR) –​ .59 (NR) VTR: SCID (first and third waves) broad
Kendler, 2007 DSM-​III-​R. Bivariate analysis with drug
use disorders.

Mazzeo et al., 2010 .62 (.18–​.76) 0 (0–​.35) .38 (.23–​.62) VTR: self-​report SCID items DSM-​IV.

Binge Eating Disorder

Mitchell et al., 2010 .45 (0–​.92) 0.13 (0–​.96) .42 (0–​1.00) VTR: self-​report SCID items DSM-​IV
(overlap with previous samples), mean
age=40.44 years (SD=8.34)

Javaras et al., 2008 .39 (NR) –​ .61 (NR) Norwegian Twin Panel (18 to
31 years): self-​report “binge eating with a
feeling of loss of control in the absence of
compensatory behaviors”.

Purging Disorder

Munn-​Chernoff, .20 (0–​.65) .24 (0–​.57) .56 (.35–​.79) Missouri Adolescent Female Twin
Keel, et al., 2015 Study (median age 15 years at first
interview): SSAGA

Peterson et al., 2016 .63 (.54–​.70) –​ .37 (.30–​.37) Swedish Twin Study of Adults (20
.85 (.04–​.99) –​ .15 (.01–​.96) to 47 years): self-​report “have you
ever made yourself vomit to control
your weight or shape?”, initiation and
progression
1
Full model or model of best fit –​not necessarily significantly better fitting than the full model;
2
All populations were female with the exception of the Rowe et al (2002) study which contained males and females;
3
This study can be interpreted as superseding the Walters and Kendler (1995) study of this population as the zygosities were corrected in the
light of genotyping, where 2 DZ twin pairs were reassigned as MZ.
Note: A=additive genetic variance; C=shared environmental variance; E=non-​shared environmental variance; CI=confidence interval; NR=not
reported; VTR=Virginia Twin Registry; SCID=Structured Clinical Interview for DSM; CIDI=Composite International Diagnostic Interview;
VTSABD=Virginia Twin Study of Adolescent Behavioural Development; MTFS=Minnesota Twin Family Study; EDI=Eating Disorder
Inventory; IBW= ideal body weight; AN=anorexia nervosa; SRQ=self-​report questionnaire; SSAGA=Semi-​structured assessment on the
genetics of alcoholism.
Table 5.2  Non-​ascertained twin studies of DSM diagnostic criteria for eating disorders
Study Criterion Model fitting results1 Population2: How was
phenotype ascertained?
A (95% CI) C (95% CI) E (95% CI)

Bulik, Sullivan, Overeating .82 (.68–​.97) –​ .18 (.03–​.32) VTR: SCID (first and third
& Kendler, waves) “Have you ever in your
1998 life had eating binges during
which you ate a lot of food in a
short period of time?” Bivariate
analysis of two waves

Sullivan, Overeating .46 (.32–​.58) –​ .54 (42–​.68) VTR: SCID (first wave).

Bulik, & Bivariate analysis with vomiting.
Kendler, 1998

Bulik, Sullivan, Overeating .49 (.38–​.61) –​ .51 (.37–​.64) VTR: SCID (first and third
& Kendler, waves). Bivariate analysis with
2003 lifetime history of obesity.

Javaras, et al., Objective .39 (.26–​.52) .61 (NR) Norwegian Twin Registry (aged
2008 binge 18–​31 years): SRQ, binge
episodes eating with loss of control in
the absence of compensatory
behaviours (females and males
combined).

Wade, Treloar, Objective .17 (.05–​.28) –​ .83 (.72–​.96) ATR (aged 28–​40 years): EDE,
& Martin, binge 2 x week for 3 month period
2008 episodes ever.

Sullivan, Bulik, Vomiting .70 (.50–​.84) –​ .30 (.16–​50) VTR (first wave): SCID “Have
& Kendler, you ever vomited?”
1998

Wade, Treloar, Vomiting .08 (.01–​.18) –​ .92 (.82–​1.0) ATR (aged 28–​40 years): EDE,
& Martin, 2 x week for 3 month period
2008 ever.

Keski-​ Intentional .66 (.55–​.75) –​ .34 (.25–​.45) Finnish Twin Registry (aged
Rahkonen, weight loss .38 (.19–​.55) –​ .62 (.45–​.81) 16–​27 years): SRQ “How many
Neale, et al., times in your life have you
2005 intentionally lost >5 kg weight?”
Bivariate analysis with BMI -​
females and males.

Reichborn-​ IWS -​ .31 (.24–​.38) .69 (.68–​.76) Norwegian Twin Registry

Kjennerud (aged 18–​31 years): SRQ “Is
et al., 2004 it important for your self-​
evaluation that you keep a
certain weight?” (females and
males combined).

Wade & Bulik, IWS .25 (.14–​.36) –​ .75 (.64–​.87) ATR (aged 28–​40 years): EDE,
2007 2 items over previous 3-​month
period. Multivariate analysis
with perfectionism.
Table 5.2 Continued
Study Criterion Model fitting results1 Population2: How was
phenotype ascertained?
A (95% CI) C (95% CI) E (95% CI)

Wilksch & IWS .15 (0–​48) .23 (0–​43) .62 (52–​73) ATR (aged 12–​15 years): EDE,
Wade, 2009 2 items over previous 3-​month
period. Univariate analysis.

Mazzeo et al., Objective .39 (.25–​.71) .04 (0–​.18) .57 (.26–​.75) VTR: self-​report SCID
2010 binge .14 (0–​.24) .33 (.22–​.48) items DSM-​IV (overlap with
episodes .02 (0–​.12) .55 (.39–​.79) previous samples)
Vomiting .53 (.41–​.74) .03 (0–​.22) .53 (.36–​.74)
Laxatives .43 (.21–​.54) .12 (0–​.32) .53 (.38–​.70)
Diuretics .44 (.23–​.62) .11 (0–​.27) .47 (.34–​.65)
Driven .13 (0–​.28) .63 (.53–​.75)
exercise .35 (.13–​.60)
Fasting .42 (.20–​.62)
IWS .24 (.10–​.46)

Mitchell et al., Objective .43 (.23–​.67) .11 (0–​.22) .46 (.25–​.74) VTR: self-​report SCID
2010 binge items DSM-​IV (overlap with
episodes previous samples), mean
age=40.44 years (SD=8.34)

Spanos, Burt, IWS .49 (.33–​.63) 0 (0–​.34) .51 (.37–​.67) Michigan State University
& Klump, Twin Registry (MSUTR),
2010 mean age 20.92 (SD=2.47),
two items from the EDE-​Q

Slane, Burt, & Overeating .41 (.23–​.56) –​ .59 (.44–​.77) MUSTR (mean age 20.92,
Klump, 2012 SD=2.47): 7 self-​report items
adapted from EDI

Koren et al., Overeating .38 (.20–​.55) –​ .62 (.45–​.80) Missouri Adolescent Female
2014 Twin Study (median
age 15 years at first
interview): SSAGA, no binge
eating (BE), BE without loss
of control, BE with loss of
control

Munn-​ Objective .43 (.33–​.52) –​ .57 (.48–​.67) Missouri Adolescent Female

Chernoff, Grant binge Twin: SSAGA, analysed with
et al., 2015 eating and alcohol use disorder
compensatory
behaviours
1
Full model or model of best fit (lowest Akaike’s Information Criteria) –​not necessarily significantly better fitting than the full model;
2
All populations were female unless otherwise noted.
Note: A=additive genetic variance; C=shared environmental variance; E=non-​shared environmental variance; CI=confidence interval; NR=not
reported; VTR=Virginia Twin Registry; SCID=Structured Clinical Interview for DSM; BMI=body mass index; ATR=Australian Twin
Registry; EDE=Eating Disorder Examination; EDE-​Q=Eating Disorder Examination Questionnaire; SRQ=self-​reported questionnaire;
IWS=importance of weight and shape; SSAGA=Semi-​structured assessment on the genetics of alcoholism.
environment (Wilksch & Wade, 2009), consistent correlation in genetic risk factors is associated with
with the Norwegian study. The shared environment a meaningful overlap in genetic risk. For example,
was also suggested by the multivariate twin study bivariate twin models of both obsessive-​compulsive
of BN summarized in Table 5.1 (Kendler et  al., disorder and AN (Cederlof et al., 2015), and gener-
1995). This is of interest given that the latent vari- alized anxiety disorder and AN (Dellava, Kendler, &
able models used in twin studies have relatively low Neale, 2011) found moderate but significant over-
power to identify any but sizable effects of environ- lap in genetic risk factors (ra = 0.52 and ra = 0.20
ment that is shared by twins (M. C. Neale, Eaves, & respectively), but most of the genetic variance was
Kendler, 1994). disorder specific for obsessive-​compulsive disorder
Table 5.3 presents a summary of 19 twin studies and AN (Cederlof et  al., 2015), while generalized
examining continuous measures of disordered eat- anxiety disorder and AN shared an appreciable
ing: behaviors and cognitions or attitudes that are amount of genetic variance, around 44% (Dellava
not included in diagnostic criteria. These studies et al., 2011). Detection of meaningful overlap can
have been conducted across eight different coun- inform development of interventions for disordered
tries (United Kingdom, Australia, United States, eating, in that treatment approaches that are effec-
Japan, Finland, Netherlands, Korea, and Sweden), tive for “shared risk phenotypes” may also be indi-
with most supporting a significant and substan- cated for eating disorders.
tial contribution of both heritability and non- A number of twin studies have examined the
shared environment to the phenotypes. Some of overlap between genetic risk factors for eating dis-
the studies indicating a contribution of the shared orders and depression and suicidality, showing that
environment (Klump, McGue, & Iacono, 2000; the association between the two phenotypes is pri-
Kamakura, Ando, Ono, & Maekawa, 2003; Keski-​ marily due to shared genetic risk factors rather than
Rahkonnen, Bulik, et  al., 2005; Klump, Burt, an overlap in environmental risk factors. The cor-
McGue, & Iacono, 2007), with findings suggesting relation between genetic risk factors contributing
that the contribution of the shared environment to both AN and major depression was 0.56 (95%
decreases substantially from childhood to adoles- CI:  0.44–​0.69) (Wade, Bulik, Neale, & Kendler,
cence (Klump et al., 2007). 2000), and the genetic correlation for BN and major
While most studies have not included males depression was 0.46 (Walters et al., 1992). An exam-
given the low prevalence of eating disorders in men ination of BN in a multivariate analysis of phobia,
compared to women, a handful of studies suggest generalized anxiety disorder, panic disorder, major
that genetic risk may work differently between males depression, and alcoholism showed that there was a
and females. Heritability for intentional weight shared genetic risk factors among BN, phobia, and
loss and disordered eating is substantially lower in panic disorder (Kendler et al., 1995). A study of dis-
males than females, with almost half the variance ordered eating and depressive symptoms concluded
(Keski-​Rahkonnen, Neale et  al., 2005; Slof-​Op ‘t that phenotypic associations between disordered
Landt et  al., 2008). Additionally, heritability does eating and depressive symptoms appear to be due
not appear to play a role in the etiology of body to common genetic factors (Slane, Burt, & Klump,
dissatisfaction or drive for thinness in males, in con- 2011), with a correlation between the genetic latent
trast to females, where it contributes around half of factors of 0.70 (95% CI: 0.50–​0.89) and a nonsig-
the variance (Keski-​Rahkonnen, Bulik et al., 2005). nificant overlap between nonshared environmental
Much more work is required in order to investigate factors. One study examining the overlap between
different specific types of environment that may any eating disorder (AN, BN, binge-​eating disor-
increase heritability for disordered eating in females der, and purging disorder) and suicidality found
compared to males. that genetic correlations between the two were sig-
nificant (0.60:  95% CI:  0.25–​1.00), but the cor-
Shared Latent Risk Factors Between Two or relations between latent sources of environmental
More Different Phenotypes influence were not (Wade, Fairweather-​Schmidt,
Twin studies are also useful for modeling the Zhu, & Martin, 2015). A trivariate examination of
overlap in genetic risk factors across multiple phe- AN, major depressive disorder, and suicide attempts
notypes and thus can increase power to identify (Thornton, Welch, Munn-​Chernoff, Lichtenstein,
potential biological pathways of action where the & Bulik, 2016) also suggested that genetic factors
associated phenotypes are more studied than eating accounted for a substantial amount of liability to all
disorders. Twin studies can also identify whether the three phenotypes (ra ranging from .48 to .77).

92 Genetic Influences
Table 5.3  Non-​ascertained twin studies of disordered eating
Study Outcome Model fitting results2 Population3: How was
Measure1 phenotype defined?
A (95% CI) C (95% CI) E (95% CI)

Rutherford, EAT .41 NR NR Institute of Psychiatry

McGuffin, EDI .44 NR NR NR Twin Registry (aged
Katz, & Murray, 18–​45 years)
1993

Wade, Martin, Dietary restraint .32 (.12–​.48) –​ .68 (.52–​.89) ATR (aged 36–​
& Tiggemann, Eating concern .46 (.30–​.58) –​ .54 (.42–​.70) 51 years): EDE four
1998 Weight concern –​ .52 (.43–​.64) .48 (.39–​.60) subscales over last
Shape concern .62 (.50–​.71) –​ .38 (.29–​.50) 3 months –​interview.

Wade, et al., Three waves of .59 (.50–​.68) –​ .41 (.32–​.51) ATR (mean age
1999 repeated measures 36.3+4.7 years at first
wave): First wave= eating
behaviours and diagnoses
ever, second wave=SCID
interview for BN ever,
third wave=EDE interview
global score over last
3 months.

Wade et al., Disordered eating .34 (NR) 0.09 (NR) 0.57 (NR) ATR: First wave= eating
2000 behaviours and diagnoses
ever in a multivariate
analysis with personality/​
family variables

Klump, EDI: 11 yr olds .82 (NR) 0.11 (NR) 0.07 (NR) MFTS: adapted EDI
McGue, & EDI: 17 yr olds .80 (NR) 0.00 (NR) 0.20 (NR) (body dissatisfaction,
Iacono, 2000 binge eating,
compensatory behaviour,
weight preoccupation

Klump, EDI .58 (.47–​.67) –​ .42 (.33–​.53) MFTS (mean

McGue, & age=17.46 years,
Iacono, 2002 SD=0.51, range=16 to
18 years): adapted EDI
in multivariate analysis
(personality)

Neale, Mazzeo, Dietary restraint 0 (0–​.30) .31 (.04–​.42) .69 (.58–​.80) VTR (aged 25–​
& Bulik, 2003 Disinhibition .45 (.32–​.57) 0 0–​.23 .55 (.43–​.68) 65 years): three subscales
Hunger .08 (0–​.38) .16 (0–​.34) .76 (.62–​.89) of the Three Factor Eating
Questionnaire.

Kamakura, Perfectionism .37 (.23–​.49) –​ .66 Keio Twin Project, Japan

Ando, Ono, Maturity fears –​ .43 (.29–​.54) (.51–​.77) (aged 14–​29 years): Five
& Maekawa, Ineffectiveness –​ .47 (.34–​.57) .57 subscales of the EDI.
2003 Interoceptive –​ (.46–​.71)
Awareness –​ .43 (.29–​.54) .53
Interpersonal (.43–​.66)
Distrust .34 (.20–​.46) .57
(.46–​.71)
.66
(.54–​.80)

(continued)
Table 5.3 Continued
Study Outcome Model fitting results2 Population3: How was
Measure1 phenotype defined?
A (95% CI) C (95% CI) E (95% CI)

Wade, Feeling Fat .72 (.68–​.75) –​ .28 (.25–​.32) ATR (mean age,
Wilkinson, & 32.4+4.2 years): Body
Ben-​Tovim, Attitudes Questionnaire.
2003 Multivariate analysis with
BMI.
Body –​ .47 (.42–​.53)
Disparagement .53 (.47–​.59) –​ .61 (.55–​.68)
Strength and –​ .61 (.54–​.68)
Fitness .39 (.32–​.46) –​ .54 (.48–​.60)
Salience of Weight –​ .48 (.43–​.55)
& Shape .39 (.33–​.46)
Attractiveness .46 (.40–​.52)
Lower Body
Fatness .52 (.45–​.57)

Keski-​ Drive for Thinness 51 (44–​58) 49 (43–​56) Finnish Twin Registry

Rahkonnen, (aged 22–​27 years): EDI
Bulik, et al.,
2005

–​ 86 (84–​88) 14 (12–​16) Females and males.

Body 59 (53–​65) 41 (35–​47)

dissatisfaction

–​ 85 (83–​87) 15 (13–​17)

Klump, Burt, EDI: 11 years .06 (0–​.26) .40 (.21–​.50) .54 (.47–​.61) MFTS: repeated measures
McGue, & EDI.
Iacono, 2007

EDI: 14 years .46 (.24–​.59) .10 (0.31) .44 (.38–​.50)

EDI: 17 years .46 (.24–​.59) .10 (0–​.31) .44 (.38–​.50)

Slof-​Op ‘t Disordered eating .65 (.58–​.71) –​ .35 (.29–​.42) Netherlands Twin Registry
Landt, et al., (aged 14–​18 years): 4
2008 items: dieting, importance
of weight & shape, fear of
weight gain, have you ever
had eating binges. Females
and males.

.39 (.28–​.49) –​ .61 (NR)

Munn et al., Weight concerns .43 (.33–​.52) –​ .57 (.48–​.68) Colorado Community
2010 and behaviours –​ .52 (.42–​.64) Twin Study (CTS) and
Emotional Longitudinal Twin Study
overeating .48 (.36–​.58) (LTS), twins and non-​
twin siblings (mean age
18.37 years, SD=1.68, 16-​
26 years): factor analysis
of 7 questions yielded a 2
factor solution
Table 5.3 Continued
Study Outcome Model fitting results2 Population3: How was
Measure1 phenotype defined?
A (95% CI) C (95% CI) E (95% CI)

Sung, Lee, Restraint .31 +/​-​ .036 –​ –​ Korean twins and

Song, Lee, & family members (aged ≥
Lee, 2010 Emotional eating .25 +/​-​ .098 –​ –​ 30 years); DEBQ scores
–​ –​ adjusted for age and sex
External eating .25 +/​-​ .060 effects: only heritability
estimates and standard
errors were reported

Spanos, Burt, & Shape concern .64 (.29–​.74) 0 (0–​.30) .36 (.26–​.50) Michigan State University
Klump, 2010 0 (0–​.26) .34 (.24–​.49) Twin Registry (MSUTR),
Weight concern .66 (.35–​.76) mean age 20.92
(SD=2.47), two sub-​scales
from the EDE-​Q

Slane, Burt, & Disordered eating .59 (.44–​.71) -​ .41 (.29–​.56) MSUTR: 30-​item, self-​
Klump, 2011 report Minnesota Eating
Behavior Survey (MEBS)
analysed with depressive
symptoms

Mitchell et al., Weight .20 (0–​.34) 0 (0–​.27) .80 (.66–​.96) VTR (aged 25–​65 years):
2011 suppression .25 (0–​.56) 0 (0–​.39) .75 (.44–​1) subtraction of current
from highest weight, no
binge eating and binge
eating with loss of control

Anckarsater Eating problems .47 (.41–​.51) 0 (0–​.04) .53 (.49–​.57) Longitudinal Child and
et al., 2011 .35 (.29–​.40) 0 (0–​.04) .65 (.60–​.70) Adolescent Twin Study
in Sweden (CATSS),
launched in 2004,
EDI collected at 9, 12,
15 years: girls and boys

Ferguson, Disordered eating .40 .07 .53 National Longitudinal

Munoz, behaviours .48 .05 .47 Study of Adolescent
Winegard, & Health (American
Winegard, 2012 adolescents, 4 waves of
data, aged 16 years at
Wave 1, SD=1.75); 13
items (females and males)
1
Self-​report questionnaire unless otherwise noted;
2
Full model or model of best fit (lowest Akaike’s Information Criteria) –​not necessarily significantly better fitting than the full model;
3
All populations were female unless otherwise noted.
Note: A=additive genetic variance; C=shared environmental variance; E=non-​shared environmental variance; CI=confidence interval; NR=not
reported; VTR=Virginia Twin Registry; SCID=Structured Clinical Interview for DSM; BMI=body mass index; ATR=Australian Twin
Registry; EDE=Eating Disorder Examination; SRQ=self-​reported questionnaire; MTFS=Minnesota Twin Family Study; EDI=Eating Disorder
Inventory; EAT=Eating Attitudes Test; DEBQ=Dutch Eating Behaviours Questionnaire.
 Numerous studies have examined comorbid- with the perfectionism measures. Similarly, while
ity between bulimic disorders and substance use genetic factors were more likely to contribute to
disorders (alcohol, illicit drug, and caffeine abuse/​ phenotypic associations between disordered eating
dependence), indicating the correlation between and the personality factors of negative and posi-
genetic risk factors is significant and moderate, tive emotionality, and constraint, shared genetic
ranging from 0.35 to 0.53 (Baker, Mazzeo, & variance between the two phenotypes was limited
Kendler, 2007; Baker, Mitchell, Neale, & Kendler, (Klump, McGue, & Iacono, 2002). While one
2010), with little overlap in environmental risk study has found that disordered eating did not share
factors. No clear conclusions could be drawn with genetic risk factors with neuroticism (Wade et  al.,
respect to AN symptoms. Bivariate modeling indi- 2000), a more recent study has shown that covari-
cated a moderate-​to-​large degree of overlap in addi- ance among binge eating episodes, neuroticism, and
tive genetic factors between alcohol use and binge conscientiousness was attributable to genetic (neu-
eating and compensatory behaviors (0.31–​0.61), roticism ra  =  0.37; conscientiousness ra  =  −0.22)
and no overlap in environmental effects (Slane, and individual-​specific environmental (neuroticism
Burt, & Klump, 2012). Munn-​Chernoff and col- re = 0.28; conscientiousness re = −0.19) influences
leagues (2013) found that one model could explain (Koren et  al., 2014). Such research can inform
variance in both females and males with respect to which personality traits may be an early indicator of
alcohol dependence and binge eating, with a sig- genetic vulnerability to a variety of eating disorder
nificant genetic correlation between the two (0.26, behaviors.
95% CI: 0.12–​0.41), and a smaller overlap between
nonshared environmental influences (0.14, 95% Gene–​Environment Interplay
CI: 0.03–​0.25). In contrast, shared genetic factors Differentiating direct environmental influences
were solely responsible for the correlation between from genetic or environmental confounds “remains
alcohol use disorder and bulimic behaviors (0.33, one of the fundamental problems facing the social
95% CI: 0.18–​0.49) as the nonshared environmen- sciences” (Turkheimer, D’Onofrio, Maes, & Eaves,
tal correlation was not significant (Munn-​Chernoff, 2005, p.  1229). Use of twin studies to investigate
Grant et al., 2015). genotype–​environment correlations and genotype–​
Such studies can also inform how different environment interactions can start to untangle these
aspects of eating disorders relate to each other. For influences, and point us in the direction of the most
example, genetic risk factors for BMI are consis- influential environments to target in interventions.
tently found to be relatively independent of dis- We know little about genotype–​environment corre-
ordered eating, including intentional weight loss lations with respect to eating disorders. One recent
(Keski-​Rahkonnen, Neale, et al., 2005), disordered study used a cotwin control design with females
eating (Slof Op ‘t Landt et  al., 2008), attitudes twins aged 8 to 14 years to examine the reasons for
toward one’s body (Wade, Wilkinson, Ben-​Tovim, the strong associations between exposure to weight-​
2003), eating attitudes and behaviors (Klump conscious peer groups and increased levels of dis-
et al., 2000), and binge eating (Bulik, Sullivan, & ordered eating (O’Connor, Klump, VanHuysse,
Kendler, 2003). A further study showed importance McGue, & Iacono, 2016). The observed asso-
of weight and shape shared little genetic risk with ciations between disordered eating and weight-​
BMI, but that BMI and body dissatisfaction shared conscious peer group association (rs ranging from
common sources of genetic risk (Wade, Zhu, & 0.36 to 0.50) were largely due to genetic and/​or
Martin, 2011). In other words, eating disorders are shared environmental selection factors rather than
not primarily weight disorders, given only 3% to pure socialization effects.
20% of the variance between BMI and disordered There is an accumulating number of informa-
eating is shared in women. tive studies of genotype–​environment interactions,
Much less work has examined shared genetic risk summarized in Table 5.4. Four of the six studies
between eating and continuous measures. Wade and show support for such an interaction with respect
Bulik (2007) found that there was very little shared to binge eating and dietary restraint, disordered
genetic risk between various dimensions of perfec- eating and puberty, disordered eating and divorce,
tionism and importance of weight and shape, where and disordered eating and weight-​related peer teas-
data supported an independent rather than com- ing. Genetic and nonshared environmental factors
mon pathways model, and only 6% of the genetic for binge eating increased at higher levels of dietary
risk factors for the importance construct overlapped restraint, present after controlling for age, BMI, and

96 Genetic Influences
Table  5.4 Summary of  studies examining genotype–​environment interactions related to  disordered eating
in twin samples
Authors N twins (age) Disordered eating indicator Self-​report Environmental
measure

Racine, Burt, Iacono, 1678 (11 to 29 years) Self-​report binge eating from Dietary restraint: Eating
McGue, & Klump, 2011 Minnesota Eating Behavior Disorder Examination
Survey (MEBS) Questionnaire

Suisman, Thompson, Keel, 1064 (8 to 25 years) Self-​report internalization of Puberty: Pubertal

Burt, Neale, Boker, Sisk, & the thin ideal: Sociocultural Development Scale
Klump, 2014 Attitudes Toward Appearance
Questionnaire-​3

Klump, Burt, McGue, & 510 (13 to 16 years) Self-​report disordered Puberty: Pubertal
Iacono, 2007 eating: Total Score from the Development Scale
MEBS

Suisman, Burt, McGue, 1810 (14 to 28 years) Self-​report disordered Parental divorce
Iacono, & Klump, 2011 eating: MEBS

O’Connor, Klump, 1534 (16 to 20 years) Self-​report body Parental divorce

VanHuysse, McGue, & dissatisfaction: Body Rating
Iacono, 2016 Scale

Fairweather-​Schmidt & 685 (3 waves between Eating Disorder Examination Peer teasing
Wade, 2017 ages 12 and 19 years) global score

genetic and environmental overlap among dietary 2007). Cross-​sectional studies also show the same
restraint and binge eating (Racine, Burt, Iacono, pattern (Klump et al., 2010), with genetic influences
McGue, & Klump, 2011). Results suggested that being modest in preadolescent twins, but becoming
genetic risk for binge eating and dietary restraint significant from early adolescence through middle
could enhance individual differences in risk for binge adulthood. In contrast, shared environmental fac-
eating. Significant increases in genetic influence on tors made the largest contributions in preadolescent
disordered eating with advancing pubertal develop- twins, then decreased with age, while nonshared
ment were also shown in an adolescent sample of environmental effects remained relatively constant
twins (Klump et al., 2007). Significantly higher her- across the different age groups.
itability of body dissatisfaction was found in twins This work has led to studies examining the
from divorced versus intact families (Suisman, Burt, role of pubertal development and hormones as an
McGue, Iacono, & Klump, 2011). Finally, as levels environmental trigger of heritability for disordered
of peer teasing increased, both genetic and environ- eating. This research has implicated higher levels
mental influences on disordered eating strength- of prenatal circulating estrogen as increasing risk
ened; genetic sources increased proportionally more for the probability of disordered eating (Culbert,
than environmental sources (Fairweather-​Schmidt Breedlove, Burt & Klump, 2008; Procopio &
& Wade, 2017). These studies have the ability to Marriott, 2007). Conversely, prenatal testosterone
identify influential environments that could be pri- exposure may decrease disordered eating attitudes
oritized as targets in prevention efforts. (Culbert, Breedlove, Sisk, Burt, & Klump, 2013).
Related research in twins has shown that estimates
Developmental Changes to Genetic and of genetic influences for emotional eating were two
Environmental Risk Factors times higher in post-​as compared with preovula-
Work with the adolescent twin registry in tion, but that preovulation was marked by a pre-
Minnesota has found that both shared and non- dominance of environmental influences, including
shared environment contributing to disordered eat- shared environmental effects that have not been
ing decreases as children become adolescents and previously detected for binge eating phenotypes in
heritability increases significantly (Klump et  al., adulthood (Klump et al., 2015). Independent tests

Wade, Bulik 97
of the hypothesis that prenatal exposure to sex hor- psychiatric disorders; namely, decades of investiga-
mones influences twins’ risk for eating disorders tions that focused on single candidate genes, fol-
(such that individuals with a female cotwin would lowed by linkage studies that despite initial promise
be more likely than individuals with a male cotwin failed to deliver implicated genes, and finally suc-
to meet diagnostic criteria for an eating disorder) has cess using hypothesis-​free genomewide association
not been supported in other twin samples (Raevuori study (GWAS) approaches. Rather than focusing on
et al., 2008; Lydecker et al., 2012), indicating that single genes driven by biological hypotheses, GWAS
this hypothesis needs to be further tested. represent an unbiased search of the entire human
A longitudinal investigation of weight and shape genome for loci that differ significantly between
concern over adolescence (12 to 16 years) found a large samples of individuals with a given trait in
model containing additive genetic contributions, comparison to similarly large samples of individuals
as well as both nonshared and shared environment, without the target trait.
was a best fit to the data (Wade, Hansell et  al., Rather than reviewing the fairly fruitless past of
2013). Consistent with previous findings, the role candidate gene studies, we refer the reader to other
of genes and nonshared environment increased as reviews that document the history of genetic stud-
twins became older, with a new source of heritability ies in eating disorders (Brandys, de Kovel, Kas,
emerging at ages 13 to 15 years. At that time, a range van Elburg, & Adan, 2015; Hinney, Scherag, &
of environmental stressors were also strongly associ- Hebebrand, 2010; Scherag, Hebebrand, & Hinney,
ated with weight and shape concerns, including peer 2010; Slof-​Op ‘t Landt et al., 2005). The same holds
teasing about weight, adverse life events, media inter- with linkage studies, which rely on large pedigrees
nalization, and pressure to be thin. A subsequent lon- with multiple affected individuals, or large samples
gitudinal study of disordered eating examining the of affected relative pairs in order to identify regions
twins from ages 12 to 19 years (Fairweather-​Schmidt of the genome that are shared identical by descent
& Wade, 2015) also showed that both genetic influ- (Grice et al., 2002; Bulik et al., 2005).
ences and nonshared environmental influences Bypassing this literature allows us to focus on
increased over adolescence, but shared environmen- rapid advances in technology and statistical genet-
tal influences decreased. While nonshared environ- ics in order to advance discovery science in eating
mental sources active at ages 12–​15 years continued disorders on a genomewide level. A GWAS remains
to contribute at 16–​19  years, new sources of both an “association” technique, meaning you determine
additive genetic and nonshared environmental risk genes that are associated with a particular trait like
were introduced at ages 16–​19 years, consistent with AN by comparing large samples of individuals with
observations that there is a later emergence of binge lifetime AN to large samples of ancestrally matched
and purge symptomatology in late adolescence and individuals who have never had an eating disorder.
early adulthood (Favaro, Caregaro, Tenconi, Bosello, The operative term in GWAS is “large,” as sample
& Santonastaso, 2009). Weight-​related peer teasing size is the critical ingredient for successful discovery.
in early-​mid adolescence was the only independent Whereas our initial candidate gene studies in eat-
predictor of increases of disordered eating in later ing disorders comprised one or a few hundred cases
adolescence. All of these studies also indicate key and controls, we now know that successful GWAS
developmental periods for the timing of preven- require tens of thousands of individuals in order
tion programs in order for them to exert maximum to identify implicated genes that are of small to
effect. moderate effect. As such, the bulk of effort that has
been put toward eating disorders genetics over the
past 5 years has been focused on increasing sample
Molecular Genetic Studies: The Current size—​and most of this work has been on anorexia
State of Play nervosa with genetic science on BN and binge-​
Rapid advances in methodology and technology eating disorder falling woefully behind.
have enabled revolutionary changes in the study of The two initial GWAS conducted on anorexia
the genetics of eating disorders. nervosa were conducted by the Children’s Hospital
of Philadelphia and Price Foundation consortia
Identifying Genes Associated with Eating (Wang et  al., 2011) and the Genetic Consortium
Disorders for Anorexia Nervosa and Wellcome Trust Case
The history of molecular genetics in eating disor- Control Consortium 3 (WTCCC3) (Boraska et al.,
ders is not dissimilar to that seen with other major 2014). Neither GWAS identified genomewide

98 Genetic Influences
significant loci for AN. Two additional GWAS has support. When accompanied by borderline per-
explored scales of the Eating Disorders Inventory sonality disorder or suicidality, BN is associated with
(Boraska et al., 2012) and 15 self-​report questions hypermethylation of certain GR exon 1C promoter
relating to the presence and treatment of different sites (Steiger, Labonte, Groleau, Turecki, & Israel,
behaviors and cognitions that define disordered eat- 2013). Women with binge-​purge eating syndromes
ing (Wade, Gordon et al., 2013) and similarly did who evidenced low levels of dopaminergic neuro-
not identify implicated loci. transmission (carriers of high function COMT and
In retrospect, there is nothing surprising about low-​function DRD4 7R alleles) showed more life-
these results, as our current understanding clarifies time substance abuse (Steiger et al., 2016). Women
that all of the samples in these investigations were with bulimia spectrum disorder with borderline
prohibitively small. Following in the footsteps of personality disorder showed small, but significant
others researching psychiatric disorders who have increases in DRD2 methylation levels compared
demonstrated the positive effect of enhancing sam- to women with no eating disorder (Groleau et al.,
ple size, we formed the Eating Disorders Working 2014). In women with bulimia spectrum disorder,
Group of the Psychiatric Genomics Consortium carriers of the s allele of 5HTTLPR showed signifi-
and united researchers and clinicians from around cantly more affective instability, behavioral impul-
the world to consolidate samples, collect additional sivity, interpersonal insecurity, comorbid borderline
samples, and achieve the sample sizes necessary for personality disorder, and lower density (Bmax) of
gene discovery. This unparalleled collaboration led paroxetine-​binding sites than women without this
to a joint analysis of the CHOP/​Price Foundation polymorphism (Steiger et al., 2005).
and GCAN/​ WTCCC3 samples, which boosted Early experience of abuse has been implicated as
statistical power by boosting sample size. As pre- a specific environment of interest. The 5HTTLPR
dicted, this approach led to the identification of polymorphism, linked to reduced serotonin uptake
the first genomewide significant “hit” for AN on activity in both women with bulimic syndromes
chromosome 12. The GWAS included 3,495 cases and their unaffected first-​degree relatives (Bruce,
and 10,982 controls. The identified gene was in a Steiger, Kin, Ng, & Israel, 2006), has been shown
gene-​rich area and harbored multiple autoimmune to interact with childhood abuse, explaining signifi-
GWAS hits for diabetes, polycystic ovary syndrome, cant variance in stimulus seeking, insecure attach-
rheumatoid arthritis, vitiligo, and asthma (Duncan ments, and borderline personality disorder in BN
et al., 2017). Sample collection is ongoing, and we (Steiger et al., 2007). This polymorphism has also
estimate that we will have ~20,000 cases by the end been associated with poorer treatment response
of 2017. Genetic correlations further supported the in BN (Steiger et al., 2008). Additionally, BN co-​
negative genetic correlation with BMI and multiple occurring with childhood abuse or borderline per-
anthropometric and metabolic phenotypes further sonality disorder is associated with a propensity
underscoring the possible role of metabolic factors toward elevated methylation at specific BDNF pro-
in the etiology of AN. This work opens new areas moter region sites (Thaler et al., 2014), and a sig-
of discovery for eating disorders and paves the way nificant BcII × abuse interaction has been detected,
for similar efforts in BN and binge-​eating disorder. suggesting stress-​ induced alterations in glucocor-
ticoid sensitivity contribute to BN and depressive
Interactions Between Specific Genetic disturbances (Steiger et al., 2012).
Variants and Environments
It has been argued that epigenetic processes (the Conclusion
impact of the environment on gene expression) In summary, genetic factors are clearly an impor-
provide a biological “platform” on which perinatal tant part of the explanatory framework for the eti-
insults, life stresses, and consequences of malnutri- ology of eating disorders; however, they do not act
tion may interact with genetic susceptibility, lead- alone. Eating disorders are complex traits influenced
ing to mood-​and impulse-​regulation problems that by multiple interacting genetic and environmen-
trigger disordered eating (Steiger & Thaler, 2016). tal factors. As with all complex traits, genetics are
The relevance of epigenetic processes to eating dis- also not destiny. Multiple factors along the devel-
orders has been shown in several studies (e.g., Booij opmental path can modify the effects and potency
et  al., 2015; Thaler et  al., 2014). The association of genetic factors. Protective genetic and environ-
between epigenetic processes and emotional regula- mental factors, which are harder to identify, can also
tion difficulties in women with eating disorders also counter the potential influence of risk genes. With

Wade, Bulik 99
the emergence of large GWAS for AN, we will soon Boraska, V., Davis, O. S., Cherkas, L. F., Helder, S. G., Harris, J.,
be extraordinarily well positioned to decode the Krug, I.,  . . .  Zeggini, E. (2012). Genome-​wide association
analysis of eating disorder-​related symptoms, behaviors, and
genetic factors that influence risk for this disorder, personality traits. American Journal of Medical Genetics. Part
with similar efforts for BN and binge-​eating disor- B, Neuropsychiatric Genetics, 159B, 803–​811.
der on the horizon. Boraska, V., Franklin, C. S., Floyd, J. A., Thornton L. M.,
Huckins, L. M., Southam, L.,   .  .  .  Bulik, C. M. (2014).
Future Directions A genome-​ wide association study of anorexia nervosa.
Molecular Psychiatry, 19, 1085–​1094.
Advancing our understanding of the etiology
Bouchard, T. J., & McGue, M. (2003). Genetic and environ-
of eating disorders will require:  (1)  large sample mental influences on human psychological differences.
sizes for GWAS for all eating disorders, (2)  closer Journal of Neurobiology, 54, 4–​45.
attention to the manner in which phenotypes are Brandys, M. K., de Kovel, C. G., Kas, M. J., van Elburg, A. A., &
assessed and defined, (3) a focus on the components Adan, R. A. (2015). Overview of genetic research in anorexia
nervosa: The past, the present and the future. International
of eating disorder diagnoses, (4) further use of fam-
Journal of Eating Disorders, 48, 814–​825.
ily and twin studies to identify endophenotypes that Bruce, K. R., Steiger, H., Kin, N. M. K., Ng, Y., & Israel, M.
can support better delineation of eating disorders, (2006). Reduced platelet [3H]paroxetine binding in anorexia
(5) further development of analytic techniques that nervosa:  Relationship to eating symptoms and personality
can help us differentiate direct environmental influ- pathology. Psychiatry Research, 142, 225–​232.
Bulik C. M., Bacanu S. A., Klump K. L., Fichter M. M., Halmi K.
ences in family and twin models such that specific
A., Keel P.,  . . .  Devlin B. (2005). Selection of eating disorders
sources of shared environment can be identified, phenotypes for linkage analysis. American Journal of Medical
and (6)  an increase in the research that examines Genetics. Part B, Neuropsychiatric Genetics, 139, 81–​87.
genotype–​environment interactions using genome- Bulik, C. M., Hebebrand, J., Keski-​Rahkonen, A., Klump, K.
wide data. Ultimately, the vision for this research L., Reichborn-​Kjennerud, T., Mazzeo, S. E., & Wade, T. D.
(2007). Genetic epidemiology, endophenotypes, and eat-
is that molecular genetic studies, supported by
ing disorder classification. International Journal of Eating
ongoing work with twins and families, will identify Disorders, 40, S52–​S60.
implicated biological pathways that will improve Bulik, C. M., Sullivan, P. F., & Kendler, K. S. (1998). Heritability
detection and risk prediction and allow us to reposi- of binge-​ eating and broadly defined bulimia nervosa.
tion existing interventions or develop new targeted Biological Psychiatry, 44, 1210–​1218.
Bulik, C. M., Sullivan, P. F., & Kendler, K. S. (2003). Genetic
interventions that can assist with recovery from
and environmental contributions to obesity and binge eat-
these challenging illnesses. ing. International Journal of Eating Disorders, 33, 293–​298.
Bulik, C. M., Sullivan, P. F., Tozzi, F., Furberg, H., Lichtenstein,
References P., & Pedersen, N. L. (2006). Prevalence, heritability, and
Anckarsater, H., Lundstrom, S., Kollberg, L., Kerekes, N., Palm, prospective risk factors for anorexia nervosa. Archives of
C., Carlstrom, E.,  . . .  Lichtenstein, P. (2011). The Child General Psychiatry, 63, 305–​312.
and Adolescent Twin Study in Sweden (CATSS). Twin Bulik, C. M., Sullivan, P. F., Wade, T. D., & Kendler, K. S. (2000).
Research and Human Genetics, 14, 495–​508. Twin studies of eating disorders:  A  review. International
Anokhin, A. P., Golosheykin, S., Grant, J. D., & Heath, A. C. Journal of Eating Disorders, 27, 1–​20.
(2010). Developmental and genetic influences on prefrontal Bulik-​Sullivan, B., Finucane, H., Anttila, V., Gusev, A., Day, F.,
function in adolescents: A longitudinal twin study of WCST ReproGen Consortium,  . . .  Neale, B. M. (2015). An atlas of
performance. Neuroscience Letters, 472, 119–​122. genetic correlations across human diseases and traits. Nature
Archibald, A. B., Linver, M. R., Graber, J. A., & Brooks-​Gunn, J. Genetics, 47, 1236–​1241.
(2002). Parent-​adolescent relationships and girls’ unhealthy Bulik-​Sullivan, B. K., Loh, P. R., Finucane, H. K., Ripke, S., &
eating:  Testing reciprocal effects. Journal of Research on Yang, J., Schizophrenia Working Group of the Psychiatric
Adolescence, 12, 451–​461. Genomics Consortium  . . .  Neale, B. M. (2015). LD Score
Baker, J. H., Mazzeo, S. E., & Kendler, K. S. (2007). Association regression distinguishes confounding from polygenicity
between broadly defined bulimia nervosa and drug use dis- in genome-​ wide association studies. Nature Genetics, 47,
orders:  Common genetic and environmental influences. 291–​295.
International Journal of Eating Disorders, 40, 673–​678. Burt, S. A. (2009). Rethinking environmental contributions to
Baker, J. H., Mitchell, K. S., Neale, M. C., & Kendler, K. S. (2010). child and adolescent psychopathology:  A  meta-​analysis of
Eating disorder symptomatology and substance use disor- shared environmental influences. Psychological Bulletin, 135,
ders:  Prevalence and shared risk in a population based twin 608–​637.
sample. International Journal of Eating Disorders, 43, 648–​658. Cassin, S. E., & von Ranson, K. M. (2005). Personality and eat-
Booij, L., Casey, K. F., Antunes, J. M., Szyf, M., Joober, R., ing disorders: A decade in review. Clinical Psychology Review,
Israel, M., & Steiger, H. (2015). DNA methylation in indi- 25, 895–​916.
viduals with anorexia nervosa and in matched normal-​eater Cederlof, M., Thornton, L. M., Baker, J., Lichtenstein, P.,
controls:  A  genome-​wide study. International Journal of Larsson, H., Ruck, C.,   .  .  .  Mataix-​ Cols, D. (2015).
Eating Disorders, 48, 874–​882. Etiological overlap between obsessive-​compulsive disorder

100 Genetic Influences

 and anorexia nervosa:  A  longitudinal cohort, multigenera- Fairburn, C. G., Cooper, Z., Bohn, K., O’Connor, M. E., Doll,
tional family and twin study. World Psychiatry, 14, 333–​338. H. A., & Palmer, R. L. (2007). The severity and status of
Collier, D. A., & Treasure, J. L. (2004). The aetiology of eating eating disorder NOS:  Implications for DSM-​V. Behaviour
disorders. British Journal of Psychiatry, 185, 363–​365. Research and Therapy, 45, 1705–​1715.
Cooper, P. J., & Fairburn, C. G. (1993). Confusion over the core Fairburn, C. G., Cowen, P. J., & Harrison, P. J. (1999). Twin
psychopathology of bulimia nervosa. International Journal of studies and the etiology of eating disorders. International
Eating Disorders, 13, 385–​389. Journal of Eating Disorders, 26, 349–​358.
Cross-​Disorder Group of the Psychiatric Genomics Consortium, Fairburn C. G., & Harrison P. H. (2003). Eating disorders.
Lee, S. H., Ripke, S., Neale, B. M., Faraone, S. V., Purcell, S. Lancet, 361(9355), 407–​416.
M.,  . . .  International Inflammatory Bowel Disease Genetics Fairweather-​Schmidt, A. K., & Wade, T. D. (2014). DSM-​5
Consortium. (2013). Genetic relationship between five psy- eating disorders and other specified eating and feeding dis-
chiatric disorders estimated from genome-​wide SNPs. Nature orders:  Is there a meaningful differentiation? International
Genetics, 45, 984–​994. Journal of Eating Disorders, 47, 524–​533.
Culbert, K. M., Breedlove, M. S., Burt, S. A., & Klump, K. L. Fairweather-​Schmidt, A. K., & Wade, T. D. (2015). Changes
(2008). Prenatal hormone exposure and risk for eating dis- in genetic and environmental influences on disordered eat-
orders:  A  comparison of opposite-​sex and same-​sex twins. ing between early and late adolescence: A longitudinal twin
Archives of General Psychiatry, 65, 329–​336. study. Psychological Medicine, 45, 3249–​3258.
Culbert, K. M., Breedlove, M. S., Sisk, C. L., Burt, S. A., & Fairweather-​Schmidt, A. K., Wade, T. D. (2017). Does peer-​
Klump, K. L. (2013). The emergence of sex differences in teasing moderate genetic and environmental risk for disor-
risk for disordered eating attitudes during puberty:  A  role dered eating? British Journal of Psychiatry, 201, 350–​355.
for prenatal testosterone exposure. Journal of Abnormal Fall, T., & Ingelsson, E. (2014). Genome-​ wide association
Psychology, 122, 420–​432. studies of obesity and metabolic syndrome. Molecular and
Culbert, K. M., Racine, S. E., & Klump, K. L. (2015). Research Cellular Endocrinology, 382, 740–​757.
review:  What we have learned about the causes of eating Favaro, A., Caregaro, L., Tenconi, E., Bosello, R., & Santonastaso,
disorders—​a synthesis of sociocultural, psychological, and P. (2009). Time trends in age of onset of anorexia nervosa
biological research. Journal of Child Psychology and Psychiatry, and bulimia nervosa. Journal of Clinical Psychiatry, 70,
56, 1141–​1164. 1715–​1721.
Dellava, J., Kendler, K. S., & Neale, M. (2011). Generalized Ferguson, C. J., Munoz, M. E., Winegard, B., & Winegard, B.
anxiety disorder and anorexia nervosa:  Evidence of shared (2012). The influence of heritability, neuroticism, maternal
genetic variation. Depression and Anxiety, 28, 728–​733. warmth and media use on disordered eating behaviors: A pro-
den Braber, A., Zilhao, N. R., Fedko, I. O., Hottenga, J. J., Pool, spective analysis of twins. Psychiatric Quarterly, 83, 353–​360.
R., Smit, D. J.,   .  .  .  Boomsma, D. I. (2016). Obsessive-​ Foley, D., Neale, M. C., & Kendler, K. S. (1998). Reliability
compulsive symptoms in a large population-​ based twin-​ of a lifetime history of major depression:  Implications for
family sample are predicted by clinically based polygenic heritability and co-​ morbidity. Psychological Medicine, 28,
scores and by genome-​wide SNPs. Translational Psychiatry 857–​870.
and Psychiatry, 6, e731. Ghaderi, A. (2003). Structural modeling analysis of prospective
Devlin, B., Bacanu, S., Klump, K. L., Bulik, C. M., Fichter, M. risk factors for eating disorder. Eating Behaviors, 3, 387–​396.
M., Halmi, K.,  . . .  Kaye, W. H. (2002). Linkage analysis Goodman, A., Heshmati, A., Malki, N., & Koupil, I. (2014).
of anorexia nervosa incorporating behavioural covariates. Associations between birth characteristics and eating disor-
Human Molecular Genetics, 11, 689–​696. ders across the life course: Findings from 2 million males and
Dring, G. (2015). Anorexia runs in families:  Is this due to females born in Sweden, 1975–​1998. American Journal of
genes or the family environment? Journal of Family Therapy, Epidemiology, 179, 852–​863.
37, 79–​92. Grice, D. E., Halmi, K. A., Fichter, M. M., Strober, M.,
Dryer, R., Tyson, G. A., & Kiernan, M. J. (2013). Bulimia ner- Woodside, D. B., Treasure, J. T.,   .  .  .  Berrettini, W. H.
vosa: Professional and lay people’s beliefs about the causes. (2002). Evidence for a susceptibility gene for anorexia ner-
Australian Psychologist, 48, 338–​344. vosa on chromosome 1. American Journal of Human Genetics,
Dryer, R., Uesaka, Y., Manalo, E., & Tyson, G. A. (2015). 70, 787–​792.
Cross-​cultural examination of beliefs about the causes of Groleau, P., Joober, R., Israel, M., Zeramdini, N., DeGuzman,
bulimia nervosa among Australian and Japanese females. R., & Steiger, H. (2014). Methylation of the dopamine D2
International Journal of Eating Disorders, 48, 176–​186. receptor (DRD2) gene promoter in women with a bulimia-​
Duncan, L., Yilmaz, Z., Gaspar, H., Walters, R., Goldstein, J., spectrum disorder:  Associations with borderline person-
Anttila, V., Bulik-​Sullivan, B.,   .  .  .  Bulik, C. M. (2017). ality disorder and exposure to childhood abuse. Journal of
Significant Locus and Metabolic Genetic Correlations Psychiatric Research, 48, 121–​127.
Revealed in Genome-​Wide Association Study of Anorexia Hebebrand, J., & Remschmidt, H. (1995). Anorexia nervosa
Nervosa. American Journal of Psychiatry, doi: 10.1176/​appi. viewed as an extreme weight condition:  Genetic implica-
ajp.2017.16121402. [Epub ahead of print]. tions. Human Genetics, 95, 1–​11.
Eisler, I. (2005). The empirical and theoretical base of fam- Hemiri, L., Kuja-​ Halkola, R., Larsson, H., & Jayaram-​
ily therapy and multiple family day therapy for adolescent Lindstrom, N. (2016). Genetic overlap between impulsivity
anorexia nervosa. Journal of Family Therapy, 27, 104–​131. and alcohol dependence:  A  large scale national twin study.
Fairburn, C. G. (1991). The heterogeneity of bulimia nervosa Psychological Medicine, 46, 1091–​1102.
and its implications for treatment. Journal of Psychosomatic Hewitt, J. K. (1997). Behavior genetics and eating disorders.
Research, 35, 3–​9. Psychopharmacology Bulletin, 33, 355–​358.

Wade, Bulik 101
Hinney, A., Friedel, S., Remschmidt, H., & Hebebrand, J. drive for thinness in young adult twins. International Journal
(2004). Genetic risk factors in eating disorders. American of Eating Disorders, 37, 188–​199.
Journal of PharmacoGenomics, 4, 209–​223. Keski-​Rahkonnen A., Neale, B. M., Bulik, C. M., Pietiläinen, K.
Hinney, A., Kesselmeier, M., Jall, S., Volckmar, A. L., Focker, H., Rose R. J., Kaprio K., & Rissanen, A. (2005). Intentional
M., Antel., J.,   .  .  .  Hebebrand, J. (2017). Evidence for weight loss in young adults: Sex-​specific genetic and environ-
three genetic loci involved in both anorexia nervosa risk and mental effects. Obesity Research, 13, 745–​753.
variation of body mass index. Molecular Psychiatry, 22(2), Kessler R. C. (2002). The categorical versus dimensional assess-
192–​201. ment controversy in the sociology of mental illness. Journal
Hinney, A., Scherag, S., & Hebebrand, J. (2010). Genetic find- of Health and Social Behavior, 43, 171–​188.
ings in anorexia and bulimia nervosa. Progress in Molecular Klump, K. L., Burt, S. A., McGue, M., & Iacono, W. G. (2007).
Biology and Translational Science, 94, 241–​270. Changes in genetic and environmental influences on disor-
Hinney, A., Vogel, C. I., & Hebebrand J. (2010). From mono- dered eating across adolescence: A longitudinal twin study.
genic to polygenic obesity: Recent advances. European Child Archives of General Psychiatry, 64, 1409–​1415.
and Adolescent Psychiatry, 19, 297–​310. Klump, K. L., Burt, S. A., Spanos, A., McGue, M., Iacono, W.
Iranzo-​Tatay, C., Gimeno-​ Clemente, N., Barbera-​ Fons, M., G., & Wade, T. D. (2010). Age differences in genetic and
Rodriguez-​ Campayo, M. A., Rojo-​ Bofill, L., Livianos-​ environmental influences on weight and shape concerns.
Aldana, L.,  . . .  Rojo-​Moreno, L. (2015). Genetic and envi- International Journal of Eating Disorders, 43, 679–​688.
ronmental contributions to perfectionism and its common Klump, K. L., Hildebrandt, B. A., O’Connor, S. M., Keel, P. K.,
factors. Psychiatry Research, 230, 932–​939. Neale, M., Sisk, C. L.,  . . .  Burt, S. A. (2015). Changes in genetic
Jacobi, C., & Fittig, E. (2010). Psychosocial risk factors for eating risk for emotional eating across the menstrual cycle: A longitu-
disorders. In W. S. Agras (Ed.), The Oxford handbook of eating dinal study. Psychological Medicine, 45, 3227–​3237.
disorders (pp. 123–​136). New York: Oxford University Press. Klump, K. L., Kaye, W. H., & Strober, M. (2001). The evolving
Jacobi, C., Hayward, C., de Zwaan, M., Kraemer, H. C., & Agras, genetic foundations of eating disorders. Psychiatric Clinics of
W. S. (2004). Coming to terms with risk factors for eating North America, 24, 215–​225.
disorders:  Application of risk terminology and suggestions Klump, K. L., McGue, M., & Iacono, W. G. (2000). Age differ-
for a general taxonomy. Psychological Bulletin, 130, 19–​65. ences in genetic and environmental influences on eating atti-
Javaras, K. N., Laird, N. M., Reichborn-​Kjennerud, T., Bulik, tudes and behaviors in preadolescent and adolescent twins.
C. M., Pope, H. G., & Hudson, J. I. (2008). Familiality and Journal of Abnormal Psychology, 109, 239–​251.
heritability of binge eating disorder: Results of a case-​control Klump, K. L., McGue, M., & Iacono, W. G. (2002). Genetic
family study and a twin study. International Journal of Eating relationships between personality and eating attitudes and
Disorders, 41, 174–​179. behaviors. Journal of Abnormal Psychology, 111, 380–​389.
Johnson, C., & Flach, A. (1985). Family characteristics of 105 Klump, K. L., Miller, K., Keel, P., McGue, M., & Iacono, W.
patients with bulimia. American Journal of Psychiatry, 142, (2001). Genetic and environmental influences on anorexia
1321–​1324. nervosa syndromes in a population-​ based twin sample.
Johnston, J., Cohen, P., Kasen, S., & Brook, J. (2002). Childhood Psychological Medicine, 31, 737–​740.
adversities associated with risk for eating disorders or weight Koren, R., Munn-​Chernoff, M. A., Duncan, A. E., Bucholz, K.
problems during adolescence or early adulthood. American K., Madden, P. A. F., Heath, A. C., & Agrawal, A. (2014). Is
Journal of Psychiatry, 159, 394–​400. the relationship between binge eating episodes and personal-
Kamakura, T., Ando, J., Ono, Y., & Maekawa, H. (2003). A twin ity attributable to genetic factors? Twin Research and Human
study of genetic and environmental influences on psycho- Genetics, 17, 65–​71.
logical traits of eating disorders in a Japanese female sample. Kortegaard, L. S., Hoerder, K., Joergensen, J., Gillberg, C., &
Twin Research, 6, 292–​296. Kyvik, K. (2001). A preliminary population-​ based twin
Kendler, K. S. (2013). What psychiatric genetics has taught us study of self-​reported eating disorder. Psychological Medicine,
about the nature of psychiatric illness and what is left to 31, 361–​365.
learn. Molecular Psychiatry, 18, 1058–​1066. Krebs, G., Waszczuk, M. A., Zavos, H. M., Bolton, D., & Eley,
Kendler, K. S., MacLean, C., Neale, M., Kessler, R. C., Heath, T. C. (2015). Genetic and environmental influences on
A., & Eaves, L. (1991). The genetic epidemiology of bulimia obsessive-​compulsive behaviour across development: A lon-
nervosa. American Journal of Psychiatry, 148, 1627–​1637. gitudinal twin study. Psychological Medicine, 45, 1539–​1549.
Kendler, K. S., Neale, M., Kessler, R. C., Heath, A., & Eaves, L. Locke, A. E., Kahali, B., Berndt, S. I., Justice, A. E., Pers, T.
(1993). The lifetime history of major depression. Reliability H., Day, F. R.,  . . .  Speliotes, E. K. (2015). Genetic studies
of diagnosis and heritability. Archives of General Psychiatry, of body mass index yield new insights for obesity biology.
50, 863–​870. Nature, 518, 197–​206.
Kendler, K. S., Thornton, L. M., & Pedersen, N. L. (2000). Lydecker, J. A., Pisetsky, E. M., Mitchell, K. S., Thornton, L. M.,
Tobacco consumption in Swedish twins reared apart and Kendler, K. S., Reichborn-​Kjennerud, T.,  . . .  Mazzeo, S. E.
reared together. Archives of General Psychiatry, 57, 886–​892. (2012). Association between co-​twin sex and eating disorders
Kendler, K. S., Walters, E. E., Neale, M. C., Kessler, R. C., in opposite sex twin pairs: Evaluations in North American,
Heath, A. C. & Eaves, L. J. (1995). The structure of the Norwegian, and Swedish samples. Journal of Psychosomatic
genetic and environmental risk factors for six major psychiat- Research, 72, 73–​77.
ric disorders in women: Phobia, generalized anxiety disorder, Lyons M. J., & Bar J. L. (2001). Is there a role for twin studies
panic disorder, bulimia, major depression, and alcoholism. in the molecular genetics era? Harvard Review of Psychiatry,
Archives of General Psychiatry, 52, 374–​383. 9, 318–​323.
Keski-​Rahkonnen, A., Bulik, C. M., Neale B. M., Rose R. J., McKnight Investigators. (2003). Risk factors for the onset of eat-
Rissanen A., & Kaprio, K. (2005). Body dissatisfaction and ing disorders in adolescent girls:  Results of the McKnight

102 Genetic Influences

 longitudinal risk factor study. American Journal of Psychiatry, Neale, M. C., Eaves, L. J., & Kendler, K. S. (1994). The power of
160, 248–​254. the classical twin study to resolve variation in threshold traits.
Maes, H. H., Neale, M. C., & Eaves, L. J. (1997). Genetic and Behavior Genetics, 24, 239–​258.
environmental factors in relative body weight and human Neiderhiser, J. M. (2001). Understanding the role of genome
adiposity. Behavior Genetics, 27, 325–​351. and environment: Methods in genetic epidemiology. British
Martin, N. G., Boomsma, D., & Machin, G. (1997). A twin-​ Journal of Psychiatry, 178 (Suppl. 40), s12–​s17.
pronged attack on complex traits. Nature Genetics, 17, 387–​392. Neumark-​Sztainer, D., Eisenberg, M. E., Fulkerson, J. A., Story,
Martin, N. G., Eaves, L. J., Kearsey, M. J., & Davies, P. (1978). M., & Larson, N. I. (2008). Family meals and disordered
The power of the classical twin study. Heredity, 40, 97–​116. eating in adolescents. Archives of Pediatric and Adolescent
Mazzeo, S. E., Mitchell, K. S., Bulik, C. M., Aggen, S. H., Medicine, 162, 17–​22.
Kendler, K. S., & Neale, M. C. (2010). A twin study of spe- O’Connor, S. M., Klump, K. L., VanHuysse, J. L., McGue, M.,
cific bulimia nervosa symptoms. Psychological Medicine, 40, & Iacono, W. (2016). Does parental divorce moderate the
1203–​1213. heritability of body dissatisfaction? An extension of previous
Mazzeo, S. E., Mitchell, K. S., Bulik, C. M., Reichborn-​ gene-​environment interaction effects. International Journal of
Kjennerud, T., Kendler, K. S., & Neale, M. C. (2009). Eating Disorders, 49, 188–​192.
Assessing the heritability of anorexia nervosa symptoms Peterson, C. M., Baker, J. H., Thornton, L. M., Trace, S. E.,
using a marginal maximal likelihood approach. Psychological Mazzeo, S. E., Neale, M. C.,   .  .  .  Bulik, C. M. (2016).
Medicine, 39, 463–​473. Genetic and environmental components to self-​ induced
Meaney, M. J., Aitken, D. H., Bodnoff, S. R., Iny, L. J., vomiting. International Journal of Eating Disorders, 49,
Tatarewicz, J. E., & Sapolsky, R. M. (1985). Early postna- 421–​427.
tal handling alters glucocorticoid receptor concentrations in Pinheiro, A. P., Sullivan, P. F., Bacaltchuck, J., Prado-​ Lima,
selected brain regions. Behavioral Neuroscience, 99, 765–​770. P. A., & Bulik, C. M. (2006). Genetics in eating disor-
Mitchell, K. S., Neale, M. C., Bulik, C. M., Aggen, S. H., ders: Extending the boundaries of research. Revista Brasileira
Kendler, K. S., & Mazzeo, S. E. (2010). Binge eating disorder, de Psiquiatria, 28, 218–​225.
a symptom-​level investigation of genetic and environmental Plomin, R. (2000). Behavioural genetics in the 21st century.
influences on liability. Psychological Medicine, 40, 1899–​1906. International Journal of Behavioural Development, 24, 30–​34.
Mitchell, K. S., Neale, M. C., Bulik, C. M., Lowe, M., Maes, H. Plomin, R., DeFries, J. C., & McClearn, G. E. (1990). Behavioral
H., Kendler, K. S., & Mazzeo, S. E. (2011). An investigation genetics: A primer (2nd ed.). New York, NY: Freeman.
of weight suppression in a population-​based sample of female Procopio, M., & Marriott, P. (2007). Intrauterine hormonal
twins. International Journal of Eating Disorders, 44, 44–​49. environment and risk of developing anorexia nervosa.
Munn, M. A., Stallings, M. C., Rhee, S. H., Sobik, L. E., Corley, Archives of General Psychiatry, 64, 1402–​1407.
R. P., Rhea, S. A., & Hewitt, J. K. (2010). Bivariate analy- Racine, S. E., Burt, S. A., Iacono, W. G., McGue, M., & Klump,
sis of disordered eating characteristics in adolescence and K. L. (2011). Dietary restraint moderates genetic risk for
young adulthood. International Journal of Eating Disorders, binge eating. Journal of Abnormal Psychology, 120, 119–​128.
43, 751–​761. Raevuori, A., Kaprio, J., Susser, E., Sihvola, E., Hoek, H. W.,
Munn-​Chernoff, M. A., Duncan, A. E., Grant, J. D., Wade, T., Rissanen, A., & Keski-​Rahkonnen, A. (2008). Anorexia and
Agrawal, A., Bucholz, K. K.,  . . .  Heath, A. C. (2013). A bulimia nervosa in same-​sex and opposite-​sex twins: Lack of
twin study of alcohol dependence, binge eating, and com- association with twin type in a nationwide study of Finnish
pensatory behaviors. Journal of Studies on Alcohol and Drugs, twins. American Journal of Psychiatry, 165, 1604–​1610.
74, 664–​673. Reichborn-​Kjennerud, T., Bulik, C. M., Kendler, K. S., Roysamb,
Munn-​Chernoff, M. A., Grant, J. D., Agrawal, A., Sartor, C. E., E., Tambs, K., Torgersen, S., & Harris, J. R. (2004). Undue
Werner, K. B., Bucholz, K. K.,  . . .  Duncan, A. E. (2015). influence of weight on self-​evaluation:  A  population-​based
Genetic overlap between alcohol use disorder and bulimic twin study of gender differences. International Journal of
behaviors in European American and African American Eating Disorders, 35, 123–​132.
women. Drug and Alcohol Dependence, 153, 335–​340. Rice, J. P., Rochberg, N., Endicott, J., Lavori, P. W., & Miller,
Munn-​Chernoff, M. A., Keel, P. K., Klump, K. L., Grant, J. D., C. (1992). Stability of psychiatric diagnoses: An application
Bucholz, K. K., Madden, P. A. F.,  . . .  Duncan, A. E. (2015). of the affective disorders. Archives of General Psychiatry, 49,
Prevalence of and familial influences on purging disorder in 824–​830.
a community sample of female twins. International Journal of Rowe, R., Pickles, A., Simonoff, E., Foley, D., Rutter, M., &
Eating Disorders, 48, 601–​606. Silberg, J. (2002). Bulimic symptoms in the Virginia twin
Munn-​Chernoff, M. A., von Ranson, K. M., Culbert, K. M., study of adolescent behavioural development. Biological
Larson, C. L., Burt, S. A., Klump, K. L. (2013). An exami- Psychiatry, 51, 172–​182.
nation of the representativeness assumption for twin studies Rutherford, J., McGuffin, P., Katz, R. J., & Murray, R. M.
of eating pathology and internalizing symptoms. Behavior (1993). Genetic influences on eating attitudes in a normal
Genetics, 43, 427–​435. female twin population. Psychological Medicine, 23, 425–​436.
Mustelin, L., Joutsi, J., Latvala, A., Pietiläinen, K. H., Rissanen, Rutter, M., & Silberg, J. (2002). Gene-​environment interplay in
A., & Kaprio, J. (2012). Genetic influences on physical activ- relation to emotional and behavioural disturbance. Annual
ity in young adults: A twin study. Medical Science, Sports, and Review of Psychology, 53, 463–​490.
Exercise, 44, 1293–​1301. Scarr, S., & McCartney, K. (1983). How people make their own
Neale, B. M., Mazzeo, S. E., & Bulik, C. M. (x). A twin study of environments: A theory of genotype → environment effects.
dietary restraint, disinhibition and hunger: An examination Child Development, 54, 424–​435.
of the Eating Inventory (Three Factor Eating Questionnaire). Scherag, A., Dina, C., Hinney, A., Vatin, V., Scherag, S., Vogel,
Twin Research, 6, 471–​478. C. I.,  . . .  Meyre, D. (2010). Two new Loci for body-​weight

Wade, Bulik 103
 regulation identified in a joint analysis of genome-​wide asso- exposures and nutritional status. Physiology and Behavior,
ciation studies for early-​onset extreme obesity in French and 162, 181–​185.
German study groups. PLoS Genetics, 6, e1000916. Steiger, H., Thaler, L., Gauvin, L., Joober, R., Labbe, A., Israel,
Scherag, S., Hebebrand, J., & Hinney, A. (2010). Eating dis- M., & Kucer A. (2016). Epistatic interactions involving
orders:  The current status of molecular genetic research. DRD2, DRD4, and COMT polymorphisms and risk of
European Child and Adolescent Psychiatry, 19, 211–​226. substance abuse in women with binge-​purge eating distur-
Singh, A. L., & Waldman, I. D. (2010). The etiology of asso- bances. Journal of Psychiatric Research, 77, 8–​14.
ciations between negative emotionality and childhood Stein, A., Woolley, H., Cooper, S. D., & Fairburn, C. G. (1994).
externalizing disorders. Journal of Abnormal Psychology, 119, An observational study of mothers with eating disorders and
376–​388. their infants. Journal of Child Psychology and Psychiatry and
Slane, J. D., Burt, S. A., & Klump, K. L. (2011). Genetic and Allied Disciplines, 35, 733–​748.
environmental influences on disordered eating and depres- Stein, A., Woolley, H., Murray, L., Cooper, P., Cooper, S.,
sive symptoms. International Journal of Eating Disorders, 44, Noble, F., Affonso, N., & Fairburn, C. G. (2001). Influence
605–​611. of psychiatric disorder on the controlling behaviour of moth-
Slane, J. D., Burt, S. A., & Klump, K. L. (2012). Bulimic behav- ers with 1-​year old infants: A study of women with maternal
iors and alcohol use:  Shared genetic influences. Behavior eating disorder, postnatal depression, and a healthy compari-
Genetics, 42, 603–​613. son group. British Journal of Psychiatry, 179, 157–​162.
Slof-​Op’t Landt, M. C.  T., Bartels, M., van Furth, E. F., van Stice, E., Spangler, D., & Agras, W. S. (2001). Exposure to
Beijsterveldt, C. E. M., Meulenbelt, I., Slagboom, P. E., & media-​portrayed thin-​ideal images adversely affects vulner-
Boomsma, D. I. (2008). Genetic influences on disordered able girls:  A  longitudinal experiment. Journal of Social and
eating are largely independent of body mass index. Acta Clinical Psychology, 20, 270–​288.
Psychiatrica Scandinavica, 117, 348–​356. Strober, M. (1991). Disorders of the self in anorexia nervosa: An
Slof-​Op ‘t Landt, M. C.  T., van Furth, E. F., Meulenbelt, I., organismic-​developmental paradigm. In C. Johnson (Ed.),
Slagboom, P. E., Bartels, M., Boomsma, D. I., & Bulik, C. Psychoanalytic theory and treatment for eating disorders.
M. (2005). Eating disorders: From twin studies to candidate New York, NY: Guilford Press.
genes and beyond. Twin Research and Human Genetics, 8, Suisman, J. L., Burt, S. A., McGue, M., Iacono, W. G., &
467–​482. Klump, K. L. (2011). Parental divorce and disordered eat-
Spanos, A., Burt, S. A., & Klump, K. L. (2010). Do weight and ing:  An investigation of a gene-​ environment interaction.
shape concerns exhibit genetic effects? Investigating discrepant International Journal of Eating Disorders, 44, 169–​177.
findings. International Journal of Eating Disorders, 43, 29–​34. Suisman, J. L., Thompson, J. K., Keel, P. K., Burt, S. A., Neale,
Spelt J., & Meyer, J. M. (1995). Genetics and eating disorders. In M., Boker, S., & Klump, K. L. (2014). Genetic and envi-
J. R. Turner, L. R. Cardon, & J. K. Hewitt (Eds.), Behaviour ronmental influences on thin-​ ideal internalization across
genetic approaches in behavioural medicine (pp. 167–​185). puberty and preadolescent, adolescent, and young adult
New York, NY: Plenum Press. development. International Journal of Eating Disorders, 47,
Steiger, H., Gauvin, L., Joober, R., Israel, M., Badawi, G., 773–​783.
Groleau, P.,  . . .  Ouelette, A. S. (2012). Interaction of the Sulek, S., Lacinova, Z., Dolinkova, M., & Haluzik, M. (2007).
BcII glucocorticoid receptor polymorphism and childhood Genetic polymorphisms as a risk factor for anorexia nervosa.
abuse in Bulimia Nervosa (BN): Relationship to BN and to Prague Medical Report, 108, 215–​225.
associated trait manifestations. Journal of Psychiatric Research, Sullivan, P. F., Bulik, C. M., & Kendler, K. S. (1998). Genetic
46, 152–​158. epidemiology of binging and vomiting. British Journal of
Steiger, H., Joober, R., Gauvin, L., Bruce, K. R., Richardson, J., Psychiatry, 173, 75–​79.
Israel, M.,  . . .  Groleau, P. (2008). Serotonin-​system poly- Sung, J., Lee, K., Song, Y.-​M., Lee, M. K., & Lee, D.-​H. (2010).
morphisms (5-​HTTLPR and -​1438G/​A) and responses of Heritability of eating behavior assessed using the DEBQ
patients with bulimic syndromes to multimodal treatments. (Dutch Eating Behavior Questionnaire) and weight-​related
Journal of Clinical Psychiatry, 69, 1565–​1571. traits: The Healthy Twin study. Obesity, 18, 1000–​1005.
Steiger, H., Joober, R., Israel, M., Young, S. N., Ng, M. K., Swanson, S. A., Brown, T. A., Crosby, R., & Keel, P. K. (2014).
Ng, Y. K., Gauvin, L.,  . . .  Torkaman-​Zehi, A. (2005). The What are we missing? The costs versus benefits of skip
5HTTLPR polymorphism, psychopathologic symptoms, rule designs. International Journal of Methods in Psychiatric
and platelet [3H-​] paroxetine binding in bulimic syndromes. Research, 23, 474–​485.
International Journal of Eating Disorders, 37, 57–​60. Thaler, L., Gauvin, L., Joober, R., Groleau, P., de Guzman,
Steiger, H., Labonte, B., Groleau, P., Turecki, G., & Israel, M. R., Ambalavanan, A.,   .  .  .  Steiger, H. (2014).
(2013). Methylation of the glucocorticoid receptor gene Methylation of BDNF in women with bulimic eating syn-
promoter in bulimic women:  Associations with borderline dromes:  Associations with childhood abuse and borderline
personality disorder, suicidality, and exposure to childhood personality disorder. Progress in Neuro-​ Psychopharmacology
abuse. International Journal of Eating Disorders, 46, 246–​255. and Biological Psychiatry, 54, 43–​49.
Steiger, H., Richardson, J., Joober, R., Gauvin, L., Israel, M.,  . . .  Thornton, L. M., Welch, E., Munn-​ Chernoff, M. A.,
Young, S. N. (2007). The 5HTTLPR polymorphism, prior Lichtenstein, P., & Bulik, C. M. (2016). Anorexia nervosa,
maltreatment and dramatic-​ erratic personality manifesta- major depression, and suicide attempts: Shared genetic fac-
tions in women with bulimic syndromes. Journal of Psychiatry tors. Suicide and Life-​Threatening Behavior. doi:  10.1111/​
and Neuroscience, 32, 354–​362. sltb.12235 [Epub ahead of print].
Steiger, H., & Thaler, L. (2016). Eating disorders, gene-​ Turkheimer, E., D’Onofrio, B. M., Maes, H. H., & Eaves, L. J.
environment interactions and the epigenome: Roles of stress (2005). Analysis and interpretation of twin studies including

104 Genetic Influences

 measures of the shared environment. Child Development, 76, Wade, T. D., Zhu, G., & Martin, N. G. (2011). Undue influence
1217–​1233. of weight and shape: Is it distinct from body dissatisfaction
Turkheimer, E., & Waldron, M. (2000). Nonshared environ- and concern about weight and shape? Psychological Medicine,
ment:  A  theoretical, methodological, and quantitative 41, 819–​828.
review. Psychological Bulletin, 126, 78–​108. Walters, E. E., & Kendler, K. S. (1995). Anorexia nervosa and
Wade, T. D., & Bulik, C. M. (2007). Shared genetic and envi- anorexic-​like syndromes in a population-​based female twin
ronmental risk factors between undue influence of body sample. American Journal of Psychiatry, 152, 64–​71.
shape and weight on self-​evaluation and dimensions of per- Walters, E. E., Neale, M. C., Eaves, L. J., Heath, A. C., Kessler,
fectionism. Psychological Medicine, 37, 635–​644. R. C., & Kendler, K. S. (1992). Bulimia nervosa and major
Wade, T. D., Bulik, C. M., Neale, M. C., & Kendler, K. S. depression: A study of common genetic and environmental
(2000). Anorexia nervosa and major depression: An exami- factors. Psychological Medicine, 22, 617–​622.
nation of shared genetic and environmental risk factors. Wang, K., Zhang, H., Bloss, C. S., Duvvuri, V., Kaye, W.,
American Journal of Psychiatry, 157, 469–​471. Schork, N. J.,  . . .  Price Foundation Collaborative Group.
Wade, T. D., Fairweather-​Schmidt, A. K., Zhu, G., & Martin, (2011). A genome-​wide association study on common SNPs
N. G. (2015). Does shared genetic risk contribute to the co-​ and rare CNVs in anorexia nervosa. Molecular Psychiatry, 16,
occurrence of eating disorders and suicidality? International 949–​959.
Journal of Eating Disorders, 48, 684–​691. Waugh, E., & Bulik, C. M. (1999). Offspring of women with
Wade, T. D., Gillespie, N., & Martin, N. G. (2007). A comparison eating disorders. International Journal of Eating Disorders, 25,
of early family life events amongst monozygotic twin women 123–​133.
with lifetime anorexia nervosa, bulimia nervosa or major depres- Whelan, E., & Cooper, P. J. (2000). The association between
sion. International Journal of Eating Disorders, 40, 679–​686. childhood feeding problems and maternal eating disor-
Wade, T. D., Gordon, S., Medland, S., Bulik, C. M., Heath, A. der: A community study. Psychological Medicine, 30, 69–​77.
C., Montgomery, G. W., & Martin, N. G. (2013). Genetic Wilksch S. M., & Wade T. D. (2009). An investigation of
variants associated with disordered eating. International temperament endophenotype candidates for early emer-
Journal of Eating Disorders, 46, 594–​608. gence of the core cognitive component of eating disorders.
Wade, T. D., Hansell, N. K., Crosby, R. D., Bryant-​Waugh, R., Psychological Medicine, 39, 811–​821.
Treasure, J., Nixon, R., Byrne, S., & Martin, N. G. (2013). Willer, C. J., Speliotes, E. K., Loos, R. J., Li, S., Lindgren,
A study of changes in genetic and environmental influences C. M., Heid, I. M.,   .  .  .  Hirschhorn, J. N. (2009). Six
on weight and shape concern across adolescence. Journal of new loci associated with body mass index highlight a neu-
Abnormal Psychology, 122, 119–​130. ronal influence on body weight regulation. Nature Genetics,
Wade, T. D., Martin, N. G., Neale, M. C., Tiggemann, M., 41, 25–​34.
Treloar, S. A., Bucholz, K. K., Madden, P. A. F., & Heath, A. Wisotsky, W., Dancyger, I., Fornari, V., Katz, J., Wisotsky, W. L.,
C. (1999). The structure of genetic and environmental risk & Swencionis, C. (2003). The relationship between eating
factors for three measures of disordered eating. Psychological pathology and perceived family functioning in eating disor-
Medicine, 29, 925–​934. der patients in a day treatment program. Eating Disorders,
Wade, T. D., Martin, N. G., & Tiggemann, M. (1998). Genetic 11, 89–​99.
and environmental risk factors for the weight and shape con- Wonderlich, S. A. (1992). Relationship of family and personality
cern characteristic of bulimia nervosa. Psychological Medicine, factors in bulimia nervosa. In J. Crowther, D. Tennenbaum,
28, 761–​771. S. Hobfoll, & M. Stephens (Eds.), The etiology of bulimia
Wade, T. D., Martin, N. G., Tiggemann, M., Abraham, S., nervosa:  The individual and familial context. Washington,
Treloar, S. A., & Heath, A. C. (2000). Genetic and envi- DC: Hemisphere.
ronmental risk factors shared between disordered eating, Woodside, D. B., Shekter-​Wolson, L., Garfinkel, P. E., Olmsted,
psychological and family variables. Personality and Individual M. P., Kaplan, A. S., & Maddocks, S. E. (1995). Family
Differences, 28, 729–​740. interactions in bulimia nervosa I:  Study design, compari-
Wade, T. D., Treloar, S. A., & Martin, N. G. (2008). Shared sons to established population norms, and changes over
and unique risk factors between lifetime purging and objec- the course of an intensive day hospital treatment program.
tive binge eating:  A  twin study. Psychological Medicine, 38, International Journal of Eating Disorders, 17, 105–​115.
1455–​1464. Wurtman, J. J. (1993). Depression and weight gain: The seroto-
Wade, T. D., Wilkinson, J., & Ben-​Tovim, D. (2003). The genetic nin connection. Journal of Affective Disorders, 29, 183–​192.
epidemiology of body attitudes, the attitudinal compo- Yilmaz, Z., Hardaway, A., & Bulik, C. (2015). Genetics and
nent of body image in women. Psychological Medicine, 33, epigenetics of eating disorders. Advances in Genomics and
1395–​1405. Genetics, 5, 131–​135.

Wade, Bulik 105
 C H A PT E R


Psychosocial Risk Factors
6 for Eating Disorders

Corinna Jacobi, Kristian Hütter, and Eike Fittig

Abstract
This chapter provides an updated overview of risk factors for eating disorders, on the basis of the risk
factor taxonomy described by (Kraemer et al., 1997). It summarizes risk factors identified in longitudinal
studies and markers and retrospective correlates from cross-​sectional studies through April 2002 for
anorexia nervosa, bulimia nervosa, and binge eating disorder, identifies new studies published between
May 2002 and June 2015, and integrates them into the earlier review. The updated review confirms
that longitudinal evidence on risk factors is strongest for nonspecific eating disorder diagnoses including
subclinical forms and weakest for participants with diagnoses of anorexia nervosa. When strict criteria
for caseness are applied, the majority of risk factors were not able to predict distinct diagnoses and
only very few risk factors were confirmed in more than one sample. Case prediction, specificity, and
replication therefore remain the biggest challenges in risk factor research for eating disorders.
Key Words:  anorexia nervosa, binge eating disorder, bulimia nervosa, Kraemer taxonomy, longitudinal
study, risk factor

Introduction methods and definitions of the proposed theoret-

The identification of modifiable risk factors ical framework were applied to the field of EDs.
not only improves our knowledge of the etiol- Because the results have been presented elsewhere
ogy of a disorder but also has important public in detail (Jacobi, 2005; Jacobi, Hayward, de
health implications, especially for chronic disor- Zwaan, Kraemer, & Agras, 2004), we only briefly
ders with poor prognosis. Owing to the incon- summarize core aspects here.
sistent use of the terms “risk” and “risk factor” In the approach of Kraemer and colleagues, prec-
and to improve communications among scien- edence represents a crucial criterion for the defini-
tists, clinicians, and politicians in different fields, tion of risk factors. Accordingly, the majority of risk
Helena Kraemer and coworkers—​as part of the factors can be assessed only in longitudinal studies.
MacArthur Foundation Research Network on Exceptions are so-​called fixed markers, that is, invar-
Psychopathology and Development—​ 20  years iable risk factors documented before the onset of
ago proposed exact definitions and methods for the ED in medical records or birth registers. These
risk and etiology factors (Kraemer et  al., 1997). factors are usually derived from cross-​ sectional
For the development of eating disorders (EDs)—​ (case-​control, family history, twin, epidemiological
comparable with other mental disorders—​a wide studies, or register studies). Variables not fulfilling
range of variables has been claimed to be risk precedence of the factor to the onset of the disor-
factors. Many of them were assessed in cross-​ der are correlates. The status of variable markers and
sectional studies that do not allow for “true” risk causal risk factors can be established only in ran-
factor identification. To separate “true” risk fac- domized clinical trials (prevention or intervention
tors for EDs from correlates of the disorders, the studies) that confirm that the modification of the

106
factors leads to a change in the risk of the outcome subclinical syndromes or eating disorders not oth-
(e.g., onset of the disorder). erwise specified (EDNOS). In six studies, the out-
In our recent meta-​analysis we included a sep- come was defined as scoring above a cutoff on an
arate category of factors in addition to those pro- ED scale, mostly the EAT-​26. A closer examination
posed in the Kraemer et al. typology: cross-​sectional of the outcomes reveals that the focus is on bulimic
studies with retrospective (risk) factor assessment and binge eating syndromes, whereas reports of ano-
before the onset of the ED according to the subjects’ rexic syndromes as outcomes were very rare. Taken
self-​reports. Retrospective risk factor assessment is together, of the 12,776 subjects in the 15 studies,
problematic because of retrospective recall or mem- 26 cases of AN, 88 cases of BN, 78 EDNOS cases,
ory bias, especially in subjects affected with the dis- and 238 partial/​subclinical or “binge eating” cases
order. However, because longitudinal studies are emerged during the follow-​up periods.
difficult and expensive to conduct and thus rare, we Samples in the studies consisted mostly of
decided to consider and include these so-​called ret- adolescents between 12 and 15  years; three stud-
rospective correlates for exploratory or hypothesis-​ ies assessed infants or younger children (Johnson,
generating reasons in the original review. Cohen, Kasen, & Brook, 2002; Kotler, Cohen,
Davies, Pine, & Walsh, 2001; Marchi & Cohen,
Method 1990), and two studies focused on young adults
The proposed taxonomy of risk factors was first (Ghaderi, 2001; Vollrath, Koch, & Angst, 1992). In
applied to the field of EDs as part of a comprehensive eight studies the samples included females only; in
meta-​analytic review (Jacobi et al., 2004). A detailed seven studies the samples consisted of both males
computerized and manual literature search of poten- and females. The duration of follow-​ups varied from
tial risk factor studies published through April 2002 1 to 18 years. The number and breadth of included
was conducted resulting in the screening of approx- risk factors, and the definitions and assessment of
imately 5,000 abstracts and inclusion of about 300 risk status or caseness (symptomatic/​asymptomatic;
(cross-​sectional and longitudinal) studies. Among cases/​noncases; high/​moderate/​low risk) varied sig-
others, the following criteria were established for nificantly: seven studies assessed ED symptoms and
study inclusion in the original review: syndromes with structured diagnostic interviews,
and eight studies defined caseness on the basis of
1.  With few exceptions (regarding epidemiolog- questionnaire cutoffs or score combinations.
ical studies) only studies with a (healthy or unaf-
fected) control group were included. Anorexia Nervosa
2.  The sample size required in the studies was at To summarize the results of our previous analy-
least 10 cases per cell. sis, the following risk factors were found for AN
3.  For longitudinal studies, a follow-​up interval (Jacobi, 2005): Twin studies suggest a genetic influ-
of at least 1 year was needed to allow enough time ence for anorexia present before birth. Additional
for symptoms of eating disturbances or disorders to fixed markers are female gender, ethnicity, season of
change or emerge. birth (between April and June), birth-​related peri-
4.  The focus was placed on risk factors for ED natal complications (cephalhematoma), premature
syndromes or ED “cases”; accordingly, longitudinal delivery (based on medical records), and obstetric
studies addressing solely (dimensional) disturbances complications. Risk factors in early and later child-
or symptoms as outcomes assessed via question- hood include maternally reported health problems
naires (e.g., EDI) were excluded. and a number of factors around childhood eating
such as picky eating, anorexic symptoms, digestive
Summary of Risk Factors for Eating and other eating-​related problems (e.g., eating con-
Disorders Through 2002: Characteristics flicts, struggles around meals, unpleasant meals).
of Longitudinal Studies Surprisingly few studies identified risk factors for
In the original review, 15 longitudinal studies AN during adolescence. Apart from adolescent age
comprising 12,776 subjects were found fulfilling and early pubertal timing (both of which are relevant
our inclusion criteria (see Table 6.1 for references). for BN as well), only a factor comprising weight and
The majority of these studies identified risk fac- shape concerns and dieting could be confirmed as a
tors for a mixture of full Diagnostic and Statistical risk factor for (binge eating subtype) AN.
Manual (DSM) syndromes of anorexia nervosa In addition to these longitudinally assessed
(AN) and bulimia nervosa (BN) and/​or partial or risk factors, pregnancy complications and shorter

Jacobi, Hüt ter, Fit tig 107

Table 6.1  Characteristics of Longitudinal Studies on Risk Factors for Eating Disorders (through 2002)
Study Sample (N ) Age at Baseline Follow-​up Cases (n) Outcome Predictors
(years) Interval (years)

1. Attie & 193 female adolescents 13.9 2 NR Emergence of eating problems, EAT-​ Height; Weight; Body fat;
Brooks-​Gunn and their mothers; initial 26 scores Pubertal timing; Body image;
(1989) symptoms controlled for Personality: Psychopathology,
Emotional Tone, Impulse Control
(SIQYA); Family relationships, EAT

2. Marchi & 659 children and mothers; Interviewed at 10 AN = 5, Problematic eating behaviors Eating behaviors (struggle over eating,
Cohen (1990) 326 girls and 333 boys three ages: 1–​10, BN = 9, and disorders (modified amount eaten, picky eater, speed of
9–​18, and 11–​21 Severe Diagnostic Interview Schedule for eating, interest in food, pica, digestive
AN and BN Children: DISC) problems)
behaviors = 65

3. Patton et al. 734 adolescent girls; initial 15 1 N = 16; BN = 4, Eating disorder (risk group selected EAT; GHQ; Putative risk
(1990) symptoms controlled for by 12 = partial EAT-​26 cutoff: 20/​21), General factors: family background, reported
subgrouping syndromes Health Questionnaire (GHQ) ≥ 5/​6; family weight and eating pattern,
subsequent semistructured interview personal background, weight history,
for assessment of clinical status/​ perceived current social stress,
caseness (categories: cases, dieters, personality (Cattell personality
nondieters) questionnaire); attitudes toward
eating, weight and shape, menstrual
history, current weight and shape
(clinical interview)

4. Vollrath 292 males and 299 females 27–​28 2 (out of 9-​ BN = 4, binge Eating problems (binge eating, Same as dependent variables
et al. (1992) of representative young adult year follow-​up) eating = 28 vomiting, dieting, weight, etc.) as part
sample (N = 4,567), divided of the Structured Psychological
into high-​and low-​risk Interview and Rating of the Social
groups according to SCL-​90 Consequences for Epidemiology
scores; controlling for initial (SPIKE); Diagnosis: computerized
symptoms not reported algorithms
5. Killen et al. 967 adolescent girls, 12.4 3 BN = 13, binge Eating disorder symptoms and Weight concerns; EDI; Dietary
(1994) community sample; subjects eating = 19 partial syndromes, classification as restraint; Pubertal development;
with initial symptoms symptomatic based on binge eating, Height; Weight; BMI, Behavior
excluded from analysis compensatory behaviors, overconcern Problem Scales (Youth Self-​Report
with weight and body shape, loss of Inventory)
control

6. Graber 116 female adolescents, 14.31 8 n = 15 Subclinical or clinical eating problems EAT; Physical development (body fat,
et al. (1994) recruited from private (cutoff: EAT 26 ≥ 20) age at menarche); Body image, self-​
schools); grouped into four image, psychopathology, emotional
groups (low to chronic risk) tone, family relationships (SIQYA);
EDI: Perfectionism, Ineffectiveness;
Affective states (Youth Self-​Report);
Family organization and functioning
(FES)

7. Leon et al. Community sample (852 Grades 7–​10 3 NR Disordered eating (risk status: 18 items Demographic variables; Eating
(1995) girls, 815 boys); initial from ED checklist, BMI <17 or >30, Disorders Checklist; EDI;
symptoms controlled for abnormal range scores on EDI Drive Health Behavior Survey; Pubertal
for Thinness and Bulimia scales) Development Scale; Personality
(MMPI): Negative and Positive
Emotionality, Constraint; General
Behavior Inventory; Autonomy;
Attitudes about sexuality

8. Killen et al. 877 high-​school girls, 14.9 4 Partial Eating disorder symptoms and Weight concerns; EDI; Dietary
(1996) (community sample); initial syndromes = 36 partial syndromes (EDE—​interview restraint; Height; Weight; BMI;
symptoms controlled for adaptation) Temperament (emotionality,
activity, sociability); Drinking
frequency; (baseline and follow-​up);
EDE-​adaptation

9. Button 594 girls; initial symptoms 11–​12 4 EAT + =47 EAT 26 score; EAT score ≥ 20; Self-​esteem; Perceived health status,
et al. (1996) not controlled for “EATpath”; Eating behavior and fatness concern, family relationships,
(FU: N = 394) concerns, weight variables; Self-​esteem; school problems, general worrying/​
HAMD anxiety and depression scale nervousness (five questions)

(continued)
Table 6.1 Continued
Study Sample (N ) Age at Baseline Follow-​up Cases (n) Outcome Predictors
(years) Interval (years)

10. Calam & 92 girls and their mothers 12.8 7 NR EAT-​26-​, BITE-​scores EAT-​26; BITE; Setting Conditions
Waller (1998) (part of study of family for Anorexia (SCANS)-​questionnaire
management of children’s (including 14 self-​esteem and 8
eating habits); initial perfectionism items); BMI; Family
symptoms not controlled for Assessment Device (FAD)

11. Leon et al. 726 girls, 698 boys, high-​risk Grades 7–​10 to 3–​4 BN = 6, Eating Disorder Risk Factor Index, Personality and temperament
(1999) subjects y1 excluded from 9–​12 AN = 1, Kiddie-​SADS: Expanded eating (MPQ;GBI), Eating Disorders
prospective analysis EDNOS = 14 disorders module Checklist, EDI; Health behaviors;
Physical and pubertal development;
Psychopathology (Kiddie-​SADS)

12. Patton 1947 male and female 14–​15 3 BN = 2, Branched Eating Disorders Test to assess Dieting (adolescent dieting scale);
et al. (1999) students, recruited in Partial symptoms and partial syndromes Exercise; Weight and height;
cohorts; initial symptoms BN = 27, Psychiatric morbidity (computerized
controlled for Partial AN = 4 interview)

13. Ghaderi & 1,157 females, random 18–​30 2 AN = 1, Survey for Eating Disorders (SEDs), Self-​Concept Questionnaire (SCQ),
Scott (2001) general population sample; BN = 11, questionnaire to assess DSM-​IV eating Body Shape Questionnaire (BSQ),
initial symptoms controlled BED = 10, disorders Perceived social support from
for by subgrouping EDNOS = 4 family (PSSFa), Ways of Coping
Questionnaire (WCQ)

14. Kotler 976 (t1)–​776 (t4) children, 6.1 (range 17 AN = 18, Eating behaviors based on maternal Early childhood eating problems
et al. (2001) 45%–​50% female 1–​10.9) BN = 29 interviews (t1–​t3); Parent (t2, t3) and (unpleasant meals, struggles over
youth (t2–​t4) versions of the Diagnostic eating, amount eaten, picky or
Interview Schedule for Children (DISC) choosy eating, interest in food)
assessed by maternal interview

15. Johnson 782 children, 397 males, 385 6 16–​18 AN = 1, Parent (t2, t3) and youth (t2–​t4) versions Childhood temperament, parental
et al. (2002) females; initial symptoms BN = 10, of the Diagnostic Interview Schedule for psychiatric problems, maladaptive
controlled for EDNOS = 41, Children (DISC) parental behavior, childhood
(atyp. AN = 9, maltreatment, other childhood
BED = 9, adversities (all assessed by the
others = 23) Disorganizing Poverty Interview)
gestational age were confirmed as retrospective cor- risk factors summarized in the bulimia section are
relates, as were feeding and gastrointestinal prob- equally relevant for binge eating disorder (BED).
lems, infant sleep difficulties, and a high-​concern Only two risk factor studies explicitly included the
parenting style during early childhood. Further, proposed criteria for BED (Ghaderi & Scott, 2001;
obsessive-​compulsive personality disorders, anxiety Johnson et  al., 2002). On the basis of these, low
disorders, and higher levels of feelings and experi- self-​esteem, high body concern, high use of escape-​
ences of loneliness, shyness, and inferiority in child- avoidance coping, low perceived social support, and
hood and adolescence were also identified. childhood experiences of sexual abuse and physical
During adolescence, retrospective correlates are neglect are risk factors for BED. Additional proba-
a high level of exercise, dieting behavior (especially ble risk factors (retrospective correlates) are negative
for the binge-​type anorexics), the presence of body self-​evaluation, major depression, marked conduct
dysmorphic disorder, increased exposure to sexual problems, deliberate self-​ harm, greater levels of
abuse and other adverse life events, a higher level of exposure to parental criticism, high expectations,
perfectionism, negative self-​evaluation, premorbid minimal affection, parental underinvolvement,
obsessive-​compulsive disorder, and acculturation. low maternal care and high overprotection, and
greater parental neglect and rejection. In addition,
Bulimia Nervosa BED women reported higher rates of sexual abuse;
In addition to genetic factors, female gender, repeated severe physical abuse; bullying; critical
ethnicity, obstetric complications, and early child- comments by family about shape, weight, or eating;
hood health problems, confirmed for both AN and and teasing about shape, weight, eating, or appear-
BN, the risk factors for bulimia can be summa- ance. In comparison to psychiatric controls the fol-
rized as follows (Jacobi, 2005):  higher body mass lowing factors turned out to be specific retrospective
index (BMI), experiences of sexual abuse or physical correlates: low parental contact; critical comments
neglect during childhood, higher levels of psychiatric about shape, weight, or eating; and childhood obe-
morbidity or negative affectivity, negative percep- sity (Dominy, Johnson, & Koch, 2000; Fairburn,
tion of parental attitudes, low interoceptive aware- Cooper, Doll, & Welch, 1999; Striegel-​ Moore,
ness, amount of alcohol consumption over the last Dohm, Pike, Wilfley, & Fairburn, 2002).
30 days, temperament-​related factors represented by
elevations on two subscales of the YSR-​Inventory Limitations of Risk Factor Research
(Unpopular, Aggressive), adolescent age and early Although the improved knowledge of risk fac-
pubertal timing, low self-​ esteem (age 13–​15), as tors now allows for a clearer distinction between
well as increased weight and shape concerns. Finally, mostly cross-​sectionally assessed correlates of EDs
perceived low social support from the family and an and “true,” longitudinally assessed risk factors, the
escape-​avoidant style of coping with stressful events application of the rigorous risk factor methodol-
of everyday life were found in late adolescence. ogy also made a number of limitations of previous
Additional retrospective correlates for BN are research evident:  Although the majority of lon-
pregnancy complications, childhood obesity, child- gitudinal studies included sample sizes of several
hood overanxious disorder, sexual abuse and adverse hundred subjects, the samples are mostly too small
life events, dieting, acculturation, and social phobia. for consistent and meaningful risk factor detection
As was the case for AN, some parental problems of full syndromes of EDs, especially AN. Studies
(alcoholism, depression, drug abuse, obesity), a including subjects already at higher risk at study
number of family environmental factors (e.g., criti- start may therefore be more promising.
cal comments on weight and shape, low contact), Longitudinal evidence on risk factors is much
other adverse family experiences, and negative self-​ stronger for BN and binge-​ related syndromes,
evaluation also represent retrospective correlates. whereas our knowledge on risk factors for AN is still
Lastly, mood-​and anxiety-​related prodromal symp- very limited. Because of the overlap of the different
toms including severe dieting were found during full and partial syndromes in longitudinal studies,
late adolescence. current research does not permit a valid differen-
tiation of risk factors for BN vs. BED vs. partial
Binge Eating Disorder syndromes. In addition, most of previous risk fac-
Because the outcome of the longitudinal stud- tor studies did not consider interactions between
ies is often a mixture of bulimic or binge eating risk factors, necessary to improve the understanding
syndromes, it can be assumed that some of the of the etiology of the disorder and to improve the

Jacobi, Hüt ter, Fit tig 111

development and effectiveness of preventive inter- The literature search resulted in over 9,000 stud-
ventions. Finally, the specificity of factors has been ies to be screened for inclusion. In addition to the
addressed for only a few retrospective correlates of aforementioned longitudinal studies, 18 new longi-
AN, BN, and BED, but not for risk factors from tudinal studies meeting our inclusion criteria were
longitudinal studies, as no other outcomes than identified. These studies comprise 12 independent
EDs were addressed. samples with N = 12,663 participants (Table 6.2).
Overall, n  =  615 ED cases (including 27 cases of
Risk Factor Update (2002–​2015) subclinical ED) emerged during the follow-​up peri-
In the following paragraphs, we present the ods from independent samples of the longitudinal
methods and results of our risk factor update studies, but exact numbers of ED cases reported in
including longitudinal and register studies pub- same-​sample studies sometimes differ slightly. Of
lished through 2015. these, n = 21 cases were full-​syndrome and subclini-
cal AN, n = 107 cases were full-​syndrome and sub-
Method and Study Characteristics clinical BN, N = 63 cases were full-​syndrome and
For our risk factor update, the criteria for study subclinical BED, n  =  22 cases fulfilled criteria for
search and inclusion as used in the original meta-​ DSM-​5 purging disorder and n = 196 were EDNOS
analysis were only slightly adapted for studies pub- cases. For the remaining cases, no specific diagnostic
lished between May 2002 and June 2015. Even category was reported. The majority of the samples
stricter than in the previous review, we placed the were adolescents and young adults between 12 and
focus of new studies on the identification of risk 30  years. Two samples included children (Hilbert
factors for ED syndromes (“cases”) and excluded et al., 2014; Moorhead et al., 2003). Only one sam-
all studies using dimensional outcomes solely. ple included males (Beato-​Fernandez, Rodriguez-​
However, newer studies were included irrespective Cano, Belmonte-​Llario, & Martinez-​Delgado,
of the number of cases found. Studies reporting on 2004). The duration of follow-​ups varied from 1.5
retrospective correlates were no longer included in to 22 years.
the update. To address the specificity of risk factors Unfortunately, despite the relatively large num-
for ED diagnoses (e.g., AN, BN, BED, and their ber of new studies, many of the previously men-
subclinical forms) and to address replication of risk tioned problems remain:  Results (and effect sizes)
factors across all studies we applied the following are often not reported separately for the different
steps:  First, we collapsed studies of the previous ED categories but only for collapsed ED categories.
review and the current update which included at Even when reported for specific diagnoses, the num-
least 75% of cases of one specific ED category, and bers often do not add up to the total number of ED
examined which risk factors were predictive in more cases, probably due to some remaining subclinical
than one sample for each ED category, that is, could forms not reported. While some studies explicitly
be shown to be replicated using these relatively strict include “subclinical” forms of a specific ED diagno-
outcome criteria. Because most studies—​if report- ses, others summarize those as EDNOS categories.
ing diagnoses at all—​included a mixture of different The EDNOS and subclinical forms of ED remain
ED diagnoses, we felt that a criterion of at least 75% the largest categories; subclinical and clinical forms
of cases of one category would represent a minimal of AN remain the smallest category.
criterion to identify specific predictors. In previous
reviews (Jacobi, 2005; Jacobi & Fittig, 2010), this Risk Factors and Markers for Anorexia
criterion had not been applied and risk factors had Nervosa from Longitudinal Studies
been assumed to be predictive for a specific diagnos- None of the previous and new longitudinal risk
tic category if at least one new case of the respective factor studies fulfilled the criterion of detecting at
category had emerged during follow-​ up. If pos- least 75% of subclinical and full-​ syndrome AN
sible, we also determined the potency of replicated cases in more than one sample. Therefore, despite
risk factors on the basis of the area under the curve very few AN-​specific results from register studies
(AUC) statistic (Jacobi et al., 2004). Second, across (see below), identifying risk factors specific for AN
studies in the previous review and the update we remains a challenge. In samples with mixed out-
also examined, which factors had been replicated comes including lower rates of AN cases, four risk
two or more times in samples including lower rates factors could be replicated in at least two independ-
of specific categories (i.e., more mixed samples, ent samples: Body dissatisfaction, dieting, negative
<75% of one category). affect/​depressive symptoms, and childhood eating

112 Psychosocial Risk Factors

Table 6.2  Risk Factor Update: Characteristics of Longitudinal Studies Published Between 2002 and 2015
Study Sample (N) Age at baseline Follow-​up Cases Outcome (dependent Predictors (independent Results
(years) interval (y) variables) variables)

1. Cervera N = 2,509 girls (out 12–​21 18 months n = 90 (EDNOS) EAT >30, Eysenck Personality Higher levels of neuroticism
et al., 2003 of 2,743) semistructured Inventory, AFA increased risk of ED, high
interview (self-​esteem) levels of self-​esteem were
(psychiatrist) protective

2. Moorhead Original sample 5–​26 22 years n = 21; lifetime AN Diagnostic Interview Health-​related childhood Mothers of women with ED
et al., 2003 N = 763 youth, or BN: n = 9, partial Schedule IV (DIS-​ predictors (emotional-​ reported more pregnancy
subsample: n = 30 syndrome ED: n = 12 IV); subset contacted behavioral predictors, complications, more health
girls EDI DT > 12 at age 27 self-​concept, family problems of daughters before
vs. n = 38 girls <10 history of ED, pregnancy age 5, more anxiety-​depression
complications at age 9

3. The N = 1,103 girls Grades 6–​9 3 years n = 32, BN: McKnight Eating McKnight Risk Factor Arizona: Hispanic ethnicity,
McKnight n = 1; partial BN: Disorder Examination Survey variables thin body preoccupation and
Investigators, n = 26; BED: n = 5 Interview; diagnoses social pressure, increase in life-​
2003 events; California: Thin body
preoccupation, social pressure
to be thin

4. Beato-​ N = 1,076 subjects 12–​13 2 years n = 40; AN: n = 1; EAT-​40 > 30; BITE+; EAT-​40; BITE; BSQ; Pathological body
Fernandez (500 males, 576 BN: n = 8; BSQ; at 2 years GHQ; RSE; Family dissatisfaction, negative
et al. 2004 females) EDNOS: n = 31 SCAN interview APGAR perception of parental attitudes
(3 males) additionally (feeling of being ignored/​not
loved enough by mother)

5. Fairburn N = 2,992 female 19.8 years 2 years n = 104; AN: n = 10; EDE interview EDE-​Q Low BMI (<19); objective
et al. 2005 dieters BN: n = 19; bulimia, purging; secret eating,
EDNOS: n = 75 fear of losing control, desire to
have empty stomach

6. Stice et al., N = 496 (100% 11–​18 1 n = 23 threshold /​ EDDI, semistructured Fasting, dietary restraint Continuous and dichotomous
2008 female) subthreshold BN; interview, adapted fasting, and continuous dietary
n = 38 binge eating from the EDE restraint predict bulimic
pathology onset

(continued)
Table 6.2 Continued
Study Sample (N) Age at baseline Follow-​up Cases Outcome (dependent Predictors (independent Results
(years) interval (y) variables) variables)

7. Allen et al., N = 1,597 (55% antenatal—​ at age 14 n = 95; AN: n = 1, 24 items adapted Five domains: Antenatal; Female sex, being perceived
2009 male) 10 years assessment BN: n = 9, BED: n = 7, from the Child Eating parent and family as overweight by one’s parent
of ED-​status EDNOS: n = 27: Partial/​ Disorder Examination functioning; child health; at age 8 and 10 strongest
subthreshold: AN: n = 4, and EDE-​Q child personal and predictors of ED; high
BN: n = 13, emotional functioning; maternal BMI at 16 weeks of
BED: n = 14, eating, weight and shape gestation, sociaI problems,
EDNOS: n = 20) related higher neurocognitive
difficulties (age 8), and low
social-​related self-​efficacy (age
10) also predictive

8. Buckner N = 1,709 (46,3% M = 16.6 14 n = 13; AN: n = 4, SCID-​I disorders (major depressive, OCD the only significant
et al., 2010 male) (1.2) BN: n = 9 separation anxiety, social predictor of AN
anxiety, panic, overanxious,
obsessive-​compulsive
disorder, specific phobia)

9. Isomaa N = 81 (9.9% male) ~ 15 3 ED: n = 20, subclinical Interview based on the four subgroups of dieters (high High-​risk dieting (depressed
et al., 2010 ED: n = 15 Eating Disorder risk: depressed and feeling-​fat and feeling-​fat dieters)
Examination /​Survey dieters; low risk: vanity and OR = 14.86
for Eating Disorders overweight dieters)

10. Jacobi N = 236 (100% M = 20.8 1–​3 n = 24; BN: n = 1, Eating Disorder 58 risk Medium potency: negative
et al., 2011 female); high-​ (2.6) subthreshold BN: n = 11, Examination factors: sociodemographic comments by coach or
risk sample with subthreshold (Interview) and parental/​family factors, teacher, depression diagnoses,
elevated weight BED: n = 7, not maximum weight, lifetime compensatory behavior, EDE-​
and shape concerns specified: n = 7) mental disorders, negative Q Eating Concern Scale;
(based on WCS) comments on weight, shape Low potency: negative
and eating, eating disorder comments by siblings, parental
attitudes and behaviors (WCS, average weight, number of
EDE-​Q, EDI), childhood alcoholic drinks in a week
maltreatment, social support,
depressed mood, self-​esteem,
social adjustment, negative life
events, alcohol use
11. Stice N = 496 (100% 12–​22 1 ED: n = 86; EDDI, semistructured Perceived sociocultural Body dissatisfaction (BD)
et al., 2011 female) AN: n = 3, subthresh. interview adapted pressure to be thin, thin-​ showed greatest predictive
AN: n = 3, BN: n = 6, from the EDE ideal internalization, body potency in identifying youth at
subthresh. BN: n = 26, dissatisfaction, dieting, risk for ED onset. Participants
BED: n = 5, negative affectivity, in upper 24% of BD showed
subthresh. BED: n = 21, depressive symptoms 4.0-​fold increase in incidence
purging disorder: n = 22 for ED onset. Among high BD
participants, those in the upper
32% of depressive symptoms
showed a 2.9-​fold increase in
ED onset. Among participants
in the low BD branch, those
in the upper 12% of dieting
showed a 3.6-​fold increase in
ED onset

12. Tanofsky-​ N = 195 (37.4% 10.4 (1.5) 4.7 (1.2) BED, full or Eating Disorder Loss of control eating Baseline reports of LOC Ever
Kraff et al., male) partial: n = 9 Examination (LOC) predicted BED
2011 (Interview)

13. Thomas N = 496 (100% 13–​19 1–​2 Subthreshold BN: n = 14 Eating Disorder Age adjusted BMI change Substantial weight gain
et al., 2011 female) Diagnostic Interview, (weight increase or or weight loss produced a
adapted from EDE decrease) sevenfold increase in risk for
interview ED onset; however, effects not
significant

14. Allen N = 526 (218 Antenatal Assessment ED: n = 98 (87% 24 items adapted Maternal vitamin D Lowest maternal vitamin D
et al., 2013 males, 308 females) of ED status female) from the Child Eating concentrations at 18 weeks quartile vs. highest quartile
at age 14, Disorder Examination gestation significantly predicted ED risk
17, and 20 and EDE-​Q in female offspring

(continued)
Table 6.2 Continued

Study Sample (N) Age at baseline Follow-​up Cases Outcome (dependent Predictors (independent Results
(years) interval (y) variables) variables)

15. Allen, N = 1,297 (52% 10 Assessment ED: n = 146 24 items adapted from Parent-​perceived childhood Support for a mediational
Byrne, & male) of ED status (BN: n = 81, the EDE-​Q overweight at age 10; early model linking parent-​perceived
Crosby, 2014 at age 14, BED: n = 43, Purging adolescent eating, weight childhood overweight at age
17, and 20 disorder: n = 22) and shape concerns at age 10, adolescent eating, weight
14 as potential mediators and shape concerns at age 14,
and risk for a binge eating
or purging disorder in later
adolescence

16. Allen, N = 1.383 (49% Antenatal—​ at age 17 ED: n = 146 Later-​onset binge Five domains: Antenatal; Female sex, (high) eating,
Byrne, Oddy, male) 14 years and 20 (BN: n = 81, eating or purging, parent and family weight, and shape concerns at
et al., 2014 assessment BED: n = 43, Purging eating disorders, functioning; child health; age 14 predicted ED onset
of ED-​status disorder: n = 22) 24 self-​report items child personal and
adapted from the emotional functioning;
EDE-​Q eating, weight and shape
related

17. Hilbert N = 120 (matched 10.77 (1.46) 5.5 (0.6) BED (full or Eating Disorder Loss of control eating Persistent LOC significantly
& Brauhardt, sample, 43% male) partial): n = 16 Examination-​ (LOC) predicted partial/​full-​syndrome
2014 Questionnaire BED onset

18. Rohde N = 496 (100% 13–​16 1–​4 ED: n = 59; AN: n = 3, EDDI, semistructured Perceived pressure to Body dissatisfaction consistent
et al., 2015 female) BN: n = 12, interview adapted be thin, thin-​ideal predictor of ED-​onset;
BED: n = 14, atyp. from the EDE internalization, body negative affectivity, perceived
AN: n = 12, subthresh. dissatisfaction, dieting, pressure to be thin, thin-​ideal
BN: n = 21, subthresh. negative affectivity, BMI internalization, dieting also
BED: n = 14; purging predictive
disorder: n = 17; 33
more than one disorder)
problems. The potency of each of these risk factors factor (thin body preoccupation and social pressure)
ranges from low to large. measuring concerns with weight, shape, and eating
Rohde, Stice, and Marti (2015) found body dis- (including media modeling, social eating, dieting,
satisfaction between ages 13 and 16 to be predictive and weight teasing) significantly predicted onset
of future ED at all four of the examined assessment of EDs at both sites (Stanford and Arizona). In the
points. This sample included 25% of mostly subclin- Stice et  al. (2008) study, fasting predicted thresh-
ical AN cases. In two other studies (Beato-​Fernandez old and subthreshold BN cases. Depending on the
et al., 2004; Ghaderi & Scott, 2001), body dissatisfac- measure (continuous or dichotomous fasting meas-
tion was also found to be predictive but in both stud- ure), the potency was low or medium. Patton et al.
ies only one case of clinical AN respectively emerged. (1999) found that female moderate to severe dieters
Dieting was confirmed to be predictive in two inde- were 5 to 18 times more likely to develop subclin-
pendent samples with mixed outcomes including ical BN than those who did not. The same authors
10%–​25% of mostly subclinical AN cases (Fairburn, also found participants with the highest levels of
Cooper, Doll, & Davies, 2005; Rohde et  al., 2015; psychiatric morbidity to be more than six times
Stice, Marti, & Durant, 2011) as were childhood eat- more likely to develop an eating disorder compared
ing and digestive problems (Kotler et al., 2001; Marchi with those with lower levels. In the McKnight
& Cohen, 1990) in samples including 20 and 47% of Longitudinal Study (The McKnight Investigators,
mostly subclinical AN cases. Finally, negative affect 2003), higher scores on a factor psychological influ-
and/​or depressive symptoms were also confirmed in ences including depressed mood turned out to be a
two independent samples including 5%–​25% of AN proxy for the factor thin body preoccupation and
cases (Leon, Fulkerson, Perry, Keel, & Klump, 1999; social pressure and predicted onset of EDs at both
Rohde et al., 2015; Stice et al., 2011). sites (Stanford and Arizona).
In addition, the following risk factors were However, several other factors could be replicated
found to be predictive in one sample respectively in studies including lower rates of BN cases: body
including 3%–​31% of subclinical and clinical dissatisfaction/​weight and shape concerns, compen-
AN cases: Pressure to be thin and thin-​ideal inter- satory behavior, negative affect/​ depressed mood,
nalization age 14 (Rohde et  al., 2015), low BMI psychiatric morbidity, low social support, negative
(Fairburn et al., 2005), obsessive-​compulsive disor- comments and social pressure to be thin, negative
der (Buckner, Silgado, & Lewinsohn, 2010), higher life events, and early childhood eating problems.
scores on the General Health Questionnaire and Body dissatisfaction, body concerns, and weight
feelings of not being loved enough or ignored by and shape concerns were found to be predictive
mother or father (Beato-​Fernandez et  al., 2004), in four samples including 20%–​56% of (subclini-
physical neglect and sexual abuse (Johnson et  al., cal and full-​syndrome) BN cases (Beato-​Fernandez
2002), and low self-​esteem, low perceived social et al., 2004; Ghaderi & Scott, 2001; Jacobi et al.,
support, and escape-​avoidance coping (Ghaderi & 2011; Rohde et  al., 2015; Stice et  al., 2011).
Scott, 2001). However, in the latter study, only one Baseline compensatory behavior and purging were
case of AN was detected. predictive in two high-​risk studies with 50% of sub-
clinical and clinical BN (Jacobi et  al., 2011) and
Risk Factors and Markers for Bulimia 18% of BN cases (Fairburn et al., 2005).
Nervosa from Longitudinal Studies Negative affect and depressed mood predicted
Three of the previous and new longitudinal 20%–​56% of subclinical and full-​syndrome BN cases
risk factor samples fulfilled the criterion of detect- in four studies:  Rohde et  al. (2015) found negative
ing at least 75% of subclinical and full-​syndrome affect at ages 14 and 15 to be predictive of mixed ED
BN cases (Patton, Selzer, Coffey, Carlin, & Wolfe, cases including 56% of (subclinical and full-​syndrome)
1999; Stice, Davis, Miller, & Marti, 2008; The BN cases, while Leon et  al. (1999) found the latent
McKnight Investigators, 2003). Replicated risk variable negative affect in both girls and boys to be
factors from these samples are dieting/​fasting, and predictive of ED cases including 29% of BN cases.
psychiatric morbidity, specifically negative affect. Beato-​Fernandez et al. (2004) found higher scores on
Unfortunately, only one study (Stice et  al., 2008) the General Health Questionnaire including symp-
allowed for a potency determination based on AUV toms associated with anxiety and depression to pre-
of the factor dieting/​fasting. dict caseness in an adolescent sample including 20%
In the McKnight Longitudinal Study (The of BN cases. Jacobi et  al. (2011) found a lifetime
McKnight Investigators, 2003), higher scores on a diagnoses of depression assessed by the Structured

Jacobi, Hüt ter, Fit tig 117

Clinical Interview for DSM IV (SCID) to predict Risk Factors and Markers for Binge Eating
ED caseness in a high-​ risk sample of college-​ age Disorder from Longitudinal Studies
women selected by high weight and shape concerns, Four of the previous and new longitudinal risk
including 50% of mostly subclinical BN cases. factor samples fulfilled the criterion of detecting at
Psychopathology or psychiatric morbidity was also least 75% of subclinical and full-​syndrome BED
confirmed in several samples: A lifetime panic disor- cases (Allen et  al., 2014; Hilbert & Brauhardt,
der and alcohol use (number of drinks) per week was 2014; Tanofsky-​Kraff et  al., 2011; Vollrath et  al.,
also predictive in the latter high-​risk sample (Jacobi 1992). Replicated risk factors from these samples
et  al., 2011) and the change in general psychiatric are loss of control eating and BMI. Loss of control
morbidity in the study by Patton et al. (1990) includ- eating was confirmed to predict (partial or full-​
ing 25% of BN cases. Lower perceived social support syndrome) BED in two studies including children
from the family was reported significantly more often (Hilbert & Brauhardt, 2014; Tanofsky-​Kraff et al.,
before ED onset (including 42% of BN cases) by par- 2011). Higher BMI at age 14 and in young adult-
ticipants in the study by Ghaderi and Scott (2001) hood was confirmed to predict subclinical BED
compared with those who did not develop ED. In a also in two studies (Allen et  al., 2014; Vollrath
second study, the feeling of not being loved enough or et al., 1992). The potency of loss of control eating
ignored by mother or father predicted ED onset in a was high. Regarding BMI, it was low for the one
sample including 20% of BN cases (Beato-​Fernandez study that allowed for an estimate of potency (Allen
et al., 2004). Three studies (Jacobi et al., 2011; Rohde et al., 2014).
et al., 2015; The McKnight Investigators, 2003) con- In addition, the following risk factors could be
firmed the role of thin body preoccupation, social replicated in two or more samples including lower
pressure to be thin and negative comments about (<75%) rates of BED cases:  Thin-​ideal internal-
weight and shape by coach, teacher, sibling, or friend ization, thin body preoccupation and pressure to
in mostly subclinical BN (50%–​84%). Two studies be thin, body concern, body dissatisfaction, and
confirmed the role of a change in negative life events weight, eating, and shape concerns were confirmed
(The McKnight Investigators, 2003) and more specifi- in five samples with 16%–​100% of subclinical and
cally, physical neglect and sexual abuse (Johnson et al., clinical BED cases (Allen et  al., 2014; Ghaderi &
2002) in samples including 19% and 85% of BN cases. Scott, 2001; Jacobi et al., 2011; Rohde et al., 2015;
Also, two studies confirmed the role of early childhood Stice et  al., 2011; The McKnight Investigators,
eating problems (pica, eating too little, reducing con- 2003). Psychiatric morbidity, specifically negative
cerns) as risk factor for samples including 53% and affect, lifetime depression and panic disorder diag-
80% of BN cases respectively (Kotler et  al., 2001; nosis, alcohol consumption (number of drinks per
Marchi & Cohen, 1990). In the Marchi and Cohen week), and general psychological influences pre-
(1990) study, reducing concerns and behavior were dicted BED in three samples including 16%, 20%,
also significantly associated with the mostly subclini- and 47% of subclinical and clinical BED cases
cal BN cases (53%). (Jacobi et al., 2011; Rohde et al., 2015; Stice et al.,
Two studies including BMI as potential risk 2011; The McKnight Investigators, 2003).
factor for BN found significant but contradictory Finally, in addition to the factors mentioned
effects. Vollrath et al. (1992) found a positive associ- before, Allen et al. (2014) found antenatal mater-
ation between BMI and ED onset (19% BN cases) nal drug use, maternal health problems, assessed
whereas (Fairburn et  al., 2005) found a negative at age 14, 14-​ year objective binge eating and
association between BMI and ED onset (18% of externalizing problems, 8-​ year social problems,
BN cases). Three other studies including 50%–​84% and 10-​year withdrawn problems to predict sub-
of BN cases did not find a significant association clinical BED cases (100% of the sample). In our
at all (Jacobi et al., 2011; Rohde et al., 2015; The own high-​risk study including 29% of subclini-
McKnight Investigators, 2003). cal BED cases (Jacobi et al., 2011) we also found
Finally, in the high-​ risk sample of dieting baseline compensatory behavior; lower parental
women, (Fairburn et  al., 2005) also found those weight; negative comments from coach, teacher,
dieters to be at highest risk for developing an eating siblings, or friends; and EDI bulimia to be pre-
disorder 2  years later who—​apart from the desire dictive. Ghaderi and Scott (2001) found low self-​
to have an empty stomach—​initially also displayed esteem, high use of escape-​avoidance coping, low
objective bulimic episodes, secret eating, and fear of perceived social support, and high body concerns
losing control. to be predictive in a sample including 38% of

118 Psychosocial Risk Factors

BED cases. Finally, one study confirmed the role markers to the ED onset was not specified in the
of a change in negative life events (The McKnight papers and the time periods for the assessment of
Investigators, 2003) in a sample including 16% risk factors and ED onset were overlapping, that is,
of BED cases, whereas one other study confirmed the risk factors could have occurred after ED onset.
the role of physical neglect and sexual abuse as The majority of the register studies examined birth-​
risk factors for a sample including 17% of BED related and sociodemographic variables (e.g., season
cases (Johnson et al., 2002). of birth, birth-​related characteristics) retrospectively
from predominantly Scandinavian birth registers,
Risk Factors and Markers for Nonspecific psychiatric case registers, population-​ based regis-
Eating Disorder Cases from Longitudinal ters, and census and other databases. With exclu-
Studies sion of overlapping samples, the register studies
Very few factors remain from longitudinal stud- comprise N = 23,407 ED cases in total (diagnoses
ies including EDNOS outcomes, that is, studies reported: AN: n = 9,149; subclinical AN: n = 201;
that did not specify individual diagnostic catego- BN: n = 2,181; EDNOS: n = 5,216).
ries. Cervera et  al. (2003) found higher levels of Season of birth for both AN and BN was exam-
neuroticism to be predictive of EDNOS, whereas ined in five studies. While the evidence for this
higher levels of self-​esteem turned out to be a pro- factor—​at least for AN—​has been inconsistent in
tective factor. Mothers of women with lifetime the past, the newer studies including larger, cross-​
EDs in the longitudinal study by Moorhead et al. national samples provide more consistent evi-
(2003) reported that their children at age 9 expe- dence: Vellisca, Latorre, Santed, and Reales (2013)
rienced more problems with anxiety-​ depression found no evidence of significant variation in month
measured by the Simmons Behavior Checklist and or season of birth in a Spanish sample of AN patients
more health problems before age 5 than non-​eating-​ when compared with the general population. In the
disordered peers. Cases were mostly partial cases of to-​date largest study carried out in different coun-
ED. In the study by Isomaa, Isomaa, Marttunen, tries across the world including N = 4,045 patients
Kaltiala-​Heino, and Bjorkqvist (2010), high-​ risk with AN, Winje, Torgalsboen, Brunborg, and Lask
dieters had a 15-​fold risk of developing a DSM-​IV (2013a) could not support the hypothesis of an
eating disorder compared to the low-​risk dieters. association between season of birth and AN.
Allen, Byrne, Forbes, and Oddy (2009) examined For BN, one study from Norway Winje,
a wide range of potential risk factors for a mix- Torgalsboen, Brunborg, and Lask (2013b) did not
ture of full and partial ED relative to general and find a statistically significant association between
psychiatric control groups in a population-​based month of birth and the distribution of births
sample of early adolescents drawn from the Western compared with women in the general population,
Australian Pregnancy Cohort. Being female and while another study (Brewerton, Dansky, O’Neil,
being perceived as overweight by one’s parent were & Kilpatrick, 2012) found the highest number
the strongest predictors in the final multivariate of births of women with BN, binge eating, and
model relative to both control groups. In addition, purging to have occurred in fall compared with
maternal BMI, social problems, low social-​related women in the general population. Finally, Button
self-​
efficacy, and neurocognitive difficulties were and Aldridge (2007), in a mixed sample of patients
predictive relative to the general control group only. with ED did not find any evidence of a significant
variation in month or season of birth in AN, BN,
Results of Risk Factors and Markers or EDNOS. Taken together, when the latest studies
for Eating Disorders from Register Studies are taken into account, the preponderance of stud-
Between 2002 and 2015, a relatively large num- ies showing a significant association between season
ber of studies trying to identify markers on the basis of birth and AN and BN cannot be confirmed.
of register studies were conducted in addition to Together with previous inconsistent findings, this
the longitudinal studies. Overall, 13 register studies factor therefore cannot be classified as risk factor
(N = at least 4,541,350 participants in independent (marker) for ED any more.
samples) and two fixed marker studies (N = 1,504 However, several newer register studies once
participants; n = 61 ED cases including n = 27 AN more support the role of pre-​and perinatal com-
cases) were included. Several other register studies plications and other birth characteristics as mark-
published between 2002 and 2015 were excluded ers for ED. Lindberg and Hjern (2003), in a large
because the precedence of the risk factors and sample of inpatients with AN (N = 1,122), found

Jacobi, Hüt ter, Fit tig 119

that premature children with a gestational age of 23 education predicted a higher rate of ED in females
to 32 and 33 to 36 weeks showed a higher risk of but not in males. In females, an increasing number
developing AN. In addition, birth traumata such as of full-​siblings was associated with lower rate of ED,
cephalhematoma (a hemorrhage of blood between whereas an increasing number of half-​siblings was
the skull and the periosteum) and premature rup- associated with a higher rate.
ture of membrane were associated with a higher Finally, two other register studies also tried
risk of developing AN, as were intercountry adop- to identify fixed markers using somewhat differ-
tion and experiences of foster care before age 13. ent methods. Romero-​Martinez and Moya-​Albiol
The role of perinatal factors was confirmed by the (2014) examined the 2d:4d finger ratio as proxy
second study by Favaro, Tenconi, and Santonastaso for prenatal exposure to testosterone in 34 parents
(2006). Several obstetric complications (maternal of offspring with AN compared with autism spec-
anemia, diabetes mellitus, preeclampsia, placental trum disorders and control parents, but found only
infarction, neonatal cardiac problems, hyporeac- partial support for their hypothesis. Mazzeo et  al.
tivity) turned out to be predictors of the develop- (2009) examined genetic and environmental con-
ment of AN. The risk of developing AN increased tributions to liability to AN symptoms in 27 AN
with the total number of obstetric complications. cases identified out of 1,430 twins of a Norwegian
Obstetric complications predictive of BN were twin register. When comparing monozygotic and
placental infarction, neonatal hyporeactivity, early dizygotic twin pairs in an ACE model they found
eating difficulties, and low birth weight for gesta- modest evidence for heritability of AN diagnoses.
tional age. Compared with the reference group,
for patients with ED in the study by Nosarti et al. Interactions Between Risk Factors
(2012), very preterm birth (<32 weeks) was asso- The lack of consideration of interactions
ciated with a more than threefold increased risk of between risk factors is one of the limitations of pre-
ICD-​8 to—​10 ED diagnoses. Lower gestational age vious risk factor studies. Considering interactions
was also confirmed as marker for AN in a Swedish may, however, improve the understanding of the
total-​
population based cohort study (Goodman, pathways of the development of the disorder, thus
Heshmati, Malki, & Koupil, 2014). Other markers enabling us to improve the effectiveness of preven-
found in this study were multiple birth for AN and tive interventions. To address the different ways of
higher birth weight for BN. interactions between risk factors (i.e., overlapping
Based on our previous review, the role of famil- factors, proxies, mediators, moderators), Kraemer,
ial/​
parental psychopathology could only be clas- Stice, Kazdin, Offord, and Kupfer (2001) proposed
sified as a retrospective correlate because existing additional definitions and methodological recom-
data were all cross-​sectional in nature. Bould et al. mendations a few years after publishing their risk
(2015) conducted the first study to examine this factor taxonomy.
relationship based on data from a Swedish total In the context of EDs, this extended methodo-
population register combined with other health logical approach has been applied in only three stud-
register data. They found higher ICD-​8 to—​10 ED ies:  The McKnight Investigators (2003) examined
incidence rates when parents had suffered from any risk factors for EDs and their interactions in more
mental illness, depression/​ anxiety, bipolar disor- than 1,000 adolescent girls followed for 3 years. At
der, or personality disorder. Bulik et al. (2006), in both sites (Arizona, Stanford), thin body preoccu-
a population-​based twin study found neuroticism, pation and social pressure to be thin (TBPSP) were
assessed 30  years earlier, to be predictive of later significantly related to ED onset. At the Arizona
subclinical and clinical AN cases. site, one proxy (general psychological influences)
Goodman, Heshmati, and Koupil (2014) also for TBPSP was found. Being Hispanic moderated
assessed a number of parental and grandparental TBPSP. In addition, change in negative life events
socioeconomic variables in their total-​population turned out to be an independent risk factor.
based cohort study. The ED (AN, BN and In the second study, Agras, Bryson, Hammer,
EDNOS) incidence in females was independently and Kraemer (2007) examined childhood precur-
predicted by greater educational level among the sors and their interactions of the most potent risk
father, mother, and maternal grandparents. Among factors for eating disorders, thin body preoccupa-
males, a similar pattern was found with for AN tion and social pressure to be thin (TBPSP) in 11-​
only. Other socioeconomic variables were assessed year old children (N = 134). Two moderators were
in the study by Ahren et al. (2013). Higher parental found identifying different groups at risk for the

120 Psychosocial Risk Factors

development of (TBPSP). A father with high body the overall number of ED cases detected increased
dissatisfaction characterized the largest group, in markedly. While numbers of AN and BN cases dif-
which TBPSP was elevated for girls who were con- fer only slightly from previous numbers, more recent
cerned about and attempted to modify their weight studies—​probably as a consequence of the introduc-
and for children with fathers who had a high drive tion of DSM-​5 (American Psychiatric Association,
for thinness. A child at risk for overweight charac- 2013)—​ included newer ED diagnoses, namely,
terized the second smaller group. Parental behaviors (clinical and subclinical) BED cases as well as few
such as overcontrol of their child’s eating, together cases of purging disorder. At the same time however,
with later pressure from parents and peers to be for about a third of all cases classified as ED in the
thin, were related to higher levels of TBPSP. individual studies, no specific subclinical or clini-
Our own study assessed risk factors and their cal diagnoses are reported. In addition, the majority
interactions within a college-​ age population of of the reported diagnoses still represent unclear and
N = 236 women already at higher risk defined by inconsistently defined subclinical conditions. A rel-
elevated weight and shape concerns at baseline atively large number of new register studies emerged
(Jacobi et al., 2011). Participants were assessed and with impressively large sample sizes. However, the
followed over up to 3  years. At 3-​year follow-​up, number of potential risk factors in these studies is
11.2% of subjects had developed a full or partial often very limited, and precedence to ED onset is
ED. Ten of 88 potential risk factors could be clas- not always given for all included variables. Although
sified as independent risk factors, eight as prox- to a yet limited extent, more recent studies begin to
ies. Comments about eating, from teacher/​coach/​ address interactions between risk factors.
siblings, and a history of depression were the most Across all studies, despite slight advantages,
potent risk factors. The incidence for participants many previously mentioned limitations remain.
with either or both of these risk factors was 34.8% They become even more obvious, when a relatively
(16/​46) compared to 4.2% (6/​144) for students strict criterion for the prediction of caseness (selec-
without these risk factors. Although the study dif- tion of studies with ≥75% of cases of one diagnos-
fered from most of the studies included in the meta-​ tic category) is applied to examine the specificity of
analysis with regard to risk status, age, and sample risk factors for the different ED diagnoses. For AN,
size, risk factors found in this high-​risk college-​age only two samples examining fixed markers fulfilled
sample are in accordance with factors from the this criterion and very few other risk factors could
meta-​analysis. be replicated in samples with much lower propor-
Using a slightly different methodology, Stice tions of AN cases. While identifying risk factors
et al. (2008) tried to identify interactions between for AN in general remains the biggest challenge,
risk factors implying distinct risk pathways in a pro- the identification of specific risk factors for AN (vs.
spective study of 496 adolescent females followed BN or BED) is restricted to factors already present
over 8 years. New ED onset cases comprised a mixed before or during birth, which—​for the individual
group of predominantly subclinical ED diagnoses. itself—​cannot be changed anymore later in life. For
Body dissatisfaction emerged as the most potent BN and BED, two risk factors respectively, could
predictor, with girls in the upper 24% of body dis- be identified in samples including at least 75% of
satisfaction showing a fourfold increased incidence cases of the respective category: dieting/​fasting and
of ED onset (24% vs. 6%). Among participants in psychiatric morbidity for BN, and loss of control
the high body dissatisfaction branch, those in the eating and (higher) BMI for BED. Currently, these
upper 32% of depressive symptoms showed a 2.9-​ factors represent the best replicated risk factors for
fold increased incidence of onset (43% vs. 15%). BN and BED diagnoses. Because the overall num-
Among participants in the low body dissatisfaction ber of studies including BED cases is still relatively
branch, those in the upper 12% of dieting showed small compared with BN, future studies might lead
a 3.6-​fold increased incidence onset (18% vs. 5%). to different conclusions regarding the specificity of
these risk factors. Given the limited number of stud-
Conclusions ies with replicated risk factors identifying more than
Our risk factor update illustrates few changes 75% of cases of one diagnostic category, assump-
in the more recent studies compared to the ear- tions on potency unfortunately also have to remain
lier studies: Despite an almost equal total number vague, ranging from low to high, and currently do
of participants examined in longitudinal studies, not allow for a meta-​analytic effect size determina-
diagnostic assessment seems to have improved and tion of risk factor-​diagnoses combinations.

Jacobi, Hüt ter, Fit tig 121

 When samples with less than 75% of a specific studies addressing interactions between risk factors
ED diagnosis are examined, the picture becomes demonstrate that individuals are at highest risk for
even more blurry and many risk factors turn out ED onset if both factors are present (Jacobi et al.,
to be predictive of all diagnostic categories of 2011; Stice et al., 2008). Few preventive interven-
ED:  Dieting/​weight and shape concerns and psy- tions assume (indirect) secondary or meditational
chiatric morbidity (especially negative affect/​ effects on negative affect or depression (Stice,
depression) have been replicated for all categories; Butryn, Rohde, Shaw, & Marti, 2013; Stice, Rohde,
early childhood eating problems have been repli- Butryn, Shaw, & Marti, 2015). To our knowledge,
cated for both AN and BN; while negative life to date, only one preventive trial has—​in addition
events have been replicated for both BN and BED. to weight and shape concerns—​directly targeted
Because psychiatric morbidity/​ negative affect negative affect/​depression by selecting individu-
and negative life events have been confirmed as a als based on current or lifetime depression (Taylor
risk factors for other mental disorders (Hayward, et  al., 2016) and tailoring the intervention to the
Killen, Kraemer, & Taylor, 2000; Hirshfeld-​Becker, most potent risk factors (Jacobi et al., 2011).
Micco, Simoes, & Henin, 2008) they are very likely
nonspecific risk factors for a range of psychiatric
Future Directions
conditions, while it seems plausible that weight-​
Despite some improvements in the more recent
and shape-​related variables and dieting are specific
studies, future studies should address the limitations
risk factors for ED. Unfortunately, only one of the
pointed out previously:
newer studies included other outcomes than ED
(Allen, Byrne, Forbes, & Oddy, 2009) and found 1.  Specific ED diagnoses should be assessed and
only two variables (child sex and parent-​perceived reported.
child overweight) to be specifically predictive of ED 2.  Because subclinical ED syndromes represent
versus psychiatric controls in the final model. the largest outcome category, the required criteria
Based on the more recent findings, the status for these syndromes should be made explicit and be
of two risk factors changed:  For AN, several new distinguishable from EDNOS/​OSFED (American
studies do not support the role of season of birth Psychiatric Association, 2013) categories.
as fixed marker any more and for BN, the current 3.  Even in studies with small numbers of cases,
database is not sufficient for a classification of this effect sizes per risk factor diagnosis combination
factor. Also, BMI cannot be supported any more by should be reported to enable future meta-​analytic
more recent studies as risk factor for BN but turned potency determination.
out to be predictive for BED in two studies. 4.  To address the specificity of risk factors for ED
For the prevention of ED, targeting variable, versus other mental disorders, other outcomes need to
potent, and replicated risk factors predicting future be assessed at follow-​up and even larger sample sizes
ED onset by preventive interventions is crucial. may be required. Although the inclusion of high-​risk
The few AN-​specific factors are not suitable as tar- samples has been suggested as one way of increasing
gets for preventive interventions, since they cannot the chance of identifying cases (Jacobi et al., 2004),
be changed. They could—​at best—​be included to to date very few studies made use of such a high-​risk
inform mothers with birth complications and pre- design (Fairburn et  al., 2005; Isomaa et  al., 2010;
term birth to be mindful of their offspring’s increased Jacobi et al., 2011), and only one of these examined
risk for AN. However, childhood loss of control eat- additional risk factors (Jacobi et al., 2011).
ing and BMI could be included in the development 5.  To facilitate case prediction and to better
of specific preventive interventions for BED, while determine the relative potency of some factors in
they may not be as relevant for other EDs. Female relation to others, more studies are needed includ-
gender, body dissatisfaction/​weight and shape con- ing a range of factors instead of just a few. These
cerns, dieting, and psychiatric morbidity/​negative might be best addressed in populations selected by
affect remain the best candidates but—​given their well-​replicated risk factors.
nonspecific nature—​are not suitable for inclusion 6.  Preventive interventions for AN and BED
the development of diagnosis-​ specific preventive might benefit from both the inclusion of the few
interventions. Currently, the focus of most preven- specific risk factors while preventive interven-
tive interventions for ED is on risk factors such as tions for ED in general might benefit from tar-
body dissatisfaction/​weight and shape concerns, and geting negative affect or depression by selecting
dieting (Stice, Shaw, & Marti, 2007), although two individuals based on these risk factors or adapting

122 Psychosocial Risk Factors

intervention content to be able to target these risk in early adulthood? A  7-​year longitudinal study. International
factors for individuals at highest risk. Journal of Eating Disorders, 24, 351–​362. doi:10.1002/​
(sici)1098-​108x(199812)24:4<351::aid-​eat2>3.0.co;2-​1
Cervera, S., Lahortiga, F., Martinez-​Gonzalez, M. A., Gual, P.,
References de Irala-​Estevez, J., & Alonso, Y. (2003). Neuroticism and
Agras, W. S., Bryson, S., Hammer, L. D., & Kraemer, H. C. low self-​esteem as risk factors for incident eating disorders
(2007). Childhood risk factors for thin body preoccupa- in a prospective cohort study. International Journal of Eating
tion and social pressure to be thin. Journal of the American Disorders, 33, 271–​280. doi:10.1002/​eat.10147
Academy of Child and Adolescent Psychiatry, 46, 171–​178. Dominy, N. L., Johnson, W. B., & Koch, C. (2000). Perception
doi:10.1097/​chi.0b01361802bd997 of parental acceptance in women with binge eating disorder.
Ahren, J. C., Chiesa, F., Koupil, I., Magnusson, C., Dalman, C., Journal of Psychology, 134, 23–​36.
& Goodman, A. (2013). We are family-​parents, siblings, and Fairburn, C. G., Cooper, Z., Doll, H. A., & Davies, B. A.
eating disorders in a prospective total-​population study of (2005). Identifying dieters who will develop an eating disor-
250,000 Swedish males and females. International Journal of der: A prospective, population-​based study. American Journal
Eating Disorders, 46, 693–​700. doi:10.1002/​eat.22146 of Psychiatry, 162, 2249–​2255.
Allen, K. L., Byrne, S. M., & Crosby, R. D. (2014). Distinguishing Fairburn, C. G., Cooper, Z., Doll, H. A., & Welch, S. L. (1999).
between risk factors for bulimia nervosa, binge eating disor- Risk factors for anorexia nervosa: Three integrated case-​control
der, and purging disorder. Journal of Youth and Adolescence, comparisons. Archives of General Psychiatry, 56, 468–​476.
44, 1580–​1591. doi:10.1007/​s10964-​014-​0186-​8 Favaro, A., Tenconi, E., & Santonastaso, P. (2006). Perinatal
Allen, K. L., Byrne, S. M., Forbes, D., & Oddy, W. H. (2009). Risk factors and the risk of developing anorexia nervosa and
factors for full—​and partial-​syndrome early adolescent eating bulimia nervosa. Archives of General Psychiatry, 63, 82–​88.
disorders: A population-​based pregnancy cohort study. Journal doi:10.1001/​archpsyc.63.1.82
of the American Academy of Child and Adolescent Psychiatry, Ghaderi, A. (2001). Eating disorders:  Prevalence, incidence, and
48, 800–​809. doi:10.1097/​CHI.0b013e3181a8136d prospective risk factors for eating disorders among young adult
American Psychiatric Association. (2013). Diagnostic and sta- women in the general population (Vol. 99). Uppsala:  Acta
tistical manual of mental disorders (5th ed.). Arlington, Universitatis Upsaliensis.
VA: American Psychiatric Association. Ghaderi, A., & Scott, B. (2001). Prevalence, incidence and prospective
Attie, I., & Brooksgunn, J. (1989). Development of eating prob- risk factors for eating disorders. Acta Psychiatrica Scandinavica,
lems in adolescent girls: A longitudinal-​study. Developmental 104, 122–​130. doi:10.1034/​j.1600-​0447.2001.00298.x
Psychology, 25, 70–​79. doi:10.1037/​0012-​1649.25.1.70 Goodman, A., Heshmati, A., & Koupil, I. (2014). Family history
Beato-​Fernandez, L., Rodriguez-​Cano, T., Belmonte-​Llario, of education predicts eating disorders across multiple gen-
A., & Martinez-​Delgado, C. (2004). Risk factors for eating erations among 2 million Swedish males and females. PLoS
disorders in adolescents: A Spanish community-​based longi- ONE, 9, 9. doi:10.1371/​journal.pone.0106475
tudinal study. European Child and Adolescent Psychiatry, 13, Goodman, A., Heshmati, A., Malki, N., & Koupil, I. (2014).
287–​294. doi:10.1007/​s00787-​004-​0407-​x Associations between birth characteristics and eating
Bould, H., Koupil, I., Dalman, C., DeStavola, B., Lewis, G., disorders across the life course:  Findings from 2 Million
& Magnusson, C. (2015). Parental mental illness and eat- males and females born in Sweden, 1975–​1998. American
ing disorders in offspring. International Journal of Eating Journal of Epidemiology, 179, 852–​863. doi:10.1093/​aje/​
Disorders, 48, 383–​391. doi:10.1002/​eat.22325 kwt445
Brewerton, T. D., Dansky, B. S., O’Neil, P. M., & Kilpatrick, Graber, J. A., Brooksgunn, J., Paikoff, R. L., & Warren, M. P.
D. G. (2012). Seasonal patterns of birth for subjects with (1994). Prediction of eating problems:  An 8-​year study of
bulimia nervosa, binge eating, and purging:  Results from adolescent girls. Developmental Psychology, 30, 823–​834.
the National Women’s Study. International Journal of Eating doi:10.1037/​0012-​1649.30.6.823
Disorders, 45, 131–​134. doi:10.1002/​eat.20898 Hayward, C., Killen, J. D., Kraemer, H. C., & Taylor, C. B.
Buckner, J. D., Silgado, J., & Lewinsohn, P. M. (2010). (2000). Predictors of panic attacks in adolescents. Journal of
Delineation of differential temporal relations between specific the American Academy of Child and Adolescent Psychiatry, 39,
eating and anxiety disorders. Journal of Psychiatric Research, 207–​214. doi:10.1097/​00004583-​200002000-​00021
44, 781–​787. doi:10.1016/​j.jpsychires.2010.01.014 Hilbert, A., & Brauhardt, A. (2014). Childhood loss of con-
Bulik, C. M., Sullivan, P. F., Tozzi, F., Furberg, H., Lichtenstein, trol eating over five-​year follow-​up. International Journal of
P., & Pedersen, N. L. (2006). Prevalence, heritability, and Eating Disorders, 47, 758–​761. doi:10.1002/​eat.22312
prospective risk factors for anorexia nervosa. Archives of Hilbert, A., Pike, K. M., Goldschmidt, A. B., Wilfley, D. E.,
General Psychiatry, 63, 305–​312. Fairburn, C. G., Dohm, F. A.,  . . .  Weissman, R. S. (2014).
Button, E., & Aldridge, S. (2007). Season of birth and eating Risk factors across the eating disorders. Psychiatry Research,
disorders:  Patterns across diagnoses in a specialized eating 220, 500–​506. doi:10.1016/​j.psychres.2014.05.054
disorders service. International Journal of Eating Disorders, Hirshfeld-​Becker, D. R., Micco, J. A., Simoes, N. A., & Henin,
40, 468–​471. doi:10.1002/​eat.20381 A. (2008). High risk studies and developmental antecedents
Button, E., Sonuga-​ Barke, E., Davies, J., & Thompson, M. of anxiety disorders. American Journal of Medical Genetics Part
(1996). A prospective study of self-​esteem in the prediction C-​Seminars in Medical Genetics, 148c, 99–​117. doi:10.1002/​
of eating problems in adolescent schoolgirls: Questionnaire ajmg.c.30170
findings. British Journal of Clinical Psychology, 35, 193–​203. Isomaa, R., Isomaa, A. L., Marttunen, M., Kaltiala-​Heino, R.,
doi:10.1111/​j.2044-​8260.1996.tb01176.x & Bjorkqvist, K. (2010). Psychological distress and risk for
Calam, R., & Waller, G. (1998). Are eating and psychosocial character- eating disorders in subgroups of dieters. European Eating
istics in early teenage years useful predictors of eating characteristics Disorders Review, 18, 296–​303. doi:10.1002/​erv.1004

Jacobi, Hüt ter, Fit tig 123

Jacobi, C. (2005). Psychosocial risk factors for eating disor- Assessing the heritability of anorexia nervosa symptoms
ders. In S. Wonderlich, J. Mitchell, M. de Zwaan, & H. using a marginal maximal likelihood approach. Psychological
Steiger (Eds.), Eating disorders review (pp. 59–​85). Oxford, Medicine, 39, 463–​473. doi:10.1017/​s0033291708003310
UK: Radcliffe. Moorhead, D. J., Stashwick, C. K., Reinherz, H. Z., Giaconia,
Jacobi, C., & Fittig, E. (2010). Psychosocial risk factors for eat- R. M., Streigel-​Moore, R. M., & Paradis, A. D. (2003).
ing disorders. In W. S. Agras (Ed.), The Oxford handbook Child and adolescent predictors for eating disorders in a
of eating disorders. (pp. 123–​136). New  York, NY:  Oxford community population of young adult women. International
University Press. Journal of Eating Disorders, 33, 1–​9. doi:10.1002/​eat.10105
Jacobi, C., Fittig, E., Bryson, S. W., Wilfley, D., Kraemer, H. Nosarti, C., Reichenberg, A., Murray, R. M., Cnattingius, S.,
C., & Taylor, C. B. (2011). Who is really at risk? Identifying Lambe, M. P., Yin, L.,  . . .  Hultman, C. M. (2012). Preterm
risk factors for subthreshold and full syndrome eating disor- birth and psychiatric disorders in young adult life. Archives of
ders in a high-​risk sample. Psychological Medicine, 41, 1939–​ General Psychiatry, 69, 610–​617.
1949. doi:10.1017/​S0033291710002631 Patton, G. C., Johnson-​Sabine, E., Wood, K., & Mann, A.
Jacobi, C., Hayward, C., de Zwaan, M., Kraemer, H. C., & Agras, H. (1990). Abnormal eating attitudes in London school-
W. S. (2004). Coming to terms with risk factors for eating girls: A prospective epidemiological study: Outcome at twelve
disorders:  Application of risk terminology and suggestions month follow-​up. Psychological Medicine, 20, 383–​394.
for a general taxonomy. Psychological Bulletin, 130, 19–​65. Patton, G. C., Selzer, R., Coffey, C., Carlin, J. B., & Wolfe, R.
Johnson, J. G., Cohen, P., Kasen, S., & Brook, J. S. (2002). (1999). Onset of adolescent eating disorders:  population
Childhood adversities associated with risk for eating dis- based cohort study over 3 years. British Medical Journal, 318,
orders or weight problems during adolescence or early 765–​768.
adulthood. American Journal of Psychiatry, 159, 394–​400. Rohde, P., Stice, E., & Marti, C. N. (2015). Development and
doi:10.1176/​appi.ajp.159.3.394 predictive effects of eating disorder risk factors during adoles-
Killen, J. D., Taylor, C. B., & Hayward, C. (1994). Pursuit of cence: Implications for prevention efforts. International Journal
thinness and onset of eating disorder symptoms in a com- of Eating Disorders, 48, 187–​198. doi:10.1002/​eat.22270
munity sample of adolescent girls: A three year prospective Romero-​Martinez, A., & Moya-​Albiol, L. (2014). Prenatal tes-
study. International Journal of Eating Disorders, 16, 227–​238. tosterone of progenitors could be involved in the etiology of
Killen, J. D., Taylor, C. B., Hayward, C., Haydel, K. F., Wilson, both anorexia nervosa and autism spectrum disorders of their
D. M., Hammer, L.,  . . .  Strachowski, D. (1996). Weight offspring. American Journal of Human Biology, 26, 863–​866.
concerns influence the development of eating disorders: A 4-​ doi:10.1002/​ajhb.22597
year prospective study. Journal of Consulting and Clinical Stice, E., Butryn, M. L., Rohde, P., Shaw, H., & Marti, C. N.
Psychology, 64, 936–​940. doi:10.1037//​0022-​006x.64.5.936 (2013). An effectiveness trial of a new enhanced dissonance
Kotler, L. A., Cohen, P., Davies, M., Pine, D. S., & Walsh, B. T. eating disorder prevention program among female college
(2001). Longitudinal relationships between childhood, ado- students. Behaviour Research and Therapy, 51, 862–​871.
lescent, and adult eating disorders. Journal of the American doi:10.1016/​j.brat.2013.10.003
Academy of Child and Adolescent Psychiatry, 40, 1434–​1440. Stice, E., Davis, K., Miller, N. P., & Marti, C. N. (2008).
doi:10.1097/​00004583-​200112000-​00014 Fasting increases risk for onset of binge eating and bulimic
Kraemer, H. C., Kazdin, A. E., Offord, D. R., Kessler, R. C., pathology:  A  5-​year prospective study. Journal of Abnormal
Jensen, P. S., & Kupfer, D. J. (1997). Coming to terms with Psychology, 117, 941–​946. doi:10.1037/​a0013644
the terms of risk. Archives of General Psychiatry, 54, 337–​343. Stice, E., Marti, C. N., & Durant, S. (2011). Risk factors for
Kraemer, H. C., Stice, E., Kazdin, A., Offord, D., & Kupfer, onset of eating disorders: Evidence of multiple risk pathways
D. (2001). How do risk factors work together? Mediators, from an 8-​year prospective study. Behaviour Research and
moderators, and independent, overlapping, and proxy Therapy, 49, 622–​627. doi:10.1016/​j.brat.2011.06.009
risk factors. American Journal of Psychiatry, 158, 848–​856. Stice, E., Rohde, P., Butryn, M. L., Shaw, H., & Marti, C. N.
doi:10.1176/​appi.ajp.158.6.848 (2015). Effectiveness trial of a selective dissonance-​based
Leon, G. R., Fulkerson, J. A., Perry, C. L., & Earlyzald, M. B. eating disorder prevention program with female college stu-
(1995). Prospective analysis of personality and behavioral dents: Effects at 2—​and 3-​year follow-​up. Behaviour Research
vulnerabilities and gender influences in the later develop- and Therapy, 71, 20–​26. doi:10.1016/​j.brat.2015.05.012
ment of disordered eating. Journal of Abnormal Psychology, Stice, E., Shaw, H., & Marti, C. N. (2007). A meta-​analytic
104, 140–​149. doi:10.1037/​0021-​843x.104.1.140 review of eating disorder prevention programs: Encouraging
Leon, G. R., Fulkerson, J. A., Perry, C. L., Keel, P. K., & Klump, findings. Annual Review of Clinical Psychology, 3, 207–​231.
K. L. (1999). Three to four year prospective evaluation of doi:10.1146/​annurev.clinpsy.3.022806.091447
personality and behavioral risk factors for later disordered Striegel-​Moore, R. H., Dohm, F. A., Pike, K. M., Wilfley, D.
eating in adolescent girls and boys. Journal of Youth and E., & Fairburn, C. G. (2002). Abuse, bullying, and discrimi-
Adolescence, 28, 181–​196. doi:10.1023/​a:1021649314458 nation as risk factors for binge eating disorder. American
Lindberg, L., & Hjern, A. (2003). Risk factors for anorexia ner- Journal of Psychiatry, 159, 1902–​1907. doi:10.1176/​appi.
vosa: A national cohort study. International Journal of Eating ajp.159.11.1902
Disorders, 34, 397–​408. doi:10.1002/​eat.10221 Tanofsky-​Kraff, M., Shomaker, L. B., Olsen, C., Roza, C. A.,
Marchi, M., & Cohen, P. (1990). Early-​childhood eating behav- Wolkoff, L. E., Columbo, K. M.,  . . .  Yanovski, J. A. (2011).
iors and adolescent eating disorders. Journal of the American A prospective study of pediatric loss of control eating and
Academy of Child and Adolescent Psychiatry, 29, 112–​117. psychological outcomes. Journal of Abnormal Psychology, 120,
doi:10.1097/​00004583-​199001000-​00017 108–​118. doi:10.1037/​a0021406
Mazzeo, S. E., Mitchell, K. S., Bulik, C. M., Reichborn-​ Taylor, C. B., Kass, A. E., Trockel, M., Cunning, D., Weisman,
Kjennerud, T., Kendler, K. S., & Neale, M. C. (2009). H., Bailey, J.,  . . .  Wilfley, D. E. (2016). Reducing eating

124 Psychosocial Risk Factors

 disorder onset in a very high risk sample with significant Vollrath, M., Koch, R., & Angst, J. (1992). Binge eating and
comorbid depression:  A  randomized controlled trial. weight concerns among young adults:  Results from the
Journal of Consulting and Clinical Psychology, 84, 402–​414. Zurich cohort study. British Journal of Psychiatry, 160, 498–​
doi:10.1037/​ccp0000077 503. doi:10.1192/​bjp.160.4.498
The McKnight Investigators. (2003). Risk factors for the onset of Winje, E., Torgalsboen, A. K., Brunborg, C., & Lask, B. (2013a).
eating disorders in adolescent girls: Results of the McKnight Season of birth bias and anorexia nervosa: Results from an
longitudinal risk factor study. American Journal of Psychiatry, international collaboration. International Journal of Eating
160, 248–​254. Disorders, 46, 340–​345. doi:10.1002/​eat.22060
Vellisca, M. Y., Latorre, J. I., Santed, M. A., & Reales, J. M. Winje, E., Torgalsboen, A. K., Brunborg, C., & Lask, B.
(2013). Absence of seasonal pattern of birth in patients with (2013b). Season of birth bias and bulimia nervosa: Results
anorexia nervosa. International Journal of Eating Disorders, from a multi-​centre collaboration. European Eating Disorders
46, 86–​88. doi:10.1002/​eat.22046 Review, 21, 170–​174. doi:10.1002/​erv.2217

Jacobi, Hüt ter, Fit tig 125

 C H A PT E R


Dieting and the Eating Disorders
7
Eric Stice and Heather Shaw

Abstract
This chapter reviews theories and empirical evidence linking dietary restraint to eating pathology.
Prospective studies suggest that dieting increases risk for future onset of bulimic pathology, yet
experiments indicate that assignment to weight loss and weight maintenance diets reduces eating
disorder symptoms. The chapter considers explanations for these inconsistent findings. The analysis
suggests that dieting is not a causal risk factor for bulimic pathology but may be a proxy risk factor. It
posits that a tendency toward overeating may lead to both dieting and eating pathology. Nonetheless,
fasting does increase the reward value of food and appears to increase the likelihood of binge eating,
suggesting that eating disorder treatments should promote consuming three healthy meals daily. The
chapter proposes studies that should help differentiate dieting behaviors that can be used for healthy
weight management versus dieting behaviors that increase risk for eating pathology.
Key Words:  binge eating, bulimia nervosa, dietary restraint, dieting, eating disorder, obesity

Introduction pathology is crucial for three reasons. First, almost

Eating disorders (EDs) affect 13%–​15% of half of adolescent girls and young women report
females and are marked by chronicity, relapse, dis- dieting for weight control (Field, Haines, Rosner,
tress, functional impairment, and risk for future & Willett, 2009), implying that a substantial subset
obesity, depression, suicide attempts, anxiety dis- of this population may be at risk for developing an
orders, substance abuse, and mortality (Allen, ED. Second, 32% of adults in the United States and
Byrne, Oddy, & Crosby, 2013; Arcelus, Mitchell, other Western countries are obese (Flegal, Carroll,
Wales, & Nielsen, 2011; Stice, Marti, & Rohde, Kit, & Ogden, 2012) and the most common treat-
2013; Swanson, Crow, Le Grange, Swendsen, & ment for obesity is negative energy balance diets
Merikangas, 2011). Accordingly, numerous pro- (Jeffery et al., 2000). Because obesity results in over
spective studies have investigated risk factors that 2.8 million deaths worldwide annually (Flegal et al.,
predict development of eating pathology and main- 2012), it is critical to determine whether interven-
tenance factors that predict persistence of eating tionists should promote weight control dieting or
pathology to elucidate the etiologic processes that alternative treatments to reduce obesity (e.g., bariat-
give rise to these disturbances and inform the design ric surgery). Third, there is emerging evidence that
of effective prevention and treatment interventions. low-​calorie diets result in weight loss and improved
Dieting is a widely studied risk factor for eat- health (e.g., Heilbronn et  al., 2006). The goals of
ing pathology, and theorists posit that it increases this chapter are to review the empirical literature
risk for onset and persistence of binge eating and that has examined the relation of dieting to eat-
bulimia nervosa (Fairburn, 1997; Polivy & Herman, ing pathology, consider factors that may account
1985). Determining whether dieting increases for inconsistent findings, and offer suggestions
risk for development and persistence of bulimic for future research that may help to advance our

126
understanding of the relation between dieting and a state of chronic hunger, especially if such weight
bulimic pathology. loss leaves the dieter at a weight below the set-​point
weight that is defended physiologically” (p.  196).
Dieting: Definition and Descriptive The chronic hunger experienced by dieters puta-
Statistics tively increases the likelihood that they may binge
Dieting is defined as volitional and sustained eat. Polivy and Herman (1985) further propose that
restriction of caloric intake for the purposes of “dieting causes binging by promoting the adoption
weight loss or weight maintenance (Herman & of a cognitively regulated eating style” (p. 193) and
Polivy, 1975; Wadden, Brownell, & Foster, 2002; argue that a reliance on cognitive control over eating,
Wilson, 2002). Dieting must result in a negative rather than a reliance on physiological cues, leaves
energy balance for weight loss (a calorie-​ deficit dieters vulnerable to uncontrolled eating when these
diet) and a balance between caloric intake and cognitive controls are disrupted. Violation of strict
caloric expenditure for weight maintenance (a dietary rules may also result in the temporary aban-
balanced-​calorie diet). donment of dietary restriction because of the absti-
Although the above definition is straightforward, nence violation effect (Marlatt & Gordon, 1985).
dieting is heterogeneous because people report Specifically, people who violate a commitment to
using various behaviors to achieve weight control. not engage in a particular behavior (e.g., eat choco-
Data suggest that 60% to 75% of weight loss diet- late) may abandon the commitment and excessively
ers combine reduced caloric intake with increased engage in that particular behavior. It has also been
physical activity in an effort to enter a negative suggested that dieting results in depletion of trypto-
energy balance, but some dieters report potentially phan, an amino acid precursor of serotonin, which
harmful weight control behaviors, such as meal may increase the likelihood of binge eating carbohy-
skipping, and a smaller percentage report unhealthy drates to restore tryptophan levels (Kaye, Gendall, &
behaviors such as fasting, vomiting, or laxative Strober, 1998). Indeed, placing individuals on short-​
abuse for weight control purposes (Emmons, 1992; term energy-​deficit diets reduces plasma tryptophan
French, Perry, Leon, & Fulkerson, 1995; Serdula (Attenburrow et al., 2003).
et  al., 1993). Data suggest that there is also con- Researchers have also suggested that dietary
siderable variation in the duration of self-​initiated restraint may serve as a maintenance factor that
weight loss diets, ranging from less than a week perpetuates binge eating and bulimic symptoms
to 6 months, with a mode of 1 month (Emmons, once these behaviors have emerged. Fairburn
1992; French, Jeffery, & Murray, 1999; Williamson, (1997) theorized that binge eating precipitates
Serdula, Anda, Levy, & Byers, 1992). redoubled dietary efforts and use of radical weight
Heatherton, Herman, Polivy, King, and McGree control techniques, such as vomiting and laxative
(1988) argued that “dieting” should be distinguished use, which may develop into the self-​maintaining
from “dietary restraint,” with the former term refer- binge-​purge cycle.
ring to individuals who consistently engage in caloric
restriction and the latter term referring to individu- Empirical Tests of the Dieting Theory
als who engage in chronic dieting that is punctu- of Eating Pathology
ated by bouts of overeating. However, because the This chapter first reviews empirical studies exam-
two terms have been used synonymously for several ining the relation of dieting to eating pathology. We
decades (e.g., Polivy & Herman, 1985, 1992) and focused on (1) prospective studies that test whether
because extant self-​report dieting measures cannot scores on self-​report dieting measures predict future
distinguish those who consistently engage in caloric changes in binge eating or bulimic symptoms or
restriction from those who show intermittent bouts predict onset or persistence of bulimic pathology
of overeating (Stice, Fisher, & Lowe, 2004), we use and (2)  controlled experiments that examined the
these terms interchangeably. impact of directly manipulated dietary restriction
on binge eating and bulimic symptoms because
Theoretical Mechanisms for the Effects these designs permit firmer inferences. Cross-​
of Dieting on Eating Pathology sectional studies were omitted because there is no
There are several mechanisms by which dieting way to determine whether eating pathology is a
may increase risk for ED symptoms. According precursor, concomitant, or consequence of dieting.
to Polivy and Herman (1985), “Successful dieting Prospective studies can provide evidence of temporal
produces weight loss, which in turn might create precedence, which helps rule out the possibility of a

Stice, Shaw 127
reverse direction of effects and should not be influ- criteria for BN (Fairburn, Cooper, Doll, Norman,
enced by demand characteristics. Yet, an unmea- & O’Connor, 2000). Elevated dietary restraint at
sured third variable might explain any prospective baseline predicted persistence of compensatory
effect observed in a longitudinal study. Randomized behavior, but not binge eating over a 1-​year period
experiments provide more rigorous causal infer- in a community-​recruited sample of young women
ences because they establish temporal precedence who initially met criteria for full threshold or sub-
and are the most effective design available to rule threshold BN (Bohon, Stice, & Burton, 2008).
out that some omitted third variable explains the Studies have used ecological momentary assess-
observed relations. Nonetheless, experiments are ment to assess the relation between dieting and
vulnerable to bias created by demand characteristics binge eating. Engelberg, Gauvin, and Steiger
and expectancies and can have questionable internal (2005) found that self-​rated binge episodes were
validity. Although it would have been ideal to focus not preceded by elevations in dietary restraint rela-
solely on experiments that included a credible pla- tive to nonbinge days among women with full and
cebo control condition, because this would reduce subthreshhold BN. Zunker and associates (2011)
the possibility that any effects on the outcomes are found that the odds of binge eating increased on
due to demand characteristics or expectancies, only the day following dietary restriction.
one experiment in this area has included a placebo In sum, prospective studies indicate that adoles-
control condition. cent girls and young women with elevated scores
on a variety of dietary restraint measures show an
Prospective Studies of the Relation increased risk for future onset of binge eating and
of Dieting to Future Eating Pathology BN and PD, but not onset of AN or BED. These
Adolescent girls with elevated scores on dietary prospective studies were generally methodologically
restraint scales or who report frequent weight loss rigorous, in that many involved large representa-
dieting are at increased risk for future onset of tive samples, used a long follow-​up period, had low
bulimic symptoms (Field et  al., 1999; Neumark-​ attrition, and used validated structured diagnostic
Sztainer et  al., 2006; Stice & Agras, 1998; Stice, interviews to assess eating pathology. There has been
Presnell, & Spangler, 2002) and onset of threshold less support for the assertion that elevated dietary
and subthreshold bulimia nervosa (BN) and purg- restraint increases risk for persistence of bulimic
ing disorder (PD) (Killen et al., 1994, 1996; Patton, symptoms, in that dieting measures did not predict
Selzer, Coffey, Carlin, & Wolfe, 1999; Stice, Davis, persistence of binge eating in three studies, but did
Miller, & Marti, 2008). However, elevated dietary predict persistence of compensatory behaviors in
restraint scores did not predict future onset of two of the three studies.
threshold or subthreshold anorexia nervosa (AN) or Because self-​reported dieting has emerged as a
binge eating disorder (BED), despite ample power risk factor for eating pathology, numerous research-
(Stice, Gau, Rohde, & Shaw, 2016; Stice, Marti, & ers and clinicians assert that dieting is causally related
Durant, 2011). Research has also found that adoles- to BN (Fairburn, 1997; Heatherton & Polivy, 1992;
cent girls with elevated dietary restraint scores who Levine & Smolak, 2006; Neumark-​Sztainer et  al.,
report frequent dieting or self-​identify as dieters are 2006). Indeed, some have called for a moratorium
at increased risk for future onset of any ED (Allen on dieting because of the belief that it contributes to
et  al., 2009; Santonastaso, Friederici, & Favaro, eating pathology and have evaluated interventions
1999; Stice, Marti, & Durant, 2011). that decrease dietary restriction (Bacon et al., 2002;
Only three prospective studies tested whether Polivy & Herman, 1992). In addition, many ED pre-
self-​reported dieting is a maintenance factor that vention programs strongly discourage dieting (e.g.,
predicts persistence of bulimic symptoms among Smolak, Levine, & Schermer, 1998; Stewart, Carter,
individuals who initially endorse these symp- Drinkwater, Hainsworth, & Fairburn, 2001).
toms with nontreatment samples. Elevated dietary
restraint scores predicted persistence of compensa- Experimental Studies of the Relation
tory behaviors, but not binge eating, over a 9-​month of Short-​Term Caloric Restriction
period in a study of female high school students to Laboratory-​Based Eating
(Stice & Agras, 1998). Elevated dietary restraint did Only two studies have examined the effects of
not predict persistence of binge eating or compensa- experimentally manipulated acute caloric depriva-
tory behaviors over a 5-​year period in a community-​ tion on bulimic symptoms. Because both focused
recruited sample of women who initially met on individuals with EDs, these experiments are

128 Dieting
considered tests of whether acute dietary restriction biases. Moreover, because they were asked by an
is a maintenance factor for binge eating. Telch and experimenter to diet for a specified period of time,
Agras (1996) found that 6-​hour caloric deprivation this may have reduced dieters’ internal motivation
did not result in significantly greater self-​reported for restricting their intake.
binge eating among women with BN, BED, or
obesity relative to a no-​deprivation control condi- Experimental Studies of the Relation
tion during a standard multi-​item buffet. Agras and of Longer-​Term Dieting to Eating Disorder
Telch (1998) found that 14-​hour caloric depriva- Symptoms
tion, relative to a no-​deprivation control condition, Randomized trials have also examined the effects
produced increases in investigator coded binge eat- of long-​term weight loss diets on changes in binge
ing, but not self-​labeled binge eating, in women eating and bulimic symptoms. One trial found that
with BED. No lab-​based experiment has investi- assignment to a 20-​week low-​calorie weight loss
gated the effects of caloric deprivation on all of the intervention, which resulted in weight loss, pro-
symptoms of BN. duced significantly greater decreases in binge eating
Experiments that examined the effects of from pre to post among overweight women relative
caloric deprivation on caloric intake in the lab to an assessment-​ only control condition (Klem,
(rather than binge eating) have also produced Wing, Simkin-​Silverman, & Kuller, 1997). This
mixed findings. Short-​ term caloric deprivation intervention focused on making modest changes
(4–​24 hours), relative to no-​deprivation control in dietary fat and cholesterol intake and prescribed
conditions, resulted in elevated ad libitum caloric a 1,300-​or 1,500-​ calorie meal plan. However,
intake among women and men without EDs in another small experiment found that assignment
two experiments (Mauler, Hamm, Weike, & to various energy-​ deficit diets did not result in
Tuschen-​Caffier, 2006; Spiegel, Shrager, & Steller, changes in binge eating relative to a weight main-
1989), but this effect did not emerge in three other tenance comparison condition (Redman, Martin,
experiments (Hetherington, Stoner, Andersen, Williamson, & Ravussin, 2008). It is noteworthy
& Rolls, 2000; Schachter, Goldman, & Gordon, that the weight loss interventions evaluated in this
1968; Spiegel et al., 1989). Experiments have also latter trial resulted in a 10% reduction in body mass,
found that short-​term caloric deprivation produces making it the study that has investigated the most
significantly elevated caloric intake among women intensive weight loss diet to date. Uncontrolled tri-
with BN (Hetherington et al., 2000; Mauler et al., als have also found that assignment to low-​calorie
2006), but this effect did not replicate in another diet interventions that produced weight loss were
sample of women with BN or for individuals with associated with decreases in binge eating among
binge/​purge subtype of AN in the experiment con- overweight men, women, and preadolescent girls
ducted by Hetherington and associates (2000). (Epstein, Paluch, Saelens, Ernst, & Wilfley, 2001;
Assignment to longer periods of dietary restriction Foster, Wadden, Kendall, Stunkard, & Vogt, 1996;
(2  days to 8 weeks), versus no-​diet control condi- Schlundt, Hill, Sbrocco, Pope-​ Cordle, & Sharp,
tions, resulted in significantly elevated acute caloric 1992; Telch & Agras, 1993; Wadden, Foster, &
intake in young women in one experiment in two Letizia, 1994; Wardle, Waller, & Rapoport, 2001),
settings chosen to represent those in which diet- with one exception (Braet, Tanghe, De Bode,
ers often overeat (e.g., after consumption of high-​ Franckx, & VanWinckel, 2003).
calorie food and during stress; Wardle & Beales, Experimental psychopathology trials have exam-
1988), but this effect did not replicate in three other ined the effects of weight loss diet interventions on
experiments (Lowe, 1992, 1994; Lowe, Foster, bulimic symptoms. Presnell and Stice (2003) found
Kerzhnerman, Swain, & Wadden, 2001). that young women assigned to a 6-​week weight loss
In sum, experiments indicate that short-​ term diet intervention, which resulted in weight loss,
caloric deprivation did not have consistent effects on showed greater reductions in bulimic symptoms and
binge eating among ED participants or on general binge eating during the dieting intervention relative
caloric intake among ED participants and healthy to wait-​list controls. Groesz and Stice (2007) found
controls. These studies have generally been meth- that young women assigned to a 6-​week weight
odologically sound, although the moderate sample loss diet intervention or a 6-​week weight loss inter-
sizes may have limited the ability to detect small vention that prescribed consuming frequent small
effects. It is also possible that participants could meals, both of which produced weight loss, showed
have altered their eating due to social desirability greater reductions in bulimic symptoms and binge

Stice, Shaw 129
eating during the dieting intervention relative to likewise found that cognitive therapy and a low-​
assessment-​only controls. calorie weight loss intervention were similarly effec-
One randomized controlled prevention trial tive in reducing binge eating. A  third randomized
found that an obesity prevention program that trial found that obese BED patients assigned to a
promoted lasting moderate reductions in caloric very-​low-​calorie (800 kcal/​day) weight loss inter-
intake and increases in exercise to balance caloric vention showed similar reductions in binge eating
intake with expenditure during 3 weekly meet- to patients assigned to a very-​low-​calorie weight loss
ings, which significantly reduced risk for weight intervention that included CBT to reduce binge
gain and onset of obesity over 3-​year follow-​up eating at post-​treatment and 1-​year follow-​up (de
relative to assessment-​only controls and two alter- Zwaan et  al., 2005). However, Grilo and Masheb
native interventions, produced reductions in ED (2005) found that guided self-​help CBT produced
symptoms and risk for future onset of any EDs, superior reductions in binge eating compared to
relative to assessment-​only controls (Stice, Marti, a guided self-​ help weight loss intervention and
Spoor, Presnell, & Shaw, 2008; Stice, Shaw, Burton, an assessment-​only control condition, suggesting
& Wade, 2006). A second trial replicated the find- that the guided self-​help modality may be more
ing that this obesity prevention program produced effective for delivering CBT than for weight loss
weight loss that was accompanied by reductions in interventions.
ED symptoms and future onset of any EDs (Stice, A trial that randomized BN patients to stan-
Rohde, Shaw, & Marti, 2013). It is important to dard CBT, a physical exercise intervention (which
note that this is the only ED prevention program to resulted in weight loss that persisted through 18-​
produce a significant reduction in future ED onset month follow-​up), a nutritional counseling con-
in two trials. trol group, or a wait-​list control group found that
Randomized treatment trials have also exam- participants in the CBT and exercise interventions
ined the effects of weight loss dieting interventions showed improvements relative to the nutritional
among individuals with BN or binge eating distur- control and wait-​ list control conditions by 18-​
bances. Two controlled trials found that assignment month follow-​ up (Sundgot-​ Borgen, Rosenvinge,
to a low-​calorie weight loss diet (e.g., 1,200 calo- Bahr, & Schneider, 2002). The exercise intervention
ries a day), relative to waitlist control conditions, produced significantly greater reductions in bulimic
resulted in greater decreases in binge eating for symptoms at posttest and 18-​month follow-​up rela-
overweight and obese women who endorsed initial tive to CBT; 62% of the exercise intervention par-
binge eating at 6 months post-​treatment (Goodrick, ticipants no longer met diagnostic criteria for BN,
Poston, Kimball, Reeves & Foreyt, 1998; Reeves compared to 36% of the CBT participants at 18-​
et al., 2001). However, the low-​calorie interventions month follow-​up.
did not result in weight loss in these trials. Another Two additional trials evaluated interventions
trial found that assignment to a 6-​week weight loss that sought to manipulate dieting. Bacon and
diet intervention, which resulted in weight loss that associates (2002) examined the effects of assign-
persisted through 3-​ month follow-​ up, produced ing obese participants to a traditional low-​calorie
significantly greater reductions in binge eating and weight loss diet or to a nondiet intervention that
compensatory behaviors, as well as higher remission promoted increased exercise and body acceptance,
rates, through 3-​month follow-​up among women presented psychoeducational material on nutrition
with threshold or subthreshold BN than observed and healthy eating, and provided social support;
in wait-​list controls (Burton & Stice, 2006). participants in both interventions showed reduc-
Interestingly, one randomized trial that com- tions in binge eating, which did not differ across
pared a group behavioral weight loss diet inter- groups at post-​treatment. Those in the nondieting
vention to cognitive-​ behavioral therapy (CBT) intervention may have shown decreases in binge
among overweight patients with BED found that eating because this intervention provided psycho-
both interventions were associated with similar educational material on healthy eating and pro-
reductions in binge eating by 12-​month follow-​up, moted healthy weight control behaviors (exercise)
though CBT produced significantly larger reduc- or because of regression to the mean. Wadden and
tions in this outcome at posttest (Munsch et  al., associates (2004) compared two low-​calorie diets to
2007). Similar effects emerged in a second trial a nondieting intervention that promoted increased
(Grilo, Masheb, Wilson, Gueorguieva, & White, physical activity, provided psychoeducational mate-
2011). Nauta, Hospers, Kok, and Jansen (2000) rial on dieting, and used cognitive interventions to

130 Dieting
promote body acceptance; however, there was no bulimic symptoms, so it is unclear whether these
evidence of differential change across conditions in findings can provide information about whether
binge eating for obese individuals who were initially dietary restriction is related to bulimic pathology.
free of binge eating. The fact that all participants In sum, nine randomized controlled trials
were free of binge eating at baseline made it impos- found that assignment to low-​calorie weight loss
sible to detect the decreases in bulimic symptoms, or weight maintenance diet interventions resulted
which were observed in the dieting conditions of in significantly greater reductions in binge eating
other trials. Irrespectively, the dieting interventions and bulimic symptoms relative to assignment to
in both trials resulted in significant reductions in assessment-​ only control conditions. Assignment
weight relative to participants in the nondieting to a weight maintenance diet  also resulted in sig-
conditions, suggesting that the low-​calorie dieting nificantly greater reductions in bulimic symptoms
conditions resulted in a greater negative energy bal- than assignment to a placebo control condition.
ance overall than the nondieting conditions, yet did Results from several noncontrolled intervention
not provoke the increases in binge eating that would trials also indicated that assignment to weight loss
be expected based on the dietary restraint theory. diet interventions produced reductions in binge
A noteworthy aspect of these null findings was that eating and bulimic symptoms, though one study
the nondieting interventions used in these trials did not replicate this effect. Three trials found that
were based on the program developed by Polivy and low-​calorie weight loss diet interventions were as
Herman (1992). effective as CBT in reducing binge eating among
obese binge eaters. However, one study that used
Animal Studies of the Effects of Caloric a self-​help format suggested that CBT produced
Deprivation on Caloric Intake superior effects, and another trial found that an
Experiments with animals have also investigated exercise intervention that resulted in weight loss
the effects of caloric deprivation on subsequent produced greater reductions in bulimic symptoms
caloric intake. Rats randomized to extreme caloric than did CBT among patients with BN. Two tri-
deprivation conditions (in which they lost 15%–​ als that compared a low-​calorie diet to a nondieting
20% of their body mass) consume significantly intervention did not produce evidence of signifi-
more calories during ad libitum feeding immedi- cantly greater increases in bulimic symptoms in the
ately after the deprivation period than nondeprived former condition relative to the latter condition, as
control rats (e.g., Ogawa et al., 2005). Rats assigned would be expected based on the dietary restraint
to a moderate caloric restriction condition (in which theory. Most of these studies were methodologically
they lost 7%–​9% of their body mass) consumed sig- rigorous, in that these trials and experiments were
nificantly more calories in the 4 hours immediately adequately powered, had low attrition, and used
after the period of deprivation but did not show state-​of-​the-​art structured diagnostic interviews to
significantly different caloric intake during the 24 assess bulimic symptoms. However, one limitation
hours after the deprivation period relative to non- of this literature is that only one of these trials used
restricted rats (Hagan, Chandler, Wauford, Rybak, a placebo control condition. It is noteworthy that
& Oswald, 2003). Several experiments found that some of these trials imply that weight loss dieting
rats assigned to cycles of caloric restriction and may represent an efficacious treatment for BN and
refeeding did not show significantly different ad that weight maintenance dieting may represent an
libitum caloric intake after refeeding relative to efficacious prevention intervention for EDs.
controls (Boggiano et al., 2005; Hagan et al., 2003;
Hagan et  al., 2002). Thus, there is some evidence Incompatible Findings
that severe caloric restriction results in elevated On the one hand, prospective studies have
caloric intake immediately after the deprivation provided evidence that individuals with elevated
period. Although experiments with animals allow scores on various dietary restraint measures show
greater experimental control over caloric restriction an increased risk for future onset of and increases
manipulations and are immune to demand charac- in binge eating, bulimic symptoms, and BN and
teristics, they have questionable generalizability to PD, which appears to support the dietary restraint
humans, given that the animal studies involve invol- model of eating pathology, yet experimental studies
untary food restriction, which may be very different produced findings that are incompatible with the
from the voluntary dietary restriction practiced by prospective studies. Although the experiments that
humans. In addition, these studies did not assess examined the effects of acute caloric restriction on

Stice, Shaw 131
laboratory-​assessed binge eating produced mixed isolate the effect of the one manipulated variable
effects that were difficult to interpret, eight con- (because potential confounds should be uncor-
trolled trials and experiments produced effects related with treatment condition). Although
indicating that assignment to longer-​term weight random assignment can fail in creating initially
loss or weight maintenance diets resulted in sig- equivalent groups, particularly if the cell sizes
nificantly greater reductions in binge eating and are small, this is the best available tool to rule
bulimic symptoms than assignment to assessment-​ out third variable confounds. Thus, the positive
only control conditions, and one weight mainte- relation of self-​reported dieting to increases in
nance diet produced greater reductions in bulimic eating pathology observed in prospective studies
symptoms than a placebo control condition. Most may have emerged because some third variable
of these trials and experiments verified that the increases the risk for both variables. One trou-
participants showed weight loss or weight main- bling aspect of this explanation for the inconsis-
tenance, confirming that dietary restraint was suc- tent findings, however, is that the findings from
cessfully manipulated. These latter findings appear the prospective and experimental studies are
incompatible with the dietary restraint model consistently in the opposite direction. Typically
of eating pathology. These contradictory find- when prospective effects are due to an omitted
ings from prospective and experimental studies confounding variable, experimental studies that
are troubling because they have opposing public manipulate the independent variable do not
health implications. If dieting increases bulimic affect the dependent variable.
symptoms, prevention programs and treatment This potential explanation for the inconsis-
interventions should attempt to decrease dieting tent findings has several implications for future
and alternative nondieting treatments for obesity research. First, it is vital to use randomized
should be developed. In contrast, if dieting reduces experiments to investigate the relation of dietary
bulimic symptoms, prevention programs and treat- restraint to eating pathology, because this design
ment interventions should help individuals diet permits greater inferential confidence than pro-
more effectively, which should yield positive effects spective studies. Researchers should maximize
for both eating pathology and obesity. Thus, it is the ecological validity of these experiments and
critical to understand why these inconsistent find- include placebo control conditions and objective
ings emerged. measures of ED symptoms, which should reduce
There are several possible explanations for these the risk that demand characteristics or expectan-
contradictory findings. In what follows, we discuss cies account for the apparent intervention effects
various explanations, review supporting evidence, and to include manipulation checks. Second, it
consider the implications of each explanation, and is also vital to search for potential third variable
propose studies that might help evaluate the verac- confounds that explain the relation between self-​
ity of these different explanations. reported dieting and future increases in bulimic
symptoms observed in the prospective studies.
Prospective Studies Are More It is also important to use objective measures of
Vulnerable to Confounding Variables potential third variable confounds, and to then
Than Experiments manipulate potential confounds in randomized
One possible explanation for the contradic- experiments to confirm their causal relation to
tory findings is that these two types of research dieting and ED symptoms. Third, this interpreta-
designs differ in their inferential power. The tion suggests that researchers should use random-
major weakness of prospective studies is that an ized experiments to confirm the causal status of
omitted third variable could account for any pro- other putative risk factors for eating pathology
spective effect observed in a longitudinal study. because it appears that an exclusive reliance on
That is, some confound may cause both elevated prospective studies may produce questionable
dieting and elevated bulimic symptoms. In con- inferences. For example, experimental trials that
trast, randomized experiments can rule out third have evaluated interventions that reduce thin-​
variable alternative explanations. The reason that ideal internalization, body dissatisfaction, and
researchers randomly assign participants to con- negative affect have found that bulimic symptoms
ditions is to create groups that are equivalent on are also reduced (Bearman, Stice, & Chase, 2003;
all potential confounds (known or unknown), Burton, Stice, Bearman, & Rohde, 2007; Stice,
which theoretically allows the investigator to Presnell, Gau, & Shaw, 2007).

132 Dieting
Dietary Restriction Interventions May Be asked to engage in their typical weight loss dieting
Unrepresentative of Real-​World Dieting behaviors for 1 month or to a condition in which
Another possible explanation for the inconsis- they were asked to refrain from engaging in their
tent results from the prospective and experimental usual weight loss dieting behaviors for 1  month
studies is that the weight loss diets evaluated in the (Presnell, Stice, & Tristan, 2007). A  manipulation
experiments involve more extreme dietary restric- check confirmed that participants in the dieting as
tion than is typical of real-​world weight loss diet- usual condition reported significantly more days of
ing. That is, perhaps real-​world dietary restriction is dieting during the 4-​week period than those in the
simply less effective than most weight loss interven- nondieting control condition, which corresponded
tions, which is why the two types of studies produce to a large effect (r = .74). Unexpectedly, participants
different effects on change in ED symptoms. assigned to the usual weight loss dieting condition
Although this explanation holds intuitive appeal, showed no weight loss, but participants assigned
it is not a satisfactory account for the inconsistent to the no dieting condition gained a significant
findings for several reasons. First, if self-​initiated amount of weight. These findings imply that typical
diets are simply less effective than the weight loss weight loss dieting may not result in weight loss and
diets evaluated in the experiments, then the effects further that most dieters may simply be transiently
from the prospective studies should be smaller than curbing an overeating tendency when they are diet-
the effects from the experiments involving pre- ing on their own in the real world. There were no
scribed diets, but in the same direction; yet, the significant differences in change in bulimic symp-
effects from the prospective studies are consistently toms over time across conditions, suggesting that
in the opposite direction relative to those from the ineffective weight loss dieting (or effective weight
experimental studies. Second, both lower intensity maintenance dieting) has no impact on bulimic
weight maintenance interventions (Klem et  al., behaviors. This is an important point because
1997; Stice, Shaw, Burton, & Wade, 2006) and it has been argued that it is unsuccessful dieting,
higher intensity weight loss interventions (Groesz rather than successful dieting, that increases risk for
& Stice, 2007; Presnell & Stice, 2003) produce bulimic pathology (Heatherton et al., 1988). These
significant reductions in binge eating and bulimic results provide evidence that weight loss diets evalu-
symptoms. This pattern of findings implies that ated in the experimental trials may not be represen-
even more modest weight control diets reduce ED tative of real-​world dieting, but provide no support
symptoms. Third, the weight loss observed in the for the assertion that real-​world dieting results in
experimental trials that evaluated prescribed weight increased bulimic symptoms (despite adequate
loss diets was small; participants lost an average of power), as suggested by the prospective studies that
0.14 kg per week (0.3 lbs). Fourth, the weight loss are ostensibly studying real-​world dieting.
and weight maintenance diets from the experiments The ecological validity of the weight loss and
were similar in duration (mode = 1.5 months, range weight maintenance diet interventions evaluated in
1–​18  months) to the reported duration of real-​ the experimental trials are also limited by the fact
world weight loss diets (mode  =  1  month, range that participants typically meet in groups with pro-
1 week–​6  months; Emmons, 1992; French et  al., fessional facilitators who guide the recommended
1999; Williamson et  al., 1992). These consider- reductions in caloric intake and increases in phys-
ations imply that it is unlikely that the experimental ical activity. Although some interventions involved
studies produced different findings from the pro- as little as 3 hours of contact, this is not something
spective studies because the weight loss interven- typical of most participants who engage in weight
tions evaluated in the former were more extreme loss or weight maintenance diets outside of treat-
than real-​world weight loss diets. ment settings. The evidence that participants typi-
Nonetheless, we conducted an experiment that cally show reductions in binge eating in behavioral
sought to provide a more ecologically valid test of weight loss interventions involving professionally-​
whether real-​world weight loss dieting results in led groups (Munsch et  al., 2007; Nauta et  al.,
decreased bulimic symptoms. We focused on young 2000), but not when participants use a self-​help
women who reported at least intermittent dieting book with similar content (Grilo & Masheb, 2005),
in the past year because we were only interested in implies that these groups may play a role in reducing
generalizing the results to women who voluntarily binge eating. In addition, the fact that participants
engage in weight loss dieting. We randomly assigned in weight loss interventions from the experimen-
participants to a condition in which they were tal trials typically showed reductions in bulimic

Stice, Shaw 133
symptoms, whereas those who dieted as usual in the & Pestone, 1992; Stice, Palmrose, & Burger, 2015).
naturalistic dieting experiment did not, might also People might also provide biased reports of dieting
be interpreted as providing evidence for the impor- behaviors. Further, studies have found that individ-
tance of contact with supportive professionals. uals with high scores on dietary restraint scales gain
In sum, there is preliminary evidence that the more weight over time than people with low scores
weight loss diet interventions evaluated in prior (Klesges, Isbell, & Klesges, 1992; Klesges, Klem, &
experiments are not representative of real-​ world Bene, 1989; Stice, Cameron, Killen, Hayward, &
weight loss dieting, in that the findings from Presnell Taylor, 1999; Stice, Presnell, Shaw, & Rohde, 2005;
et  al. (2007) suggest that most real-​world dieters Tanofsky-​Kraff et al., 2007). Similarly, there is evi-
may be transiently curbing an overeating tendency dence that self-​identified weight loss dieters show
when they diet for weight loss purposes, but do not elevated rates of weight gain and onset of obesity
typically experience weight loss. However, there relative to those who do not identify as weight
was no support for the assertion that this explains loss dieters (French, Jeffery, Forster, et  al., 1994;
the inconsistent findings that have emerged from Neumark-​Sztainer et al., 2006). These data suggest
prospective and experimental studies, as ineffec- that dieting measures are not identifying individuals
tive real-​world weight loss dieting did not result in who are engaging in dietary restriction.
increased bulimic symptoms. It will be vital to rep- We conducted a series of unobtrusive observa-
licate the experimental evidence that most people tional studies that examined the validity of dietary
engaging in real-​world weight loss dieting do not restraint scales. We investigated the Restraint
achieve the negative energy balance necessary for Scale (RS; Polivy et al. 1978), Three Factor Eating
weight loss and that they may be temporarily curb- Questionnaire Restraint scale (TFEQ-​R; Stunkard
ing an overeating tendency because these two find- & Messick, 1985), Dutch Restrained Eating Scale
ings explain several vexing patterns of findings that (DRES; van Strien, Frijters, van Staveren, Defares, &
have emerged in this literature. It would be useful Deurenberg, 1986), Eating Disorder Examination-​
for future experiments that manipulate real-​world Questionnaire-​Restraint scale (EDEQ-​R; Fairburn
weight loss dieting to involve adolescent girls, as this & Beglin, 1994), and the Dietary Intent Scale (DIS;
is typically when bulimic symptoms emerge. These Stice, 1998). We used caloric intake as the criterion
types of studies may also afford a unique approach because the original validity studies concluded that
for investigating whether particular dieting behav- these scales were valid measures of dietary restriction
iors practiced in the real world are more effective because they correlated inversely with self-​reported
than others in producing weight loss and whether caloric intake (e.g., French, Jeffery, & Wing, 1994;
particular dieting behaviors predict increases in Neumark-​Sztainer, Jeffery, & French, 1997; van
bulimic symptoms. Strien et al., 1986).
Another factor that undermines the ecologi- We conducted multiple studies that varied in
cal validity of the weight loss diet interventions is food types consumed, settings examined, and pop-
that they typically involved group meetings with ulations studied. Commonly used dietary restraint
supportive professionals. It would be useful for scales were not inversely correlated with caloric
future controlled trials to investigate the impact of intake, as suggested by the original validity stud-
group meetings with professionals on change in ED ies that relied on self-​reported caloric intake. These
symptoms among those assigned to weight loss diet dieting scales only showed weak correlations with
interventions. caloric intake during eating episodes across four
studies and virtually none of these correlations were
Researchers May Have Used Invalid statistically significant (mean r = –​.07, range: –​.34
Measures of Dietary Restraint to .20; Stice et  al., 2004). Other independent
A third possible explanation for the conflicting studies have likewise found that dietary restraint
findings is that the prospective studies that found scales did not correlate inversely with objectively
dieting predicted future onset of bulimic symptoms measured caloric intake during single eating epi-
and BN may have used invalid dieting measures. sodes (Hetherington et  al., 2000; Jansen, 1996;
This is suggested by evidence that people routinely Ouwens, van Strien, & van der Staak, 2003; Sysko,
under-​report caloric intake and that under-​reporting Walsh, & Wilson, 2007; van Strien, Cleven, &
is greater for those with elevated scores on dieting Schippers, 2000).
scales and elevated body weight (Bandini, Schoeller, Research has also examined multiple objective
Cyr, & Dietz, 1990; Lichtman, Pisarska, Berman, assessments of caloric intake, which should provide

134 Dieting
a more stable index of eating behaviors. Rolls and with low scores. Similar findings emerged in stud-
colleagues (1997) found that normal weight adults ies that used objective measures of caloric intake
with high versus low scores on the TFEQ-​R did not during single eating episodes, intake during mul-
show differences in caloric intake during three meals tiple eating episodes, intake over a 2-​week obser-
and a snack consumed during a 20-​hour monitor- vational period, or intake at workplace cafeterias
ing period in the laboratory. Jansen and associates over a 3-​month period. These data imply that die-
(2003) found the Eating Disorder Examination-​ tary restraint scales do not assess dietary restric-
Restraint scale (EDE-​R, Fairburn & Cooper, 1993) tion as suggested by validation studies that relied
did not correlate with observed caloric intake dur- on self-​reported caloric intake (e.g., French, Jeffery,
ing three separate taste tests involving snack foods & Wing, 1994; Neumark-​Sztainer et al., 1997; van
for normal weight preadolescents. Martin and asso- Strien et  al., 1986) and the item content of these
ciates (2005) found that the TFEQ-​restraint scale scales. These findings are troubling because they
did not correlate with observed caloric intake dur- suggest that virtually all studies on dietary restraint
ing four healthy meals that were eaten in a labora- have used invalid dieting measures, which should
tory setting by normal weight young women over a be considered when interpreting this literature.
1-​month period. Sysko, Walsh, Schebendach, and These findings provide no support for Heatherton
Wilson (2005) found that the TFEQ-​restraint scale, and associates’ (1988) suggestion that the Restraint
DIS, EDEQ-​restraint scale, and EDE-​restraint scale Scale assesses unsuccessful dietary restriction and
did not correlate with observed caloric intake of a that other dietary restraint measures, such as the
yogurt shake eaten in a lab setting by young women Dutch Restrained Eating scale (van Strien et  al.,
with AN or normal weight control women dur- 1986), and the TFEQ restraint scale (Stunkard &
ing two separate sessions. Stice, Cooper, Schoeller, Messick, 1985), assess successful dieting; the valid-
Tappe, and Lowe (2007) found that the TFEQ-​ ity data suggest that none of these widely used
restraint scale did not correlate with objectively measures identify successful dieters.
measured caloric intake during lunch meals con- One explanation for the evidence that dietary
sumed in work cafeterias over a 3-​month period. restraint scales show significant inverse correlations
Still other studies used doubly labeled water with self-​reported caloric intake, but nonsignificant
(DLW) to assess longer-​term caloric intake. DLW relations with objectively measured caloric intake is
uses isotopic tracers to measure total carbon dioxide social desirability bias. Because weight management
production, which can be used to generate accu- is socially valued in Western culture and obesity is
rate estimates of total caloric intake over a 2—​week stigmatized, people may over-​report dieting behav-
period. Bathalon et  al. (2000) found that normal iors. This interpretation converges with evidence
weight women with high scores on the TFEQ-​ that under-​reporting of caloric intake is greater for
restraint scale did not consume fewer calories over overweight versus lean individuals (Prentice et  al.,
a 2-​week period than weight-​matched women with 1986; Stice, Palmrose, & Burger, 2015). However,
low scores on this scale. Tuschl, Platte, Laessle, it is also likely that honest forgetting contributes to
Stichler, and Pirke (1990) found a nonsignificant underreporting of caloric intake.
correlation between the TFEQ-​restraint scale and These validity findings provide a compelling
DLW assessments of total caloric intake over a 2-​ explanation for why findings from the prospec-
week period. Stice, Cooper, et  al. (2007) found tive and experimental studies have been inconsis-
that the TFEQ-​restraint scale did not correlate with tent; although the experiments were examining the
DLW assessments of total caloric intake over a 2-​ effects of confirmed calorie-​deficit diets, the pro-
week period in a sample of overweight women or spective studies were examining individuals who do
in a sample of normal-​weight women. Stice, Sysko, not appear to be on a caloric-​deficit diet necessary
Roberto, C.  A., and Allison (2010) found that for weight loss. That is, the experiments appear to
the Restraint Scale, TFEQ-​restraint scale, Dutch have been examining individuals on energy-​deficit
Restrained Eating Scale, and the Dietary Intent weight loss diets, whereas the prospective studies
Scale did not correlate with total caloric intake were not.
over a 2-​week period in a sample of average-​to-​ One puzzling aspect of this explanation, how-
overweight young women. ever, is why individuals who are either unable to
Collectively, these findings indicate that indi- achieve a true caloric-​ deficit diet, who want to
viduals with elevated scores on dietary restraint give the impression that they are on such a diet, or
measures do not consume less food than individuals commonly forget about foods they consume, are at

Stice, Shaw 135
increased risk for onset of binge eating, BN, and scores often increase when individuals are placed
PD. If the dietary restraint measures are invalid, on low-​calorie diets relative to controls who are
such that individuals with high versus low dietary not placed on weight loss diets (e.g., Groesz &
restraint scores do not eat less, it seems that the pro- Stice, 2007; Williamson et al., 2007). Others have
spective studies should simply have produced null suggested that many dieters are eating less than they
findings with regard to the relation between dietary desire rather than restricting their dietary intake suf-
restraint scales and future bulimic pathology. ficiently to produce a negative energy balance and
However, the fact that the prospective studies have therefore only perceive that they are on a weight
observed positive relations between initial scores loss diet (Lowe & Levine, 2005). The evidence that
on dietary restraint measures and future increases individuals with elevated dieting scores are more
in bulimic pathology implies that these scales assess likely to gain weight over time relative to indi-
a latent construct, other than energy-​deficit diet- viduals with lower dietary restraint scores (French,
ing, which increases the risk for future onset of Jeffery, Forster, et  al., 1994; Klesges et  al., 1992;
bulimic pathology. That is, although it is tempting Stice, Cameron, et al., 1999; Tanofsky-​Kraff et al.,
to dismiss the findings from the prospective studies 2007) suggests that they are not only unsuccessful
because they appear to have used invalid measures at reducing their caloric intake below their energy
of dietary restriction, one fact remains: These scales needs on a sustained basis, but are also unable to
have shown predictive validity for future develop- consistently avoid consumption beyond their
ment of bulimic pathology. energy needs and therefore gain weight over time
This analysis suggests that self-​reported dietary (Lowe & Levine, 2005).
restriction is a proxy risk factor for bulimic pathol-
ogy, but that the nature of the true latent construct Only Certain Dietary Behaviors Increase
that is tapped by these scales is unclear. A  proxy Risk for Bulimic Pathology
risk factor is a variable that predicts a pathological Another possible explanation for the conflicting
outcome not because it has any causal relation to findings is that the weight loss interventions evalu-
the development of pathology, but because it corre- ated in the experiments promote healthy dietary
lates with a true causal risk factor for the pathology behaviors, but that it is unhealthy dietary behaviors,
(Kraemer, Stice, Kazdin, Offord, & Kupfer, 2001). which may be more common in real-​world weight
Thus, we argue that a key research priority will be loss diets, that lead to bulimic pathology onset. That
to elucidate the latent factor, or factors, tapped by is, if weight loss efforts outside of treatment settings
dietary restraint scales, which increase risk for onset typically involve unhealthy weight control behav-
of bulimic pathology. iors, this might explain why self-​reported dietary
One other intriguing implication of the valid- restriction that does not occur in the context of
ity findings is that dietary restraint scales may weight loss interventions increases risk for eating
assess relative dietary restriction rather than abso- pathology onset.
lute dietary restriction. That is, these scales may be One way to explore this possibility is to sys-
identifying people who are curbing an overeating tematically manipulate suspected unhealthy
tendency, but who are not actually achieving the weight control behaviors experimentally because
negative energy balance necessary for weight loss. this approach would provide a rigorous test of
Because these individuals are eating less than they whether these behaviors are related to bulimic
normally eat or less then they desire, they may symptoms. The National Task Force on the
perceive this relative restriction as dietary restraint Prevention and Treatment of Obesity (2000)
despite the fact that they are not achieving the categorized meal skipping as an unhealthy die-
negative energy balance necessary for weight loss. tary technique because it is associated with poor
This interpretation is consistent with the evidence nutritional intake and increased consumption
that (1)  intermittent dieters temporarily arrest a of higher calorie foods at subsequent feedings
weight gain trajectory while they are attempting (de Castro & Elmore, 1988; Morgan, Zabik, &
to engage in a weight loss diet, but do not show Stampley, 1986), and may increase risk for binge
weight loss (Presnell et al., 2007); (2) individuals eating because it results in greater reward from
with elevated dietary restraint scores consumed eating and increases attention to food (Placanica,
more calories than those with low dietary restraint Faunce, & Job, 2002). Experiments confirm that
scores in one study, but did not feel that they had enforced periods of caloric deprivation result in
overeaten (Jansen, 1996); and (3) dietary restraint greater reinforcement value of food, as assessed by

136 Dieting
operant tasks that measure how hard participants and increases risk for bulimic pathology (Stice,
will work to earn food and as assessed by caloric Davis, et  al., 2008). Given that full threshold
intake (Epstein, Truesdale, Wojcik, Paluch, & BN emerges in only about 2% of young women,
Raynor, 2003; Raynor & Epstein, 2003). Acute but that 40% to 60% of adolescent girls report
caloric deprivation also increases brain reward engaging in weight loss dieting, some particularly
region response to palatable food cues, antici- unhealthy weight control behavior practiced by
pated receipt, and actual receipt (Fuhrer, Zysset, a small subset of dieters, such as fasting, could
& Stumvoll, 2008; Goldstone et al., 2009; Leidy account for the emergence of bulimic symptoms.
et  al., 2011; Stice, Burger, & Yokum, 2013; Animal studies suggest that marked caloric restric-
Uher, Treasure, Heining, Brammer, & Campbell, tion increases risk for subsequent overeating (e.g.,
2006). Acute caloric deprivation reduces mRNA Ogawa et al., 2005). In support of this interpreta-
for dopamine transporter and extracellular dopa- tion, fasting predicted future onset of binge eating,
mine (Patterson et al., 1998; Pothos, Hernandez, BN, and PD (Stice, Davis, et al., 2008; Stice, Gau,
& Hoebel, 1995), which may explain why dieting et al., 2016).
increases the reinforcing value of food. Data indi- Additional prospective and experimental tests
cate that 35% of adult dieters and 50% of adoles- of the effects of fasting are warranted to resolve
cent dieters report skipping meals for weight loss the nature of the relation between dieting and
purposes (Emmons, 1992; French et  al., 1999; bulimic pathology. First, researchers should pro-
Wardle, Griffith, Johnson, & Rapoport, 2000), spectively test whether self-​reported fasting and
suggesting that this may be the most common other unhealthy weight loss behaviors that result
unhealthy weight control behavior. in documented weight loss increase risk for future
One study that manipulated meal skipping ran- increases in binge eating and bulimic symptoms.
domly assigned obese women to a 12-​week low-​ Second, this account would predict that among
calorie diet that either prescribed consumption of those in the dieting as usual condition of natu-
three meals a day or to a 12-​week low-​calorie diet ralistic dieting experiments (e.g., Presnell et  al.,
that that prescribed consumption of two meals a 2007), the subset of participants that showed
day (Schlundt et al., 1992). A manipulation check marked weight loss should show greater pre-​to-​
indicated that participants in the meal skipping post increases in binge eating than those who do
condition reported eating significantly fewer meals, not show weight loss. More generally, the natural-
but there was no effect on change in binge eating istic dieting experimental paradigm offers a useful
or body mass over the 12-​week study. A  second way of testing whether real-​world dieting behav-
study manipulated meal frequency by randomizing iors predict increases in binge eating and bulimic
young women to a weight loss diet intervention that symptoms. Because prospective studies are vulner-
encouraged consumption of many small meals (4–​ able to third-​variable alternative explanations, it is
5) throughout the day or to a standard weight loss important to examine the effects of experimentally
intervention, in which it was assumed that many manipulating any potentially unhealthy weight
participants would skip meals based on prior find- control behaviors suggested by the prospective
ings (Groesz & Stice, 2007). Participants were told analyses, such as fasting (preferably by decreasing
to consume 1,200 calories a day in both conditions. these behaviors).
A manipulation check indicated that participants in In sum, there is no experimental support for the
the many small meals condition consumed signif- supposition that unhealthy dieting, characterized by
icantly more meals per day on average (3.8 meals meal skipping or frequency, increases risk for binge
per day) than participants in the latter condition eating or bulimic symptoms. However, only two
(2.8 meals per day), which accounted for 42% of experiments have manipulated these behaviors, and
the variance in meal frequency. However, there no experiments have investigated the effects of other
was no effect on change on binge eating, bulimic unhealthy weight control behaviors, such as fasting.
symptoms, or weight over the 6-​week intervention Thus, it is premature to rule out this possible expla-
period. nation for the inconsistent findings from prospec-
Another unhealthy weight control method tive and experimental studies regarding the impact
is fasting (skipping two or more meals in a row of dieting on bulimic symptoms. Future prospective
for weight loss purposes). A small subset of indi- and experimental studies should examine the rela-
viduals with elevated dietary restraint scales may tion of fasting and other extreme weight loss behav-
engage in fasting that produces severe weight loss iors on bulimic symptoms.

Stice, Shaw 137
Reductions in Bulimic Symptoms contributes to weight gain. Fourth, not all experi-
in Experimental Trials Are Due to Demand ments in which demand characteristics would
Characteristics be possible produced reductions in binge eating
Demand characteristics could also explain the or bulimic symptoms (e.g., Presnell et  al., 2007).
inconsistency in the findings from prospective and Fifth, the fact that 71% of ED prevention programs
experimental studies. Participants in the experi- evaluated to date did not result in significant reduc-
ments evaluating weight loss and weight mainte- tions in ED symptoms relative to assessment-​only
nance diets may simply report reductions in binge control conditions implies that demand character-
eating and bulimic symptoms because they feel istics alone are not sufficient to produce reductions
that they are expected to show reductions in eat- in bulimic symptoms (Stice, Shaw, & Marti, 2007).
ing, including overeating. Methodologists have Sixth, one weight maintenance diet produced sig-
noted that participants in the intervention condi- nificantly greater reductions in bulimic symptoms
tion of randomized trials may overreport changes than an active placebo control group (Stice et  al.,
in the target behavior in the expected direction 2006). Finally, it appears that perhaps only individ-
(Baranowski, Klesges, Cullen, & Himes, 2004). uals with BED show large responses to a pill placebo
For example, one hypertension management trial condition (Pearlstein et al., 2003), as there was little
found that intervention participants underreported evidence of a placebo response among patients with
sodium intake relative to the underreporting of BN (Carter et al., 2003; Mitchell et al., 1990). In
controls, as measured by biological assays of sodium addition, because the former trial did not include an
intake (Espeland et al., 2001). A related explanation assessment-​only control condition, it is impossible
is that the reported reductions in bulimic symp- to parse the effects of a placebo response from the
toms are due to participant expectancies (i.e., are a effects of regression to the mean. Nonetheless, this
placebo response). Pearlstein and associates (2003) possible alternative explanation for the inconsistent
found a remission rate from binge eating of 50% findings of the prospective versus the experimental
over a 3-​month period among patients assigned to studies cannot be ruled out. It will be important for
a pill placebo condition of a randomized trial of a future experiments to use collateral reports of ED
pharmacologic treatment for BED. However, one symptoms, objective measures of binge eating and
study found that patients with BN showed only compensatory behaviors, and placebo control con-
a slight reduction in binge eating and compensa- ditions to provide estimates of intervention effects
tory behaviors in response to pill placebo treatment that are not influenced by placebo response.
(Mitchell et  al., 1990), and another found that a
placebo self-​ help intervention did not result in Implications Regarding Possible
greater reductions in binge eating or compensatory Explanations for the Inconsistent Findings
behaviors among patients with BN relative to wait-​ Several interesting findings emerged from our
list controls (Carter et al., 2003). review of studies addressing possible explanations
Although this explanation is important to con- for the inconsistent findings regarding the relation
sider, it seems untenable for several reasons. First, of dietary restriction to bulimic symptoms from
many of the interventions evaluated in the experi- prospective and experimental studies. First, given
ments and trials were weight loss interventions that the evidence that the dietary restraint measures
were not portrayed as likely to affect eating disor- used in the literature appear to be invalid measures
dered symptoms. Second, this account does not of dietary restriction, it is not surprising that stud-
explain why participants on many of these trials ies using these scales find different effects relative to
reported reductions in the compensatory behaviors experiments that investigate the effects of confirmed
(Burton & Stice, 2006; Groesz & Stice, 2007). It weight loss diets. However, it is unlikely that this
is our impression that these weight loss or weight explanation accounts for the fact that the prospec-
maintenance interventions would not produce tive and experimental findings are in the opposite
demand characteristics for reductions in compen- direction. If self-​report dietary restraint scales were
satory behaviors, particularly fasting and exces- simply invalid measures of actual dietary restriction
sive exercise. Third, most of the experiments that and do not identify people who are engaging in
produced reductions in binge eating and bulimic calorie-​deficit diets necessary for weight loss, then
symptoms reduced objectively measured body mass, the prospective studies should have simply pro-
which partially validates that participants showed duced null effects rather than consistently observed
actual reductions in binge eating since binge eating positive relations between elevated dietary restraint

138 Dieting
scores and risk for future onset of bulimic symp- It logically follows that the only way that experi-
toms. The fact that dietary restraint scales have pre- mental studies of weight loss diets can produce
dictive validity for future bulimic symptom onset effects that are opposite of those observed in the
suggests that these scales assess some unidentified prospective studies is that the weight loss diet inter-
third variable that truly increases risk for bulimic ventions directly reduce the omitted third variable.
symptoms. That is, dietary restraint scales seem That is, if it is necessary to reduce an overeating ten-
to identify a population at high risk for bulimic dency to decrease bulimic symptoms, the omitted
pathology because they tap some characteristic that third variable may be a tendency to overeat, which
is causally related to the development of this eating could be the third variable that explains the effects
disturbance. emerging from prospective studies. We have posited
Second, there was evidence that real-​world weight that individuals with a tendency toward overeating
loss dieting is not as effective as weight loss inter- may attempt to curb this through dieting because of
ventions where participants meet with weight loss undesired weight gain and that this overeating ten-
professionals. Although these findings require repli- dency also increases risk for eventual onset of binge
cation, they also may explain why prospective stud- eating and bulimic symptoms (Stice, Cameron,
ies examining real-​world dieting produce different et al., 1999). Others have also proposed that dietary
effects than experiments investigating the effects of restraint scales identify people with a tendency to
weight loss diet interventions; the former do not pro- overeat, which may increase risk for BN (Lowe &
duce weight loss, whereas the latter do. Again, how- Kral, 2006; van Strien et al., 2000). Based on the
ever, it appears that this explanation cannot account present analysis, we propose this as the most logical
for the fact that the prospective and experimental interpretation of the results that were reviewed to
studies produce effects that are in the opposite direc- explain the inconsistent findings from the prospec-
tion. If real-​world dieting is less effective than weight tive and experimental studies. Consistent with this
loss diet interventions, the prospective studies should interpretation, overeating predicts future onset of
have produced null effects or effects that were smaller, BN, BED, and PD, producing the largest effect of
but in the same direction, as those emerging from any of the other 12 other risk factors examined for
the experiments. The fact that prospective studies BN and BED (Stice et al., 2016).
indicate that individuals who perceive themselves as The working hypothesis that dietary restraint
engaging in weight loss dieting are at elevated risk for scales are a proxy risk factor for onset of bulimic
onset of bulimic symptoms also suggests that these pathology because these scales identify individuals
scales tap some third variable that is causally related with an overeating tendency could explain several
to the development of eating pathology. perplexing findings. This account might explain
Third, an analysis of the inferential power of pro- why weight loss diet interventions that result in a
spective and experimental studies also could explain documented energy deficit diet successfully reduce
the inconsistent findings. Although experiments bulimic symptoms (Burton & Stice, 2006; Groesz
may have limited ecological validity, they are more & Stice, 2007; Klem et al., 1997; Presnell & Stice,
immune to third variable alternative explanations, 2003) and why individuals with elevated scores on
whereas it is always possible that relations found dietary restraint scales often gain more weight over
in prospective studies can be explained by some time than those with lower scores (Klesges et  al.,
unmeasured third variable. Thus, it is possible that 1989, 1992; Stice, Cameron, et al., 1999; Stice et al.,
some omitted third variable that causes both diet- 2005; Tanofsky-​Kraff et  al., 2007). This account
ing and eventual onset of bulimic pathology may also may explain why individuals who typically
explain the relation between initial scores on dieting engage in weight loss dieting tend to gain weight
measures and future increases in bulimic symptoms. over time when they are not actively engaging in
Given that the considerations discussed herein what they perceive as weight loss dieting (Presnell
suggest that dietary restraint scales identify people et al., 2007) and why individuals with high scores
with some characteristic that increases risk for on dietary restraint measures typically weigh more
bulimic symptom onset, we think it may be fruitful than those with low scores (e.g., Nederkoorn &
to search for this third variable. Identifying this vari- Jansen, 2002; Roefs, Herman, MacLeod, Smulders,
able would help advance our understanding regard- & Jansen, 2005).
ing factors that cause onset of bulimic pathology The assertion that dietary restraint scales iden-
and in the development of preventive and treatment tify people with overeating tendency also accords
interventions for this eating pathology. with several other findings. First, this may explain

Stice, Shaw 139
why such a wide variety of experimental manipu- Several studies appear to be consistent with the
lations can trigger overeating among individuals thesis that individuals with BN show heightened
with elevated scores on dietary restraint scales. reward sensitivity. Women with bulimic symptoms
Experiments have found that these individu- report greater reward sensitivity in general on sur-
als overeat in response to negative mood induc- veys relative to controls (Davis & Woodside, 2002;
tions (Schotte, Cools, & McNally, 1990), positive Kane, Loxton, Staiger, & Dawe, 2004; Loxton &
mood inductions (Cools, Schotte, & McNally, Dawe, 2006, 2007; Nederkoorn, van Eijs, & Jansen,
1992), stress/​threat inductions (Polivy, Herman, & 2004). Relative to controls, individuals with BN pre-
McFarlane, 1994), listening to the radio (Bellisle fer sweeter tastes (Franko, Wolfe, & Jimerson, 1994)
& Dalix, 2001), performing cognitively distract- and higher-​fat foods (Sunday & Halmi, 1990), con-
ing tasks (Hofmann, Rauch, & Gawronski, 2007), sume more artificially sweetened solution in a mod-
consuming alcohol (Polivy & Herman, 1976), con- ified sham-​feeding paradigm and report elevated
suming high-​ calorie food (Jansen, Merckelbach, intake of low-​ calorie artificially sweetened foods
Oosterlaan, Tuiten, & Van den Hout, 1988), con- (Klein, Boudreau, Devlin, & Walsh, 2005, 2006),
suming food that is perceived to be high-​calorie and engage in sham-​feeding types of behaviors, such
(Spencer & Fremouw, 1979), the smell of food as chewing and spitting out food (Eckern, Stevens,
(Fedoroff, Polivy, & Herman, 1997), simply think- & Mitchell, 1999; Guarda et  al., 2004; Kovacs,
ing of food (Fedoroff et al., 1997), or to exposure Mahon, & Palmer, 2002). Laboratory studies have
to dieting commercials (Strauss, Doyle, & Kreipe, found that individuals with versus without BN con-
1994). This account is also consistent with evidence tinue to eat after reaching satiety, report persistent
that individuals with elevated dietary restraint urges to eat and hunger after completing meals, and
scores report more difficulty controlling their caloric often show increased rate of consumption during
intake than individuals with lower dietary restraint an eating episode (Guss, & Kissileff, 2000; Sunday
scores (Jansen et al., 1988). & Halmi, 1996). Although people typically show
a diminished preference for food that is presented
Origins of a Chronic Tendency Toward repeatedly over time, including a reduction in sali-
Overconsumption vary response, individuals with BN do not show
Based on the idea that the chronic overconsump- sensory specific satiety when consuming one food
tion thesis may resolve several puzzling findings in type (LaChaussee, Kissileff, Walsh, & Hadigan,
the literature, it is useful to consider individual dif- 1992) or habituation of salivary response to tastes
ferences that might lead to this overeating tendency. of palatable food relative to nondisordered controls
Investigating factors that may contribute to an over- (Wisniewski, Epstein, Marcs, & Kaye, 1997). In
eating tendency is vital because this could advance addition, an avid sucking style during feeding in
our understanding of the etiologic processes that the first month of life, which may reflect greater
lead to eating pathology and obesity. In the sections reinforcement from food intake, predicted future
that follow, we propose three potential individual onset of overeating in childhood (Stice, Agras, &
difference factors that may contribute to overeating, Hammer, 1999) and higher body mass by age 6
which may represent the omitted third variable that (Agras, Kraemer, Berkowitz, & Hammer, 1990).
explains the relation between dietary restraint scores Data also suggests that obese versus lean individ-
and bulimic pathology. We also consider studies uals experience greater reward from eating. Obese
that have addressed these individual difference fac- individuals report that food intake is more reinforc-
tors. Finally, we propose studies that should advance ing than lean individuals (Jacobs & Wagner, 1984;
our understanding of the causal risk factors for eat- Johnson, 1974; Westenhoefer & Pudel, 1993).
ing pathology and obesity. Obese relative to lean adults work harder for food
and work for more food (Johnson, 1974; Saelens &
Greater Consummatory Food Reward Epstein, 1996), suggesting that the former find food
It could be that individuals who experience more reinforcing. Obese relative to lean children
greater reward from food intake are at elevated risk more often eat in the absence of hunger (Fisher &
for onset of binge eating and obesity. Some indi- Birch, 2002). Preferences for foods high in fat and
viduals may experience greater activation of the sugar predict an elevated rate of subsequent weight
mesolimbic reward system in response to the con- gain during childhood and an increased risk for
sumption of food, which might contribute to over- obesity onset in adulthood (Stunkard, Berkowitz,
consumption (Dawe & Loxton, 2004). Stallings, & Schoeller, 1999; Westerterp-​Plantenga,

140 Dieting
Ijederma, & Wijckmans-​Duijsens, 1996). One pro- risk for overeating and binge eating (Roefs et  al.,
spective fMRI study found that elevated response 2005). Theoretically, the elevated reward from food
in reward region (e.g., substantia nigra, ventral teg- intake that potentially characterizes those at risk for
mental area, ventral pallidum, and nucleus accum- overeating would be expected to increase anticipated
bens) to high-​calorie beverage tastes (Geha et  al. reward from eating and produce craving for palat-
2013) predicted future weight gain, which suggests able foods (Dawe & Loxton, 2004). The condition-
a vulnerability to overeating (Geha et  al., 2013), ing model of binge eating postulates that through
but other studies were not able to replicate this classical conditioning over time, cues such as the
effect (Stice, Burger, & Yokum, 2015; Stice, Spoor, sight and smell of food eventually elicit physiologi-
Bohon, & Small, 2008). cal responses that are experienced as food craving,
Although these data provide some support for which putatively increase the risk for binge eating
the notion that individuals with BN and obese indi- (Jansen, 1998). Repeating pairings of these cues
viduals may experience greater reward from food (which elicit craving) and binge eating are thought
intake, few studies have tested whether this also to strengthen the link between the cues and binge
holds true for those with elevated dietary restraint eating behaviors, which serves to maintain binge
scores. Taste test studies found that individuals eating and BN.
with elevated dietary restraint scores did not differ Relative to nondisordered controls, individuals
from those with lower scores on how pleasant they with BN or recurrent binge eating rate pictures of
rated palatable foods (Ahern, Field, Spoor, Bohon, food as more interesting and arousing and report a
& Stice, 2010; Fedoroff et  al., 1997; Goldfield & greater desire to eat, even when sated (Karhunen,
Legg, 2006; Roefs et al., 2005). In addition, indi- Lappalainen, Tammela, Turpeinen, & Uusitupa,
viduals with elevated dietary restraint scores did not 1997; Mauler et  al., 2006). Yet individuals with
work harder for snack foods (Ahern et  al., 2010) BN did not differ from dieters in terms of food
or work harder for snack foods than for healthy craving after exposure to pictures of palatable
foods (Goldfield & Legg, 2006) relative to those foods (Bossert-​Zaudig, Laessle, Meiller, Elllgring,
with lower dietary restraint scores. However, a brain & Pirke, 1991). Individuals with BN, relative to
imaging study found that individuals with high ver- controls, report greater urges to binge and less con-
sus low dietary restraint scores showed greater acti- fidence in their ability to control their food intake
vation in the orbitofrontal cortex and dorsolateral after exposure to the sight, smell, and taste of food
prefrontal cortex, which have both been implicated (Bulik, Lawson, & Carter, 1996; Staiger, Dawe,
in reward, in response to receiving high-​calorie bev- & McCarthy, 2000). Individuals with BN report
erage (Burger & Stice, 2011). greater urges to binge eat in response to both palat-
In sum, research supports the notion that elevated able and unpalatable control foods (Staiger et  al.,
reward from food intake is related to binge eating, 2000), suggesting that they may anticipate greater
bulimic symptoms, and obesity. However, there is reward from eating any food types, rather than
currently only mixed evidence that individuals with just palatable food. Moreover, individuals with
elevated dietary restraint rate food as more palat- BN, relative to controls, report persistent urges
able or rewarding than those who report low dietary to eat and hunger after completing meals (Guss &
restraint. Future research in this area is needed Kissileff, 2000).
before firm conclusions can be drawn. Hopefully Psychophysiology studies that examined the
additional studies will use objective techniques salivary response to food cues of individuals with
(e.g., brain imaging) to test whether individuals BN or recurrent binge eating versus controls have
with BN or obesity and individuals with elevated produced mixed results, with some finding that the
dietary restraint scores show elevated reward from former show more (Legenbauer, Vogele, & Ruddel,
food intake relative to controls. Further, prospective 2004; LeGoff, Leichner, & Spigelman, 1988), less
studies should test whether elevated reward from (Bulik et al., 1996; Karhunen et al., 1997), or similar
food intake increases risk for future increases in self-​ (Staiger et al., 2000) salivary response to food cues
reported dieting, binge eating, and weight gain. relative to controls. Salivary response to food pres­
entation appears to reflect food craving, as it corre-
Greater Anticipatory Food Reward lates positively with self-​reported hunger and desire
It is also possible that greater anticipated reward to binge eat (Legenbauer et al., 2004). Studies that
from food intake is what differentiates those with have examined other physiological measures have
high versus low dietary restraint scores and increases also produced inconsistent findings. For example,

Stice, Shaw 141
one study found that individuals with BN, relative food may lead to the overconsumption of food in
to controls, showed reduced startle response to pic- general, which results in the positive energy bal-
tures of food and increased corrugator facial muscle ance necessary for obesity onset. It might be useful
responses to these same images, but no differences if future studies used pictures of food or real food,
with regard to skin conductance and heart rate rather than food words, as this may provide a more
response (Mauler et  al., 2006). However, another sensitive test of implicit attitudes toward food in
study found that individuals with BN did not show these two groups.
a reduced startle response while viewing pictures of Self-​
reported food cravings correlated posi-
food relative to controls (Friederich et al., 2006). tively with body mass (Delahanty, Meigs, Hayden,
This pattern of findings implies that individuals Williamson, & Nathan, 2002) and objectively
with and without BN do not show reliable differ- measured caloric intake in the lab (Nederkoorn,
ences in these physiological responses to food cues, Smulders, & Jansen, 2000). Further, obese relative
that these measures are unreliable, or that the small to lean individuals report stronger craving of high-​
sample sizes produced inconsistent findings. It fat and high-​ sugar foods (Drewnowski, Krahn,
is also possible that individuals with BN show an Demitrack, Nairn, & Gosnell, 1992; Drewnowski,
approach-​avoidance response to food stimuli that Kurth, Holden-​Wiltse, & Saari, 1992; White,
results from an increased drive to consume food Whisenhunt, Williamson, Greenway, & Netemeyer,
coupled with negative feelings toward food because 2002). Obese versus lean humans also show greater
they often binge eat, which moves them further responsivity of brain regions associated with reward
from the thin-​ideal to which they often aspire. Such and motivation (striatum, amygdala, orbitofrontal
an approach–​avoidance response may lead to both cortex) to pictures of high-​calorie foods versus low-​
positive and negative emotional responses to food calorie foods and control images (e.g., Bruce et  al.
cues, which are being assessed by these physiological 2012; Frankort et  al., 2012; Holsen et  al., 2012;
measures. Consistent with this notion, individuals Martin et  al., 2010). Critically, individuals who
with BN often report more negative feelings while show elevated reward region response to food images
looking at, smelling, or touching food (Bulik et al., and cues show elevated future weight gain, which is
1996; Legenbauer et al., 2004; Mauler et al., 2006; indicative of overeating (Demos et al., 2012; Stice,
Staiger et al., 2000; Uher et al., 2004). Burger, & Yokum, 2015; Yokum et al., 2011, 2014).
Research has compared obese and lean indi- Individuals with high dietary restraint scores
viduals using cognitive psychology paradigms to report greater cravings for palatable foods than
determine whether the former show a more posi- those with lower scores (Gendall, Joyce, Sullivan,
tive attitude toward higher caloric density foods, & Bulik, 1998; Pelchat, 1997; Polivy, Coleman,
on the basis that elevated food craving would result & Herman, 2005), though this relation did not
in positive implicit attitudes toward these foods. replicate in other studies (Fedoroff et al., 1997;
One study used the Implicit Association Test (IAT; Hill, Weaver, & Blundell, 1991; Rodin, Mancuso,
Greenwald, McGhee, & Schwartz, 1998) to objec- Granger, & Nelbach, 1991; White et al., 2002).
tively test whether obese women show more posi- Individuals with elevated dietary restraint scores
tive attitudes toward food than lean women (Roefs report a greater likelihood to give in to food crav-
& Jansen, 2002). Unexpectedly, both obese and ings and consume the craved food (Fedoroff et al.,
lean participants showed more negative associations 1997; Polivy et al., 2005). Results imply that those
with high-​fat foods than with low-​fat foods, with with elevated dietary restraint scores only report
this effect being more pronounced for the obese moderately greater food craving, which is why
women. A second study used the Extrinsic Affective the effects are inconsistent. It is also possible that
Simon Task (EAST; De Houwer, 2003) to test food craving is only elicited when the individuals
whether obese youth show more positive implicit encounter palatable foods, which might suggest that
attitudes toward unhealthy foods (versus healthy future studies should involve presentation of real
foods) in relation to lean youth (Craeynest et  al., food. Consistent with this reasoning, simply smell-
2005). Obese relative to lean youth showed a more ing food (Fedoroff et al., 1997) or thinking about
pronounced positive implicit attitude toward food eating food (Fedoroff et al., 1997) leads to greater
in general, though there were no differences with caloric intake among individuals with high but not
regard to explicit attitudes toward unhealthy versus low dietary restraint scores. A brain imaging study
healthy foods. The evidence that obese individu- found that individuals with high versus low dietary
als have a more positive implicit attitude toward restraint scores showed greater activation in reward

142 Dieting
regions (orbitofrontal cortex and dorsolateral pre- (Presnell et al., 2007), whereas those on monitored
frontal cortex) in response to palatable food images weight loss diets do (Groesz & Stice, 2007; Presnell
(Coletta et al., 2009). These data seem consistent & Stice, 2003), it could be that calorie deficit diet-
with the notion that the former have elevated food ing reduces food craving. Indeed, uncontrolled
cravings that are easily triggered. obesity treatment trials indicate that assignment
Studies have also used implicit measures to to low-​calorie diets that result in weight loss pro-
explore whether individuals with elevated dietary duce reductions in self-​reported food cravings and
restraint scores show a positive emotional response hunger for adults (Harvey, Wing, & Mullen, 1993;
to food stimuli than individuals with lower dietary Lappalainen, Sjoden, Hursti, & Vesa, 1990; Martin,
restraint scores. One study used the affective prim- Makris, et  al., 2006; Martin, O’Neil, & Pawlow,
ing paradigm (Fazio, Sanbonmatsu, Powell, & 2006; Wing, Marcus, Blair, & Burton, 1991).
Kardes, 1986) and the Extrinsic Affective Simon Moreover, participants on more restrictive very low-​
Task (EAST; De Houwer, 2003) to test whether calorie diets show significantly greater reductions in
those with high dietary restraint scores showed a self-​reported food craving than those on less restric-
greater automatic positive attitude toward palat- tive low-​ calorie diets (Lappalainen et  al., 1990;
able foods than to unpalatable foods relative to their Martin, O’Neil, et al., 2006). Assignment to a low-​
lower-​scoring counterparts (Roefs et  al., 2005). calorie diet that results in weight loss produced sig-
Although both groups showed greater positive nificantly greater reductions in self-​reported hunger
attitudes toward palatable foods relative to low-​fat relative to an assessment-​ only control condition
foods on both paradigms, there were no differences (Groesz & Stice, 2007).
in the response between individuals with high ver- There is also evidence that restriction from a
sus low dietary restraint scores. Another study that particular food type results in reductions in craving
used a version of the IAT (Greenwald et al., 1998) for that particular food type (Harvey et al., 1993;
that presented pictures of candy found that dietary Martin, Makris et al., 2006). One weight loss trial
restraint scores positively correlate with implicit found that carbohydrate cravings decreased more
positive attitudes toward candy (Hofmann et  al., for participants on a low-​carbohydrate diet than for
2007). However, another study found that individ- those on a low-​fat diet, whereas cravings for high-​fat
uals with high and low dietary restraint scores did foods decreased more for participants on the low-​fat
not differ on implicit attitudes toward pictures of diet than for participants on the low-​carbohydrate
palatable foods versus unpalatable foods or control diet (Martin, Makris, et al., 2006). These findings
images (Ahern et al., 2010). Future studies should suggest that caloric deprivation reduces food crav-
consider using actual food in these paradigms to see ing and that deprivation of particular food types
whether this provides a more sensitive measure of produces even greater decreases in craving for
food craving. those particular foods. These results provide more
Other studies have compared salivary reactivity evidence that experiments that manipulate dietary
in response to presentation of food and food cues restriction can produce findings that are opposite of
among those with high and low dietary restraint studies that use self-​report dietary restraint scales.
scores. Individuals with elevated dietary restraint There are two theoretical accounts that may
scores show greater salivary response to the sight and explain why low-​calorie diets produce reductions
smell of real food than their lower-​scoring counter- in food craving. First, according to Jansen (1998),
parts (Legenbauer et al., 2004), though null effects conditioned food cues will elicit food craving as
have also emerged (Bulik et al., 1996; Nederkoorn & long as they remain a reliable predictor of excessive
Jansen, 2002). This pattern of findings may suggest food intake. Theoretically, craving will be reduced
that individuals with elevated dietary restraint scores when the cues do not reliably predict intake of high-​
show only moderately greater food craving than calorie foods. It is possible that low-​calorie diets that
their lower-​scoring counterparts, which is why the involve restricted intake of high-​calorie foods effec-
effects are not consistently observed. Alternatively, tively produce deconditioning. Perhaps individuals
it may signal that procedural aspects of the studies, on low-​calorie diets are still exposed to many food
such as the duration of exposure to food, impact cues but are not engaging in excessive intake of the
whether elevated salivary response is observed. high-​fat and high-​sugar foods. A  second explana-
Given the evidence that individuals with elevated tion for why low-​calorie diets reduce food craving
dietary restraint scores do not typically achieve a is that low-​calorie diets often have a limited vari-
negative energy balance necessary for weight loss ety of foods (Martin, O’Neil, et al., 2006). Access

Stice, Shaw 143
to more versus less variety of food types promotes Obese individuals have also shown greater
greater overall caloric intake (McCrory, Suen, & impulsivity than lean individuals on self-​ report
Roberts, 2002). measures (Chalmers, Bowyer, & Olenick, 1990;
In sum, the evidence is somewhat mixed Ryden et al., 2003). Relative to lean women, obese
regarding whether individuals with obesity, women show more difficulties with response inhibi-
bulimic pathology, or elevated dietary restraint tion on a stop-​signal task (Nederkoorn, Smulders,
scores show greater anticipatory reward from Havermans, Roefs, & Jansen, 2006). Obese adults,
food than do controls, with self-​report measures, relative to lean adults, show a preference for high
implicit tests, and psychophysiology measures immediate gain, but larger future losses in a gam-
producing inconsistent effects. Future studies bling task, rather than lower immediate reward and
that use more objective and validated measures less future loss (Davis, Levitan, Muglia, Bewell, &
may be warranted. Interestingly, there was con- Kennedy, 2004). However, other studies have not
sistent evidence that caloric deprivation generally found a preference for immediate rewards relative
reduces food craving and that deprivation of par- to delayed rewards between obese and lean adults
ticular food types produces even greater decreases (Forzano & Logue, 1992; Logue & King, 1991;
in craving for those particular foods. Nederkoorn, Smulders, et al., 2006). Obese relative
to lean children are less effective in response inhi-
Greater Impulsivity bition in a stop-​signal task and are more sensitive
It has also been hypothesized that impulsive to reward in a gambling task than lean children
individuals may be more vulnerable to the omni- (Nederkoorn, Braet, van Eijs, Tanghe, & Jansen,
present temptation of palatable foods in the current 2006). Among obese children, response inhibi-
obesogenic environment, which may increase risk tion deficits on a stop-​signal task correlated posi-
for weight gain, as well as for onset of binge eating tively with BMI (Nederkoorn, Jansen, Mulkens, &
and BN (Nederkoorn et al., 2004). Among individ- Jansen, 2007). Initially elevated impulsivity is asso-
uals with inhibitory control deficits, impulses and ciated with less weight loss during obesity treatment
immediate reward will play a more important role (Best et al., 2012; Nederkoorn, Braet, et al., 2006;
in determining behaviors than longer-​term adverse Nederkoorn, Jansen et  al., 2007; Pauli-​Pott et  al.,
consequences of the behavior (Nederkoorn, et  al., 2010). Individuals with elevated scores on self-​
2004). It is possible that the elevated consumma- reported impulsivity scales show greater objectively
tory or anticipatory food rewards contribute to measured caloric intake than their lower-​ scoring
more impulsive responses to food cues. peers (Guerrieri et al., 2007). Critically, individuals
Individuals who exhibit binge eating or BN who show immediate reward bias for high-​calorie
typically have elevated scores on self-​report mea- foods show elevated future weight gain (Evans,
sures of trait impulsivity (Kane et al., 2004; Nasser, Fuller-​Rowell, & Doan, 2012; Francis & Susman,
Gluck, & Geliebter, 2004; Nederkoorn et  al., 2009; Schlam, Wilson, Shoda, Mischel, & Ayduk,
2004; Vervaet, Audenaert, & Heeringen, 2003). 2013; Seeyave et al., 2009; Sutin et al., 2011).
Bulimic symptoms correlate positively with self-​ In terms of neuroimaging findings, obese ver-
reported impulsivity among women (Guerrieri, sus lean teens showed less activation of prefrontal
Nederkoorn, & Jansen, 2007; Loxton & Dawe, regions (dorsolateral prefrontal cortex [dlPFC], ven-
2007), particularly scales assessing a tendency to tral lateral prefrontal cortex [vlPFC]) when trying
act rashly (Fischer, Smith, & Anderson, 2003). to inhibit responses to high-​calorie food images and
Self-​reported impulsivity correlated positively with behavioral evidence of reduced inhibitory control
test meal caloric intake among individuals seek- (Batterink et  al., 2010), though participants who
ing weight loss treatment (Nasser et  al., 2004) showed less recruitment of inhibitory regions did
and individuals with BED (Galanti, Gluck, & not show excess future weight gain. Another study
Geliebter, 2007). Bulimic symptoms also correlate found that participants who showed less recruit-
positively with impulsive behaviors, including self-​ ment of inhibitory control regions (inferior, mid-
mutilation, suicidal ideation, and heavy substance dle, and superior frontal gyri) during difficult versus
use (Penas-​Lledo & Waller, 2001; Pidcock, Fischer, easy choices on a delay-​discounting task showed ele-
Forthun, & West, 2000; Vervaet et al., 2003) and vated future weight gain (Kishinevsky et al., 2012).
delinquency and substance abuse predicted future Results converge with evidence that obese versus
onset of bulimic symptoms (Wonderlich, Connolly, lean adults showed reduced gray matter volume in
& Stice, 2004). the prefrontal cortex (Pannacciulli et  al., 2006), a

144 Dieting
region that modulates inhibitory control, and with symptoms and greater weight loss than assignment
a marginal trend for reduced gray matter volume in to assessment-​only control conditions. Our analy-
the prefrontal cortex to predict weight gain over 1-​ sis of potential explanations for these consistently
year follow-​up (Yokum, Ng, & Stice, 2011). incompatible findings identified several plausible
Studies that compared individuals with high and explanations.
low dietary restraint scores have also provided find- One possibility that follows from considering
ings consistent with the impulsivity theory. Logue the different inferential strength of prospective ver-
and King (1991) found that those who indicated sus experimental studies is that some omitted third
that they were currently on a diet, relative to those variable explains the results from the prospective
who were not, exhibited significantly more impul- studies that suggest that individuals who report
sive responding on an immediate versus delayed dietary restraint are at increased risk for future
food reward operant conditioning task. They also bulimic symptoms. Another possibility is that the
found that impulsive responding positively corre- prospective studies used invalid measures of dietary
lates with dietary restraint scores. Nederkoorn and restraint that do not identify individuals who are in
associates (2004) found that individuals with high a negative energy balance, which explains why they
dietary restraint scores exhibited significantly worse produced different results than experiments that
inhibition of basic non-​food-​related motor responses manipulated energy-​deficit dieting. A  third possi-
on a stop-​signal task and scored higher on a self-​ bility is that the weight loss dietary interventions
report impulsivity scale than their lower-​ scoring examined in the experiments are more extreme
counterparts. However, one study found that dietary than real-​world weight loss dieting and have limited
restraint scores did not significantly correlate with ecological validity because participants meet with
self-​
reported impulsivity (Guerrieri et  al., 2007). weight loss professionals.
The evidence that individuals with high versus low However, it appears that only the first explana-
dietary restraint scores typically overeat in response tion can logically explain the inconsistent findings.
to presentation of food cues, such as smelling a pal- Our analysis suggested that the only plausible expla-
atable food (e.g., Jansen & Van den Hout, 1991; nation is that dietary restraint scales identify a sub-
Rogers & Hill, 1989) or even just thinking about group at increased risk for future onset of bulimic
a palatable food (Fedoroff et al., 1997), might sug- pathology. Based on the evidence, we propose that
gest that the former are more likely to yield to these true energy deficit dieting results in a reduction in
tempting food cues and respond by overeating. bulimic symptoms, and that these scales identify
In sum, there is mounting evidence that indi- individuals with a chronic overeating tendency.
viduals with BN, binge eating disturbances, obe- We posit that this overeating tendency results in
sity, and elevated dietary restraint scores show attempts to curb caloric intake (self-​initiated diet-
greater impulsivity in general, as well as in response ing) and also increases risk for future onset of an
to food stimuli, relative to controls. Thus, future ED characterized by uncontrollable bouts of over-
studies should continue to examine the potential eating. This working hypothesis appears to explain
role of impulsivity in the etiology of binge eating, the inconsistent pattern of findings, as well as sev-
BN, and obesity. It will also be important for stud- eral other anomalous findings. Further, we explored
ies to explore factors that interact with impulsivity the available evidence that individual differences in
in the prediction of binge eating and weight gain. reinforcement from food intake, anticipated rein-
For instance, it has been suggested that a tendency forcement from food intake, and impulsivity may
toward disinhibited eating interacts with dietary give rise to this overeating tendency.
restriction to increase risk for overeating (Ouwens Despite this working hypothesis, there are two
et al., 2003). other possible explanations for the inconsistent find-
ings that warrant further research. First, it is possible
Summary of the Review of Possible that the reductions in bulimic symptoms are a prod-
Explanations for the Inconsistent Findings uct of the demand characteristics of experiments in
Prospective studies have found that individuals which participants are assigned to a weight loss diet
who report dietary restraint are at increased risk for condition. Future studies should address this possi-
future onset of binge eating, bulimic pathology, and bility with experiments that use confederate reports
EDs. However, experiments have found that assign- of bulimic symptoms, objective measures of these
ment to weight loss diets results in significantly symptoms, or placebo control conditions. Second,
greater reductions in binge eating and bulimic it is conceivable that some small subset of dieters

Stice, Shaw 145
engage in a particularly unhealthy form of dieting, available can we hope to develop valid etiologic mod-
such as extreme fasting, which really does increase els for psychopathology and design optimally effec-
their risk for bulimic symptoms. Because only a tive prevention and treatment interventions.
small subset of dieters eventually develop bulimic
pathology, we think this is an important possibil- Future Directions
ity to explore as well, though it will be vital to use This chapter suggests several important direc-
objective measures of these unhealthy weight con- tions for future research. First, future research
trol behaviors and to conduct randomized experi- should attempt to identify the third variable
ments that manipulate candidate dieting behaviors. that is inadvertently tapped by dietary restraint
scales that truly increases risk for future onset of
Conclusion bulimic pathology. Thus, future studies should
In conclusion, experiments from multiple labs attempt to experimentally manipulate any poten-
that have evaluated weight loss and weight main- tial third variables, such as a chronic tendency
tenance diet interventions suggest that neither toward overeating, in an effort to provide more
causes eating pathology. This is key because it sug- rigorous inferences about any putative cause of
gests that obesity prevention and treatment inter- bulimic symptoms. It will also be vital to make
ventions that prescribe weight loss dieting do not these experiments as ecologically valid as pos-
produce iatrogenic effects. Indeed, an interven- sible. Second, it will be important to investigate
tion that promotes weight maintenance through individual difference factors that might give rise
reduced caloric intake and increased exercise is to potential third variables under consideration.
the only prevention program that has significantly We suspect it would be useful to investigate indi-
reduced future onset of EDs in two trials. There vidual differences in consummatory food reward,
is also evidence that a weight loss dieting inter- anticipatory food reward, and impulsivity as fac-
vention was an efficacious treatment for this ED. tors that may give rise to both efforts at curbing an
These data indicated that the dietary restraint overeating tendency and eventual bulimic pathol-
model of bulimic pathology is in need of revi- ogy. Third, it will also be important to conduct
sion. One remarkable feature of this literature is experiments that manipulate particular weight
the confidence that has been placed in the asser- loss behaviors. As noted, we believe that it is pos-
tion that dieting causes bulimic pathology in the sible that some subset of individuals with elevated
absence of any experimental evidence that actual dietary restraint scores engage in some particularly
dietary restriction increases risk for bulimic symp- unhealthy weight loss behaviors that may increase
toms (e.g., Fairburn, 1997; Heatherton & Polivy, risk for bulimic pathology, such as extreme fast-
1992; Polivy & Herman, 1985; Levine & Smolak, ing. For ethical reasons, it would be best to reduce,
2006; Neumark-​Sztainer et al., 2006). rather than increase, these unhealthy behaviors
Nonetheless, we feel it would be useful to identify experimentally. Fourth, it would be valuable to
the third variable that is tapped by dietary restraint investigate the possibility that demand character-
scales, which truly increases risk for future onset of istics lead to the reductions in bulimic symptoms
bulimic pathology, because this will advance our in experiments investigating the effects of weight
understanding of the etiologic processes that give loss diets on this outcome. Such studies will need
rise to eating pathology. More generally, the lit- to use confederate reports from individuals who
erature reviewed here illustrates the importance of are blinded to the condition of the experiment,
documenting the validity of measures with methods objective measures of bulimic symptoms (e.g.,
that are less subject to distortion by social desirabil- electrolyte abnormalities), or placebo control con-
ity. Finally, this analysis underscores the hazards of ditions. Fifth, it is important for future research to
relying on prospective studies when making etiologic develop valid measures of dietary restriction, with-
inferences. The fact that it is impossible to rule out out which it will be virtually impossible to make
third-​variable alternative explanations for prospec- accurate inferences regarding the consequences of
tive findings from longitudinal studies serves as a dieting. Finally, given the evidence that weight loss
cogent reminder of the importance of using random- dieting curbs bulimic symptoms in prevention and
ized experiments that manipulate putative etiologic treatment trials, it is important to develop even
factors to confirm causal relations suggested by pro- more effective interventions that promote weight
spective studies (Hinshaw, 2002; Stice, 2002). Only maintenance diets for individuals at a healthy
through the use of the most rigorous research designs weight and weight loss diets for individuals who

146 Dieting
are overweight. The increasing prevalence of obe- Bellisle, F., & Dalix, A. (2001). Cognitive restraint can be offset
sity suggests that this is a particularly pressing need by distraction, leading to increased meal intake in women.
American Journal of Clinical Nutrition, 74, 197–​200.
from a public health perspective. Prevention pro- Best, J. R., Theim, K. R., Gredysa, D. M., Stein, R. I., Welch,
grams that affect two adverse outcomes (EDs and R. R., Saelens, B. E.,  . . .  Wilfley, D. E. (2012). Behavioral
obesity) clearly have greater public health signifi- economic predictors of overweight children’s weight loss.
cance than programs that affect only one of these Journal of Consulting and Clinical Psychology, 80, 1086–​1096.
outcomes. Boggiano, M. M., Chandler, P. C., Viana, J. B., Oswald, K. D.,
Maldonado, C. R., & Wauford, P. K. (2005). Combined
dieting and stress evoke exaggerated responses to opioids in
References binge-​eating rats. Behavioral Neuroscience, 119, 1207–​1214.
Agras, W. S., Kraemer, H., Berkowitz, R., & Hammer, L. (1990). Bohon, C., Stice, E., & Burton, E. (2008). Maintenance fac-
Influence of early feeding style on adiposity by 6 years of age. tors for persistence of bulimic pathology:  A  community-​
Journal of Pediatrics, 116, 805–​809. based natural history study. International Journal of Eating
Agras, W. S., & Telch, C. F. (1998). The effects of caloric depriva- Disorders, 42, 173–​178.
tion and negative affect on binge eating in obese binge-​eating Bossert-​Zaudig, S., Laessle, R., Meiller, C., Ellgring, H., & Pirke,
disordered women. Behavior Therapy, 29, 491–​503. K. M. (1991). Hunger and appetite during visual perception
Ahern, A., Field, M., Spoor, S., Bohon, C., & Stice, E. (2010). of food in eating disorders. European Psychiatry, 6, 237–​242.
Relation of dietary restraint scores to cognitive biases and Braet, C., Tanghe, A., De Bode, P., Franckx, H., & VanWinckel,
reward sensitivity. Appetite, 55, 61–​68. M. (2003). Inpatient treatment of obese children: A multi-
Allen, K., Byrne, S., Forbes, D., & Oddy, W. (2009). Risk component programme without stringent calorie restriction.
factors for full-​and partial-​syndrome early adolescent eat- European Journal of Pediatrics, 162, 391–​396.
ing disorders: A population-​based pregnancy cohort study. Bruce, J., Hancock, L., Bruce, A., Lepping, R., Martin, L.,
Journal of the Academy of Child and Adolescent Psychiatry, Lundgren, J.,  . . .  Savage, C. R. (2012). Changes in brain
48, 800–​809. activation to food pictures after adjustable gastric banding.
Allen, K. L., Byrne, S. M., Oddy, W. H., & Crosby, R. D. Surgery for Obesity and Related Diseases, 8, 602–​608.
(2013). DSM-​IV-​TR and DSM-​5 eating disorders in ado- Bulik, C. M., Lawson, R. H., & Carter, F. A. (1996). Salivary
lescents: Prevalence, stability, and psychosocial correlates in reactivity in restrained and unrestrained eaters and women
a population-​based sample of male and female adolescents. with bulimia nervosa. Appetite, 27, 15–​24.
Journal of Abnormal Psychology, 122, 720–​732. Burger, K., & Stice, E. (2011). Relation of dietary restraint
Arcelus, J., Mitchell, A., Wales, J., & Nielsen, S. (2011). scores to activation of reward-​ related brain regions in
Mortality rates in patients with anorexia nervosa and response to food intake, anticipate intake, and food pictures.
other eating disorders:  A  meta-​ analysis of 36 studies. NeuroImage, 55, 233–​239.
Archives of General Psychiatry, 68, 724–​731. doi:10.1001/​ Burton, E. M., & Stice, E. (2006). Evaluation of a healthy-​weight
archgenpsychiatry.2011.74 treatment program for bulimia nervosa:  A  preliminary ran-
Attenburrow, M. J., Williams, C., Odontiadis, J., Powell, J., van de domized trial. Behaviour Research and Therapy, 44, 1727–​1738.
Ouderaa, F., Williams, M., & Cowen, P. (2003). The effect of a Burton, E. M., Stice, E., Bearman, S. K., & Rohde, P. (2007). An
nutritional source of tryptophan on dieting-​induced changes in experimental test of the affect-​regulation model of bulimic
brain 5-​HT function. Psychological Medicine, 33, 1381–​1386. symptoms and substance use:  An affective intervention.
Bacon, L., Keim, N. L., Van Loan, M. D., Derricote, M., Gale, International Journal of Eating Disorders, 40, 27–​36.
B., Kazaks, A., & Stern, J. S. (2002). Evaluating a non-​diet Carter, J. C., Olmsted, M. P., Kaplan, A. S., McCabe, R. E.,
wellness intervention for improvement of metabolic fitness, Mills, J. S., & Aime, A. (2003). Self-​help for bulimia ner-
psychological well-​being and eating and activity behaviors. vosa:  A  randomized controlled trial. American Journal of
International Journal of Obesity, 26, 854–​865. Psychiatry, 160, 973–​978.
Bandini, L. G., Schoeller, D. A., Cyr, H. N., & Dietz, W. H. (1990). Chalmers, D. K., Bowyer, C. A., & Olenick, N. L. (1990).
Validity of reported energy-​intake in obese and nonobese ado- Problem drinking and obesity: A comparison in personality
lescents. American Journal of Clinical Nutrition, 52, 421–​425. patterns and life-​style. International Journal of Addiction, 25,
Baranowski, T., Klesges, L. M., Cullen, K. W., & Himes, J. H. 803–​817.
(2004). Measurement of outcomes, mediators, and mod- Coletta, M., Platek, S., Mohamed, F. B., van Steenburgh, J.
erators in behavioral obesity prevention research. Preventive J., Green, D., Lowe, M. R., (2009). Brain activation in
Medicine, 38, S1–​S13. restrained and unrestrained eaters: An fMRI study. Journal of
Bathalon, G. P., Tucker, K. L., Hays, N. P., Vinken, A. G., Abnormal Psychology, 118, 598–​609.
Greenberg, A. S., McCrory, M. A., & Roberts, S. B. (2000). Cools, J., Schotte, D. E., & McNally, R. J. (1992). Emotional
Psychological measures of eating behavior and the accuracy arousal and overeating in restrained eaters. Journal of
of 3 common dietary assessment methods in healthy post- Abnormal Psychology, 101, 348–​351.
menopausal women. American Journal of Clinical Nutrition, Craeynest, M., Crombez, G., De Houwer, J., Deforche, B.,
71, 739–​745. Tanghe, A., & De Bourdeaudhij, I. (2005). Explicit and
Batterink, L., Yokum, S., & Stice, E. (2010). Body mass correlates implicit attitudes towards food and physical activity in
inversely with inhibitory control in response to food among childhood obesity. Behaviour Research and Therapy, 43,
adolescent girls: An fMRI study. NeuroImage, 52, 1696–​1703. 1111–​1120.
Bearman, S. K., Stice, E., & Chase, A. (2003). Effects of body Davis, C., Levitan, R. D., Muglia, P., Bewell, C., & Kennedy,
dissatisfaction on depressive and bulimic symptoms: A longi- J. L. (2004). Decision-​making deficits and overeating: A risk
tudinal experiment. Behavior Therapy, 34, 277–​293. model for obesity. Obesity Research, 12, 929–​935.

Stice, Shaw 147
Davis, C., & Woodside, D. B. (2002). Sensitivity to reward- Fairburn, C. G., & Beglin, S. J. (1994). Assessment of eating dis-
ing effects of food and exercise in the eating disorders. orders:  Interview or self-​report questionnaire? International
Comprehensive Psychiatry, 43, 189–​194. Journal of Eating Disorders, 16, 363–​370.
Dawe, S., & Loxton, N. J. (2004). The role of impulsivity in Fairburn, C. G., & Cooper, Z. (1993). The eating disorder
the development of substance use and eating disorders. examination. In C. G. Fairburn & G. T. Wilson (Eds.), Binge
Neuroscience and Biobehavioral Review, 28, 343–​351. eating:  Nature, assessment, and treatment (pp. 317–​360).
de Castro, J. M., & Elmore, D. K. (1988). Subjective hunger New York, NY: Guilford Press.
relationships with meal patterns in the spontaneous feed- Fairburn, C. G., Cooper, Z., Doll, H. A., Norman, P., &
ing behavior of humans: Evidence for a causal connection. O’Connor, M. (2000). The natural course of bulimia ner-
Physiology and Behavior, 43, 159–​165. vosa and binge eating disorder in young women. Archives of
De Houwer, J. (2003). The extrinsic affective Simon task. General Psychiatry, 57, 659–​665.
Experimental Psychology, 50, 77–​85. Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F.
Delahanty, L. M., Meigs, J. B., Hayden, D., Williamson, D. A., R. (1986). On the automatic activation of attitudes. Journal
& Nathan, D. M. (2002). Psychological and behavioral cor- of Personality and Social Psychology, 50, 229–​238.
relates of baseline BMI in the diabetes prevention program. Fedoroff, I. C., Polivy, J., & Herman, C. P. (1997). The effect of
Diabetes Care, 25, 1992–​1998. pre-​exposure to food cues on the eating behavior of restrained
Demos, K., Heatherton, T., & Kelley, W. (2012). Individual and unrestrained eaters. Appetite, 28, 33–​47.
differences in nucleus accumbens activity to food and Field, A. E., Camargo, C. A., Taylor, C. B., Berkey, C. S.,
sexual images predict weight gain and sexual behavior. Frazier, A. L., Gillman, M. W., & Colditz, G. A. (1999).
Journal of Neuroscience, 32, 5549–​5552. doi:10.1523/​ Overweight, weight concerns, and bulimic behaviors among
JNEUROSCI.5958-​11.2012 girls and boys. Journal of the American Academy of Child and
De Zwaan, M., Mitchell, J. E., Crosby, R. D., Mussell, M. P., Adolescent Psychiatry, 38, 754–​760.
Raymond, N. C., Specker, S. M., & Seim, H. C. (2005). Field, A., Haines, J., Rosner, B., & Willett, W. (2009). Weight-​
Short-​term cognitive behavioral treatment does not improve control behaviors and subsequent weight change among
outcome in a comprehensive very-​ low-​
calorie diet pro- adolescents and young adult females. American Journal of
gram in obese women with binge eating disorder. Behavior Clinical Nutrition, 91, 147–​153.
Therapy, 36, 89–​99. Fischer, S., Smith, G. T., & Anderson, K. G. (2003). Clarifying
Drewnowski, A., Krahn, D. D., Demitrack, M. A., Nairn, K., the role of impulsivity in bulimia nervosa. International
& Gosnell, B. A. (1992). Taste responses and preferences Journal of Eating Disorders, 33, 406–​411.
for sweet high-​fat foods:  Evidence for opioid involvement. Fisher, J. O., & Birch, L. L. (2002). Eating in the absence of hun-
Physiology and Behavior, 51, 371–​379. ger and overweight in girls from 5 to 7 years of age. American
Drewnowski, A., Kurth, C., Holden-​ Wiltse, J., & Saari, J. Journal of Clinical Nutrition, 76, 226–​231.
(1992). Food preferences in human obesity: Carbohydrates Flegal, K. M., Carroll, M. D., Kit, B. K., & Ogden, C. L. (2012).
versus fats. Appetite, 18, 207–​221. Prevalence of obesity and trends in the distribution of body mass
Eckern, M., Stevens, W., & Mitchell, J. E. (1999). Brief index among US adults, 1999–​2010. JAMA, 307, 491–​497.
report: The relationship between rumination and eating dis- Forzano, L. B., & Logue, A. W. (1992). Predictors of adult
orders. International Journal of Eating Disorders, 26, 414–​419. humans’ self-​control and impulsiveness for food reinforcers.
Emmons, L. (1992). Dieting and purging behavior in black and Appetite, 19, 33–​47.
white high school students. Journal of the American Dietetic Foster, G. D., Wadden, T. A., Kendall, P. C., Stunkard, A. J., &
Association, 92, 306–​312. Vogt, R. A. (1996). Psychological effects of weight loss and
Engelberg, M. J., Gauvin, L., & Steiger, H. (2005). A natural- regain:  A  prospective evaluation. Journal of Consulting and
istic evaluation of the relation between dietary restraint, the Clinical Psychology, 64, 752–​757.
urge to binge, and actual binge eating. International Journal Francis, L., & Susman, E. J. (2009). Self-​regulation failure and
of Eating Disorders, 38, 355–​360. rapid weight gain in children from age 3 to 12 years. Archives
Epstein, L. H., Paluch, R. A., Saelens, B. E., Ernst, M. M., & of Disease in Children and Adolescents, 163, 297–​302.
Wilfley, D. E. (2001). Changes in eating disorder symp- Franko, D. L., Wolfe, B. E., & Jimerson, D. C. (1994). Elevated
toms with pediatric obesity treatment. Journal of Pediatrics, sweet taste pleasantness ratings in bulimia nervosa. Physiology
139, 58–​65. and Behavior, 56, 969–​973.
Epstein, L. H., Truesdale, R., Wojcik, A. Paluch, R. A., & Raynor, Frankort, A., Roefs, A., Siep, N., Roebroeck, A., Havermans,
H. A. (2003). Effects of deprivation on hedonics and reinforc- R., & Jansen, A. (2011–​2012). Reward activity in satiated
ing value of food. Physiology and Behavior, 78, 221–​227. overweight women is decreased during unbiased viewing but
Espeland, M. A., Kumanyika, S., Wilson, A. C., Wilcox, S., increased when imaging taste: An event-​related fMRI study.
Chao, N., Bahnson, J.,   .  .  .  Zheng, B. (2001). Lifestyle International Journal of Obesity, 36, 1–​11.
interventions influence relative errors in self-​reported diet French, S. A., Jeffery, R. W., Forster, J. L., McGovern, P. G.,
intake of sodium and potassium. Annuals of Epidemiology, Kelder, S. H., & Baxter, J. E. (1994). Predictors of weight
11, 85–​93. change over two years among a population of working
Evans, G., Ruller-​Rowell, R., & Doan, S. (2012). Childhood adults: The Healthy Worker Project. International Journal of
cumulative risk and obesity:  The mediating role of self-​ Obesity, 18, 145–​154.
regulatory ability. Pediatrics, 129, e68. French, S., Jeffery, R. W., & Murray, D. (1999). Is dieting good for
Fairburn, C. G. (1997). Eating disorders. In D. M. Clark & C. you? Prevalence, duration, and associated weight and behav-
G. Fairburn (Eds.), Science and practice of cognitive behaviour ior changes for specific weight loss strategies over four years
therapy (pp. 209–​241). Oxford, UK: Oxford University Press. in US adults. International Journal of Obesity, 23, 320–​327.

148 Dieting
French, S. A., Jeffery, R. W., & Wing, R. R. (1994). Food intake Hagan, M. M., Chandler, P. C., Wauford, P. K., Rybak, R. J., &
and physical activity:  A  comparison of three measures of Oswald, K. D. (2003). The role of palatable food and hun-
dieting. Addictive Behaviors, 19, 401–​409. ger as trigger factors in an animal model of stress induced
French, S. A., Perry, C. L., & Leon, G. R., & Fulkerson, J. A. binge eating. International Journal of Eating Disorders, 34,
(1995). Dieting behaviors and weight change history in 183–​197.
female adolescents. Health Psychology, 14, 548–​555. Hagan, M. M., Wauford, P. K., Chandler, P. C., Jarrett, L. A.,
Friederich, H. C., Kumari, V., Uher, R., Riga, M., Schmidt, U., Rybak, R. J., & Blackburn, K. (2002). A new animal model
Campbell, I. C.,  . . .  Treasure, J. (2006). Differential moti- of binge eating: Key synergistic role of past caloric restriction
vational responses to food and pleasurable cues in anorexia and stress. Physiology and Behavior, 77, 45–​54.
nervosa and bulimia nervosa:  A  startle reflex paradigm. Harvey, J., Wing, R. R., & Mullen, M. (1993). Effects on food
Psychological Medicine, 36, 1327–​1335. cravings of a very low calorie diet or a balanced, low calorie
Fuhrer, D., Zysset, S., & Stumvoll, M. (2008). Brain activity in diet. Appetite, 21, 105–​115.
hunger and satiety: An exploratory visually stimulated fMRI Heatherton, T. F., Herman, C. P., Polivy, J., King, G. A., &
study. Obesity, 16, 945–​950. McGree, S. T. (1988). The (mis)measurement of restraint: An
Galanti, K., Gluck, M. E., & Geliebter, A. (2007). Test meal analysis of conceptual and psychometric issues. Journal of
intake in obese binge eaters in relation to compulsivity and Abnormal Psychology, 97, 19–​28.
impulsivity. International Journal of Eating Disorders, 40, Heatherton, T. F., & Polivy, J. (1992). Chronic dieting and
727–​732. eating disorders:  A  spiral model. In J. H. Crowther, D. L.
Geha, P. Y., Aschenbrenner, K., Felsted, J., O’Malley, S. S., & Tennenbaum, S. E. Hobfold, & M. A. Parris (Eds.), The eti-
Small, D. M. (2013). Altered hypothalamic response to ology of bulimia nervosa: The individual and familial context
food in smokers. American Journal of Clinical Nutrition, (pp. 133–​155). Washington, DC: Hemisphere.
97, 15–​22. Heilbronn, L. K., de Jonge, L., Frisard, M. I., DeLany, J. P.,
Gendall, K. A., Joyce, P. R., Sullivan, P. F., & Bulik, C. M. Larson-​Meyer, D. E., Rood, J.,   .  .  .  Ravussin, E. (2006).
(1998). Food cravers:  Characteristics of those who binge. Effect of 6-​month calorie restriction on biomarkers of lon-
International Journal of Eating Disorders, 23, 353–​360. gevity, metabolic adaption, and oxidative stress in overweight
Goldfield, G. S., & Legg, C. (2006). Dietary restraint, anxiety, individuals:  A  randomized controlled trial. JAMA, 295,
and the relative reinforcing value of snack food in non-​obese 1539–​1548.
women. Eating Behaviors, 7, 323–​332. Herman, P., & Poliy, J. (1975). Anxiety, restraint, and eating
Goldstone, A. P., Prechtl de Hernandez, C. G., Beaver, J. D., behavior. Journal of Abnormal Psychology, 66–​72.
Muhammed, K., Croese, C., Bell, G.,  . . .  Bell, J. D. (2009). Hetherington, M. M., Stoner, S. A., Andersen, A. E., & Rolls,
Fasting biases brain reward systems towards high-​ calorie B. J. (2000). Effects of acute food deprivation on eating
foods. European Journal of Neuroscience, 30, 1625–​1635. behavior in eating disorders. International Journal of Eating
Goodrick, G. K., Poston, W. S., Kimball, K. T., Reeves, R. S., & Disorders, 28, 272–​283.
Foreyt, J. P. (1998). Nondieting versus dieting treatments for Hill, A. J., Weaver, C. F., & Blundell, J. (1991). Food craving,
overweight binge-​eating women. Journal of Consulting and dietary restraint and mood. Appetite, 17, 187–​197.
Clinical Psychology, 66, 363–​368. Hinshaw, S. P. (2002). Intervention research, theoretical mecha-
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. (1998). nisms, and causal processes related to externalizing behavior
Measuring individual differences in implicit cognition: The patterns. Development and Psychopathology, 14, 789–​818.
Implicit Association Test. Journal of Personality and Social Hofmann, W., Rauch, W., & Gawronski, B. (2007). And deplete
Psychology, 74, 1464–​1480. us not into temptation: Automatic attitudes, dietary restraint,
Grilo, C. M., & Masheb, R. M. (2005). A randomized con- and self-​regulatory resources as determinants of eating behav-
trolled comparison of guided self-​help cognitive behavioral ior. Journal of Experimental Social Psychology, 43, 497–​504.
therapy and behavioral weight loss for binge eating disorder. Holsen, L., Savage, C., Martin, L., Bruce, A., Lepping, R., Ko,
Behaviour Research and Therapy, 43, 1509–​1525. E.,  . . .  Goldstein, J. M. (2012). Importance of reward and
Grilo, C. M., Masheb, R. M., Wilson, G. T., Gueorguieva, R., & prefrontal circuitry in hunger and satiety: Prader-​Willi syn-
White, M. A. (2011). Cognitive behavioral therapy, behav- drome vs simple obesity. International Journal of Obesity, 36,
ioral weight loss, and sequential treatment for obese patients 638–​647.
with binge eating disorder:  A  randomized controlled trial. Jacobs, S. B., & Wagner, M. K. (1984). Obese and nonobese
Journal of Consulting and Clinical Psychology, 79, 675–​685. individuals:  Behavioral and personality characteristics.
Groesz, L. M., & Stice, E. (2007). An experimental test of the Addictive Behaviors, 9, 223–​226.
effects of dieting on bulimic symptoms: The impact of eating Jansen, A. (1996). How restrained eaters perceive the amount
episode frequency. Behaviour Research and Therapy, 45, 49–​62. they eat. British Journal of Clinical Psychology, 35, 381–​392.
Guarda, A. S., Coughlin, J. W., Cummings, M., Marinilli, A., Jansen, A. (1998). A learning model of binge eating: Cue reactiv-
Haug, N., Boucher, M., & Heinberg, L. J. (2004). Chewing ity and cue exposure. Behaviour Research and Therapy, 36,
and spitting in eating disorders and its relationship to binge 257–​272.
eating. Eating Behaviors, 5, 231–​239. Jansen, A., & VandenHout, M. (1991). On being led into temp-
Guerrieri, R., Nederkoorn, C., & Jansen, A. (2007). How impul- tation: Counterregulation of dieters after smelling a preload.
siveness and variety influence food intake in a sample of Addictive Behaviors, 16, 247–​253.
healthy women. Appetite, 48, 119–​122. Jansen, A., Merckelbach, H., Oosterlaan, J., Tuiten, A., & van
Guss, J. L., & Kissileff, H. R. (2000). Microstructural analyses of den Hout, M. (1988). Cognitions and self-​ talk during
human ingestive patterns: From description to mechanistic food intake of restrained and unrestrained eaters. Behaviour
hypotheses. Neuroscience Biobehavioral Review, 24, 261–​268. Research and Therapy, 26, 393–​398.

Stice, Shaw 149
Jansen, A., Theunissen, N., Slechten, K., Nederkoorn, C., Boon, LaChaussee, J. L., Kissileff, H. R., Walsh, B. T., & Hadigan, C.
B., Mulkens, S., & Roefs, A. (2003). Overweight children over- M. (1992). The single item meal as a measure of binge-​eating
eat after exposure to food cues. Eating Behaviors, 4, 197–​209. behavior in patients with bulimia nervosa. Physiology and
Jeffery, R., Drewnowski, A., Epstein, L. H., Stunkard, A. Behavior, 51, 593–​600.
J., Wilson, G. T., Wing, R. R., & Hill, D. (2000). Long-​ Lappalainen, R., Sjoden, P., Hursti, T., & Vesa, V. (1990). Hunger/​
term maintenance of weight loss:  Current status. Health craving responses and reactivity to food stimuli during fasting
Psychology, 19, 5–​16. and dieting. International Journal of Obesity, 14, 679–​688.
Johnson, W. G. (1974). Effect of cue prominence and subject Legenbauer, T., Vogele, C., & Ruddel, H. (2004). Anticipatory
weight on human food-​ directed performance. Journal of effects of food exposure in women diagnosed with bulimia
Personality and Social Psychology, 29, 843–​848. nervosa. Appetite, 42, 33–​40.
Kane, T. A., Loxton, N. J., Staiger, P. K., & Dawe, S. (2004). LeGoff, D., Leichner, P., & Spigelman, M. (1988). Salivary
Does the tendency to act impulsively underlie binge eat- responses to olfactory food stimuli in anorexics and bulim-
ing and alcohol use problems? An empirical investigation. ics. Appetite, 11, 15–​25.
Personality and Individual Differences, 36, 83–​94. Leidy, H., Lepping, R., Savage, C., Harris, C. (2011). Neural
Karhunen, L. J., Lappalainen, R. I., Tammela, L., Turpeinen, A. responses to visual food stimuli after a normal vs. higher
K., & Uusitupa, M. I. (1997). Subjective and physiological protein breakfast in breakfast-​skipping teens: A pilot fMRI
cephalic phase responses to food in obese binge-​eating women. study. Obesity, 19, 2019–​2025.
International Journal of Eating Disorders, 21, 321–​328. Levine, M. P., & Smolak, L. (2006). The prevention of eating
Kaye, W., Gendall, K., & Strober, M. (1998). Serotonin neuro- problems and eating disorders:  Theory, research, and practice.
nal function and selective serotonin reuptake inhibitor treat- Mahwah, NJ: Erlbaum.
ment in anorexia and bulimia nervosa. Biological Psychiatry, Lichtman, S. W., Pisarska, K., Berman, E. R., & Pestone, M.
44, 825–​838. (1992). Discrepancy between self-​reported and actual caloric
Killen, J. D., Hayward, C., Wilson, D. M, & Taylor, C. B., intake and exercise in obese subjects. New England Journal of
Hammer, L. D., Litt, I.   .  .  .  Haydel, F. (1994). Factors Medicine, 327, 1893–​1898.
associated with eating disorder symptoms in a community Logue, A. W., & King, G. R. (1991). Self-​control and impulsive-
sample of 6th and 7th grade girls. International Journal of ness in adult humans when food is the reinforcer. Appetite,
Eating Disorders, 15, 357–​367. 17, 105–​120.
Killen, J. D., Taylor, C. B., Hayward, C., Haydel, K. F., Wilson, Lowe, M. R. (1992). Staying on versus going off a diet: Effects
D. M., Hammer, L.,  . . .  Strachowski, D. (1996). Weight on eating in normal weight and overweight individuals.
concerns influence the development of eating disorders: A 4-​ International Journal of Eating Disorders, 12, 417–​424.
year prospective study. Journal of Consulting and Clinical Lowe, M. R. (1994). Putting restrained and unrestrained non-
Psychology, 64, 936–​940. dieters on short-​ term diets—​ effects on eating. Addictive
Kishinevsky, F., Cox, J., Murdaugh, D., Stoeckel, L., Cook, E., Behaviors, 19, 349–​356.
& Weller, R. (2012). fMRI reactivity on a delay discount- Lowe, M. R., Foster, G. D., Kerzhnerman, I., Swain, R. M.,
ing task predicts weight gain in obese women. Appetite, 58, & Wadden, T. A. (2001). Restrictive dieting vs. “undiet-
582–​592. ing”:  Effects on eating regulation in obese clinic attenders.
Klein, D. A., Boudreau, G. S., Devlin, M. J., & Walsh, B. T. Addictive Behaviors, 26, 253–​266.
(2005). Use of artificial sweetened products in eating disor- Lowe, M. R., & Kral, T. V. E. (2006). Stress-​induced eating in
ders. Appetite, 42, 374. restrained eaters may not be caused by stress or restraint.
Klein, D. A., Boudreau, G. S., Devlin, M. J., & Walsh, B. T. Appetite, 46, 16–​21.
(2006). Artificial sweetener use among individuals with eat- Lowe, M. R., & Levine, A. S. (2005). Eating motives and the
ing disorders. International Journal of Eating Disorders, 39, controversy over dieting: Eating less than needed versus less
341–​345. than wanted. Obesity Research, 13, 797–​806.
Klem, M. L., Wing, R. R., Simkin-​Silverman, L., & Kuller, L. Loxton, N. J., & Dawe, S. (2006). Reward and punishment
H. (1997). The psychological consequences of weight gain sensitivity in dysfunctional eating and hazardous drinking
prevention in healthy, premenopausal women. International women: Associations with family risk. Appetite, 47, 361–​371.
Journal of Eating Disorders, 21, 167–​174. Loxton, N. J., & Dawe, S. (2007). How do dysfunctional eat-
Klesges, R. C., Isbell, T. R., & Klesges, L. M. (1992). Relationship ing and hazardous drinking women perform on behavioural
between dietary restraint, energy intake, physical activity, measures of reward and punishment sensitivity? Personality
and body weight: A prospective analysis. Journal of Abnormal and Individual Differences, 42, 1163–​1172.
Psychology, 101, 668–​674. Marlatt, G. A., & Gordon, J. R. (1985). Relapse preven-
Klesges, R. C., Klem, M. L., & Bene, C. R. (1989). Effects of tion:  Maintenance strategies in the treatment of addictive
dietary restraint, obesity, and gender on holiday eating behaviors. New York, NY: Guilford Press.
behavior and weight gain. Journal of Abnormal Psychology, Martin, L. E., Holsen, L. M., Chambers, R. J., Bruce, A. S.,
98, 499–​503. Brooks, W. M., Zarcone, J. R.,  . . .  Savage, C. R. (2010).
Kovacs, D., Mahon, J., & Palmer, R. L. (2002). Chewing Neural mechanisms associated with food motivation in obese
and spitting out food among eating-​ disordered patients. and healthy weight adults. Obesity, 18, 254–​260.
International Journal of Eating Disorders, 32, 112–​115. Martin, C. K., Makris, A., Brill, C., Stein, R., Bailer, B.,
Kraemer, H. C., Stice, E., Kazdin, A., Offord, D., & Kupfer, D. Rosenbaum, D.,  . . .  Foster, G. (2006). Restriction of cer-
(2001). How do risk factors work? Mediators, moderators, tain types of foods during low calorie/​low fat vs. low carbo-
independent, overlapping, and proxy risk factors. American hydrate diets results in decreased cravings and preferences for
Journal of Psychiatry, 158, 848–​856. restricted foods. Obesity Research, 14, A1–​A318.

150 Dieting
Martin, C. K., O’Neil, P. M., & Pawlow, L. (2006). Changes in eating disorders in a longitudinal study of adolescents: How
food cravings during low-​calorie and very-​low-​calorie diets. do dieters fare 5 years later? Journal of the American Dietetic
Obesity, 14, 115–​121. Association, 106, 559–​568.
Martin, C. K., Williamson, D. A., Geiselman, P. J., Walden, Ogawa, R., Strader, A. D., Clegg, D. J., Sakai, R. R., Seeley,
H., Smeets, M., Morales, S., & Redmann, S. Jr. (2005). R. J., & Woods, S. C. (2005). Chronic food restriction and
Consistency of food intake over four eating sessions in the reduced dietary fat:  Risk factors for bouts of overeating.
laboratory. Eating Behaviors, 6, 365–​372. Physiology and Behavior, 86, 578–​585.
Mauler, B. I., Hamm, A. O., Weike, A. I., & Tuschen-​Caffier, Ouwens, M. A., van Strien, T., & van der Staak, C. P. F. (2003).
B. (2006). Affect regulation and food intake in bulimia ner- Tendency toward overeating and restraint as predictors of
vosa:  Emotional responding to food cues after deprivation food consumption. Appetite, 40, 291–​298.
and subsequent eating. Journal of Abnormal Psychology, 115, Pannacciulli, N., Del Parigi, A., Chen, K., Le, D. S., Reiman, E.
567–​579. M., & Tataranni, P. A. (2006). Brain abnormalities in human
McCrory, M. A., Suen, V. M., & Roberts, S. B. (2002). obesity: A voxel-​based morphometric study. Neuroimage, 31,
Biobehavioral influences on energy intake and adult weight 1419–​1425.
gain. Journal of Nutrition, 132, 3830S–​3834S. Patterson, T. A., Brot, M. D., Zavish, A., Schenk, J. O., Snot,
Mitchell, J. E., Pyle, R. L., Eckert, E. D., Hatsukami, D., P., & Figlewicz, D. P. (1998). Food deprivation decreases
Pomeroy, C., & Zimmerman, R. (1990). A comparison mRNA and activity of the rat dopamine transporter.
study of antidepressants and structured intensive group psy- Neuroendrocrinology, 68, 11–​20.
chotherapy in the treatment of bulimia nervosa. Archives of Patton, G. C., Selzer, R., Coffey, C., Carlin, J. B., & Wolfe, R.
General Psychiatry, 47, 149–​157. (1999). Onset of adolescent eating disorders:  Population
Morgan, K. J., Zabik, M. E., & Stampley, G. L. (1986). The role based cohort study over 3 years. BMJ, 318, 765–​768.
of breakfast in diet adequacy of the US adult population. Pauli-​Pott, U., Albayrak, O., Hebebrand, J., & Pott, W. (2010).
Journal of the American College of Nutrition, 5, 551–​563. Does inhibitory control capacity in overweight and obese
Munsch, S., Biedert, E., Meyer, A., Michael, T., Schlup, B., children and adolescents predict success in a weight reduc-
Tuch, A., & Margraf, J. (2007). A randomized comparison tion program? European Child and Adolescent Psychiatry, 19,
of cognitive behavioral therapy and behavioral weight loss 135–​141.
treatment for overweight individuals with binge eating dis- Pearlstein, T., Spurell, E., Hohlstein, L. A., Gurney, V., Read,
order. International Journal of Eating Disorders, 40, 102–​113. J., Fuchs, C., & Keller, M. B. (2003). A double-​ blind,
Nasser, J. A., Gluck, M. E., & Geliebter, A. (2004). Impulsivity placebo-​controlled trial of fluvoxamine in binge eating dis-
and test meal intake in obese binge eating women. Appetite, order: A high placebo response. Archives of Women’s Mental
43, 303–​307. Health, 6, 147–​151.
National Task Force on the Prevention and Treatment of Obesity. Pelchat, M. L. (1997). Food cravings in young and elderly adults.
(2000). Dieting and the development of eating disorders in Appetite, 28, 103–​113.
overweight and obese adults. Archives of Internal Medicine, Penas-​Lledo, E., & Waller, G. (2001). Bulimic psychopathol-
160, 2581–​2589. ogy and impulsive behaviors among nonclinical women.
Nauta, H., Hospers, H., Kok, G., & Jansen, A. (2000). A com- International Journal of Eating Disorders, 29, 71–​75.
parison between a cognitive and a behavioral treatment for Pidcock, B. W., Fischer, J. L., Forthun, L. F., & West, S. L.
obese binge eaters and obese non-​ binge eaters. Behavior (2000). Hispanic and Anglo college women’s risk factors for
Therapy, 31, 441–​461. substance use and eating disorders. Addictive Behavior, 25,
Nederkoorn, C., Braet, C., van Eijs, Y., Tanghe, A., & Jansen, 705–​723.
A. (2006). Why obese children cannot easily resist food: The Placanica, J. L., Faunce, G. J., & Job, R. F. S. (2002). The effect
role of impulsivity. Eating Behaviors, 7, 315–​322. of fasting on attentional biases for food and body shape/​
Nederkoorn, C., Jansen, E., Mulkens, S., & Jansen, A. (2007). weight words in high and low eating disorder inventory
Impulsivity predicts treatment outcome in obese children. scores. International Journal of Eating Disorders, 32, 79–​90.
Behaviour Research and Therapy, 45, 1071–​1075. Polivy, J., Coleman, J., & Herman, C. P. (2005). The effect
Nederkoorn, C., & Jansen, A. (2002). Cue reactivity and regula- of deprivation on food cravings and eating behavior in
tion of food intake. Eating Behaviors, 3, 61–​72. restrained and unrestrained eaters. International Journal of
Nederkoorn, C., Smulders, F. T., Havermans, R. C., Roefs, A., Eating Disorders, 38, 301–​309.
& Jansen, A. (2006). Impulsivity in obese women. Appetite, Polivy, J., & Herman, C. P. (1976). Effects of alcohol on eat-
47, 253–​256. ing behavior: Influence of mood and perceived intoxication.
Nederkoorn, C., Smulders, F. T., & Jansen, A. (2000). Cephalic Journal of Abnormal Psychology, 85, 601–​606.
phase responses, cravings and food intake in normal subjects. Polivy, J., & Herman, C. P. (1985). Dieting and binge eat-
Appetite, 35, 45–​55. ing: A causal analysis. American Psychologist, 40, 193–​204.
Nederkoorn, C., van Eijs, Y., & Jansen, A. (2004). Restrained Polivy, J., & Herman, C. P. (1992). Undieting:  A  program to
eaters act on impulse. Personality and Individual Differences, help people stop dieting. International Journal of Eating
37, 1651–​1658. Disorders, 11, 261–​268.
Neumark-​Sztainer, D., Jeffery, R. W., & French, S. A. (1997). Polivy, J., Herman, C. P., & McFarlane, T. (1994). Effects of anxiety
Self-​reported dieting:  How should we ask? What does it on eating: Does palatability moderate distress-​induced overeat-
mean? Associations between dieting and reported energy ing in dieters? Journal of Abnormal Psychology, 103, 505–​510.
intake. International Journal of Eating Disorders, 22, 437–​449. Polivy, J., Herman, C. P., & Warsh, S. (1978). Internal and
Neumark-​Sztainer, D., Wall, M., Guo, J., Story, M., Haines, external components of emotionality in restrained and unre-
J., & Eisenberg, M. (2006). Obesity, disordered eating, and strained eaters. Journal of Abnormal Psychology, 87, 497–​504.

Stice, Shaw 151
Pothos, E. N., Hernandez, L., & Hoebel, B. G. (1995). Chronic Schlundt, D. G., Hill, J. O., Sbrocco, T., Pope-​Cordle, J., &
food deprivation decreases extracellular dopamine in the Sharp, T. (1992). The role of breakfast in the treatment of
nucleus accumbens: Implications for a possible neurochemi- obesity:  A  randomized clinical trial. American Journal of
cal link between weight loss and drug abuse. Obesity Research, Clinical Nutrition, 55, 645–​651.
3, 525s–​529s. Schotte, D. E., Cools, J., & McNally, R. J. (1990). Film-​induced
Prentice, A. M., Black, A. E., Coward, W. A., Davies, H. L., negative affect triggers overeating in restrained eaters. Journal
Goldberg, G. R., Murgatroyd, P. R.,  . . .  Whitehead, R. G. of Abnormal Psychology, 99, 317–​320.
(1986). High-​levels of energy-​expenditure in obese women. Seeyave, D., Coleman, S., Appugliese, D., Corwyn, R., Bradley,
British Medical Journal, 292, 983–​987. R., Davidson, N.,   .  .  .  Lumeng, J. C. (2009). Ability to
Presnell, K., & Stice, E. (2003). An experimental test of the effect delay gratification at age 4 years and risk for overweight at
of weight-​loss dieting on bulimic pathology:  Tipping the age 11  years. Archives of Pediatric and Adolescent Medicine,
scales in a different direction. Journal of Abnormal Psychology, 163, 303–​308.
112, 166–​170. Serdula, M. K., Collins, M. E., Williamson, D. F., Anda, R. F.,
Presnell, K., Stice, E., & Tristan, J. (2007). An experimental Pamuk, E., & Byter, T. E. (1993). Weight control practices
investigation of the effects of naturalistic dieting on bulimic of United-​States adolescents and adults. Annals of Internal
symptoms:  Moderating effects of depressive symptoms. Medicine, 119, 667–​671.
Appetite, 50, 91–​101. Smolak, L., Levine, M., & Schermer, F. (1998). A controlled
Raynor, H. A., & Epstein, L. H. (2003). The relative-​reinforcing evaluation of an elementary school primary prevention pro-
value of food under differing levels of food deprivation and gram for eating problems. Journal of Psychosomatic Research,
restriction. Appetite, 40, 15–​24. 44, 339–​353.
Redman, L. M., Martin, C. K., Williamson, D. A., & Ravussin, Spencer, J. A., & Fremouw, W. J. (1979). Binge eating as a
E. (2008). Effect of caloric restriction in non-​obese humans function of restraint and weight classification. Journal of
on physiological, psychological, and behavioral outcomes. Abnormal Psychology, 88, 262–​267.
Physiology and Behavior, 94, 643–​648. Spiegel, T. A., Shrager, E. E., & Steller, E. (1989). Responses of
Reeves, R. S., McPherson, R. S., Nichaman, M. Z., Harrist, R. lean and obese subjects to preloads, deprivation, and palat-
B., Foreyt, J. P., & Goodrick, G. K. (2001). Nutrient intake ability. Appetite, 13, 45–​69.
of obese female binge eaters. Journal of the American Dietetic Staiger, P., Dawe, S., & McCarthy, R. (2000). Responsivity
Association, 101, 209–​215. to food cues in bulimic women and controls. Appetite,
Rodin, J., Mancuso, J., Granger, J., & Nelbach, E. (1991). Food 35, 27–​33.
cravings in relation to body mass index, restraint, and estra- Stewart, D. A., Carter, J. C., Drinkwater, J., Hainsworth, J., &
diol levels:  A  repeated measures study in healthy women. Fairburn, C. G. (2001). Modification of eating attitudes and
Appetite, 17, 177–​185. behavior in adolescent girls: A controlled study. International
Roefs, A., Herman, C. P., MacLeod, C. M., Smulders, F. T., & Journal of Eating Disorders, 29, 107–​118.
Jansen, A. (2005). At first sight:  How do restrained eaters Stice, E. (2002). Risk and maintenance factors for eating pathol-
evaluate high-​fat palatable foods? Appetite, 44, 103–​114. ogy: A meta-​analytic review. Psychological Bulletin, 128, 825–​848.
Roefs, A., & Jansen, A. (2002). Implicit and explicit attitudes Stice, E., & Agras, W. S. (1998). Predicting onset and cessation
toward high-​ fat foods in obesity. Journal of Abnormal of bulimic behaviors during adolescence:  A  longitudinal
Psychology, 111, 517–​521. grouping analysis. Behavior Therapy, 29, 257–​276.
Rogers, P. J., & Hill, A. J. (1989). Breakdown of dietary restraint Stice, E., Agras, W. S., & Hammer, L. (1999). Factors influenc-
following mere exposure to food stimuli: Interrelationships ing the onset of childhood eating disturbances:  A  five-​year
between restraint, hunger, salivation, and food intake. prospective study. International Journal of Eating Disorders,
Addictive Behaviors, 14, 387–​397. 25, 375–​387.
Rolls, B. J., Castellanos, V. H., Shide, D. J., Miller, D. L., Stice, E., Burger, K., & Yokum, S. (2013). Caloric deprivation
Pelkman, C. L., Thorwart, M. L., & Peters, J. C. (1997). increases responsivity of attention and reward regions to
Sensory properties of a nonabsorbable fat substitute did intake, anticipated intake, and images of palatable foods.
not affect regulation of energy intake. American Journal of NeuroImage, 67, 322–​330.
Clinical Nutrition, 65, 1375–​1383. Stice, E., Burger, K., & Yokum, S. (2015). Reward region respon-
Ryden, A., Sullivan, M., Torgerson, J. S., Karlsson, J., Lindroos, sivity predicts future weight gain and moderating effects of
A. K., & Taft, C. (2003). Severe obesity and personal- the TaqIA allele. Journal of Neuroscience, 35, 10316–​10324.
ity:  A  comparative controlled study of personality traits. Stice, E., Cameron, R., Killen, J. D., Hayward, C., & Taylor,
International Journal of Obesity, 27, 1534–​1540. C. B. (1999). Naturalistic weight reduction efforts prospec-
Saelens, B. E., & Epstein, L. H. (1996). The reinforcing value of tively predict growth in relative weight and onset of obesity
food in obese and non-​obese women. Appetite, 27, 41–​50. among female adolescents. Journal of Consulting and Clinical
Santonastaso, P., Friederici, S., & Favaro, A. (1999). Full and partial Psychology, 67, 967–​974.
syndromes in eating disorders:  A  1-​year prospective study of Stice, E., Cooper, J. A., Schoeller, D. A., Tappe, K., & Lowe,
risk factors among female students. Psychopathology, 32, 50–​56. M. R. (2007). Are dietary restraint scales valid measures of
Schachter, S., Goldman, R., & Gordon, A. (1968). Effects of moderate-​to long-​term dietary restriction? Objective biolog-
fear, food deprivation, and obesity on dieting. Journal of ical and behavioral data suggest not. Psychological Assessment,
Personality and Social Psychology, 10, 91–​97. 19, 449–​458.
Schlam, T. R., Wilson, N. L., Shoda, Y., Mischel, W., & Ayduk, Stice, E., Davis, K., Miller, N. P., & Marti, C. N. (2008).
O. (2013). Preschoolers’ delay of gratification predicts their Fasting increases risk for onset of binge eating and bulimic
body mass 30 years later. Journal of Pediatrics, 162, 90–​93. pathology:  A  5-​year prospective study. Journal of Abnormal
doi:10.1016/​j.jpeds.2012.06.049 Psychology, 117, 941–​946.

152 Dieting
Stice, E., Fisher, M., & Lowe, M. R. (2004). Are dietary restraint Stunkard, A. J., & Messick, S. (1985). The three-​factor eating
scales valid measures of acute dietary restriction? Unobtrusive questionnaire to measure dietary restraint, disinhibition and
observational data suggest not. Psychological Assessment, hunger. Journal of Psychosomatic Research, 29, 71–​83.
16, 51–​59. Sunday, S. R., & Halmi, K. A. (1990). Taste perceptions and
Stice, E., Gau, J., Rohde, P., & Shaw, H. (2016). Risk factors hedonics in eating disorders. Physiology and Behavior, 113, 173.
that predict future onset of each DSM-​ 5 eating disor- Sunday, S. R., & Halmi, K. A. (1996). Micro-​and macroanalyses
der:  Predictive specificity in high-​risk adolescent females. of patterns within a meal in anorexia and bulimia nervosa.
Journal of Abnormal Psychology, 126, 38–​51. Appetite, 26, 21–​36.
Stice, E., Marti, N., & Durant, S. (2011). Risk factors for onset Sundgot-​Borgen, J., Rosenvinge, J. H., Bahr, R., & Schneider,
of eating disorders: Evidence of multiple risk pathways from L. S. (2002). The effect of exercise, cognitive therapy, and
an 8-​year prospective study. Behaviour Research and Therapy, nutritional counseling in treating bulimia nervosa. Medicine
49, 622–​627. and Science in Sports and Exercise, 34, 190–​195.
Stice, E., Marti, C., & Rohde, P. (2013). Prevalence, incidence, Sutin, A., Ferrucci, L., Zonderman, A., & Terracciano, A. (2011).
impairment, and course of the proposed DSM-​5 eating dis- Personality and obesity across the adult life span. Journal of
order diagnoses in an 8-​year prospective community study of Personality and Social Psychology, 101, 579–​592.
young women. Journal of Abnormal Psychology. doi:10.1037/​ Swanson, S., Crow, S., Le Grange, D., Swendsen, J., &
a0030679 Merikangas, K. (2011). Prevalence and correlates of eat-
Stice, E., Marti, N., Spoor, S., Presnell, K., & Shaw, H. (2008). ing disorders in adolescents:  Results from the national
Dissonance and healthy weight eating disorder prevention comorbidity survey replication adolescent supplement.
programs:  Long-​term effects from a randomized efficacy Archives of General Psychiatry, 68, 714–​723. doi:10.1001/​
trial. Journal of Consulting and Clinical Psychology, 76, archgenpsychiatry.2011.22
329–​340. Sysko, R., Walsh, T. B., Schebendach, J., & Wilson, G. T.
Stice, E., Palmrose, C., & Burger, K. (2015). Elevated BMI and (2005). Eating behaviors among women with anorexia ner-
male gender are associated with greater underreporting of vosa. American Journal of Clinical Nutrition, 82, 296–​301.
caloric intake as assessed by doubly labeled water. Journal of Sysko, R., Walsh, B. T., & Wilson, G. T. (2007). Expectancies,
Nutrition, 145, 2412–​2418. dietary restraint, and test meal intake among undergraduate
Stice, E., Presnell, K., Gau, J., & Shaw, H. (2007). Testing media- women. Appetite, 49, 30–​37.
tors of intervention effects in randomized controlled trials: An Tanofsky-​Kraff, M., Cohen, M. L., Yanovski, S. Z., Cox, C.,
evaluation of two eating disorder prevention programs. Theim, K. R., Keil, M.,  . . .  Yanovski, J. A. (2007). A pro-
Journal of Consulting and Clinical Psychology, 75, 20–​32. spective study of psychological predictors of body fat gain
Stice, E., Presnell, K., Shaw, H., & Rohde, P. (2005). among children at high risk for adult obesity. Pediatrics, 117,
Psychological and behavioral risk factors for obesity onset in 1203–​1209.
adolescent girls:  A  prospective study. Journal of Consulting Telch, C. F., & Agras, W. S. (1993). The effect of a very-​low-​
and Clinical Psychology, 73, 195–​202. calorie diet on binge-​eating. Behavior Therapy, 24, 177–​193.
Stice, E., Presnell, K., & Spangler, D. (2002). Risk factors for Telch, C. F., & Agras, W. S. (1996). The effects of short-​term
binge eating onset:  A  prospective investigation. Health food deprivation on caloric intake in eating-​disordered sub-
Psychology, 21, 131–​138. jects. Appetite, 26, 221–​233.
Stice, E., Rohde, P., Shaw, H., & Marti, N. (2013). Efficacy trial Tuschl, R. J., Platte, P., Laessle, R. G., Stichler, W., & Pirke, K.
of a selective prevention program targeting both eating dis- M. (1990). Energy expenditure and everyday eating behav-
orders and obesity among female college students:  1-​and ior in healthy young women. American Journal of Clinical
2-​year follow-​up effects. Journal of Consulting and Clinical Nutrition, 52, 81–​86.
Psychology, 81, 183–​189. Uher, R., Murphy, T., Brammer, M. J., Dalgleish, T., Phillips, M.
Stice, E., Shaw, H., Burton, E., & Wade, E. (2006). Dissonance L., Ng, V. W.,   .  .  .  Treasure, J. (2004). Medial prefrontal
and healthy weight eating disorder prevention pro- cortex activity associated with symptom provocation in eating
grams: A randomized efficacy trial. Journal of Consulting and disorders. American Journal of Psychiatry, 161, 1238–​1246.
Clinical Psychology, 74, 263–​275. Uher, R., Treasure, J., Heining, M., Brammer, M., & Campbell,
Stice, E., Shaw, H., & Marti, N. (2007). A meta-​analytic review I. (2006). Cerebral processing of food-​related stimuli: Effects
of eating disorder prevention programs:  Encouraging find- of fasting and gender. Behavioural Brain Research, 169,
ings. Annual Review of Clinical Psychology, 3, 207–​231. 111–​119.
Stice, E., Spoor, S., Bohon, C., & Small, D. (2008). Relation van Strien, T., Cleven, A., & Schippers, G. (2000). Restraint,
between obesity and blunted striatal response to food is mod- tendency toward overeating, and ice cream consumption.
erated by the TaqIA1 gene. Science, 322, 449–​452. International Journal of Eating Disorders, 28, 333–​338.
Stice, E., Sysko, R., Roberto, C. A., & Allison, S. (2010). Are van Strien, T., Frijters, J. E., van Staveren, W. A., Defares,
dietary restraint scales valid measures of dietary restriction? P. B., & Deurenberg, P. (1986). The predictive validity of
Additional objective behavioral and biological data suggest the Dutch Restrained Eating Scale. International Journal of
not. Appetite, 54, 331–​339. Eating Disorders, 5, 747–​755.
Strauss, J., Doyle, A. E., & Kreipe, R. E. (1994). The paradoxical Vervaet, M., Audenaert, K., & van Heeringen, C. (2003).
effect of diet commercials on reinhibition of dietary restraint. Cognitive and behavioural characteristics are associated with
Journal of Abnormal Psychology, 103, 441–​444. personality dimensions in patients with eating disorders.
Stunkard, A. J., Berkowitz, R. I., Stallings, V. A., & Schoeller, D. European Eating Disorder Review, 11, 363–​378.
A. (1999) Energy intake, not energy output, is a determinant Wadden, T. A., Brownell, K. D., & Foster, G. D. (2002).
of body size in infants 1,2,3. American Journal of Clinical Obesity:  Responding to the global epidemic. Journal of
Nutrition, 69, 524–​530. Consulting and Clinical Psychology, 70, 510–​525.

Stice, Shaw 153
Wadden, T. A., Foster, G. D., & Letizia, K. A. (1994). One-​ Measurement of dietary restraint: Validity tests of four ques-
year behavioral treatment of obesity:  Comparison of mod- tionnaires. Appetite, 48, 183–​192.
erate and severe caloric restriction and the effects of weight Williamson, D., Serdula, M., Anda, R., Levy, A., & Byers, T.
maintenance therapy. Journal of Consulting and Clinical (1992). Weight loss attempts in adults: Goals, duration, and
Psychology, 62, 165–​171. rate of weight loss. American Journal of Public Health, 82,
Wadden, T. A., Foster, G. D., Sarwer, D. B., Anderson, D. 1251–​1257.
A., Gladis, M., Sanderson, R. S.,   .  .  .  Phelan, S. (2004). Wilson, G. T. (2002). The controversy over dieting. In C. G.
Dieting and the development of eating disorders in obese Fairburn & K. D. Brownell (Eds.), Eating disorders and
women: Results of a randomized controlled trial. American obesity:  A  comprehensive handbook (2nd ed., pp. 93–​97).
Journal of Clinical Nutrition, 80, 560–​568. New York, NY: Guilford Press.
Wardle, J., & Beales, S. (1988). Control and loss of control over Wing, R. R., Marcus, M. D., Blair, E. H., & Burton, L. R.
eating:  An experimental investigation. Journal of Abnormal (1991). Psychological responses of obese type II diabetic sub-
Psychology, 97, 35–​40. jects to a very-​low-​calorie-​diet. Diabetes Care, 14, 596–​599.
Wardle, J., Griffith, J., Johnson, F., & Rapoport, L. (2000). Wisniewski, L., Epstein, L. H., Marcs, M. D., & Kaye, W.
Intentional weight control and food choice habits in (1997). Differences in salivary habituation to palatable
a national representative sample of adults in the UK. foods in bulimia nervosa patients and controls. Psychosomatic
International Journal of Obesity, 24, 534–​540. Medicine, 59, 427–​433.
Wardle, J., Waller, J., & Rapoport, L. (2001). Body dissatisfac- Wonderlich, S. A., Connolly, K. M., & Stice, E. (2004).
tion and binge eating in obese women, the role of restraint Impulsivity as a risk factor for eating disordered behav-
and depression. Obesity Research, 9, 778–​787. ior: Assessment implications with adolescents. International
Westenhoefer, J., & Pudel, V. (1993). Pleasure from Journal of Eating Disorders, 36, 172–​182.
food: Importance for food choice and consequences of delib- Yokum, S., Gearhardt, A., Harris, J., Brownell, K., & Stice, E.
erate restriction. Appetite, 20, 246–​249. (2014). Individual differences in striatum activity to food
Westerterp-​Plantenga, M. S., Ijederma, M. J., & Wijckmans-​ commercials predict weight gain in adolescents. Obesity, 22,
Duijsens, N. E. (1996). The role of macronutrient selection in 2544–​2551.
determining patterns of food intake in obese and non-​obese Yokum, S., Ng, J., & Stice, E. (2011). Attentional bias for food
women. European Journal of Clinical Nutrition, 50, 580–​591. images associated with elevated weight and future weight
White, M. A., Whisenhunt, B. L., Williamson, D. A., Greenway, gain: An fMRI study. Obesity, 19, 1775–​1783.
F. L., & Netemeyer, R. G. (2002). Development and vali- Zunker, C., Peterson, C. B., Crosby, R. D, Ciao, L., Engel, S.
dation of the Food-​Craving Inventory. Obesity Research, 10, G., Mitchell, J. E., & Wonderlich, S. A. (2011). Ecological
107–​114. momentary assessment of bulimia nervosa:  Does die-
Williamson, D. A., Martin, C. K., York-​ Crowe, E., Anton, tary restriction predict binge eating? Behavior Research and
S. D., Redman, L. M., Han, H., & Ravussin, E. (2007). Therapy, 49, 714–​717.

154 Dieting
 CH A PT E R


Mood, Emotions, and Eating Disorders
8
Claus Vögele, Annika P. C. Lutz, and E. Leigh Gibson

Abstract
Mood and emotions are intrinsically involved with eating. This chapter discusses basic mechanisms,
findings, and models that help our understanding of the interactions between eating and emotions, in
both clinical and nonclinical populations. The finding that negative affect predicts EDs transdiagnostically,
and that comorbidity with depressive disorders and anxiety disorders is the norm among patients with
EDs suggests that EDs may not necessarily be restricted to domains of eating behavior and body image
but may also be associated with significant difficulties in affective functioning. This chapter reviews
the evidence relating to the notion that EDs are disturbances of mood regulation, in which regulatory
strategies specifically related to eating and the body are used to diminish negative affect associated with
food, body image, or stress.
Key Words:  disinhibition, dopamine, emotion regulation, emotional eating, negative affect, opioid,
restrained eating, reward, serotonin, stress

Introduction and, as a consequence, enhance intake of sweet

As with other fundamental drives that motivate and high-​fat foods.
us to seek pleasurable goals, eating has always been c.  In normal eating, emotions affect
associated with mood and emotion. This chapter eating in congruence with their cognitive and
considers these questions: In what ways can eating motivational features (e.g., food appears more
and emotions be linked, and what are the impli- pleasing during positive moods compared with
cations for understanding eating disorders (EDs)? negative moods).
In a recent review of the links between eating and
Macht used a flow diagram to help understand
emotions, Macht (2008) has proposed a “five-​way
the structure and predictions from the model,
model” with the following components:
which is reproduced here (Figure 8.1). Some of
1.  Emotions aroused by food stimuli affect food these concepts are considered in more detail in the
choice. text that follows, but there are clearly several ways in
2.  Emotions high in arousal or intensity suppress which eating and emotions can interact. For exam-
eating due to incompatible emotional responses. ple, mood could influence food choice via a change
3.  Emotions moderate in arousal or intensity of appetite, or by changing other behavior that con-
affect eating depending on motivations to eat: strains or alters food availability. On the other hand,
alteration of mood may be an outcome—​perhaps
a.  In restrained eating, negative and
even consciously sought—​ of food choice. Thus,
positive emotions enhance food intake due to
moods or emotions could provide internal stimuli
impairment of cognitive control.
or states that elicit beneficial, for example correc-
b.  In emotional eating, negative emotions
tive, food choice. Further, eating a particular food,
elicit the tendency to be regulated by eating

155
 Emotion

yes Preference
Food-induced Control of food
Craving
emotion? choice
Avoidance

no

yes
Intense emotion? Suppression Decreased food intake

no Impairment of
Restrained cognitive eating Increased food intake
eating controls
yes
Eating habits related
to emotions?
Emotional Eating to Increased intake of
no eating regulate emotions sweet, high-fat foods

Emotion features yes Congruent Change of eating congruent

relevant to eating? modulation with emotion features

no
No change of eating

Figure 8.1  Flow diagram showing predicted changes in eating from Macht’s “five-​way model” of interactions between emotions and
eating.  Macht, M. How emotions affect eating: A five-​way model. Appetite, 50, 1–​11. © 2008. Reprinted with permission from the author and
Elsevier.

or combination, can alter emotions via sensory or punishment (fear, anxiety), and the other with
(including hedonic) effects, associated social con- the absence of (expected) reward (anger, frustration)
text, cognitive expectations, changes in appetite, or or punishment (relief ) (Figure 8.2). Clearly, in the
nutritional modulation of brain function, for exam- context of EDs, food may be both a reward and a
ple. These possibilities are discussed in the text that punisher: an object of pleasure or fear (see the sec-
follows. First, we need to consider what is meant by tion “Emotional Responses to Food Cues”).
mood and emotion. In this chapter, both moods and emotions
Mood is typically characterized as a psychologi- are considered in relation to food, as there is evi-
cal arousal state lasting at least several minutes and dence for involvement of both types of affect, and
usually longer, with dimensions related to energy, instances where the distinction is unclear; research
tension, and pleasure (hedonic tone) (Matthews & on food and mood lags behind neuropsycholog-
Deary, 1998; Reid & Hammersley, 1999; Thayer, ical research on mood and emotion (Hammersley
1989). Moods have been distinguished from emo- & Reid, 2008; Small, Zatorre, Dagher, Evans, &
tions, in that emotions can be defined as short-​term Jones-​Gotman, 2001). The term “affect” is meant
affective responses to appraisals of particular stim- here to refer to either mood or emotion. Food may
uli, situations, or events having reinforcing poten- alter or induce emotions by rapid sensory stimu-
tial, whereas moods may appear and persist in the lation or relief of hunger, or as a result of cogni-
absence of obvious stimuli, and may be more cov- tive appraisal of the change in internal state or its
ert to observers (Matthews & Deary, 1998; Rolls, expectation, but may also alter mood by slower
2007). However, this distinction has been more the- changes in brain chemistry. In fact, for food to be
oretical than empirical (Fredrickson, 2004). Perhaps maximally rewarding, all stages of processing should
the most relevant definition of emotions to this be intact, from expectation and initial sensory con-
topic is that provided by Rolls (2007): “Emotions tact, via longer orosensory stimulation, to gastro-
are states elicited by rewards and punishers, that is, intestinal and hepatic recognition (Booth, 1994;
by instrumental reinforcers.” Rolls then classifies Hetherington et al., 2013).
different emotions on two dimensions (with inten- Macht’s five-​ way model implicates individ-
sity increasing toward the ends of the axes), one ual differences in the emotion–​eating interaction.
associated with delivery of reward (pleasure, elation) Recent theories of mood and emotion suggest that

156 Mood, Emotions, and Eating Disorders

 S+

Ecstasy

Elation

Pleasure
Rage Anger Frustration Relief

S+ or S+! S– or S–!

Grief Sadness
Apprehension

Fear

Terror

S–

Figure 8.2  Emotions represented on two dimensions, based on the definition of emotions as states elicited by rewards and punishers
(Rolls, 2007): Intensity increases away from the center of the diagram. The vertical axis describes emotions associated with the delivery of
a reward (up: S+) or punisher (down: S–​). The horizontal axis describes emotions associated with the nondelivery of an expected reward
(left: S+ = omission; S+! = termination) or the nondelivery of an expected punisher (right: S–​= omission; S–​! = termination).  Reproduced
with permission from Rolls, E. T. [2007]. Emotion explained [Fig. 2.1., p. 14]. Oxford, UK: Oxford University Press.

mood changes do not necessarily follow predictably of negative moods (sad, ashamed, anxious, sleepy)
from changes in neurophysiological arousal. Instead, increased with increasing energy density of the
the affective significance of a given level of arousal foods, and more so for overweight than normal
will depend on the person’s current subjective and weight women. Moreover, medium-​and high-​
motivational state (Reid & Hammersley, 1999). The energy foods were rated less healthy and more dan-
interaction of physiology, arousal, and emotion may gerous than low-​energy foods. These effects were
also be moderated by personality factors, which independent of rated pleasantness of the foods. It
have even been shown to influence effects of stress is most likely that these effects were psychological
on taste perception, for example (Dess & Edelheit, rather than physiological in nature, given the small
1998). In particular, the major personality traits of amounts of food eaten, and the immediacy of the
extraversion and neuroticism are known to moder- ratings. The negative effects of the high-​energy foods
ate mood changes (Matthews & Deary, 1998), and presumably reflect concerns about their impact
to interact with mood and responses to emotional on health and weight gain. Interestingly, though,
stimuli (Canli, Amin, Haas, Omura, & Constable, stronger increases in negative mood were seen
2004). Thus, personality and cognitive factors could for women reporting greater tendencies to eat in
substantially modulate any impact of physiological response to emotional state (see sections “Negative
change induced by food:  Indeed, Macht (2008) Affect, Eating Attitudes, and Comfort Eating”).
points out that the interaction of eating and emo- This would imply that any reinforcing effect of eat-
tions can vary both between and within individuals. ing such foods on prior emotional state must occur
One way in which cognitive factors influence during rather than after eating.
emotional responses to food is via beliefs and expec- These results are similar to the finding that self-​
tations (Hetherington et al., 2013). In a laboratory identified chocolate “addicts” felt guiltier after eat-
study, Macht, Gerer, and Ellgring (2003) asked ing chocolate than did a control group (Macdiarmid
women to rate various emotions immediately after & Hetherington, 1995). The chocolate “addicts”
eating small amounts (5 g) of nine different foods, also reported lower positive and higher nega-
three being low in energy, three medium, and three tive affect before eating. By contrast, in healthy
high in energy (in counterbalanced order). Intensity men, experimental induction of sadness decreased

Vögele, Lutz, Gibson 157

appetite, whereas when they were cheerful, choco- rapidly detected by the brain, as afferent information
late tasted more pleasant and stimulating and they is conveyed by the vagus (tenth cranial) nerve, from
ate more of it (Macht, Roth, & Ellgring, 2002). The the gut and liver. Such vagal afferent activity is now
sex difference might reflect attitudinal differences or known to influence higher brain centers including
culturally influenced differences in acquired coping those involved in emotions (Zagon, 2001):  Even
strategies. simple distention of the stomach by balloon has
been shown, by brain imaging, to activate an area
Basic Mechanisms of the brain, the anterior cingulate cortex, known
In this section, we consider ways in which to underlie depressive symptoms (Faris et al., 2006)
moods and emotions can influence, and be influ- as well as binge eating tendency (Geliebter, Benson,
enced by, eating behavior in a non-​clinical popula- Pantazatos, Hirsch & Carnell, 2016). Indeed, arti-
tion. Underlying neuroendocrine mechanisms are ficial cervical vagal nerve stimulation (VNS) is now
discussed, and then how these basic mechanisms used as a treatment for depression (George et  al.,
might contribute to development of common eat- 2002), following a successful history as a treatment
ing conditions such as comfort eating that could for refractory epilepsy (Morris & Mueller, 1999).
adversely affect mental and physical health. The Moreover, VNS of the splanchnic branch of the
importance of individual differences in susceptibil- vagus nerve has been developed—​ and approved
ity is also considered. by the US Food and Drug Administration—​ as
a treatment for obesity (Apovian et  al., 2016;
General Effects of Hunger and Eating Lebovitz, 2016).
on Emotions: Implications for Eating A link among vagal activity, emotions, and EDs
Disorders has been most extensively developed for bulimia
Perhaps the most reliable way in which food nervosa (BN), where sufferers show increased risk of
interacts with emotional states is the change in anxiety and depression (Braun, Sunday, & Halmi,
mood and arousal that occurs from before to after 1994) (see section “Comorbidity of Affective
eating a meal. Many animals, including humans, Disorders with Eating Disorders” and ­chapter 17).
tend to be aroused, alert, and even irritable when Faris et  al. (2008) have argued that vagal activity
hungry:  hunger is, after all, a strong motivational may be elevated in BN, but with associated desen-
state aimed at encouraging the search for food. sitization of the central nervous system to vagal
Hunger is also a powerful modulator of emotional afferent information, as a result of adaptation to
responses (Rolls, 2007), such that the same events frequent afferent stimulation following prolonged
or stimuli may provoke quite different emotional binge eating and vomiting. This is in line with evi-
responses in one who is hungry versus one who dence of higher vagal activity in relation to dietary
is sated. restriction (Vögele, Hilbert, & Tuschen-​Caffier,
In contrast to the hungry state, after eating a sati- 2009) and results in a higher threshold for satiation
ating meal, we typically become calm, lethargic, and (Halmi & Sunday, 1991), including that mediated
may even sleep. Normal eaters are of course “sat- by the gut hormone cholecystokinin. Vagal afferent
isfied” postprandially, and relief of hunger is typi- involvement is further supported by evidence that
cally a positive and rewarding experience, involving bulimic symptoms are reduced by treatment with
activation of central reward pathways (Grigson, the antiemetic 5-​HT3 (5-​hydroxytryptamine; ser-
2002)—​just as well for the survival of the species. otonin) antagonist, ondansetron, which is known
In support of this, when mood and eating context to inhibit vagal afferent stimulation (Faris et  al.,
were randomly sampled 10 times a day for a week, 2008). Furthermore, this treatment ameliorated the
eating a meal was more likely to result in a posi- depression that otherwise tended to increase cycli-
tive mood than either a neutral or negative mood cally before onset of binge/​vomiting episodes (Faris
(Macht, Haupt, & Salewsky, 2004), at least in the et  al., 2006) (see section “Negative Emotionality/​
short term. Mood as a Maintaining Factor for Disordered Eating
Behavior”). Faris et al. (2008) suggest that the ini-
Vagal Nerve Activity, Bulimia Nervosa, and tially voluntary binge/​purging leads to disruption of
Depression the normal regulation of vagal activity, via vagovagal
A key component of this shift in arousal state is feedback, resulting in a vicious circle of depressive
likely to be postprandial changes in autonomic neu- episodes, weak satiation, and short-​term euphoria
ral activity. During a meal, nutrient absorption is from binge/​purging. Despite this apparent success,

158 Mood, Emotions, and Eating Disorders

however, vagal afferent inhibition does not appear early afternoon in subjects not eating lunch (Smith
to be widely used to treat BN, as opposed to obe- & Miles, 1986). Moreover, susceptibility to this
sity (McClelland, Bozhilova, Campbell, & Schmidt, decline depends on anxiety level: The more anxious
2013). Nevertheless, a recent meta-​ analysis con- one is feeling before lunch, or the more neurotic
cluded that BN is indeed associated with higher the personality trait, the less one will experience any
resting vagal tone (Peschel et al., 2016). post-​lunch dip (Craig, Baer, & Diekmann, 1981;
Smith & Miles, 1986). Given the tendency for EDs
Impact of Meal Size, Timing, and Habit to be associated with higher levels of anxiety, the
on Mood and Emotions inference is that ED patients will be less susceptible
Normal eaters are typically very habitual in their to this afternoon decline in arousal.
choice of food and size and timing of meals:  It is
the essence of their normality. As a result, they have Neural Substrates Shared by Sensory
learned a set of beliefs and expectations about the Reward and Emotions
impact of their habitual dietary regimen, and these Brain pathways involved in food-​related reward
are likely to mitigate any impact of physiological have been dissociated into two functional systems;
changes on emotions. Therefore, to some extent, one underlying motivational aspects of eating and
dietary experiences that differ from a person’s food salience (“wanting”), the other involved in
habitual eating could cause their behavior to change hedonic evaluation of food sensory stimuli (food
through cognitive rather than, or as well as, physi- pleasantness or “liking”). Berridge (2009) points
ological influences. For example, although there is out that this separation of “wanting” from “lik-
some evidence that larger meals may reduce arousal ing” allows for “irrational desires that could under-
and alertness, this effect can depend on the meal lie some pathologies of appetite.” The dopamine,
size being different from that habitually consumed opioid, endocannabinoid, and benzodiazepine/​
(Craig, 1986). In fact, meal size per se seems to gamma-​ aminobutyric acid neurotransmitter sys-
have little impact on mood unless too little is eaten tems dominate these reward processes, with differ-
(Gibson & Green, 2002), whereas Macht (1996) ential effects depending on the particular pathways
found that a larger meal prevented deterioration in and nuclei involved (Berridge, 2009; Berridge &
emotion in people being stressed by noise: this may Robinson, 1998). These systems are also involved in
reflect competing influences of parasympathetic emotions and responses to stress, and findings that
and sympathetic drives, with perhaps suppression link some of them to both eating and emotions are
of the latter by the former after a larger meal than is considered here.
habitually consumed.
One implication of these findings for EDs is that Opioids, Stress, Appetite, and Positive Mood
initial changes in eating habits, for example, meal Endogenous opioid neuropeptides are released
size, may have some impact on emotions, but this during stress, and are known to be important for
impact may soften as the disordered eating becomes adaptive effects such as resistance to pain and induc-
more habitual. For example, short-​term studies in tion of positive mood, possibly by relief from aver-
healthy individuals show that omitting breakfast sive memories (Koepp et  al., 2009). They are also
is likely to increase autonomic activity during the involved in motivational and hedonic processes in
morning and produce cognitive effects reminiscent eating behavior, such as stimulation of appetite by
of increased anxiety (Conners & Blouin, 1983), such palatable foods (Berridge, 2009; Doyle, Berridge,
as attention to irrelevant stimuli (Dusek, Mergler, & & Gosnell, 1993; Mercer & Holder, 1997). One
Kermis, 1976). However, it is not known to what might therefore expect a link between opioid action,
extent adaptation occurs to such effects when break- emotions, and eating behavior. Indeed, in animals
fast is repeatedly avoided. and human infants, the ingestion of sweet and fatty
Arousal levels typically peak during late morn- foods, including milk, alleviates crying and other
ing, followed by a drop in arousal and ability to sus- behavioral signs of distress (Blass, Shide, & Weller,
tain attention after the midday meal that has been 1989; Upadhyay et al., 2004). This effect depends
termed the “post-​lunch dip” (Folkard & Monk, on sweet taste rather than calories, as nonnutritive
1985). However, this dip at least partly reflects an sweeteners also reduce crying (Barr et al., 1999).
underlying circadian rhythm that is confounded This stress-​reducing effect of sweet taste can be
with the effect of a midday meal:  thus, vigilance blocked by opioid antagonists, and opioid analge-
has also been found to decline from late morning to sia can be enhanced by chronic intake of sucrose

Vögele, Lutz, Gibson 159

solutions or fat (Blass et al., 1989; D’Anci, Kanarek, but not the hedonics of “liking” (Berridge, 2009).
& Marks-​Kaufman, 1997; Kanarek, White, Biegen, Moreover, evidence from in vivo microdialysis stud-
& Marks-​Kaufman, 1991). Opioid blockade also ies in rats ingesting sucrose suggests distinct roles
reduces consumption of preferred foods (Kanarek for dopamine release in two areas of the nucleus
et  al., 1991; Yeomans & Wright, 1991), whereas accumbens, the shell and the core (Bassareo et al.,
repeated intake of a sweet, fatty energy-​dense food 2015). That is, dopamine release in the nucleus
was found to down-​ regulate an opioid pathway accumbens core appears to represent unlearned
(ventral striatum) involved in food reward in rats rewarding experience, such as provided by the sweet
(Kelley, Will, Steininger, Zhang, & Haber, 2003). taste of sucrose, whereas dopamine in the shell func-
Animal models of stress-​related binge eating also tions seems to signal the reinforcing value of learned
show that the binge eating is dependent on opioid cues that predict delivery of sucrose and promote
pathways (Boggiano et al., 2005). Intriguingly, ED operant responding for such rewards. Furthermore,
patients who engage in binge eating and purging such nucleus accumbens shell dopamine activa-
behaviors show raised thresholds for pain, which tion appears to suppress core dopamine release,
might suggest up-​ regulation of opioid pathways perhaps preventing impulsive actions and allow-
(and desensitization to vagal afferents) (Papezova, ing more efficient goal-​directed behavior (Bassareo
Yamamotova, & Uher, 2005), and would be in line et  al., 2015). An intriguing possibility is that this
with a drive to binge on palatable food. However, this shell suppression of core dopamine release may be
appears to conflict with evidence above that chronic less effective in people with a tendency to overeat
consumption of such food would down-​regulate palatable foods.
reward-​related opioid pathways. Also, altered opioi- An early influential finding that could be rel-
dergic sensitivity may be an indirect consequence of evant to binge eating disorder (BED) was that the
other neuroendocrine disturbances in EDs, includ- availability of dopamine D2 receptors (DRD2)
ing insulin and leptin (Figlewicz & Benoit, 2009). in the striatum was inversely correlated to body
Nevertheless, disturbed opioid receptor networks mass index (BMI; Wang et  al., 2001), suggesting
in the prefrontal cortex (PFC) and other regions of a neurochemical predisposition to overeat palatable
the limbic and appetite-​regulating areas of the brain foods so as to enhance dopamine release (see sec-
have been proposed to underlie impulsive appetitive tion “Binge Eating: Emotional Responses to Food
behavior that might contribute to disorders such as Cues”). Although this is in line with the “reward
binge eating and bulimia (Baldo, 2016), particularly deficiency syndrome” (Blum et  al., 2016), which
in an obesogenic environment laden with cues to has been used to explain addiction in terms of a
hedonic eating and easy food availability (Giuliano need to correct insufficient dopamine activity in
& Cotone, 2015). brain reward pathways (Wang, Volkow, Thanos, &
The ability of sweet, fatty food to alleviate pain Fowler, 2004), it is difficult to align with abundant
seems to become more idiosyncratic with age behavioral and brain imaging evidence that obese
(Gibson, 2006): indeed, even in infants, after a few people, including those with binge eating tenden-
months, sweet taste becomes less effective at calm- cies or simply overeating and gaining weight, are
ing than does pacifier sucking, which might reflect not only highly motivated to consume palatable
a maturational separation of taste and emotion foods but also experience enhanced pleasure in
(Blass & Camp, 2003), or a difference in opportu- their sensory anticipation and experience (Seymour,
nities to learn the instrumental emotional value of Reinblatt, Benson, & Carnell, 2015; Stice & Yokum,
the two experiences. Thus, in adults, the ability of 2016). For example, energy-​dense snack food rein-
palatable food to alleviate aversive emotional states forced greater effort to obtain it in obese than in
may depend on childhood learning opportunities, nonobese women (Saelens & Epstein, 1996), and
perhaps interacting with biological predispositions. young children of obese parents showed greater
“enjoyment of food” as well as higher preference for
Dopamine, Reward, and Overeating high-​fat, energy-​dense foods than did the offspring
Dopamine, as the dominant neurotransmitter of nonobese parents (Wardle, Guthrie, Sanderson,
in the brain’s mesocorticolimbic system, was for Birch, & Plomin, 2001). Moreover, this “enjoyment
many years considered to be a major substrate of of food” measure is strongly and positively related
pleasure. Nevertheless, the weight of the evidence to adiposity in young children over the entire con-
now supports its role in “wanting” of rewards, or in tinuum of body size (Carnell & Wardle, 2008).
mediating incentive salience of reward-​related cues, Therefore, it seems more probable that DRD2 may

160 Mood, Emotions, and Eating Disorders

be down-​regulated after chronic overconsumption recent study that induced a probable deficit in dopa-
of palatable energy-​dense (e.g., high fat, high sugar) mine in recovered AN patients exposed to pictures
food, akin to the well-​evidenced down-​regulation of emaciated or physically active women found that
seen in chronic use of addictive drugs (Martinez the aberrant appetitive emotional response to those
et al., 2012). Alternatively, the detuned dopamine pictures was absent compared to when dopamine
system may result from neuroendocrine feedback function was not disrupted (O’Hara et al., 2016).
signals of excess adiposity, such as leptin (Figlewicz
& Benoit, 2009). Stress may also be relevant, since Negative Affect, Eating Attitudes, and
it reduces dopamine output in reward pathways, Comfort Eating
and moreover, this effect can be prevented in rats There is now a robust, if complex, literature on
by training to obtain palatable sweet food (Nanni eating as a means to modify emotional states. Its
et  al., 2003). Furthermore, if DRD2 deficiency is origins derive from early psychosomatic and psy-
fundamental to obesity, one would expect BMI to choanalytic clinical models of overeating and obe-
be greater in those known to have lower DRD2 den- sity, based on the notion that obese people may
sity. This hypothesis was recently tested by Benton overeat by confusing emotional arousal with hunger
and Young (2016), who conducted a meta-​analysis and/​or seeking comfort or distraction from emo-
of studies reporting BMI and Taq1A genetic poly- tional distress by eating (Bruch, 1974; Schachter,
morphism variants. People either homozygous or Goldman, & Gordon, 1968). This concept of
heterozygous for the Taq1A A1 allele are known to “comfort eating,” that is, eating to reduce negative
have a 30%–​40% lower density of DRD2 in the emotions, has enjoyed a resurgence of interest, for
striatum (Pohjalainen et  al., 1998) and to be sus- example, in its role in problematic eating and poor
ceptible to substance misuse (Munafò, Matheson, control of weight, perhaps inspired by the Zeitgeist
& Flint, 2007). The meta-​analysis, however, found of the “obesogenic environment” (Chua, Touyz, &
no overall difference in the likelihood of being obese Hill, 2004; Fulkerson, Sherwood, Perry, Neumark-​
between those with or without the A1 allele, imply- Sztainer, & Story, 2004; Gibson, 2012; Waters,
ing that decreased DRD2 density in the striatum Hill, & Waller, 2001). Furthermore, a recent longi-
does not cause obesity. Indeed, given that DRD2 tudinal study of at-​risk female adolescents has con-
receptors can be both autoreceptors and postsynap- firmed that negative affect is a key transdiagnostic
tic receptors, the implications for control of dopa- risk factor for eating disorders (Stice, Gau, Rohde,
mine release and signaling are complex (Gibson, & Shaw, 2016).
2012), and even more so given the evidence of a It is clear from numerous studies, including
role for a dopamine D4 genotype in obesity risk observational, survey, and experimental designs, that
(Stice, Yokum, Bohon, Marti & Smolen, 2010). negative affect, for example, induced by stress, can
Nevertheless, altered dopamine function in interac- adversely affect eating (reviewed by Gibson, 2006,
tion with other neurotransmitter systems, such as 2012; Wardle & Gibson, 2002). The outcome will
the mu opioid system involved in hedonic experi- depend on an interaction between the nature and
ence, and the serotonin system involved in appetite severity of the stressor, or cause of the negative emo-
and affect, remains a likely contributor to risk of tions, stress-​related constraints on behavior, and the
overeating, including binge eating (Baldo, 2016; individual’s propensity (and perhaps learned strat-
Gibson, 2012; Tuominen et al., 2015). egies) for coping. Particularly with severe stress or
Given that EDs are associated with greater anhedonia, eating of any sort may be suppressed.
risk of substance misuse, as well as susceptibility Nevertheless, very often the result is an increase in
to negative affect (Yanovski, Nelson, Dubbert, & consumption of energy-​dense, especially sweet or
Spitzer, 1993) (see section “Clinical Evidence” and fatty foods, or both, and not surprisingly there is
­chapter 17), and that dopamine is also involved in evidence to link stress eating with risk of overweight
stress sensitivity and depression (Pani, Porcella, & (Gibson, 2012; Torres & Nowson, 2007), including
Gessa, 2000), it seems logical to expect abnormal a small but significant predictive effect of stress on
dopaminergic function in EDs. A range of findings weight gain in a meta-​analysis of longitudinal stud-
from various methods including measuring dopa- ies (Wardle, Chida, Gibson, Whitaker & Steptoe,
mine metabolites in cerebrospinal fluid, genetic 2011). Moreover, animal studies suggest that selec-
polymorphisms, and neuroimaging do indeed sup- tion of energy-​dense foods during chronic stress,
port dopamine dysfunction in anorexia nervosa and the associated increase in (central) adiposity,
(AN) and BN (Kaye, 2008; see ­chapter 6). Indeed, a together with hyperinsulinemia, actually help to

Vögele, Lutz, Gibson 161

ameliorate the impact of stress by limiting activity of revised 18-​item Three Factor Eating Questionnaire;
the limbic hypothalamic-​pituitary-​adrenal (LHPA) Karlsson, Persson, Sjostrom, & Sullivan, 2000).
axis (Dallman et  al., 2003; Peters et  al., 2007). Cognitive restraint may still be a risk factor for
Such an effect may reduce stress-​ related depres- overeating during stress, as there is evidence that
sion and anxiety. Activation of the LHPA axis may it may interact with and enhance emotional eat-
also increase incentive salience of cues to reward ing (Haynes, Lee, & Yeomans, 2003; Macht &
(Pecina, Schulkin, & Berridge, 2006), and the over- Müller, 2007a; Williams et  al., 2002). However,
active LHPA axis and other neuroendocrine distur- dietary restraint per se may elicit eating in response
bances seen in EDs (Kaye, 2008) may contribute to challenging situations because of disruption to
to reinforcing the aberrant behaviors (Södersten, cognitive “diet monitoring,” and a focus of atten-
Nergardh, Bergh, Zandian, & Scheurink, 2008). tion on the salient food cues, with little emotional
involvement (Lattimore & Caswell, 2004; Wallis
Restrained Versus Emotional Eating & Hetherington, 2004; Ward & Mann, 2000).
Early evidence suggested a key prerequisite Moreover, Macht (2008) points out that restrained
for overeating during negative affect was to be a eaters can therefore be expected to show increased
“restrained eater” (one who consciously attempts to eating in response to both negative and positive
restrict food intake and eat less than is wanted) or a emotions. These ideas would seem to be reflected
dieter (Heatherton, Herman, & Polivy, 1991; Oliver in the task-​based neuroimaging evidence that recov-
& Wardle, 1999; Weinstein, Shide, & Rolls, 1997). ered AN women show exaggerated activation of
This is consistent with a number of experimental planning areas but no dissociation of positive and
studies, which were based on the premise that the negative feedback (Wagner et al., 2007).
adaptive response to stress should be reduced appe-
tite and eating, but that the obese, overweight, or Susceptibility to Stress and Comfort Eating
restrained eaters may be unresponsive to their inter- Interestingly, in relation to EDs, there is accu-
nal physiological influences on appetite (Craighead mulating evidence that emotional eaters may
& Allen, 1995; Rodin, 1981), or normally domi- be particularly susceptible, both physiologically
nant cognitive strategies for restraint may be disin- and emotionally, to the effects of stress; that is,
hibited or overridden during negative affect or stress they seem to be high stress reactors. For example,
(Herman & Polivy, 1975; Schachter et  al., 1968). women who ate more from a selection of snack
However, the first measure of restrained eating foods after a stressful task also showed the great-
(Herman & Polivy, 1980) appears to have included est release of the stress-​sensitive hormone, corti-
aspects subsequently recognized as disinhibited sol, and more stress-​induced negative affect (Epel,
(uncontrolled) or emotional eating, and by defini- Lapidus, McEwen, & Brownell, 2001). These high
tion this latter concept is probably most relevant reactors also showed a preference for sweet foods. In
to comfort eating (Lowe & Fisher, 1983; Oliver, another experimental study in men and women, the
Wardle, & Gibson, 2000; Williams et al., 2002). In stressful condition induced greater intake of sweet
particular, in an experimental study in which both foods, and to a larger extent, the more anxious the
eating attitudes were measured, Oliver et al. (2000) person was and the higher their disinhibited eat-
found that emotional eaters did not eat more food ing tendency (Rutters, Nieuwenhuizen, Lemmens,
overall under stress, but instead selectively ate more Born, & Westerterp-​Plantenga, 2009). In a large
sweet and fatty foods:  restraint, however, did not survey of French men (n  =  7,378) and women
significantly influence stress eating. Similarly, (n = 22,862), emotional eating was associated with
Gibson and Harris (unpublished data) have recently both greater intake of sweet fatty foods and depres-
replicated survey findings that stress selectively sive symptoms in women, but conversely in men
increases consumption of energy-​dense, especially the association was with the absence of depressive
sweet and fatty, foods (Gibson, 2006; Kandiah symptoms (Camilleri et al., 2014). This sex differ-
et al., 2006; Zellner et al., 2006), and extended it ence in the effect of affective state on emotional eat-
to show that the stress eating tendency is strongly ing is similar to the effect of positive emotion on
predicted by both emotional eating (ρ (135) = .59, promoting chocolate eating in men discussed ear-
p < .001) and uncontrolled (disinhibited) eating (ρ lier (Macht et al., 2002). In an observational study
(135) = .33, p < .001) as well as by the perception of of daily hassles, mood, and snacking, only women
being overweight (ρ (135) = .23, p < .01), but not with high cortisol responses to a laboratory stressor
by cognitive restraint (ρ (135) = .07, ns) (using the reported eating more snacks in association with

162 Mood, Emotions, and Eating Disorders

increased daily hassles (Newman, O’Connor, & low-​TRP-​containing proteins (alphalactalbumin
Conner, 2007). This may well be relevant to binge and casein, respectively) (Markus et  al., 2000). It
eating:  Goldfield, Adamo, Rutherford, and Legg was argued that, because stress increases serotonin
(2008) used a computer-​based food choice task to activity, the poor stress coping of this sensitive
determine whether stress affected choice in binge group might indicate a deficit in serotonin synthesis
eaters. Like the previous studies, only binge eaters that is improved by this dietary intervention. More
who were classified as high stress reactors (based on recently, administration of a TRP-​rich peptide sup-
perceived feelings of negative affect) altered food plement improved well-​being in women undergo-
choice under stress, preferring unhealthy snacks ing stressful laboratory tasks (Gibson et al., 2014).
over fruit and vegetables:  Non-​binge-​eaters were These findings are particularly relevant, because
unaffected in this paradigm (see section “Binge EDs are associated with anxiety, depression, impul-
Eating: Emotional Responses to Food Cues”). sivity, and poor stress coping, all of which have been
Despite this evidence for an apparent “affect linked to serotonin. Further, there is now ample evi-
reduction” model, there is an alternative explana- dence for a dysfunctional serotonin system in brain
tion for a link between stress and sweet fatty food emotional circuits in EDs, including reduced sero-
intake that does not require such food to be com- tonin release, and postsynaptic sensitization, during
forting. Pool, Delplanque, Coppin, and Sander food restriction (Kaye, 2008). Genetic studies of the
(2015) propose that stress may actually reduce the role of the affect-​related 5-​HT transporter-​linked
pleasure of eating highly palatable foods:  instead polymorphic region (5-​ HTTLPR) suggest that
stress amplifies learned motivational and attentional serotonin transporter deficits may underlie these
responses to the presence of such foods (possibly findings (Schepers & Markus, 2015).
dopamine-​mediated). This is at the expense of more
cognitively demanding goal-​dependent control on Chocolate and Emotions: Immediate and
(healthy) eating, and allows our habitual and long-​ Delayed Effects
established (energy-​dense) food preferences to be It is notable that chocolate, often chosen during
evoked. stress, is high in sugar, which, when combined with
Further, this stress susceptibility associated effects of cocoa, strongly stimulates insulin release
with binge eating and comfort eating may sug- (Brand Miller, Holt, de Jong, & Petocz, 2003),
gest another mechanism, which could encourage but has only 3% to 6% of its energy in the form
selection of sweet and fatty foods in an attempt to of protein. Thus, if eaten in sufficient amounts on
regulate emotions, that is, dietary manipulation of an empty stomach, chocolate might increase TRP
the brain serotonin system. Synthesis of the neu- availability to the brain, and so allow improved
rotransmitter serotonin (5-​HT) depends on dietary mood via enhanced serotonin release. Macht (2008)
availability of the precursor essential amino acid, has argued that such a TRP-​dependent mechanism
tryptophan (TRP):  Uptake into the brain of TRP would be too slow acting (probably taking 1 to 2
in turn depends on its ratio to the large neutral, hours after eating). However, it is still possible that
primarily branched-​chain, amino acids (LNAAs), a delayed mood enhancement via increased sero-
with which it competes for transport from blood tonin could reinforce a learned liking for chocolate
to brain. In brief, insulinogenic high-​carbohydrate (or sweet, fatty foods), and a tendency to choose
meals, with little protein, raise this ratio and can it during negative emotions:  Such longer-​ lasting
result in increased serotonin synthesis in the brain effects may be particularly important in establishing
(Fernstrom & Fernstrom, 1995) (see ­chapter 6). disordered eating, such as binge eating, as Macht
In the context of stress susceptibility and (2008) acknowledges. After all, Macht and Mueller
negative emotions, there is some intriguing evi- (2007b) showed that the lessening of film-​induced
dence: When participants were divided into high or negative mood by eating chocolate, which depended
low stress-​prone groups, as defined by a question- on sensory palatability, lasted only a few min-
naire measure of neuroticism, carbohydrate-​rich/​ utes: Even then, this effect was greater in emotional
protein-​poor meals (which raised plasma TRP/​ eaters. Furthermore, a survey of 931 Californian
LNAA ratios) before a stressful task were found to men and women who had completed the Center
block task-​induced depressive feelings and release for Epidemiological Studies Depression Scale found
of the glucocorticoid stress hormone cortisol, but that those with moderate to high depression scores
only in the high stress-​prone group (Markus et al., reported eating far more chocolate than those with
1998). This finding was replicated using high-​versus low depression scores (Rose, Koperski, & Golomb,

Vögele, Lutz, Gibson 163

2010). One could infer that (American) chocolate this comorbidity, however, is unclear. Depression
was ineffective as an antidepressant, but clearly and anxiety may precede, occur simultaneously, or
consumers find something reinforcing in its con- follow the appearance of the ED (Polivy & Herman,
sumption. The mood-​ enhancing actions demon- 2002). It has also been suggested that EDs and
strated for caffeine and theobromine in chocolate mood disorders have common familial causal fac-
(Smit, Gaffan, & Rogers, 2004), together with sen- tors (Hudson et al., 2003; Mangweth et al., 2003).
sory, caloric, and social reinforcement, are clearly In addition, the presence of negative emotionality
also likely to be important mechanisms. Finally, or psychiatric comorbidity in those with EDs may
given the conflict and ambivalence that food can be associated with greater ED psychopathology
create in ED patients, it is notable that images of (Grilo, White, & Masheb, 2009) or medical com-
chocolate can induce simultaneously appetitive plications (Takimoto, Yoshiuchi, & Akabayashi,
and aversive motivational states (assessed by physi- 2008), or both. The overlap of ED and bipolar dis-
ological responses) in chocolate cravers (Rodriguez, order has been argued to be due to both conditions
Fernandez, Cepeda-​Benito, & Vila, 2005). sharing common genetic factors (Winham et  al.,
2014), and/​or to exposure to certain psychotropics
Clinical Evidence (i.e., medication-​induced EDs; e.g., Martin, Han,
In this section, we consider the clinical evi- Anton, Greenway, & Smith, 2009).
dence relating mood and eating disorders, focus-
ing on Bulimia nervosa, Binge eating disorder and Negative Emotionality and Eating
Anorexia nervosa. The first part describes the epi- Disorders: Risk Factor, Retrospective
demiological evidence linking eating disorders with Correlate, or Consequence?
affective disorders, and the evidence concerning It has been argued that the increased co-​
emotion dysregulation in people with eating disor- occurrence of ED and mood disorders (particularly
ders. The second part discusses emotional responses depression) indicates that mood disturbances are a
to food cues and body image exposure, the role of risk factor for EDs. In order to be ascertained as a
negative affect and emotional processing deficits for risk factor, negative emotionality has to precede the
each of the three main eating disorders separately. ED and increase the likelihood of its manifestation.
Negative emotionality as a predictor for EDs has
Comorbidity of Affective Disorders been assessed in several longitudinal studies. Leon,
with Eating Disorders Fulkerson, Perry, Keel, and Klump (1999) used a
Comorbidity with mental disorders is reportedly composite measure of negative affectivity including
the norm among patients with ED (Keski-​Rahkonen negative emotionality, depression, ineffectiveness,
& Mustelin, 2016). For example, mood and anxiety and body dissatisfaction, and found a significant but
disorders have been shown to be extremely common moderate association with ED risk 3 to 4 years later.
in a large clinical cohort (n = 11,588) of Swedish ED In the two studies by Patton and coworkers (Patton,
patients (Ulfvebrand, Birgegard, Norring, Hogdahl, Johnson-​Sabine, Wood, Mann, & Wakeling, 1990;
& von Hausswolff-​Juhlin, 2015), but high levels of Patton, Selzer, Coffey, Carlin, & Wolfe, 1999), psy-
comorbidity between depressive and ED symptoms chiatric morbidity predicted the onset of EDs. In
have also been reported for adolescents, both boys a recent study by Stice et al. (2016) negative affect
and girls (Herpertz-​Dahlmann et  al., 2015; Rojo-​ and functional impairment predicted onset of EDs
Moreno et al., 2015). The most common comorbid across ED diagnostic categories (AN, BN, BED,
disorders are anxiety disorders and mood disorders, and purging disorder) over a 3-​year period, suggest-
followed in frequency by substance use disorders ing negative affect to increase the risk for the full
(Keski-​Rahkonen & Mustelin, 2016). The DSM-​ spectrum of EDs, and possibly other psychopathol-
5-​defined BED and BN are common in patients ogies. In contrast to these results, four studies (Attie
with bipolar disorder, possibly more common & Brooks-​ Gunn, 1989; Graber, Brooks-​ Gunn,
than DSM-​ IV-​
defined BED and BN (McElroy Paikoff, & Warren, 1994; Killen et al., 1996; Leon,
et al., 2016). Fulkerson, Perry, & Early-​Zald, 1995) failed to find
These findings suggest that EDs may not negative emotionality or psychopathology to pre-
be pathologies that are necessarily restricted to dict EDs in multivariate analyses.
domains of eating behavior and body image but Although this evidence seems inconclusive in
may also be related to significant difficulties in affec- terms of a specific role of negative emotionality for
tive functioning. The exact nature and etiology of EDs, it is likely that negative affect increases risk for

164 Mood, Emotions, and Eating Disorders

binge eating and compensatory weight control behav- Negative Emotionality/​Mood as a
iors (Lavender et  al., 2016), which in turn predict Maintaining Factor for Disordered Eating
onset of EDs (Stice et al., 2016). Yet another alterna- Behavior
tive may be that negative affectivity or depression and Several investigators view EDs as disturbances of
EDs share some underlying etiological mechanisms mood regulation that use regulatory strategies spe-
with respect to emotion regulation: A recent review cifically related to eating and the body (e.g., binge
(Donofry, Roecklein, Wildes, Miller, & Erickson, eating, dieting, purging, exercising excessively) as
2016) suggests that depression and EDs exhibit techniques aimed at diminishing unpleasant affect
common structural and functional alterations in (Taylor, Bagby, & Parker, 1997; Telch, 1997). This
brain regions involved in emotion regulation (see seems to be a unifying feature of all ED types, as
next section), including the amygdala, ventral stria- the core symptoms and behaviors are shared, even if
tum, nucleus accumbens, anterior cingulate cortex, the diagnostic criteria differ. Indeed, whether or not
insula, and dorsolateral prefrontal cortex. to distinguish between different types of EDs has
Some cross-​ sectional studies have investigated been debated vigorously in the literature, and some
chronology of onset retrospectively; however, only authors argue that EDs are one syndrome with dif-
two of those included control groups. Rastam ferent manifestations, a view that is summarized
(1992) found premorbid obsessive-​compulsive per- under the “spectrum hypothesis” (Van der Ham,
sonality disorder to be significantly more prevalent Meulman, VanStrien, & van Engeland, 1997).
in patients with AN than in control participants Nevertheless, differences in psychological traits
(35% vs. 4%). Using logistic regression, Bulik, and clinical symptoms exist. Impulsivity is one
Sullivan, Fear, and Joyce (1997) showed that the crucial feature that distinguishes AN on the one
risk for AN increased in the presence of obsessive-​ hand from BN and BED on the other. Increased
compulsive disorder and overanxious disorder rel- impulsivity in those with BN or BED is reflected
ative to controls. The risk for BN was increased in in their relative failure to resist food under certain
the presence of social phobia and overanxious dis- circumstances, and to experience loss-​ of-​
control
order relative to controls. This was true, however, over eating, resulting in a binge. Other examples
also for women with major depression. The authors for increased impulsivity concern sexual promiscu-
conclude from these findings that overanxious dis- ity, suicide attempts, drug abuse, and stealing or
order in childhood is a nonspecific risk factor for shoplifting, which are frequently reported for BN
later psychopathology, albeit the relationship with patients (Matsunaga, Kiriike, Iwasaki, Miyata, &
AN is particularly strong. Similarly, social phobia Matsui, 2000). Individuals with AN (restricting
may precede a range of psychopathologies, but with type), on the other hand, are remarkably success-
a strong link to BN. ful in maintaining restricted food intake over long
In their review, Jacobi, Hayward, de Zwaan, periods of time, so are excessive in their capacity
Kraemer, and Agras (2004) concluded that prior to control and withstand hunger signals. It has
psychiatric morbidity and negative emotionality are been suggested that AN is but one expression of a
variable risk factors of unclear potency. Results from broader phenotype characterized by perfectionism,
longitudinal studies suggest that premorbid anxiety rigidity, and the propensity for behavioral con-
disorders and negative affectivity are risk factors for straint (Lilenfeld et al., 1998). Such differences in
other mental disorders including affective disorders personality traits may also play a role in emotion
and EDs, and should, therefore, be considered as regulation. The literature on mood, emotions, and
nonspecific risk factors. EDs emphasizes the relationship between BN or
Of course, negative emotionality may not cause BED and affect dysregulation on the one hand, and
an ED so much as to be a consequence or main- AN and alexithymia on the other.
taining factor, or both. Stice, Hayward, Cameron,
Killen, and Taylor (2000), for example, found ED Bulimia Nervosa
symptoms and dietary restraint to predict subse- The most salient feature of BN is an abnormal
quent depression in initially nondepressed individu- eating pattern characterized by alternating periods
als. Whichever is first, these findings suggest that of fasting and episodes of binge eating. In addition,
negative affect can contribute to ED symptoms. there is recurrent behavior to prevent weight gain
Several models have been put forward for how neg- including purging directly after binge eating (e.g.,
ative affect impacts on ED symptoms, and these are vomiting or use of laxatives) or excessive exercise.
discussed in the following sections. In his dual pathway model of bulimic pathology,

Vögele, Lutz, Gibson 165

Stice (1994, 2001) hypothesized that elevated pres- Retrospective Correlate, or Consequence?”). That
sure to be thin fosters body dissatisfaction in BN. BN individuals perceive food cues as negative
This increased body dissatisfaction then promotes or threatening, even at a preconscious level (i.e.,
restrained eating and negative affect, especially outside of the individual’s awareness), has been
depression. It is further assumed that individuals demonstrated using the startle probe methodol-
with BN use binge eating and purging as a means ogy:  Mauler, Hamm, Weike, and Tuschen-​Caffier
of regulating negative mood states. From the per- (2006) found persons with BN to respond with
spective of self-​regulation (Hofmann, Rauch, & significantly larger startle responses during viewing
Gawronski, 2007; Mann & Ward, 2004) dieting of food cues compared with control participants,
or fasting can be construed as an attempt to reduce indicating a defensive motivational state. Food
negative affect elicited by cues that are associated deprivation for 24 hours attenuated this response,
with food intake and the fear of gaining weight. but eating before exposure to food cues augmented
These cues can be related to two distinct classes of it. These findings show similarities to the enhanced
stimuli: food cues and body shape. Both have been startle response to chocolate images seen in choco-
investigated in relation to emotional responses and late cravers (who share with BN a struggle to con-
emotion regulation in BN. trol eating) compared with noncravers (Rodriguez
et al., 2005; section “Anorexia Nervosa”).
Emotional Responses to Food Cues These results support the notion that food cues
Emotional responses of individuals with BN to are perceived as threatening by BN individuals
food cues differ from those of healthy study par- and that dieting or purging may be used to avoid
ticipants. When purging was rendered impossible, or reduce the negative affect associated with food
persons with BN responded with increased negative exposure and eating. In other words, individuals
affect after eating compared with healthy control with BN make food less frightening or threatening
participants, with increased fear of gaining weight, by demonstrating control over their consumption of
tenseness, and feelings of being depressed and irri- it. If the restrained eating pattern collapses, that is,
tated (Buree, Papaeorgis, & Hare, 1990; Staiger, if control fails with subsequent eating or binge eat-
Dawe, & McCarthy, 2000). Legenbauer, Vögele, ing, food once again becomes threatening. This pat-
and Rüddel (2004) exposed women with BN and tern of results is consistent with cognitive models of
non-​eating-​disordered participants to their favorite bulimic pathology suggesting that core beliefs such
binge or snack foods, and gave them the opportu- as “control over eating” are central for BN (Cooper,
nity to eat ad libitum after the end of the 20-​minute Wells, & Todd, 2004). Rigid self-​regulation is fun-
exposure trial. Those with BN responded with damental to the self-​esteem of women with BN
increased distress and feelings of tension and inse- (Fairburn, Cooper, & Shafran, 2003), and if this
curity to the exposure, while women in the control self-​regulation is violated the negative affect induced
group showed a decrease in these ratings. In addi- by food cues increases.
tion, women with BN ate less than women in the
control group after exposure but responded with Emotional Responses to Body Image Exposure
increases in feelings of guilt and shame that were Frequently, patients with BN are excessively con-
positively associated with the amount of food eaten. cerned with body weight and shape, and these fac-
With a view to appetite and hunger regulation tors are assumed to be important for self-​evaluation
in non-​eating-​disordered individuals, these results (DSM-​5) and the maintenance of disordered eat-
may seem counterintuitive, as food cues normally ing. Body dissatisfaction may arise from perceptual
have reinforcing, appetitive qualities (see section body size distortion or cognitive-​ evaluative dis-
“Comorbidity of Affective Disorders with Eating satisfaction (Lautenbacher, Roscher, Strian, Pirke,
Disorders”). Indeed, predigestive physiological & Krieg, 1993). While patients with AN and BN
responses (i.e., cephalic phase responses, e.g., sali- overestimate their body size to similar degrees, stud-
vation) to food cues in BN are more marked and, ies on the cognitive-​evaluative aspects of body size
therefore, seem to indicate the opposite pattern generally find that those with BN are dissatisfied
(Legenbauer et  al., 2004; Tuomisto et  al., 1999). with their body and would like to be thinner, even
The BN individuals’ attitude to food is, there- more so than patients with AN (Cash & Deagle,
fore, ambivalent:  desired but feared, or wanted 1997). Nevertheless, there is evidence that this may
but not liked (Berridge, 2009; section “Negative be true for non-​eating-​disordered women as well
Emotionality and Eating Disorders:  Risk Factor, (Goldfein, Walsh, & Midlarsky, 2000). Concerns

166 Mood, Emotions, and Eating Disorders

about body weight and shape are widespread in emission computed tomography to investigate the
adult women ranging in age from 18 to 70  years effects of body shape exposure in BN and AN on
(e.g., Ricciardelli, Tate, & Williams, 1997), in high brain activation. When exposed to a video of their
socioeconomic status sites around the world (Swami own figure, BN individuals responded with hyper-
et  al., 2010) and also adolescent boys and girls activity in brain regions implicated in emotion
(Vögele & Woodward, 2005). processing (i.e., right temporal and occipital areas),
The dual pathway model (Stice, 1994, 2001) suggesting a “fear response.” Increased emotional
stipulates that negative affect (and restrained eat- and self-​referential processing in BN was also sug-
ing) is the consequence of body dissatisfaction and gested by another study, in which viewing over-
mediates the association between body dissatisfac- weight bodies led to a larger activation of the rostral
tion and bulimic eating behaviors. Pinhas, Toner, anterior cingulate cortex in BN than in healthy con-
Ali, Garfinkel, and Stuckless (1999) showed that trol women (Spangler & Allen, 2011).
exposure to slides of women representing the ideal
of female beauty in Western culture led to an imme- Negative Affect and Binge Eating
diate negative effect on mood in a sample of non- Continued dieting can lead to a relative state
clinical university students. These results suggest of deprivation, which increases the incentive value
that internalized images of ideal beauty have a det- of food cues (Baker, Piper, McCarthy, Majeskie,
rimental effect on women and may play a mediat- & Fiore, 2004). This increased incentive value of
ing role (e.g., through negative social comparison food cues together with overall negative affective
processes) between body image exposure and nega- state may trigger binge eating in BN, comparable
tive affect. The question whether exposure to one’s to the motivational basis of drug cues trigger-
own body image leads to negative affect in BN has ing relapse. In addition, body dissatisfaction may
been investigated in only a small number of stud- increase negative affect and thus lead to binge eat-
ies. Tuschen-​Caffier, Vögele, Bracht, and Hilbert ing or purging, or both. There is an increasing body
(2003) confronted BN women and non-​ eating-​ of evidence that such negative mood changes may
disordered female volunteers with their body shape precede binge-​eating episodes. This has been inves-
and body-​related appearance using a video recording tigated in controlled experiments (Agras & Telch,
and an imagery task. In addition, they were asked 1998; Telch & Agras, 1996), and studies monitor-
to describe their physical appearance. Both groups ing food consumption in BN after mood induction
responded with similar increases in self-​ reported have tended to find systematic associations between
anxiety, tension, insecurity, and sadness to both the negative affect and caloric content or amount
video confrontation and imagery task. This suggests of food eaten (e.g., Legenbauer et  al., 2004).
that the confrontation with body shape is a stress- Nevertheless, whether the amounts of food eaten
ful and upsetting experience even for non-​eating-​ would amount to an “objective” binge episode,
disordered women. The groups differed, however, or whether participants’ sense of loss-​ of-​
control
with respect to the time taken for body description. over eating would resemble that experienced in a
The BN individuals took less time to describe their natural environment remains doubtful, probably
waists, hips, and bottoms compared with non-​ because it is notoriously difficult in laboratory set-
eating-​ disordered participants, and the authors tings to re-​create the situational context in which
interpreted this result as avoidance behavior. In line a binge-​eating episode occurs. Researchers have,
with other areas where exposure has been shown therefore, turned increasingly to investigations of
to be a successful treatment approach, also in BN binge eating in the natural environment. Using
repeated body image exposure over several sessions retrospective assessments, some studies have found
results in an improvement in negative emotions that individuals with BN report significantly more
(Trentowska, Bender, & Tuschen-​Caffier, 2013). negative mood in general and prior to a binge epi-
There are an increasing number of studies sode compared with non-​ eating-​
disordered con-
employing neuroimaging techniques to investigate trols (Kjelsas, Borsting, & Gudde, 2004; Mitchell
the brain areas that are implicated in EDs (Uher et  al., 1999; Waters et  al., 2001). Such retrospec-
et  al., 2005). Most of these studies have investi- tive reports are problematic, however, as they are
gated patients diagnosed with AN or mixed groups susceptible to memory bias, and there is evidence
of EDs, indicating differences in brain activation in that retrospective reporting of binges is especially
response to visual food cues or eating. One study inaccurate (Bardone, Krahn, Goodman, & Searles,
(Beato-​Fernández et al., 2009) used single photon 2000). The concurrent assessment of mood changes

Vögele, Lutz, Gibson 167

and eating behavior using ecological momentary In summary, these studies provide convinc-
assessment (EMA) provides a method for the inves- ing evidence that mood and stress in daily life are
tigation of the temporal sequence of negative affect related to binge eating in women with BN, and that
and binge-​eating episodes (with or without subse- purging is used to reduce the negative affect associ-
quent purging), which is less susceptible to retro- ated with food cues and eating. There is conflicting
spective recall effects. Alpers and Tuschen-​Caffier evidence as to the reinforcing effects of binge eat-
(2001) assessed mood, feelings of hunger, and the ing in BN: Alpers and Tuschen-​Caffier (2001) and
desire to eat hourly over 2 consecutive days in BN Hilbert and Tuschen-​Caffier (2007) found mood
patients and a clinical (panic disorder patients) to deteriorate after binge eating, whereas Smyth
and a healthy control group. Overall, both patient et al. (2007) demonstrated a similar pattern of rapid
samples reported more negative feelings than the mood recovery after both binge eating and vomit-
healthy control group. The BN patients rated most ing. The reasons for these discrepancies in findings
feelings more negatively in the hour before binge are not clear, although differences in research design
eating compared with the rest of the day. Also, BN (e.g., assessment time points, mode and frequency
patients’ general mood state worsened after binge of assessment, length of observation period, etc.)
eating but returned to prebinge levels after purg- and sample sizes may have contributed to this sce-
ing. This suggests that negative mood states pre- nario. The former two studies involved smaller sam-
cede binge-​eating episodes in BN, and that mood ples (40 and 20 per group, respectively) and shorter
deteriorates after binge eating but is alleviated by observation periods (2 consecutive days) with more
purging. By including a clinical control group, the frequent assessments (32 assessments over 16 hours)
authors were able to establish the specificity of the than the latter (131 participants, 2 weeks, 6 assess-
association of negative mood states and the desire ments over 13.5 hours). It is conceivable, for exam-
to eat in BN. Using a similar design, Hilbert and ple, that the mood-​enhancing effects of binge eating
Tuschen-​Caffier (2007) confirmed these findings or vomiting follow different temporal sequences, so
in that negative mood increased prior to binge eat- that results are partly determined by assessment
ing, and worsened after binge eating in both BN protocols. Alternatively, binge-​eating episodes may
and BED. Results on mood after purging were not not invariably be followed by either mood improve-
reported, however, so it remains unclear whether ment or mood deterioration. Rather, as yet to be
purging may be reinforced by mood improvement identified situational or individual factors (e.g.,
as suggested previously. A  caveat for both studies cognitions) may determine which emotional conse-
is the fact that mood was assessed and analyzed in quences arise from binge eating in individuals with
very close proximity to the binge-​eating episode, so BN. For example, it could be argued that a binge-​
that the implied direction of negative mood leading eating episode is experienced as less negative and
to binge eating may have been obscured (i.e., the therefore potentially mood-​ improving compared
negative mood immediately prior to binge eating to prebinge levels if the possibility for subsequent
may have been a consequence of the realization of purging is given. Equally, if purging is rendered
the impending binge-​eating episode). Smyth et al. impossible, the experience of binge eating may
(2007) addressed this problem by removing mood have the described negative effect on mood (Staiger
assessments from their analyses that may have been et al., 2000).
temporally consequential to the binge event (i.e., Current meta-​ analytic evidence suggests that
those within 10 minutes of the event). The results negative affect increases following binge eating, but
confirm that decreasing positive affect and increas- decreases after purging (Haedt-​Matt & Keel, 2011).
ing negative affect including anger and hostility reli- Certainly, within the affective processing model of
ably preceded binge eating and subsequent purging. negative reinforcement (Baker et  al., 2004) both
Moreover, mood significantly improved after binge binge eating and purging can be conceptualized as
eating or purging, and this recovery was more rapid behaviors that allow escape or avoidance of negative
than the mood deterioration before the binge-​ affect. Results from a study by Berg et  al. (2013)
eating episode. Smyth et al. (2007) also showed that suggest that facets of global negative affect could
mood was significantly worse on days when binge be differentially related to bulimic behaviors such
eating and purging occurred compared with non- as binge eating and purging:  fear, guilt, hostility,
binge days. Even the most positive mood rating on and sadness increased prior to and decreased follow-
binge days was still more negative than the average ing binge-​only and binge/​purge events. The same
mood on a nonbinge day. was true for purge-​only events, with the exception

168 Mood, Emotions, and Eating Disorders

that hostility did not increase significantly prior to physiological responses that are experienced as crav-
purging. Ratings of guilt were higher than those ing. It is interesting in this context that the type of
of the other negative affects at the time of binge-​ food seems to be important for binge eating in BED
only and binge/​ purge events. Guilt increased patients, as they prefer high-​caloric food with high
prior to and decreased following the three behav- sugar and fat content (De Zwaan, 2001). Based on
ior types, even after controlling for fear, hostility, these results it has been suggested that individuals
and sadness, which—​according to the authors—​ with BED differ in reward sensitivity from non-​
suggests a particular role for feelings of guilt for the eating-​disordered individuals. In the following sec-
maintenance of BN. tions we review the evidence in relation to these
models.
Binge Eating Disorder
Binge eating disorder (BED) resembles BN, but Emotional Responses to Food Cues
binge eating occurs in the absence of regular com- Emotional responses to food cues in binge eaters
pensatory behaviors to prevent weight gain, such as have mostly been investigated using experimental
fasting or purging. Consequently, overweight and procedures, in both clinical and nonclinical sam-
obesity are common among those diagnosed with ples. McNamara, Hay, Katsikitis, and Chur-​Hansen
BED (Grucza, Przybeck, & Cloninger, 2007). Also, (2008) compared emotional responses to the pres-
BED is associated with co-​occurring physical and entation of food images in relation to body image
mental illnesses, as well as impaired quality of life concerns and ED symptoms, both assessed by
and social functioning (Wilfley, Wilson, & Agras, self-​report questionnaires. The nonclinical sample
2003). Results from a community sample suggest included children, adolescents, and young adults,
that BED contributes to these comorbidities over both male and female, so that age and sex were also
and above those expected for obesity (Gruzca et al., included in the analysis. The results show a decrease
2007). Among mental illnesses, depressive disor- in the emotive responses (i.e., fear, disgust, happi-
ders and anxiety disorders are the most prevalent ness) to food pictures with increasing age. Males had
comorbid lifetime and current diagnoses for indi- the more positive emotive responses, but not neces-
viduals with BED (Dingemans, Spinhoven, & Van sarily less negative emotions toward food, whereas
Furth, 2007; Grilo et al., 2009). Those with current in adult females a fear-​emotive response was associ-
comorbidity have higher levels of current ED psy- ated with eating concern and body dissatisfaction.
chopathology, negative affect, and lower self-​esteem The authors concluded that the more positive emo-
compared to BED patients without lifetime or cur- tional response in males may be a protective factor
rent comorbidity. that reduces their vulnerability to EDs.
Taken together, these studies provide evidence Vögele and Florin (1997) examined emo-
that depressive symptoms (trait), acute negative tional and psychophysiological responses to food
mood (state), and binge eating behavior are related, exposure in self-​ identified female binge eaters,
but little is known about their causal relationships. and found higher sympathetic arousal levels as
It has been suggested that binge eating in BED indicated by blood pressure and electrodermal
reflects difficulties in regulating negative emo- activity throughout the exposure compared with
tions (Whiteside et  al., 2007), and—​as for BN—​ female non-​ binge-​eaters. This was accompanied
that binge eating is an attempt to reduce negative by increases in ratings of hunger, desire to binge,
affect associated with food, body shape, or stress, sadness, and nervousness. Heart rate during food
in particular as negative emotions are reported to exposure predicted the relative amount of food
often precede binge eating episodes in BED (Stice, eaten after exposure across all participants, but this
Akutagawa, Gaggar, & Agras, 2000). An impor- relationship was more pronounced in binge eaters.
tant difference to BN concerns the role of dieting. In summary, the results indicate that binge eaters
Whereas in BN dieting nearly always precedes a responded with greater subjective and physiolog-
binge-​eating episode, this pattern is usually reversed ical arousal to food cues than non-​binge-​eaters,
in BED, in that sporadic dieting attempts follow which was experienced as aversive and relieved
eating binges. by eating. This supports the affective processing
Another approach follows a conditioning model model of negative reinforcement (Baker et  al.,
of binge eating (CBE) (Jansen, 1994), in which it is 2004) and indicates that binge eaters use binge
postulated that through classical conditioning, cer- eating to down-​regulate the strong negative emo-
tain cues such as the sight and smell of food elicit tions elicited by food cues.

Vögele, Lutz, Gibson 169

 The foods chosen by participants in Vögele but feared (because of its presumed consequences,
and Florin’s experiment (1997) did not differ in i.e., fattening, etc.), as is the case for BN and
amount or calories between binge eaters and non-​ probably chocolate cravers (see section “Anorexia
binge-​eaters. Nevertheless, binge eaters are reported Nervosa”).
to prefer high-​caloric foods with high fat and sugar Findings related to the appetitive qualities of
content (De Zwaan, 2001), and this preference food cues are reported by Goldfield and coworkers
might be mediated by the reward value of these (2008). This study differs from the previous investi-
types of food. There is a literature on the neural cor- gations in that not the emotional responses to food
relates of visual food stimuli of high-​caloric food in cues per se but changes in the relative reinforcing
normal-​weight women (e.g., Killgore & Yurgelun-​ value of snack foods were assessed in relation to par-
Todd, 2007), which indicates that activity in brain ticipants’ stress reactivity and (questionnaire-​based)
regions important for early processing of these cues binge eating status. Although there were no differ-
can be influenced by their hedonic value as well as ences between low stress-​reactive binge and non-​
the affective state of the individual. Greater positive binge-​eaters in the way they rated the reinforcing
affect is associated with increased activity within qualities of food under stress, stress-​reactive binge
the primary visual cortex during visual perception eaters found food more rewarding under stress as
of calorie rich and highly flavorful foods, but not opposed to high-​ reactive non-​ binge-​
eaters, who
for less pleasurably satisfying low-​ calorie foods. rated food less reinforcing. This effect was indepen-
Such findings suggest a mechanism whereby affec- dent of BMI and dietary restraint. These findings
tive state may affect the early stages of sensory proc- suggest that the experience of stress has opposite
essing, possibly influencing subsequent perceptual effects on the relative reinforcing value of food and,
experience of a stimulus. therefore, food intake, depending on binge eater
Schienle, Schäfer, Hermann, and Vaitl (2009) status (see also section “Binge Eating Disorder”).
used functional magnetic resonance imaging to In their recent, comprehensive review of neu-
explore the neural correlates of visual food exposure roimaging and neurophysiological studies, genetic
after an overnight fast in four groups of partici- and heritability studies and cognitive studies,
pants: overweight BED, overweight healthy control, Kessler et  al. (2016) conclude that patients with
normal-​weight healthy control, and normal-​weight BED exhibit increased impulsivity, compulsivity,
BN. The BED patients reported enhanced reward and altered reward sensitivity, along with enhanced
sensitivity and showed stronger activation in the attentional biases directed toward food cues, and
medial orbitofrontal cortex (OFC) while viewing impaired cognitive function. These patterns may be
food pictures than all other groups. Activity in the related to maladaptation of the corticostriatal cir-
OFC is assumed to reflect the hedonic value and cuitry regulating motivation and impulse control
reward relevance of food. The conclusion drawn by similar to that found in other impulsive/​compulsive
Schienle et al. (2009) is that the heightened medial disorders.
OFC reactivity to food cues observed for BED sug-
gests increased reinforcement sensitivity, which may Emotional Responses to Body Image Exposure
lead to binge eating. It is interesting to note in this In contrast to AN and BN, concerns about body
context that the study by Schienle et al. (2009) and weight and shape are currently not included in the
also the studies by Killgore and coworkers (e.g., diagnostic criteria for BED (DSM-​5). Nevertheless,
2007) seem to suggest a positive emotional response there is increasing evidence that a negative body
to visual food stimuli, even in BED (Schienle et al., image is also important in BED, and that they have
2009), although this would contradict the previous similar levels of overvaluation of shape and weight as
reports on negative affect in response to food cues. those with BN and AN (Wilfley, Schwartz, Spurell,
On closer inspection it becomes clear, however, that & Fairburn, 2000). This effect is not merely due to a
the positive aspects referred to in Schienle et  al. higher body weight, as overweight individuals with
(2009) and Killgore et al. (2007) refer to appetitive BED tend to have higher weight and shape concerns
characteristics of food pictures, whereas the negative and to desire a slimmer body than overweight indi-
emotional responses reported by McNamara et  al. viduals without BED (Lewer, Nasrawi, Schroeder,
(2008) and Vögele and Florin (1997) refer to emo- & Vocks, 2016). Moreover, overvaluation of body
tions such as fear, sadness, and nervousness. In BED weight and shape has been shown to predict the
these emotions seem complementary rather than degree of eating-​related psychopathology and psy-
mutually exclusive: Food is desired (i.e., appetitive) chological disturbance (Hrabosky, Masheb, White,

170 Mood, Emotions, and Eating Disorders

& Grilo, 2007). In summary, there is accumulat- Several laboratory studies have investigated this
ing evidence that overvaluation of body weight and question by comparing food consumption under
shape and a negative body image are as important in conditions of negative mood induction with a con-
BED as they are in AN and BN (Grilo et al., 2008). trol condition. In Telch and Agras (1996), mood
In the studies by Hrabosky et  al. (2007) and induction by tape-​recorded vivid imagery did not
Grilo et  al. (2008) overvaluation was significantly affect food consumption in obese BED when com-
related to depression levels in BED. This points to pared with a non-​eating-​disordered control group,
the possibility that negative affect mediates the asso- although BED participants consumed more calories
ciation between a negative body image and eating-​ overall. Chua, Touyz, and Hill (2004) randomly
related pathology. Further support for this notion allocated obese BED to a sad or neutral film con-
comes from correlational studies reporting a system- dition before monitoring food consumption in an
atic association between physical appearance-​related alleged taste test (chocolate). Participants in the sad
teasing experiences and depression levels in BED condition ate significantly more when compared
(Jackson, Grilo, & Masheb, 2002). Nevertheless, to with those in the control condition. The authors’
our knowledge there is only one study that exam- hypothesis that those with a restrained eating style
ined the effects of body shape exposure on emo- in the negative mood condition would eat most
tional responses in an experimental design. Hilbert, failed to reach statistical significance. Munsch,
Tuschen-​Caffier, and Vögele (2002) exposed a Michael, Biedert, Meyer, and Margraf (2008)
sample of female BED individuals to their physical randomly allocated obese BED to a negative or
appearance in a whole-​body mirror using a manual neutral mood induction (using guided imagery).
to guide participants in describing their appear- Participants also received either a balanced diet
ance. In comparison with age-​and weight-​matched (rich in carbohydrates) or a poor diet (high in fat)
non-​ eating-​
disordered control participants, those over 3 consecutive days preceding laboratory test-
with BED responded with more negative mood ing. Although the mood induction was successful,
(sad, tense, insecure, anxious, disgusted), although there were no differences between experimental
a decrease in mood was apparent in both groups. groups in the amount of food eaten in the alleged
When the exposure session was repeated 2 days later, taste test (dessert cream). Dingemans, Martijn,
emotional responses across groups were less negative Jansen, and van Furth (2009) tested the assumption
than during the first exposure. Nevertheless, the dif- that it is not the experience of negative affect per se
ferences between BED and controls remained, with that prompts overeating in BED, but the attempt
BED showing more intense negative emotions. to regulate it. Models of self-​regulation provide the
Appearance self-​esteem was lower in BED during theoretical background for this hypothesis (e.g.,
both exposure sessions than in controls, and BED Muraven & Baumeister, 2000), which stipulates
described their bodies as fatter than non-​eating-​ that self-​control is a limited resource that can be
disordered participants. Interestingly, levels of depleted. When, for example, a situation demands
anxiety were lower and less affected by the repeated two consecutive acts of self-​control, performance on
exposure than feelings of tension, sadness, insecu- the second act (e.g., resist eating) is usually impaired
rity, and disgust. Body image disturbance in BED because of energy depletion. Accordingly, half of the
may, therefore, be more related to a composite of sample was instructed to suppress their emotions
negative mood rather than anxiety. while watching a sad film clip, while the other half
were told to respond naturally. The main hypothesis
Negative Affect and Binge Eating that those in the emotion-​suppression condition
Affect models, in contrast to dietary restraint would eat most was not confirmed. Nevertheless,
models, postulate that binge eating is triggered by individuals with BED who had severe depressive
negative affect (Meyer, Waller, & Waters, 1998), symptoms consumed more calories than those with
and that binge eating acts as a negative reinforcer as no or mild depressive symptoms, independent of
it provides short-​term relief by reducing or numb- the experimental condition. Mood improved after
ing negative emotions or distraction from aversive eating to levels comparable to those before watch-
mood states. These models are clearly more appro- ing the film.
priate for BED than restraint models, as dieting The notion that emotion regulation may be
usually follows a binge but does not precede it. more important for binge eating than negative
The evidence for the notion that negative affect affect per se has found some support from cross-​
triggers binge eating in BED is, however, mixed. sectional studies. Whiteside et al. (2007) evaluated

Vögele, Lutz, Gibson 171

difficulties in regulating negative emotions in rela- non-​eating-​disordered individuals monitored their
tion to binge eating in a cross-​sectional study of eating and mood for 6 consecutive days using a
male and female undergraduate university students. hand-​held computer. Poor mood, feelings of poor
Hierarchical regression results indicated that dif- eating control, and craving sweets all preceded binge
ficulties in emotion regulation as assessed by self-​ eating episodes in BED. Although mood was worse
report questionnaire accounted for a significant overall for women with BED, it was especially poor
amount of the variance in binge eating, over and before a binge-​eating episode. An unexpected find-
above sex, dietary restraint, and overevaluation of ing was the relatively high percentage of women in
body weight and shape. The specific types of emo- the control group (i.e., not identified as BED) but
tion regulation difficulties most strongly associated who still reported binge eating over the observation
with binge eating were limited access to emotion period and who—​in terms of negative affect—​were
regulation strategies and difficulty identifying and between BED and those who did not experience
making sense of emotional states. any binge episodes.
The sample investigated by Whiteside et  al. Deaver, Miltenberger, Smyth, Meidinger,
(2007) included predominantly young adults. and Crosby (2003) investigated a sample of self-​
Nevertheless, there is evidence that the association identified female binge eaters and compared their
between emotion regulation and binge eating also responses on an affect grid at 2-​minute intervals
applies to much younger individuals. Czaja, Rief, before, during, and after binge eating episodes and
and Hilbert (2009) examined a community sample regular meals to those of individuals not reporting
of 8-​to 13-​year-​old children who had experienced any binge eating. The observation period was 4 to
at least one episode of loss of control over eating. 7  days, depending on whether at least two binge-​
Control participants had not experienced any loss eating episodes had occurred. The results show
of control over eating or other disordered eating that negative affect preceded binge-​eating episodes,
behavior and were matched for age, sex, percentile which became less negative during the binge, and
BMI, education, and socioeconomic status of the again increased after the binge. This pattern was
mother. The results show that children who had the same for normal meals, but differed in level in
experienced loss of control over eating made higher that negative affect was less marked during normal
use of dysfunctional emotion regulation strategies, meals. Participants in the control group experienced
especially for the regulation of anxiety. The use of similar levels of negative affect before meals than
maladaptive strategies was associated with greater did BED before meals; however, they reported sig-
depressiveness. nificantly more pleasant affect than did binge eaters
In summary, there is some evidence linking neg- during and after eating regular meals.
ative affect or its failed regulation to overeating in Women meeting DSM-​ IV criteria for BED
BED, although results from controlled laboratory took part in the study of Stein et al. (2007). Over
experiments are inconclusive. One reason for this 7 consecutive days participants were prompted
may be the artificial laboratory setting, and it has by hand-​held computers at six intervals during
been suggested that it is unclear how well laboratory the day to answer a series of questions pertaining
eating behavior mimics naturalistic eating behavior to eating and mood since the last data entry. As
(Walsh & Boudreau, 2003). For example, the fre- in previous studies, negative mood and hunger
quently used taste tests in these investigations usu- were significantly higher before binges compared
ally involve the presentation of only very few food to nonbinge times. In contrast to previous results
items (sweets, chocolate, or dessert cream) even (Deaver et al., 2003), however, negative mood was
though typical binge foods can be quite idiosyn- even worse after binge-​eating episodes, a finding
cratic. Moreover, the negative affect experienced in that is shared by the previously mentioned study
a typical laboratory setting may be far less intense by Hilbert and Tuschen-​Caffier (2007; see sec-
than that encountered in a natural setting. tion “Bulimia Nervosa: Negative Affect and Binge
As is the case for studies of binge eating in BN, Eating”). An interesting finding in Hilbert and
the concurrent assessment of mood changes and Tuschen-​Caffier (2007) was that negative mood
eating behavior in the natural environment offers antecedent to binge eating was in part attributable
an attractive alternative for the investigation of to general psychopathology, suggesting that anx-
antecedents and consequences of binge eating in iety and depression predispose individuals with
BED. In Greeno, Wing, and Shiffman (2000) BED to more negative mood proximal to binge
women with BED and weight-​and age-​matched eating.

172 Mood, Emotions, and Eating Disorders

 We can summarize, therefore, that there is evi- kept an electronic diary over 7 consecutive days
dence that negative affect triggers binge eating in and recorded daily hassles type of stress, desire to
BED and those at risk (e.g., obese individuals). The eat, negative affect, and rumination. The results of
question exactly how negative affect triggers overeat- mixed regression modeling showed that daily has-
ing, however, has not been resolved yet, and several sles predicted the desire to eat; however, the predic-
pathways have been suggested. Affect-​driven models tive value further increased when negative affect and
include trade-​off theory, which posits that in the face rumination were accounted for. This finding sug-
of negative mood, binge eating serves to substitute gests a significant contribution of ruminative think-
a less aversive affective condition (e.g., guilt after ing to the mechanisms of negative-​affect-​induced
binge eating) for the more aversive emotional state eating. With ruminative thinking being a key fea-
(e.g., depression) that preceded the binge (Kenardy, ture of depression it links in with the previously
Arnow, & Agras, 1996). There is the escape from reported finding (Hilbert & Tuschen-​Caffier, 2007)
self-​awareness model, which proposes that the act that negative mood antecedent to binge eating is in
of binge eating allows narrowing the attentional part attributable to general psychopathology, in par-
focus on the immediate stimuli (i.e., food) in order ticular depression and anxiety.
to block out negative emotions (Heatherton & As is the case for BN, however, there is conflict-
Baumeister, 1991; Nolen-​Hoeksema, Stice, Wade, ing evidence as to the reinforcement model, which
& Bohon, 2007). The previously described self-​ for BED predicts that binge eating functions as at
regulation model assumes that self-​regulation is a least temporary relief from negative affect experi-
limited resource that can be depleted with a resulting enced prior to binge eating. Deaver et  al. (2003)
loss of self-​control (Muraven & Baumeister, 2000). found support for this notion, whereas the results of
For example, negative affect might trigger overeat- Stein et al. (2007) and Hilbert and Tuschen-​Caffier
ing because it distracts people from or depletes the (2007) suggest that individuals with BED feel even
resources used for their dietary vigilance (Stice, worse after a binge eating episode. Notwithstanding
Akutagawa, et  al., 2000). Another affect-​ related methodological differences between these studies
model is masking theory, according to which binge (e.g., time sampling protocols), it could be argued
eating serves as an attribution for negative affect that that these discrepancies in findings might reflect
masks other problems (Polivy & Herman, 1999). If differences in situational or individual factors (e.g.,
negative emotions can be blamed on binge eating, subtypes of binge eaters) that determine whether
this could be perceived as more tolerable than other mood improves after binge eating or not. Recent
sources of distress. results by Berg et al. (2015) are consistent with the
These models are not mutually exclusive, but affect regulation model of binge eating and suggest
only emphasize different aspects of overeating as that binge eating may function to regulate global
dysfunctional coping strategy in response to nega- negative affect. They echo the results reported for
tive affect (Schachter et al., 1968; Spoor, Bekker, Van BN, but in obese adults experiencing binge epi-
Strien, & van Heck, 2007), with ruminative think- sodes, in particular in highlighting the role of feel-
ing probably playing a key role. Ruminative think- ings of guilt, suggesting the importance of targeting
ing can be defined as uncontrollable perseverative negative affect in the treatment of binge eating in
thinking about past or present events, in particular the context of obesity.
“dwelling on negative affect” (Nolen-​Hoeksema,
1991; Thomsen, 2006). It may contribute to over- Anorexia Nervosa
eating in at least two ways: First, it might be an indi- The core feature of AN is a relentless pursuit of
cator for disinhibition, as ruminative thinking has thinness and a refusal to maintain a body weight at
also been linked to a general lack of behavioral inhi- a minimally acceptable standard for age and height.
bition (Siegle & Thayer, 2004). Second, ruminat- The diagnostic criterion in the DSM-​5 stipulates
ing may lead to a prolonged dwelling on negative that weight should be significantly low, that is, less
affect, thus increasing its intensity and possibly acti- than minimally normal or, for children and adoles-
vating eating as a dysfunctional coping or “escap- cents, less than minimally expected. To determine
ist” behavior. Kubiak, Vögele, Siering, Schiel, and whether the weight criterion is met, the DSM-​5
Weber (2008) investigated the role of ruminative recommends consideration of available numerical
thinking for emotional eating in a sample of obese, guidelines for body weight or BMI, the individu-
female adolescents who reported at least one binge-​ al’s body build, weight history, and physiological
eating episode over the past 12 months. Participants disturbances.

Vögele, Lutz, Gibson 173

 Individuals with AN have an essential fear studied the neural correlates of the processing of
of gaining weight or becoming fat, place undue visual food stimuli in EDs and found that both AN
value on weight, and typically show disturbed per- and BN showed decreased activation in areas that
ceptions of their own body shape and size. The are associated with appetitive behavior (inferior pari-
DSM-​ 5 and the ICD-​ 10 distinguish two clini- etal lobe, occipital cortex) and increased activation
cal subtypes:  Those with restricting AN (AN-​R) in areas that are associated with cognitive control
accomplish weight loss primarily through fasting or (medial prefrontal cortex), compared with non-​
excessive exercise, or both. Those with AN of the eating-​disordered controls. Gordon et  al. (2001)
binge eating/​purging type (AN-​BP) reduce weight used positron emission tomographic measurements
through self-​induced vomiting and use of laxatives. of regional cerebral blood flow (rCBF) to compare
There is some crossover between the two subgroups, responses of AN and healthy participants while
and it remains unclear as to whether these are two viewing pictures of high-​and low-​calorie foods, and
enduring subtypes, or different phases of the same nonfood items. In contrast to healthy participants,
illness (Eddy et al., 2002). Results from a retrospec- who reported a desire to eat while viewing pictures
tive study of a large sample of AN patients suggest of high-​calorie foods, AN individuals responded
that the switch from the AN-​R type to the AN-​BP with elevated anxiety and increased heart rates. This
type is associated with a large increase in the preva- was associated with exaggerated responses in visual
lence of suicide attempts (Foulon et al., 2007). The association cortices, similar to those observed in
authors argue the affected patients could experience studies of specific phobias.
binge eating as a failure to control appetite, and To control for differences in hunger and sati-
purging as a loss of previous asceticism. Switching ety, participants in these studies were required to
from AN-​R to AN-​BP could, therefore, be associ- have eaten at a specified time before the experiment
ated with higher negative affectivity or an increase started. It is not clear, therefore, in what way hun-
in impulsivity, both of which could contribute to ger or satiety might affect such results. This was
the observed higher rates of suicide among persons addressed in a functional magnetic resonance imag-
with AN-​BP. ing (fMRI) study by Santel, Baving, Krauel, Münte,
In terms of clinical symptoms and impulsivity, and Rotte (2006) in which patients with AN and
individuals with AN-​BP, therefore, have much in healthy controls were tested in both a hungry and
common with individuals with BN, who have simi- a satiated state. Hunger scores of AN patients were
lar rates of suicide attempts. Both groups engage far lower than those of healthy participants, with a
in an abnormal eating pattern characterized by trend for this effect to be more marked in the hun-
alternating periods of fasting and episodes of binge gry condition. And AN patients rated food stimuli
eating, together with purging directly after binge as less pleasant than healthy controls, independ-
eating (e.g., vomiting or use of laxatives) or exces- ent of the condition of hunger or satiety. This was
sive exercise. The only distinguishing feature is the accompanied by decreased somatosensory process-
abnormally low body weight in AN. ing in the satiated state (i.e., decreased activation in
the inferior parietal lobe) and decreased attention to
Emotional Responses to Food Cues food stimuli in the hungry state (i.e., reduced occip-
As previously described, non-​eating-​disordered ital activation). The authors interpret these results
individuals tend to assign positive emotional in terms of reduced responsiveness to the pleasant
valence to food stimuli that is enhanced by hun- aspects of food stimuli that are highly appetizing to
ger (Lozano, Crites, & Aikmann, 1999). Patients healthy individuals. This altered cognitive process-
with AN, who are preoccupied with food irrespec- ing of visual food stimuli may facilitate fasting. For
tive of hunger, however, report negative valence of example, the decreased attention to food stimuli
food stimuli (Vaz, Alcaina, & Guisado, 1998), and while hungry could be interpreted as avoidance
in this respect do not differ from BN or BED. Self-​ behavior.
starvation can thus be negatively reinforced, and Such cross-​sectional investigations of AN patients
may become rewarding due to several factors, such provide important information about abnormali-
as social reinforcement related to weight loss, devel- ties in emotional experiences and associated brain
opment of a sense of inner mastery over hunger and functions with this disorder. Nevertheless, the
weight, or alleviation of negative affect. causal relationships between brain alterations and
Uher and coworkers (Uher, Brammer, et  al., the clinical syndrome are unclear because primary
2003; Uher, Murphy, et al., 2004) have extensively neural disturbances cannot be differentiated from

174 Mood, Emotions, and Eating Disorders

phenomena secondary to the disorder. In an effort own body (Vocks, Legenbauer, Wächter, Wucherer,
to overcome this problem, Uher et al. (2003) inves- & Kosfelder, 2007), and persons with AN are no
tigated responses to food stimuli and pictures from exception. Neuroimaging studies investigating this
the International Affective Picture System (IAPS) phenomenon typically expose participants to dis-
with fMRI in chronically ill AN-​R and in long-​ torted images of their own body, that is, a heavier
term recovered AN-​R and healthy controls. Instead version of themselves. Although this paradigm
of a prospective design, which would be difficult to was intended to provoke fear of fatness, results are
realize, Uher et al. (2003) argued that by obtaining inconsistent regarding the involvement of struc-
evidence from people fully recovered from AN trait tures related to the “fear circuit” (e.g., amygdala),
factors could be identified without the confounding with some studies reporting increased amygdala
evidence from acute starvation, metabolic altera- activation (Miyake et al., 2010; Seeger, Braus, Ruf,
tions, and treatment-​related stress. The main find- Goldberger, & Schmidt, 2002) and others not
ing of this study was that food-​related brain activity (Wagner, Ruf, Braus, & Schmidt, 2003). Moreover,
in recovered AN-​R was a combination of responses increased amygdala activation in response to nega-
seen in currently ill AN-​R (increased activity in the tive body-​related stimuli does not seem to be spe-
medial frontal and anterior cingulated cortices) and cific to AN, but has also been reported in healthy
those recorded in healthy controls (increased activ- women (Kurosaki, Shirao, Yamashita, Okamoto,
ity in the apical and lateral prefrontal cortices). & Yamawaki, 2006; Friederich et al., 2007; Shirao,
Food pictures were rated as less pleasant and more Okamoto, Okada, Okamoto, & Yamawaki, 2003).
aversive by currently ill AN-​R compared to either Summarizing neuroimaging findings from the
recovered AN-​R or healthy controls. No differences exposure to fat-​distorted body images and unfa-
in brain activation or subjective ratings were found vorable self–​other body comparisons, Gaudio
for the IAPS stimuli. Although cross-​sectional in and Quattrocchi (2012) conclude that a network
research design, by comparing recovered and cur- involving the prefrontal cortex, insula, and amyg-
rently ill patients the study results suggest that the dala is implicated in the affective component of
medial prefrontal response to food stimuli is present body image disturbance in AN.
in both active illness and recovery and may, there- A symptom specific to AN is the positive valu-
fore, represent a trait factor of the disease. ation of not merely slim, but underweight, even
O’Hara, Campbell, and Schmidt (2015) pro- emaciated, female bodies associated with a drive
vide a narrative review of neurobiological and psy- for thinness. In this context, exposure to images
chophysiological evidence supporting a role for of underweight bodies has been shown to activate
altered reward processes in the development and structures of the ventral-​striatal “reward circuit” in
maintenance of AN. In AN, there does not appear participants with AN (Fladung et al., 2010). Other
to be a generalized inability to experience reward, studies have been able to demonstrate an implicit
rather than a reluctance to gain weight, leading association of underweight bodies with positive
to an aversive appraisal of food-​and taste-​related affect in women with AN, and women scoring high
stimuli. Neuroimaging studies involving food-​and on drive for thinness, using an implicit-​association
taste-​related stimuli report differences in prefron- test (Ahern, Bennett, & Hetherington, 2008) and
tal, cingulate, insular, and striatal regions relative an affective startle modulation paradigm (Reichel
to healthy controls. These suggest that interactions et  al., 2014). It appears, therefore, that images of
between higher order cognitions and motivational underweight bodies have an immediate positive and
reward circuits are important in the pathology of rewarding effect on AN, thereby possibly contrib-
AN. Studies reporting hyperactivity in frontocingu- uting to the relentless pursuit of thinness in these
late and insular-​striatal neurocircuitry in response patients.
to food-​related cues suggest that in AN there is Altered body-​ related affective processing also
increased salience attribution to such cues, irre- plays a role in body size estimation. Individuals
spective of valence. This hyperactivity may reflect with AN, on average, overestimate their body size
the effort involved in controlling exposure to over- (Cash & Deagle, 1997; Sepúlveda, Botella, & León,
whelming food stimuli. 2002; Smeets, 1997). In addition, using electroen-
cephalography, a recent study demonstrated altera-
Emotional Responses to Body Image Exposure tions in the basic visual processing of images of
Individuals with EDs in general experience one’s own body in AN, occurring already 100 ms
negative emotions when confronted with their after stimulus onset (Lutz et al., 2015). The cause of

Vögele, Lutz, Gibson 175

this perceptual distortion is unclear, although it is and in differentiating between feelings and physical
certainly of a complex nature. Body size perception sensations. It involves an impaired capacity to con-
is influenced by affective, cognitive, and other fac- struct mental representations of emotions, which
tors, such as hunger (Farrell, Lee, & Shafran, 2005; is the prerequisite for the cognitive processing of
Slade, 1994). In addition, a recent study suggests emotional experiences and their communication to
the reciprocal relationship of body size perception others. Gilboa-​Schechtmann, Avnon, Zubery, and
and affective factors, in that size perception affected Jeczmien (2006) compared groups of patients with
body satisfaction: Preston and Ehrsson (2014) used AN (AN-​R and AN-​BP) and BN to non-​eating-​
a full-​body illusion to demonstrate that endorsing disordered individuals. Despite some differences
an enlarged body image increased body dissatisfac- between the ED groups in emotional awareness,
tion. While altered visual processing of body images AN and BN individuals appeared to be more simi-
appears to be amenable to treatment (Vocks et al., lar in their profile of emotional deficiencies than
2010), altered brain activity associated with emo- previously suggested. Importantly, the differences
tional response to body image exposure seems to between the ED groups and non-​eating-​disordered
persist, even after treatment (Rodriguez-​Cano et al., controls were mediated by emotional distress, a
2009; Vocks et al., 2010). This highlights the neces- finding that confirms a previous result by Espina
sity for further targeting emotional responses to Eizaguirre, Saenz de Cabezón, Ochoa de Alda,
body image exposure in research and treatment of Joaristi Olariaga, and Juaniz (2004). Only a mea-
body image disturbance in AN. sure of rumination remained significantly correlated
with a continuous measure of ED pathology when
controlling for anxiety and depression (see also the
Emotional Processing Deficits previous discussion of rumination and general psy-
The literature on mood and emotions in AN has chopathology in BN and BED).
largely concentrated on the concept of alexithymia. As most studies investigating emotional pro-
This concept was initially developed by observing cessing deficits in ED have used the TAS, there is
patients with psychosomatic disorders and hypoth- a preponderance of questionnaire-​ based, cross-​
esizes that alexithymia is a risk factor for organic sectional studies. Nevertheless, some recent psycho-
disorders (Sifneos, 1991). Alexithymia is conceptu- physiological studies have confirmed aspects of the
alized as a cognitive–​affective deficit comprising the assumption that emotional processing is impaired
following impairments: (1) difficulties in identifying in AN. Pollatos, Herbert, Schandry, and Gramann
feelings and distinguishing emotions from physical (2008) found marked differences in evoked poten-
sensations, (2) difficulties in communicating emo- tials and emotion recognition performance in
tional states to others, (3)  restricted daydreaming, response to emotional faces between AN individu-
and (4) a concrete/​externally oriented style of think- als and non-​eating-​disordered controls. The authors
ing (Taylor, Bagby, & Parker, 1991). Alexithymia conclude that differences in brain dynamics might
has been almost exclusively investigated using the contribute to difficulties in the identification of
Toronto Alexithymia Scale (TAS; Bagby, Taylor, & facially expressed emotions and deficits in social
Parker, 1994). functioning.
Several studies suggest that alexithymia, as opera- Yet another approach in the study of emotional
tionalized by the TAS, is an important factor in EDs processing in AN was used by Doba et al. (2007).
(e.g., Beales & Dolton, 2000; Guilbaud et al., 2000; The authors investigated the temporal organization
Zonnevijlle-​Bender, van Goozen, Cohen-​Kettenis, of emotional expression in autobiographical speech.
& van Engeland, 2002), with AN-​R individuals The results show that AN individuals expressed
typically reporting greater difficulties in identifying more negative emotions and fewer neutral states,
their emotions and in describing their feelings to together with negative emotional perseveration,
others when compared with non-​eating-​disordered when compared with healthy participants.
women (Taylor, Parker, Bagby, & Bourke, 1996) In summary, individuals with AN appear to have
and also individuals with BN (Taylor et al., 1997), difficulties in representing their own emotional
although this latter result could not be replicated in experiences and those of others. This may thwart
a more recent study (Bydlowski et al., 2005). patients from seeking healthcare, reduce their treat-
As a concept, alexithymia relies on a deficit ment adherence, contribute to the chronic nature
model of emotional processing in that it emphasizes of the disorder, and increase the risk for potential
difficulties in identifying and describing feelings relapse. Another consequence of these emotional

176 Mood, Emotions, and Eating Disorders

processing deficits is increased negative affect, which aftermath may lead to emotion-​dependent habits.
AN individuals may try to alleviate by starvation, Foods that are themselves capable of shifting emo-
hyperactivity, binge eating, and purging. Alternative tional states may provide reinforcement through
or additional strategies to cope with increased levels escape or distraction from aversive emotions. More
of negative affect could consist of modifying access moderate emotions can be generated by consump-
to autobiographical emotional memories by retriev- tion of palatable foods, and in turn influence our
ing memories less specifically. Autobiographical appreciation of foods. However, even in the non-
memories are considered to represent transient men- clinical population, our emotional involvement
tal constructions that are generated from a database with foods is complex and conflicting. A key theme
of autobiographical knowledge, which is hierar- in the literature is that the nature and consequences
chically organized from specific to general memo- of emotional interactions with food depends on
ries (Conway & Pleydell-​Pearce, 2000). Previous individual predispositions.
results suggest that there is a systematic relationship The interaction among mood, emotions, and
between the intensity of experienced negative events EDs is, however, more complex. The finding that
and general autobiographical memory such that the negative affect predicts eating disorders transdiag-
more aversive the experience the lower the spec- nostically, and that comorbidity with depressive
ificity of memories (Raes, Hermans, De Decker, disorders and anxiety disorders is the norm among
Eelen, & Williams, 2003). This was investigated patients with EDs suggests that EDs may not be
in a sample of AN-​R by Nandrino, Doba, Lesne, necessarily restricted to domains of eating behav-
Christophe, and Pezard (2006). Patients with AN-​R ior and body image but also may be related to sig-
recalled more general memories than controls in the nificant difficulties in affective functioning. It has
autobiographical memory test, but had no deficit been suggested that EDs are disturbances of mood
in explicit memory. Moreover, the overgeneraliza- regulation in which regulatory strategies specifically
tion of autobiographical memories included both related to eating and the body are used to diminish
negative and positive memories, and this pattern negative affect associated with food, body image, or
increased significantly with illness duration. These stress. In particular, three eating styles or tendencies
results suggest a general impairment in the access to have been identified: (1) restrained eating, (2) emo-
emotional memories in AN. Patients with AN may tional eating, and (3) uncontrolled or disinhibited
be prone to suppress or control not only negative eating. In brief, restrained eating is characterized
affect but also positive affect. The positive correla- by attempts at cognitive control over how much is
tion between the number of general responses and eaten. This strategy may be more or less successful,
illness duration shows that this deficit in experienc- although in many populations it may intensify as
ing and integrating emotional events is reinforced as excess weight is gained. Where the cognitive com-
the disorder becomes chronic. ponent is strong, emotional influences on eating
seem to be suppressed, although intense emotions,
Conclusion and other distractors, may nevertheless disrupt the
Mood and emotions are intrinsically involved cognitive control through competition for atten-
with eating, as would be expected for any funda- tional resources. This tendency clearly has implica-
mentally rewarding behavior:  for example, the tions for understanding AN.
pleasure from delicious food, or the disgust from By contrast, both emotional and uncontrolled
foul tasting food, are very real emotions. The ques- eating are common tendencies that may be present
tion is in what ways do these normal emotional in exaggerated form in BN and BED. Emotional
relations with food contribute to, or shed light on, eating is the tendency to overeat provoked by
the development of abnormal relations with food negative affect, whereas uncontrolled eating is
that eventually become clinical EDs. This chapter the tendency for sensory, social, or physiologi-
has first considered basic mechanisms, findings, and cal stimuli associated with food to elicit eating in
models that help our understanding of the interac- excess of need. These two facets are often intercor-
tions between eating and emotions, in both clinical related, and may share some common processes,
and nonclinical populations. This allows considera- such as altered function in opioid and dopaminer-
tion of normal food–​emotion interactions and how gic pathways. In both cases, highly palatable sweet
these may evolve into more aberrant behavior. or fatty foods are specifically capable of inducing
Intense emotions are often incompatible with overeating: However, for emotional eating, stress
eating, at least in the short-​term, although their or negative affect are prerequisites, whereas for

Vögele, Lutz, Gibson 177

uncontrolled eating, sensitization of hedonic and American Psychiatric Association (APA). (2013). Diagnostic and
reward pathways, reflecting genetic and/​or behav- statistical manual of mental disorders (5th ed.). Arlington, VA,
American Psychiatric Association.
ioral history, may be critical. This has led to the Apovian, C. M., Shah, S. N., Wolfe, B. M., Ikramuddin, S.,
notion of EDs as one possible manifestation of Miller, C. J., Tweden, K. S.,   .  .  .  Shikora, S. A. (2016).
behavioral addictions (Shaffer et  al., 2004). The Two-​year outcomes of vagal nerve blocking (vBloc) for the
frequent co-​occurrence of polysubstance abuse in treatment of obesity in the ReCharge Trial. Obesity Surgery.
EDs would further support this hypothesis. doi: 10.1007/​s11695-​016-​2325-​7
Attie, I., & Brooks-​ Gunn, J. (1989). Development of eat-
Approaches for defining and modeling these ten- ing problems in adolescent girls:  A  longitudinal study.
dencies continue to evolve:  Nevertheless, they are Developmental Psychology, 25, 70–​79.
helping to understand and predict the transition Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994). The twenty-​
from normal to abnormal eating and the particular item Toronto Alexithymia Scale—​ II. Convergent, dis-
role that mood and emotions may play. criminant, and concurrent validity. Journal of Psychosomatic
Research, 38, 33–​40.
Baker, T. B., Piper, M. E., McCarthy, D. E., Majeskie, M. R.,
Future Directions & Fiore, M. C. (2004). Addiction motivation reformu-
The findings reviewed here raise several key ques- lated:  An affective processing model of negative reinforce-
tions for future research to address: ment. Psychological Review, 111, 35–​51.
Baldo, B. A. (2016). Prefrontal cortical opioids and dysregulated
1.  What characteristics of an individual’s motivation:  A  network hypothesis. Trends in Neurosciences,
emotion regulation lead him or her to change 39, 366–​377.
from merely emotional involvement with food to Bardone, A. M., Krahn, D. D., Goodman, B. M., & Searles, J.
the severely exaggerated and dysfunctional eating S. (2000). Using interactive voice response technology and
timeline follow-​back methodology in studying binge eating
associated with a clinical diagnosis of an ED? and drinking behavior: Different answers to different forms
2.  What are the implications of research on of the same question? Addictive Behaviors, 25, 1–​11.
emotions and EDs for understanding the common Barr, R. G., Pantel, M. S., Young, S. N., Wright, J. H., Hendricks,
transitions between ED diagnostic categories, and L. A., & Gravel, R. (1999). The response of crying newborns
for predicting remission? to sucrose: Is it a “sweetness” effect? Physiology and Behavior,
66, 409–​417.
3.  What is the role of brain reward systems in Bassareo, V., Cucca, F., Musio, P., Lecca, D., Frau, R., & Di
EDs? Chiara, G. (2015). Nucleus accumbens shell and core dopa-
4.  Can improved knowledge of brain reward mine responsiveness to sucrose in rats:  Role of response
systems help to understand the etiology of EDs contingency and discriminative/​conditioned cues. European
better? Journal of Neuroscience, 41, 802–​809.
Beales, D. L., & Dolton, R. (2000). Eating disordered
5.  Can successful treatments be devised patients:  Personality, alexithymia, and implications for pri-
based on knowledge of the role of opioidergic, mary care. British Journal of General Practice, 50, 21–​26.
dopaminergic, and serotonergic systems in Beato-​Fernández, L., Rodríguez-​Cano, T., García-​Vilches, I.,
emotion regulation, normal eating, and EDs? García-​Vicente, A., Poblete-​García, V., Castrejon, A. S., &
6.  Can knowledge of the role of the vagus Toro, J. (2009). Changes in regional cerebral blood flow after
body image exposure in eating disorders. Psychiatry Research,
nerve in emotion regulation, appetite, and EDs 171, 129–​137.
be further developed and combined to improve Benton, D., & Young, H. A. (2016). A meta-​analysis of the
treatment strategies for EDs? relationship between brain dopamine receptors and obe-
sity:  A  matter of changes in behavior rather than food
References addiction? International Journal of Obesity, 40(Suppl 1),
Agras, W. S., & Telch, C. F. (1998). The effects of caloric depriva- S12–​S21.
tion and negative affect on binge eating in obese binge-​eating Berg, K. C., Crosby, R. D., Cao, L., Crow, S. J., Engel, S. G.,
disordered women. Behavior Therapy, 29, 491–​503. Wonderlich, S. A., & Peterson, C. B. (2015). Negative affect
Ahern, A. L., Bennett, K. M., & Hetherington, M. M. (2008). prior to and following overeating-​only, loss of control eating-​
Internalization of the ultra-​thin ideal: Positive implicit associa- only, and binge eating episodes in obese adults. International
tions with underweight fashion models are associated with drive Journal of Eating Disorders, 48, 641–​653.
for thinness in young women. Eating Disorders, 16, 294–​307. Berg, K. C., Crosby, R. D., Cao, L., Peterson, C. B., Engel, S. G.,
Alpers, G. W., & Tuschen-​Caffier, B. (2001). Negative feelings Mitchell, J. E., & Wonderlich, S. A. (2013). Facets of nega-
and the desire to eat in bulimia nervosa. Eating Behaviors, tive affect prior to and following binge-​only, purge-​only, and
2, 339–​352. binge/​purge events in women with bulimia nervosa. Journal
American Psychiatric Association (APA). (1994). Diagnostic and of Abnormal Psychology, 122, 111–​118.
statistical manual of mental disorders (4th ed.). Washington, Berridge, K. C. (2009). “Liking” and “wanting” food
DC: Author. rewards:  Brain substrates and roles in eating disorders.
Physiology and Behavior, 97, 537–​550.

178 Mood, Emotions, and Eating Disorders

Berridge, K. C., & Robinson, T. E. (1998). What is the role of Conners, C. K., & Blouin, A. G. (1983). Nutritional effects
dopamine in reward:  Hedonic impact, reward learning, or on behavior of children. Journal of Psychiatric Research, 17,
incentive salience? Brain Research: Brain Research Reviews, 28, 193–​201.
309–​369. Conway, M., & Pleydell-​Pearce, C. (2000). The construction
Blass, E. M., & Camp, C. A. (2003). Changing determinants of autobiographical memories in the self-​memory system.
of crying termination in 6-​to 12-​week-​old human infants. Psychological Review, 107, 261–​288.
Developmental Psychobiology, 42, 312–​316. Cooper, M., Wells, A., & Todd, G. (2004). A cognitive model of
Blass, E. M., Shide, D. J., & Weller, A. (1989). Stress-​reducing bulimia nervosa. British Journal of Clinical Psychology, 43, 1–​16.
effects of ingesting milk, sugars, and fats: A developmental Craig, A. (1986). Acute effects of meals on perceptual and cogni-
perspective. Annals of the New York Academy of Sciences, 575, tive efficiency. Nutrition Reviews, 44, 163–​171.
292–​305. Craig, A., Baer, K., & Diekmann, A. (1981). The effects of lunch on
Blum, K., Febo, M., Badgaiyan, R. D., Braverman, E. R., Dushaj, sensory-​perceptual functioning in man. International Archives
K., Li, M.,   .  .  .  Demetrovics, Z. (2016). Neuronutrient of Occupational and Environmental Health, 49, 105–​114.
amino-​acid therapy protects against reward deficiency syn- Craighead, L. W., & Allen, H. N. (1995). Appetite awareness
drome:  Dopaminergic key to homeostasis and neuroplasti- training: A cognitive behavioral intervention for binge eat-
city. Current Pharmaceutical Design, July 19 [Epub ahead of ing. Cognitive and Behavioral Practice, 2, 249–​270.
print]. Czaja, J., Rief, W., & Hilbert, A. (2009). Emotion regulation
Boggiano, M. M., Chandler, P. C., Viana, J. B., Oswald, K. D., and binge eating in children. International Journal of Eating
Maldonado, C. R., & Wauford, P. K. (2005). Combined Disorders, 42, 356–​362.
dieting and stress evoke exaggerated responses to opioids in Dallman, M. F., Pecoraro, N., Akana, S. F., la Fleur, S. E.,
binge-​eating rats. Behavioral Neuroscience, 119, 1207–​1214. Gomez, F., Houshyar, H.,  . . .  Manalo, S. (2003). Chronic
Booth, D. A. (1994). Psychology of nutrition. London, UK: Taylor stress and obesity: A new view of “comfort food”. Proceedings
and Francis. of the National Academy of Sciences of the United States of
Brand Miller, J. C., Holt, S. H., de Jong, V & Petocz, P. (2003). America, 100, 11696–​11701.
Cocoa powder increases postprandial insulinemia in lean D’Anci, K. E., Kanarek, R. B., & Marks-​Kaufman, R. (1997).
young adults. Journal of Nutrition, 133, 3149–​3152. Beyond sweet taste: Saccharin, sucrose, and polycose differ in
Braun, D. L., Sunday, S. R., & Halmi, K. A. (1994). Psychiatric their effects upon morphine-​induced analgesia. Pharmacology
comorbidity in patients with eating disorders. Psychological Biochemistry and Behavior, 56, 341–​345.
Medicine, 24, 859–​867. Deaver, C. M., Miltenberger, R. G., Smyth, J., Meidinger, A., &
Bruch, H. (1974) Eating disorders: Anorexia, obesity and the person Crosby, R. (2003). An evaluation of affect and binge eating.
within. London, UK: Routledge and Kegan Paul. Behavior Modification, 27, 578–​599.
Bulik, C. M., Sullivan, P. F., Fear, J. L., & Joyce, P. R. (1997). Dess, N. K., & Edelheit, D. (1998). The bitter with the sweet: The
Eating disorders and antecedent anxiety disorders:  A  con- taste/​stress/​temperament nexus. Biological Psychology, 48,
trolled study. Acta Psychiatrica Scandinavica, 96, 101–​107. 103–​119.
Buree, B. U., Papageorgis, D., & Hare, R. D. (1990). Eating in De Zwaan, M. (2001). Binge eating and obesity. International
anorexia nervosa and bulimia nervosa: An application of the Journal of Obesity, 25, 51–​55.
tripartite model of anxiety. Canadian Journal of Behavioural Dingemans, A. E., Martijn, C., Jansen, A., & van Furth, E. F.
Science, 22, 207–​218. (2009). The effect of suppressing negative emotions on eat-
Bydlowski, S., Corcos, M., Jeammet, P., Paterniti, S., Berthoz, S., ing behavior in binge eating disorder. Appetite, 52, 51–​57.
Laurier, C.,  . . .  Consoli, S. M. (2005). Emotion-​processing Dingemans, A. E., Spinhoven, P., & van Furth, E. F. (2007).
deficits in eating disorders. International Journal of Eating Predictors and mediators of treatment outcome in patients
Disorders, 37, 321–​329. with binge eating disorder. Behaviour Research and Therapy,
Camilleri, G. M., Méjean, C., Kesse-​ Guyot, E., Andreeva, 45, 2551–​2562.
V. A., Bellisle, F., Hercberg, S., & Péneau, S. (2014). The Doba, K., Pezard, L., Lesne, A., Vignau, J., Christophe, V., &
associations between emotional eating and consumption of Nandrino, J.-​L. (2007). Dynamics of emotional expression
energy-​dense snack foods are modified by sex and depressive in autobiographic speech of patients with anorexia nervosa.
symptomatology. Journal of Nutrition, 144, 1264–​1273. Psychological Reports, 101, 237–​249.
Canli, T., Amin, Z., Haas, B., Omura, K., & Constable, R. Donofry, S. D., Roecklein, K. A., Wildes, J. E., Miller, M. A.,
T. (2004). A double dissociation between mood states & Erickson, K. I. (2016). Alterations in emotion generation
and personality traits in the anterior cingulate. Behavioral and regulation neurocircuitry in depression and eating disor-
Neuroscience, 118, 897–​904. ders: A comparative review of structural and functional neu-
Carnell, S., & Wardle, J. (2008). Appetite and adiposity in chil- roimaging studies. Neuroscience and Biobehavioral Reviews,
dren: Evidence for a behavioral susceptibility theory of obe- 68, 911–​927.
sity. American Journal of Clinical Nutrition, 88, 22–​29. Doyle, T. G., Berridge, K. C., & Gosnell, B. A. (1993). Morphine
Cash, T. F., & Deagle, E. A. (1997). The nature and extent of enhances hedonic taste palatability in rats. Pharmacology
body-​image disturbance in anorexia nervosa and bulimia Biochemistry and Behavior, 46, 745–​749.
nervosa:  A  meta-​analysis. International Journal of Eating Dusek, J. B., Mergler, M. L., & Kermis, M. D. (1976). Attention,
Disorders, 22, 107–​125. encoding, and information processing in low-​and high-​test-​
Chua, J. L., Touyz, S., & Hill, A. J. (2004). Negative mood-​ anxious children. Child Development, 47, 201–​207.
induced overeating in obese binge eaters:  An experimental Eddy, K. T., Keel, P. K., Dorer, D. J., Delinsky, S. S., Franko,
study. International Journal of Obesity, 28, 606–​610. D. L., & Herzog, D. B. (2002). Longitudinal comparison

Vögele, Lutz, Gibson 179

 of anorexia nervosa subtypes. International Journal of Eating Geliebter, A., Benson, L., Pantazatos, S. P., Hirsch, J., & Carnell,
Disorders, 31, 191–​201. S. (2016). Greater anterior cingulate activation and connec-
Epel, E., Lapidus, R., McEwen, B., & Brownell, K. (2001). tivity in response to visual and auditory high-​calorie food
Stress may add bite to appetite in women:  A  labora- cues in binge eating:  preliminary findings. Appetite, 96,
tory study of stress-​induced cortisol and eating behavior. 195–​202.
Psychoneuroendocrinology, 26, 37–​49. George, M. S., Nahas, Z., Li, X. B., Kozel, F. A., Anderson, B., &
Espina Eizaguirre, A., Saenz de Cabezón, A. O., Ochoa de Alda, Yamanaka, K. (2002). Potential new brain stimulation thera-
I., Joaristi Olariaga, L., & Juaniz, M. (2004). Alexithymia pies in bipolar illness: Transcranial magnetic stimulation and
and its relationships with anxiety and depression in eat- vagus nerve stimulation. Clinical Neuroscience Research, 2,
ing disorders. Personality and Individual Differences, 36, 256–​265.
321–​331. Gibson, E. L. (2006). Emotional influences on food
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive choice:  Sensory, physiological and psychological pathways.
behaviour therapy for eating disorders:  A  “transdiagnostic” Physiology and Behavior, 89, 53–​61.
theory and treatment. Behaviour Research and Therapy, 41, Gibson, E. L. (2012). The psychobiology of comfort eat-
509–​528. ing:  Implications for neuropharmacological interventions.
Faris, P. L., Eckert, E. D., Kim, S. W., Meller, W. H., Pardo, J. Behavioural Pharmacology, 23, 442–​460.
V., Goodale, R. L., & Hartman, B. K. (2006). Evidence for Gibson, E. L., & Green, M. W. (2002). Nutritional influences on
a vagal pathophysiology for bulimia nervosa and the accom- cognitive function:  Mechanisms of susceptibility. Nutrition
panying depressive symptoms. Journal of Affective Disorders, Research Reviews, 15, 169–​206.
92, 79–​90. Gibson, E. L., Vargas, K., Hogan, E., Holmes, A., Rogers, P. J.,
Faris, P. L., Hofbauer, R. D., Daughters, R., VandenLangenberg, Wittwer, J.,  . . .  Mohajeri, M. H. (2014). Effects of acute
E., Iversen, L., Goodale, R. L.,  . . .  Hartman, B. K. (2008). treatment with a tryptophan-​rich protein hydrolysate on
De-​stabilization of the positive vago-​vagal reflex in bulimia plasma amino acids, mood and emotional functioning in
nervosa. Physiology and Behavior, 94, 136–​153. older women. Psychopharmacology, 231, 4595–​4610.
Farrell, C., Lee, M., & Shafran, R. (2005). Assessment of body Gilboa-​Schechtman, E., Avnon, L., Zubery, E., & Jeczmien, P.
size estimation: A review. European Eating Disorders Review, (2006). Emotional processing in eating disorders:  Specific
13, 75–​88. impairment or general distress related deficiency? Depression
Fernstrom, M. H., & Fernstrom, J. D. (1995). Brain tryptophan and Anxiety, 23, 331–​339.
concentrations and serotonin synthesis remain responsive to Giuliano, C., & Cottone, P. (2015). The role of the opioid system
food consumption after the ingestion of sequential meals. in binge eating disorder. CNS Spectrums, 20, 537–​545.
American Journal of Clinical Nutrition, 61, 312–​319. Goldfein, J. A., Walsh, T., & Midlarsky, E. (2000). Influence
Figlewicz, D. P., & Benoit, S. C. (2009). Insulin, leptin, and of shape and weight on self-​evaluation in bulimia nervosa.
food reward:  Update 2008. American Journal of Physiology International Journal of Eating Disorders, 27, 435–​445.
Regulatory Integrative and Comparative Physiology, 296, Goldfield, G. S., Adamo, K. B., Rutherford, J., & Legg, C.
R9–​R19. (2008). Stress and the relative reinforcing value of food in
Fladung, A.-​K., Grön, G., Grammer, K., Herrnberger, B., Schilly, female binge eaters. Physiology and Behavior, 93, 579–​587.
E., Grasteit, S.,  . . .  von Wietersheim, J. (2010). A neural Gordon, C. M., Dougherty, D. D., Fischman, A. J., Emans, S.
signature of anorexia nervosa in the ventral striatal reward J., Grace, E., Lamm, R.,  . . .  Rauch, S. L. (2001). Neural
system. American Journal of Psychiatry, 167, 206–​212. substrates of anorexia nervosa: A behavioral challenge study
Folkard, S., & Monk, T. H. (1985). Hours of work: Temporal fac- with positron emission tomography. Journal of Pediatrics,
tors in work scheduling. Chichester, UK: Wiley. 139, 51–​57.
Foulon, C., Guelfi, J. D., Kipman, A., Adès, J., Romo, L., Graber, J. A., Brooks-​Gunn, J., Paikoff, R. L., & Warren, M.
Houdeyer, K.,  . . .  Gorwood, P. (2007). Switching to the P. (1994). Prediction of eating problems:  An 8–​year study
binge/​ purging subtype of anorexia nervosa is frequently of adolescent girls. Developmental Psychology, 30, 823–​834.
associated with suicidal attempts. European Psychiatry, 22, Greeno, C. G., Wing, R. R., & Shiffman, S. (2000). Binge ante-
513–​519. cedents in obese women with and with out binge eating disor-
Fredrickson, B. L. (2004). The broaden-​and-​build theory of pos- der. Journal of Consulting and Clinical Psychology, 68, 95–​102.
itive emotions. Philosophical Transactions of the Royal Society Grigson, P. S. (2002). Like drugs for chocolate:  Separate
of London B: Biological Sciences, 359, 1367–​1378. rewards modulated by common mechanisms? Physiology and
Friederich, H.-​C., Uher, R., Brooks, S., Giampietro, V., Brammer, Behavior, 76, 389–​395.
M., Williams, S. C. R.,  . . .  Campbell, I. C. (2007). I’m not as Grilo, C. M., Hrabosky, J. I., White, M. A., Allison, K. C.,
slim as that girl: Neural bases of body shape self-​comparison Stunkard, A. J., & Masheb, R. M. (2008). Overvaluation of
to media images. NeuroImage, 37, 674–​681. shape and weight in binge eating disorder and overweight
Fulkerson, J. A., Sherwood, N. E., Perry, C. L., Neumark-​ controls:  Refinement of a diagnostic construct. Journal of
Sztainer, D., & Story, M. (2004). Depressive symptoms Abnormal Psychology, 117, 414–​419.
and adolescent eating and health behaviors: A multifaceted Grilo, C. M., White, M. A., & Masheb, R. M. (2009). DSM-​IV
view in a population-​based sample. Preventive Medicine, 38, psychiatric disorder comorbidity and its correlates in binge
865–​875. eating disorder. International Journal of Eating Disorders, 42,
Gaudio, S., & Quattrocchi, C. C. (2012). Neural basis of a 228–​234.
multidimensional model of body image distortion in ano- Grucza, R. A., Przybeck, T. R., & Cloninger, C. R. (2007).
rexia nervosa. Neuroscience and Biobehavioral Reviews, 36, Prevalence and correlates of binge eating disorder in a com-
1839–​1847. munity sample. Comprehensive Psychiatry, 48, 124–​131.

180 Mood, Emotions, and Eating Disorders

Guilbaud, O., Corcos, M., Chambry, J., Paterniti, S., Loas, disorders:  Application of risk terminology and suggestions
G., & Jeammet, P. (2000). Alexithymie et dépression dans for a general taxonomy. Psychological Bulletin, 130, 19–​65.
les troubles des conduites alimentaires. [Alexithymia and Jansen, A. (1994). The learned nature of binge eating. In C. R.
depression in eating disorders]. Encéphale, 26, 1–​6. Legg & D. A. Booth (Eds.), Appetite: Neural and behavioural
Haedt-​Matt, A. A., & Keel, P. K. (2011). Revisiting the affect bases (pp. 193–​211). Oxford, UK: Oxford University Press.
regulation model of binge eating:  A  meta-​analysis of stud- Kanarek, R. B., White, E. S., Biegen, M. T., & Marks-​Kaufman, R.
ies using ecological momentary assessment. Psychological (1991). Dietary influences on morphine-​induced analgesia
Bulletin, 137, 660–​681. in rats. Pharmacology Biochemistry and Behavior, 38, 681–​684.
Halmi, K. A., & Sunday, S. R. (1991). Temporal patterns of hun- Kandiah, J., Yake, M., Jones, J., & Meyer, M. (2006). Stress
ger and fullness ratings and related cognitions in anorexia influences appetite and comfort food preferences in college
and bulimia. Appetite, 16, 219–​237. women. Nutrition Research, 26, 118–​123.
Hammersley, R., & Reid, M. (2008). Theorising transient mood Karlsson, J., Persson, L. O., Sjostrom, L., & Sullivan, M. (2000).
after ingestion. Neuroscience and Biobehavioral Reviews, 33, Psychometric properties and factor structure of the Three-​
213–​222. Factor Eating Questionnaire (TFEQ) in obese men and
Haynes, C., Lee, M. D., & Yeomans, M. R. (2003). Interactive women:  Results from the Swedish Obese Subjects (SOS)
effects of stress, dietary restraint, and disinhibition on appe- study. International Journal of Obesity and Related Metabolic
tite. Eating Behaviour, 4, 369–​383. Disorders, 24, 1715–​1725.
Heatherton, T. F., & Baumeister, R. F. (1991). Binge eating as Kaye, W. H. (2008). Neurobiology of anorexia and bulimia ner-
escape from self-​awareness. Psychological Bulletin, 110, 86–​108. vosa. Physiology and Behavior, 94, 121–​135.
Heatherton, T. F., Herman, C. P., & Polivy, J. (1991). Effects of Kelley, A. E., Will, M. J., Steininger, T. L., Zhang, M., & Haber,
physical threat and ego threat on eating behaviour. Journal of S. N. (2003). Restricted daily consumption of a highly palat-
Personality and Social Psychology, 60, 138–​143. able food (chocolate Ensure®) alters striatal enkephalin gene
Herman, C. P., & Polivy, J. (1975). Anxiety, restraint and eating expression. European Journal of Neuroscience, 18, 2592–​2598.
behavior. Journal of Abnormal Psychology, 84, 666–​672. Kenardy, J., Arnow, B., & Agras, W. S. (1996). The aversive-
Herman, C. P., & Polivy, J. (1980). Restrained eating. In A. J. ness of specific emotional states associated with binge-​eating
Stunkard (Ed.), Obesity (pp. 208–​225). Philadelphia:  W. in obese subjects. Australian and New Zealand Journal of
B. Saunders. Psychiatry, 30, 839–​844.
Herpertz-​Dahlmann, B., Dempfle, A., Konrad, K., Klasen, F., Keski-​Rahkonen, A., & Mustelin, L. (2016). Epidemiology of
Ravens-​Sieberer, U. & BELLA Study Group (2015). Eating eating disorders in Europe: Prevalence, incidence, comorbid-
disorder symptoms do not just disappear: The implications ity, course, consequences, and risk factors. Current Opinion
of adolescent eating-​disordered behaviour for body weight in Psychiatry, 29, 340–​345.
and mental health in young adulthood. European Child and Kessler, R. M., Hutson, P. H., Herman, B. K., & Potenza, M. N.
Adolescent Psychiatry, 24, 675–​684. (2016). The neurobiological basis of binge-​eating disorder.
Hetherington, M. M., Cunningham, K., Dye, L., Gibson, E. L., Neuroscience and Biobehavioral Reviews, 63, 223–​238.
Gregersen, N. T., Halford, J. C.,  . . .  Van Trijp, H. C. (2013). Killen, J. D., Taylor, C. B., Hayward, C., Haydel, K. F., Wilson,
Potential benefits of satiety to the consumer: Scientific con- D. M., Hammer, L. D.,  . . .  Strachowski, D. (1996). Weight
siderations. Nutrition Research Reviews, 26, 22–​38. concerns influence the development of eating disorders: A 4-​
Hilbert, A., & Tuschen-​Caffier, B. (2007). Maintenance of binge year prospective study. Journal of Consulting and Clinical
eating through negative mood:  A  naturalistic comparison Psychology, 64, 936–​940.
of binge eating disorder and bulimia nervosa. International Killgore, W. D., & Yurgelun-​Todd, D. A. (2007). Positive affect
Journal of Eating Disorders, 40, 521–​530. modulates activity in the visual cortex to images of high
Hilbert, A., Tuschen-​Caffier, B., & Vögele, C. (2002). Effects of calorie foods. International Journal of Neuroscience, 117,
prolonged and repeated body image exposure in binge-​eating 643–​653.
disorder. Journal of Psychosomatic Research, 52, 137–​144. Kjelsas, E., Borsting, I., & Gudde, C. B. (2004). Antecedents
Hofmann, W., Rauch, W., & Gawronski, B. (2007). And deplete and consequences of binge eating episodes in women with an
us not into temptation: Automatic attitudes, dietary restraint, eating disorder. Eating and Weight Disorders, 9, 7–​15.
and self-​regulatory resources as determinants of eating behav- Koepp, M. J., Hammers, A., Lawrence, A. D., Asselin, M. C.,
ior. Journal of Experimental Social Psychology, 43, 497–​504. Grasby, P. M., & Bench, C. J. (2009). Evidence for endoge-
Hrabosky, J. I., Masheb, R. M., White, M. A., & Grilo, C. M. nous opioid release in the amygdala during positive emotion.
(2007). Overvaluation of shape and weight in binge eating Neuroimage, 44, 252–​256.
disorder. Journal of Consulting and Clinical Psychology, 75, Kubiak, T., Vögele, C., Siering, M., Schiel, R., & Weber, H.
175–​180. (2008). Daily hassles and emotional eating in obese adoles-
Hudson, J. I., Mangweth, B., Pope, H. G., Jr., De Col, C., cents under restricted dietary conditions: The role of rumina-
Hausmann, A., Gutweniger, S.,  . . .  Tsuang, M. T. (2003). tive thinking. Appetite, 51, 206–​209.
Family study of affective spectrum disorder. Archives of Kurosaki, M., Shirao, N., Yamashita, H., Okamoto, Y., &
General Psychiatry, 60, 170–​177. Yamawaki, S. (2006). Distorted images of one’s own body
Jackson, T. D., Grilo, C. M., & Masheb, R. M. (2002). Teasing activates the prefrontal cortex and limbic/​paralimbic sys-
history and eating disorder features: An age-​and body mass tem in young women:  A  functional magnetic resonance
index–​matched comparison of bulimia nervosa and binge-​ imaging study. Biological Psychiatry, 59, 380–​386.
eating disorder. Comprehensive Psychiatry, 43, 108–​113. Lattimore, P., & Caswell, N. (2004). Differential effects of active
Jacobi, C., Hayward, C., de Zwaan, M., Kraemer, H. C., & Agras, and passive stress on food intake in restrained and unre-
W. S. (2004). Coming to terms with risk factors for eating strained eaters. Appetite, 42, 167–​173.

Vögele, Lutz, Gibson 181

Lautenbacher, S., Roscher, S., Strian, F., Pirke, K. M., & Krieg, Macht, M., & Müller, J. (2007b). Immediate effects of choco-
J. C. (1993). Theoretical and empirical considerations on the late on experimentally induced mood states. Appetite, 49,
relation between body image, body scheme and somatosen- 667–​674.
sation. Journal of Psychosomatic Research, 37, 447–​454. Macht, M., Roth, S., & Ellgring, H. (2002). Chocolate eating in
Lavender, J. M., Utzinger, L. M., Cao, L., Wonderlich, healthy men during experimentally induced sadness and joy.
S. A., Engel, S. G., Mitchell, J. E., & Crosby, R. D. (2016). Appetite, 39, 147–​158.
Reciprocal associations between negative affect, binge eating, Mangweth, B., Hudson, J. I., Pope, H. G., Hausmann, A., De
and purging in the natural environment in women with buli- Col, C., Laird, N. M.,  . . .  Tsuang, M. T. (2003). Family
mia nervosa. Journal of Abnormal Psychology, 125, 381–​386. study of the aggregation of eating disorders and mood disor-
Lebovitz, H. E. (2016). Interventional treatment of obesity and ders. Psychological Medicine, 33, 1319–​1323.
diabetes: An interim report on gastric electrical stimulation. Mann, T., & Ward, A. (2004). To eat or not to eat: Implications
Reviews of Endocrine and Metabolic Disorders, 17, 73–​80. of the attentional myopia model for restrained eaters. Journal
Legenbauer, T., Vögele, C., & Rüddel, H. (2004). Anticipatory of Abnormal Psychology, 113, 90–​98.
effects of food exposure in women diagnosed with bulimia Markus, C. R., Olivier, B., Panhuysen, G. E. M., Van der Gugten,
nervosa. Appetite, 42, 33–​40. J., Alles, M. S., Tuiten, A.,  . . .  de Haan, E. E. (2000). The
Leon, G. R., Fulkerson, J. A., Perry, C. L., & Early-​Zald, M. bovine protein alphalactalbumin increases the plasma ratio
B. (1995). Prospective analysis of personality and behavioral of tryptophan to the other large neutral amino acids, and
influences in the later development of disordered eating. in vulnerable subjects raises brain serotonin activity, reduces
Journal of Abnormal Psychology, 104, 140–​149. cortisol concentration, and improves mood under stress.
Leon, G. R., Fulkerson, J. A., Perry, C. L., Keel, P. K., & Klump, American Journal of Clinical Nutrition, 71, 1536–​1544.
K. L. (1999). Three to four year prospective evaluation of Markus, C. R., Panhuysen, G., Tuiten, A., Koppeschaar, H.,
personality and behavioral risk factors for later disordered Fekkes, D., & Peters, M. L. (1998). Does carbohydrate-​rich,
eating in adolescent girls and boys. Journal of Youth and protein-​poor food prevent a deterioration of mood and cog-
Adolescence, 28, 181–​196. nitive performance of stress-​prone subjects when subjected
Lewer, M., Nasrawi, N., Schroeder, D., & Vocks, S. (2016). Body to a stressful task? Appetite, 31, 49–​65.
image disturbance in binge eating disorder: A comparison of Martin, C. K., Han, H., Anton, S. D., Greenway, F. L., & Smith, S.
obese patients with and without binge eating disorder regard- R. (2009). Effect of valproic acid on body weight, food intake,
ing the cognitive, behavioral and perceptual component of physical activity and hormones: Results of a randomized con-
body image. Eating and Weight Disorders, 21, 115–​125. trolled trial. Journal of Psychopharmacology, 23, 814–​825.
Lilenfeld, L. R., Kaye, W. H., Greeno, C. G., Merikangas, K. R., Martinez, D., Saccone, P. A., Liu, F., Slifstein, M., Orlowska,
Plotnicov, K., Pollice, C.,  . . . . Nagy, L. (1998). A controlled D., Grassetti, A.,   .  .  .  Comer, S. D. (2012). Deficits in
family study of anorexia nervosa and bulimia nervosa: psychi- dopamine D receptors and presynaptic dopamine in heroin
atric disorders in first-​degree relatives and effects of proband dependence: Commonalities and differences with other types
comorbidity. Archives of General Psychiatry, 55, 603–​610. of addiction. Biological Psychiatry, 71, 192–​198.
Lowe, M. R., & Fisher, E. B. (1983). Emotional reactivity, emo- Matsunaga, H., Kiriike, N., Iwasaki, Y., Miyata, A., & Matsui, T.
tional eating and obesity:  A  naturalistic study. Journal of (2000). Multi-​impulsivity among bulimic patients in Japan.
Behavioral Medicine, 6, 135–​148. International Journal of Eating Disorders, 27, 348–​352.
Lozano, D. I., Crites, S. L., & Aikmann, S. N. (1999). Changes Matthews, G., & Deary, I. J. (1998). Personality traits. Cambridge,
in food attitudes as a function of hunger. Appetite, 32, UK: Cambridge University Press.
207–​218. Mauler, B. I., Hamm, A. O., Weike, A. I., & Tuschen-​Caffier,
Lutz, A. P. C., Herbert, C., Schulz, A., Voderholzer, U., Koch, B. (2006). Affect regulation and food intake in bulimia ner-
S., & Vögele, C. (2015). Body or cup? Alterations in featural vosa:  Emotional responding to food cues after deprivation
and configural body image processing in anorexia nervosa. and subsequent eating. Journal of Abnormal Psychology, 115,
Psychophysiology, 52(Suppl. 1), S123. 567–​579.
Macdiarmid, J. I., & Hetherington, M. M. (1995). Mood McClelland, J., Bozhilova, N., Campbell, I., & Schmidt, U.
modulation by food:  An exploration of affect and cravings (2013). A systematic review of the effects of neuromodulation
in “chocolate addicts.” British Journal of Clinical Psychology, on eating and body weight: Evidence from human and animal
34, 129–​138. studies. European Eating Disorders Reviews, 21, 436–​455.
Macht, M. (1996). Effects of high-​and low-​energy meals on McElroy, S. L., Crow, S., Blom, T. J., Biernacka, J. M., Winham,
hunger, physiological processes and reactions to emotional S. J., Geske, J.,  . . .  Frye, M. A. (2016). Prevalence and cor-
stress. Appetite, 26, 71–​88. relates of DSM-​5 eating disorders in patients with bipolar
Macht, M. (2008). How emotions affect eating:  A  five-​way disorder. Journal of Affective Disorders, 191, 216–​221.
model. Appetite, 50, 1–​11. McNamara, C., Hay, P., Katsikitis, M., & Chur-​Hansen, A.
Macht, M., Gerer, J., & Ellgring, H. (2003). Emotions in over- (2008). Emotional responses to food, body dissatisfaction
weight and normal-​weight women immediately after eat- and other eating disorder features in children, adolescents
ing foods differing in energy. Physiology and Behavior, 80, and young adults. Appetite, 50, 102–​109.
367–​374. Mercer, M. E., & Holder, M. D. (1997). Food cravings, endog-
Macht, M., Haupt, C., & Salewsky, A. (2004). Emotions and eat- enous opioid peptides, and food intake: A review. Appetite,
ing in everyday life: Application of the experience-​sampling 29, 325–​352.
method. Ecology of Food and Nutrition, 43, 327–​337. Meyer, C., Waller, G., & Waters, A. (1998). Emotional states
Macht, M., & Müller, J. (2007a). Interactive effects of emotional and bulimic psychopathology. In H. Hoek, M. Katzman, &
and restrained eating on responses to chocolate and affect. J. Treasure (Eds.), The neurobiological basis of eating disorders
Journal of Nervous and Mental Disease, 195, 1024–​1026. (pp. 271–​289). Chichester, UK: Wiley.

182 Mood, Emotions, and Eating Disorders

Mitchell, J. E., Mussell, M. P., Peterson, C. B., Crow, S., Patton, G. C., Johnson-​Sabine, E., Wood, K., Mann, A. H., &
Wonderlich, S. A., Crosby, R. D.,  . . .  Weller, C. (1999). Wakeling, A. (1990). Abnormal eating attitudes in London
Hedonics of binge eating in women with bulimia nervosa schoolgirls: A prospective epidemiological study: Outcome at
and binge eating disorder. International Journal of Eating twelve month follow-​up. Psychological Medicine, 20, 383–​394.
Disorders, 26, 165–​170. Patton, G. C., Selzer, R., Coffey, C., Carlin, J. B., & Wolfe, R.
Miyake, Y., Okamoto, Y., Onoda, K., Kurosaki, M., Shirao, N., (1999). Onset of adolescent eating disorders:  Population
Okamoto, Y., & Yamawaki, S. (2010). Brain activation dur- based cohort study over 3 years. British Medical Journal, 318,
ing the perception of distorted body images in eating disor- 765–​768.
ders. Psychiatry Research: Neuroimaging, 181, 183–​192. Pecina, S., Schulkin, J., & Berridge, K. C. (2006). Nucleus
Morris, G. L., 3rd, & Mueller, W. M. (1999). Long-​term treat- accumbens corticotropin-​ releasing factor increases cue-​
ment with vagus nerve stimulation in patients with refractory triggered motivation for sucrose reward: Paradoxical positive
epilepsy. The Vagus Nerve Stimulation Study Group E01–​ incentive effects in stress? BMC Biology, 4, 8. doi: 10.1186/​
E05. Neurology, 53, 1731–​1735. 1741–​7007-​4–​8
Munafò, M. R., Matheson, I. J., & Flint, J. (2007). Association Peschel, S. K., Feeling, N. R., Vögele, C., Kaess, M., Thayer, J.
of the DRD2 gene Taq1A polymorphism and alcohol- F., & Koenig, J. (2016). A meta-​analysis on resting state high-​
ism: A meta-​analysis of case-​control studies and evidence of frequency heart rate variability in bulimia nervosa. European
publication bias. Molecular Psychiatry, 12, 454–​461. Eating Disorders Reviews, 24, 355–​365.
Munsch, S., Michael, T., Biedert, E., Meyer, A. H., & Margraf, Peters, A., Pellerin, L., Dallman, M. F., Oltmanns, K. M.,
J. (2008). Negative mood induction and unbalanced nutri- Schweiger, U., Born, J., & Fehm, H. L. (2007). Causes of
tional style as possible triggers of binges in binge eating disor- obesity:  Looking beyond the hypothalamus. Progress in
der (BED). Eating and Weight Disorders, 13, 22–​29. Neurobiology, 81, 61.
Muraven, M., & Baumeister, R. F. (2000). Self-​regulation and Pinhas, L., Toner, B. B., Ali, A., Garfinkel, P. E., & Stuckless,
depletion of limited resources: Does self-​control resemble a N. (1999). The effects of the ideal of female beauty on
muscle? Psychological Bulletin, 126, 247–​259. mood and body satisfaction. International Journal of Eating
Nandrino, J.-​L., Doba, K., Lesne, A., Christophe, V., & Pezard, Disorders, 25, 223–​226.
L. (2006). Autobiographical memory deficit in anorexia ner- Pohjalainen, T., Rinne, J. O., Nagren, K., Lehikoinen, P., Anttila,
vosa:  Emotion regulation and effect of duration of illness. K., Syvalahti, E. K., & Hietala, J. (1998). The A1 allele of the
Journal of Psychosomatic Research, 61, 537–​543. human D2 dopamine receptor gene predicts low D2 recep-
Nanni, G., Scheggi, S., Leggio, B., Grappi, S., Masi, F., Rauggi, tor availability in healthy volunteers. Molecular Psychiatry, 3,
R., & De Montis, M. G. (2003). Acquisition of an appeti- 256–​260.
tive behavior prevents development of stress-​induced neuro-​ Polivy, J., & Herman, C. P. (1999). Distress and eating:  Why
chemical modifications in rat nucleus accumbens. Journal of do dieters overeat? International Journal of Eating Disorders,
Neuroscience Research, 73, 573–​580. 26, 153–​164.
Newman, E., O’Connor, D. B., & Conner, M. (2007). Daily Polivy, J., & Herman, C. P. (2002). Causes of eating disorders.
hassles and eating behaviour:  The role of cortisol reactivity Annual Review of Psychology, 53, 187–​213.
status. Psychoneuroendocrinology, 32, 125–​132. Pollatos, O., Herbert, B. M., Schandry, R., & Gramann, K.
Nolen-​Hoeksema, S. (1991). Responses to depression and their (2008). Impaired central processing of emotional faces in
effects on the duration of depressive episodes. Journal of anorexia nervosa. Psychosomatic Medicine, 70, 701–​708.
Abnormal Psychology, 100, 569–​582. Pool, E., Delplanque, S., Coppin, G., & Sander, D. (2015). Is
Nolen-​Hoeksema, S., Stice, E., Wade, E., & Bohon, C. (2007). comfort food really comforting? Mechanisms underlying stress-​
Reciprocal relations between rumination and bulimic, sub- induced eating. Food Research International, 76, 207–​215.
stance abuse, and depressive symptoms in female adoles- Preston, C., & Ehrsson, H. H. (2014). Illusory changes in body
cents. Journal of Abnormal Psychology, 116, 198–​207. size modulate body satisfaction in a way that is related to
O’Hara, C. B., Campbell, I. C., & Schmidt, U. (2015). A non-​clinical eating disorder psychopathology. PLoS ONE, 9,
reward-​centred model of anorexia nervosa: A focussed narra- e85773.
tive review of the neurological and psychophysiological liter- Raes, F., Hermans, D., De Decker, A., Eelen, P., & Williams
ature. Neuroscience and Biobehavioral Reviews, 52, 131–​152. J. (2003). Autobiographical memory specificity and affect
O’Hara, C. B., Keyes, A., Renwick, B., Giel, K. E., Campbell, I. regulation: An experimental approach. Emotion, 3, 201–​206.
C., & Schmidt, U. (2016). Evidence that illness-​compatible Rastam, M. (1992). Anorexia nervosa in 51 Swedish adoles-
cues are rewarding in women recovered from anorexia ner- cents:  Premorbid problems and comorbidity. Journal of the
vosa: A study of the effects of dopamine depletion on eye-​ American Academy of Child and Adolescent Psychiatry, 31,
blink startle responses. PLoS One, 11, e0165104. 819–​829.
Oliver, G., & Wardle, J. (1999). Perceived effects of stress on Reichel, V. A., Schneider, N., Grünewald, B., Kienast, T.,
food choice. Physiology and Behavior, 66, 511–​515. Pfeiffer, E., Lehmkuhl, U., & Korte, A. (2014). “Glass fair-
Oliver, G., Wardle, J., & Gibson, E. L. (2000). Stress and food ies” and “bone children”: Adolescents and young adults with
choice:  A  laboratory study. Psychosomatic Medicine, 62, anorexia nervosa show positive reactions towards extremely
853–​865. emaciated body pictures measured by the startle reflex para-
Pani, L., Porcella, A., & Gessa, G. L. (2000). The role of stress in digm. Psychophysiology, 51, 168–​177.
the pathophysiology of the dopaminergic system. Molecular Reid, M., & Hammersley, R. (1999). The effects of carbohy-
Psychiatry, 5, 14–​21. drates on arousal. Nutrition Research Reviews, 12, 3–​23.
Papezova, H., Yamamotova, A., & Uher, R. (2005). Elevated Ricciardelli, L. A., Tate, D., & Williams, R. J. (1997). Body dis-
pain threshold in eating disorders: Physiological and psycho- satisfaction as a mediator of the relationship between dietary
logical factors. Journal of Psychiatric Research, 39, 431–​438. restraint and bulimic eating patterns. Appetite, 29, 43–​54.

Vögele, Lutz, Gibson 183

Rodin, J. (1981). Current status of the internal/​external hypoth- Sifneos, P. E. (1991). Affect, emotional conflict, and deficit: An
esis for obesity: What went wrong? American Psychologist, 36, overview. Psychotherapy and Psychosomatics, 56, 116–​122.
361–​372. Slade, P., & Brodie, D. (1994). Body-​image distortion and eating
Rodriguez, S., Fernandez, M. C., Cepeda-​Benito, A., & Vila, disorder:  A  reconceptualization based on the recent litera-
J. (2005). Subjective and physiological reactivity to choco- ture. European Eating Disorders Review, 2, 32–​46.
late images in high and low chocolate cravers. Biological Small, D. M., Zatorre, R. J., Dagher, A., Evans, A. C., & Jones-​
Psychology, 70, 9–​18. Gotman, M. (2001). Changes in brain activity related to
Rodriguez-​ Cano, T., Beato-​ Fernandez, L., Garcia-​ Vilches, I., eating chocolate:  From pleasure to aversion. Brain, 124,
Garcia-​Vicente, A., Poblete-​Garcia, V., & Soriano-​Castrejon, 1720–​1733.
A. (2009). Regional cerebral blood flow patterns of change Smeets, M. A.  M. (1997). The rise and fall of size estimation
following the own body image exposure in eating disor- research in anorexia nervosa: A review and reconceptualiza-
ders: A longitudinal study. European Psychiatry, 24, 275–​281. tion. European Eating Disorders Review, 5, 75–​95.
Rojo-​Moreno, L., Arribas, P., Plumed, J., Gimeno, N., Garcia-​ Smit, H. J., Gaffan, E. A., & Rogers, P. J. (2004). Methylxanthines
Blanco, A., Vaz-​Leal, F.,  . . .  Livianos, L. (2015). Prevalence and are the psychopharmacologically active constituents of choc-
comorbidity of eating disorders among a community sample of olate. Psychopharmacology, 176, 412–​419.
adolescents: 2-​year follow-​up. Psychiatry Research, 227, 52–​57. Smith, A. P., & Miles, C. (1986). Acute effects of meals, noise
Rolls, E. T. (2007). Emotion explained. Oxford, UK:  Oxford and nightwork. British Journal of Psychology, 77, 377–​387.
University Press. Smyth, J. M., Wonderlich, S. A., Heron, K. E., Sliwinski, M.
Rose, N., Koperski, S., & Golomb, B. A. (2010). Mood J., Crosby, R. D., Mitchell, J. E., & Engel, S. G. (2007).
food: Chocolate and depressive symptoms in a cross-​sectional Daily and momentary mood and stress are associated with
analysis. Archives of Internal Medicine, 170, 699–​703. binge eating and vomiting in bulimia nervosa patients in
Rutters, F., Nieuwenhuizen, A. G., Lemmens, S. G., Born, J. the natural environment. Journal of Consulting and Clinical
M., & Westerterp-​Plantenga, M. S. (2009). Acute stress-​ Psychology, 75, 629–​638.
related changes in eating in the absence of hunger. Obesity, Södersten, P., Nergardh, R., Bergh, C., Zandian, M., &
17, 72–​77. Scheurink, A. (2008). Behavioral neuroendocrinology and
Saelens, B. E., & Epstein, L. H. (1996). Reinforcing value of treatment of anorexia nervosa. Frontiers in Neuroendocrinology,
food in obese and non-​obese women. Appetite, 27, 41–​50. 29, 445–​462.
Santel, S., Baving, L., Krauel, K., Münte, T. F., & Rotte, M. Spangler, D. L., & Allen, M. D. (2011). An fMRI investigation
(2006). Hunger and satiety in anorexia nervosa: fMRI during of emotional processing of body shape in bulimia nervosa.
cognitive processing of food. Brain Research, 1114, 138–​148. International Journal of Eating Disorders, 45, 17–​25.
Schachter, S., Goldman, R., & Gordon, A. (1968). Effects of Spoor, S. T.  P., Bekker, M. H.  J., van Strien, T., & van Heck,
fear, food deprivation, and obesity on eating. Journal of G. L. (2007). Relations between negative affect, coping, and
Personality and Social Psychology, 10, 91–​97. emotional eating. Appetite, 48, 368–​376.
Schepers, R., & Markus, C. R. (2015). Gene x cognition Staiger, P., Dawe, S., & McCarthy, R. (2000). Responsivity to food
interaction on stress-​induced eating:  Effect of rumination. cues in bulimic women and controls. Appetite, 35, 27–​33.
Psychoneuroendocrinology, 54, 41–​53. Stein, R. I., Kenardy, J., Wiseman, C. V., Zoler Dounchis, J.,
Schienle, A., Schäfer, A., Hermann, A., & Vaitl, D. (2009). Arnow, B. A., & Wilfley, D. E. (2007). What’s driving the
Binge-​eating disorder:  Reward sensitivity and brain activa- binge in binge eating disorder? A prospective examination of
tion to images of food. Biological Psychiatry, 65, 654–​661. precursors and consequences. International Journal of Eating
Seeger, G., Braus, D. F., Ruf, M., Goldberger, U., & Schmidt, M. Disorders, 40, 195–​203.
H. (2002). Body image distortion reveals amygdala activa- Stice, E. (1994). Review of the evidence for a sociocultural model
tion in patients with anorexia nervosa: A functional magnetic of bulimia nervosa and an exploration of the mechanisms of
resonance imaging study. Neuroscience Letters, 326, 25–​28. action. Clinical Psychology Review, 14, 633–​661.
Sepúlveda, A. R., Botella, J., & León, J. A. (2002). Body-​image Stice, E. (2001). A prospective test of the dual-​pathway model of
disturbance in eating disorders:  A  meta-​analysis. Psychology bulimic pathology: Mediating effects of dieting and negative
in Spain, 6, 83–​95. affect. Journal of Abnormal Psychology, 110, 124–​155.
Seymour, K. E., Reinblatt, S. P., Benson, L., & Carnell, S. Stice, E., Akutagawa, D., Gaggar, A., & Agras, W. S. (2000).
(2015). Overlapping neurobehavioral circuits in ADHD, Negative affect moderates the relation between dieting and
obesity, and binge eating:  Evidence from neuroimaging binge eating. International Journal of Eating Disorders, 27,
research. CNS Spectrums, 20, 401–​411. 218–​229.
Shaffer, H. J., LaPlante, D. A., LaBrie, R. A., Kidman, R. C., Stice, E., Gau, J. M., Rohde, P., & Shaw, H. (2016). Risk factors
Donato, A. N., & Stanton, M. V. (2004). Toward a syn- predict future onset of each DSM-​5 Eating Disorder: Predictive
drome model of addiction:  Multiple expressions, common specificity in high-​risk adolescent females. Journal of Abnormal
etiology. Harvard Review of Psychiatry, 12, 367–​374. Psychology, Oct 6 [Epub ahead of print].
Shirao, N., Okamoto, Y., Okada, G., Okamoto, Y., & Yamawaki, Stice, E., Hayward, C., Cameron, R. P., Killen, J. D., & Taylor,
S. (2003). Temporomesial activation in young females asso- C. B. (2000). Body-​image and eating disturbances predict
ciated with unpleasant words concerning body image. onset of depression among female adolescents: A longitudi-
Neuropsychobiology, 48, 136–​142. nal study. Journal of Abnormal Psychology, 109, 438–​444.
Siegle, G. J., & Thayer, J. T. (2004). Physiological aspects of Stice, E., & Yokum, S. (2016). Gain in body fat is associated
depressive rumination. In C. Papageorgiou & A. Wells with increased striatal response to palatable food cues,
(Eds.), Depressive rumination nature theory and treatment (pp. whereas body fat stability is associated with decreased striatal
79–​104). New York, NY: Wiley. response. Journal of Neuroscience, 36, 6949–​6956.

184 Mood, Emotions, and Eating Disorders

Stice, E., Yokum, S., Bohon, C., Marti, N., & Smolen, A. neuroanatomy of body shape perception in healthy and
(2010). Reward circuitry responsivity to food predicts future eating-​disordered women. Biological Psychiatry, 58, 990–​997.
increases in body mass:  Moderating effects of DRD2 and Ulfvebrand, S., Birgegard, A., Norring, C., Hogdahl, L. & von
DRD4. Neuroimage, 50, 1618–​1625. Hausswolff-​ Juhlin, Y. (2015). Psychiatric comorbidity in
Swami, V., Frederick, D. A., Aavik, T., Alcalay, L., Allik, J., women and men with eating disorders: Results from a large
Anderson, D.,  . . .  Zivcic-​Becirevic, I. (2010). The attrac- clinical database. Psychiatry Research, 230, 294–​299.
tive female body weight and female body dissatisfaction Upadhyay, A., Aggarwal, R., Narayan, S., Joshi, M., Paul, V. K.,
in 26 countries across 10 world regions:  Results of the & Deorari, A. K. (2004). Analgesic effect of expressed breast
International Body Project I. Personality and Social Psychology milk in procedural pain in term neonates:  A  randomized,
Bulletin, 36, 309–​325. placebo-​controlled, double-​blind trial. Acta Paediatrica, 93,
Takimoto, Y., Yoshiuchi, K., & Akabayashi, A. (2008). Effect of 518–​522.
mood states on QT interval and QT dispersion in eating disor- Van der Ham, T., Meulman, J. J., van Strien, D. C., & van
der patients. Psychiatry and Clinical Neurosciences, 62, 185–​189. Engeland, H. (1997). Empirically based subgrouping of
Taylor, G. J., Bagby, R. M., & Parker, J. D. A. (1991). The alexi- eating disorders in adolescents:  A  longitudinal perspective.
thymia construct:  A  potential paradigm for psychosomatic British Journal of Psychiatry, 170, 363–​368.
medicine. Psychosomatics, 32, 153–​164. Vaz, F. J., Alcaina, T., & Guisado, J. A. (1998). Food aversions
Taylor, G. J., Bagby, R. M., & Parker, J. D. A. (1997). Disorders in eating disorders. International Journal of Food Sciences and
of affect regulation: Alexithymia in medical and psychiatric ill- Nutrition, 49, 181–​186.
ness. Cambridge, UK: Cambridge University Press. Vocks, S., Busch, M., Schulte, D., Grönemeyer, D., Herpertz,
Taylor, G. J., Parker, J. D. A., Bagby, R. M., & Bourke, M. P. S., & Suchan, B. (2010). Effects of body image therapy on
(1996). Relationships between alexithymia and psychologi- the activation of the extrastriate body area in anorexia ner-
cal characteristics associated with eating disorders. Journal of vosa: An fMRI study. Psychiatry Research, 183, 114–​118.
Psychosomatic Research, 41, 561–​568. Vocks, S., Legenbauer, T., Wächter, A., Wucherer, M., &
Telch, C. F. (1997). Skills training treatment for adaptive Kosfelder, J. (2007). What happens in the course of body
affect regulation in a woman with binge-​eating disorder. exposure? Emotional, cognitive, and physiological reac-
International Journal of Eating Disorders, 22, 77–​81. tions to mirror confrontation in eating disorders. Journal of
Telch, C. F., & Agras, S. W. (1996). Do emotional states influ- Psychosomatic Research, 62, 231–​239.
ence binge eating in the obese? International Journal of Eating Vögele, C., & Florin, I. (1997). Psychophysiological responses
Disorders, 20, 271–​279. to food exposure:  An experimental study in binge eaters.
Thayer, R. E. (1989). The biopsychology of mood and arousal. International Journal of Eating Disorders, 21, 147–​157.
Oxford, UK: Oxford University Press. Vögele, C., Hilbert, A., & Tuschen-​Caffier, B. (2009). Dietary
Thomsen, D. K. (2006). The association between rumination restriction, cardiac autonomic regulation and stress reactivity
and negative affect:  A  review. Cognition and Emotion, 20, in bulimic women. Physiology and Behavior, 98, 229–​234.
1216–​1235. Vögele, C., & Woodward, H. (2005). Körperbild,
Torres, S. J., & Nowson, C. A. (2007). Relationship between Diätverhalten und körperliche Aktivität bei 9–​10 jähri-
stress, eating behavior, and obesity. Nutrition, 23, 887–​894. gen Kindern [Body image, dietary behaviour and phys-
Trentowska, M., Bender, C., & Tuschen-​ Caffier, B. (2013). ical activity in 9–​10  year old children]. Kindheit und
Mirror exposure in women with bulimic symptoms:  How Entwicklung (Themenheft Essstörungen bei Kindern und
do thoughts and emotions change in body image treatment? Jugendlichen), 14, 229–​236.
Behaviour Research and Therapy, 51, 1–​6. Wagner, A., Aizenstein, H., Venkatraman, V. K., Fudge, J., May,
Tuominen, L., Tuulari, J., Karlsson, H., Hirvonen, J., Helin, S., J. C., Mazurkewicz, L.,  . . .  Kaye, W. H. (2007). Altered
Salminen, P.,  . . .  Nummenmaa, L. (2015). Aberrant meso- reward processing in women recovered from anorexia ner-
limbic dopamine-​opiate interaction in obesity. Neuroimage, vosa. American Journal of Psychiatry, 164, 1842–​1849.
122, 80–​86. Wagner, A., Ruf, M., Braus, D., & Schmidt, M. (2003). Neuronal
Tuomisto, T., Hetherington, M. M., Morris, M. F., Tuomisto, M. activity changes and body image distortion in anorexia ner-
T., Turjanmaa, V., & Lappalainen, R. (1999). Psychological vosa. NeuroReport, 14, 2193–​2197.
and physiological characteristics of sweet food “addiction.” Wallis, D. J., & Hetherington, M. M. (2004). Stress and eat-
International Journal of Eating Disorders, 25, 169–​175. ing: The effects of ego-​threat and cognitive demand on food
Tuschen-​Caffier, B., Vögele, C., Bracht, S., & Hilbert, A. (2003). intake in restrained and emotional eaters. Appetite, 43, 39–​46.
Psychological responses to body shape exposure in patients Walsh, B. T., & Boudreau, G. (2003). Laboratory studies of binge
with bulimia nervosa. Behaviour Research and Therapy, 41, eating disorder. International Journal of Eating Disorders, 34
573–​586. (Suppl), 30–​38.
Uher, R., Brammer, M. J., Murphy, T., Campbell, I. C., Ng, V. Wang, G. J., Volkow, N. D., Logan, J., Pappas, N. R., Wong, C.
W., Williams, S. C. R., & Treasure, J. (2003). Recovery and T., Zhu, W.,  . . .  Fowler, J. S. (2001). Brain dopamine and
chronicity in anorexia nervosa: Brain activity associated with obesity. Lancet, 357, 354–​357.
differential outcomes. Biological Psychiatry, 54, 934–​942. Wang, G. J., Volkow, N. D., Thanos, P. K., & Fowler, J. S.
Uher, R., Murphy, T., Brammer, M. J., Dalgleish, T., Phillips, M. (2004). Similarity between obesity and drug addiction as
L., Ng, V. W.,  . . .  Treasure, J. (2004). Medial prefrontal cor- assessed by neurofunctional imaging:  A  concept review.
tex activity associated with symptom provocation in eating Journal of Addictive Diseases, 23, 39–​53.
disorders. American Journal of Psychiatry, 161, 1238–​1246. Ward, A., & Mann, T. (2000). Don’t mind if I do: Disinhibited
Uher, R., Murphy, T., Friederich, H. C., Dalgleish, T., Brammer, eating under cognitive load. Journal of Personality and Social
M. J., Giampietro, V.,  . . .  Treasure, J. (2005). Functional Psychology, 78, 753–​763.

Vögele, Lutz, Gibson 185

Wardle, J., Chida, Y., Gibson, E. L., Whitaker, K. L., & Steptoe, Williams, J. M. G., Healy, H., Eade, J., Windle, G., Cowen,
A. (2011). Stress and adiposity: A meta-​analysis of longitudi- P. J., Green, M. W., & Durlach, P. (2002). Mood, eat-
nal studies. Obesity, 19, 771–​778. ing behaviour and attention. Psychological Medicine, 32,
Wardle, J., & Gibson, E. L. (2002). Impact of stress on 469–​481.
diet:  Processes and implications In S. Stansfeld & M. G. Winham, S. J., Cuellar-​ Barboza, A. B., McElroy, S. L.,
Marmot (Eds.), Stress and the heart: Psychosocial pathways to cor- Oliveros, A., Crow, S., Colby, C. L.,  . . .  Biernacka, J. M.
onary heart disease (pp. 124–​149). London, UK: BMJ Books. (2014). Bipolar disorder with comorbid binge eating his-
Wardle, J., Guthrie, C., Sanderson, S., Birch, L., & Plomin, R. tory:  A  genome-​wide association study implicates APOB.
(2001). Food and activity preferences in children of lean and Journal of Affective Disorders, 165, 151–​158.
obese parents. International Journal of Obesity and Related Yanovski, S. Z., Nelson, J. E., Dubbert, B. K., & Spitzer, R. L.
Metabolic Disorders, 25, 971–​977. (1993). Association of binge eating disorder and psychiatric
Waters, A., Hill, A., & Waller, G. (2001). Bulimic’s responses to comorbidity in obese subjects. American Journal of Psychiatry,
food cravings: I. Binge-​eating a product of hunger or emo- 150, 1472–​1479.
tional state? Behaviour Research and Therapy, 39, 877–​886. Yeomans, M. R., & Wright, P. (1991). Lower pleasantness of
Weinstein, S. E., Shide, D. J., & Rolls, B. J. (1997). palatable foods in nalmefene-​ treated human volunteers.
Changes in food intake in response to stress in men and Appetite, 16, 249–​259.
women: Psychological factors. Appetite, 28, 7–​18. Zagon, A. (2001). Does the vagus nerve mediate the sixth sense?
Whiteside, U., Chen, E., Neighbors, C., Hunter, D., Lo, T., & Trends in Neurosciences, 24, 671–​673.
Larimer, M. (2007). Difficulties regulating emotions:  Do Zellner, D. A., Loaiza, S., Gonzalez, Z., Pita, J., Morales, J.,
binge eaters have fewer strategies to modulate and tolerate Pecora, D., & Wolf, A. (2006). Food selection changes under
negative affect? Eating Behaviors, 8, 162–​169. stress. Physiology and Behavior, 87, 789–​793.
Wilfley, D. E., Schwartz, M. B., Spurrell, E. B., & Fairburn, Zonnevijlle-​Bender, M. J., Van Goozen, S. H., Cohen-​Kettenis,
C. G. (2000). Using the Eating Disorder Examination to P. T., & Van Engeland, H. (2002). Do adolescent anorexia
identify the specific psychopathology of binge eating disor- nervosa patients have deficits in emotional functioning?
der. International Journal of Eating Disorders, 27, 259–​269. European Child and Adolescent Psychiatry, 11, 38–​42.
Wilfley, D. E., Wilson, G. T., & Agras, W. S. (2003). The clinical
significance of binge eating disorder. International Journal of
Eating Disorders, 34 (Supplement), S96–​S106.

186 Mood, Emotions, and Eating Disorders

 CH A PT E R


Cultural Influences on Body Image and
9 Eating Disorders

Eileen P. Anderson-​Fye

Abstract
Sociocultural factors have long been implicated in body image and eating disorders. Decades of data,
drawn from multiple disciplines, consistently demonstrate the influence of culture on body image and
eating disorders across several levels of analysis. This chapter engages the rich empirical literature on this
subject to retheorize the role and importance of these contextual factors in light of anthropological and
related social theories relevant to contemporary circumstances. Specifically, this chapter first analyzes
and operationalizes what we mean by “culture” in body image and eating disorder scholarship, describes
trends in salient sociocultural factors, and highlights the varying impacts of globalization where societies
are increasingly interconnected. It also urges research that builds on current understandings by increasing
collaborations among not only multiple disciplines within the social sciences but also biological and clinical
sciences.
Key Words:  body image, culture, cross-​cultural, eating disorder, globalization, interdisciplinary,
intersectionality

Introduction body image and disordered eating behavior. Finally

Sociocultural factors have long been implicated increased collaborative interdisciplinary research is
in body image and eating disorders. Data support- proposed.
ing the influence of culture on body image and our
contemporary forms of disordered eating come What Is Culture?
from multiple streams of evidence that vary both Any exploration of these subjects first requires
interculturally (i.e., across cultures) and intracultur- understanding of what we mean by “culture” and
ally (i.e., within any given culture) over time. This “sociocultural factors.” Related disciplines use
chapter reviews the multidisciplinary evidence for these terms in distinct ways. These differences are
the importance of culture on macro, meso, and meaningful in an inherently interdisciplinary area
micro levels of analysis as relates to both body image of study that includes psychiatry and several other
and disordered eating. It describes the structural, biomedical specialties—​among them, gastrointesti-
behavioral, and ideological aspects of culture that nal medicine and nutrition—​as well as psychology,
influence body image and eating in many types of anthropology, sociology, history, and gender studies.
societies. The field has accumulated several decades Anthropology, arguably the discipline most respon-
of examination of cultural variations in body image sible for studying culture, itself has many internal
and eating outcomes interculturally and intracultur- definitions of the term. However, at the core, most
ally. This chapter draws on these collective findings of these definitions hold three components. First,
to suggest shifts in the way that the field consid- culture has semiotic, meaning-​based components
ers and investigates demographic and other cul- including ideas, ideals, and beliefs. Second, it has
tural factors and contexts that systematically impact behavioral components such as actions, rituals, and

187
habits. Finally, these meaning-​and behavior-​based understanding health challenges and mental illness
components are shared by a social group. Therefore, widely applied to body image and eating concerns
this chapter takes culture to include beliefs and (Maine, Samuels, & Tantillo, 2015). In it, health
behaviors shared by a meaningful social group. outcomes are considered to emerge from a com-
Contemporary psychiatry and psychology bination of factors, including biology or genetics,
are more likely to describe “culture” as a variable psychology or individual-​level attributes, and social
belonging to racial or ethnic groups. In most of these or group-​level factors. Since its introduction nearly
fields’ texts on body image and eating disorders, four decades ago, this tripartite structure diversified
culture may include nationality, but usually reflects psychological analysis and provided more attention
a society’s major racial groups—​such as African-​ to multilevel interactions that gained traction with
descended or Black, Hispanic or Latino, Asian clinicians (Borrell-​Carrió, Suchman, & Epstein,
and Pacific Islander, Native American, and White 2004; Engel, 1980). While the model remains
or Caucasian (cf. Betancourt & López, 1993). The robust, medical and psychological anthropologists,
texts may further subdivide these groups into eth- among others, continue to argue that it underes-
nic divisions such as Italian American, Caribbean timates the impact that culture has on all of these
British, or Middle Eastern origin. Religious levels (Burkett, 1991; Lester, 2007; Phillips et  al.,
affiliation—​for example, Jewish or Irish Catholic—​ 2012; Taylor 2016).
can also constitute “culture” in these fields. A bene- From an anthropological viewpoint, every aspect
fit of this sort of group-​level analysis is that racial or of human functioning is deeply cultural, including
ethnic identity often includes certain kinds of food the institutions that humans create. From intergen-
and eating beliefs and behaviors that are relevant. erational transmission of genetics and customs, to
A  drawback of this definition of culture is that it prenatal shaping, to human lifespan development,
can oversimplify complex issues and thus obscure cultural context impacts all aspects of biology and
key theoretical or clinical factors in body image and psychology. Further, models of healthy human
eating disorders. development are deeply imbued with cultural val-
For example, African Americans in the United ues. Anthropologists see these models of typical
States are a diverse racial grouping that can include or ideal psychological functioning as ethnopsy-
individuals descended from slaves, recent immi- chologies (Gaines, 1992; Lillard, 1998; Lutz, 1983;
grants from sub-​Saharan Africa or the Caribbean, or White, 1992). That is, they are models of develop-
those of ethnically and racially diverse backgrounds ment bounded by space and time. This assumption
that are generalized based on phenotype, especially is in contrast to dominant academic psychology,
but not solely based on skin color. This group- which attempts to provide models and mechanisms
ing also may include wide variation in wealth and of universal human behavior.
access to resources as well as differences in national For example, an increase in autonomy across
region and urban, suburban, and rural location. As adolescence is an important developmental task
discussed later regarding intersectionality, actions in contemporary psychology. In fact, unsuccess-
taken based on such oversimplification of ideas can ful individuation from parents is implicated as
lead to profoundly negative health consequences an important factor in disordered eating. When
for subgroups and individuals with respect to body we look cross-​ culturally, however, we see highly
image and eating. One example of this problem variable models of autonomy in comparison to
involves a point of conventional wisdom prevalent interconnectedness. Many cultures in Asia, Latin
in psychological literature until recently:  “African America, and the Pacific Islands are thought to be
American women prefer larger body sizes and there- more sociocentric than individualistic (Marcus &
fore have more positive body image” (cf. McClure, Kitayama, 1991; Triandis, 1995, 2001). Therefore,
2017). Because of this stereotype, African American “healthy” adolescent development in these contexts
women have been underdiagnosed when suffering means navigating interconnectedness with close and
with eating disorders; as a result, by the time of extended family. Such models are reflected in insti-
diagnosis the disease often is more severe (Gordon, tutional treatment for disordered eating across con-
Brattole, Wingate, & Joiner, 2006; Gordon, Perez, texts (e.g., Lester, 2007, Kempa & Thomas, 2007).
& Joiner, 2002). Conflicts around these models of “healthy” devel-
Psychology has provided scholars and prac- opment and autonomy arise intraculturally as well,
titioners the biopsychosocial approach (Engel, especially in cases of immigrant and refugee youth,
1977), a dominant interdisciplinary model for who now make up approximately one-​quarter of all

188 Cultural Influences on Body Image

school-​aged children in the United States (Child each form has variable emergence and prevalence
Trends, 2014) and who are at particular risk for per society through cultural history (Anderson-​Fye
disordered eating (Kimber, Couturier, Georgiades, & Becker, 2003). These histories provide insight
Wahoush, & Jack, 2014). Contemporary waves of into societal conditions that make eating disorders
immigrant youth tend to come from societies clas- more likely. In short, eating disorders are temporally
sified as more sociocentric than the United States, and spatially unique expressions of distress that are
Canada, Australia, and western Europe. Families not found in all societies or throughout Western1
have usually sacrificed to provide these children history. Across contexts, these disorders tend to
greater opportunity in a wealthier or less volatile incorporate suffering of mood, anxiety, familial ten-
country. In these cases (discussed in more detail sions, and gendered ideals, among other concerns.
later), an individual young person may experience Taken together, the meanings and uses of “cul-
conflict between familial and institutional models ture” and related terms, such as “sociocultural,” are
of development that in turn leads to body image important vehicles by which to better understand
or eating problems (National Research Council and intervene in body image and eating disorders.
& Institute of Medicine, 1998; Kim, Chen, Li, No one field can describe adequately the full impact
Huang, & Moon, 2009;). In other cases, psycholog- of cultural aspects of health and illness. Rather,
ical therapeutic models may need to become more scholars in many disciplines contribute useful ways
pluralistic to treat diverse youth successfully. to think about and work with related issues on mul-
Sociologists and historians also have contributed tiple levels from the individual clinical encounter to
substantially to understandings of the structural group-​level policies.
and time-​bounded limits of body image and eat-
ing disorders. Medical sociologists are interested How Does Culture Impact Body Image
in the structural (societal) components of health and Eating Concerns?
inequalities. They study how race, class, gender, Given what we know about culture from
sexuality, and other categories affect not only health such disciplinary perspectives, how does culture
outcomes like eating disorders, but also the profes- shape and affect body image and eating concerns?
sional cultures that treat sick individuals (Boulware, Providing a meaningful answer to this question
Cooper, Ratner, LaVeist, & Powe, 2003; Williams requires multiple levels of analysis across several
& Mohammed, 2009; Williams & Sternthal, domains of relevant impact.
2010). Sociological scholarship on eating disorders
addresses differential access to healthcare (Becker, Large-​Scale Environmental and
Hadley Arrindell, Perloe, Fay, & Striegel-​Moore, Political-​Economic Variables
2010; Williams & Mohammed, 2009; Williams & On the broadest level, environmental and
Sternthal, 2010) and structural risk factors for the political-​economic factors can help create environ-
disorders (Nagel & Jones, 1992; Stice, 2001). These ments for the emergence of eating disorders as well
sociological categories of inquiry are engaged along as their patterning. Key environmental influences
with culture per se in the commonly used term may include climate, terrain, and urban versus rural
“sociocultural.” residence. Such elements relate to communities’ or
For their parts, historians illustrate how notions families’ ability to grow enough healthful food, and
of health, disease, healing, and bodies can vary in also shape variations in lifestyle (e.g., in a desert,
the same culture over time. In this way, they help mountainous region, or plains). Political-​economic
establish understanding of the sociocultural compo- variables of prime importance to body image and
nents of body image and eating disorders. In the eating disorders include colonial and postcolonial
field of body image and eating concerns, scholars status, industrialization level, and wealth disparities.
such as Joan Jacobs Brumberg (1988) and Carolyn Yet other large-​scale components related to disease
Walker Bynum (1988) have been particularly influ- and political economy (such as disease burden) also
ential in documenting when eating disorders such as influence body image and eating patterns.
anorexia nervosa entered the cultural repertoire—​as One apt example of the kinds of effects of such
well as detailing some of the cultural politics that factors involves societies that have persistent food
gave rise to this form of self-​abnegation. Body ideals insecurity, where eating disorders with food restric-
change, as do forms of expression of distress, includ- tion are rare (Holtkamp, Hebebrand, & Herpertz-​
ing eating disorders. While multiple forms of disor- Dahlmann, 2004). A society that has not achieved
dered eating have been found across many societies, subsistence food production does not typically

Anderson-Fye 189
evidence contemporary forms of eating disorders shaping the context of eating disorders rather than
(Bemporad, 1997; World Health Organization, determining them.
2005) for an understandable reason: If people reg- Others, meanwhile, have offered hypotheses
ularly do not have enough food to eat, they gener- regarding the impact of climate and related clothing
ally are unlikely to restrict it willingly. However, customs on eating disorders variation (Sloan, 2002;
two kinds of intracultural variation lead to excep- Torres-​McGehee, Monsma, Dompier, & Washburn,
tions to this pattern. The first involves eating 2012). Certainly it would seem logical to think that
disorders among the wealthiest members of a food-​ areas with warmer climates, where people wear fewer
insecure society; these have been documented in clothes, would see increased body consciousness,
places such as South Africa (Szabo, 1999). The body monitoring, and disordered eating compared
second centers on upwardly mobile young people to places with colder climates. Yet research does not
who want to “pass” as wealthier than they are in support such assumptions (cf. Miller & Pumariega,
situations where slender body shapes are associated 2001). Conversely, warm climates where full-​body
with wealth and/​or obesity with poverty (Brewis, covering is customary are not spared eating disor-
et  al. 2011; McClure, Poole, & Anderson-​Fye, ders (Trainer, 2017). Finally, colder climates show
2012). Even if individuals do not have enough significant variation in the extent of eating disor-
healthful food to eat on a consistent basis, they ders across populations, but they remain a persistent
still may restrict their intake to achieve a thin problem in places like Canada and Scandinavia.
body akin to those of more affluent members of These data again allow us to conclude that while
a society (see “Symbolic Body Capital” section). weather and clothing customs can shape the par-
This behavior is more likely to be seen in places ticular instantiations of eating disorders, no loca-
with clear wealth disparities and only a portion tion is entirely immune. Further, factors beyond
of the population experiencing food insecurity. climate are more important to explain patterns of
Food-​insecure eating disorders related to upward disordered eating.
mobility also are more likely where obesity is asso- One of the major structural factors related to
ciated with poverty, such as in the United States body image and eating disorders around the world
(Brewis, et  al. 2011). This association of poverty is gender (discussed further later). Women over-
with obesity (as opposed to thinness) is becoming all are at risk for poorer health than men in many
increasingly common globally, including within countries related to structural inequalities (Farmer
the developing world (e.g., Hruschka, 2017). & Connors, 1996), and eating disorders are no
Large-​scale factors also can influence the over- exception. Everywhere gender has been examined,
all prevalence of eating disorders, but again, schol- it is related to body image, eating, and eating disor-
ars are wise to refrain from broad generalizations. ders, albeit not in identical ways across cultures. The
A few decades ago, for example, the prevailing per- majority of contemporary eating disorders are found
ception was that eating disorders were more com- among females (Hsu, 1989; Striegel-​Moore et  al.,
mon in urban and suburban regions than rural ones 2009), although evidence indicates a recent and
(Azuma & Henmi, 1982; Nadaoka, et  al., 1996). substantial rise of disordered eating and body dys-
Over time, however, studies of rural areas in such morphia among males around the world (cf. Pope,
countries as the United States (Miller, Verhegge et al., 2000; Monaghan & Atkinson, 2014; Phillips,
Miller, & Pumariega, 1999), Mexico (Bojorquez & 1996). For decades, scholars have attempted to
Unikel, 2004), and Italy (Preti et al., 2007) reported understand fully the cultural aspects of gender that
rates of disordered eating just as high as those found put girls and women at precariously high risk for
in more densely populated and developed areas. body image dissatisfaction and disordered eating in
Research in other nations found even higher rates many places in the world (cf. Becker, 2004; Bruch,
of disordered eating in rural areas (e.g., Lee et al., 1973, 1978; Casper 2017; Steiner-​Adair, 1986). At
1990). These complicated findings ultimately illus- the extreme end of these investigations, some cul-
trate that no level of urbanization is free from eat- tural studies scholars have argued that anorexia is in
ing disorder risk. As Preti et al. (2007) wrote, “it is fact a logical outcome of being female in contempo-
evident that the factors influencing the distribution rary Western cultures. That is, the ideal woman as
of eating disorder symptoms and their psychological constructed by media and paradoxical cultural roles
correlates by place of residence are far more com- should be so slight that she should cease to exist
plex than currently thought.” As a result, urban- (Bordo, 1993). In the section “From Gender and
ization status should be thought of as potentially Race to Intersectionality” in what follows, gender

190 Cultural Influences on Body Image

is reexamined as a major factor in relation to oth- early childhood behaviors related to eating, toilet-
ers at the demographic level, such as race, class, and ing, and physical activity are structured differently
sexuality. across cultures, thereby imbuing habits and values
Disease context is another environmental fac- (Tobin, Wu, & Davidson, 1989). For example,
tor affected by political economy that can shape while children in a typical Chinese preschool all
body image, ideals, and disordered eating. Evidence line up to visit the bathroom at the same time,
shows that past epidemics such as tuberculosis (also those in an analogous US preschool learn to raise
known then as “consumption”) contributed to their hands on an as-​needed individual basis (Tobin
new ideals for beauty (Byrne, 2011; Sontag, 1978). et al., 1989). In this case, the Chinese children are
Poets, musicians, and even physicians romanti- learning to orient toward communalism even in
cized tuberculosis as a disease that consumed deli- basic physiological needs, while the American chil-
cate, beautiful, and creative souls. The pale, slender dren are learning to pay attention to their individual
tubercular aesthetic persisted for a century or more bodies and expression of personal needs. Over time,
in Western cultures, especially among middle-​class systematic behavioral choices become part of social-
white women. ized cultural repertoires that then feel like habit or
In contrast, the emergence of HIV (human “common sense,” and are central to both body image
immunodeficiency virus) in more recent decades and eating behaviors throughout the life course.
led to a highly stigmatized condition that beauty A striking example of communal socialization
ideals opposed. In Cape Town, South Africa, for of food and body that affects later experiences and
example, one in three young women is HIV posi- behavior is found in Becker’s work in Fiji (Becker,
tive (Shisana et al., 2005); beauty ideals there have 2004, 2017). Becker, both a psychiatrist and an
to navigate carefully against signaling this disease. anthropologist, found that in rural Fiji, bodies were
Upwardly mobile black women will endorse slen- social objects, shared by close others. It was the job
der beauty ideals, but not those so thin that their of close kin and friends to feed someone, rather
appearance might imply that they are HIV positive than for that person to pursue an individual task.
(Latner, Stunkard, & Wilson, 2005; Matoti-​Mvalo Meals were communal, and involved feeding others
& Puoane, 2011). In this context, weightlifting and relatively large amounts, a behavior that indicated
lithe muscularity developed as an ideal, but then care. Such practices and rituals around meals set up
another bind materialized. Antiretroviral therapies the construction of a body as a shared object. Rather
create a characteristic pattern of muscularity; as a than one body equaling one person, the responsibil-
result, young upwardly mobile women in this con- ity for one healthy body spanned significant oth-
text intentionally develop other patterns of mus- ers (e.g., Anderson-​Fye, 2011; Becker, 2004). This
culature lest they be assumed to be taking these communal orientation toward bodies and body
medications. Such narrow bands of acceptable body image shaped young peoples’ experiences of body
ideals create a context that can give rise to disordered image and eating behavior over time. In this case,
eating among young women (Lee, 2013, Spettigue indigenous purgatives gave young women a cultur-
& Henderson, 2004). In short, even large-​scale fac- ally understandable way to span both traditional
tors change over time, creating a constantly shifting values and new transnational goals of svelte bodies
demographic terrain for eating disorder risk. and aspirational lives (Becker, 2017).
Shared behaviors can be meaningful at subcul-
Practices, Behaviors, and Habits tural levels that affect both body image and eating
The elements of culture that entail behavior concerns. A subculture is a smaller salient group than
include group-​ level routines, rituals, and habits an entire society that shares certain practices and
as well as individual participation and innovation. semiotic systems. Some of the subcultural groups
Cultural feeding and eating practices from infancy that have been shown to be important to body image
are known to shape later eating behaviors and out- and disordered eating due to behavioral patterning
comes. For example, infant feeding “on demand” include religion, sports, jobs, and college cultures,
versus “on schedule” remains hotly contested in among others. None of these subcultural groups is
Western societies because of both their biological without meaning; however, these routine, everyday
and behavioral correlates longitudinally (Iacovou practices mean that individuals experience their lives
& Sevilla, 2013). Children learn to eat in differ- in various ways with important outcomes over time.
ent culturally patterned ways around the world, Religious rituals and beliefs also have been
whether they use a fork or chopsticks. Similarly, long known to affect body image development

Anderson-Fye 191
and disordered eating. For example, religions such of eating disorders; in one study, 45.2% of female
as Catholicism and Judaism include fasting as a college swimmers cited the team suit as a weight
key role in ritual behavior; these practices have stressor, the highest for any factor (Reel & Gill,
been affiliated with disordered eating (Boyatzis & 2001). In each sport subculture, the regular pattern-
Quinlan, 2008). Women participating in religious ing of behavior over days, weeks, and years plays an
ritual where the abnegation of food is loaded as important role in body image and eating outcomes.
morally “good” are at higher risk for disordered eat- Similar to sport subcultures, job subcultures
ing (Marsden, Karagianni, & Morgan, 2007). can also create behaviors that impact body image
Sports subcultures have proven to have both and eating disorder risk. Scholars have found that
positive and detrimental effects on body image and ornamentally focused jobs like modeling experi-
eating concerns. Specifically, instrumental sports—​ ence heightened rates of body checking and disor-
where accomplishment is based on performance dered eating behavior (Preti, Usai, Miotto, Petretto,
regardless of body size—​ appear to be protective Masala, 2008, Treasure, Wack, & Roberts, 2008).
for youth, and girls in particular (Sundgot-​Borgen Similarly, research involving flight attendants con-
& Torstveit, 2004, Zucker, Womble, Williamson, sistently demonstrates that body size matters a great
& Perrin, 1999). These types of sports emphasize deal in that career (Roehling, 2002), with negative
movement and task accomplishment, which may implications for their sense of body image. While
be achieved across a range of sizes and shapes. some airlines have eased their regulations regard-
Instrumental sports include soccer, basketball, and ing the weights of flight attendants, some have not;
hockey, among others. Research has consistently this history and uneven application of enhanced
found these types of activities to be associated with flexibility creates expectations of thinness that has
healthy body image development for young women been associated with body image disturbance (Tyler
(Gutgesell, Moreau, & Thompson, 2003; Reinking & Abbott, 1998). Such pressures are not limited to
& Alexander, 2005). While less research aimed at particular careers; stigma against obesity also has
questions regarding body image and eating behavior been found to play a role in employment markets
has been conducted with young men in instrumen- around the world, also with body image conse-
tal sports, they also appear to benefit (Baum, 2006; quences (Anderson-​Fye & Brewis, 2017; Pagan &
Wilkins, Boland, & Albinson, 1991). Davila, 1997; see “Symbolic Body Capital” section).
Ornamental sports—​where body shape and size Another common subculture that patterns daily
are importance to success—​are far more perilous behavior is the college campus. When young people
for females and males (Anderson & Petrie, 2012; attend college, and especially a residential college, their
Beals & Manore, 2004). These types of sports entail daily routines and eating behaviors change (Racette,
aspects of high-​level performance, but also empha- Deusinger, Strube, Highstein, & Deusinger, 2005;
size body appearance and weight. Specifically, Rudd & Lennon, 2000). The enduring prevalence
some pursuits—​ for example, dance, gymnastics, of the phrase “freshman 15”—​referencing pounds of
wrestling, jockeying, and ice skating—​ often can weight gain among first-​year students—​signals the
require participants to experience close monitor- importance placed on appearance, particularly for
ing and measurement (Byrne & McLean, 2002; women, in setting where a wide array of foods is read-
Currie, 2010). Bodies may be monitored for rea- ily available. Mealtimes often are important sources
sons of aesthetics or the weight classes established of sociality, where peer behavior becomes increasingly
for competition in such sports. Estimates of eating influential. A particularly pernicious example of resi-
disorders and risks are as high as 42% for female dential living environments contributing to negative
athletes in aesthetic sports such as dance or gym- eating outcomes involves sororities or other groups
nastics (Greenleaf, Petrie, Carter, & Reel, 2009; where young women participate in eating disordered
Sundgot-​Borgen, 1994 and 11% for males in sports behavior together; examples include forced vomiting
with distinct weight classes like rowing or wrestling after meals (Boskind-​Lodahl & White, 1978) or tak-
(Baum, 2006). Some sports, like swimming, have ing diuretics, laxatives, or other medications for the
important elements of both categories. While swim- purpose of reducing weight (Mintz & Betz, 1988).
ming focuses on performance across body types, College campuses are important sites of behavioral
some young women can feel self-​conscious wear- change with respect to eating behaviors and physi-
ing bathing suits on a regular basis. This emotional cal activity.
reaction in turn can impact body image negatively, At the extreme in the behavioral category are
and in some cases contribute to the development theories of disordered eating that posit anorexia

192 Cultural Influences on Body Image

nervosa as an “activity disorder” (O’Connor & toward their bodies more negatively (Cusumano &
Esterik, 2015). In this view, contemporary cul- Thompson, 1997; Durkin & Paxton, 2002). Similar
tures offer the opportunity for ascetic behaviors studies have been performed with Internet expo-
that individuals pursue to degrees that result in sure, in which young people browse the Web in
eating disorders. Rather than drawing on physi- between pre-​and post-​tests (Fardouly, Diedrichs,
ological feedback loops well known to eating dis- Vartanian, & Halliwell, 2015).
orders, advocates for this thinking argue that the Given these studies showing the insidious effects
behaviors alone create the problem. These authors of thin media ideals, some countries recently have
posit explicitly that the disorder has no meaning legislated minimum body mass index (BMI) criteria
and no prior psychopathology or history. Instead, for models to try to limit these images’ negative cul-
they argue that the disorders are solely a function tural effects on girls and women in particular. For
of participating in culturally normative behaviors example, in Israel and France, underweight mod-
such as undereating and overexercising. While such els are banned from print and catwalk modeling
a view is provocative, it has been soundly criticized (Reynolds, 2013; Stampler, 2015). Driven by medi-
by clinicians and sociocultural researchers alike cal professionals who have eating disordered clients
(e.g., Lester, 2015). Behavioral components are who strive for model-​like ideals, these initiatives are
better understood as one piece of the complex eat- one of many prongs aimed to curtail cultural contri-
ing disorders puzzle. butions to ill health. These innovative initiatives are
some of the first to try to control ideals and images
Ideas, Ideals, and Images toward body image health. At the least they send a
People’s beliefs about their bodies have long strong message to counter long-​prevailing trends of
been found to be important in their actions regard- increasing slenderness as a feminine ideal.
ing their eating, exercise, and other related behav- Unfortunately, these new efforts to encourage
iors. Often, these behaviors are directed toward healthier body ideals have been accompanied by the
becoming more like an ideal image that is drawn introduction of technologies that allow the emer-
from media and local ideals. Sometimes, behaviors gence of Internet-​based subcultures that have per-
are also intended to steer an individual away from nicious sociocultural effects on individuals. A stark
stigmatizing images. Ideas, ideals, and images are example is pro–​eating disorder websites. On these
discrete factors in body image and eating disorder sites, members share “best practices” with regard
development. More, the way that these variables to behaviors for self-​starvation and other modes of
interact can also vary by cultural context (Anderson-​ self-​harm (Borzekowski, Schenk, Wilson, & Peebles,
Fye & Brewis, 2017). 2010; Rouleau & von Ranson, 2011). Research has
Media images of ideal female and male bodies shown that these sites give individuals a sense of
have been shown to have a profound impact on connection and support for participation in disor-
body image and eating behaviors (Borzekowski & dered and dangerous eating behavior. The rise of
Bayer, 2005; Derenne & Beresin, 2006). Television, such websites holds some irony, given that most eat-
print media, and Internet images are all impli- ing disorders involve secrecy and shame as hallmark
cated. Both in the areas of media production and characteristics (Hayaki, Friedman, & Brownell,
in areas near and far of media consumption, the 2002; Kelly & Carter, 2013). In addition, research-
svelte privileged ideals of favored media set difficult ers recently have documented a surge in image-​
standards that many compare themselves against based pro-​anorexia and “thinspiration” social media
(Cohn & Adler, 1992), feel badly about (Aubrey, posts (Ging & Garvey, 2017). On the other hand,
2006; Groesz, Levine, & Murnen, 2002), and act on the same technologies can be employed in support
(Harrison, 2000; Harrison & Cantor, 1997). Some of eating disorders recovery (Aardoom, Dingemans,
of the most interesting studies proving the negative Boogaard, & Van Furth, 2014, McNamara &
impact of magazine images on body image have Parsons, 2016) and movements such as “health at
been done in psychology laboratories. In these stud- any size” (Brown, 2009). The Internet and related
ies, young women and men are given a pretest on media have opened new avenues for consumption
body image satisfaction. Then they are given con- of images, ideas, and ideals, although the valences
temporary fashion and popular magazines to read. of impact are many and complex (cf. Anderson-​Fye,
Finally, they are post-​tested again on body image 2003; Ging & Garvey, 2017).
satisfaction. After short exposure to these images, Studies have shown convincingly that images
participants evaluate their own bodies and feelings of ideal bodies can impact body image negatively

Anderson-Fye 193
across many types of cultural contexts internation- and appearance (Shugart, 2008), however, other
ally, but they do not always do so, nor are the effects studies posit that men of all sexual orientations
uniform. Cross-​culturally, it was initially thought and locales are experiencing an uptick in pressures
that Western-​produced images of ideal female bod- based on appearance in work and dating (Frith &
ies seen in the media were absorbed uncritically by Gleeson, 2004). Lesbian women were once thought
other women, particularly those in poorer and devel- to be more resistant to body image dissatisfaction
oping nations (Craig, Swinburn, Matenga-​Smith, and eating disorders (Bergeron & Senn, 1998),
Matangi, & Vaughn, 1996). However, research has however other studies show them to be at equal or
shown that local ethnopsychologies have a substan- even increased risk compared with straight women
tial effect in terms of differential impact. Notably, (Morrison, Morrison, & Sager, 2004). A  recent
young women may filter the visual ideals presented review of the Youth Risk Health Behavior Studies
in the media based on their own models of beauty showed that sexual minority students suffer from
and success; the choices and attitudes of young increased rates of disordered eating, a troubling
women in Belize in the late 1990s provide a compel- trend worthy of further study (Kann, 2016).
ling example of this differentiation (cf. Anderson-​ In any culture, body ideals are upheld, created,
Fye, 2004; Anderson-​Fye & Lin, 2009). That is, if and recreated not solely through imagery but rather
a young woman (and her cultural milieu) disagrees through multiple means. For example, Taylor’s
with a media ideal, that ideal will be less salient to recent work (2016) details how body ideals are cre-
her. However, if she endorses and aspires to that ated and monitored through verbal means in a US
ideal, it may become paramount. More, increas- high school context. By examining what both male
ing evidence exists that Western-​produced images and female students do and do not say—​as well as
may interact with local ideals to create hybrid ideals how they say it—​Taylor explicates the daily micro-
(Kim, 2011, Poorani, 2012). cultural scaffolding that both sustains and polices
Subculturally in diverse Western nations, data ideals among diverse youth (Taylor, 2016). For
are mixed on the effects of media ideals for those example, Taylor found that an individual targeted
from nondominant groups. Some studies indicate or being too fat could remove the spotlight from
that mainstream media images of slender women herself by mentioning someone who was heavier—​
are less important to African American and Latina who in turn become the focus of discussion (Taylor,
women (Rubin, Fitts, & Becker, 2003; Schooler, 2017). By discussing ideals such as “six-​pack abs”
2008). More recent work indicates that instead, and stigmatized conditions such as “too fat,” the
young African American and Latina women draw students monitored themselves and others in a
on a more diverse set of images showing multiple tightly controlled range of body ideals.
body ideals (McClure, 2017; Taylor, 2016). Internal An important distinction in the realm of ideas,
variation within groups such as African American ideals, and images that Taylor and others increas-
and Latina women is of central importance to ingly examine is the distinction between ideals and
understanding their experiences of risk and out- anti-​ideals or stigmatized conditions. For females,
comes with respect to eating (see “From Gender and this tension is often seen in aspirations toward a thin
Race to Intersectionality” section). ideal and away from being “too fat” (e.g., Anderson-​
While pressures regarding appearance among Fye et  al., 2017). For males, the tension is more
straight men are thought to be increasing (Pope, likely to manifest as a pull toward muscularity while
Phillips, Olivardia, 2000), gay male subcultures also trying to avoid fatness (McCreary, 2012; Pope
are thought to have even more exacting standards et al., 2000). This notion of aspiring toward an ideal
for ideal images of attractiveness. Gay and bisexual versus avoiding a culturally stigmatized body shape
men have been found to suffer from body image may be more important to the field of body image
dissatisfaction at higher rates than straight men for than previously recognized and warrants additional
decades (Kaminski, Chapman, Haynes, & Own, research (Lester & Anderson-​Fye, 2017).
2005; Williamson & Hartley, 1998). More, they In particular, as rates of obesity skyrocket around
have had to respond with body ideals in opposition the world (NCD Risk Factor Collaboration, 2016),
to HIV-​positive status due to stereotyped stigma. they have contributed to profound concerns about
Recent studies have posited that straight men are global health. These fears have led some bioethicists
also experiencing increased body image pressure. and public health advocates to actually promote
Terms like “metrosexual” have emerged to describe a strategy of shaming fat individuals (Callahan,
straight men who invest resources in grooming 2013). In a set of studies from several fields, obesity

194 Cultural Influences on Body Image

stigma has been shown to be dangerous to individu- non-​ Western nations struggle with the tension
als in terms of body image and eating behavior both between young people desiring clothing meeting
in Western nations (e.g., Casper, 2017; Greenhalgh, transnational Hollywood or Bollywood standards,
2012, 2015; Puhl & Brownell, 2001) and around and adults’ preferences for more traditional choices
the world (Anderson-​Fye & Brewis, 2017; Brewis, (Becker, 2004; Trainer, 2017). Whatever the local
Wutich, Falletta-​Cowden, & Rodriguez-​Soto, situation, available clothing choices provide a site
2011). Some have indicated that these negative for body consciousness for young people as they
ideas toward fat may be at least as harmful as obe- operate within familial, cultural, and transnational
sity itself (Brewis, 2014; Greenhalgh, 2012; Puhl & ideals (Taylor, 2016).
Heuer, 2010). Ideas and images regarding ideal bodies come
A subtle but powerful image-​based subcultural through many sociocultural modalities and through
influence that has been shown to impact body ide- multiple routes. Media images from television,
als is marketing of children’s toys and clothing. magazines, and newspapers have long been known
Toys like Barbie dolls, with their impossible-​ to-​ to affect individuals’ body image satisfaction and
achieve physiques, have been shown to negatively eating behaviors (Becker, Burwell, Gilman, Herzog,
affect girls’ body image (Brownell & Napolitano, & Hamburg, 2002). Such depictions achieve
1995). The subtle but powerful messages that these salience through direct consumption as well as
figures communicate to young people have been indirect means such as explicit and implicit verbal
shown quantitatively (Dittmar, Halliwell, & Ive, monitoring by peers and adults (Nichter, 2001).
2006) and qualitatively (Nichter & Nichter, 1991). The Internet and cellular phones have made access
In more recent years dolls that represent more typi- to images even more constant. More, ubiquitous
cal female proportions have been introduced, albeit social media offer additional means by which peo-
with varying degrees of success. While a doll alone is ple’s bodies can be compared and judged (Fardouly
unlikely to introduce eating pathology, it is part of et  al., 2015). These technologies have increased
a cultural milieu that sends unhealthy messages to transnational connections to the extent that those in
young women. Popular media coverage of contro- even the most remote parts of the world have access.
versy about the dolls appears to have raised aware- While the full effects of our new routes of connect-
ness of the issue of subtle messages to young people edness are not yet fully understood, they are worthy
about ideal shapes. An analog for boys includes of continuing study—​including whether sociocul-
action figures that have gotten increasingly muscu- tural interventions such as limiting the lower end of
lar over time in the United States (Pope, Olivardia, model BMI and reducing computer manipulation
Gruber, & Borowiecki, 1999). Even young boys’ of models’ bodies in visual media can be effective.
Halloween costumes are now padded with fake
muscles (Alexander, 2014), thus sending them mes- Symbolic Body Capital
sages about a bulked-​up male ideal. These exemplars An important concept in understanding how
parallel the shaping of bodies and features of ideal culture impacts body image, eating behavior, and
cartoon characters on children’s television (Klein & use of other body-​shaping technologies is symbolic
Shiffman, 2006). For both girls and boys, the mes- body capital. Symbolic body capital allows under-
sages provided through toys, heroes, and television standing of how the meanings of physical attributes
shows may be subtle and indirect. Nevertheless, they of human appearance are leveraged within specific
continue to persist and contribute to body image cultural contexts to achieve gain in economic or
dissatisfaction (Knauss, Paxton, & Alsaker, 2007). dating markets (Anderson-​ Fye & Brewis, 2017).
Available and affordable children’s clothing is Symbolic body capital appears to have explicit
shaped culturally, and often consistent with popu- and implicit components that impact body-​related
lar media ideals. Girls face increasingly sexualized ideals and behaviors (Brewis, 2017; Lester &
clothing choices at younger ages in the United States Anderson-​Fye, 2017).
(Graff, Murnen, & Smolak, 2012). American boys Across contexts, for example, people can
also face limited choices with emphasis on options describe and identify types of bodies that are more
that provide bulk and coverage (Taylor, 2016). likely than others to attain successful and desirable
Some nations, such as Sweden, have made deliberate employment (Brewis, 2017; Anderson-​Fye et  al.,
strides to push comfortable, durable unisex cloth- 2017). In the United States, social science research
ing, albeit with limited efficacy (Friesen, Holmqvist, is well aware of the bias against overweight work-
& Anderson-​ Fye, 2013). Many developing and ers in hiring and promotion (Rothblum, Brand,

Anderson-Fye 195
Miller, Oetjen, 1990; Fikkan & Rothblum, 2005). the local environment, it was rare in the global data
Thin and fit individuals are likely to fare better, (Anderson-​Fye & Brewis, 2017). It appears univer-
and the effects of weight bias in the workplace are sal that body shapes, sizes, and aspects carry mean-
more pronounced for females. In tourism-​focused ing that matters centrally to processes of upward
areas in the developing world, young women who mobility in terms of employment and dating or
fit transnational ideals of thinness are more likely mating. It is equally widespread that local context
to he hired for service-​sector jobs than are women sets the stage for the meaning imbued, and that
who fit curvier local ideals (Anderson-​Fye, 2004, meaning also can change over historical time peri-
Anderson-​Fye et  al., 2017; Becker, 2017). Such a ods (Lester & Anderson-​Fye, 2017). This meaning
situation can change shared cultural ideals over time is part of how culture shapes actual bodies and eat-
or introduce new ones (Anderson-​Fye et al., 2017; ing behaviors.
Becker, 2017; Katzman, Hermans, Van Hoeken, &
Hoek, 2004). Shifts in How Cultural Factors in Body
What may be symbolic regarding body capital Image and Eating Disorders Are
may be overall body shape or size or specific fea- Understood
tures. For example, some have posited that a muscu- Historically, review articles regarding culture and
lar body may indicate moral character of discipline body image/​eating disorders discuss classical struc-
and hard work, a desirable trait in a worker (Pope tural factors including gender, race, class, religion,
et al., 2000; Thompson & Cafri, 2007). However, and region. However, in more recent years social
in other contexts, the same muscularity could indi- science research has progressed beyond these cat-
cate self-​absorption and time spent on recreational egories to focus on complex and shifting dynamics
pursuits instead of industrious ones (Anderson-​Fye reflective of contemporary society. This section con-
et  al., 2017). Thus, the operation of the capital centrates on three major areas important to body
related to specific traits is embedded in local mean- image and eating disorders, including shifts from
ing systems. As indicated in this example, the pro- (1)  gender and race to intersectionality; (2)  class
cess accomplishing the desired physical attributes to upward mobility; and (3) cross-​cultural data to
may be implicated as much as the end state itself transnational systems.
(Edmonds & Mears, 2017).
In addition to employment markets, symbolic From Gender and Race to Intersectionality
body capital is active in dating and mating choices Scholarship on body image and eating disorders
(Anderson-​Fye & Brewis, 2017). Again, when has long been subjected to gender-​based analysis.
examined through the lenses of social science, Similarly, whether explicit or implicit, race has
research participants can identify and describe the also been of central importance. In earlier work,
types of bodies and traits that are more likely to be gender and race were seen as determinative; that
attractive to potential heterosexual mates in most is, white women of middle class and above status
locales. Interestingly, what each gender endorses were seen at primary risk for eating disorders such
as the other’s ideal in a heterosexual context is not as anorexia and bulimia nervosa. While it remains
always aligned. For example, while a Belizean male debatable whether these statistics were once true,
may believe his potbelly makes him unattractive to it is clear now that females and males from many
potential dates, a female may view him as more sexu- backgrounds suffer from body image and eating dis-
ally faithful and hardworking at his job than a fitter orders (Smolak & Levine, 2010).
man, and therefore highly desirable (Anderson-​Fye A theoretical innovation in understanding which
et al., 2017). The local contexts of meaning of body women and men suffer from these disorders is the
shape, size, and features again emerge as paramount concept of intersectionality. While the theory of
(Becker, 2004, Edmonds & Mears, 2017). intersectionality first emerged more than a quarter
There are exceptions to the widespread impor- century ago (Crenshaw, 1989), only recently have
tance of symbolic body capital for romantic rela- researchers recognized the power of this approach
tionships. In Nepal, upwardly mobile students to illuminate experiences, causes, and potential
could not understand questions aimed at pairing solutions for the inequalities individuals face with
body shape and size with dating. The students regard to healthcare and results. Some of the delay
explained that partnerships were based on charac- in its full use can be found in the concept itself.
terological traits, not on bodily ones (Anderson-​Fye That is, the impact of various aspects of identity in
et  al., 2017). While this exception was robust for individual and social contexts cannot be considered

196 Cultural Influences on Body Image

as simply additive—​for example, the effects of being in lower wealth categories in the United States
African American and female extend beyond a cal- and therefore have less access to health insurance
culation that simply joins together racial and gen- than Whites (Smedley et  al., 2003). However,
der discrimination. Rather, this theory posits that great intragroup variation exists among African
these qualities will interact with one another, and Americans in terms of wealth, healthcare use, gen-
in turn with factors ranging from geography and der identity, school culture, familial dynamics, and
education to economic class and age, in their effects many other factors shown to be important to both
on individuals and groups (Crenshaw, 1989)—​ body image and eating disorders. Our prior modes
including with regard to health outcomes (Schulz of analyzing “Body Image and Eating Disorders
& Mullings, 2006). Among Blacks” deemphasize these differences even
Such intersections are critically important when though they may, in fact, have powerful impacts on
it comes to understanding body image and eat- these individuals. Rather than a simplified analysis
ing disorders (McClure, 2017). While females still that, in effect, collapses and racializes social and cul-
account for 90% or more of eating disorder diagno- tural influences, these various important categories
ses, that percentage has been decreasing over time should be considered in tandem.
as more males struggle with body image and eat- An excellent recent example of intersectionality
ing. For some disorders, such as body dysmorphic in the case of African American adolescent female
disorder, the gender balance is about even (Phillips, body image comes from the work of McClure, who
2012). A consistent clinical problem when percent- conducted a year of mixed-​methods research in a
ages of one binary group far outweigh its counter- Rust Belt public high school known for strong aca-
part is underdiagnosis within the statistical minority demics and racial diversity (McClure 2013, 2017).
group. Such diagnostic issues can have serious clini- This work showed how intersections of race, class,
cal consequences for more advanced disease. The gender, neighborhood, family history, and school
aforementioned cultural issues in structural inequal- culture impacted these young women’s sense of
ities; behavioral norms; and ideas, ideals, and images body image and physicality. More, McClure did not
all are implicated in higher rates of body dissatisfac- rely on measures and categories normed on White
tion and eating issues for women. Cultural assump- middle-​class girls, but rather allowed for categori-
tions on the part of healthcare professionals can also cal representations to emerge systematically from
affect diagnosis and treatment (Gordon, Perez, & the young women themselves. She found a diverse
Joiner, 2002). Thus, white males and black females typology embedded in historical trends and imbued
with anorexia may be diagnosed later and have even with cultural and individual differences that were
higher mortality rates than white females. Given exemplified through popular culture representa-
that anorexia nervosa is already the most deadly tions (McClure 2013, 2017). This scholarship is
psychiatric disease (Arcelus, Mitchell, Wales, & an example of advancement in biocultural theory
Nielsen, 2011), this further increase in fatalities—​at and empirical research that moves the field beyond
least partially due to cultural stereotypes—​is deeply anachronistic thinking toward meaningful innova-
concerning. tion in consonance with contemporary theory.
In the case of African American females, many Another sophisticated example of research
other factors complicate the situation. For example, related to body image, eating concerns, and inter-
socioeconomic status may impact willingness to sectionality is Taylor’s ethnography in a diverse
approach biomedical healthcare practitioners in a Southwestern public high school (Taylor, 2016,
timely way. African Americans have been found to 2017). Through sociolinguistic methods, Taylor
trust biomedicine less than whites (Jacobs, Rolle, examines what is said—​and how it is said—​and
Ferrans, Whitaker, & Warnecke, 2006) and to access what is left unsaid in terms of monitoring young
clinical care at lower rates (LeCook, McGuire, people’s body shapes, sizes, ideals, and stigmatized
& Zaslavsky, 2012). Such mistrust comes from anti-​
ideals. Taylor details a heretofore hypoth-
a long history of mistreatment of Black people in esized moment-​ to-​
moment adjusting of micro-
the United States by biomedical research and prac- cultural situations where young people are held
tice (Jacobs et al, 2006; Smedley, Stith, & Nelson, up to media standards as well as ever-​shifting peer
2003). Higher education allows continued critical comparison. In addition, she shows how gender
engagement with healthcare institutions, while also differences develop further in microinteractions
providing increased social capital to seek out trust- of mixed-​gender peer groups. Her work elucidates
worthy doctors. Also, Blacks are overrepresented gender, class, race, ethnicity, sexuality, school, and

Anderson-Fye 197
local contexts as ever-​shifting close contexts for Analysis of socioeconomic class per se has per-
the construction of body image and eating behav- sisted for decades (McClelland & Crisp, 2001). It
iors. Studies like these center the ways our salient is difficult to know whether these earlier findings
cultural categories intersect in to create contin- indicated a true absence of disordered eating among
gent outcomes. In so doing, they offer impor- lower classes of women, or whether they represented
tant evidence for the value of intersectionality to flawed sampling methods and clinical identifica-
advancing understanding of culture, body image, tions. Meanwhile, other studies started to reflect
and eating disorders. disordered eating attitudes and behaviors across
While psychologists have called for increased class groups. By 1996, Gard and Freeman called the
attention in intersectionality, such investiga- assumed relationship between socioeconomic status
tions still lag behind in body image and eating and eating disorders “a myth.” This work proceeded
disorders research. In an influential article in to document the socioeconomic diversity underem-
American Psychologist, Cole (2009) proposed that phasized in the eating disorders literature and impli-
for psychological research topics where gender, cated stereotypes, bias, and conflation of anorexia
race, ethnicity, social class, and sexuality make and bulimia nervosa for the inaccurate representa-
a difference, three questions should be asked at tion (Gard & Freeman, 1996). However, they main-
each stage of the research process. First, “Who tained that socioeconomic status remained relevant;
is included within this category?”; next, “What it just was not in itself the deciding factor.
role does inequality play?”; and finally, “Where Around the turn of the millennium, scholars
are there similarities?” (Cole, 2009, p. 170). Such began to posit that in fact, upward mobility may
questions would make implicit assumptions about be a more important risk factor than class per se
categories explicit and show the variable relation- (cf. Anderson-​Fye & Becker, 2003). The rapid rise
ships among salient social factors. This type of of disordered eating around the world and among
methodology further unifies the social sciences immigrant youth provided additional support
of psychology, anthropology, and sociology—​and for the hypothesis that the process of striving for
also draws on historical analysis. In addition, it more—​rather than class background—​was impli-
also operationalizes more refined conceptual tools cated in behaviors consistent with eating disorders,
as compared to the legacy of monolithic group particularly anorexia nervosa (cf. Becker, 2004;
analysis. While many popular media outlets and Katzman et  al., 2004; Nasser & Katzman, 1999).
blogs have called for intersectional analysis spe- For example, in a discussion of immigrant body
cifically with respect to body image disordered image in the United Kingdom, Nasser and Katzman
eating, the empirical literature continues to trail (1999) posited the importance of achieving the new
those exhortations. country’s thin aspirational body ideal and leaving
behind curvier shapes of countries of origin in the
From Socioeconomic Status to Upward process of acculturation.
Mobility Similarly, young women from lower classes in
Not unlike gender and race, class was thought to the United States have reported aspiring to thin-
be determinative in early work on eating disorders ner body sizes and employing disordered eating
(Bruch, 1978; Brumberg, 1988). That is, females and exercise in pursuit of “passing” as coming from
who were middle class and above were susceptible to wealthier backgrounds (Thompson, 1992). More,
the “golden cage” (Bruch, 1978) of eating disorders. the competitive processes of advanced graduate
These disorders represented middle-​ class White education and exclusive work environments have
women’s struggles with gender roles during second-​ been shown to support behaviors of disordered eat-
wave feminism (Orbach, 1980). Specific problems ing (Szweda & Thorne, 2002; Edmonds & Mears,
described in the literature included dysfunctional 2017). Reviews of the empirical literature on socio-
gendered family dynamics with absent fathers and economic class, body image, and eating disorders
enmeshed mothers (Bruch, 1978), as well as a drive have begun to exhibit consonance on the important
for perfectionism as educational and employment role of upward mobility (e.g., Anderson-​Fye, 2008;
opportunities were expanding for women (Steiner-​ Bojorquez & Unikel, 2004; McLaren & Kuh, 2004;
Adair, 1986). White working-​class and poor women Smolak & Levine, 2010).
were omitted from these analyses and thought to be Recently, fat stigmatization has also been linked
immune from the risk, along with women of color to upward mobility (Anderson-​ Fye et  al., 2017;
from any class background. Brewis, 2017; Puhl & Heuer, 2010; Sobal, 2017).

198 Cultural Influences on Body Image

This relationship is theorized to operate in both that the World Health Organization (WHO)
directions. That is, increasingly around the world, determined that eating disorders would consti-
thinner bodies carry higher prestige (Hrushka, tute “priority disorder” designation in its report
2017). Therefore those with obese bodies are stig- on children and adolescents with mental disorders
matized from attaining higher levels of status in (Saraceno, 2003). This inclusion was based on cri-
employment and educational settings (e.g., Puhl & teria of prevalence, severity, and the possibility that
Heuer, 2010; Sobal, 2017). Conversely, those who effective primary healthcare could make a signifi-
actively aspire to attain higher status in a society cant positive improvement for those suffering with
may stigmatize obesity more heavily than others the disorders.
(Anderson-​Fye & Brewis, 2017). In a multicountry Since that time, many reviews have chronicled
comparative study of young adults, Anderson-​Fye increasing rates of eating disorders in non-​Western
et al. (2017) found that college students endorsed and developing countries (e.g., Anderson-​Fye, 2008,
higher levels of stigma against obese bodies than did 2009; Anderson-​Fye & Becker, 2003; Edmonds,
non-​ college-​
bound students of matching gender 2012; Smolak & Levine 2010). More, they have
and age. Among young women in rural Fiji, Becker established region-​by-​region lists of the status of
found a positive relationship between upwardly eating disorders in each continent or subcontinen-
mobile career aspirations and both disordered eat- tal region. Thus we have multiple characterizations
ing and obesity stigma (Becker, 2017). Upward of the state of knowledge about disordered eating
mobility appears to play an important role in sym- in eastern Europe, sub-​Saharan Africa, the Middle
bolic body capital, pursuit of ideal bodies, and dis- East and North Africa, South Asia, East Asia, Latin
dain for fat bodies. America, the Caribbean, the Pacific Islands, and
A limitation of the work on upward mobility is so on. Such cataloging is useful, both to establish
the paucity of measures that are both reliable and the nature and extent of the disorders and also for
valid across contexts. While many scholars are pilot- researchers working in a specific area.
ing measures intended to hold up across both com- However, at this historical moment, more may
plex societies and developing nations (e.g., Hadley be gained by again turning to contemporary theory
& Weaver, in press), consensus does not yet exist on in the fields that are invested in these studies. Social
best choices. Multimethod measures that include scientific theories of globalization, transnational sys-
subjective and objective elements represent a prom- tems, and global health can further our knowledge
ising step toward addressing these shortcomings. of aspects of the disorders that may be biological
or social universals as well as meaningful variations.
From Cross-​Cultural Cataloging to Global Such knowledge aids work on intervention at large
Systems and Regional Trends structural, medial institutional, and individual clin-
Since culture has been implicated centrally ical levels (Becker, 2017). Current knowledge about
in body image and eating disorders from at least national and regional variation is substantial, even
midcentury scholarship, researchers studying though all cultures continue to experience change.
these issues have long been interested not only Understanding globalization, transnational net-
in historical data in Western cultures but also in works of related systems, and significant instances
cross-​cultural data (e.g., Hooper & Garner, 1986; of discrepant data sometimes can tell more about
Pumarino & Vivanco, 1987). Early reports showed important structural and underlying factors related
clear distinctions between countries in terms of to body image and eating disorders than will another
eating disorder prevalence. It is from these reports catalog of updated regional variation.
that notions began to emerge that eating disorders Two characterizations of contemporary global-
were confined to Western, complex, late industrial ization are particularly useful for thinking about
societies. At the same historical time, the world was body image and eating disorders. Suárez-​Orozco
undergoing significant and rapid globalization and & Suárez-​Orozco (2001, p.  347) describes global-
increased transnational connection (Appadurai, ization as “processes of change, generating at once
1996). As part of these trends, important concepts centrifugal (qua the borders of nation states) and
including individualized bodies, control over body centripetal (qua the post-​national) forces that result
size, and valuation of “work” on the body were also in the deterritorialization of important economic,
shared and embraced to differing degrees (Becker, social, and cultural practices from their traditional
1995; Lee, 2004). By 2003, eating disorders were moorings in the nation state.” This paradox—​where
widespread enough on the six inhabited continents local context is both more and less important—​is

Anderson-Fye 199
apt for current cross-​cultural trends. For example, In Nepal, gendered standards of attractive bod-
ideal body image trends toward fit, muscular males ies draw on Bollywood more than Hollywood,
and slender but athletic females are increasingly although like young women in Fiji, these standards
common (Anderson-​Fye & Brewis, 2017). So too are valued for employment mobility, rather than
has been the spread of conceptualizing bodies as dating (Anderson-​Fye et al., 2017). In urban con-
individuated and controllable entities, even in socio- texts there, men’s desired muscularity is extant, but
centric societies (Becker, 1995, 2017). However, the reflects a far leaner standard than those in the United
local instantiations of widespread global trends are States, Samoa, and others. Korean standards, mean-
extremely important to consider from both scien- while, are even more exacting. There, men endorse
tific and public policy perspectives. slender, angular bodies that may even require plas-
For example, the importance of nonfat muscular- tic surgery (Shin, 2011). Women also aspire to have
ity among young men in Samoa is striking. Samoa, slender, fit-​appearing bodies; in this case, however,
a society with one of the highest average body mass ideal shapes may require cosmetic surgeries such
indices (BMI) in the world, previously valued very as shaving jawbones into a “V”-​shape or removing
large male bodies (Brewis & McGarvey, 2000). the lowest rib (Stone, 2013). Lee and colleagues
The relatively rapid change to a value on slender (1993) documented the rise of eating disorders
muscularity offers insight regarding how quickly among Chinese people in Hong Kong, though
previously lesser exposed populations may adopt he discovered a non-​fat-​phobic variant that called
new ideals (Lipinski & Pope, 2002). It is impor- into question one of the essential diagnostic criteria
tant to recognize, though, that these Samoan men and hallmark characteristics of anorexia. Through
endorsed ideal fat and muscularity images that were understanding key exceptions, scholars can begin to
even more muscular than those US men chose in a identify which features of body image and eating
comparison study (Lipinski & Pope, 2002). While concerns are universal, and which are variable. Such
this study did not ethnographically investigate the differences are critical to understand for particularly
meaning of muscularity, the changed but shared pernicious problems such as eating disorders.
local value on hypermuscularity may be indicative On the other side of the world, Jamaican women,
of the continuing value placed on large bodies; a long thought to be “fat loving” (Sobo, 1993), have
desire to overcompensate for distance from global been reporting a substantial increase in anorexia
markets; or perhaps some other locally meaningful and bulimia nervosa (James, 2012). More, they too
hybrid. want bodies considered fit and not fat (Anderson-​
In nearby Fiji, Becker has been chronicling the Fye et al., 2017). However, the local model of femi-
changes in body conceptions, body ideals, and eat- nine ideal shape includes a voluptuous “booty” that
ing behaviors among young women for more than moves well whether walking or dancing. While
two decades. While transnational media and eco- Fijian, Nepali, Korean, and Jamaican women would
nomic participation through the travel and tour- all endorse a “slender and fit” ideal, the local instan-
ism industry have increased, so too have reports tiation of that ideal, as well as the means to possess
of dieting and disordered eating (Becker, 1995, it, are vitally important for any understanding of
2004, 2017). However, the means by which these health intervention with the potential to be locally
women control their weights—​ namely, herbal meaningful and effective. Thus, Suárez-​Orozco’s
purgatives—​are deeply meaningful locally (Becker, centrifugal and centripetal forces remain important
2017). Ethnic Fijians long have used these herbal to understanding n body image and eating behav-
purgatives for a host of maladies. By adopting iors globally.
this particular approach to manage weight, young Appadurai describes important routes of trans-
women have taken traditionally important means national connection (Appadurai, 1996). He terms
and applied them to globalized ends; that is, they them “scapes” including ethno-​, media-​, techno-​,
aspire for thinner bodies to support participation finance-​and ideoscapes. Each of these transnational
in global tourism jobs (such as becoming a flight “flows” is relevant to both body image and eating
attendant or resort service worker). Employment behaviors. Ethnoscapes include the categorization
in these jobs tends to favor those who meet stan- and identification of people with a particular group
dards of svelte global sophistication. Young women heritage. These associations may be local or dia-
weave together Fijian and transnational practices sporic; regardless, they have ideological and behav-
as they navigate growing into adults in a rapidly ioral meaning that can impact identity, habitual
changing landscape. practices, and eating outcomes. For example, Israeli

200 Cultural Influences on Body Image

Jews and Jewish Americans both have higher than gain access to a wider array of actual and semiotic
average rates of disordered eating, a finding that has resources that allow for the reduction of suffer-
been linked to religious and cultural practices and ing. More, cultural change in and of itself can be
beliefs as mentioned previously (Latzer, Witztum, a risk factor for disordered eating (Anderson-​Fye &
& Stein, 2008). For example, Judaism includes fast- Becker, 2003; Nasser, et al., 2001). All that said, total
ing as part of rituals, a practice that is associated quality of life may improve over time as a result of
with disordered eating for certain subgroups. While some of these choices. Satellite technologies and the
these two groups live far apart and have variegated Internet have also increased exponentially the global
daily contexts, their eating outcomes have striking exposure of media messages and cultural practices
and meaningful similarities. that impact body image and eating. First, through
The impact of global media has been men- global Internet access and then through individual
tioned previously. However, it is important to cellular phones, the human population—​even in
recognize that while much of the world may have remote and impoverished places—​has unparalleled
access to similar content (whether it is produced access to a central pool of ideas, images, and values.
in Hollywood or Bollywood), the local uptake of While the valence of such pooled information is
those media with respect to body image and eat- highly debatable, the effects are undeniable.
ing depends on salient ethnopsychology (Anderson-​ In terms of finance, the results of transnational
Fye, 2003), beliefs (Becker et  al., 2002), and flows of money cannot be overstated for effects
available material culture (Bojorquez & Unikel, on food and drug markets, conceptions of bodies,
2004; Hooper & Garner, 1986). Media images, management of bodies, and other macrostructural
ideals, and ideas all carry important consistent factors. In some tropical nations, the lush fresh pro-
themes, but the way those themes are integrated duce grown is exported, leaving locals with little
and applied—​ or locally reconfigured into what fresh, healthy food in addition to other problematic
Collier and Ong call “assemblages” (2005)—​can pricing problems (Krueger, 1991). Thus, countries
vary. While young women in the high-​BMI context like Belize are skyrocketing up global charts of “fat-
of the United Arab Emirates, for example, aspire to test nations” (WHO, 2016). The flow of money
global slender ideals, they also have to navigate a also is tightly related to power in a society. Thus,
highly gender segregated terrain. This environment when multi-​national corporations enter a develop-
includes particular generational and physical spaces ing country, the ideals they bring in terms of work
of accepted discussion and behavior regarding such ethic, bodily expectations, daily routines, and gen-
topics (Trainer, 2017). In Belize, upon the arrival dered values can influence profoundly the local
of both globally produced television and Internet landscape. The individuated ideal worker, often
access, young women proved highly resistant to gendered, has been perpetuated by global capitalism
thin images from US-​based media. Research found (Kondo, 1990). In countries like China, the com-
that this response related to a strong ethnopsychol- pensation a young woman may receive in a transna-
ogy of self-​protection and self-​care (Anderson-​Fye, tional corporate factory far outweighs her earning
2003, 2004). The young women interpreted global potential in a rural community. Thus, ironically, she
media messages to mean that it was appropriate for is put forward communally to be socialized into an
females to make themselves suffer to achieve certain individuated personhood and body that may cre-
physical ideals, a concept completely unaccept- ate habitual strains, as well as global aspirations of
able to their overall ethos (Anderson-​Fye, 2004). upward mobility requiring thin ideal bodies. While
Therefore, mediascapes are important in their wide- myriad ways exist in which global economic flows
spread access, but local interpretation and uptake is may impact individuals in terms of urban or trans-
at least as significant—​and often more so. national migration, eating behaviors, and activity
Technologies also have been essential to global- patterns, the notion of the individualized body that
ization (Appadurai, 1996). The expansion of air can be controlled in a Cartesian “mind over matter”
travel to previously unserved or less-​served areas fashion tends to be a widespread cultural influence
has accelerated related processes by moving people, affecting generations (Becker, 1995).
products, and media more quickly. Such move- The global proliferation of biomedicine is con-
ment has both positive and problematic effects on sistent with contemporary economic development
body image and eating concerns. Individuals may and its implicit assumptions about personhood and
desire upward class movement that appears to mechanistic bodies. The near worldwide expan-
require bodily control and discipline. They may also sion of biomedical technologies and practices have

Anderson-Fye 201
advanced population health, increased lifespan, otherwise pernicious disorder was present). In these
and impacted other positive outcomes of human situations, some treatment modalities from Western
well-​being. However, these systems have also been cultures were effective, but other local variants also
documented to further ensconce global ideals about were needed. Within East Asian nations, cultural
individual controllable bodies (Becker, 1995) that variations such as the dominant importance of “K-​
can (and should) be modified by pharmaceutical pop” (Korean pop music) culture help explain ide-
intervention (Ecks, 2013; Kitanaka, 2012). Several als and behaviors such as extreme plastic surgeries
recent studies have focused on the techniques global (Stone, 2013). The resistance to disordered eating
pharmaceutical companies use to market drugs such that young women in Belize demonstrated, particu-
as SSRIs to previously resistant cultures such as in larly in circumstances where most others around the
Japan (Kitanaka, 2012) and India (Ecks, 2013). world were susceptible, informed the global health
Such trends impart messages consistent with striv- community that such disorders were not a uni-
ing toward ideal transnational bodies—​that is, they formly inevitable part of development (Anderson-​
should be constantly active and achieving—​ and Fye, 2004; Lester, 2004; Smolak & Levine, 2010).
therefore slender but strong. Thus, many kinds of Thus, specific variants of body image and eating
transnational markets, from food to manufacturing disorders that hold regional patterning are still
to “big pharma,” introduce and reinforce messages important to understand along with the underlying
of individuation, productivity, and power differen- structural dynamics of globalization that paradoxi-
tials that have been implicated in eating disorders. cally create simultaneous homogeneity and hetero-
Finally, the transnational flow of ideas, ideals, geneity in cultural factors.
and ideology that occurs via the other four routes Taken together, globalization theories and data
of migration and travel, multifaceted media out- urge examination of powerful transnational trends
lets, new technologies, and transnational economic that shape the local context for many regard-
institutions is important for body image and eating ing body image and healthy eating. Within those
disorders in an increasingly interconnected world. powerful structural scaffolds, important cultural,
These ideological or symbolic components have institutional and micro-​level differences allow dif-
been discussed above and span the very conception ferent sorts of bodies and eating behaviors to pro-
of what is a body through to aesthetic and moral liferate. While local body preferences, trends, and
ideals. Understanding the global systems and pro- histories have been underexamined, they must be
cesses that drive outcomes in contemporary interna- taken into account along with these macrostructural
tional, regional, and community variation in body trends in order to understand outcomes (Lester &
image and eating outcomes is an essential advance- Anderson-​Fye, 2017).
ment in the field to understand the multilevel
impacts of sociocultural factors on health (Farmer, Toward Increased Multidisciplinary
1999; Nichter, 2008). Collaborative Research
These transnational dynamics of globalization do This chapter has sought to reflect on the mean-
not negate the importance of region and place. To ing and multilevel impacts of cultural factors on
the contrary, local adaptations or hybrid forms of normative and pathological body image and eating
body image and eating issues that also have transna- behaviors. Culture has been named as a component
tional aspects may be incredibly influential for both of disordered body image and eating behaviors for
scientific understanding and clinical outcomes. over a half-​century. Yet, its consideration in research
For example, Lee and colleagues documented non-​ and clinical settings is highly variable as models of
fat-​phobic anorexia in Singapore and other East biological universal mechanisms can eclipse those
Asian societies (Lee, Ho, & Hsu, 1993). Similarly, that leave room for the conceptualization of culture.
Korea—​a culture known for widespread eating dis- Over this same time period, multiple social sciences
orders and plastic surgery—​shows surprisingly low have continued to theorize culture as relevant to
fat phobia (Sun, n.d.). Fat phobia was previously health within and across societies. It is time to bring
a key diagnostic characteristic of anorexia nervosa closer our best knowledge of body image and eating
(DSM). The East Asian work showing a non-​fat-​ disorders with that of cultural processes in order to
phobic variant not only was important clinically improve scientific understanding, global health, and
but also helped the scientific community under- individual clinical outcomes across contexts.
stand which aspects of anorexia may still be cultur- While the field has some excellent models
ally shaped (and therefore variable even though the of collaborations among researchers spanning

202 Cultural Influences on Body Image

backgrounds from genetics to clinical bioscience to Anderson-​Fye, E. P. (2011). Body images in non-​western cul-
psychology to anthropology (e.g., Katzman et  al., tures. In T. Cash & L. Smolak (Eds.), Body image: A hand-
book of theory, research, and clinical practice. New  York,
2004, Becker, Keel, Anderson-​ Fye, & Thomas, NY: Guilford Press.
2004), these teams typically represent exceptions. Anderson-​ Fye, E. P., & Becker, A. E. (2003). Sociocultural
Promising directions in research in neighboring aspects of eating disorders. In J. K. Thompson (Ed.),
psychiatric concerns are building analogous teams Handbook of eating disorders and obesity (pp. 565–​589).
that are advancing understanding of systemic and Hoboken, NJ: Wiley.
Anderson-​ Fye, E. P., & Brewis, A. A., Eds. (2017). Fat
processual models of gene, behavior, and envi- planet:  Obesity, culture, and symbolic body capital (pp. 49–​
ronmental interactions (e.g., Worthman, Plotsky, 78). Santa Fe, NM:  School for Advanced Research Press;
Schecter, & Cummings, 2010) and epigenetics Albuquerque: University of New Mexico Press.
(Ptak & Petronis, 2010), among other multilevel Anderson-​Fye, E., & Lin, J. (2009). Belief and behavior aspects
models that take cultural factors seriously (Suomi, of the EAT-​26:  The case of schoolgirls in Belize. Culture,
Medicine, and Psychiatry, 33, 623–​638.
2011). Multidisciplinary and mixed-​ methods Anderson-​Fye, E. P., McClure, S. M., Floriano, M., Bharati, A.,
research is needed both to understand and inter- Chen, Y. & James, C. (2017). Fat and too fat: risk and protec-
vene effectively in multilevel cultural factors related tion for obesity stigma in three countries. In E. P. Anderson-​
to body image and eating disorders. Future work in Fye & A. A. Brewis (Eds.), Fat planet: Obesity, culture, and
this area should seek to enhance and support col- symbolic body capital (pp. 49–​78). Santa Fe, NM: School for
Advanced Research Press; Albuquerque:  University of New
laborative teams across biosciences, clinical sciences, Mexico Press.
and social sciences that bring the various epistemol- Appadurai, A. (1996). Modernity at large cultural dimensions of
ogies, theories, methods, and data that would allow globalization. Minneapolis: University of Minnesota Press.
better handling of multilevel understanding of these Arcelus, J., Mitchell, A. J., Wales, J., & Nielsen, S. (2011).
complex issues. Mortality rates in patients with anorexia nervosa and other
eating disorders:  A  meta-​analysis of 36 studies. Archives of
General Psychiatry, 68, 724–​731.
Note Aubrey, J. S. (2006). Effects of sexually objectifying media
1. The term “Western” is an inaccurate gloss primarily refer- on self-​ objectification and body surveillance in under-
ring to the United States, Canada, and western Europe. Body graduates:  Results of a two-​ year panel study. Journal of
image trends in these areas tend to be similar to those in Communication, 56, 625–​625.
Australia and increasingly eastern Europe and other regions. Azuma, Y., & Henmi, M. (1982). A study on the incidence of
Even though this term is not precise, it is widely accepted anorexia nervosa in schoolgirls. Annual Report of Research
and economical to employ in a broad discussion. Precise Group in Eating Disorders, 30–​34.
countries, regions, or cultural groups are used when discuss- Baum, A. (2006). Eating disorders in the male athlete. Sports
ing specific studies. Medicine, 36, 1–​6.
Beals, K., & Manore, M. M. (2004). Behavioral, physical, and
References psychological characteristics of female athletes with subclini-
Aardoom, J. J., Dingemans, A. E., Boogaard, L. H., & Van Furth, cal eating disorders. International Journal of Sport Nutrition
E. F. (2014). Internet and patient empowerment in individu- and Exercise Metabolism, 10, 128–​143.
als with symptoms of an eating disorder:  A  cross-​sectional Becker, A. E. (1995). Body, self, and society: The view from Fiji.
investigation of a pro-​recovery focused e-​community. Eating Philadelphia: University of Pennsylvania Press.
Behaviors, 15, 350–​356. Becker, A. E. (2004). Television, disordered eating and young women
Alexander, S. M. (2014). The corporate masquerade: Branding in Fiji: Negotiating body image and identity during rapid social
masculinity through Halloween costumes. Journal of Men’s change. Culture, Medicine and Psychiatry, 28, 533–​559.
Studies, 22, 180–​193. Becker, A. E. (2017). Body size, social standing, and weight
Anderson, C., & Petrie, T. A. (2012). Prevalence of disordered management: The view from Fiji. In E. P. Anderson-​Fye &
eating and pathogenic weight control behaviors among A. A. Brewis (Eds.), Fat planet:  Obesity, culture, and sym-
NCAA division I female collegiate gymnasts and swimmers. bolic body capital (pp. 79–​96). Santa Fe, NM:  School for
Research Quarterly for Exercise and Sport, 83, 120–​124. Advanced Research Press; Albuquerque:  University of New
Anderson-​ Fye, E. P. (2003). Never leave your- Mexico Press.
self:  Ethnopsychology as mediator of psychological global- Becker, A. E., Burwell, R., Gilman, S., Herzog, D., & Hamburg,
ization among Belizean schoolgirls. Ethos, 31, 77–​112. P. (2002). Eating behaviours and attitudes following pro-
Anderson-​ Fye, E. P. (2004). A “Coca-​Cola” shape:  Cultural longed exposure to television among ethnic Fijian adolescent
change, body image and disorders in San Andrés, Belize. girls. British Journal of Psychiatry, 180, 509–​514.
Culture, Medicine, & Psychiatry, 28, 561–​595. Becker, A. E., Hadley Arrindell, A., Perloe, A., Fay, K., & Striegel-​
Anderson-​Fye, E. P. (2008). Cross-​cultural issues in body image Moore, R. H. (2010). A qualitative study of perceived social
and eating problems among children and adolescents. In L. barriers to care for eating disorders: Perspectives from ethni-
Smolak & J. K. Thompson (Eds.), Body image, eating disor- cally diverse health care consumers. International Journal of
ders, and obesity in youth: Assessment, prevention, and treatment Eating Disorders, 43, 633–​647.
(pp. 144–​174). Washington, DC:  American Psychological Becker, A. E., Keel, P., Anderson-​Fye, E., & Thomas, J. (2004).
Association Press. Genes and/​or jeans? Genetic and socio-​cultural contributions

Anderson-Fye 203
 to risk for eating disorders. Journal of Addictive Diseases, 23, Burkett, G. L. (1991). Culture, illness and the biopsychosocial
81–​103. model. Family Medicine, 23, 287–​291.
Bemporad, J. R. (1997). Cultural and historical aspects of eating Byrne, K. (2011). Tuberculosis and the Victorian literary imagina-
disorders. Theoretical Medicine, 18, 401–​420. tion. Cambridge, UK: Cambridge University Press.
Betancourt, H., & López, S. R. (1993). The study of culture, Byrne, S. M., & McLean, N. J. (2002). The cognitive-​behavioral
ethnicity and race in American Psychology. American model of bulimia nervosa: A direct evaluation. International
Psychologist, 48, 629–​637. Journal of Eating Disorders, 31, 17–​31.
Bergeron, S. M., & Senn, C. Y. (1998). Body image and socio- Callahan, D. (2013). Obesity: Chasing an elusive epidemic. The
cultural norms:  A  comparison of heterosexual and lesbian Hastings Report, 43, 34–​40.
women. Psychology of Women Quarterly, 22, 385–​401. Casper, M. J. (2017). Excess gains and losses:  Maternal obe-
Bojorquez, I., & Unikel, C. (2004), Presence of disordered eating sity, infant mortality, and the biopolitics of blame. In E. P.
among Mexican teenage women from a semi-​urban area: Its Anderson-​Fye & A. A. Brewis (Eds.), Fat planet:  Obesity,
relation to the cultural hypothesis. European Eating Disorders culture, and symbolic body capital (pp. 149–​170).
Review, 12, 197–​202. Santa Fe, NM:  School for Advanced Research Press;
Bordo, S. (1993). Unbearable weight: Feminism, western culture, Albuquerque: University of New Mexico Press.
and the body. Berkeley: University of California Press. Child Trends. (2014). Immigrant Children: Indicators on Children
Borrell-​Carrió, F., Suchman, A. L., & Epstein, R. M. (2004). and Youth. Washington, DC.
The biopsychosocial model 25  years later:  Principles, prac- Cohn, L. D., & Adler, N. E. (1992). Female and male percep-
tice, and scientific inquiry. Annals of Family Medicine, 2, tions of ideal body shapes: Distorted views among Caucasian
576–​582. college students. Psychology of Women Quarterly, 16, 69–​79.
Borzekowski, D. L., & Bayer, A. M. (2005). Body image and Cole, E. R. (2009). Intersectionality and research in psychology.
media use among adolescents. Adolescent Medicine Clinics, American Psychologist, 64, 170–​180.
16, 289. Collier, S. J., & Ong, A. (2005). Global assemblages, anthropo-
Borzekowski, D. L., Schenk, S., Wilson, J. L., & Peebles, R. logical problems. In S. J. Collier & A. Ong (Eds.), Global
(2010). e-​Ana and e-​Mia: A content analysis of pro–​eating assemblages: Technology, politics, and ethics as anthropological
disorder web sites. American Journal of Public Health, 100, problems (pp. 3–​21). Malden, MA: Blackwell.
1526–​1534. Craig, P. L., Swinburn, B. A., Matenga-​Smith, T., Matangi, H.,
Boskind-​Lodahl, M., & White, W. C., Jr. (1978). The definition & Vaughn, G. (1996). Do Polynesians still believe that big is
and treatment of bulimarexia in college women: A pilot study. beautiful? Comparison of body size perceptions and prefer-
Journal of American College Health Association, 27, 84–​97. ences of Cook Islands, Maori and Australians. New Zealand
Boulware, L. E., Cooper, L. A., Ratner, L. E., LaVeist, T. A., & Medical Journal, 109, 200–​203.
Powe, N. R. (2003). Race and trust in the health care system. Crenshaw, K. (1989). Demarginalizing the intersection of race
Public Health Reports, 118, 358–​365. and sex: a Black feminist critique of antidiscrimination doc-
Boyatzis, C. J., & Quinlan, K. B. (2008). Women’s body trine, feminist theory and antiracist politics. The University of
image, disordered eating, and religion:  A  critical review of Chicago Legal Forum, 140, 139–​167.
the literature. In R. Piedmont (Ed.), Research in the social Currie, A. (2010). Sport and eating disorders:  Understanding
scientific study of religion (Vol. 19, pp. 183–​208). Leiden, and managing the risks. Asian Journal of Sports Medicine,
Netherlands: Brill. 1, 63–​68.
Brewis, A. (2014). Stigma and the perpetuation of obesity. Social Cusumano, D. L., & Thompson, J. K. (1997). Body image and
Science and Medicine, 118, 152–​158. body shape ideals in magazines:  Exposure, awareness, and
Brewis, A. (2017). Introduction:  Making sense of new global internalization. Sex Roles, 37, 701–​721.
body norms (pp. 1–​14). In E. P. Anderson-​Fye & A. A. Brewis Derenne, J. L., & Beresin, E. V. (2006). Body image, media, and
(Eds.), Fat planet: Obesity, culture, and symbolic body capital eating disorders. Academic Psychiatry, 30, 257–​261.
(pp. 79–​96). Santa Fe, NM: School for Advanced Research Dittmar, H., Halliwell, E., & Ive, S. (2006). Does Barbie make
Press; Albuquerque: University of New Mexico Press. girls want to be thin? The effect of experimental exposure to
Brewis, A., & McGarvey, S. (2000). Body image, body size, and images of dolls on the body image of 5-​to 8-​year-​old girls.
Samoan ecological and individual modernization. Ecology of Developmental Psychology, 42, 283.
Food and Nutrition, 39, 105–​120. Durkin, S. J., & Paxton, S. J. (2002). Predictors of vulnerabil-
Brewis, A., Wutich, A., Falletta-​Cowden, A., & Rodriguez-​Soto, ity to reduced body image satisfaction and psychological
I. (2011). Body norms and fat stigma in global perspective. wellbeing in response to exposure to idealized female media
Current Anthropology, 52, 269–​276. images in adolescent girls. Journal of Psychosomatic Research,
Brown, L. B. (2009). Teaching the “health at every size” para- 53, 995–​1005.
digm benefits future fitness and health professionals. Journal Ecks, S. (2013). Eating drugs: Psychopharmaceutical pluralism in
of Nutrition Education and Behavior, 41, 144–​145. India. New York: New York University Press.
Brownell, K. D., & Napolitano, M. A. (1995). Distorting reality Edmonds, A. (2012). Body image in non-​western societies. In T.
for children: Body size proportions of Barbie and Ken dolls. Cash (Ed.), The encyclopedia of body image and human appear-
International Journal of Eating Disorders, 18, 295–​298. ance (pp. 238–​242). New York, NY: Academic Press.
Bruch, H. (1973). Eating disorders: Obesity, anorexia nervosa, and Edmonds, A., & Mears, A. (2017). Managing body capital in
the person within. New York, NY: Basic Books. the fields of labor, sex, and health. In E. P. Anderson-​Fye
Bruch, H. (1978). The golden cage: The enigma of anorexia ner- & A. A. Brewis (Eds.), Fat planet: Obesity, culture, and sym-
vosa. Cambridge, MA: Harvard University Press. bolic body capital (pp. 33–​48). Santa Fe, NM:  School for
Brumberg, J. J. (1988). Fasting girls: The history of anorexia ner- Advanced Research Press; Albuquerque:  University of New
vosa. Cambridge, MA: Harvard University Press. Mexico Press.

204 Cultural Influences on Body Image

Engel, G. L. (1977). The need for a new medical model: A chal- Research methods for anthropological studies of food and nutri-
lenge for biomedicine. Science, 196, 129–​136. tion. Oxford, UK: Berghahn.
Engel, G. L. (1980). The clinical application of the biopsycho- Harrison, K. (2000). The body electric:  Thin‐ideal media and
social model. American Journal of Psychiatry, 137, 535–​544. eating disorders in adolescents. Journal of Communication,
Fardouly, J., Diedrichs, P. C., Vartanian, L. R., & Halliwell, E. 50, 119–​143.
(2015). Social comparisons on social media:  The impact Harrison, K., & Cantor, J. (1997). The relationship between
of Facebook on young women’s body image concerns and media consumption and eating disorders. Journal of
mood. Body Image, 13, 38–​45. Communication, 47, 40–​67.
Farmer, P. (1999). Infections and inequalities: The modern plagues. Hayaki, J., Friedman, M. A., & Brownell, K. D. (2002).
Berkeley: University of California Press. Shame and severity of bulimic symptoms. Eating Behaviors,
Farmer, P., & Connors, M. (1996). Women, Poverty and 3, 73–​83.
AIDS:  Sex, Drugs and Structural Violence. Monroe, Holtkamp, K., Hebebrand, J., & Herpertz-​Dahlmann, B. (2004).
ME: Common Courage Press. The contribution of anxiety and food restriction on physical
Fikkan, J. & Rothblum, E. D. (2005). Weight bias in employment. activity levels in acute anorexia nervosa. International Journal
In K. D. Brownell, R. M. Puhl, M. B. Schwartz, & L. Rudd of Eating Disorders, 36, 163–​171.
(Eds.), Bias, stigma, discrimination, and obesity (pp. 13–​28). Hooper, M., & Garner, D. (1986). Application of the Eating
New York: Guilford Press. Disorders Inventory to a sample of black, white, and mixed-​
Friesen, A., Holmqvist, K., & Anderson-​Fye, E. (2013). Body race girls in Zimbabwe. International Journal of Eating
image and child well-​being. Chapter 83. In A. B. Ben-​Arieh, Disorders, 5, 161–​168.
F. Casas, I. Frones, J. Korbin (Eds.), Handbook of child well-​ Hruschka, D. J. (2017). From thin to fat and back again: A dual
being. New York, NY: Springer. process model of the big body mass reversal. In E. P. Anderson-​
Frith, H., & Gleeson, K. (2004). Clothing and embodi- Fye & A. A. Brewis (Eds.), Fat planet: Obesity, culture, and
ment: Men managing body image and appearance. Psychology symbolic body capital (pp. 193–​204). Santa Fe, NM: School
of Men and Masculinity, 5, 40–​48. for Advanced Research Press; Albuquerque:  University of
Gaines, A., ed. (1992). Ethnopsychiatry: The cultural constructions New Mexico Press.
of professional and folk psychiatries. Albany: State University Hsu, L. K. G. (1989). The gender gap in eating disorders: Why
of New York Press. are the eating disorders more common among women?
Gard, M. C., & Freeman, C. P. (1996). The dismantling of a Clinical Psychology Review, 9, 393–​407.
myth: Review of eating disorders and socioeconomic status. Iacovou, M., & Sevilla, A. (2013). Infant feeding: The effects of
International Journal of Eating Disorders, 20, 1–​12. scheduled vs. on-​demand feeding on mothers’ wellbeing and
Ging, D., & Garvey, S. (2017). Written in these scars are the children’s cognitive development. European Journal of Public
stories I  can’t explain:  A  content analysis of pro-​ana and Health, 23, 13–​19.
thinspiration image sharing on Instagram. New Media and Jacobs, E., Rolle, I., Ferrans, C., Whitaker, E., & Warnecke,
Society, 19, 1–​20. R. (2006). Understanding African Americans’ views of the
Gordon, K. H., Brattole M. M., Wingate, L. R., & Joiner, T. E. trustworthiness of physicians. Journal of General Internal
(2006). The impact of client race on clinician detection of Medicine, 21, 642–​647.
eating disorders. Behavior Therapy, 37, 319–​325. James, C. (2012). Dying to be beautiful. Mona, Jamaica: University
Gordon, K. H., Perez, M., & Joiner, T. E. (2002). The impact of West Indies.
of racial stereotypes on eating disorder recognition. Kaminski, P. L., Chapman, B. P., Haynes, S. D., & Own, L. (2005).
International Journal of Eating Disorders, 32, 219–​224. Body image, eating behaviors, and attitudes toward exercise
Graff, K., Murnen, S. K., & Smolak, L. (2012). Too sexualized among gay and straight men. Eating Behaviors, 6, 179–​187.
to be taken seriously? Perceptions of a girl in childlike vs. Kann, L. (2016). Sexual identity, sex of sexual contacts, and
sexualizing clothing. Sex Roles, 66, 764–​775. health-​related behaviors among students in grades 9–​12—​
Greenhalgh, S. (2012). Weighty subjects: The biopolitics of the United States and selected sites, 2015. MMWR. Surveillance
U.S. war on fat. American Ethnologist, 39, 471–​487. Summaries, 65.
Greenhalgh, S. (2015). Fat talk nation:  The human costs of Katzman, M. A., Hermans, K. M. E., Van Hoeken, D., & Hoek,
America’s war on fat. Ithaca, NY: Cornell University Press. H. (2004). Not your “typical island woman”: Anorexia ner-
Greenleaf, C., Petrie, T. A., Carter, J., & Reel, J. J. (2009). vosa is reported only in subcultures in Curaçao. Culture,
Female collegiate athletes:  Prevalence of eating disorders Medicine, and Psychiatry, 28, 463–​492.
and disordered eating behaviors. Journal of American College Kelly, A. C., & Carter, J. C. (2013). Why self‐critical patients
Health, 57, 489–​495. present with more severe eating disorder pathology:  The
Groesz, L. M., Levine, M. P., & Murnen, S. K. (2002). The effect mediating role of shame. British Journal of Clinical Psychology,
of experimental presentation of thin media images on body 52, 148–​161.
satisfaction: A meta‐analytic review. International Journal of Kempa, M. L., & Thomas, A. J. (2007). Culturally sensitive
Eating Disorders, 31, 1–​16. assessment and treating of eating disorders. Eating Disorders,
Gutgesell, M. E., Moreau, K. L., & Thompson, D. L. (2003). 8, 17–​30.
Weight concerns, problem eating behaviors, and problem Kim, S. Y., Chen, Q., Li, J., Huang, X., & Moon, U. J. (2009).
drinking behaviors in female college athletes. Journal of Parent–​child acculturation, parenting, and adolescent
Athletic Training, 38, 62–​66. depressive symptoms in Chinese immigrant families. Journal
Hadley, C., & Weaver, L. J. (In press). Designing food insecu- of Family Psychology, 23, 426–​437.
rity scales from the ground up: An introduction and work- Kim, Y. (2011). Idol republic:  The global emergence of girl
ing example of building and using food insecurity scales industries and the commercialization of girl bodies. Journal
in anthropological research. In J. Brett & J. Chrzan (Eds.), of Gender Studies, 20, 333–​345.

Anderson-Fye 205
Kimber, M., Couturier, J., Georgiades, K., Wahoush, O., & Maine, M. D., Samuels, K. L., & Tantillo, M. (2015). Eating
Jack, S. M. (2014). Body image dissatisfaction among immi- disorders in adult women biopsychosocial, developmental,
grant children and adolescents in Canada and the United and clinical considerations. Advances in Eating Disorders, 3,
States:  A  scoping review. International Journal of Eating 133–​143.
Disorders, 47, 892–​897. Marcus, H. R., & Kitayama, S. (1991). Culture and the
Kitanaka, J. (2012). Depression in Japan:  Psychiatric cures for a self:  Implications for cognition, emotion and motivation.
society in distress. Princeton, NJ: Princeton University Press. Psychological Review, 98, 224–​253.
Klein, H., & Shiffman, K. S. (2006). Messages about physical Marsden, P., Karagianni, E., & Morgan, J. F. (2007). Spirituality
attractiveness in animated cartoons. Body Image, 3, 353–​363. and clinical care in eating disorders:  A  qualitative study.
Knauss, C., Paxton, S. J., & Alsaker, F. D. (2007). Relationships International Journal of Eating Disorders, 40, 7–​12.
amongst body dissatisfaction, internalisation of the media Matoti-​Mvalo, T., & Puoane, T. (2011). Perceptions of body size
body ideal and perceived pressure from media in adolescent and its association with HIV/​AIDS. South African Journal of
girls and boys. Body Image, 4, 353–​360. Clinical Nutrition, 24, 40–​45.
Kondo, D. (1990). Crafting selves:  Power, gender, and discourses McClelland, L., & Crisp, A. (2001). Anorexia nervosa and social
of identity in a Japanese workplace. Chicago, IL:  University class. International Journal of Eating Disorders, 29, 150–​156.
of Chicago Press. McClure, S. M. (2017). Symbolic body capital of an “other”
Krueger, A. O. (1991). The political economy of agriculture price kind:  African American females as a bracketed subunit in
policy: Volume 5. A synthesis of the political economy in develop- female body valuation. In E. P. Anderson-​Fye & A. A. Brewis
ing countries. Baltimore, MD:  Johns Hopkins Press for the (Eds.), Fat planet: Obesity, culture, and symbolic body capital
World Bank. (pp. 97–​124). Santa Fe, NM: School for Advanced Research
Latner, J. D., Stunkard, A. J., & Wilson, G. T. (2005). Press; Albuquerque: University of New Mexico Press.
Stigmatized students:  Age, sex, and ethnicity effects in the McClure, S. M. (2013). “It’s just gym”:  Physicality and iden-
stigmatization of obesity. Obesity Research, 13, 1226–​1231. tity among African American adolescent girls. PhD diss.,
Latzer, Y., Witztum, E., & Stein, D. (2008). Eating disorders Cleveland: Case Western Reserve University.
and disordered eating in Israel: An updated review. European McClure, S. M., Poole, M., & Anderson-​Fye, E. P. (2012).
Eating Disorders Review, 16, 361–​374. Race, ethnicity, and human appearance. In T. F. Cash (Ed.),
LeCook, B., McGuire, T., & Zaslavsky, A. (2012). Measuring Encyclopedia of body image and human appearance (Vol. 2, pp.
racial/​ethnic disparities in health care: Methods and practical 707–​710). San Diego: Academic Press.
issues. Health Services Research, 47, 1232–​1254. McCreary, D. R. (2012). Muscularity and body image. In T. F.
Lee, C. K., Kwak, Y. S., Yamamoto, J., Rhee, H., Kim, Y. S., Cash (Ed.), Encyclopedia of body image and human appearance
Han, J. H., Choi, J. O., Lee, Y. H. (1990). Psychiatric epi- (Vol 2, pp. 561–​567). San Diego: Academic Press.
demiology in Korea:  Part II:  Rural and urban differences. McLaren, L., & Kuh, D. (2004). Body dissatisfaction in midlife
Journal of Nervous & Mental Disease. women. Journal of Women and Aging, 16(1–​2), 35–​54.
Lee, K. (2013). Engaging in peer conversation about slim- McNamara, N., & Parsons, H. (2016). Everyone here wants eve-
ming predicts body dissatisfaction in Chinese college ryone else to get better’:  The role of social identity in eat-
women: A study in Hong Kong. Social Influence, 8, 1–​17. ing disorder recovery. British Journal of Social Psychology, 55,
Lee, S. (2004). Engaging culture:  An overdue task for eat- 662–​680.
ing disorders researchers. Culture, Medicine, Psychiatry, 28, Miller, M., & Pumariega, A. (2001). Culture and eating disor-
617–​621. ders:  A  historical and cross-​cultural review. Psychiatry, 64,
Lee, S., Ho, T. P., & Hsu, L. K. G. (1993). Fat phobic and non-​ 93–​110.
fat phobic anorexia nervosa:  A  comparative study of 70 Miller M., Verhegge R., Miller B., & Pumariega, A. (1999).
Chinese patients in Hong Kong. Psychological Medicine, 23, Assessment of risk of eating disorders among adolescents in
999–​1017. Appalachia. Journal of the American Academy of Child and
Lester, R. J. (2004). Eating disorders and the problem of ‘cul- Adolescent Psychiatry, 38, 437–​443.
ture’ in acculturation. Culture, Medicine and Psychiatry, 28, Mintz, L. B., & Betz, N. E. (1988). Prevalence and correlates of
607–​615. eating disordered behaviors among undergraduate women.
Lester, R. J. (2007). Critical therapeutics: Cultural politics and Journal of Counseling Psychology, 35, 463.
clinical reality in two eating disorder treatment centers. Monaghan L. F., & Atkinson M. (2014). Challenging myths of mas-
Medical Anthropology Quarterly, 21, 369–​387. culinity: Understanding physical cultures. Surrey, UK: Ashgate.
Lester, R. J. (2015). Book review: From virtue to vice: Negotiating Morrison, M. A., Morrison, T. G., & Sager, C. L. (2004). Does
anorexia. Medical Anthropology Quarterly, 30. body satisfaction differ between gay men and lesbian women
Lester, R., & Anderson-​Fye, E. P. (2017). Fat matters: Capital, and heterosexual men and women? A meta-​analytic review.
markets, and morality. In E. P. Anderson-​Fye & A. A. Brewis Body Image, 1, 127–​138.
(Eds.), Fat planet: Obesity, culture, and symbolic body capital Nadaoka, T., Oiji, A., Takahashi, S., Morioka, Y., Kashiwakura,
(pp. 15–​34). Santa Fe, NM: School for Advanced Research M., & Totsuka, S. (1996). An epidemiological study of eat-
Press; Albuquerque: University of New Mexico Press. ing disorders in a northern area of Japan. Acta Psychiatrica
Lillard, A. (1998). Ethnopsychologies:  Cultural variations in Scandinavica, 93, 305–​310.
theories of mind. Psychological Bulletin, 123, 3–​32. Nagel, K. L., & Jones, K. H. (1992). Sociological factors in the
Lipinski, J. P., & Pope, H. G., Jr. (2002). Body ideals in young development of eating disorders. Adolescence, 105, 107–​113.
Samoan men: A comparison with men in North America and Nasser, M., & Katzman, M. (1999). Eating disor-
Europe. International Journal of Men’s Health, 1, 163–​171. ders:  Transcultural perspectives inform prevention. In N.
Lutz, C. (1983). Parental goals, ethnopsychology, and the devel- Piran, M. P. Levine & C. Steiner-​Adair (Eds.), Preventing eat-
opment of emotional meaning. Ethos, 11, 246–​262. ing disorders (pp. 26–​44). Hove, UK: Brunner/​Mazel.

206 Cultural Influences on Body Image

Nasser, M., Katzman, M. A., & Gordon, R. A. (2001). Eating disor- dietary patterns during freshman and sophomore years of
ders and cultures in transition. Hove, UK: Brunner-​Routledge. college. Journal of American College Health, 53, 245–​251.
National Research Council and Institute of Medicine, Committee Reel, J. J., & Gill, D. L. (2001). Slim enough to swim? Weight
on the Health and Adjustment of Immigrant Children and pressures for competitive swimmers and coaching implica-
Families. (1998). In D. J. Hernandez & E. Charney (Eds.), tions. Sport Journal, 4. www.thesportjournal.org
From generation to generation:  The health and well-​being of Reinking, M. F., & Alexander, L. E. (2005). Prevalence of
children in immigrant families. Washington, DC:  National disordered-​eating behaviors in undergraduate female colle-
Academies Press. giate athletes and nonathletes. Journal of Athletic Training,
NCD Risk Factor Collaboration. (2016). Trends in adult body-​ 40, 47–​51.
mass index in 200 countries from 1975 to 2014: A pooled Reynolds, E. (2013). Underweight models banned in Israel to fight
analysis of 1698 population-​based measurement studies with anorexia, Daily Mail. London. Accessed April 30. 2017. http://​
19·2 million participants. Lancet, 387, 1377–​1396. www.dailymail.co.uk/​news/​article-​2256025/​Underweight-​
Nichter, M. (2001). Fat talk:  What girls and their parents say models-​BANNED-​Israel-​fight-​anorexia-​New-​law-​forces-​
about dieting. Cambridge, MA: Harvard University Press. women-​ads-​stay-​healthy-​faked-​images-​identified-​too.html
Nichter, M. (2008). Global health: Why cultural perceptions, social Roehling, M. V. (2002). Weight discrimination in the American
representations, and biopolitics matter. Tucson:  University of workplace:  Ethical issues and analysis. Journal of Business
Arizona Press. Ethics, 40, 177–​189.
Nichter, M., & Nichter, M. (1991). Hype and weight. Medical Rothblum, E. D., Brand, P. A., Miller, C. T., & Oetjen, H. A.
Anthropology, 13, 249–​284. (1990). The relationship between obesity, employment dis-
O’Connor, R. A., & van Esterik, P. (2015). From virtue to crimination, and employment-​related victimization. Journal
vice: Negotiating anorexia. New York, NY: Bergahn Books. of Vocational Behavior, 37, 251–​266.
Orbach, S. (1980). Fat is a feminist issue. New  York, Rouleau, C. R., & von Ranson, K. M. (2011). Potential risks of
NY: Berkeley Press. pro-​eating disorder websites. Clinical Psychology Review, 31,
Pagan, J. A., & Davila, A. (1997). Obesity, occupational attain- 525–​531.
ment, and earnings. Social Science Quarterly, 78, 756–​770. Rubin, L. R., Fitts, M. L., & Becker, A. E. (2003). “Whatever
Phillips, K. A. (1996). The broken mirror: Understanding and treat- feels good in my soul”:  Body ethics and aesthetics among
ing body dysmorphic disorder. Oxford: Oxford University Press. African American and Latina women. Culture, Medicine, and
Phillips, K. A. (2012). Body dysmorphic disorder. In T. Cash Psychiatry, 27, 49–​75.
et  al. (Eds.), The encyclopedia of body image and human Rudd, N. A., & Lennon, S. J. (2000). Body image and
appearance (pp. 74–​81). New York, NY: Elsevier. appearance-​ management behaviors in college women.
Phillips, K. A., Pinto, A., Hart, A. S., Coles, M. E., Eisen, J. L., Clothing and Textiles Research Journal, 18, 152–​162.
Menard, W., & Rasmussen, S. A. (2012). A comparison of Saraceno, B. (2003). Caring for children and adolescents with men-
insight in body dysmorphic disorder and obsessive compul- tal disorders: Setting WHO directions. Geneva: World Health
sive disorder. Journal of Psychiatric Research, 46, 1293–​1299. Organization.
Poorani, A. (2012). Who determines the ideal body? A summary Schooler, D. (2008). Real women have curves a longitudi-
of research findings on body image. New Media and Mass nal investigation of TV and the body image development
Communication, 2, 1–​12. of Latina adolescents. Journal of Adolescent Research, 23,
Pope, H. G., Olivardia, R., Gruber, A., & Borowiecki, J. (1999). 132–​153.
Evolving ideals of male body image as seen through action Schulz, A. J., & Mullings, L., (Eds). (2006). Intersectionality
toys. International Journal of Eating Disorders, 26, 65–​72. and health: An introduction. In Schulz, A. J., & Mullings, L.
Pope, H. G., Phillips, K. A., & Olivardia, R. (2000). The Adonis (Eds.). Gender, race, class, and health: Intersectional approaches
complex:  The secret crisis of male body obsession. New  York, (pp. 3–​17). San Francisco, CA: Jossey-​Bass.
NY: The Free Press. Shin, M. (2011). Present and future of aesthetic plastic surgery
Preti, A., Pinna, C., Nocco, S., Pilia, S., Mulliri, E., Micheli, in Korea. Journal of the Korean Medical Association, 54, 581.
V.,  . . .  Masala, C. (2007). Rural/​urban differences in the Shisana, O., Rehle, T., Simbayi, L. C., Parker, W., Zuma, K.,
distribution of eating disorder symptoms among adoles- Bhana, A.,  . . .  Onoya, D. (2005). South African National
cents from community samples. Australia and New Zealand HIV prevalence, HIV incidence, behaviour and communication
Journal of Psychiatry, 1, 525–​535. survey. Cape Town: Human Sciences Research Council Press.
Preti, A., Usai, A., Miotto, P., Petretto, D., & Masala, C. Shugart, H. (2008). Managing masculinities:  The metrosexual
(2008). Eating disorders among professional fashion models. moment. Communication and Critical/​Cultural Studies, 5,
Psychiatry Research, 159, 86–​94. 280–​300.
Ptak, C., & Petronis, A. (2010). Epigenetic approaches to psychi- Sloan, D. M. (2002). Does warm weather climate affect eating
atric disorders. Dialogues in Clinical Neuroscience, 12, 25–​35. disorder pathology? International Journal of Eating Disorders,
Puhl, R. M., & Heuer, C. A. (2010). Obesity stigma: Important 32, 240–​244.
considerations for public health. American Journal of Public Smedley, B., Stith, A., & Nelson, A. (Eds.). (2003). Unequal
Health, 100, 1019–​1028. treatment:  Confronting racial and ethnic disparities in health
Puhl, R. M., & Brownell, K. D. (2001). Bias, discrimination, care. Washington, DC: National Academies Press.
and obesity. Obesity Research, 9, 788–​805. Smolak, L., & Levine, M. (2010). Cultural influences on body
Pumarino, H., & Vivanco, N. (1987). Appetite and eating disor- image and the eating disorders. In Agras, W. S. (Eds.). Oxford
ders: An increasing pathology? Revista Medica de Chile, 110, handbook of eating disorders (pp. 223–​246). Oxford: Oxford
1081–​1092. University Press.
Racette, S. B., Deusinger, S. S., Strube, M. J., Highstein, G. R., Sobal, J. (2017). Sociological analysis of the stigmatisation of obe-
& Deusinger, R. H. (2005). Weight changes, exercise, and sity. In J. Germov & L. Williams (Eds.). A sociology of food

Anderson-Fye 207
 and nutrition: The social appetite (pp. 259–​274). Melbourne, Thompson, J. K., & Cafri, G. (2007). The muscu-
Australia: Oxford University Press. lar ideal:  Psychological, social and medical perspectives.
Sobo, E. (1993). One blood:  The Jamaican body. Albany:  State Washington, DC: American Psychological Association.
University of New York Press. Tobin, J., Wu, D., & Davidson, D. (1989). Preschool in three
Sontag, S. (1978). Illness as metaphor. New  York, NY:  Farrar, cultures:  Japan, China, and the United States. New Haven,
Straus and Giroux. CT: Yale University Press.
Spettigue, W., & Henderson, K. A. (2004). Eating disorders Torres-​McGehee, T. M., Monsma, E. V., Dompier, T. P., &
and the role of the media. Canadian Child and Adolescent Washburn, S. A. (2012). Eating disorder risk and the role of
Psychiatry Review, 13, 16–​19. clothing in collegiate cheerleaders’ body images. Journal of
Stampler, L. (2015). France just banned ultra-​ thin models. Athletic Training, 47, 541–​548.
Time. Accessed April 30, 2017. http://​time.com/​3770696/​ Trainer, S. (2017). Glocalizing beauty: Weight and body image in
france-​banned-​ultra-​thin-​models/​ the new Middle East. In E. P. Anderson-​Fye & A. A. Brewis
Steiner-​Adair, C. (1986). The body politic: Normal female ado- (Eds.), Fat planet: Obesity, culture, and symbolic body capital
lescent development and the development of eating disorders. (pp. 171–​192). Santa Fe, NM: School for Advanced Research
Journal of the American Academy of Psychoanalysis, 14, 95–​114. Press; Albuquerque: University of New Mexico Press.
Stice, E. (2001). Risk factors for eating pathology:  Recent Treasure, J. L., Wack, E. R., & Roberts, M. E. (2008). Models
advances and future directions. In R. H. Striegel Moore as a high-​risk group: The health implications of a size zero
& L. Smolak (Eds.), Eating disorders:  Innovative direc- culture. British Journal of Psychiatry, 192, 243–​244.
tions for research and practice (pp. 51–​73). Washington, Triandis, H. C. (1995). Individualism and collectivism: New direc-
DC: American Psychological Association. tions in social psychology. Boulder, CO: Westview Press.
Striegel-​Moore, R. H., Rosselli, F., Perrin, N., DeBar, L., Wilson, Triandis, H. C. (2001). Individualism-​collectivism and personal-
G. T., May, A., & Kraemer, H. C. (2009). Gender difference ity. Journal of Personality, 69, 907–​924.
in the prevalence of eating disorder symptoms. International Tyler, M., & Abbott, P. (1998). Chocs away: Weight watching in
Journal of Eating Disorders, 42, 471–​474. the contemporary airline industry. Sociology, 32, 433–​450.
Stone, Z. (2013, May 24). The K-​Pop plastic surgery obsession. Walker Bynum, C. (1988). Holy feast and holy fast: The religious
Atlantic Monthly. https://​www.theatlantic.com/​health/​archive/​ significance of food to medieval women. Berkeley:  University
2013/​05/​the-​k-​pop-​plastic-​surgery-​obsession/​276215/​ of California Press.
Suárez-​Orozco, C., & Suárez-​Orozco, M. (2001). Children of White, G. (1992). Ethnopsychology. In T. Schwartz, G. White,
immigration. Cambridge, MA: Harvard University Press. & C. Lutz (Eds.), New directions in psychological anthropology.
Sun, E. J. (n.d.). Fat-​phobia: XS in America, XL in Korea: Impact Cambridge, UK: Cambridge University Press.
of American culture exposure on body ideals of Korean Wilkins, J. A., Boland, F. J., & Albinson, J. (1991). A compari-
students (unpublished manuscript). Cleveland, OH:  Case son of male and female university athletes and nonathletes
Western Reserve University. on eating disorder indices: Are athletes protected? Journal of
Sundgot-​Borgen, J. (1994). Eating disorders in female athletes. Sport Behavior, 14, 129–​143.
Journal of Sports Medicine, 17, 176. Williams, D. R., & Mohammed, S. A. (2009). Discrimination
Sundgot-​Borgen, J., & Torstveit, M. K. (2004). Prevalence of and racial disparities in health: Evidence and needed research.
eating disorders in elite athletes is higher than in the general Journal of Behavioral Medicine, 32, 20.
population. Clinical Journal of Sport Medicine, 14, 25–​32. Williams, D. R., & Sternthal, M. (2010). Understanding racial/​
Suomi, S. J. (2011). Risk, resilience, and gene-​ environment ethnic disparities in health:  Sociological contributions.
interplay in primates. Journal of the Canadian Academy of Journal of Health and Social Behavior, 51, S15–​S27.
Child and Adolescent Psychiatry, 20, 289–​297. Williamson, I., & Hartley, P. (1998). British research into the
Szabo C. P. (1999). Eating attitudes among black South Africans. increased vulnerability of young gay men to eating distur-
American Journal of Psychiatry, 156, 981. bance and body dissatisfaction. European Eating Disorders
Szweda, S., & Thorne, P. (2002). The prevalence of eating disor- Review, 6, 160–​170.
ders in female health care students. Occupational Medicine, World Health Organization. (2005). Nutrition in adoles-
52, 113–​119. cence: Issues and challenges for the health sector. Geneva: World
Taylor, N. L. (2017). Fat is a linguistic issue: Discursive nego- Health Organization.
tiation of power, identity, and the gendered body among World Health Organization. (2016). Global database on body
youth. In E. P. Anderson-​Fye & A. A. Brewis (Eds.), Fat mass index. http://​apps.who.int/​bmi/​index.jsp
planet: Obesity, culture, and symbolic body capital (pp. 125–​ Worthman, C., Plotsky, P., Schecter, D., & Cummings, C.
148). Santa Fe, NM:  School for Advanced Research Press; (2010). Formative experiences:  The interaction of caregiv-
Albuquerque: University of New Mexico Press. ing, culture, and developmental psychobiology. Cambridge,
Taylor, N. L. (2016). Schooled on fat:  What teens tell us about UK: Cambridge University Press.
gender, body image, and obesity. New York, NY: Routledge. Zucker, N. L., Womble, L. G., Williamson, D. A., & Perrin, L.
Thompson, B. (1992). A way outa’ no way:  Eating problems A. (1999). Protective factors for eating disorders in female
among African-​ American, Latina, and White women. college athletes. Journal of Treatment and Prevention, 7,
Gender and Society, 6, 546–​561. 207–​218.

208 Cultural Influences on Body Image

 PART 
3
Assessment and
Comorbidities of the
Eating Disorders
 CH A PT E R


Psychological Assessment of the
10 Eating Disorders

Drew A. Anderson, Joseph Donahue, Lauren E. Ehrlich, and Sasha Gorrell

Abstract
Clinicians and researchers have several approaches with which to assess eating disorder and related
symptomatology, including interviews, self-​report instruments, and behavioral measures. The purpose of
this chapter is to describe a process, based on a functional approach, that will help assessors to develop
assessments and choose instruments for eating disorders and eating-​related problems. This approach
takes into account both theoretical and practical concerns and allows assessors to individualize their
assessments depending on their particular needs. This process starts with broad considerations about
the context in which the assessment is to be given and ends with the choice of specific instruments to
be used.
Key Words:  assessment, behavioral assessment, eating disorder, functional, interview, self-​report

Introduction instruments. The steps can be summarized by the

Discussions of assessment are often limited to use of a few key questions: “where,” “why,” “what,”
reviews of particular instruments and their psy- and “how.”
chometric properties, which reduces the topic to
its components. Assessment, however, is a proc- Step One—​“Where”: Understanding
ess, and there is no one assessment instrument or the Context of the Assessment
test battery that is useful or appropriate in all cir- The first step is in this process is to take into
cumstances. Instead, any choices concerning spe- account the broad context in which the assessment
cific instruments are secondary to the function of occurs. Context can be thought of as the “where” of
the assessment. The purpose of this chapter is to assessment. Specifically, under what circumstances
describe a process, based on a functional approach, is the assessment taking place? For instance, is the
that will help assessors to individualize their own assessment taking place in a psychiatric treatment
assessments based on their particular circumstances, facility as part of regular clinical practice, or is it
taking into account both theoretical and practical part of a controlled research trial? This broad con-
concerns. Thinking in terms of the function or goals text shapes the larger goals of the entire assessment
of the assessment frees the assessor to consider how process. As an example, if the assessment is being
to best go about achieving those goals. In this chap- conducted as part of a controlled clinical trial the
ter we focus on assessment within the broad context assessor often will have the opportunity to use
of psychotherapy, but the process is applicable to lengthier, more time-​intensive instruments, includ-
other contexts. ing semistructured interviews, and to include more
This process starts with broad considerations elaborate assessment batteries than he or she might
about the context in which the assessment is to use in a purely clinical setting. There are also dis-
be given and ends with the choice of specific crete dependent variables specified in the research

211
design that must be assessed, which will dictate to be more appropriate instruments. Conversely, in
some degree the exact measures that will be used. the example of the school nurse given previously,
Further, the psychometric properties of the instru- in the context of a school setting the function of
ments themselves will be of paramount importance. the assessment would be for screening purposes.
Conversely, psychologists in private practice are Thus, a brief screening instrument would be a more
often faced with a number of limitations in their appropriate choice compared to longer and more
ability to conduct elaborate assessments, includ- elaborate clinical interviews. The function of the
ing time constraints and difficulties in obtaining assessment also helps dictate how often assessments
reimbursement from third-​ party payers (Eisman should occur. For example, if the function of an
et  al., 2000; Turchik, Karpenko, Hammers, & assessment is to track treatment progress, then peri-
McNamara, 2007; Wright et al., 2016). Thus, the odic or even weekly administration of a brief self-​
assessor might not want to limit him-​or herself to report instrument would be indicated. However,
a semistructured interview devoted entirely to eat- in the context of screening such a schedule would
ing disorder symptomatology, but instead conduct clearly be excessive.
an unstructured clinical interview that covers a Although there are a number of potential func-
broader range of topics and supports the diagnos- tions for an assessment, they can be grouped into
tic and treatment recommendations with a few easy a few common categories, including screening,
to score but psychometrically sound instruments diagnosis, treatment planning, and outcome. These
that the patient can fill out in the waiting room. functions can overlap; for example, a suicide screen-
(For further suggestions, see Turchik et al., 2007). ing can be embedded in a larger diagnostic inter-
As another example, school nurses are operating in view, but we discuss them separately.
a context in which they generally do not need to
conduct elaborate assessments. They may have only Screening
the need, or the time, to ask a few brief screening Screening tests are quick, easy to use, and inex-
questions of an adolescent they suspect of an eating pensive procedures given to an entire relevant
disorder to refer the individual for further evalua- population to determine which apparently healthy
tion or treatment. individuals are actually at high risk for a particular
The notion of context is intimately connected disorder (Evans, Galen, & Britt, 2005). It is beyond
with the function or purpose of the assessment. The the scope of this chapter to detail the statistical and
second step in designing an assessment is to exam- mathematical principles involved in determining
ine the assessment’s intended function. the efficacy of screening tests; Grimes and Schultz
(2002) provide a concise overview of these issues.
Step Two—​“Why”: Determining However, we should note one issue that is relevant
the Function of the Assessment for the evaluation of eating disorders screenings.
The function of an assessment is the “why” of the The issue is that the positive predictive value (PPV)
assessment process. The assessor needs to determine of a screening test (i.e., the proportion of indi-
why exactly the assessment is being conducted. viduals with a positive test who actually have the
As mentioned previously, the context of the condition) varies with the prevalence rate of the
assessment (e.g., an intake at an inpatient psychi- condition, so even good screening tests can have
atric facility, an outpatient treatment session, or a poor PPV when applied to low-​prevalence popu-
forensic evaluation) often dictates the function of lations (Grimes & Schultz, 2002; Nielsen & Lang,
the assessment. However, an assessment can have a 1999). Thus, because the rate of eating disorders is
number of potential functions, and the function of relatively low in the general population, the inci-
the assessment will dictate, in large part, the assess- dence of false positives will be relatively high. As an
ment measures and procedures that will be used. example, one study of the SCOFF (Morgan, Reid,
For example, it makes little sense to use a screening & Lacey, 2000; Parker, Lyons, & Bonner, 2005), a
measure for eating disorders at an eating disorder screening test for eating disorders discussed later in
inpatient treatment program, since presumably the this chapter, yielded a positive predictive value of
individual being assessed has already acknowledged 24.4% (Luck et  al., 2002). In this study, only 11
eating-​related pathology. In that context, the func- of the 45 cases identified by the SCOFF actually
tion of the assessment would more likely be for had an eating disorder, and the authors noted the
diagnosis and treatment planning, and a clinical low prevalence of eating disorders in their sample
interview and longer self-​report instruments would as an explanation for this finding. A  false-​positive

212 Psycholo gical Assessment

result on a screening test can lead to potential harm will be the diagnostic criteria themselves. The iden-
(Grimes & Schultz, 2002), so individuals using tification of domains of interest for treatment plan-
screening tests for eating disorders should be aware ning and evaluation of outcome, however, can be
of this fact. much more complicated.
An assessment for treatment planning and
Diagnosis outcome typically includes many of the same
Generally, issues of diagnosis are subsumed domains of interest that would be evaluated in
under the function of treatment planning and out- screening or diagnosis functions. It goes fur-
come, discussed later in this chapter. There are some ther, however, in that it allows for the assessor
instances, however, when an assessor simply needs to take into account process or change variables
to verify whether an individual meets diagnostic as well as more theoretical considerations about
criteria for an eating disorder. For example, third-​ the eating disorders and associated psychopathol-
party payers may require formal documentation of ogy. As part of this process, the assessor should
a diagnosis in order to reimburse for services. Also, reflect on the criteria for treatment success that
some therapies minimize the necessity of a specific will be used. Some questions that are useful at
diagnosis for treatment, so diagnosis is not always this stage of the process include:  What would
a necessary part of an assessment for treatment have to change for this person to be considered
planning purposes (e.g., Fairburn, 2008). As such, to be “cured” or “improved?” What would indi-
diagnosis does represent a distinct function of the cate that the individual is not improving? What
assessment process. would be desirable but not essential for treatment
to be considered a success? The answers to these
Treatment Planning and Outcome questions involve considerations of the theoreti-
Treatment planning and outcome represents the cal assumptions underlying treatment, discussed
broadest function for the assessment of eating disor- in the next section.
ders. This includes such things as making treatment
decisions (e.g., which treatment will work best for Underlying Theoretical Assumptions
this individual?) and evaluating treatment outcome It is extremely important to pay attention to
(e.g., did this patient improve, and if so by how the theoretical basis of the therapeutic approach
much?). This type of assessment can be done more being used when determining domains of interest.
or less frequently, depending on the circumstances Therapy and therapeutic techniques do not come
of the assessment. For example, some instruments out of a vacuum; they are developed from theoret-
may be used only at the beginning and end of treat- ical assumptions about the nature of personality,
ment, while others may be used more frequently to psychopathology, and principles of effective change
track treatment progress more closely. (Kanfer & Schefft, 1988; Pachankis & Goldfried,
Once the assessor determines the context and 2007). These assumptions influence what domains
function of the assessment, he or she must deter- will be considered to be of interest. For example,
mine the exact domains or constructs that are rel- the original cognitive-​behavioral account of bulimia
evant to the assessment question. Determining this nervosa posits that five core symptom domains are
is the next step in the assessment process. responsible for the maintenance of the disorder: low
self-​
esteem, overconcern with body weight and
Step 3—​“What”: Determining the Domains shape, extreme dietary restraint, binge eating, and
or Constructs of Interest purgative behavior (Fairburn, Marcus, & Wilson,
Determining the domains or constructs to be 1993). According to this model, an individual
assessed can be thought of as the “what” of assess- with bulimia nervosa being treated with cognitive-​
ment. In this step the assessor asks what specific behavioral therapy (CBT) should show improve-
thoughts, attitudes, and behaviors should be ment in all five of these domains to be considered
assessed. For screening and diagnosis this decision a treatment success, with reductions in overconcern
can be fairly straightforward. As mentioned previ- with body weight and shape being particularly
ously, screening measures generally ask broadly important for long-​term success (Fairburn, 1997a;
about the most common signs and symptoms of Fairburn et  al., 1993). Accordingly, if an individ-
eating disorders (e.g., underweight status, com- ual with bulimia nervosa is being treated within a
pensatory behavior, binge eating, and body image traditional CBT framework, measurement of these
disturbance). For diagnosis, the domains of interest domains will be an essential part of the assessment

Anderson, Donahue, EHrlich, Gorrell 213

of treatment progress and outcome (Anderson & Thresholds for Recovery
Maloney, 2001). Further, the newest revision of The assessor should also consider the threshold at
CBT for eating disorders, called enhanced CBT which a change in a given domain could be consid-
(CBT-​E), hypothesizes that additional mechanisms ered an indicator of success. This issue can become
may be operating in some individuals with eating quite difficult when evaluating eating disorders. For
disorders (Fairburn, 2008; Fairburn, Cooper, & example, body image disturbance is both a com-
Shafran, 2003). Thus, an assessor evaluating patients mon symptom and one of the diagnostic criteria
being treated with CBT-​E may have to add instru- for anorexia and bulimia nervosa. However, some
ments over and above what might be necessary in a degree of body dissatisfaction is normative in the
traditional CBT evaluation, depending on a given general population (Duarte, Ferreira, Trindade, &
individual’s case formulation. As another exam- Pinto-​Gouveia, 2016; Rodin, Silberstein, & Striegel-​
ple, interpersonal psychotherapy (IPT) posits that Moore, 1985). Given this “normative discontent,”
interpersonal problems underlie the maintenance should an individual being treated for an eating dis-
of bulimia nervosa (Fairburn, 1997b). As such, if order have to demonstrate a total absence of body
an individual with bulimia nervosa is being treated dissatisfaction to be considered a treatment success?
from within an IPT framework, measurement of Treatment studies suggest that this may be unlikely
interpersonal domains will be an essential part of (Lundgren, Danoff-​Burg, & Anderson, 2004), and
the assessment of treatment progress and outcome. some have argued that a return to normal function-
As can be seen from these examples, some of the ing is an unnecessarily stringent criterion for success
domains identified as essential from one therapeu- (Wise, 2004). Is it sufficient then to merely require
tic perspective might not be considered essential or them to reach a level of “normative discontent?”
even relevant from another therapeutic perspective, While more achievable, some would suggest that this
and even a single therapeutic modality may have represents a less-​than-​ideal outcome. Further, some
variations depending on individual patient charac- treatment approaches do not believe any reduc-
teristics. Assessment should reflect these differences, tion in body dissatisfaction is necessary for treat-
and all the essential domains thought to be nec- ment success (e.g., Acceptance and Commitment
essary for change in a given model of psychother- Therapy; Heffner & Eifert, 2004) and thus would
apy should be measured as part of the assessment see the issue as irrelevant. As another example, if an
process. individual with bulimia nervosa, purging subtype
improves from daily bulimic episodes to the point
Secondary Domains that she is only purging once a month on average,
One must also consider not only the core should this be considered a treatment success? These
domains of interest, however, but also other sec- examples illustrate the complexities the assessor can
ondary domains that might impact the onset and face in determining thresholds for change.
course of eating disorders as well. While it is not There are, however, a number of conceptual and
possible to evaluate every possible factor that might mathematical strategies for determining clinically
be related to the eating disorders, certain disor- significant change (Follette & Callaghan, 2001;
ders and life events have consistently been shown Jacobson, Roberts, Berns, & McGlinchey, 1999).
to be associated with eating-​related problems. For Perhaps the most widely known strategy for deter-
example, some mood disorders, anxiety disorders, mining clinically significant change was proposed
substance use disorders, personality disorders, by Jacobson and colleagues (Jacobson, Follette, &
and sexual abuse have repeatedly been found to Revenstorf, 1984; Jacobson et  al., 1999; Jacobson
be associated with eating disorders, and have also & Truax, 1991). In this methodology, two criteria
been hypothesized to play a causal role in their are necessary for clinically significant change. The
development (Becker & Grilo, 2015; Cassin & von first criterion is that the dependent measure showed
Ranson, 2005; Fichter, Quadflieg, & Hedlund, real change (i.e., the change was not due simply to
2006, 2008; O’Brien & Vincent, 2003; Sihvola measurement error). Jacobson and colleagues pro-
et  al., 2009). Further, these comorbidities may posed the reliable change index (RCI) as a statisti-
be associated with poorer treatment outcome for cal approach to determine if the change observed
eating disorders (Fichter et  al., 2006, 2008; Keel reflects more than the fluctuations of an imprecise
et al., 2003; Peterson, Becker, Treasure, Shafran, & measuring instrument (Jacobson & Truax, 1991).
Bryant-​Waugh, 2016). The RCI was originally designed to be conducted

214 Psycholo gical Assessment

on individual-​level data, but has been adapted for Step 4—​“How”: Determining the Specific
use on group-​level data (Jacobson & Truax, 1991). Instruments to Be Used in the Assessment
The second criterion is that by the end of treat- Once the assessor has covered the where, why,
ment clients should end up in a range that renders and what questions, he or she needs to answer
them indistinguishable from those in the normal “how.” Specifically, how should he or she assess
population; a number of different cutoff points for the domains identified in the previous step? One
determining when this occurs have been developed, important issue to consider is the format or formats
depending on the information available and the par- of the assessment instruments that will be used.
ticular needs of the researcher or clinician (Jacobson Interviews have traditionally been seen as the
et al., 1999). Jacobson and colleagues (1999) sum- method of choice for assessment of eating-​related
marized the possible outcomes using this approach. problems, particularly when evaluating more com-
They note that patients can be considered “recov- plex constructs such as binge eating (Fairburn &
ered” if the magnitude of change is statistically reli- Beglin, 1990, 1994). Individuals typically report
able and the client ends up within normal limits lower levels of eating-​ related pathology when
on the variable of interest or “improved but not responding via interview versus questionnaire
recovered” if the patient shows statistically reliable (Anderson & Maloney, 2001; Berg, Peterson,
change but ends treatment still somewhat dysfunc- Frazier, & Crow, 2011; Fairburn & Beglin, 1990,
tional. If the client ends up within normal limits 1994; French et  al., 1998; Keel, Crow, Davis, &
by the end of therapy but the magnitude of change Mitchell, 2002), and this has been interpreted as
is not statistically reliable, then the clinical signifi- evidence that individuals overendorse symptoms
cance of the change cannot be determined. Finally, on questionnaires (Fairburn & Beglin, 1990, 1994;
if the patient shows neither statistically reliable French et  al., 1998). However, a growing body of
change nor a recovery to within normal limits, the research has questioned this assumption, suggesting
patient would be considered “not recovered,” and if that individuals may respond more honestly if they
he or she showed statistically reliable change in the feel more anonymous or do not have to directly
direction opposite of recovery he or she would be face the person conducting the assessment. This has
classified as “deteriorated.” been shown in both clinical (Berg et al., 2011; Keel
A similar procedure for evaluating the clin- et al., 2002) and laboratory (Anderson, Simmons,
ical significance of change at the group level Milnes, & Earleywine, 2007; Berg et  al., 2011;
has been developed by Kendall and colleagues French et  al., 1998; Perry et  al., 2002; Lavender
(Kendall & Grove, 1988; Kendall, Marrs-​Garcia, & Anderson, 2008, 2009) contexts. This effect
Nath, & Sheldrick, 1999), which compares data presumably occurs because eating-​related pathol-
on treated individuals with that of normative ogy such as binge eating and purging are shameful
individuals. A  value of one standard deviation and embarrassing (Hamburg, Herzog, Brotman, &
around the normative mean has often been used Staisor, 1989; Hayaki, Friedman, & Brownell, 2002;
as the definition of clinically equivalent, although Vitousek, Daly, & Heiser, 1991) and it is therefore
other cutoff points may be used (Kendall et  al., easier to not have to admit to these behaviors in a
1999; Sheldrick, Kendall, & Heimberg, 2001; face-​to-​face setting. Thus, interview and self-​report
Wise, 2004). methods should not be used interchangeably, but it
These procedures may not be necessary for all is currently not clear which method of assessment
individuals. In particular, clinicians may not have produces the most valid assessment of eating-​related
the time to calculate an RCI on every patient. problems. Until this issue is resolved, we suggest
Also, the use of many of these procedures requires that assessors use both interview and self-​ report
that norms be available for the assessment instru- questionnaires in the assessment process.
ment being used, which is not always the case. A considerable number of instruments have
Nevertheless, these procedures provide theoreti- been developed to assess eating-​related constructs;
cal frameworks by which assessors can help deter- in fact, entire books have been devoted to the topic
mine what constitutes “improved” or “recovered” in (e.g., Allison, 2009; Mitchell & Peterson, 2005;
patients being assessed. In any event, even if these Williamson, 1990). It is beyond the scope of this
procedures are not used the assessor should think chapter to review all of these measures or to discuss
deeply about these issues when assessing treatment their psychometrics in great detail. It is also beyond
outcome. the scope of this chapter to discuss basic issues in

Anderson, Donahue, EHrlich, Gorrell 215

psychometrics (for a brief review of psychometric menstruation was removed from the scale, the time
issues, see Anderson & Paulosky, 2004a). We only frame for diagnostic items changed from 6 months
note in passing that data collected via self-​report, to 3 months, and the weight requirement to meet
as is commonly done in adult assessment, is subject criteria for AN was removed. Otherwise, the EDE-​
to substantial error and bias, which can affect both 17.0D is identical to the older EDE 16.0D (Fairburn
reliability and validity (Korotitsch & Nelson-​Gray, & Beglin, 2008). The EDE 16.0D and previous ver-
1999). We will, however, review some of the mea- sions have demonstrated adequate reliability and
sures specifically developed to assess eating-​related validity, and norms are available (Anderson, De
pathology that are widely used in the literature or Young, & Walker, 2009; Berg, Peterson, Frazier, &
that have good psychometric properties. It is impor- Crow, 2012; Fairburn & Cooper, 1993). While psy-
tant to keep in mind, however, that assessors will chometric data on the newly released EDE 17.0D
likely need to supplement these measures based on have not yet been published, given its similarity to
their specific needs determined in the earlier other its predecessors it should also demonstrate good
steps of the assessment process. Indeed, the use of psychometric properties.
multiple measures is common among both clini- The EDE is unique in that the expert interviewer,
cians and researchers (Anderson & Paulosky, 2004b; not the interviewee, decides whether a particular
Williamson, Anderson, & Gleaves, 1996). eating episode is a binge or not. This is particularly
helpful because the DSM-​5 (American Psychiatric
Interview Association [APA], 2013) requires that a binge be
Interviews are generally more appropriate for objectively large and involve a loss of control, but
diagnosis, treatment planning, and evaluation than many laypersons do not use this definition when
they are for screening. In the appropriate context, describing an eating episode as a binge (Beglin &
however, they are arguably the most important Fairburn, 1992; Johnson, Boutelle, Torgrud, Davig,
means of data collection in the entire assessment & Turner, 2000; Telch, Pratt, & Niego, 1998). The
process (Groth-​Marnat, 1997). Interviews vary on EDE does have some shortcomings, however. In
their degree of structure. Unstructured interviews particular, the authors note that training is essen-
are more flexible and can be easily adapted to a par- tial if the EDE 17.0D is being used for research
ticular client and his or her unique circumstances, purposes (Fairburn et al., 2014), but such training
but they usually have poor or unknown psychomet- does not appear to be widely available. (The train-
ric properties (Groth-​Marnat, 1997; Korotitsch & ing schedule, as well as a copy of the EDE 17.0D
Nelson-​Gray, 1999). Unstructured interviews for itself, is available at http://​www.credo-​oxford.com).
eating-​related pathology are extremely common A second concern is that the EDE can also be too
in clinical practice (Anderson & Paulosky, 2004b); lengthy to complete in some contexts. Nevertheless,
Crowther and Sherwood (1997) and Peterson the EDE remains a valuable tool for the assessment
(2005) provide helpful guidelines for conduct- of eating-​related pathology.
ing such an interview. Semistructured interviews Other interviews for assessing eating-​ related
are more commonly used in research as opposed pathology for diagnosis and treatment evaluation
to clinical settings (Anderson & Paulosky, 2004a, and planning are available, although none are as
2004b). Although less flexible than unstructured popular as the EDE. Reviews of these instruments
interviews, they have more psychometric precision, can be found elsewhere (D. A.  Anderson et  al.,
and many have norms available (Groth-​ Marnat, 2009; Grilo, 2005).
1997; Korotitsch & Nelson-​Gray, 1999). The SCOFF (Morgan et al., 2000; Parker et al.,
The Eating Disorder Examination (EDE; 2005), a simple, easy-​to-​remember instrument for
Fairburn & Cooper, 1993; Fairburn, Cooper, & eating disorders, can be administered orally in a
O’Connor, 2014) is probably the most highly brief interview format. It contains only five items
regarded semistructured interview specific to and is appropriate for use by nonspecialists such as
eating-​related pathology. The EDE has four sub- general medical practitioners. It has been shown to
scales: Restraint, Eating Concern, Weight Concern, have good psychometric properties (Siervo, Boschi,
and Shape Concern. The most recent version, the Papa, Bellini, & Falconi, 2005), although recent
EDE 17.0D (Fairburn et al., 2014) is a revision of work suggests the SCOFF may have lower sensitivity
the widely used EDE 16.0D (Fairburn & Cooper, in older, ethnically diverse samples (Solmi, Hatch,
1993). The 17.0D has been updated in accordance Hotopf, Treasure, & Micali, 2015). Furthermore,
with DSM-​5 diagnostic criteria; an item assessing there is some disagreement about the optimal cutoff

216 Psycholo gical Assessment

score to indicate probable eating pathology (Siervo (i.e., the EDE and Structured Clinical Interview
et al., 2005). for DSM-​IV Axis 1 Diagnoses (SCID; Stice et al.,
The ESP (Cotton, Ball, & Robinson, 2003) is 2000b; Stice, Fisher, & Martinez, 2004), and thus
a four-​item instrument developed for eating disor- may be a good alternative where time or training
der screening purposes that can be delivered orally. considerations preclude the use of an interview.
Like the SCOFF, it has good psychometric proper-
ties and can be delivered by nonspecialists (Cotton Treatment Planning and Evaluation
et al., 2003). There are a number of self-​ report measures
that can be used throughout the therapeutic pro-
Self-​Report Questionnaires cess, from initial assessment to evaluation of treat-
Self-​report questionnaires can be used for every ment outcome. Detailed reviews of many of these
assessment-​ related function, from screening to measures can be found elsewhere (Allison, 2009,
diagnosis to treatment planning and evaluation. It Peterson & Mitchell, 2005); we will focus on some
is beyond the scope of this chapter to review all the of the more widely used measures in the literature.
available self-​report questionnaires for eating dis- In addition to being used as a screening measure,
orders; readers should consult Allison (2009) and the EAT (Garner & Garfinkel, 1979; Garner et al.,
Peterson and Mitchell (2005) for more detailed 1982) can also be used repeatedly to track treatment
discussions of a number of well-​validated instru- progress (Garner, 1997). The EDDS has also been
ments. We do, however, provide some examples of used to measure changes in treatment, although in
measures that are particularly suited to each assess- one study it was less sensitive to change than the
ment function, with a focus on measures specifically EDE (Stice et al., 2004).
designed to measure eating-​related pathology. The Eating Disorders Inventory-​ 3 (Garner,
2004), is the latest revision of this questionnaire
Screening (Garner, 1991; Garner, Olmsted, & Polivy, 1983),
For screening, the SCOFF (Morgan et al., 2000; previous versions of which have been widely used
Parker et al., 2005) has been administered in writ- in both clinical and research contexts (Anderson &
ten format. Also, the Eating Attitudes Test (EAT), Paulosky, 2004a, 2004b). It contains three eating-​
one of the most commonly used self-​report ques- disorder-​specific scales and nine psychological scales
tionnaires for eating-​related pathology (Anderson & that assess psychopathology common in eating dis-
Paulosky, 2004a, 2004b), was designed to be used as order patients. It also has six composite scores; one
a screening measure. It is available in 40-​item (EAT-​ that is eating disorder specific and five that are gen-
40; Garner & Garfinkel, 1979) and 26-​item (EAT-​ eral integrative psychological constructs, as well as
26; Garner, Olmsted, Bohr, & Garfinkel, 1982) three response style indicators. Its previous editions
versions. Total scores above 30 on the EAT-​40 and have shown good psychometric properties, and this
20 on the EAT-​26 indicate probable eating-​related latest revision appears to be no exception (Anderson
pathology (Garfinkel & Newman, 2001). Both ver- et al., 2009; Garner, 2004).
sions have good psychometric properties (Anderson The Multifactorial Assessment of Eating
et al., 2009; Garfinkel & Newman, 2001). Disorder Symptoms scale (MAEDS; Anderson,
Williamson, Duchmann, Gleaves, & Barbin, 1999)
Diagnosis was developed as a self-​report inventory to assess
One self-​ report questionnaire, the Eating domains of eating disorder symptoms necessary
Disorder Diagnostic Scale (EDDS; Stice, Telch, & for successful treatment (depression, binge eating,
Rizvi, 2000a, 2000b) was developed specifically to purgative behavior, fear of fatness, restrictive eating,
diagnose anorexia nervosa, bulimia nervosa, and and avoidance of forbidden foods). It was developed
binge eating disorder according to DSM-​IV (APA, specifically as a treatment outcome measure and has
1994) criteria. It does not use the term “binge,” been shown to have good psychometric properties
which minimizes any problems with idiosyncratic (Anderson et al., 2009; Anderson et al., 1999).
definitions respondents might have for this term. Its The Eating Disorder Examination Questionnaire
scoring has been adapted for use with the DSM-​ (EDE-​Q; Fairburn & Beglin, 1994, 2008) is a self-​
5 diagnostic criteria (Bohon & Stice, 2015) and it report questionnaire derived from the EDE. As
has demonstrated good psychometric properties expected, individuals tend to score higher on the
(Krabbenborg et al., 2012) and very good agreement EDE-​Q than the EDE (Anderson et  al., 2009).
with longer, more involved diagnostic interviews Because the items on the EDE-​Q are almost identical

Anderson, Donahue, EHrlich, Gorrell 217

to the EDE and it generates the same subscales as commonly practiced. For example, a review of con-
the EDE, it can be useful for treatment planning trolled treatment studies of CBT for BN found that
and evaluation, particularly if the assessor or treat- only three of the 16 studies reviewed assessed all five
ment team is familiar with EDE terminology. of the core domains identified in the CBT model
of the maintenance of that disorder (Anderson &
Self-​Monitoring Maloney, 2001). Improving assessment procedures
Self-​
monitoring, broadly defined, is one of can improve all phases of treatment, from identi-
the most widely used assessment procedures in fication to determining long-​term outcome, how-
the assessment of eating pathology (Anderson & ever, and we hope that this chapter will spur interest
Paulosky, 2004a). Direct aspects of food intake in developing individualized assessment batter-
and eating pathology (e.g., the amount of food ies based on both practical needs and theoretical
eaten and whether a purge followed) as well as the considerations.
larger context surrounding each eating episode
(e.g., mood, location) can be assessed in this way. References
Allison, D. B. (2009). Handbook of assessment methods for eat-
While examples of standardized forms are available
ing behaviors and weight-​related problems (2nd ed.). Newbury
(e.g., Schlundt, 1995; Williamson, 1990), self-​ Park, CA: Sage.
monitoring is a very flexible procedure and can be American Psychiatric Association (APA). (1994). Diagnostic and
customized to fit the specific needs of the assessor statistical manual of mental disorders (4th ed.) Washington,
and treatment team. Self-​monitoring can be used at DC: Author.
American Psychiatric Association (APA). (2013). Diagnostic and
any stage of the assessment process, but is particu-
statistical manual of mental disorder (5th ed.). Washington,
larly useful for tracking treatment progress; many DC: Author.
therapies routinely have patients self-​monitor their Andersen, A. E. (1995). A standard test meal to assess treat-
food intake (e.g., Fairburn, 2008; Garner, 1997), so ment response in anorexia nervosa patients. Eating
these data are often readily available. Disorders: Journal of Treatment and Prevention, 3, 47–​55.
Anderson, D. A., De Young, K. P., & Walker, D. C. (2009).
Assessment of eating disordered thoughts, feelings, and
Test Meals behaviors. In D. B. Allison (Ed.), Handbook of assessment
One shortcoming with both interview and ques- methods for eating behaviors and weight-​related problems (2nd
tionnaire modalities is that they both rely on accu- ed.). Newbury Park, CA: Sage.
rate responding from the individual being assessed. Anderson, D. A., & Maloney, K. C. (2001). The efficacy of
cognitive-​behavioral therapy on the core symptoms of buli-
Denial and minimization are significant problems in
mia nervosa. Clinical Psychology Review, 21, 971–​988.
the eating disorders (Anderson & Paulosky, 2004a), Anderson, D. A., & Paulosky, C. A. (2004a). Psychological
and it is relatively easy for individuals to minimize assessment of eating disorders and related features. In J. K.
pathology on self-​ report assessment instruments. Thompson (Ed.), Handbook of eating disorders and obesity
Test meals require that an individual actually eat (pp. 112–​129). New York, NY: John Wiley & Sons.
Anderson, D. A., & Paulosky, C. A. (2004b). A survey of the
a food or meal under controlled conditions; the
use of assessment instruments by eating disorder profes-
amount eaten as well as the behavior of the individ- sionals in clinical practice. Eating and Weight Disorders, 9,
ual being assessed can be directly measured. Some 238–​241.
have suggested that test meals can be extremely use- Anderson, D. A., Simmons, A. M., Milnes, S., M., & Earleywine,
ful in all stages of the assessment process (Andersen, M. (2007). The effect of response format on endorsement
of eating disordered attitudes and behaviors. International
1995; Anderson & Paulosky, 2004a; Williamson,
Journal of Eating Disorders, 40, 90–​93.
1990), although they are probably most appropri- Anderson, D. A., Williamson, D. A., Duchmann, E. G., Gleaves,
ate for evaluating treatment progress and outcome. D. H., & Barbin, J. M. (1999). Development and validation
They do not appear to be commonly used, however of a multifactorial treatment outcome measure for eating dis-
(Anderson & Maloney, 2001; Anderson & Paulosky, orders. Assessment, 6, 7–​20.
Becker, D. F., & Grilo, C. M. (2015). Comorbidity of mood
2004b). Nevertheless, test meals represent one way
and substance use disorders in patients with binge-​eating
to overcome some of the problems associated with disorder:  Associations with personality disorder and eating
self-​reported food intake. disorder pathology. Journal of Psychosomatic Research, 79,
159–​164.
Conclusions Beglin, S. J., & Fairburn, C. G. (1992). What is meant by the
term “binge”? American Journal of Psychiatry, 149, 123–​124.
Although the theory-​based approach outlined
Berg, K. C., Peterson, C. B., Frazier, P., & Crow, S. J. (2011).
in this chapter represents a comprehensive and sys- Convergence of scores on the interview and questionnaire
tematic approach to the assessment of eating disor- versions of the Eating Disorder Examination:  A  meta-​
ders and related problems, it does not appear to be analytic review. Psychological Assessment, 23, 714.

218 Psycholo gical Assessment

Berg, K. C., Peterson, C. B., Frazier, P., & Crow, S. J. (2012). Fairburn, C. G., Marcus, M. D., & Wilson, G. T. (1993).
Psychometric evaluation of the eating disorder examina- Cognitive-​ behavioral therapy for binge eating and buli-
tion and eating disorder examination‐questionnaire: A sys- mia nervosa:  A  comprehensive treatment manual. In:  C.
tematic review of the literature. International Journal of G. Fairburn, & G. T. Wilson (Eds.), Binge eating:  Nature,
Eating Disorders, 45, 428–​438. assessment, and treatment (pp. 361–​404). New  York,
Bohon, C., & Stice, E. (2015). Eating Disorder Diagnostic Scale. NY: Guilford Press.
In T. Wade (Ed.). Encyclopedia of feeding and eating disorders Fichter, M. M., Quadflieg, N., & Hedlund, S. (2006). Twelve-​
(pp. 1–​4). Singapore: Springer. year course and outcome predictors of anorexia nervosa.
Cassin, S. E., & von Ranson, K. M. (2005). Personality and eat- International Journal of Eating Disorders, 39, 87–​100.
ing disorders: A decade in review. Clinical Psychology Review, Fichter, M. M., Quadflieg, N., & Hedlund, S. (2008). Long-​
25, 895–​916. term course of binge eating disorder and bulimia ner-
Cotton, M., Ball, C., & Robinson, P. (2003). Four simple ques- vosa:  Relevance for nosology and diagnostic criteria.
tions can help screen for eating disorders. Journal of General International Journal of Eating Disorders, 41, 577–​586.
Internal Medicine, 18, 53–​56. Follette, W. C., & Callaghan, G. M. (2001). The evolution of
Crowther, J. H., & Sherwood, N. E. (1997). Assessment. In clinical significance. Clinical Psychology: Science and Practice,
D. M. Garner & P. E. Garfinkel (Eds.), Handbook of treat- 8, 431–​435.
ment for eating disorders (2nd ed., pp. 34–​49). New  York, French, S. A., Peterson, C. B., Story, M., Anderson, N., Mussell,
NY: Guilford Press. M. P., & Mitchell, J. E. (1998). Agreement between survey
Duarte, C., Ferreira, C., Trindade, I. A., & Pinto-​Gouveia, J. and interview measures of weight control practices in adoles-
(2016). Normative body dissatisfaction and eating psychopa- cents. International Journal of Eating Disorders, 23, 45–​56.
thology in teenage girls: The impact of inflexible eating rules. Garfinkel, P., & Newman, A. (2001). The Eating Attitudes
Eating and Weight Disorders-​Studies on Anorexia, Bulimia and Test:  Twenty-​five years later. Eating and Weight Disorders,
Obesity, 21(1), 41–​48. 6, 1–​24.
Eisman, E., Dies, R., Finn, S. E., Eyde, L. D., Kay, G. G., Garner, D. M. (1991). Eating Disorder Inventory-​2 professional
Kubiszyn, T. W.,  . . .  Moreland, K. L. (2000). Problems and manual. Odessa, FL: Psychological Assessment Resources.
limitations in the use of psychological assessment in contem- Garner, D. M. (1997). Psychoeducational principles in treat-
porary health care delivery. Professional Psychology:  Research ment. In D. M. Garner & P. E. Garfinkel (Eds.), Handbook
and Practice, 31, 131–​140. of treatment for eating disorders (2nd ed., pp. 145–​177).
Evans, M. I., Galen, R. S., & Britt, D. W. (2005). Principles of New York, NY: Guilford Press.
screening. Seminars in Perinatology, 29, 364–​366. Garner, D. M. (2004). EDI-​3 Eating Disorder Inventory-​3 pro-
Fairburn, C. G. (1997a). Eating disorders. In D. M. Clark & fessional manual. Odessa, FL:  Psychological Assessment
C. G. Fairburn (Eds.), The science and practice of cognitive Resources.
behaviour therapy (pp. 209–​242). Oxford, UK:  Oxford Garner, D. M., & Garfinkel, P. E. (1979). The Eating Attitudes
University Press. Test:  An index of the symptoms of anorexia nervosa.
Fairburn, C. G. (1997b). Interpersonal psychotherapy for buli- Psychological Medicine, 9, 273–​279.
mia nervosa. In D. M. Garner & P. E. Garfinkel (Eds.), Garner, D. M., Olmsted, M. P., Bohr, Y., & Garfinkel, P. E.
Handbook of treatment for eating disorders. (2nd ed., pp. 278–​ (1982). The eating attitudes test: Psychometric features and
294). New York, NY: Guilford Press. clinical correlates. Psychological Medicine, 12, 871–​878.
Fairburn, C. G. (2008). Cognitive behavior therapy and eating dis- Garner, D. M., Olmstead, M. P., & Polivy, J. (1983). Development
orders. New York, NY: Guilford Press. and validation of a multidimensional eating disorder inven-
Fairburn C. G., & Beglin, S. J. (1990). Studies of the epidemiol- tory of anorexia nervosa and bulimia. International Journal of
ogy of bulimia nervosa. American Journal of Psychiatry, 147, Eating Disorders, 2, 14–​34.
401–​408. Grilo, C. G. (2005). Structured instruments. In J. E. Mitchell
Fairburn C. G., & Beglin, S. J. (1994). Assessment of eating dis- & C. B. Peterson (Eds.), Assessment of eating disorders (pp.
orders:  Interview or self-​report questionnaire? International 79–​97). New York, NY: Guilford Press.
Journal of Eating Disorders, 16, 363–​370. Grimes, D. A., & Schulz, K. F. (2002). Uses and abuses of screen-
Fairburn C. G., & Beglin, S. (2008). The eating disorder exami- ing tests. Lancet, 359, 881–​884.
nation questionnaire 16.0. In C. G. Fairburn, Cognitive Groth-​Marnat, G. (1997). Handbook of psychological assessment
behavior therapy and eating disorders (pp. 309–​314). (3rd ed.). New York, NY: John Wiley & Sons.
New York, NY: Guilford Press. Hamburg, P., Herzog, D. B., Brotman, A. W., & Stasior, J. K.
Fairburn, C. G., & Cooper, Z. (1993). The eating disorder (1989). The treatment of resistant eating disordered patient.
examination. In C. G. Fairburn & G. T. Wilson (Eds.), Binge Psychiatric Annals, 19, 494–​499.
eating:  Nature, assessment, and treatment (pp. 317–​360). Hayaki. J., Friedman, M. A., & Brownell, K. D. (2002).
New York, NY: Guilford Press. Shame and severity of bulimic symptoms. Eating Behaviors,
Fairburn, C. G., Cooper, Z., & O’Connor, M. (2014). The 3, 73–​83.
eating disorder examination 17.0D. In C. G. Fairburn, Heffner, M., & Eifert, G. (2004). The anorexia workbook: How
Cognitive behavior therapy and eating disorders (pp. 265–​ to accept yourself, heal your suffering, and reclaim your life.
308). New York, NY: Guilford Press. Oakland, CA: New Harbinger.
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive Jacobson, N. S., Follette, W. C., & Revenstorf, D. (1984).
behaviour therapy for eating disorders:  A  “transdiagnostic” Psychotherapy outcome research: Methods for reporting var-
theory and treatment. Behaviour Research and Therapy, 41, iability and evaluating clinical significance. Behavior Therapy,
509–​528. 15, 336–​352.

Anderson, Donahue, EHrlich, Gorrell 219

Jacobson, N. S., Roberts, L. J., Berns, S. B., & McGlinchey J. Nielsen, C., & Lang, R. S. (1999). Principles of screening.
B. (1999). Methods for defining and determining the clini- Medical Clinics of North America, 83, 1323–​1337.
cal significance of treatment effects:  Description, applica- O’Brien, K. M., & Vincent, N. K. (2003). Psychiatric comorbid-
tion, and alternatives. Journal of Consulting and Clinical ity in anorexia and bulimia nervosa: Nature, prevalence, and
Psychology, 67, 300–​307. causal relationships. Clinical Psychology Review, 23, 57–​74.
Jacobson, N. S., & Truax, P. (1991). Clinical significance: A sta- Pachankis, J. E., & Goldfried, M. R. (2007). An integra-
tistical approach to defining meaningful change in psy- tive, principle-​based approach to psychotherapy. In S. G.
chotherapy research. Journal of Consulting and Clinical Hoffman & J. Weinberger (Eds.), The art and science of psy-
Psychology, 59, 12–​19. chotherapy (pp. 49–​68). New York, NY: Routledge.
Johnson, W. G., Boutelle, K. N., Torgrud, L., Davig, J. P., & Parker, S. C., Lyons, J., & Bonner, J. (2005). Eating disorders
Turner, S. (2000). What is a binge? The influence of amount, in graduate students: Exploring the SCOFF questionnaire as
duration, and loss of control criteria on judgments of a simple screening tool. Journal of American College Health,
binge eating. International Journal of Eating Disorders, 27, 54, 103–​107.
471–​479. Perry, L., Morgan, J., Reid, F., Brunton, J., O’Brien, A., Luck,
Kanfer, F. H., & Schefft, B. K. (1988). Guiding the process of A., & Lacey, H. (2002). Screening for symptoms of eating
therapeutic change. Champaign, IL: Research Press. disorders: Reliability of the SCOFF screening tool with writ-
Keel, P. K., Crow, S., Davis, T. L., & Mitchell, J. E. (2002). ten compared to oral delivery. International Journal of Eating
Assessment of eating disorders:  Comparison of interview Disorders, 32, 466–​472.
and questionnaire data from a long-​term follow-​up study Peterson, C. B. (2005). Conducting the diagnostic interview. In
of bulimia nervosa. Journal of Psychosomatic Research, 53, J. E. Mitchell & C. B. Peterson (Eds.), Assessment of eating
1043–​1047. disorders (pp. 32–​58). New York, NY: Guilford Press.
Keel, P. K., Dorer, D. J., Eddy, K. T., Franko, D., Charatan, D. Peterson, C. B., Becker, C. B., Treasure, J., Shafran, R., & Bryant-​
L., & Herzog, D. B. (2003). Predictors of mortality in eating Waugh, R. (2016). The three-​legged stool of evidence-​based
disorders. Archives of General Psychiatry, 60, 179–​183. practice in eating disorder treatment: research, clinical, and
Kendall, P. C., & Grove, W. M. (1988). Normative comparisons patient perspectives. BMC Medicine, 14(1), 1.
in therapy outcome. Behavioral Assessment, 10, 147–​158. Peterson, C. B., & Mitchell, J. E. (2005). Self-​report measures.
Kendall, P. C., Marrs-​Garcia, A., Nath, S. R., & Sheldrick, R. C. In J. E. Mitchell & C. B. Peterson (Eds.), Assessment of eating
(1999). Normative comparisons for the evaluation of clini- disorders (pp. 98–​119). New York, NY: Guilford Press.
cal significance. Journal of Consulting and Clinical Psychology, Rodin, J., Silberstein, L. R., & Striegel-​Moore, R. H. (1985). Women
67, 285–​299. and weight:  A  normative discontent. In T. B. Sonderegger
Korotitsch, W., & Nelson-​Gray, R. O. (1999). Self-​report and (Ed.), Psychology and gender: Nebraska symposium on motivation
physiological measures. In S. C. Hayes, D. H. Barlow, & R. (pp. 267–​307). Lincoln: University of Nebraska Press.
O. Nelson-​ Gray (Eds.), The scientist-​practitioner:  Research Schlundt, D. G. (1995). Assessment of specific eating behaviors
and accountability in the age of managed care (pp. 320–​352). and eating style. In D. B. Allison (Ed.), Methods for the assess-
Boston, MA: Allyn and Bacon. ment of eating behaviors and weight-​related problems (pp. 142–​
Krabbenborg, M. A., Danner, U. N., Larsen, J. K., Veer, N., 302). Newbury Park, CA: Sage.
Elburg, A. A., Ridder, D. T.,  . . .  Engels, R. C. (2012). The Sheldrick, R. C., Kendall, P. C., & Heimberg, R. G. (2001).
Eating Disorder Diagnostic Scale:  Psychometric features The clinical significance of treatments:  A  comparison of
within a clinical population and a cut‐off point to differenti- three treatments for conduct disordered children. Clinical
ate clinical patients from healthy controls. European Eating Psychology, 8, 418–​430.
Disorders Review, 20, 315–​320. Siervo, M., Boschi, V., Papa, A., Bellini, O., & Falconi, C.
Lavender, J. M., & Anderson, D. A. (2008). A novel assessment (2005). Application of the SCOFF, Eating Attitudes Test
of behaviors associated with body dissatisfaction and disor- 26 (EAT 26) and Eating Inventory (TFEQ) Questionnaires
dered eating. Body Image, 5, 399–​403. in young women seeking diet-​therapy. Eating and Weight
Lavender, J. M., & Anderson, D. A. (2009). Effect of perceived Disorders, 10, 76–​82.
anonymity in assessments of eating disordered behaviors and Sihvola, E., Keski-​Rahkonen, A., Dick, D. M., Hoek, H. W.,
attitudes. International Journal of Eating Disorders, 42(6), Raevuori, A., Rose, R. J.,  . . .  Kaprio, J. (2009). Prospective
546–​551. associations of early-​onset Axis I disorders with developing
Luck, A. J., Morgan, J. F., Reid, F., O’Brien, A., Brunton, J., eating disorders. Comprehensive Psychiatry, 50, 20–​25.
Price, C.,  . . .  Lacey, J. H. (2002). The SCOFF question- Solmi, F., Hatch, S. L., Hotopf, M., Treasure, J., & Micali, N.
naire and clinical interview for eating disorders in general (2015). Validation of the SCOFF questionnaire for eat-
practice:  Comparative study. British Medical Journal, 325, ing disorders in a multiethnic general population sample.
755–​756. International Journal of Eating Disorders, 48, 312–​316.
Lundgren, J. D., Danoff–​Burg, S., & Anderson, D. A. (2004). Stice, E., Fisher, M., & Martinez, M. (2004). Eating Disorder
Cognitive-​behavioral therapy for bulimia nervosa: An empir- Diagnostic Scale: Additional evidence of reliability and valid-
ical analysis of clinical significance. International Journal of ity. Psychological Assessment, 16, 60–​71.
Eating Disorders, 35, 262–​274. Stice, E., Telch, C. F., & Rizvi, S. L. (2000a). Correction to Stice
Mitchell, J. E., & Peterson, C. B. (2005). Assessment of eating et al. (2000). Psychological Assessment, 12, 252.
disorders. New York, NY: Guilford Press. Stice, E., Telch, C. F., & Rizvi, S. L. (2000b). Development and
Morgan, J. F., Reid, F., & Lacey, J. H. (2000). The SCOFF ques- validation of the Eating Disorder Diagnostic Scale: A brief
tionnaire: A new screening tool for eating disorders. Western self-​report measure of anorexia, bulimia, and binge-​eating
Journal of Medicine, 172, 164–​165. disorder. Psychological Assessment, 12, 123–​131.

220 Psycholo gical Assessment

Telch, C. F., Pratt, E. M., & Niego, S. H. (1998). Obese Williamson, D. A., Anderson, D. A., & Gleaves, D. G. (1996).
women with binge eating disorder define the term binge. Anorexia and bulimia:  Structured interview methodolo-
International Journal of Eating Disorders, 24, 313–​317. gies and psychological assessment. In K. Thompson (Ed.),
Turchik, J. A., Karpenko, V., Hammers, D., & McNamara, J. Body image, eating disorders, and obesity: An integrative guide
R. (2007). Practical and ethical assessment issues in rural, for assessment and treatment (pp. 205–​223). Washington,
impoverished, and managed care settings. Professional DC: American Psychological Association.
Psychology, 38, 158–​168. Wise, E. A. (2004). Methods for analyzing psychotherapy out-
Vitousek, K. B., Daly, J., & Heiser, C. (1991). Reconstructing comes: A review of clinical significance, reliable change, and
the internal world of the eating-​ disordered individ- recommendations for future directions. Journal of Personality
ual:  Overcoming denial and distortion in self-​ report. Assessment, 82, 50–​59.
International Journal of Eating Disorders, 10, 647–​666. Wright, C. V., Beattie, S. G., Galper, D. I., Church, A. S., Bufka, L.
Williamson, D. A. (1990). Assessment of eating disor- F., Brabender, V. M., & Smith, B. L. (2016). Assessment prac-
ders:  Obesity, anorexia, and bulimia nervosa. Elmsford, tices of professional psychologists: Results of a national survey.
NY: Pergamon Press. Professional Psychology. http://​dx.doi.org/​10.1037/​pro0000086

Anderson, Donahue, EHrlich, Gorrell 221

 C H A PT E R


Medical Complications of Anorexia Nervosa
11 and Bulimia Nervosa

Philip S. Mehler

Abstract
Medical complications are commonly found in patients with anorexia nervosa and bulimia nervosa. In the
case of anorexia nervosa, the complications are a direct result of weight loss and malnutrition and can
affect every body system. As weight loss becomes more severe and the disorder becomes increasingly
chronic, there is an increased likelihood for the development of medical complications. Bulimia nervosa
also has many medical complications. In general, they relate to the mode and frequency of purging
behaviors, which are inherent to the illness. Many of them are serious and thus contribute to the excess
morbidity and mortality that exists in patients with bulimia nervosa. This chapter will discuss the medical
complications that occur as a consequence of anorexia nervosa including the purging subtype, and
bulimia nervosa.
Key Words:  gastrointestinal, bone marrow, cardiac, osteoporosis, hypoglycemia, hormone

Anorexia Nervosa also have an attenuated blood pressure response

As opposed to all other mental illness, anorexia to exercise with exuberant tachycardia at minimal
nervosa (AN), is unique in that it is inextricably levels of exertion, especially as the disease becomes
associated with many different medical complica- more severe.
tions that are due to weight loss and resultant mal- For a long time there was a sentiment that QT
nutrition. No body system is immune from the prolongation was operative in AN and caused the
progressive medical ravages of AN. As the weight heightened sudden death risk via malignant ven-
loss and malnutrition, which define AN, becomes tricular arrhythmias, including torsades de pointes;
more severe, the medical complications also become this is likely not accurate because there is emerging
more prevalent and significant. evidence that prolonged QT is not inherent to AN
(Krantz et al., 2011). Rather when QT prolongation
Cardiac Complications is noted in a patient with AN, a thorough search
Anorexia nervosa is known to have the high- for electrolyte aberrations or potentially offending
est mortality rate of any psychiatric illness. While medications should ensue (Facchini et  al., 2006).
some of the excess mortality is due to suicide, it is In addition, atherosclerotic heart disease caus-
also substantially attributable to sequelae of med- ing heart attacks is not causing the sudden death
ical complications and most notably sudden car- because patients with AN have not been found to
diac death (Nielsen, 2001). Not unexpectedly with have obstructive coronary artery disease on autopsy
weight loss, there is thinning of the cardiac muscle studies (Arnette et al., 1979). Thus the exact causal
mass and resultant weakened cardiac contractility mechanism of sudden death of AN is not currently
(Lamzabi et al., 2015). These patients therefore may known. Myocardial coronary artery fibrosis and QT
complain of fatigue and dyspnea on exertion. They dispersion abnormalities are ongoing areas of focus.

222
 Other cardiac abnormalities related to the in the abdomen which cushions the SMA and keeps
weight loss and malnutrition include an increased it from moving medially and constricting the lumen
prevalence of pericardial effusions, which has only of the duodenum. The symptoms that are noted
recently been identified and remains heretofore not with its presence are similar to those seen with gas-
well characterized (Kastner et al., 2012). Mitral valve troparesis with the exception that SMA syndrome
prolapse may affect up to 30% of these patients, causes significant epigastric-​type pain 15 or 20
and is reversible with weight gain as supporting minutes after commencing to eat as a result of the
cardiac collagenous structures are reconstituted. small bowel-​type obstruction which ensues as solid
Lastly, bradycardia is the norm for these patients, food progressively is impeded in its passage through
reflecting heightened vagal tone. It may represent the duodenum (Le Moigne et  al., 2010). Once
an energy-​conserving mechanism (Yahalom et  al., again this will correct itself with nutritional reha-
2013). When the heart rate is below 40, inpatient bilitation and restoration of the intestinal fat pad.
admission is recommended to further evaluate for An upper GI series or a CT scan of the abdomen
cardiac stability. Hypotension generally accommo- are the recommended radiology tests to diagnose
dates the bradycardia. Systolic blood pressures less SMA syndrome. Difficulty swallowing may also be
than 100 mmHg are common. present as AN worsens due to weakness in the pha-
ryngeal muscles involved in swallowing (Holmes,
Pulmonary Gudridge, Gaudiani, & Mehler, 2012). As a result
The lungs are relatively protected from the rav- these patients may be at risk for aspiration and may
ages of AN. Recently there have been scattered complain of coughing while eating. A modified bar-
case reports of spontaneous pneumomediastinum ium swallow test is the best way to evaluate a com-
and pneumothorax (Biffl, Narayanan, Gaudiani, & plaint of dysphagia.
Mehler, 2010). Notably it appears that these con- Liver function test abnormalities are often seen
ditions may be difficult to heal and result in pro- as the malnutrition of AN worsens. This is referred
longed air leaks. Thus, caution is in order before to as apoptosis and represents programed cell
insertion of any central intravenous catheter. There death of the hepatocytes as a result of malnutrition
have also been recent concerns raised about the (Narayanan, Gaudiani, Harris, & Mehler, 2010). In
presence of pulmonary function test abnormalities, contrast to alcoholic liver disease which classically
akin to those seen in emphysema, but its natural presents with elevations of aspartate aminotransfer-
history or cause remains unelucidated (Gardini, ase (AST) and alanine aminotransferase (ALT) with
Boni, & Todisco, 2009). AST>ALT, in the apoptosis of AN, ALT is gener-
ally more abnormally elevated than is AST. No
Gastrointestinal specific evaluation of mild transaminase elevations
As a result of weight loss and malnutrition is needed (5–​10 times elevated), as resumption of
there is an overall slowing of gastrointestinal func- food intake should quickly resolve this. There is a
tion. Delayed gastric emptying is frequent in these direct correlation between the severity of these ele-
patients, especially as weight loss becomes more vations, the degree of weight loss and malnutrition
severe. This manifests in complaints of bloating, and the risk for the potentially dangerous entity
early satiety, and nausea (Benini et  al., 2004). referred to as refeeding hypophosphatemia (Brown,
Similarly, acute gastric dilatation, wherein the true Sabel, Gaudiani, & Mehler, 2015). Typically, eleva-
stomach acutely dilates to a dangerously large size tions of bilirubin and alkaline phosphatase are not
with risk of perforation, is found as well (Mascolo, noted with AN.
Dee, Townsend, Brinton, & Mehler, 2015). In a Lastly, the onset of diarrhea, as a patient with
related manner, the entire bowel has slowing, which AN begins to refeed, is another gastrointestinal
causes constipation, bloating, and distention, even complication that can interfere with attempts
in the absence of any previous history of laxative to restore caloric intake. It develops as a result
abuse (Zipfel et al., 2006). Teleologically, this may of small intestinal villous atrophy with resistant
simply represent the body’s attempt to maximize diminution of the normal surface area necessary
its opportunity to derive rare calories from ingested for food absorption. While there certainly can be
food. Superior mesenteric artery (SMA) syndrome independent reasons for the development of diar-
is another gastrointestinal condition, not uncom- rhea, such as infections or causes of malabsorp-
monly seen as weight becomes increasingly lower. tion such as celiac disease, the aforementioned
Normally, in the fed state, there is a fat pad present cause, related to villous atrophy and reduced

Mehler 223
absorptive area, can be diagnosed by demonstrat- levels of sex hormones, both in males and females.
ing elevated levels of blood diamine oxidase levels Therefore male patients are often noted to have low
(Takmoto, Yoshiuchi, Shimodaira, & Akabayashi, testosterone levels and females have low levels of
2014). However, although patients with AN con- estrogen with resultant amenorrhea. There is rever-
veniently contend that they have gluten sensitivity, sion to a prepubertal state with the defect being
a recent study suggests that the incidence of celiac present at the level of the hypothalamus (Miller,
disease is not increased in patients with AN (Kaltsa Grinspoon, & Gleysteen, 2004). Euthyroid sick
et al., 2015). syndrome is commonly found in AN and thus thy-
roxine levels can mislead the clinician into believing
Hematologic a patient with AN also has hypothyroidism and is in
The bone marrow is adversely affected by the need of thyroid hormone replacement.
malnutrition of AN. There is trilinear hypoplasia In the past there was concern that fertility might
with reductions in all three cell lines defined by the be permanently impeded by a past history of AN.
presence of leukopenia, anemia, and thrombocyto- That does not seem to be accurate. Rather a big-
penia, in that order of frequency (Sabel, Gaudiani, ger issue is the false assumption of a lack of need
Statland, & Mehler, 2013). It is due to gelatinous for contraception with the amenorrheic state and
marrow transformation with deposition of a thick then if they do accidentally become pregnant there
gelatinous polysaccharide in the bone marrow, which is a troublesome increased risk for small gestational
replaces the normal fat content of marrow (Muhajir, age births, preterm infants, and miscarriages. Other
2013). The anemia is nonspecific and can be asso- less common and likely less significant endocrine
ciated with normal or macrocytic red cell indices. abnormalities include elevated serum corticoid lev-
Interestingly, the low white blood cell counts do els, abnormalities in antidiuretic hormone (ADH)
not seem to be associated with an increased risk of levels manifesting as diabetes insipidus and hyper-
infections in contrast to other states of neutropenia. natremia, and resistance to growth hormone sug-
Also, falsely high levels of vitamin B-​12 and folate gested by high levels of growth hormone in the
have been described in AN and are a result of the serum with low levels of insulin-​growth factor IGF-​
apoptosis of AN with leakage of stored amounts of 1 (Golden et al., 1994).
these vitamins (Corbetta et al., 2015). There is no
indication for the use of growth factors to stimulate Bone Metabolism
the bone marrow; rather, all cell lines will reconsti- One of the most troubling medical compli-
tute simply with weight restoration. cations of AN is that of reduced bone density
in the form of osteopenia and osteoporosis. The
Neurologic development of this state of reduced bone den-
There are both central and peripheral ner- sity can be rapid and severe notwithstanding the
vous system complications associated with AN. relatively young age of these patients (Faje et al.,
Neuroradiology studies such as MRI, CT, and PET 2014). It adversely affect males and females with
scans have demonstrated the presence of both gray AN (Kraeft, Uppot, & Heffess, 2013; Mehler,
and white matter atrophy. This may be somewhat Sabel, Watson, & Andersen, 2008). The patho-
permanent even with nutritional rehabilitation and physiology of this aggressive rate of loss of bone
can thus adversely diminish cognitive functioning density is based on the uncoupling of the normal
(Roberto et al., 2011). The peripheral nervous sys- dual processes that are responsible for bone health,
tem can also be injured with AN and can manifest namely increased resorption of old bone and
as local neuropraxias due to compression neuropa- decreased formation of new bone. This is in con-
thies as a result of loss of protective subcutaneous trast to the osteoporosis of older women, wherein
fat cushion above the peripheral nerves. Overall, there is mainly increased bone resorption (Miller
muscle strength can also be diminished, both et  al., 2011). This is a second medical complica-
from malnutrition as well as from the progressive tion that can result in a permanent problem mani-
deconditioning that evolves as the patient becomes fested with an increased risk of fragility fractures,
increasingly weak. spinal compression fractures, loss of height, and
a kyphotic posture. The exact pathophysiological
Endocrine cause for this loss of bone mineral is purportedly
Most endocrine processes are impaired in patients multifactorial. Some putative factors involved
with AN. The most common are reductions in include low levels of leptin, high levels of cortisol,

224 Medical Complications

growth hormone resistance, and low levels of sex Normally, there is a “gate”-​type mechanism,
hormones to name but a few. It affects trabecular called the lower esophageal sphincter, which when
bone more than cortical bone. healthy, acts as a one-​way valve that permits food to
travel down the esophagus and then into the true
Dermatologic stomach. When self-​induced vomiting is operative
There are a number of distressing, albeit not on a frequent basis, the integrity of this normally
dangerous dermatological complications of AN. intact one-​way valve is breached, which may be
These include brittle nails, thinning hair, cyanotic manifested by painful and frequent gastroesopha-
extremities, pruritus due to loss of subcutaneous tis- geal acid reflux disease and its dyspepsia constella-
sue, the development of decubitus ulcers for a simi- tion of symptoms including heartburn, reflux, and
lar reason, and lanugo hair growth on the face and chest pain (Denholm & Jankowski, 2011). Mallory-​
along the spine (Strumia, 2012). None of these are Weiss tears in the esophagus can occur secondary to
signs of virilization but rather, in part, attempts by forceful vomiting, and complaints of hematemesis
the body to conserve heat to counteract the typical may be noted.
hypothermia of AN. In addition, there are many other local adverse
events that ensue as the frequency self-​ induced
Bulimia Nervosa vomiting increases. These include dysphagia (diffi-
Although the mortality rate associated with culty swallowing) due to scarring and inflammation
bulimia nervosa is much less than that reported for to the oral swallowing mechanism from the acidic
anorexia nervosa, it remains elevated compared to gastric contents that normally are retained within
a control population. Much of this excess mortality the stomach and not in the esophagus. If a patient
rate can be attributed to medical complications that with bulimia is complaining of dysphagia, they
are unique to the mode of purging behavior used should be referred for upper endoscopy to make
by the patients with bulimia nervosa. Most of the sure that Barrett’s esophagus has not developed as
medical complications of bulimia nervosa are local- a result of recurrent acid reflux from a lax sphincter
ized to the body systems that are directly involved and from the inflammatory changes that develop in
in the act of purging (Forney, Buchman-​Schmitt, the esophagus due to the abnormal contact between
Keel, & Frank, 2016), but there remains a danger- the acidic vomitus content and the walls (mucosal
ous set of acid-​base and blood electrolyte aberra- surface) of the esophagus. Barrett’s is a precancerous
tions that are also unique to the mode of purging process within the esophagus that may portend an
abused. Some types of laboratory abnormalities increased risk of esophageal carcinoma (Spechler &
are common in all modes of purging and are the Souza, 2014).
main contributors to the premature death of this As the acid continues backward, up the gastro-
generally young population of afflicted patients. intestinal tract, it then comes into contact with
In this chapter these complications are reviewed in the oral cavity, where it adversely affects the teeth,
reference to each of the two main types of purg- gums, parotid glands, and face. Specifically, den-
ing behaviors, namely self-​induced vomiting and tal erosions result from repeated acid exposure,
stimulant laxative abuse. The lab abnormalities that and these can cause permanent harm to dentition
occur with diuretic medication abuse are very simi- (Uhlen, Tveit, Stenhagen, & Mulic, 2014). There
lar to those seen with self-​induced vomiting. also seems to be an increased risk of dental car-
riers in this population. Furthermore, the parotid
Self-​Induced Vomiting glands hypertrophy in response to excessive vomit-
Self-​induced vomiting is by far the most fre- ing. This condition is referred to as sialadenosis.
quent type of purging behavior seen in patients There is usually bilateral parotid gland enlarge-
with bulimia nervosa. Its complications are broadly ment, which is temporary and gives the appear-
divided into the local harmful effects of this type of ance of “chipmunk” faces. However, its onset
purging behavior along with those seen as a result usually occurs 2–​3  days after the cessation of
of the electrolyte and acid-​base abnormalities that self-​induced vomiting (Coleman, Altini, Nayler,
ensue as a direct result of the excessive vomiting & Richards, 1998). Moreover, epistaxis is seen
(Westmoreland, Krantz, & Mehler, 2016). As the with recurrent forceful vomiting. It should trig-
frequency of purging increases, the chance for a sig- ger inquiry about covert bulimia in a young per-
nificant medical complication also rises in concert son with multiple visits for otherwise unexplained
therein. epistaxis. Lastly, with regard to the face, angular

Mehler 225
chelosis may be present in patients with bulimia colon is distal to the small intestine, wherein most
who have poor hygiene, as a result of the macer- caloric absorption has already occurred. Stimulant
ation of the corners of the mouth from exposure laxative abuse involves the excessive ingestion of
to vomitus. OTC laxative compounds, which contain as their
As previously mentioned, there are unique active ingredient either senna, cascara, or bisacodyl
acid-​base and electrolyte abnormalities that and which act by direct stimulation of the nerve
develop as purging behaviors become frequent. plexus in the colon to activate peristalsis and the
With self-​ induced vomiting, the classic blood generation of diarrheal stools. With ongoing and
work abnormalities are hypokalemia, metabolic excessive abuse of the laxatives, local gastrointesti-
alkalosis, hyponatremia, and hyperamylasemia. nal complications develop, including hematoche-
Hypokalemia can be down to a critically low level zia, hemorrhoids, and even rectal prolapse (Xing &
(Jensen, Brabrand, Vinholt, Hallas, & Lassen, Soffer, 2001). Whether chronic stimulant abuse is
2015). It is due to both loss of potassium in the associated with development of colorectal cancer is
vomitus and also in the urine as result of chroni- a matter of uncertainty (Dukas et al., 2000).
cally elevated, adrenally secreted, serum aldosterone One additional troubling complication seen
levels, in response to the dehydration and hypo- with laxative abuse is the potential development of
tension that develop as a direct result of this purg- a state of dependence on them in order to be able to
ing behavior. This entity is referred to as Pseudo defecate. The theory is that with chronic abuse, the
Bartter’s syndrome and is the cause of a vexing patient needs progressively larger quantities of laxa-
propensity toward edema formation in these tives to cause the same amount of diarrhea. This
patients especially when purging behaviors sud- has been termed the “cathartic colon” syndrome
denly cease (Bahia, Mascolo, Gaudiani, & Mehler, and connotes the development of a state wherein
2012). Edema formation can be made even worse the colon is relegated to being an inert tube, inca-
if necessary intravenous fluids are administered too pable of propagation of stool. This might result in
rapidly to treat symptoms of dehydration (Trent, severe obstipation and perhaps, rarely, the need
Moreira, Colwell, & Mehler, 2013). The com- for a colectomy and the placement of an ostomy
monly found metabolic alkalosis is of the sodium mechanism, due to death of the myenteric plexus
chloride-​responsive type and is directly attribut- in the colon from chronic exposure to these laxa-
able to dehydration from vomiting and said ele- tives. However, it is not totally clear that this syn-
vated aldosterone levels (Mascolo, Trent, Colwell, drome truly exists; but it is clear that a relative state
& Mehler, 2012). Similarly, hyponatremia is most of dangerous dependence can occur with marked
often due to hypovolemia from loss of fluids and usage of stimulant laxatives. Osmotic laxatives,
can be very dangerous once the serum sodium con- in contrast, are relatively safe and should be judi-
centration falls to levels below 126 meq/​L and lower ciously encouraged in place of the stimulant type
(Corona et al., 2016). Of note, the aforementioned (Chu et al., 2012).
abnormal electrolyte pattern is exactly replicated There are also a myriad of electrolyte and acid-​
in those patients with bulimia who abuse diuretics base abnormalities that arise as a result of excessive
as their compensatory purging behavior. But, the abuse of stimulant laxatives. Just as with self-​induced
most severe cases of metabolic alkalosis, with serum vomiting, hypokalemia and hyponatremia are often
bicarbonate levels in excess of 40 meq/​L, are limited present as a direct result of the loss of electro-
to patients who purge via vomiting. Interestingly, lytes in the diarrhea and its resultant dehydration.
the abuse of diuretics is mostly limited to health- However, in contrast to the metabolic alkalosis that
care professionals who have surreptitious access to is frequently seen with self-​induced vomiting as well
potent diuretics, versus those that are available over as diuretic abuse, with laxative abuse the acid-​base
the counter (OTC) and tend to be fairly weak. abnormality is a hyperchloremic metabolic acido-
sis, also known as a non-​gap metabolic acidosis. It
Laxative Abuse is pathognomonic for covert laxative abuse. Also,
Laxative abuse is the second most common mode rarely, some patients with bulimia induce diarrhea
of purging used by patients with bulimia (Koracs & via the inappropriate use of enemas. The complica-
Plamer, 2004). It is worth counseling these patients tions therein are similar to laxative abuse with the
that the abuse of laxatives is quite dangerous, as is additional possibility of dangerous hyperphospha-
described herein, but it is also an ineffective means temia from the use of sodium phosphate-​containing
to lose weight, as the laxatives’ site of action in the enemas (Ori, Rozen-​Zoi, & Chagnac, 2012).

226 Medical Complications

Conclusion movement, laxative use and risk of colorectal cancer among
women. American Journal of Epidemiology, 151, 958–​964.
For AN every body system is at risk for medi-
Facchini, M., Sala, L., Malfatto, G., Bragato, R., Redaelli, G.,
cal complications as a result of weight loss and mal- & Invitti, C. (2006). Low-​K+ dependent QT prolongation
nutrition. The encouraging news is that most of and risk for ventricular arrhythmia in anorexia nervosa.
these are reversible with nutritional rehabilitation if International Journal of Cardiology, 106, 170–​176.
performed in a timely fashion. There may be two Faje, A. T., Fazeli, P. K., Miller, K. K., Katzman, D. K.,
Ebrahimi, S., Lee, H.,  . . .  Klibanski, A. (2014). Fracture
notable exceptions with regard to bone density and
risk and areal bone mineral density in adolescent females
neurocognitive functioning. with anorexia nervosa. International Journal of Eating
Bulimia nervosa, like AN is inextricably tied Disorders, 47, 458–​466.
to many different medical complications. The spe- Forney, K. J., Buchman-​Schmitt, J. M., Keel, P. K., & Frank,
cific complications seen are related to the mode G. K. (2016). The medical complications associated with
purging. International Journal of Eating Disorders, 49,
and the frequency of the specific purging behav-
249–​259.
iors used by the patient with bulimia. Once again, Gardini, G. G., Boni, E., & Todisco, P. (2009). Respiratory
effective and definitive treatment of the eating dis- function in patients with anorexia nervosa. Chest, 136,
order is the primary focus to prevent these com- 1356–​1360.
plications that can cause serious morbidity and Golden, N. H., Kreitzer, P., Jacobson, M. S., Chasalow, F. I.,
Schebendach, J., Freedman, S. M., & Shenker, I. R. (1994).
even mortality issues.
Disturbance in growth hormone secretion and action in
adolescents with anorexia nervosa. Journal of Pediatrics, 125,
References 655–​660.
Arnette, E. N., Isner, J. M., Redwood, D. R., Kent, K. M., Baker, Holmes, S. R., Gudridge, T. A., Gaudiani, J. L., & Mehler, P. S.
W. P., Ackerstein, H., & Roberts, W. C. (1979). Coronary (2012). Dysphagia in severe anorexia nervosa: a case report.
artery narrowing in coronary heart disease: A comparison of International Journal of Eating Disorders, 45, 463–​466.
cineangiographic and necropsy findings. Annals of Internal Jensen, H. K., Brabrand, M., Vinholt, P. J., Hallas, J., & Lassen,
Medicine, 91, 350–​356. A. T. (2015). Hypokalemia in acute medical patients:  risk
Bahia, A., Mascolo, M., Gaudiani, J. L., Mehler, P. S. (2012). factors and prognosis. American Journal of Medicine,
PseudoBartter syndrome in eating disorders. International 128, 60–​67.
Journal of Eating Disorders, 45, 150–​153. Kaltsa, M., Garoufi, A., Tsitsika, A., Tsirogianni, A.,
Benini, L., Todesco, T., Dalle Grave, R., Deiorio, F., Salandini, Papasteriades, C., & Kossiva, L. (2015). Patients with eating
L., & Vantini, I. (2004). Gastric emptying in patients disorders showed no signs of coeliac disease before and after
with restricting and binge/​purging subtypes of anorexia nutritional intervention. Acta Paediatrica, 104, e319–​e323.
nervosa. The American Journal of Gastroenterology, 99, Kastner, S., Salbach-​Andrae, H., Renneberg, B., Pfeiffer, E.,
1448–​1454. Lehmkuhi, U., & Schmitz, L. (2012). Echocardiographic
Biffl, W., Narayanan, V., Gaudiani, J. L., & Mehler, P. S. (2010). findings in adolescents with anorexia nervosa at beginning
The management of pneumothorax in patients with anorexia of treatment and after weight recovery. European Child &
nervosa. Patient Safety in Surgery, 4, 1–​4. Adolescent Psychiatry, 21, 15–​21.
Brown, C. A., Sabel, A. L., Gaudiani, J. L., & Mehler, P. S. Koracs, D., & Plamer, R. L. (2004). The association between
(2015). Predictors of hypophosphatemia during refeeding of laxative abuse and other symptoms among adults with
patients with severe anorexia nervosa. International Journal of anorexia nervosa. International Journal of Eating Disorders,
Eating Disorders. doi: 10.1002/​eat.22406 36, 224–​228.
Chu, E. S., Gaudiani, J. L., Mascolo, M., Statland, B., Sabel, A., Kraeft, J. J., Uppot, R. N., & Heffess, A. M. (2013). Imaging find-
Carroll, K., & Mehler, P. S. (2012). ACUTE center for eating ings in anorexia nervosa. American Journal of Roentgenology,
disorders. Journal of Hospital Medicine, 7, 340–​344. 200, 328–​334.
Coleman, H., Altini, M., Nayler, S., & Richards, A. (1998). Krantz, M. J., Sabel, A., Sagar, U., Long, C. S., Barbey, J. T.,
Sialadenosis: A presenting sign in bulimia. Head & Neck, 20, White, K.,  . . .  Mehler, P. S. (2011). Factors influencing QT
758–​762. prolongation in hospitalized patients with severe anorexia
Corbetta, F., Tremolizzo, L., Conti, E., Ferrarese, C., Neri, F., nervosa. General Hospital Psychiatry, 34, 173–​177.
Bomba, M., & Nacinovich, R. (2015). Paradoxical increase Lamzabi, I., Syed, S., Reddy, V. B., Jain, R., Harbhajanka, A.,
of plasma vitamin B12 and folates with disease severity in & Arunkumar, P. (2015). Myocardial changes in a patient
anorexia nervosa. International Journal of Eating Disorders, with anorexia nervosa: A case report and review of literature.
48, 317–​322. American Journal of Clinical Pathology, 143, 734–​737.
Corona, G., Giuliani, C., Parenti, G., Colombo, G. L., Sforza, Le Moigne, F., Lamboley, J. L., Vitry, T., Stoltz, A., Galoo, E.,
A., Maggi, M.,   .  .  .  Peri, A. (2016). The economic bur- Salamand, P.,  . . .  Farthouat, P. (2010). Superior mesenteric
den of hyponatremia: Systematic review and meta-​analysis. artery syndrome: A rare etiology of upper intestinal obstruc-
American Journal of Medicine, 129, 823–​835. tion in adults. Gastroenterologie Clinique Et Biologique, 34,
Denholm, M., & Jankowski, J. (2011). Gastrointestinal reflux 403–​406.
disease and bulimia nervosa—​ A review of the literature. Mascolo, M., Dee, E., Townsend, R., Brinton, J. T., & Mehler,
Diseases of the Esophagus, 24, 79–​85. P. S. (2015). Severe gastric dilatation due to superior mes-
Dukas, L., Willett, W. C., Colditz, G. A., Fuchs, C. S., Rosner, enteric artery syndrome in anorexia nervosa. International
B., & Giovannucci, E. L. (2000). Prospective study of bowel Journal of Eating Disorders. doi: 10.1002/​eat.22385

Mehler 227
Mascolo, M., Trent, S., Colwell, C., & Mehler, P. S. (2012). tissue volume changes following weight gain in anorexia ner-
What the emergency room department needs to know vosa. International Journal of Eating Disorders, 44, 406–​410.
when caring for patients with eating disorders. International Sabel, A. L., Gaudiani, J. L., Statland, B., & Mehler, P. S. (2013).
Journal of Eating Disorders, 45, 977–​981. Hematological abnormalities in severe anorexia nervosa.
Mehler, P. S., Sabel, A. L., Watson, T., & Andersen, A. E. Annals of Hematology, 92, 605–​613.
(2008). High risk of osteoporosis in male patients with Spechler, S. J., & Souza, R. F. (2014). Barrett’s esophagus. The
eating disorders. International Journal of Eating Disorders, New England Journal of Medicine, 371, 836–​845.
41, 666–​672. Strumia, R. (Ed.). (2012). Eating disorders and the skin.
Miller, K. K., Lee, E. E., Lawson, E. A., Misra, M., Minihan, J., New York, NY: Springer.
Grinspoon, S. K., . . . Klibanski, A. (2011). Determinants of Takmoto, Y., Yoshiuchi, K., Shimodaira, S., & Akabayashi, A.
skeletal loss and recovery in anorexia nervosa. The Journal of (2014). Diamine oxidase activity levels in anorexia nervosa.
Clinical Endocrinology & Metabolism, 91, 2931–​2937. International Journal of Eating Disorders, 47, 203–​205.
Miller, K. K., Grinspoon, S., & Gleysteen, S. (2004). Preservation Trent, S. A., Moreira, M. E., Colwell, C. B., & Mehler, P. S.
of neuroendocrine control of reproductive function despite (2013). ED management of patients with eating disorders.
under nutrition. The Journal of Clinical Endocrinology & American Journal of Emergency Medicine, 31, 859–​865.
Metabolism, 89, 4434–​4438. Uhlen, M. M., Tveit, A. B., Stenhagen, K. R., & Mulic, A.
Muhajir, M. (2013). Gelatinous transformation of bone marrow (2014). Self-​induced vomiting and dental erosion—​A clini-
in a patient with anorexia nervosa. International Journal of cal study. BMC Oral Health, 14, 92.
Hematology, 97, 157–​158. Westmoreland, P., Krantz, M. J., & Mehler, P. S. (2016). Medical
Narayanan, V., Gaudiani, J. L., Harris, R. H., & Mehler, P. complications of anorexia nervosa and bulimia. American
S. (2010). Liver function test abnormalities in anorexia Journal of Medicine, 129, 30–​37.
nervosa—​ Cause or effect. International Journal of Eating Xing, J. H., & Soffer, E. E. (2001). Adverse effects of laxatives.
Disorders, 43, 378–​438. Diseases of the Colon & Rectum, 44, 1201–​1209.
Nielsen, S. (2001). Epidemiology and mortality of eating disor- Yahalom, M., Spitz, M., Sandler, L., Heno, N., Roguin, N.,
ders. Psychiatric Clinics of North America, 24, 201–​214. & Turgeman, Y. (2013). The significance of bradycardia
Ori, Y., Rozen-​Zoi, B., & Chagnac, A. (2012). Fatalities and in anorexia nervosa. International Journal of Angiology,
severe metabolic disorders associated with the use of sodium 22, 83–​94.
phosphate enemas. Archives of Internal Medicine, 172, Zipfel, S., Sammet, I., Rapps, N., Herzog, W., Herpertz, S., &
263–​265. Martens, U. (2006). Gastrointestinal disturbances in eating
Roberto, C. A., Mayer, L. E., Brickman, A.   M., Barnes, A., disorders: A clinical and neurobiological aspects. Autonomic
Muraskin, J., Yeung, L.-​K.,   .  .  .  Walsh, T. (2011). Brain Neuroscience, 129, 99–​106.

228 Medical Complications

 CH A PT E R


Psychological Comorbidities
12 of Eating Disorders

Katherine A. Halmi

Abstract
Psychological comorbidity of eating disorders may be conceptualized in varying facets including
psychiatric diagnosis, specific behaviors, traits, affect regulation, and cognitive characteristics. Although
the Diagnostic and Statistical Manual, fifth edition (DSM-​5) modified some criteria for psychiatric
diagnoses, these modifications should have little effect over the previous rates of DSM-​IV comorbidities
and thus do not necessitate repeat large sample comorbidity studies. This chapter presents facets of
psychological comorbidities of the three major eating disorders: anorexia nervosa (AN), bulimia nervosa
(BN), and binge eating disorder (BED). The most comprehensive comorbid psychiatric diagnosis study
from the US national comorbidity survey replication revealed at least one lifetime comorbid psychiatric
DSM-​IV disorder was present in 56.2% AN, 94.5% BN, and 78.9% BED. Affect regulation, negative
affect, perfectionism, cognitive-​behavioral flexibility, and impulse control are common comorbid features
present in these disorders.
Key Words:  comorbidity, affective, disorder, anxiety, impulse control, personality, substance use,
personality trait, affect regulation, cognitive process

Introduction psychiatric-​ psychological comorbidities presented

This chapter focuses on the psychiatric-​ include psychiatric diagnoses, specific behaviors,
psychological comorbidities of the three major traits, affect regulation, and cognitive characteristics.
eating disorders, anorexia nervosa (AN), bulimia
nervosa (BN), and binge eating disorder (BED), Comorbid Psychiatric Disorders
because the most robust data concern these dis- Affective Disorders
orders. Because there are few robust eating dis- Affective disorders are the most prevalent of the
order comorbidity studies using Diagnostic and comorbid psychiatric disorders associated with eat-
Statistical Manual, fifth edition (DSM-​5), criteria, ing disorders. A study of Asians with anorexia ner-
most of the information presented is from DSM-​ vosa entering an eating disorder clinic in Singapore
IV (American Psychiatric Association [APA], 1994, revealed that depression was the most common
2013). The changes in criteria for AN, BN, and comorbid condition, affecting 25.4% of that sam-
BED from DSM-​IV to DSM-​5 are few, and are ple (Lee, Lee, Pathy, & Chan, 2005). Thus race and
expected to mainly effect an increase in prevalence. ethnicity do not seem to influence the predom-
The amenorrhea criterion was deleted for AN and inance of depression as a comorbid condition of
the frequency of binge eating and compensatory eating disorders. Table 12.1 presents data on Axis
behaviors reduced from twice to once a week for I  comorbid disorders from studies that used valid
3 months for BN and BED. Differences in the vari- structured interviews for DSM-​III-​R or DSM-​IV
ous DSM criteria over the years may produce some diagnoses. Prevalence of lifetime affective disor-
variation in the results of comorbidity studies. The ders across population-​and clinical-​sample-​based

229
studies is fairly similar. One study comparing bipo- compared with AN women without overanxious
lar patients with and without a lifetime history of disorder of childhood. The overanxious disorder
eating disorders found those with eating disorders was more common among the purging anorectics
were heavier, rated more symptomatic on a clinical and anorectics with binge-​purge behavior. Although
global impression severity scale, had a higher num- the overanxious disorder of childhood does not exist
ber of lifetime depressive episodes, and had greater in DSM-​IV, this assessment does suggest that pat-
psychiatric comorbidity excluding eating and mood terns of early onset of anxiety disorders may be a
disorders. This study suggested the eating disor- risk factor for AN. Strober (2004) suggested that
der comorbidity with bipolar disorder created an AN might manifest itself as a heightened sensitivity
increased symptom load and illness burden in the to fear conditioning with resistance to extinction in
bipolar disorder (Wildes et al., 2007). those women who also have anxiety disorders. The
An affective disorder was not shown to influence anxiety symptoms may be related to altered stria-
the response to treatment in BN (Walsh, Hadigan, tal dopamine function (Frank et al., 2005). Those
& Devlin, 1991) or to affect treatment acceptance individuals with a history of overanxious disorder of
or completion in AN (Halmi et al., 2005). childhood, compared with those without, engaged
in more serious weight-​control practices and dis-
Anxiety Disorders played greater body dissatisfaction, higher drive for
The largest number of eating disorder individu- thinness, more eating preoccupation, and a longer
als assessed for the presence of anxiety disorders duration of illness. Likewise, those who had overanx-
was from site study genetic study of eating disor- ious disorder were significantly more likely to have
ders. (Kaye, Bulik, Thornton, Barbarich, & Master, generalized anxiety disorder, OCD, specific phobia,
2004). In this study the eating disorder subtypes had social phobia, and panic disorder. The authors of
similar rates of anxiety disorders, with the excep- this study concluded that although the observations
tion of PTSD, which was approximately three times were based on retrospective recall, they suggest the
greater in those with BN or the combination of importance of early detection of overanxious disor-
anorexia and bulimia compared with the restricting der as a means of averting the later development of
type of AN patients. Two-​thirds of the participants anxiety and eating disorder symptoms.
in this study reported one or more anxiety disorders In the study by Braun, Sunday, and Halmi
in their lifetime with the most common diagnosis (1994), social phobia most commonly preceded the
being obsessive-​compulsive disorder (OCD) (41%) eating disorder in 52% of the patient reports. Most
and social phobia (20%). In the majority of these of these patients also had a history of major depres-
persons the onset of OCD, social phobia, specific sion. In those patients, the social phobia preceded
phobia, or generalized anxiety disorder occurred in the depression.
childhood before the emergence of their eating dis-
order. About 42% of the participants in this study Substance Abuse Disorders
had the onset of one or more of the anxiety disor- The association of substance abuse and eat-
ders in childhood. This figure is substantially higher ing disorders with binge eating has been reported
than the overall anxiety disorder prevalence in child- in many clinical observation studies (Holderness,
hood, which ranges from 4.7% to 17.7% (Costello Brook-​Gunn, & Warren, 1994). Adolescents engag-
& Angold, 1995). A  childhood disorder of OCD ing in binging and purging behavior have higher
occurred in 23% of these participants, compared rates of substance use and greater psychological
with community samples of 2 to 3% (Piacentini & distress than their nonpurging peers (Ross & Ivis,
Bergman, 2000). 1999). Another study found that binge eating in
In another study from the same multinational, adolescents predicted later incidents of substance
multisite, collaborative group, 39% of 249 women use disorders (Weiderman & Pryor, 1996). In this
with a lifetime history of AN reported a history of study about one-​third of the girls with BN were
over anxious disorder of childhood, a DSM-​III-​R engaged in smoking tobacco, using marijuana, and/​
diagnosis. Of these, 94% met criteria for overanx- or drinking alcohol at least weekly. In the bulimic
ious disorder of childhood before the onset of AN. adolescents substance use was also related to other
Those women with both AN and the overanxious impulsive behaviors such as attempted suicide,
disorder self-​ reported more extreme personality stealing, and sexual promiscuity.
traits and attitudes and engaged in more compensa- In contrast to bulimics and binge eaters, restrict-
tory behaviors such as purging (Raney et al., 2008) ing anorectics have low rates of comorbid substance

230 Psycholo gical Comorbidities

Table 12.1  Axis I Comorbid Psychiatric Diagnoses
Authors Comorbid Diagnosis Anorexia Nervosa Bulimia Nervosa Binge Eating Disorder
% % %

Any Mood Disorder 42.1 70.7 46.4

Hudson et al. 1987 70

Halmi et al. 1991 80 92.9

Braun et al. 1994 41 78

Hudson et al. 2007 Major Depressive 39.1 50.1 32.3

Disorder

Hudson et al. 1987 55

Halmi et al. 1991 40 78.6

Braun et al. 1994 32 45

Hudson et al. 2007 Bipolar I–​II 3 17.1 12.5

Disorders

Hudson et al. 1987 11

Halmi et al. 1991 0 7.1

Braun et al. 1994 0.33 13

Hudson et al. 2007 Any Anxiety Disorder 47.9 80.6 65.1

Hudson et al. 1987 43

Halmi 1991 60 50

Braun et al. 1994 41 50

Godart et al. 2002 72 65

Kaye et al. 2004 55 68

Hudson et al. 2007 Panic Disorder 3 16.2 13.2

Halmi et al. 1991 8

Braun et al. 1994 12 10

Godart et al. 2002 7 15

Kaye et al. 2004 9 11

Hudson et al. 2007 Social Phobia 24.8 41.3 31.9

Halmi et al. 1991 33.9

Braun et al. 1994 12 20

Godart et al. 2002 55 59

Kaye et al. 2004 22 16

Hudson et al. 2007 Specific Phobia 26.5 50 37.1

Halmi et al. 1991 12.9

(continued)
Table 12.1  Continued
Authors Comorbid Diagnosis Anorexia Nervosa Bulimia Nervosa Binge Eating Disorder
% % %

Braun et al. 1994 14 5

Godart et al. 2002 34 21

Kaye et al. 2004 14 12

Hudson et al. 2007 Generalized Anxiety 12 45.4 26.3

Disorder

Brewerton et al. 1995 12

Braun et al. 1994 6 0

Godart et al. 2002 24 23

Kaye et al. 2004 13 8

Hudson et al. 2007 Post-​Traumatic Stress 0 17.4 8.2

Disorder

Brewerton et al. 1995 3

Kaye et al. 2004 5 13

Hudson et al. 2007 Obsessive 7.5 3.5 12.2

Compulsive Disorder

Halmi et al. 1991 25.8

Braun et al. 1994 21 17

Godart et al. 2002 21 0

Kaye et al. 2004 35 40

Hudson et al. 2007 Any Substance Use 27 36.8 23.3

Disorder

Halmi et al. 1991 18

Braun et al. 1994 11.8 51.6

Hudson et al. 2007 Alcohol Abuse or 24.5 33.7 21.4

Dependence

Halmi et al. 1991 8

Braun et al. 1994 5.9 41.9

Bulik et al. 2004 16.8 46.1

Drug Abuse or
Dependence

Hudson et al. 2007 17.7 26.0 19.4

Halmi et al. 1991 8

Braun et al. 1994 0 12.5

232 Psycholo gical Comorbidities

Table 12.1  Continued
Authors Comorbid Diagnosis Anorexia Nervosa Bulimia Nervosa Binge Eating Disorder
% % %

Hudson et al. 2007 Impulsive Control 30.8 63.8 43.3

Disorder

Fernandez 2008 1.7—​Restricting 21.8

AN

Hudson et al. 2007 Attention Deficit 16.2 34.9 19.8

Hyperactivity
Disorder

Hudson et al. 2007 Oppositional Defiant 10.5 26.9 18.9

Disorder

Hudson et al. 2007 Conduct Disorder 9.8 26.5 20.0

Axis II Diagnoses with Eating Disorders

Braun et al. 1994 Any Cluster B 0 28

Borderline Personality 0 33
Disorder 0 11
Histrionic 0 20
Antisocial

Any Personality Entire Eating

Disorder Disorder Sample

Powers et al. 1988 77

Wonderlich et al. 1990 62

Gartner et al. 1989 61

Schmidt & Telch 1990 43

Rossiter et al. 1993 33

Steiger et al. 1994 28

Braun et al. 1994 69

Matsunaga 2000 26

Braun et al. 1994 Any Cluster C 29 50

abuse (Stock, Goldberg, Corbett, & Katzman, 2002). eating disorders and alcohol abuse in the same year.
Another study found dieting severity was positively Alcohol involvement was significantly associated
associated with the prevalence, frequency, and inten- with increased prevalence of major depressive dis-
sity of substance abuse (Weiderman & Pryor, 1996). order. Eighty percent of those with major depres-
The study with the largest eating-​disorder sam- sive disorder, compared with 67% of those without
ple size, 672 persons, showed alcohol abuse and it, had an alcohol abuse/​dependence disorder. This
dependence was significantly lower in prevalence study also found that the alcoholism was associated
in individuals with AN compared with those who with increased risk of OCD, a variety of anxiety dis-
had BN (Bulik et  al., 2004; see Table 12.1). Of orders, and Cluster B diagnoses, particularly bor-
those 253 persons with alcohol abuse/​dependence, derline personality disorder.
32% had the onset of alcoholism prior to the eating An earlier study (Braun et al., 1994) found that
disorder. Only 9% experienced the onset of both the AN restricting subgroup was significantly less

Halmi 233
likely to be alcohol or drug dependent compared 0.3%, and pyromania 0.3%. In the entire sample
with the bulimic subtypes. A comparison of age of of 118 individuals who had both eating disorder
onset showed 57% developed the eating disorder and ICD, the ICDs were present in only one par-
first and 27% developed alcohol dependence first. ticipant with restricting type of AN and two with
It is of interest to note that it is more typical for purging AN. Thus, of the 118 cases of eating dis-
the onset of eating disorders to precede the onset of order and ICD, all but three were associated with
alcohol abuse/​dependence disorders and more likely eating disorder subtypes that included binge eating.
for anxiety disorders to predate the onset of eating This study also demonstrated the presence of ICDs
disorders. in eating disorders was significantly associated with
greater severity of eating disorder, reflected in the
Impulse Control Disorders use of maladaptive compensatory behaviors such as
The DSM-​ 5 section “Disruptive, Impulse laxatives, diuretics, appetite suppressants, and fast-
Control, and Conduct Disorders” includes all ing. In addition there was a greater general psychi-
the diagnoses listed under the DSM-​IV section “ atric comorbidity and psychopathology including
Impulse Control Disorders,” with little change in depression and anxiety disorders, Cluster B person-
criteria. The former emphasizes, “these problems ality disorder, avoidant personality disorder, and
are manifested in behaviors that violate the rights specific personality traits such as higher impulsivity,
of others.” The three large and adequate sample size harm avoidance, neuroticism, cognitive impulsivity
studies presented here include the many diagnoses and lower self-​directedness in those who had ICD
common to both classifications. as well as their eating disorder compared with indi-
In DSM-​IV (APA, 1994) the impulse control viduals who had an eating disorder without ICD.
disorders (ICDs) are classified as pathological gam- It is of interest to note that in the above studies
bling, kleptomania, intermittent explosive disorder, 62% of the ICD occurred before the onset of the
trichotillomania, pyromania, and ICD not other- eating disorder and 40% experienced the onset of
wise specified, which includes compulsive Internet both disorders within the same 3-​year window. It
use, compulsive sexual behavior, and compulsive is likely differential personality traits exist among
buying. The ICDs are characterized by repetitive the various ICDs. For example, in this sample of
occurrence of impulsive behavior, which includes eating disorders with ICD, those individuals were
core features of compulsive engagement in a behav- three times more likely to have comorbid OCD
ior despite adverse consequences; failure to resist compared with eating disorder individuals without
the impulse, urge, or craving state before engage- ICD. Pathological gambling one might expect to be
ment in the impulsive act; and a sense of pleasure associated with higher novelty seeking. The authors
and gratification or release at the time the behavior suggest that problems with removing unwanted
is committed. Other than a few case studies, there thoughts and impaired decision-​ making may be
only three large sample size studies examining ICD a link between OCD and ICD and may partially
in eating disorders. In a study by Fernandez-​Aranda explain the association of ICD and OCD in this
et al. (2006), the prevalence of lifetime ICD in 227 eating disorder sample. Those eating disorder indi-
bulimia patients was 23.8%, with compulsive buy- viduals with ICD were also three times more likely
ing and intermittent explosive disorder as the most to have comorbid borderline personality disorder.
frequently reported ICDs. Those individuals with These findings led the authors to suggest that a sub-
BN and lifetime ICD have more extreme personal- type of BN exists whose development in association
ity profiles, especially on novelty seeking and impul- with ICD, affective disturbance, substance use, and
sivity and greater general psychopathology than personality disturbance may be an expression of
those with BN without ICD. An analysis of data genetic variance that predisposes to high levels of
from the multisite, international Price Foundation disinhibition and impulsivity.
Genetic Studies of Eating Disorders by Fernandez-​ The publication of prevalence and correlates of
Aranda et al. (2008) showed a lifetime prevalence of eating disorders in the national comorbidity survey
all ICDs in this entire sample was 16.6%. The most replication (Hudson et al., 2007) showed an unusu-
common diagnosis was compulsive buying disorder ally high prevalence of total ICDs in AN 30.8%,
11.8%, followed by kleptomania 4.5%, with 17 BN 63.8%, and BED 43.3%.
participants having both diagnoses. The remaining Further analyses of these ICDs were not pre-
diagnoses were trichotillomania 1.85%, intermit- sented in that publication. It is likely that ICDs
tent explosive disorder 0.6%, compulsive gambling are underdiagnosed by clinicians who might

234 Psycholo gical Comorbidities

acknowledge the behaviors but not regard them as a compared with the anorectic restrictors. Percentages
specific diagnosis. of DSM-​ III-​
R bulimics who had at least one
personality disorder ranged from 28% (Steiger,
Personality Disorders Thibaudeau, Leung, Houle, & Ghadirian, 1994) to
The DSM-​5 identifies the same three clusters 77% (Powers, Coovert, Brikwell, & Stevens, 1988).
of personality disorders present in DSM-​IV. There The following intermediate percentages were studies
is little change in criteria. Presented here are the by S. Wonderlich, Swift, Slotnick, and Goodman
adequate large sample size studies conducted with (1990) 52%; Gartner, Marcus, Halmi, and Loranger
validated structured interviews in the past two (1989) 61%; Schmidt and Telch (1990) 43%; and
decades. According to DSM-​IV (APA, 1994) per- Rossiter, Agras, Telch, and Schneider (1993) 33%.
sonality disorders are differentiated from personal- In a study by Braun et  al. (1994), 69% of the
ity traits in that the latter are enduring patterns of patients had at least one personality disorder and of
perceiving the environment and are exhibited in a those 93% had an Axis I  comorbidity. Thirty-​one
wide range of social and personal contexts. When percent of the bulimic subgroups and none of the
these traits are inflexible and maladaptive and cause anorectic restrictors had Cluster B disorders. The
significant functional impairment or subjective most prominent of the latter was borderline per-
distress they then constitute a personality disorder. sonality disorder, which was present in 25% of the
Thus, in DSM-​IV a personality disorder is defined bulimic subgroups and the most common Cluster
as “an enduring pattern of inner experience and B condition. Cluster C personality disorders were
behavior that deviates markedly from expectation present in 29.5% of these patients, and in this cat-
of the individual’s culture, is pervasive and inflex- egory avoidant personality disorder was the most
ible, has an onset in adolescence or early adulthood, common at 14.3%, followed by dependent 10.5%,
is stable over time, and leads to distress or impair- obsessive-​compulsive 6.7%, and passive-​aggressive
ment” (APA, 1994, p. 633). Diagnosis of personal- 4.8%. The prevalence of Cluster C personality dis-
ity disorders is more difficult and less accurate in orders did not vary according to eating disorder
that information is often necessary from informants subtype.
other than the target individual, who may not admit In a study by Hertzog, Keller, Lavori, Kenny,
to problematic behaviors or think they are an issue. and Sack (1992), the most commonly observed
This may explain the observation that many of the personality disorder in 210 patients with eat-
studies of personality disorders and eating disorders ing disorders was borderline personality disorder
are contradictory. Personality disorders are further (9%). Higher rates of borderline personality dis-
grouped into three descriptive clusters: order were found in the bulimic and anorectic-​
bulimic groups, whereas avoidant personality
Cluster A—​individuals often appear odd or
disorder was the most prevalent among the ano-
eccentric, and this includes the paranoid, schizoid,
rectic and anorectic-​bulimic groups. Matsunaga
and schizotypal personality disorders.
et  al. (2000) assessed personality disorders in
Cluster B—​individuals often appear dramatic
patients recovered from eating disorders and
and emotional or erratic, and this includes the
found a prevalence of at least one personality
antisocial, borderline, histrionic, and narcissistic
disorder in 26% with Cluster B associated with
personality disorders.
the bulimic subtypes. When Steiger et al. (2001)
Cluster C—​individuals often appear anxious or
compared bulimics with and without borderline
fearful, and this includes the avoidant, dependent,
personality disorder they found the borderline
and obsessive-​compulsive personality disorders.
showed elevated motor impulsivity, dissociation,
The DSM-​IV classification of personality dis- and higher rates of sexual abuse. In another study
orders is similar to the previous DSM-​III-​R clas- Steiger et  al. (1994) found that personality dis-
sification with the exception that passive-​aggressive order classification did not predict severity of
personality disorder was deleted from DSM-​ IV. bulimic symptoms or responsiveness to treatment.
Many of the large sample structured interview per- In the Braun et al. study (1994), an attempt was
sonality disorder studies of eating disorders were made to assess personality disorders with a more
conducted in the late 1980s to 1990s using DSM-​ sensitive dimensional measure. This was done by
III-​R criteria. Most of those studies showed a high adding the total number of questions to which a
preponderance of Cluster B (impulsive) personal- subject made a threshold response in each of the
ity disorders associated with the bulimic subtype SCID-​II personality disorder modules. A weighted

Halmi 235
personality score was arbitrarily assigned to each Psychiatric comorbidity was assessed in 850
subthreshold response within a particular module. BED cases in the Swedish national register using
A value of 1 was given to a subthreshold response, the combined International Classification of Diseases,
and each threshold response was given a value of ninth revision (ICD-​9) for years 1987–​1996 and
2.  Correlations were then compared across Axis the ICD-​10 for years 1997 to the present (Welch
I  disorders and the weighted personality scores. et  al., 2016). The percent of comorbid frequency
There were significant relationships between major was:  major depressive disorder 23.9%, any anxi-
depression and avoidant personality scores, OCDs ety disorder 17.2%, post-​traumatic stress disorder
and weighted obsessive-​compulsive scores, social 9.4%, bipolar disorder 4.2%, alcohol use disorder
phobia and avoidant personality scores, and affec- 3.2%, and drug use disorder 2.1%.
tive disorder with borderline personality scores.
There were significant correlations between alcohol Personality Traits
abuse and antisocial personality scores as well. The Many broad categories may be encompassed in
dimensional profile suggested that bulimics with a the definition of personality traits as enduring pat-
history of AN and those currently ill AN bulimic terns of perceiving the environment and are exhib-
subtypes may have different personality features ited in a wide range of social and personal contexts
from those patients who have always had exclu- (APA, 2013). This may include thought processes
sively restricting AN or exclusive BN. The groups and behaviors. For example, the research domain
with both eating disorder diagnoses, the bulimics criteria (RDoC) describe five fundamental primary
with a history of AN and the AN bulimic subtypes behavior components (domains):  negative valence
were high in antisocial, borderline, histrionic, systems (NVSs), positive valence systems (PVSs),
and self-​defeating weighted scores. The authors cognitive processes, social processes and arousal/​
hypothesized that anorectic restrictors with more modulation (Insel et al., 2010). Specific studies in
self-​defeating and impulsive personality features eating disorder research often overlap in these cat-
may be at a greater risk of developing bulimia over egories and thus are presented here under the sec-
time (i.e., becoming AN bulimic subtypes or nor- tions “Temperament,” “Affect Regulation,” and
mal weight BN with a history of AN). “Cognitive Processes.”
A prospective study would be helpful to demon-
strate whether specific personality disorder profiles Temperament
actually predispose patients to develop particular Perfectionism is one of the personality features
eating disorders. On the other hand it would be initially identified with AN. Over a century ago,
of interest to know whether the development of Charles Laseque described in these patients an unre-
an eating disorder during the adolescent years may lenting pursuit of unusually rigid standards of pro-
have a formative effect on personality. priety (Laseque, 1873). Perfectionism may predate
the onset of eating disorders (Fairburn, Cooper,
DSM-​5 and ICD-​10 Comorbidity Eating Dell, & Welch, 1999), typify the acute phase of
Disorder Studies eating disorders (Halmi et  al., 2000), and persist
Two large sample size studies were recently pub- after recovery from eating disorders (Kaye et  al.,
lished using DSM-​5 and International Classification 2004; Sutandar-​Pinnock, Blake, Carter, Olmsted,
of Diseases, 10th revision (ICD-​10) diagnostic cri- & Kaplan, 2003). In a international multicenter
teria. Patients with bipolar I (n = 699) and bipolar genetic study of AN, 322 women with a history of
II (n = 393) disorder from the Mayo Clinic Bipolar AN were assessed with the Frost Multidimensional
Biobank were evaluated with the Eating Disorder Perfectionism Scale (Frost & Steketee, 1997). Those
Diagnostic Scale, (EDDS), a validated patient-​rated with AN were distinguished from healthy compar-
questionnaire that generates DSM-​IV and DSM-​5 ison subjects by greater preoccupations and efforts
diagnoses of BED, BN, and AN (McElroy et  al., at avoidance of mistakes in daily life, parental crit-
2016). Twenty-​seven percent of these patients had icism, doubts over the correctness of actions, and
a current DSM-​5 eating disorder diagnosis:  12% more extreme adherence to personal and parental
BED, 15% BN, and 0.2% AN. Rates of DSM-​5 standards of excellence. This study found that AN
compared with DSM-​IV eating disorder diagnoses patients who engaged in purging without binge eat-
were slightly higher for all three major eating disor- ing had higher parental criticism scores than did
ders: BED 12% versus 10%, BN 15% versus 10%, those who engaged only in restricting behaviors.
and AN 0.2% versus 0. Greater perfectionism was associated with lower

236 Psycholo gical Comorbidities

body weight and greater prominence of eating pre- worrying to being relaxed, optimistic, bold, and
occupations and rituals as well as the diminished confident. Reward dependence ranges from senti-
motivation to change. Thus greater severity of eat- mental, dedicated, attached, and dependent to prac-
ing disorder symptoms was associated with greater tical, cold, detached, and independent. Persistence
perfectionism (Halmi et al., 2003). ranges from being industrious, diligent, ambitious,
Studies on relation between perfectionism and perfectionistic, and overachieving to being inactive,
obsessive-​compulsive traits reveal a significant cor- pragmatic, indolent, and underachieving. Three
relation between some perfectionism subscales and character dimensions assessed by the TCI are cooper-
obsessive-​compulsive traits (Frost & Steketee, 1997; ativeness, self-​directiveness, and self-​transcendence.
Tozzi et al., 2005). This relationship was examined Cooperativeness covers the range from socially tol-
in 607 individuals with AN and BN participat- erant, empathetic, compassionate, and principled
ing in an international genetic study. They were to critical, opportunistic, socially intolerant, and
assessed for perfectionism, obsessive-​ compulsive revengeful. Self-​directiveness ranges from mature,
personality disorder (OCPD), and OCD (Halmi, responsible, reliable, self-​accepted, and resourceful
Tozzi, et  al., 2005). There was no significant dif- to immature, fragile, unreliable, self-​striving, and
ference in comorbidity frequency among the eating ineffective. Self-​ transcendence ranges from being
disorder subtypes; OCD 20%, OCPD 13%, and a wise, patient, creative, and forgetful to impatient,
combination of OCD/​OCPD, 16%. Perfectionism unimaginative, and self-​conscious.
was most severe in those with OCPD or the combi- In this meta-​analysis, compared with controls,
nation of OCPD/​OCD. The pairing of perfection- persistence was significantly higher in all eating
ism with OCPD may be a relevant core behavioral disorders except BED, with highest levels in AN,
feature for vulnerability to develop eating disorders. which also had the highest effect sizes for harm
Obsessions and compulsions unrelated to core avoidance. Novelty seeking was significantly ele-
eating disorder symptomology were assessed in per- vated relative to controls only in BN. Recovered AN
sons with a history of AN who were participating patients had significantly lower harm avoidance and
in an international multicenter genetic study. The higher reward dependence than those who remained
obsessive compulsive symptoms were assessed using ill. A  study by Klump et  al. (2004) revealed that
the Y-​BOCS (Goodman et al., 1989), a semistruc- women with binge-​ purge behaviors had higher
tured interview. Lifetime obsessions and compul- harm avoidance scores than control women. They
sions occurred in 86% of the AN restricting type also had significantly lower self-​directedness and
and in 79.1% of the AN binge-​purge type. The cooperativeness scores than control women. This
AN subgroups did not differ from OCD controls study suggests consistent personality disturbances
in the frequency of obsessions in the symmetry and in women with BN and AN with bulimic behav-
somatic categories or in the compulsion categories iors that are present during the acute phase of the
of ordering and hoarding. This finding suggests disorder and also after recovery. The combination
there may be some common phenotype character- of high harm avoidance and low self-​directedness
istic shared by most AN and obsessive compulsive may diminish the ability to cope with stressful life
patients and indicates there may also be some com- events (Fassino et  al., 2004). Other studies have
mon brain behavioral pathways. also found women ill with BN to have the high-
Trait anxiety is a common feature of AN. Greater est novelty-​ seeking scores of all eating disorder
trait anxiety predicted a lower likelihood of recovery groups (Brewerton, Hand, & Bishop, 1993; Bulik,
in a multisite genetic study of 680 women with AN Sullivan, Joyce, & Carter, 1995; Fassino et al., 2003;
(Zerwas et al., 2013). Kleifield, Sunday, Hurt, & Halmi, 1994). Brown,
The Temperament and Character Inventory Haedt-​Matt, and Keel (2011) found BN patients to
(TCI) was used to assess personality characteristics have greater impulsivity, trait anxiety, and Cluster
in eating disorder patients in 14 studies included in C (anxious personality) disorders compared with
a meta-​analysis by Atiye, Miettunen, and Raevuori-​ controls. In eating disorder patients with binge eat-
Helkama (2015). This inventory covers four tem- ing, high novelty seeking and low self-​directedness
perament dimensions. Novelty seeking ranges from were related to emotional eating. In all patients,
exploratory, curious, and impulsive nature to indif- increased harm avoidance was correlated with sever-
ferent, frugal, orderly, and regimented. Harm avoid- ity of illness (Rotella, Fioravanti, & Godini, 2015).
ance ranges from the tendency to inhabit behavior Studies comparing anorexia to healthy controls
to avoid punishment, fearful, shy, pessimistic, and found increased perfection, interpersonal distrust,

Halmi 237
maturity fears, and persistence and decreased nov- binge eating disorder symptomatology included
elty seeking (Fornasari et  al., 2014) and increased high novelty seeking, high harm avoidance and low
harm avoidance with decreased self-​ directedness self-​directedness (Grucza et al., 2007. The authors
and novelty seeking (Amianto, Abbate-​ Daga, suggested a heterogeneity may exist among indi-
Morando, Sobrero, & Fassino, 2011). viduals with binge eating disorder; for example
These personality characteristics of women with there may be an impulsive novelty seeking behavior
eating disorders may represent enduring tempera- for some and others with a predominance of harm
mental traits that contribute to eating disorder avoidance may have an underlying mood dysregula-
pathogenesis. Nonetheless, a few studies found that tion. In a study of personality and attitudes towards
the temperamental dimensions of some of these food, harm avoidance predicted greater likelihood
traits have changed with treatment. Novelty seek- to continue eating when satiated and novelty seek-
ing and self-​ directedness significantly increased ing predicted lack of dietary control (VandenBree,
and harm avoidance significantly decreased in both Przybeck, & Cloninger, 2006).
AN and BN (Segura-​Garcia, Chiodo, Sinopoli, & Latent Profile Analysis, an extension of latent
De Fazio, 2013). In another study of cognitive-​ class analysis, applied to 154 AN patients revealed
behavioral therapy in BN patients, harm avoidance three personality subtypes; undercontrolled—​
and self-​transcendence decreased after treatment, elevated scores on self-​harm, aggression, and impul-
whereas reward dependence and self-​directedness sivity; overcontrolled—​elevated scores on self-​harm,
increased (Aguera et al., 2012). low scores on exhibitionism and impulsivity; and
Factor analysis was employed to derive pheno- low psychopathology—​ normative scores on the
types from personality and behavioral traits in a Schedule for Nonadaptive and Adaptive Personality
large sample of individuals with eating disorders (Wildes et al., 2011).
(Price Foundation Collaborative Group, 2001). The Another study found that eating disorders with
most influential factor was one of trait anxiety, harm nonsuicidal self-​ injury were positively associated
avoidance, perfectionism, obsessive-​ compulsive with harm avoidance and self-​transcendence but neg-
behaviors, and diminished self-​ directedness. atively with reward dependence, self-​directedness,
Further discriminant analysis showed an 80% rate and cooperativeness (Islam et al., 2015).
of accurate classification of those individuals with a The relationship between personality/​tempera-
diagnosis of restricting-​type AN. ment and eating disorders can be studied from
A latent class analysis applied to the same sample different perspectives including predisposition
of patients described above revealed a larger group (personality/​temperament is a vulnerable feature
characterized by greater perfectionism, obsessions, for developing eating disorders), common cause
compulsions, rigidity, conscientiousness, lower (both share a common etiology) and interaction
levels of novelty seeking and higher levels of harm (personality/​temperament and eating disorders
avoidance. This group was consistent in patients modify the presentation and course of each other).
with restricting type of AN (Keel et  al., 2004). Research to date suggests high harm avoidance, low
Diagnostic crossover from BN to AN and from self-​directedness, perfectionism, and neuroticism
AN to BN in this same population was consistently (anxiety spectrum) are risk factors for developing
associated with low self-​directedness (Tozzi et  al., and maintaining an eating disorder. Cluster, latent
2005). The authors suggested that individuals with class, and taxometric analytic techniques may fur-
low self-​directedness, independent of diagnosis, ther identify empirically based phenotypes of eating
may be characterized by an inability to regulate disorders.
behaviors and affect adequately. They suggested this
instability may then lead to alternations between Affect Regulation
the cognitive and behavioral restraint common to Studies of affect regulation in eating disor-
restricting-​type AN and the disinhibition present in ders have mainly focused on negative affect and
BN. Parental criticism was a salient factor for indi- BN. Negative affect is a broad concept including
viduals with AN who crossed over to having BN. depression, anxiety, anger, fear, hostility, and guilt
Low scores on novelty seeking and the presence of and thus may be considered as a component of the
alcohol abuse or dependence were important in the NVS domain. Ecological momentary assessment
crossover from BN to AN. (EMA) found negative affect increased the likeli-
In a study of binge eating disorder in a com- hood of binging (Engleberg, Steiger, Gauvin, &
munity sample, personality traits associated with Wonderlich, 2007) and was correlated with laxative

238 Psycholo gical Comorbidities

use, vomiting, and exercising as well as binging in 16.7% recovered AN, 10% unaffected AN sisters,
both AN and BN (Wonderlich et al., 2013). Eating and 30% BN sisters. Those with poor compared
disorder inpatients who had less change in nega- to intact set-​shifting had a significant longer dura-
tive affect over treatment reported less reduction in tion of illness, had a higher Yale-​Brown-​Cornell
body dissatisfaction and drive for thinness at dis- Eating Disorder Scale (YBC-​ EDS) rituals score,
charge (Olatuni, Cox, Ebesutani, & Wall, 2015). and were twice as likely to have a diagnosis of social
Negative affect lability, the degree to which negative phobia, alcohol/​substance abuse, and self-​harming
affect shifts from measurement to measurement, behaviors (Roberts, Tchanturia, & Treasure, 2010).
and neuroticism predicted variability in weekly The authors suggested poor set-​shifting is not due
binge-​eating frequency for individuals with BN and to starvation, since it was not related to BMI and
BED (Zander & Young, 2014). Using functional occurred as frequently in BN as in AN patients.
magnetic resonance imaging (fMRI) Bohan and Poor (32%) and intact (65%) set-​shifting in a
Stice (2012) found women with BN showed a posi- group of AN patients differed in that those with
tive correlation between negative affect and activ- poor set-​shifting had significantly more depression
ity in the putamen, caudate, and pallidum during and lower reward dependence and cooperative-
anticipated receipt of milkshake versus a tasteless ness (Abbate-​Daga, Buzzichelli, Marzola, Amianto,
solution. Negative affect increased with increased & Fassino, 2014). Weight-​recovered AN women
responsivity of reward regions to anticipated intake were compared with control women in a category-​
of palatable food. The authors suggested negative learning task that required rule acquisition and then
affect may increase the reward value of food in BN a set-​shift following a rule change. Compared with
individuals or that negative affect may be a con- control women the AN weight-​recovered group dis-
ditioned cue due to a history of binge eating in a played steeper learning curves before the rule shift
negative mood. but greater difficulty in learning new categories after
the rule shift. The hyperlearning and set-​shifting
Cognitive Processes deficits were not correlated with clinical measures
Attention to examining cognitive processes in (Filoteo et  al., 2014). These findings point to the
eating disorders initially centered on AN due to complexity of cognitive processes in AN. Two stud-
the rigid and stereotyped behaviors these individu- ies have demonstrated that set-​shifting deficits do
als use to control eating and weight and their high not exist in adolescents with AN in contrast to
prevalence of obsessive-​ compulsive personality findings in adult AN (Buhren et  al., 2012; Shott
traits and perfectionism. Cognitive flexibility is one et  al., 2012). Since there is evidence that matura-
of the executive functions that influence behavior tion of corticostriatal networks does not reach the
flexibility. Inhibition, planning, attention con- adult state until the third decade (Paus & Toro,
trol, and working memory are also components of 2009), and set-​shifting abilities are modulated by
cognitive-​executive control. Cognitive flexibility is frontostriatal brain circuits involving maturation of
a broad concept and includes the ability to bypass the dopaminergic system, which continues beyond
habitual strategies and shift cognition and behavior puberty (Somerville & Casey, 2010), it is likely the
for changing demands. Set-​shifting, which involves brain circuits that support set-​shifting are not fully
various cognitive functions, has been assessed in eat- mature in adolescent AN.
ing disorder patients with several different tasks by Behavioral response shifting, inhibiting a prepo-
different researchers, making comparisons between tent impulse in favor of an alternative, was impaired
studies difficult. Thus, this is an area requiring in AN patients but showed some improvement after
extensive future research. There are salient findings long-​term recovery (Zastrow et al., 2009).
from the limited research to date. A temporal sequence of perception begins with
In a large sample of 98 women with a current seeing the global structure followed by perceiving
eating disorder, 30 recovered AN, 30 with BN, the finer details. Central coherence is defined as the
and 88 healthy controls, a composite set-​shifting overall balance between global over detail, and weak
variable from four neuropsychological assessments central coherence is defined as the tendency to pro-
revealed about one-​third of those with a current cess information in parts rather than as a whole, with
eating disorder displayed poor set-​ shifting com- a relative difficulty in global or integrative thinking
pared with 9% of the healthy controls; specifically (Happe, Briskman, & Firth, 2001). Studies of both
restricting AN 22.9%, binge-​purge AN 45.5%, and ill and recovered AN patients (Autreve et al., 2013;
BN 36.7%. Poor set-​shifting was also present in Harrison, Tchanturia, & Treasure, 2011) and BN

Halmi 239
and BED patients have found impairment in cen- specific comorbidity eating disorder profiles will be a
tral coherence (van den Eynde et  al., 2011). This meaningful focus for future research. This endeavor
trait is also present in unaffected sisters of eating dis- should be expedited by additional knowledge from
order individuals (Roberts et al., 2010) and suggests molecular genetic studies to determine specific gene
that weak central coherence may be an innate risk function involving cognitive and behavioral traits
factor for eating disorders. and brain activation patterns.
Emotional processing constructs include emo-
tional intelligence—​the ability to perceive, express, References
assimilate, and regulate emotion, and emotional Abbate-​Daga, G., Buzzichelli, S., Marzola, E., Amianto, F., &
Fassino, S. (2014). Clinical investigation of set-​sifting sub-
recognition—​ the ability to label complex emo-
types in anorexia nervosa. Psychiatry Research, 219, 592–​597.
tions (Mayer, Salovey, & Caruso, 2000). Emotion Aguera, Z., Krug, I., Sanchez, I., Granero, R., Penelo, E., Peñas-​
regulation strategies consist of cognitive reappraisal, Lledó, E.,   .  .  .  Fernández-​Aranda, F. (2012). Personality
reframing and accepting the emotion, and emo- changes in bulimia nervosa after a cognitive behavior ther-
tion suppression, which produces negative emotion apy. European Eating Disorders Review, 20, 379–​385.
American Psychiatric Association. (1987). Diagnostic and
(Gross, 2002). Attentional biases to social and angry
statistical manual of mental disorders, [DSM-​ 111-​
R].
threat stimuli and emotion recognition are present Washington, DC.
in all categories of eating disorders both in the ill American Psychiatric Association. (1994). Diagnostic and sta-
and recovered stages (Cardi, di Matteo, Corfield, & tistical manual of mental disorders, 4th edition (DSM-​IV).
Treasure, 2012). Considerable research is necessary Washington, DC: Author.
American Psychiatric Association. (2013). Diagnostic and
to elucidate the complexities of emotional process-
Statistical Manual of Mental Disorders, 5th edition. (DSM-​5).
ing in eating disorders. Washington, DC: Author.
Reward sensitivity and motivational behaviors Amianto, F., Abbate-​Daga, G., Morando, S., Sobrero, C., &
reflect a balance between approach and avoidance Fassino, S. (2011). Personality development characteristics
systems. Although all eating disorders are more of women with anorexia nervosa. Psychiatry Research, 187,
401–​408.
sensitive to punishment compared with controls,
Atiye, M., Miettunen, J., & Raevuori-​Helkama, A. (2015). A
restricting anorectics are least sensitive to reward meta-​analysis of temperament in eating disorders. European
and have higher levels of behavioral inhibition. Eating Disorders Review, 23, 89–​99.
Bulimic types are most sensitive to reward and have Autreve, S., De Baene, W., Baeken, C., van Heeringen, C., &
higher levels of behavioral activation (Harrison Vervaet, M. (2013). Do restrictive and bingeing/​purging
subtypes of anorexia nervosa differ on central coherence
et al., 2010; Jappe et al., 2011).
and set-​shifting? European Eating Disorders Review, 21,
Functional MRI with a target-​ detection task 308–​313.
revealed impaired behavioral response shifting in Bohan, C., & Stice, E. (2012). Negative affect and neu-
anorexia nervosa to be associated with hypoacti- ral response to palatable food intake in bulimia nervosa.
vation in the ventral anterior cingulate-​ striatal-​ Appetite, 58, 964–​970.
Braun, D. L., Sunday, S. R., & Halmi, K. A. (1994). Psychiatric
thalamic loop that is involved in motivation
comorbidity in patients with eating disorders. Psychological
behavior. Predominant activation of frontoparietal Medicine, 24, 859–​867.
networks reflecting effortful and supervisory cog- Brewerton, T., Hand, L., & Bishop, B. (1993). The tridimen-
nitive control occurred during task performance sional personality questionnaire in eating disorder patients.
(Zastrow et al., 2009). Monoamines in response to International Journal of Eating Disorders, 14, 213–​218.
Brown, T. A., Haedt-​Matt, A. A., & Keel, P. (2011). Personal
stress, mood, and motivations modulate the func-
pathology in purging disorder and bulimia nervosa.
tioning of the striatocortical loops. Dopamine is International Journal of Eating Disorders, 44, 735–​740.
active in the ventral striatum for set-​shifting tasks Buhren, K., Mainz, V., Herpertz-​Dahlmann, B., Schäfer, K.,
(Robbins, 2005), and norepinephrine is involved Kahraman-​Lanzerath, B., Lente, C., & Konrad, K. (2012).
in response inhibition and sustained attention Cognitive flexibility in juvenile anorexia patients before and
after weight recovery. Journal of Neural Transmission, 119,
(Chamberlain et al., 2009).
1047–​1057.
Bulik, C., Klump, K., Thornton, L., Kaplan, A., Devlin, B.,
Conclusions Fichter, M. M.,  . . .  Kaye, W. H. (2004). Alcohol use dis-
Comorbidities of eating disorders with other order comorbidity in eating disorders: A multicenter study.
psychiatric disorders as well as with distinct person- Journal of Clinical Psychiatry, 65, 1000–​106.
Bulik, C., Sullivan, P., Joyce, P., & Carter, F. (1995). Temperament,
ality traits and features are common in eating disor-
character, and personality disorder in bulimia nervosa.
ders and often typify or characterize eating disorder Journal of Nervous and Mental Disease, 183, 593–​598.
subtypes. Comorbidity profiles may be associated Cardi, V., di Matteo, R., Corfield, F., & Treasure, J. (2012).
with treatment outcome. Developing treatments for Social reward and rejection sensitivity in eating disorders.

240 Psycholo gical Comorbidities

 World Journal of Biological Psychiatry. doi:10.3109/​15622975 Gross, J. J. (2002). Emotion regulation: Affective, cognitive and
10; 20–​24. social consequences. Psychophysiology, 39, 281–​291.
Chamberlain, S. R., Hampshire, A., Muller, U., Rubia, K., Grucza, R., Przybeck, T., & Cloninger, R. (2007). Prevalence and
Del Campo, N., Craig, K.,   .  .  .  Sahakian, B. J. (2009). correlates of binge eating disorder in a community sample.
Atomoxetine modulates right inferior frontal activa- Comprehensive Psychiatry, 48, 124–​131.
tion during inhibitory control. Biological Psychiatry, 65, Halmi, K. A., Agras, W. S., Crow, S., Mitchell, J., Wilson, G.
550–​555. T., Bryson, S. W., & Kraemer, C. (2005). Predictors of
Costello, E., & Angold, A. (1995). Epidemiology. In J. March treatment acceptance and completion in anorexia nervosa.
(Ed.), Anxiety disorders in children and adolescents (pp.  1–​ Archives of General Psychiatry, 62, 776–​781.
124). New York, NY: Guilford. Halmi, K. A., Eckert, E., Marchi, P., Sampugno, V., Apple, R.,
Engleberg, M., Steiger, H., Gauvin, L., & Wonderlich, S. (2007). & Cohen, J. (1991). Comorbidity of psychiatric diagnoses in
Binge antecedents in bulimic syndromes:  An examination anorexia nervosa. Archives of General Psychiatry, 48, 712–​718.
of dissociation and negative affect. International Journal of Halmi, K. A., Sunday, S., Klump, K., Strober, M., Leckman, J.,
Eating Disorders, 40, 531–​536. Fichter, M.,  . . .  Kaye, W. H. (2003). Obsessions and com-
Fairburn, C., Cooper, Z., Dell, H., & Welch, S. (1999). Risk pulsions in anorexia nervosa subtypes. International Journal
factors for anorexia nervosa:  Three integrated case-​control of Eating Disorders, 33, 308–​319.
comparisons. Archives of General Psychiatry, 56, 468–​476. Halmi, K. A., Sunday, S., Strober, M., Kaplan, A., Woodside,
Fassino, S., Amianto, F., Abbate-​ Daga, G., Leombruni, P., D. B, Fichter, M.,  . . .  Kaye, W. H. (2000). Perfectionism
Garzaro, L., Levi, M., & Rovera, G. G. (2003). Bulimic in anorexia nervosa: Variation by clinical subtype, obsession-
family dynamics: Role of parent’s personality; A controlled ality, and pathological eating behavior. American Journal of
study with the temperament and character inventory. Psychiatry, 1157, 1799–​1805.
Comprehensive Psychiatry, 44, 70–​77. Halmi, K. A., Tozzi, F., Thornton, L., Crow, S., Fichter, M. M.,
Fassino, S., Amianto, F., Gramaglia, C., Facchini, F., & Abbate Kaplan, A. S.,  . . .  Bulik, C. M. (2005). The relation among
Daga, G. (2004). Temperament and character in eating dis- perfectionism, obsessive compulsive personality disorder
orders:  Ten years of studies. Eating and Weight Disorders, and obsessive compulsive disorder in individuals with eat-
9, 81–​90. ing disorders. International Journal of Eating Disorders, 38,
Fernandez-​Aranda, F., Gimenez-​Murcia, S., Alvarez, E., Granero, 371–​374.
R., Vallejo, J., & Bulik, C. M. (2006). Impulse control dis- Happe, F., Briskman, J., & Firth, U. (2001). Exploring the cog-
orders in eating disorders: Clinical and therapeutic implica- nitive phenotype of autism: Weak central coherence in par-
tions. Comprehensive Psychiatry, 47, 482–​488. ents and siblings of children with autism. Journal of Child
Fernandez-​Aranda, F., Pinheiro, A., Thornton, L., Berrettini, W., Psychology and Psychiatry, 42, 299–​307.
Crow, S., Fichter, M. M.,  . . .  Bulik, C. M. (2008). Impulse Harrison, A., O’Brien, N., Lopez, C., & Treasure, J. (2010).
control disorders in women with eating disorders. Psychiatry Sensitivity to reward and punishment. Psychiatric Research,
Research, 157, 147–​157. 177, 1–​11.
Filoteo, J. V., Paul, E. J., Ashby, F. G., Frank, G. K., Helie, S., Harrison, A., Tchanturia, K., & Treasure, J. (2011). Measuring
Rockwell, R.,  . . .  Kaye, W. H. (2014). Simulating category state trait properties of detail processing and global integra-
learning and set shifting deficits in patients weight-​restored tion ability in eating disorders. World Journal of Biological
from anorexia nervosa. Neuropsychology, 28, 741–​751. Psychiatry, 12, 462–​472.
Fornasari, L., Gregoraci, G., Isola, M., Negri, G., Rambaldelli, Herzog, D., Keller, M., Lavori, P., Kenny, G., & Sack, S. (1992).
G., Cremaschi, S.,  . . .  Brambilla, P. (2014). Psychological The prevalence of personality disorders in 210 women with
and personality traits underlie decision making in recent eating disorders. Journal of Clinical Psychiatry, 53, 147–​152.
onset medication naïve anorexia nervosa, Psychiatric Research, Holderness, C., Brook-​ Gunn, J., & Warren, M. (1994).
216, 89–​96. Comorbidity of eating disorders in substance abuse review
Frank, G., Bailer, U., Henry, S., Drevets, W., Meltzer, C. C., of the literature. International Journal of Eating Disorders,
Price, J. C.,  . . .  Kaye, W. H. (2005). Increased dopamine 16, 1–​34.
D2/​D3 receptors binding after recovery from anorexia ner- Hudson, J., Pope, H., Yurgelun-​ Todd, D., Johnas, J., &
vosa measured by positron emission tomography and 11C Frankenburg, F. (1987). A controlled study of lifetime preva-
Raclopride. Biological Psychiatry, 58, 908–​912. lence of affective and other psychiatric disorders in bulimic
Frost, R. O., & Steketee, G. (1997). Perfectionism in obses- outpatients. American Journal of Psychiatry, 144, 1283–​1287.
sive compulsive disorder patients. Behaviour Research and Insel, T., Cuthbert, B. N., Garvey, M. A., Heinssen, R. K., Pine,
Therapy, 35, 291–​296. D. S., Quinn, K.,  . . .  Wang, P. (2010). Research Domain
Gartner, A., Marcus, R., Halmi, K., & Loranger, A. (1989). Criteria (RDoC):  Toward a new classification framework
DSM-​III-​R personality disorders in patients with eating dis- for mental disorders. American Journal of Psychiatry, 167,
orders. American Journal of Psychiatry, 146, 1585–​1591. 748–​751.
Godart, N. T., Flament, M. F., Perdereau, F., Jeammet, P. (2002). Hudson, J., Hiripi, E., Pope, H. (2007). The prevalence and
Comorbidity between eating disorder and anxiety disor- correlates of eating disorders in the National Comorbidity
ders: A review. International Journal of Eating Disorders, 32, Survey Replication. Biological Psychiatry, 61, 348–​358.
253–​270. International Statistical Classification of Diseases and Related
Goodman, W., Price, L., Rasmussen, S., Mazure, C., Health –​ 10th Revision [ICD-​ 10], 1992. World Health
Fleischmann, R. L., Hill, C. L.,  . . .  Charney, D. S. (1989). Organization, Geneva.
The Yale-​Brown Obsessive Compulsive Scale: Development, Islam, M., Steiger, H., Jimenez-​ Murcia, S., Israel, M.,
use and reliability. Archives of General Psychiatry, 46, Granero, R., Agüera, Z.,   .  .  .  Fernández-​Aranda, F.
1006–​1011. (2015). Non-​suicidal self-​injury in different eating disorder

Halmi 241
 types:  Relevance of personality traits and gender. European Ross, H., & Ivess. (1999). Binge eating and substance abuse
Eating Disorders Review. doi:10.1002/​erv.2374 12; 8–​12. among male and female adolescents. International Journal of
Jappe, L. M., Frank, G., Shott, M., Rollin, M. D. H., Pryor, T., Eating Disorders, 26, 245–​260.
Hagman, J. O.,  . . .  Davis, E. (2011). Heightened sensitivity Rossiter, E., Agras, W., Telch, C., & Schneider, J. (1993). Cluster
to reward and punishment in anorexia nervosa. International B personality disorder characteristics predict outcome in the
Journal of Eating Disorders, 44, 317–​324. treatment of bulimia nervosa. International Journal of Eating
Kaye, W. H., Bulik, C. M., Thornton, L., Barbarich, N., & Disorders, 13, 349–​357.
Master, K. (2004). Comorbidity of anxiety disorders with Rotella, F., Fioravanti, G., & Godini, L. (2015). Temperament
anorexia and bulimia nervosa. American Journal of Psychiatry, and emotional eating: A crucial relationship in eating disor-
161, 2215–​2221. ders. Psychiatry Research, 225, 452–​457.
Keel, P., Fichter, M., Quadflieg, N., Bulik, C., Baxter, M. G., Schmidt, N., & Telch, M. (1990). Prevalence of personality
Thornton, L.,   .  .  .  Kaye, W. H. (2004). Application of a disorders among bulimic, non-​ bulimic binge eaters, and
latent class analysis to empirically define eating disorder phe- normal controls. Journal of Psychopathology and Behavioral
notypes. Archives of General Psychiatry, 61, 192–​200. Assessment, 12, 170–​185.
Kleifield, E., Sunday, S., Hurt, S., & Halmi, K. (1994). The Segura-​Garcia, C., Chiodo, D., Sinopoli, F., & De Fazio, P.
tridimensional personality questionnaire: An exploration of (2013). Temperamental factors predict long-​term modifica-
personality traits in eating disorders. Journal of Psychiatric tions of eating disorders after treatment. BMC Psychiatry, 13,
Research, 28, 413–​423. 288–​295.
Klump, K., Strober, M., Bulik, C., Thornton, L., Johnson, C., Shott, M. E., Filoteo, J. V., Bhatnagar, K., Peak, N. J., Hagman,
Devlin, B.,   .  .  .  Kaye, W. H. (2004). Personality charac- J. O., Rockwell, R.,   .  .  .  Frank, G. K. (2012). Cognitive
teristics of women before and after recovery from an eating set-​shifting in anorexia nervosa. European Eating Disorders
disorder. Psychological Medicine, 34, 1407–​1418. Review, 20, 343–​349.
Laseque, C. (1873). De l’anorexie hysterique. Archives générales Sommerville, L. H., & Casey, B. J. (2010). Developmental neu-
de Médecine, 1, 385–​403. robiology of cognitive control and motivational systems.
Lee, H. Y., Lee, E. L., Pathy, P., & Chan, Y. H. (2005). Anorexia Current Opinion in Neurobiology, 20, 236–​241.
nervosa in Singapore: An 8 year retrospective study. Singapore Steiger, H., Thibaudeau, J., Leung, F., Houle, L., & Ghadirian,
Medical Journal, 46, 275–​282. A. (1994). Eating and psychiatric symptoms as a function
Matsunaga, H., Kaye, W., McConaha, C., Plotnicov, K., Pollice, of Axis II comorbidity in bulimic patients. Psychosomatics,
S., & Rao, R. (2000). Personality disorder among subjects 35, 41–​49.
recovered from eating disorders. International Journal of Steiger, H., Young, S. N., Kim, N., Koerner, N., Isreal, M.,
Eating Disorders, 27, 353–​357. Lageix, P., & Paris, J. (2001) Implications of impulsive and
Mayer, J. D., Salovey, P., & Caruso, D. (2000). Models of emo- affective symptoms for serotonin function in bulimia ner-
tional intelligence. In R. J. Sternberg (Ed.), The handbook vosa. Psychological Medicine, 31, 85–​95.
of intelligence (pp. 396–​400). New  York, NY:  Cambridge Stock, S., Goldberg, D., Corbett, S., & Katzman, D. (2002).
University Press. Substance use in female adolescents with eating disorders.
Olantunji, B., Cox, R., Ebesutani, C., & Wall, D. (2015). Self-​ Journal of Adolescent Health, 31, 176–​182.
harm history predicts resistance to inpatient treatment of Strober, M. (2004). Clinical and research forum pathological
body shape aversion in women with eating disorders: The role fear conditioning and anorexia nervosa:  On the search for
of negative affect. Journal of Psychiatric Research, 65, 37–​46. novel paradigms. International Journal of Eating Disorders,
Paus, T., & Toro, R. (2009). Could sex differences in white matter 35, 504–​508.
be explained by g ratio? Frontiers in Neuroanatomy, 3, 14–​20. Sutander-​Pinnock, K., Blake, W., Carter, J., Olmsted, M., &
Piacentini, J., & Bergman, R. (2000). Obsessive-​compulsive dis- Kaplan, A. (2003). Perfectionism in anorexia nervosa: A 6-​
order in children. Psychiatric Clinics of North America, 23, 24 month study. International Journal of Eating Disorders, 33,
519–​533. 225–​229.
Powers, P., Coovert, D., Brikwell, D., & Stevens, B. (1988). Tozzi, F., Thornton, L., Klump, K., Ficheter, M., Halmi, K. A.,
Other psychiatric disorders among bulimic patients. Kaplan, A. S.,  . . .  Kaye, W. H. (2005). Symptom fluctua-
Comprehensive Psychiatry, 29, 503–​508. tion in eating disorders:  Correlates of diagnostic crossover.
Price Foundation Collaborative Group. (2001). Deriving behav- American Journal of Psychiatry, 162, 732–​740.
ioral phenotypes in an international, multi-​center study of VandenBree, M., Przybeck, T., & Cloninger, C. (2006). Diet
eating disorders. Psychological Medicine, 31, 635–​645. and personality: Associations in a population-​based sample.
Raney, T., Thornton, L., Berrettini, W., Brandt, H., Crawford, Appetite, 46, 177–​188.
S., Fichter, M. M.,   .  .  .  Bulik, C. M. (2008). Influence Van den Eynde, F., Guillaume, S., Broadbent, H., Stahl, D.,
of overanxious disorder of childhood on the expression of Campbell, I. C., Schmidt, U., & Tchanturia, K. (2011).
anorexia nervosa. International Journal of Eating Disorders, Neurocognition in bulimic eating disorders:  A  systematic
41, 326–​332. review. Acta Psychiatrica Scandinavica, 124, 120–​140.
Robbins, T. W. (2005). Chemistry of the mind: Neurochemical Walsh, B. T., Hadigan, C. M., & Devlin, M. J. (1991). Long-​
modulation of prefrontal cortical function. Journal of term outcome of antidepressant treatment for Bulimia
Comparative Neurology, 493, 465–​475. Nervosa. American Journal of Psychiatry, 148, 1206–​1212.
Roberts, M., Tchanturia, K., & Treasure, J. (2010). Exploring the Weiderman, M., & Pryor, T. (1996). Substance use among
neurocognitive signature of poor set-​shifting in anorexia and women with eating disorders. International Journal of Eating
bulimia nervosa. Journal of Psychiatric Research, 44, 464–​470. Disorders, 20, 163–​168.

242 Psycholo gical Comorbidities

Weiderman, M., & Pryor, T. (1996). Substance use and impul- momentary assessment measures of negative affect and eating
sive behaviors among adolescents with eating disorders. disorder behaviors. International Journal of Eating Disorders,
Addictive Behaviours, 21, 269–​272. 48, 305–​311.
Welch, E., Jangmo, A., Thornton, L., Norring, C., von Wonderlich, S., Swift, W., Slotnick, H., & Goodman, S.
Hausswolff-​Juhlin, Y., Herman, B. K.,   .  .  .  Bulik, C. M. (1990). DSM-​III-​R Personality disorders in eating disor-
(2016). Treatment seeking patients with BED in the Swedish der subtypes. International Journal of Eating Disorders, 9,
national register: Clinical course and psychiatric comorbid- 607–​616.
ity. BMC Psychiatry, 16, 163–​171. Zander, M., & Young, K. (2014). Individual differences in nega-
Wildes, J., Marcus, M., & Fagioline, A. (2007). Eating disorders tive affect and weekly variability in binge eating frequency.
and illness burden in patients with bipolar spectrum disor- International Journal of Eating Disorders, 47, 296–​301.
ders. Comprehensive Psychiatry, 48, 516–​521. Zastrow, A., Kaiser, S., Stippich, C., Walther, S., Herzog, W.,
Wildes, J., Ringham, R., Crosby, R., Ringham, R. M., Dapelo, Tchanturia, K.,   .  .  .  Friederich, H. C. (2009). Neural
M. M., Gaskill, J. A., & Forbush, K. T. (2011). The clini- correlates of impaired cognitive-​behavioral flexibility in
cal utility of personality subtypes in patients with anorexia anorexia nervosa. American Journal of Psychiatry, 166,
nervosa. Journal of Consulting and Clinical Psychology, 79, 608–​616.
665–​674. Zerwas, S., Lund, B. C., Holle, A. V., Thornton, L., Berrettini,
Wonderlich, J., Lavender, J., Wonderlich, S., Peterson, C. W. H., Brandt, H.,  . . .  Bulik, C. M. (2013). Factors asso-
B., Crow, S. J., Engel S. G.,   .  .  .  Crosby, R. D. (2013). ciated with recovery from anorexia nervosa. Journal of
Examining convergence of retrospective and ecological Psychiatric Research, 47, 972–​979.

Halmi 243
 PART 
4
Prevention and
Treatment
 CH A PT E R


Prevention: Current Status and
13 Underlying Theory

C. Barr Taylor, Ellen E. Fitzsimmons-​Craft, and Neha J. Goel

Abstract
Eating disorders (EDs) are important and common problems among adolescents and young women,
and preventing them would be an important public health achievement. Fortunately, several recent
studies, informed by cross-​sectional, longitudinal, and clinical risk factor research, have demonstrated a
significant decrease in ED risk factors, with several programs also achieving a significant reduction in ED
onset within at-​risk females. This chapter reviews and evaluates the state of ED prevention research,
highlighting current theoretical approaches and effective programs, emphasizing emerging empirical
support for cognitive dissonance, Internet, school-​based, media literacy, and combined ED and obesity
prevention programs. Conclusions about how to enhance recent progress in the field of EDs are
provided.
Key Words:  cognitive dissonance, moderator, mediator, prevention, psychoeducation, risk factor

Introduction as discussed later, ED risk factors begin early, even

Eating disorders (EDs) are important and com- during childhood, and the peak onset of EDs is
mon problems among adolescents and young during adolescence. For this reason, although EDs
women in particular. The attitudes and behaviors are common in adults, prevention should prima-
associated with EDs and that may predispose young rily focus on adolescents and young women. Very
women to EDs can have serious psychological and little has been written about prevention in males.
physical consequences and are associated with loss Discussion focuses primarily on recent progress
of confidence and self-​esteem, shame, and other in this field, given that extensive work in the past
psychological problems. There is some evidence decade has resulted in several effective preventive
that once EDs become established, they are more interventions, at least for older adolescents and
difficult to treat, at least in the case of anorexia college-​age women, with several programs evidenc-
nervosa (AN; Guarda, 2008), and the longer the ing a significant reduction in ED onset among
disorder lasts the more likely it is to have adverse high-​ risk women (Martinsen et  al., 2014; Stice,
effects. Proponents of prevention argue, then, that Marti, Spoor, Presnell, & Shaw, 2008; Stice, Shaw,
it is important both to reduce the prevalence of risk Burton, & Wade, 2006; Taylor et al., 2006; Taylor
factors and to keep EDs from emerging and that, et al., 2016).
given the prevalence of disordered eating attitudes
and behaviors in young women, this is of major Defining Prevention
public health importance. In this chapter, we review There are three basic categories of prevention
issues related to preventing EDs, with our discus- programs:  universal, targeted (selective), and indi-
sion focusing almost entirely on young women, as cated. Universal prevention programs target the risk
females are most often the focus of studies. Also, factors in a whole population even though most of

247
the individuals in that population may be at no or onset of the disorder (see ­chapter  6 for a detailed
little risk. For instance, a universal prevention pro- discussion of these issues). In fact, a risk factor can
gram might focus on reducing the impact of media only be shown to be causative if reduction in the
on increasing all students’ desire to be very thin or risk factor is associated with reduction of the inci-
on reducing unhealthy comments or pressure to be dence of the disorder. Most of the modifiable risk
thin within schools, groups, or peers. The assump- factors for EDs are continuous measures, and it is
tions are that the reduction of a risk factor in a pop- unknown how much or for how long the measure
ulation, including those not at risk, will reduce the needs to be reduced to have a significant effect on
onset of the disorder in those at risk. The recom- reducing EDs.
mended intervention should be of potential bene- Another important factor when considering
fit and little risk to the population. Ironically, one causative issues and related prevention program
of the controversies in ED prevention is the role of development is recognition of the multiple risk fac-
the universal public health campaign to encourage tors involved in the onset of the target disorder and
weight reduction in the US population as rates of the required duration for reduction in risk factors
overweight and obesity in the United States have to truly determine that risk is no longer present.
increased dramatically over the past few decades. In theory, not only should a risk factor be reduced
In 2011–​2014, the prevalence of obesity was just through prevention programs, but it should also
over 36% in adults and 17% in youth, with higher remain reduced over time. For instance, to the
rates found in women (Ogden, Carroll, Fryar, & extent that elevated weight/​ shape concerns are
Flegal, 2015). Some proponents of ED preven- a common risk factor for the onset of EDs, then
tion have argued that such campaigns may have a a sustained reduction of weight/​ shape concerns
negative effect (O’Dea, 2005). The impact of these should be necessary to reduce onset in a high-​risk
campaigns on the prevalence of EDs remains an population. The StudentBodies prevention program,
important, unanswered question. However, given discussed in the text that follows, was designed with
the evidence that weight teasing can lead to EDs, it this assumption, and there was a sustained reduc-
is important that public health campaigns encour- tion in weight/​shape concerns in the intervention
aging weight loss do so without stigmatizing indi- sample as well as a reduction of the incidence of
viduals (Ikeda, Crawford, & Woodward-​Lopez, EDs in some subgroups (Taylor et al., 2006; Taylor
2006; Puhl & Heuer, 2010). et  al., 2016). However, as risk for EDs is multi-
Targeted or selective prevention interventions focus factorial, with different pathways and interactions
on individuals or a segment of the population whose among risk factors possibly contributing to ED
risk of developing EDs is significantly higher than onset (Allen, Byrne, Crosby, & Stice, 2016; Stice,
average. Most of the ED interventions have focused Presnell, et al., 2007), it may be that reduction of
on populations at “high risk” of developing an ED; several risk factors simultaneously may have a more
these programs are considered to be more effective pronounced effect than focusing on reducing one
than universal programs, as is discussed later (Ciao, risk factor. For example, a young woman at risk for
Loth, & Neumark-​Sztainer, 2014; Stice, Presnell, ED onset might have sustained weight/​shape con-
Gau, & Shaw, 2007; Stice, Shaw, & Marti, 2007; cerns, but an insensitive comment from a coach or
Watson et al., 2016). Indicated prevention programs teacher about her weight at a time when she is also
target individuals at very high risk or those who feeling depressed might lead to episodes of purging
have early features of the disorder. In the ED world, that elevate her to high risk for ED onset, while for
the distinctions among targeted and indicated inter- another young woman, having high weight/​shape
ventions are often arbitrary, as there may be a con- concerns alone may lead to ED onset. Thus, follow-
tinuum between low-​level behaviors and subclinical ing on the earlier statement regarding duration of
and clinical disorders. risk reduction and considering the diverse array of
risk factors involved in the etiology of EDs, should
Theory of Prevention all relevant risk factors be reduced for prolonged
The guiding theory for prevention programs is periods to truly prevent EDs? This is an important
that reduction of causative, modifiable risk factors consideration for both the development and evalua-
will reduce the incidence of a disorder. Thus, the tion of prevention programs.
foundation of prevention is the identification of Finally, it should be noted that from a pub-
such “causative risk factors,” that is, risk factors that lic health standpoint, “harm reduction,” such as
have been shown in prospective studies to cause the reduced exposure to environmental risk factors, may

248 Prevention
be an important aspect for ED prevention. Examples including EDs (Jacobi, Hayward, et  al., 2004;
of harm reduction in other fields include efforts to Pike et  al., 2007; Sanci et  al., 2008; Speranza
minimize exposure to second-​hand tobacco smoke et  al., 2003). If so, prevention of adverse child-
and drinking alcohol more safely (e.g., no access to hood experiences would likely reduce the onset
car) or in smaller amounts. In regard to ED preven- of EDs (and other mental health problems) but is
tion, harm reduction efforts have focused on ban- a goal beyond ED prevention. Other risk factors
ning “skinny models” and, more controversially, that need to be considered include personal char-
reducing or restricting public health campaigns for acteristics such as gender and age, as these may fur-
“obesity prevention.” However, many more ado- ther identify specific groups at risk for ED onset.
lescents and young men and women are at risk of Gender should be considered an important risk
obesity than are at risk for EDs, and from a public factor given the preponderance of EDs in females
health standpoint, obesity may pose a more impor- (Jacobi, Hayward, et al., 2004). In terms of onset
tant health issue. Thus ED prevention researchers and course, AN typically begins in adolescence or
should work to unite their efforts with those related young adulthood. Early work described a bimodal
to obesity prevention given that reduction of over- distribution of age at onset for AN, with two
weight and obesity, a critical societal need, would peaks at 14 and 18 years (Halmi, Casper, Eckert,
likely assist in reducing risk for concurrent and Goldberg, & Davis, 1979). More recent work has
future eating pathology (as discussed later, programs identified peak age of onset for AN to be between
that target multiple outcomes, such as overweight 15 and 19  years (Micali, Hagberg, Petersen, &
and disordered eating, are desirable from a public Treasure, 2013) or 18 and 20 years (Stice, Marti, &
health and possibly a harm reduction standpoint) Rohde, 2013; Volpe et al., 2016). Bulimia nervosa
(Haines & Neumark-​ Sztainer, 2006; Neumark-​ (BN) also typically begins in adolescence or young
Sztainer et al., 2007). adulthood, with the peak age of onset between 15-​
20  years (Micali et  al., 2013; Stice, Marti, et  al.,
Risk Factors That Inform Prevention 2013; Volpe et  al., 2016). These data suggest that
Programs preventive programs should begin by at least age 12,
In recent years, a number of prospective studies if not earlier, and that preventive programs are rel-
have been conducted to help identify risk factors for evant into the mid-​20s.
EDs. These prospective studies, along with an even There has been some interest as to whether or
larger literature on retrospective, cross-​ sectional, not acculturation to Western society may trig-
and clinical studies have identified a number of ger EDs. However, both greater and lesser accul-
risk factors (Jacobi, Hayward, de Zwaan, Kraemer, turation have been identified as risk factors for the
& Agras, 2004; Stice & Shaw, 2004). Elevated per- development of an ED, and this varies depending
ceived pressure to be thin from family, peers, and the on the group studied as well as how accultura-
media, internalization of the thin-​ideal espoused for tion and culture change are conceptualized and
women by Western culture, body mass index (BMI), measured (Doris et  al., 2015). Studies have yet to
and body dissatisfaction have predicted future eat- show a change in culture may affect ED onset rates,
ing pathology in multiple studies (Culbert, Racine, except as applied to very contained environments.
& Klump, 2015; Field, Camargo, Taylor, Berkey, Finally, genetic and biological studies may eventu-
& Colditz, 1999; Killen et al., 1996; Stice, Chase, ally identify populations at particular risk and in
Stormer, & Appel, 2001; Stice, Presnell, et al., 2007; need of targeted prevention efforts (Bulik, Kleiman,
Wertheim, Koerner, & Paxton, 2001; Wichstrom, & Yilmaz, 2016).
2000). The role of dieting as a risk factor is less clear,
and the results have been inconsistent, perhaps Screening
because of problems associated with measurement Although a number of screens have been devel-
(Stice & Shaw, 2004) and/​or because it so common oped to identify individuals with EDs, less work
as to be nonspecific. has been done on identifying individuals at risk for
In addition to the ED-​ specific risk factors but without EDs. Furthermore, few studies have
described earlier, there are additional nonspecific evaluated the psychometric properties of high-​risk
risk factors to consider. For example, a history of screens, particularly in longitudinal studies (Jacobi,
adverse childhood experiences (including sexual Abascal, & Taylor, 2004). The Weight Concerns
abuse) appears to be a risk factor for the devel- Scale (WCS) (Killen et  al., 1994) is the most
opment of a number of mental health problems, well-​studied studied risk factor screen. The WCS

Taylor, Fitzsimmons-Craft, Goel 249

consists of five questions that assess worry about Psychoeducation
weight/​shape, fear of gaining 3 lbs., frequency of Many preventive programs have been based on
dieting, importance of weight, and feelings of fat- the assumption that information about EDs, includ-
ness; it was derived from a principal components ing the consequences of EDs, can be used to reduce
analysis of a set of self-​report questions used to risk factors. This model is widely implemented
assess ED symptoms (Killen et al., 1993). In ear- in school settings as part of universal educational
lier studies, the investigators found that students programs and appears to be the most common
scoring in the upper 25% were significantly more approach to ED prevention for younger students
likely to develop an ED than those scoring in the (i.e., elementary and middle school students). It
lower 75% range (Killen et al., 1993; Killen et al., should be noted, though, that without long-​term
1994). Jacobi, Abascal, and colleagues (2004) assessment of programs delivered to younger popu-
reanalyzed the data from the initial sample (Killen lations, it may be hard to determine the effective-
et al., 1993), finding that a WCS score of greater ness given that onset of EDs would not typically
than 47 had a sensitivity of 79%, a specificity of be expected until mid-​to late adolescence. Overall,
67%, and a predictive value/​efficiency of 13% for psychoeducation approaches alone, particularly
identifying cases. those with limited interactions with participants
For a population-​ based study, a screen also (e.g., didactic programs) and delivered to individu-
needs to be able to separate out “real cases” from als at low risk for ED onset, may have limited effec-
those at high risk of becoming a case. The ideal tiveness (Stice, Shaw, et al., 2007). Furthermore, a
screen would be very sensitive and specific. A sen- psychoeducation-​only approach to the prevention
sitive screen with low specificity means that many of EDs does not appear effective.
“false positives” are identified. Unfortunately, the
cost of then determining whether these false posi- Social Learning Theory
tives are “true cases,” such as through the use of Social learning theory assumes that behavior
a clinical interview, can be expensive. To address is affected by both external and internal processes
this problem, Jacobi, Abascal, et  al. (2004) rec- (Bandura, 1986), and following from this, dis-
ommend a two-​step process, with the first step ordered eating is thought to result from several
focusing on sensitivity and the second on specific- processes (e.g., pressure to be thin from family
ity. Online screens can be designed to do just that members and peers, exposure to maternal and peer
and minimize subject burden—​only subjects who weight and shape concerns, the individual’s inter-
answer questions suggesting that they may have an nalization of the thin ideal, history of disordered
ED are provided follow-​up questions to improve eating attitudes and behaviors, history of depression
specificity. The Stanford-​ Washington University and anxiety). Thus, from a social learning theory
Eating Disorder Screen (SWED) was designed in perspective, four factors are particularly important
this way to sort individuals into risk (low, high) in influencing attitudes and behavior and need to
and symptom (subclinical, clinical) categories. be addressed in preventive interventions: (1) mod-
It does so by assessing ED diagnostic criteria, in eling, (2)  information, (3)  instructions/​ persua-
addition to factors that confer increased risk (e.g., sion from authorities, and (4) previous experience.
weight/​shape concerns, dieting, BMI). Sensitivity Effective interventions incorporate each of these
and specificity are very high for cases of DSM-​IV components. For example, social learning theory
and DSM-​5 EDs (most > .90) (Wilfley, Agras, & has generally been used to guide the preventive
Taylor, 2013). approaches used in StudentBodies, an Internet-​based
approach for at-​risk college women and adolescents
Theories and Models of Preventive that employs cognitive-​behavioral strategies shown
Interventions to be effective in reducing symptoms in individu-
As discussed in the preceding text, the overall als with clinical EDs (Clarke et al., 2001; Fairburn
theory of prevention is that reduction in risk factors & Cooper, 1993; Horowitz & Garbe, 2006; Killen
reduces risk. Another set of theories has been used et al., 1993; Lynch et al., 2005; Mann et al., 1997;
to determine the nature of the intervention itself. Striegel-​Moore et  al., 2005; Wilfley & Cohen,
The risk factors chosen for a particular intervention 1997; Wilson & Fairburn, 1993) (discussed in more
partly determine the nature of the intervention. detail in what follows). Recently, Martinsen et  al.
Five general models guiding ED preventions are (2014) examined the effects of a 1-​ year school-​
summarized below. based intervention program, based mostly on social

250 Prevention
learning theory, to prevent the development of new & Roberts, 1997), has also been used to inform
ED cases among adolescent female and male elite feminist approaches to ED research in women (e.g.,
athletes. Sixteen schools were randomized to treat- Cash, 1997; Peterson, Grippo, & Tantleff-​Dunn,
ment or control. Among females, there were no new 2008; Tiggemann & Kuring, 2004), and approaches
ED cases in the intervention schools, while 13% at to prevention (Tiggemann, 2013). A feminist per-
the control schools had developed an ED. spective on ED prevention considers the impact
of power and privilege related to gender on the
Dissonance Theory individuals’ experience of social power, worth, and
Stice and colleagues (e.g., Stice et al., 2000; Stice, their body, with the goal being to transform social
Marti, Spoor, et al., 2008; Stice, Rohde, Gau, & systems toward equity and places where bodies are
Shaw, 2009; Stice, Rohde, Shaw, & Gau, 2011) have sites of rights, agency, and freedom (Piran, 2010).
developed interventions rooted in the theory that For example, Piran (1999) implemented a feminist-​
individuals become motivated to change their atti- informed prevention program for girls and boys in
tudes and behaviors, such as unhealthy expectations a ballet school that included: addressing gender as
about weight and appearance and disordered eating a higher level risk factor, linking gender to expe-
behaviors, when faced with messages that contradict riences in the body, transforming the social envi-
these very attitudes and behaviors (Aronson, 1980; ronment, and engaging participants in creating a
Festinger, 1957, 1962; Leippe, 1994). Thus, dissonance “local” critical understanding of their experience in
programs focus on providing participants with skills the body domain. The program was associated with
to counteract the abundant weight-​and appearance-​ reductions in weight and shape concerns; however,
related messages prevalent in adolescents’ daily lives a control group was not used.
(Stice et al., 2001; Stice, Marti, Spoor, et al., 2008;
Stice, Mazotti, Weibel, & Agras, 2000; Stice, Shaw, Media Literacy and Advocacy
Becker, & Rohde, 2008; Stice, Presnell, et al., 2007; Media literacy and advocacy interventions are
Stice, Trost, & Chase, 2003). These programs typi- based on the theory that the mass media plays a
cally involve at least two 1-​hour sessions and more major role in perpetuating ED risk and that both
often three or four sessions, using trained program gaining an understanding of this risk and develop-
leaders. Efficacy trials conducted by independent ing strategies to resist media messages will reduce
teams have shown that this approach reduces ED risk factors. The focus of these programs is to
risk factors, ED symptoms, functional impair- develop or enhance skills to resist social persuasion
ment, mental health service use, and ED onset (e.g., (i.e., messages about thinness), with the goal of
Halliwell & Diedrichs, 2014; Stice, Marti, Spoor, reducing internalization of media messages about
et al., 2008; Stice et al., 2011). Effectiveness trials have thinness and appearance (Stice & Agras, 1998).
also shown that dissonance-​ based prevention pro- Media literacy approaches appear to have some
grams produce similar effects when delivered under support in studies using pre-​and early-​adolescent
ecologically valid conditions and by endogenous pro- samples (Neumark-​ Sztainer, Sherwood, Coller,
viders or peer leaders (e.g., Becker et al., 2010; Stice & Hannan, 2000; Sherwood, Harnack, & Story,
et al., 2009; Stice et al., 2011). The dissonance-​based 2000; Wade, Davidson, & O’Dea, 2003; Wilksch
approach is discussed in more detail below. & Wade, 2009) and college women (Becker, Bull,
Schaumberg, Cauble, & Franco, 2008; Becker,
Feminist Theory Smith, & Ciao, 2005; Coughlin & Kalodner,
Feminist theory assumes that gender roles influ- 2006), although one study using a midadolescent
ence the onset of EDs and that helping young sample failed to find evidence for this approach
women become aware of gender role expectations (Wilksch, Durbridge, & Wade, 2008). Recently,
may reduce weight and shape concerns. In other Wilksch et al. (2015) compared three school-​based
words, such approaches focus on providing partici- programs and a no-​intervention control group. A
pants with skills to critically think about and eval- total of 1,316 grade 7 and 8 girls and boys (mean
uate gendered issues related to body image and on age = 13.2 years) across three Australian states were
teaching acceptance of and support for healthier randomly allocated to Media Smart (which focuses
norms for weight and appearance. Objectification on media literacy); Life Smart (which focuses on
theory, which states that women come to view their reducing risk factors for both obesity and EDs); the
bodies as objects to be looked at, focusing more on Helping, Encouraging, Listening and Protecting Peers
how their body looks than how it feels (Frederickson (HELPP) initiative; or control (usual school class).

Taylor, Fitzsimmons-Craft, Goel 251

Media Smart girls had half the rate of onset of clin- to females (vs. both sexes), offered to participants
ically significant concerns about shape and weight over age 15 (vs. younger participants), and delivered
compared with control girls at 12-​month follow-​ by professional interventionists (vs. endogenous
up, and Media Smart was the only program to show providers). Programs with body acceptance and
benefit on both disordered eating and obesity risk dissonance-​induction content and without psycho-
factors. McLean, Paxton, and Wertheim (2016) education were also associated with larger effects.
systematically reviewed the literature examining There are many limitations with this analysis
empirical relationships between levels of media that need to be considered in interpreting the data.
literacy and body dissatisfaction and disordered First, a meta-​analysis using effect sizes to compare
eating. Importantly, media-​ literacy-​
based inter- universal and targeted interventions is questionable,
ventions revealed improvements in media literacy given that the goals and expected outcomes/​change
constructs of influence of media, realism skepti- in risk would likely vary between universal and tar-
cism, awareness of media motives for profit, and geted programs, as no change in risk factors may be
improvements in body-​related variables, but not a positive outcome for universal programs but not
disordered eating (McLean et al., 2016). for targeted programs. For instance, if a measure of
mood was used to determine the effects of a univer-
Transdiagnostic sal depression prevention study, would we expect it
Transdiagnostic theories of EDs, or probably to change in the normal population? Second, meta-​
more accurately, transdiagnostic approaches to analyses can obscure important comparisons. For
prevention, assume that there are common risk example, in a study comparing dissonance theory
factors for EDs and other disorders or that inter- to psychoeducation (Becker, Smith, & Ciao, 2005),
vention strategies (such as mindfulness or cogni- the results were very similar between the two in
tive restructuring) can simultaneously address a terms of changes in body dissatisfaction, dieting,
number of issues, such as anxiety, depression, and and eating pathology. This controlled comparison
EDs. Johnson et al. (2016) examined the effects of contradicts some of the moderator analyses men-
a mindfulness-​based school-​based prevention pro- tioned above in which psychoeducation programs
gram on reducing anxiety, depression, weight/​shape were deemed less effective. Third, most universal
concerns, and well-​ being, finding no significant programs focus on younger students, where rates of
effects on any of these variables. Musiat et al. (2014) EDs are low and related risk factors may also be low,
also used a transdiagnostic approach to reduce com- such that it would be difficult to detect, particularly
mon mental health disorders in college students. in the short-​term follow-​up periods typical of ED
There were significant effects for anxiety and depres- prevention research, if programs effectively reduced
sion symptoms but not eating behaviors. Thus, the outcomes not expected to be high or even fully
effectiveness for transdiagnostic approaches remains manifest in a younger age group. A more mean-
unclear. ingful analysis would be to assess targeted/​selective
versus indicated interventions within a population.
Effectiveness of Prevention Programs Overall, however, the results of the meta-​analysis—
Several meta-​ analyses and reviews (Beintner, ​that prevention programs can reduce ED risk fac-
Jacobi, & Taylor, 2012; Ciao, Loth, & Neumark-​ tors or current or future eating pathology—​are in
Sztainer, 2014; Loucas et  al., 2014; Pratt & concordance with several long-​ term controlled
Woolfenden, 2002; Stice & Shaw, 2004; Stice, studies and other meta-​analyses and reviews.
Shaw, et al., 2007; Watson et al., 2016) have pro- Another way to examine the effects of preven-
vided evidence that some types of prevention pro- tion programs is to look at the numbers needed to
grams are effective for reducing ED risk factors treat (NNTs) to prevent the onset of a new ED. For
or ED symptoms themselves. In an early but very primary outcome analyses of reducing ED onset,
comprehensive meta-​ analysis, Stice, Shaw, et  al. only two studies (both discussed above) have dem-
(2007) found that of the 51 ED prevention pro- onstrated that a preventive intervention yields a
grams evaluated, 51% resulted in significant reduc- significant reduction in rate of onset of EDs (Stice,
tions in at least one risk factor for eating pathology, Marti, Spoor, et al., 2008; Martinsen et al., 2014).
and 29% resulted in significant reductions in eating In the Stice, Marti, Spoor, et al. (2008) study, rates
pathology. Larger effects occurred for programs that of ED onset were reduced from about 16% in the
were selected (vs. universal), interactive (vs. didac- control group to about 7% in the treatment groups,
tic), multisession (vs. single session), only offered with the NNT equaling about 11. In the Martinsen

252 Prevention
et al. (2014) study, among females, there were no conducted using older adolescents and college stu-
new cases of ED in the intervention schools, while dents have largely reduced ED risk factors rather
13% at the control schools had developed and ful- than increased them.
filled the DSM-​IV criteria for ED not otherwise
specified (EDNOS) or BN, with the NNT equal- Moderators and Mediators of Eating
ing about 8. Taylor et al. (2016) found that ED Disorder Prevention Programs
onset rate was 27% lower in the intervention versus Examination of moderators and mediators in
the control group (NNT = 15), but this difference prevention studies is important for the progress and
was not significant. Importantly, the authors found future success of prevention programs. Moderators
that in the group at highest risk (i.e., with the study refer to study or participant characteristics present
entry criteria plus EDE-​Q Shape Concern scores at baseline (e.g., overweight status, risk status, age,
>3.0), ED onset rate was significantly lower in the gender, race/​ethnicity, program format) that predict
intervention group, with the NNT equal to 5. differential response to an intervention. Mediators
refer to process variables that should change prior to
Is Prevention Harmful? noted change in the outcome (e.g., thin-​ideal inter-
An important consideration to address when nalization) and that thus provide information on
discussing ED prevention programs is the concern mechanisms of change. Hence, moderators specify
that such programs may inadvertently be harmful, for whom or under what conditions the prevention
as opposed to beneficial, for participants. (This is program works, and mediators identify mecha-
a different issue as to whether or not public health nisms through which a prevention program might
obesity/​weight loss programs may confer some risk achieve its effects (Kraemer, Wilson, Fairburn, &
for increasing EDs, as discussed previously.) This Agras, 2002).
concern stems from two early prevention research As noted by Stice, Shaw, et  al. (2007), there is
studies in which results indicated an increase in ED consistent evidence that participant characteristics
risk factors over the course of the study (Carter, do moderate outcomes, with stronger effects for
Stewart, Dunn, & Fairburn, 1997; Mann et al., programs targeting individuals who were at higher
1997). However, a meta-​analysis later concluded risk and age 15 or older. This is likely because the
that there is no evidence that prevention programs inclusion of low-​risk participants and/​or younger
are harmful (Pratt & Woolfenden, 2002). Many participants reduces the ability to detect statistically
prevention studies have been conducted since, and significant and clinically meaningful changes in
there is no evidence that they are associated with eating pathology due to already low levels in these
adverse effects. groups. In addition, older adolescents likely benefit
There has also been some concern (e.g., from more from current prevention programs, as these
human subjects committees) that asking young programs often focus on skills, such as cognitive
adolescents about “ED behaviors” might be harm- dissonance (CD) and media literacy, that require
ful in that it would expose adolescents to attitudes more advanced cognitive skills that younger indi-
or behaviors they had previously not considered or viduals may not have fully developed. It has also
heighten their focus on weight and shape issues. In been hypothesized, but not empirically examined,
response to this concern, Celio, Bryson, Killen, and that individuals at risk for ED onset, which includes
Taylor (2003) compared results from 115 sixth-​ older adolescents, may be more motivated to par-
grade girls who responded to questions on risky ticipate and engage in prevention program material,
weight control behaviors and attitudes at baseline leading to greater effects for targeted preventions
and at 12-​ month follow-​ up with the responses and better outcomes for high-​risk subgroups within
of 107 girls who had not been part of the base- universal programs. Stice, Shaw, et  al. (2007) also
line assessment. Results revealed no differences in found that programs targeting female-​only popu-
scores between the two groups on the “follow-​up” lations were more effective but only for body dis-
assessment, and rates of unhealthy weight regula- satisfaction and dieting; there were no significant
tion behaviors decreased over time in the group differences in effects between female-​ only and
assessed on two occasions. Thus, there is no empiri- mixed-​sex programs for BMI, thin-​ideal internal-
cal support suggesting that surveys of ED risk fac- ization, negative affect, or overall eating pathology.
tors and behaviors increase risk for such outcomes. Thus, while female-​only programs may be gener-
Further, and of critical importance to countering ally more effective, this appears to depend on the
these concerns, the many large prevention trials outcome assessed. As with programs focusing on

Taylor, Fitzsimmons-Craft, Goel 253

high-​risk and older samples, the larger effects for with higher baseline BMI or lower baseline purg-
female-​only programs may be due to the greater ing and that effects on ED pathology were weaker
level of eating disturbance associated with females for participants with higher baseline purging
and to higher levels of motivation to participate or restrictive eating. No moderators of effect on
in ED prevention programs. In regard to program restrictive eating were identified. In a review of tri-
characteristics that moderate outcomes, there is als of the StudentBodies program, Beintner, Jacobi,
some evidence to suggest that interactive program- and Taylor (2014) found that after adjusting for
ming (as opposed to didactic) is more effective, as program adherence, intervention outcomes were
are programs delivered by trained interventionists only moderated by participant’s age, with smaller
(Stice, Shaw, et  al., 2007). Additionally, programs effects in one sample of adolescents. Thus, these
that promote body acceptance and CD skills appear more recent findings are inconsistent with past
more effective than programs without such skills, work on intervention moderators. Future research
while programs that focus on more general risk fac- should therefore focus on replication of existing
tors without a focus on specific EDs risk factors, findings using comparable measures so as to work
such as programs teaching coping skills or self-​ toward building a more consistent and complete
esteem promotion, seem to be less effective. picture of moderators in ED prevention. Finally,
However, while meta-​analyses provide an over- Manwaring et  al. (2008) investigated the pos-
all description of moderators as they pertain to sibility that adherence to specific StudentBodies
the field of ED prevention, relatively few studies components might mediate outcome, finding that
directly examine moderators, limiting our ability total weeks participation and frequency of using
to determine if (and which) programs are most the online pages/​ journals predicted changes in
effective for specific populations. Taylor et  al. restraint but not other ED symptoms, but that in
(2006) did examine moderators of StudentBodies participants with some compensatory behaviors,
and found that the program was most effective for discussion board and booster session use was actu-
individuals with BMIs greater than 25 at baseline ally associated with lower weight/​shape concerns
and, at one study site, that the program was more during follow-​up.
effective for women with baseline compensatory Stice, Presnell, and colleagues (2007) exam-
behaviors. ined mediators of the CD and healthy weight
Stice, Marti, Shaw, et al. (2008) also examined programs, finding some evidence to support thin-​
moderators of both their CD and healthy weight ideal internalization as a partial mediator for the
regulation programs and found support for two CD program. Of interest, these researchers noted
general program moderators (baseline level of risk as that in about one-​third of their CD participants,
indicated by body image distress and bulimic symp- change in thin-​ideal internalization occurred after
toms for both programs) and for program specific change in the outcome measures and that change
moderators, including thin-​ideal internalization for in the healthy weight mediators was inconsistently
the CD program and emotional eating and body associated with outcomes, suggesting that contin-
mass for the healthy weight regulation program. ued examination of mediators, and elaboration on
Overall, these findings reported by Taylor et  al. the expected change pathways/​causal pathways is
(2006) and Stice et  al. (Stice, Marti, Shaw, et  al., needed, as additional mediators likely contribute to
2008; Stice, Shaw, et al., 2007) suggest that initial change in the ED risk factor outcomes. As noted by
elevations in general ED risk factors, in a popula- the researchers, examination of demand character-
tion that may be more motivated to change, and in istics may also provide insight into the inconsistent
heightened levels of program-​specific/​target risk fac- findings. Overall, as concluded by Stice, Presnell,
tors moderate outcomes. These findings are of note et al. (2007) these findings have important implica-
given that both studies found moderating effects for tions for prevention programs in that although both
bulimic behaviors (e.g., compensatory behaviors) programs had program-​ specific mediators, both
and for elevated weight status, suggesting that it is programs did result in reduction of ED risk factors
possible to reduce onset of EDs in more than one in the long-​term. This lends support to the notion
high-​risk group. that different pathways may be involved in the onset
However, Völker, Jacobi, Trockel, and Taylor of EDs and that future programs may want to find
(2014) examined moderators and mediators of the ways to maximally target more than one risk path-
StudentBodies program, finding that intervention way in order to achieve even greater reduction of
effects on binge eating were weaker for participants risk factors and ED onset.

254 Prevention
Examples of Effective Approaches findings, particularly given that healthy weight par-
In the following section we discuss three pro- ticipants in the second study showed improvements
grams:  a CD program (the Body Project), an along with CD participants. In this study, Stice and
Internet-​based program (StudentBodies), and peer/​ colleagues (2003) used a larger sample and longer
school-​based prevention programs (e.g., Healthy follow-​up periods; participants were also somewhat
Schools, Healthy Kids [HS-​ HK]) that have been younger (mean age 17) than those in the previous
shown, using long-​term data (e.g., at least 1-​year studies (mean ages 18 and 19, respectively) in order
long-​term follow-​up data provided) from controlled to deliver the program at a time of peak ED onset.
clinical studies with sufficiently large samples Findings again provided support for CD, with
needed to have adequate power to detect differ- participants in CD evidencing greater short-​term
ences. We also discuss an emerging area of preven- (post-​test) reduction in thin-​ideal internalization,
tion programming that targets both the reduction negative affect, bulimic symptoms, and body dis-
of EDs and overweight and review recent work sup- satisfaction than healthy weight controls and greater
porting such programs. long-​term (6-​month) reduction in body dissatisfac-
tion and thin-​ideal internalization in comparison
Cognitive Dissonance Programs to healthy weight controls. Both active conditions
For the past 15 years, Stice and colleagues have showed greater short-​and long-​term reductions in
been developing, evaluating, and refining a CD pro- bulimic symptoms and negative affect relative to the
gram, the Body Project, designed to reduce thin-​ wait-​list control condition (Stice et al., 2003).
ideal internalization and other ED risk factors (e.g., Additional efficacy trials have demonstrated that
body dissatisfaction, negative affect) in females CD produces greater reductions in ED risk factors
who indicate body image concerns. As previously and symptoms, impairment, mental health service
described, CD programs are relatively short-​term use, and ED onset over 3-​year follow-​up relative
(two to four 1-​hour sessions) and focus on reduc- to assessment-​ only control conditions and three
ing thin-​ideal internalization using dissonance tech- alternative interventions. For example, Stice, Marti,
niques that require participants to take standpoints Spoor, et al. (2008) compared their CD program
that are counter to their beliefs; over time, in order to a healthy weight program, an expressive writing
to reduce the distress associated with supporting program, and assessment-​only control condition in
an opinion counter to their original beliefs, par- a sample of adolescent girls with body dissatisfac-
ticipants change their beliefs (e.g., less adherence to tion. Findings from the 1-​year, 2-​year, and 3-​year
and internalization of the thin ideal). follow-​up studies revealed support for the CD pro-
The first CD study from Stice and colleagues gram. Of note was that, compared with the assess-
evaluated three 1-​hour sessions of CD administered ment-​only group, there was a 60% reduction in
to at-​risk college women as compared with a wait-​ ED onset for CD participants (6% onset in CD
list control group, with findings indicating that group vs. 15% in assessment only controls) through
CD participants had greater decreases in ED risk 3-​year follow-​up. Randomized controlled trials from
factors, including body dissatisfaction, thin-​ ideal other research groups further established the short-​
internalization, bulimic symptoms, and negative term efficacy of CD in reducing ED pathology in
affect (Stice et al., 2000). A second study included high risk and unselected samples of college women
an active control condition as well as a wait-​list con- (e.g., Becker, Smith, & Ciao, 2005; Green, Scott,
trol condition in order to verify that the previous Diyankova, Gasser, & Pederson, 2005; Roehrig,
findings were not due to demand characteristics; Thompson, Brannick, & van den Berg, 2006).
this active control group included healthy weight Effectiveness trials have also shown that CD
regulation materials that focused on reducing body produces similar effects when delivered by endog-
image concerns by providing healthy weight control enous providers under real-​world conditions. For
skills. Findings again provided support for CD with example, Stice et al. (2011) demonstrated that CD,
participants in the CD condition evidencing greater implemented by school-​ based counselors, pro-
reductions in thin-​ideal internalization and body duced significantly greater decreases in body dis-
dissatisfaction in comparison to wait-​list and healthy satisfaction at 2-​year follow-​up and ED symptoms
weight control participants and greater reductions at 3-​year follow-​up than an educational brochure
in dieting, bulimic symptoms, and negative affect condition among female high school students with
in comparison with wait-​list controls (Stice et  al., body image concerns. College-​aged peer leaders can
2001). A third study was designed to replicate these also be trained to lead the intervention (e.g., Becker,

Taylor, Fitzsimmons-Craft, Goel 255

Bull, Schaumberg, Cauble, & Franco, 2008; Becker, psychoeducational programs (Celio et  al., 2000;
Smith, & Ciao, 2006; Becker et  al., 2010). These Springer et al., 1999; Winzelberg et al., 2000).
effectiveness studies report significant reductions These studies suggested that an Internet-​based
in ED risk factors and pathology, although effect program can reduce risk factors in college-​ age
sizes are typically smaller compared to trials where women and set the stage for a more ambitious
research staff implement the intervention. To address study. In the next study, 480 college-​age women
this issue, an enhanced version of CD was recently were randomized to StudentBodies or a wait-​ list
created, where college clinicians receive enhanced group and were followed for 3 years (Taylor et al.,
training and supervision, and it was shown to pro- 2006). There was a significant reduction in WCS
duce larger intervention effects at 1-​year follow-​up scores in the StudentBodies group compared with
compared with previous effectiveness studies (Stice, the control group at post-​intervention (p < .001),
Butryn, Rohde, Shaw, & Marti, 2013). 1 year (p < .001), and 2 years (p < .001). The slope
Cognitive dissonance, through the Body Project for reducing WCS score was significantly greater
manual (Stice, Rohde, & Shaw, 2012), is now in the treatment compared with the control group
widely available, and intervention materials are (p = .02). Over the course of follow-​up, 43 partici-
also available online (www.bodyprojectsupport. pants developed subclinical or clinical EDs. While
org). Indeed, the program appears highly dissem- there was no overall significant difference in onset of
inable and was adopted for use within sororities on EDs between the intervention and control groups,
college campuses across the United States through the intervention significantly reduced the onset of
the Reflections program, a peer-​led CD program EDs in two subgroups identified through modera-
that can be delivered in two 2-​hour sessions (Perez, tor analyses: (1) participants with an elevated BMI
Becker, & Ramirez, 2010). Other exciting adapta- (≥ 25, calculated as weight in kilograms divided
tions of CD have emerged in recent years as well, by height in meters squared) at baseline and (2) at
including a version of the Body Project that is deliv- one site, participants with baseline compensatory
ered over the Internet (eBody Project). The eBody behaviors (e.g., self-​induced vomiting, laxative use,
Project also reduces ED risk factor and symptoms diuretic use, diet pill use, driven exercise). No inter-
and weight gain prevention effects that persisted vention participants with an elevated baseline BMI
through 2-​year follow-​up (Stice, Rohde, Durant, & developed an ED, while the rates of onset of ED in
Shaw, 2012; Stice, Durant, Rohde, & Shaw, 2014). the comparable BMI control group (based on sur-
vival analysis) were 4.7% at 1  year and 11.9% at
Stanford/​Washington University 2 years. In the subgroup with a BMI of 25 or higher,
StudentBodies™ Studies the cumulative survival incidence was significantly
During the past 20 years, researchers at Stanford lower at 2 years for the intervention compared with
and Washington University in St. Louis have the control group (95% confidence interval, 0%
been exploring issues related to prevention using for intervention group; 2.7% to 21.1% for con-
Internet-​based, psychoeducational, interactive pro- trol group). For the San Francisco Bay Area site
grams. In an early study from this group, Killen sample with baseline compensatory behaviors, 4%
et al. (1993) evaluated the effectiveness of a preven- of participants in the intervention group developed
tion curriculum designed to modify the eating atti- EDs at 1 year and 14.4% by 2 years. Rates for the
tudes and unhealthful weight regulation practices comparable control group were 16% and 30.4%,
of young adolescent girls. Nine hundred sixty-​seven respectively. In sum, this study demonstrated
6th-​and 7th-​grade girls were randomized to experi- that, among college-​age women with high weight
mental healthy weight regulation curriculum or and shape concerns, an 8-​ week, Internet-​ based
no-​treatment control classes. There was a significant cognitive-​behavioral intervention can significantly
increase in knowledge among girls receiving the reduce weight and shape concerns for up to 2 years
intervention but no overall effect. This study and a and decrease risk for the onset of EDs, at least in
subsequent longitudinal prospective study in older some high-​risk groups.
adolescents demonstrated that students with higher Jacobi et  al. (2011) further analyzed the data
scores on a measure of weight and shape concerns from Taylor et al. (2006) and identified three fac-
were at risk for developing EDs. tors that moderated the intervention in individuals
The group then began to focus on targeted/​ with high weight and shape concerns: a history of
selective programs. Early efforts along this line being teased, current or lifetime depression, and/​
demonstrated the effectiveness of computer-​based or nonclinical levels of compensatory behaviors.

256 Prevention
Taylor et  al. (2016) then examined the effects of et al. (2005) provided online assessment and feed-
StudentBodies, adapted to specifically address these back to 174 10th-​grade students based on weight/​
three risk factors (based on Taylor et  al., 2006). shape concerns and overweight risk. The algorithm
Four hundred thirty-​ nine college and university identified 111 no-​risk (NR), 36 ED risk (EDR),
students with one or more of these risk factors were 16 overweight risk (OR), and 5 both risks. Fifty-​
randomized to the intervention or wait-​list control six percent of the EDR and 50% of the OR groups
and followed for up to 2  years. The ED attitudes elected to receive the recommended targeted curri-
and behaviors improved more in the intervention cula. Significant improvements in weight and shape
than control group (p  =  .02, d  =  .31). Although concerns were observed in all groups, and the study
ED onset rate was 27% lower in the intervention demonstrated that an Internet-​delivered program
group, this difference was not significant (p = .28, can be used to assess risk and provide simultaneous
NNT = 15). In the subgroup with very high weight universal and targeted interventions in classroom
shape concerns, about 50% of the sample, ED onset settings.
rate was significantly lower in the intervention than In the meantime, the protocol has been expanded
control group (20% vs. 42%, p = .025). For the 27 to include boys and to address overweight (Jones
individuals with depression at baseline, depressive et  al., 2008). In a pilot study, 100 8th-​grade boys
symptomatology improved more in the interven- and girls were sorted by an Internet-​ based algo-
tion than control group (p = .016, d = .96). Thus, rithm into two groups based on risk for developing
this study suggests that StudentBodies might be most an ED; male students participated as one general
effective when applied to subgroups of individuals group. Participants in each group were also assigned
who have very high weight and shape concerns and to an online discussion group that corresponded
additional attributes. to their group assignment and were encouraged to
In a companion study, Kass et al. (2014) sought post messages to group members of similar risk. All
to determine whether adding “guided self-​help” three groups showed significant increases in knowl-
to StudentBodies provided to students with high edge related to the program content and reported
weight and shape concerns, but at lower risk than increased physical activity levels from pre-​to post-​
those identified for the Taylor et  al. (2016) study, intervention. Females in the high-​risk group also
improved outcomes. Women with high weight/​ showed significant reductions in weight and shape
shape concerns (N  =  151) were randomized to concerns. Participants were enthusiastic about using
StudentBodies with a guided discussion group the online health program; almost all reported that
(n = 74) or no discussion group (n = 77). Regression they would prefer an online format to a traditional
analyses showed weight/​ shape concerns were classroom format. Another study showed that a
reduced significantly more among guided discus- paper-​and-​pencil program designed to reduce family/​
sion group than no discussion group participants parental critical comments about eating and shape
(p  =  .002; d  =  .52); guided discussion group par- resulted in reduced critical comments from par-
ticipants had 67% lower odds of having high-​risk ents (by their report) (Bruning-​Brown, Winzelberg,
weight/​shape concerns post-​intervention (p = .02). Abascal, & Taylor, 2004). Taken together, these stud-
There were no differences in binge eating at post-​ ies suggest that universal and targeted prevention
intervention between the two groups, and no mod- programs can be provided simultaneously and might
erators emerged as significant. Results suggest the benefit from involvement of education programs
guided discussion group improves the efficacy of aimed at parents/​families.
StudentBodies in reducing weight/​shape concerns in In Germany, Jacobi and her group translated
college students at high risk for an ED. StudentBodies into German and adapted media
In other studies, the Stanford-​ Washington more relevant to the German population (Jacobi
University in St. Louis research group returned et  al., 2007). Because many of the American
to the issue of how to provide both universal and StudentBodies programs have also been provided to
targeted interventions. The goal was to find ways German populations and examined in controlled
to provide general healthy weight regulation pro- studies, Beintner et  al. (2008) undertook a meta-​
grams to all high-​school age students while pro- analysis of studies using StudentBodies done in
viding targeted interventions to students at higher the United States and in Germany. Pre–​post data
risk. Following a study by Abascal et al. (2004) that from these programs across a large number of eat-
showed that students could be successfully allocated ing pathology variables resulted in effect sizes in the
to various risk groups using an online program, Luce moderate range for most variables. The respective

Taylor, Fitzsimmons-Craft, Goel 257

controlled effect sizes for the German and US stud- McVey and colleagues (2004) examined a six-​
ies were:  drive for thinness, 0.42, 0.35; bulimia, session life-​skills program focused on promoting
0.22, 0.29; restraint, 0.30, 0.26; WCS, 0.58, 0.36. self-​esteem and improving body image satisfaction
This analysis suggests that programs with similar and global self-​esteem, with the additional goals
content and format, addressing similar populations of reducing negative attitudes shown to be associ-
in Germany and the United States, have similar ated with risk for EDs, including thoughts related
outcomes. to unhealthy eating and perfectionism. Participants
Overall, data from this group has demonstrated included 258 preadolescent girls (n  =  182 in pre-
that an Internet-​based program can reduce risk fac- vention group; mean age = 11.8). Findings revealed
tors for high-​risk individuals and even reduce onset short-​term (pre–​post) reductions in unhealthy diet-
of EDs. The Stanford-​ Washington University in ing attitudes and improvement in self-​esteem and
St. Louis group has since combined the prevention body satisfaction; these changes were not main-
programs into an integrated universal, targeted, tained at the long-​ term (12-​month) follow-​up.
indicated prevention/​intervention for EDs (Wilfley This work was followed by two additional studies,
et al., 2013). again focused on young girls in grades 7 and 8,
who participated in 10-​session support/​discussion
Peer Support/​School-​Based Programs groups facilitated by a school nurse and by trained
Following on several key aspects of ecologi- peers. Sessions included more focused discussion
cal and social systems theories, in which inclusion related to healthy body image messages, promoting
and targeting of critical social relationships and healthier norms within peer groups, and combating
interactions is indicated in order to achieve endur- unhealthy appearance messages and expectations.
ing change on ED risk factors, several studies have Findings from these two studies revealed inconsis-
focused on the school and peer environments. For tent support for this program as one study with 214
example, Becker and colleagues (Becker et al., 2005, girls found decreases in dieting and improved body
2006, 2008) have worked with sorority systems, esteem over time for program participants but not
which are believed to be high-​risk populations, to in control participants (McVey et al., 2004), while
develop and examine the effectiveness of peer-​led another similar replication study involving 282 girls
media advocacy (MA) and CD programs on the risk failed to find differences in improvements on these
for EDs in sorority members. The focus on peer-​led outcomes between the control and intervention
programs at the college level stems from research groups over time (McVey, Lieberman, Voorberg,
indicating that peers contribute to body image satis- Wardrope, Blackmore, & Tweed, 2003).
faction and, more broadly, from the notion that col- A large-​scale study involving 982 middle school
leges tend to promote peer leadership, particularly students expanded on this earlier work by including
in organized groups such as those in the athletic males and including a focus on teachers and parents
domain or in the sorority and fraternity systems. (McVey, Tweed, & Blackmore, 2007). Materials
Findings provide support for both the CD and MA were provided not only to students but also to
peer-​led programs for high-​risk females and support teachers, and to some extent, parents, with the idea
for the peer-​led CD program for both high-​and that it is important to address and reduce teachers’
low-​risk females. These findings provide particular and parents’ own unhealthy eating and weight atti-
support for a peer-​led CD program, as this appears tudes and perceptions so that they may more effec-
to have a positive effect on both low-​and high-​risk tively address and, over time, combat unhealthy
participants. attitudes and behaviors, including teasing, in stu-
McVey, Davis, Tweed, and Shaw (2004) have dents. Of note was that this program also included
also developed and examined a universal, school-​ a focus on males, as these researchers noted that
based program that draws from several ecological males are indeed at growing risk for ED attitudes
approaches to reducing risk behaviors. In these and behaviors, and due to evidence that males are
studies, described below, an ecological approach more likely than females to initiate teasing (Stein,
was used to address the multiple systems (e.g., 1999). Findings revealed that the program was asso-
teachers, peers, parents) that may influence risk for ciated with lower internalization of media messages
ED onset in youth. The program was implemented for males and females, decreases in the number of
in a number of different classes throughout the aca- students trying to lose weight, and decreased dis-
demic year (e.g., math, drama, English, and health ordered eating scores for female students, although
classes). effects on weight loss behaviors did not persist at

258 Prevention
6-​month follow-​up. Analyses on a high-​risk group, studies and reviews that indicate that overweight
defined as students currently trying to lose weight, children and adolescents may already be endors-
revealed a significantly greater decrease in body dis- ing unhealthy eating attitudes and behaviors, some
satisfaction, media internalization, and disordered have argued that the attention put on reducing and
eating for these students in comparison to low-​risk maintaining weight, particularly in those in the
students; of note was that low-​risk students did not higher BMI ranges, has increased the rate of EDs.
evidence much increase in body dissatisfaction over In addition, others feel that the message that the
time, which is important as such dissatisfaction majority of young people need to remain “thin” or
may have been expected to increase as students were be thinner reinforces the thin-​body ideal and would
entering puberty. Findings, however, failed to show seem to counter the general ED prevention message
a reduction in weight-​related teasing or improve- that self-​esteem should not be determined by body
ments in body size acceptance. Additional analyses size (Neumark-​Sztainer, 2005).
also failed to reveal program influence on teachers’ However, despite these concerns, it is becom-
perceptions of the school environment or on their ing critical to comprehensively respond to the high
own behaviors and attitudes; a number of reasons, prevalence of obesity, EDs, and disordered eating
including time and other demands on teachers’ behaviors among youth, and researchers in both
abilities to learn and deliver program material may the obesity and ED fields have proposed using an
have contributed to the lack of findings for teachers, integrated approach to prevention that addresses
as well as the limited findings for student outcomes the spectrum of weight-​ related disorders within
such as teasing. interventions (Haines & Neumark-​Sztainer, 2006;
Overall, despite the limited findings for some Kass, Jones, et al., 2017). The identification of risk
outcomes and the lack of intervention effect for factors that are shared between these weight-​related
teachers reported by McVey et al. and the somewhat disorders is an essential step to developing effec-
inconsistent effects on disordered eating across tive prevention interventions. Neumark-​ Sztainer
studies (McVey et al., 2004; McVey, Lieberman, et al. (2007) provide preliminary support for the
Voorberg, Wardrope, & Blackmore, 2003; McVey, existence of shared risk factors for obesity and EDs,
Lieberman, Voorberg, Wardrope, Blackmore, & with recent empirical research supporting this con-
Tweed, 2003; McVey et al., 2007), these studies tention. Specifically, the authors examined and
and the work from Becker and colleagues (Becker found preliminary evidence that dieting, media use,
et al., 2005, 2006, 2008) suggest that programs body image dissatisfaction, and weight-​related teas-
that include peer-​ based and/​ or school-​ based ing may have relevance for the development of the
approaches to ED prevention may be important spectrum of weight-​related disorders. Future etio-
steps toward reducing risk for EDs in several sys- logical research designed to specifically test these
tems that influence attitudes and behaviors (e.g., and other potentially shared risk factors across dif-
peer and school environments). These programs ferent age and racial/​ethnic groups remains war-
may be particularly appealing to institutions that ranted and would provide important insights into
want to deliver effective universal programs given the relevant factors to be addressed in interventions
that Becker et al. found equal improvements for aimed at preventing a broad spectrum of weight-​
low-​and high-​risk participants who received the related disorders.
CD program (Becker et al., 2006) and McVey Several preventive interventions have been devel-
et al. (2007) found equal effects on many out- oped based on addressing both ED and obesity
comes for males and females. prevention. Austin and colleagues (Austin, Field,
Wiecha, Peterson, & Gortmaker, 2005; Austin et al.,
Eating Disorder and Obesity Prevention 2007) evaluated the effects of a program designed
In some, but not all studies, elevated weight sta- to promote healthful nutrition and physical activity
tus has been associated with increased risk of onset on disordered weight-​control behaviors in early ado-
of BN and binge ED (BED; Jacobi, Hayward, et al., lescent girls and boys (Planet Health). Austin et al.
2004); weight status has also been linked to disor- (2007) randomized 749 girls and 702 boys in grades
dered eating attitudes and behaviors in young chil- 6 and 7 from 13 middle schools to the intervention
dren and adolescents (Goldschmidt, Aspen, Sinton, (5-​2-​1 Go!). At follow-​up in girls, there was a signifi-
Tanofsky-​Kraff, & Wilfley, 2008), yet most adoles- cant effect for disordered weight-​control behaviors
cent obesity prevention programs do not assess eat- for girls, with 3.6% (15 of 422) of girls in control
ing pathology (Carter & Bulik, 2008). Despite these schools compared with 1.2% (4 of 327) of girls

Taylor, Fitzsimmons-Craft, Goel 259

in intervention schools reporting engaging in disor- target the unique pressures associated with high-​
dered weight-​control behaviors (p = .04). No inter- risk activities, such as ballet, gymnastics, cheerlead-
vention effect was observed in boys. Other studies ing, and other sports. Furthermore, it should be
also provide support for targeting both overweight noted that research has demonstrated that certain
and ED risk through one program. Jones et al. racial and cultural groups can be particularly at-​risk
(2014) provided a program to 9th-​grade students of developing EDs. In this section we review some
which had two tracks for universal and targeted preventive activities in these domains.
delivery and was designed to enhance healthy liv-
ing skills, encourage healthy weight regulation, and Dance
improve weight/​shape concerns among high school Piran (1999) developed and evaluated a program
adolescents. Three hundred thirty-​six 9th-​grade stu- aimed at preventing EDs in a world-​class, residen-
dents in two high schools in the San Francisco Bay tial ballet school for female and male students ages
area and in St. Louis were invited to participate. 10 to 18. The focus of the intervention was on the
Students who were overweight (BMI > 85th per- school environment, which was reengineered to
centile) were offered the weight management track; allow students to “feel” comfortable with the pro-
students who were normal weight were offered the cesses of puberty and growth and to promote their
healthy habits track. The BMI percentile and zBMI right to feel both safe and positive in their diverse
significantly decreased among students in the weight bodies. The focus of body shape was replaced with
management track. The BMI percentile and zBMI an emphasis on stamina and body condition.
did not significantly change among students in the Teachers were prohibited from making evaluative
healthy habits track, demonstrating that these stu- comments about body shape to students, and staff
dents maintained their weight. Weight/​shape con- members were available to help students regarding
cerns significantly decreased among participants in concerns about body shape. The effectiveness of the
both tracks who had elevated weight/​shape concerns program was examined by comparing scores from
at baseline. The results are promising but need to two cohorts of 7th-​to 9th-​and 10th-​to 12th-​
be replicated and examined in controlled trials. Stice grade students who had been in the school follow-
and colleagues (Stice, Marti, Spoor, et al., 2008; ing implementation of the program with an earlier
Stice et al., 2006) in their studies of CD also found cohort that had not been exposed to the program.
support for their healthy weight regulation program Scores on two items measuring weight and shape,
as an effective ED prevention approach. While this the EAT and EDI, were lower in the intervention
program was originally intended to serve as an active cohorts compared to the baseline cohort. These data
control condition only, participation in the healthy provide evidence that a schoolwide intervention
weight program was found to be associated with a can be effective in reducing ED risk in a high-​risk
55% reduction in the onset of overweight and 61% environment.
reduction of ED onset in comparison to the assess-
ment-​only control group and to be associated with Cheerleading
reduction in other ED risk factors, such as bulimic Whisenhunt, Williamson, Drab-​ Hudson, and
symptoms; these findings are in concordance with Walden (2008) targeted cheerleading coaches as
earlier studies suggesting that healthy weight skills potential change agents by training them to rec-
may indeed promote a reduction in ED risk (Killen ognize the symptoms of EDs and reduce the pres-
et al., 1996; Matusek et al., 2004; Stice et al., 2003). sures for thinness among their squads. At 8-​month
Thus, a clear direction for future studies would be to follow-​up, there was self-​
reported improvement
evaluate a CD program that includes healthy weight in coaches’ behavior, but no sustained knowledge
regulation material. Taken together, these stud- about EDs.
ies suggest that interventions designed to promote
healthy weight regulation and/​or prevent EDs may Athletes
help with weight maintenance and do not seem to As a subgroup, some types of athletes (e.g.,
be harmful. swimmers, gymnasts, figure skaters) are much more
likely to develop EDs than the regular population
Settings and Special Populations (Bonci et al., 2008). Athletes may be at particular
The apparent association between ED risk and risk because of the pressure from coaches and others
certain activities such as sports has been long noted, to lose weight or maintain an often very low weight
with some programs having been developed to for real or perceived need to improve performance

260 Prevention
(Bratland-​Sanda & Sundgot-​Borgen, 2012). Bonci (Meyer, 2003). Research has demonstrated that in
et al. (2008) estimated that 62% of female athletes comparison to heterosexual males, self-​ identified
and 33% of male athletes are affected by disordered bisexual and gay men tend to exhibit higher rates
eating. Martinsen and Sundgot-​ Borgen (2013) of body image dissatisfaction (M. Morrison,
assessed the prevalence of EDs among adolescent T. Morrison, & Sager, 2004), ED behaviors (Austin
elite athletes in comparison to nonathletes and et al., 2009), and EDs (Feldman & Meyer, 2007),
determined that there was a significantly higher esti- however, the prevalence rates for these domains for
mated prevalence of ED among the female athletes lesbian and bisexual females in comparison to het-
at 14% in comparison with female nonathletes at erosexual females is less clear (Feldman & Meyer,
5.1% and an estimated rate of prevalence of 3.2% 2007; Wichstrom, 2006). Brown and Keel (2015)
for the male athletes (data on the males was lim- investigated the acceptability and efficacy of a 2-​
ited due to the relatively low number of cases of session CD intervention (the PRIDE Body Project)
ED identified in the male sample). Martinsen et al. compared to wait-​ list control in reducing ED
(2014) demonstrated that an intervention to reduce risk factors among gay college males. Participants
ED risk was effective among the female athletes, and found the intervention highly acceptable, and the
subsequently, Martinsen et  al. (2015) developed a CD intervention was associated with significantly
program to teach coaches how to identify and pre- greater decreases in a variety of ED risk factors from
vent EDs. Coaches taking the course demonstrated pre-​to post-​intervention, with nearly all improve-
greater knowledge about these issues compared with ments maintained at 4-​week follow-​up.
coaches who did not take the course.
Buchholz, Mack, McVey, Feder, and Barrowman Racial/​Ethnic Minorities
(2008) evaluated the effectiveness of a selective Researchers have determined that EDs can affect
prevention program designed to reduce pressures individuals across various racial and ethnic strati-
to be thin in sports, and to promote positive body fications (Shaw, Ramirez, Trost, Randall, & Stice,
image and eating behaviors in young female ath- 2004). Shuttlesworth and Zotter (2011) deter-
letes belonging to gymnastic clubs. The interven- mined that low levels of ethnic identity represented
tion focused on competitive female gymnasts (ages a risk factor for the development of binge eating
11–​18 years), parents, and coaches. Four clubs were and bulimic pathology in Black women, while high
randomized to receive a 3-​month intervention pro- levels of ethnic identity represented a risk factor for
gram and three to a control group, with a total of both binge eating and more global eating psycho-
62 female gymnasts (intervention n  =  31; control pathology for White women. As another example,
n  =  31) completing the self-​report post-​test. The Regan and Cachelin (2006) evaluated the frequency
program resulted in athletes perceiving a reduc- of ED symptoms among a multiethnic sample and
tion in pressure from their sports clubs to be thin, determined that binge eating was significantly more
though no changes were found in body esteem, the common among Black and Hispanic women than
EAT, or the SATAQ. No significant change was their male counterparts and that Black, White, and
observed over time on mothers’ measures. Hispanic women were more likely to engage in
purging behaviors compared with Asian women.
Diabetes We could find no examples of prevention pro-
In terms of common medical conditions, pre- grams specifically targeting these or other minority
venting EDs in females with type I or type II dia- groups, or evidence of adapting existing prevention
betes may be particularly important. Females with programs to racial/​ethnic minority status. However,
type I  or II diabetes are at particular risk of EDs Rodriguez, Marchand, Ng, & Stice (2008), found
(Crow, Kendall, Praus, & Thuras, 2000; Jones, that their CD program reduced ED pathology and
Lawson, Daneman, Olmsted, & Rodin, 2000). risk factors similarly across their diverse sample of
Unfortunately, we could find no prevention studies. White, Asian-​American, and Hispanic female ado-
lescents (though sample sizes for minority partici-
Sexual Minorities pants were small).
Sexual minorities are another population particu-
larly vulnerable to developing EDs. The Minority Public Health/​Policy and Mass
Stress Model posits that this may be the case due to Media Models
unique stressors that sexual minorities face as a result Many prevention researchers have argued for the
of inhabiting a predominantly heterosexual culture need to apply public health approaches to reduce

Taylor, Fitzsimmons-Craft, Goel 261

EDs (e.g., Austin, 2012). Such approaches focus launch, the DOVE “Real Beauty Sketches” was the
on changing norms, environment, and policies that most viewed online ad (111 million views). DOVE
may reduce pressure to be thin and other factors in has also developed a Self-​Esteem Fund, which spon-
the environment that might promote EDs. sors self-​esteem building workshops with “inspi-
One major public health focus has been on rational celebrities,” and online tools to educate
reducing exposure of populations to very thin mod- parents, mentors, and young women. In the United
els on the assumption that very thin models serve as States, this company also sponsors ME!, a program
media role models for young women and increase associated with the Girl Scouts designed to build
the risk of AN. A  meta-​analysis found significant self-​confidence in girls ages 8 to 17 with educational
effect sizes between exposure to media images resources and hands-​on activities designed to pro-
depicting the thin-​ideal body and body image con- mote self-​esteem and self-​confidence.
cerns in women (Grabe, Ward, & Hyde, 2008). The impact of these programs on improving
Following the death of two South American mod- self-​esteem is not known, but the campaign was
els, one of whom had a BMI of 13.4 and another associated with a 13% increase in worldwide sales
of whom died on the catwalk, and under pressure of DOVE skin and hair products, suggesting some
from legislators, fashion organizers in Italy, Spain, positive response to these strategies. In addition, in
and Brazil banned models with BMIs lower than one small study, Oswalt and Wyatt (2007) evalu-
18.5. In addition, models also have to carry a medi- ated a campaign that included defining 10 messages
cal certificate to prove that they are healthy. Similar related to sabotaging body image and 10 ways to
regulations have since been adopted by many other enhance body image. These messages were then dis-
countries. Some experts have recommended that played on campus buses, billboards, and magnets.
the fashion industry should be a more highly regu- Following the campaign, students were asked to
lated workplace to ensure the health of the employ- evaluate the impact of these messages, with results
ees, including requiring mandated healthy weights indicating a modest impact. For instance, only 36%
(Record & Austin, 2016). of the sample agreed or strongly agreed with the
In an effort to use mass media to improve body statement, “The messages in this campaign posi-
image and resist the thin-​body ideal, DOVE has tively influenced the way I think about my body.”
undertaken a number of important activities as part Thus, although it is important for these societal-​level
of their Campaign for Real Beauty (www.campaign- programs to be initiated and maintained in order to
forrealbeauty.com). In 2004, DOVE launched an achieve enduring reduction in ED risk factors, the
ad campaign featuring “real women” whose appear- impact of initial efforts is not yet established and
ances were outside the stereotypical norms of beauty. remains a source for future research. Indeed, these
The ads asked viewers to judge the women’s looks programs provide an important opportunity for
(Oversized? Outstanding? Or Wrinkle? Wonderful). researchers to possibly disseminate effective preven-
In 2005, DOVE introduced a second phase, which tion materials by working with companies and orga-
included advertising featuring six women of vary- nizations whose goals and mission statements may
ing sizes and shapes. (Ironically, an article suggested support such programs.
that many of the “real women” presented in the Although many other “environmental” and
DOVE campaign had actually been digitally altered public health approaches have been recommended,
[Collins, 2008].) The next phase, launched in except as applied to specific settings and popula-
2007, addressed issues of women and aging. Other tions as discussed above, we could find no evidence
efforts have focused on helping young girls realize as to their effectiveness.
that what they see in movies and magazines repre-
sents an unrealistic standard of beauty, including Proanorexia and Bulimia Websites and
through the use of “viral films.” The “viral” films Social Media
include Evolution, which depicts the transforma- The Internet hosts thousands of websites that
tion of a real women into a model and promotes promote AN (pro-​ANA) and/​or BN (pro-​MIA).
awareness of how unrealistic perspectives of beauty Such websites may include pictures of very thin
are created, and Onslaught, a film that shows the women who are described as “beautiful” and also
barrage of beauty images that girls absorb ever day. provide tips on how to hide EDs. A Belgian study
DOVE’s “Real Beauty Sketches” ad, launched in that sampled 711 7th-​, 9th-​, and 11th-​grade stu-
2013, examines the issue of how we view ourselves dents found that about 12.6% of girls and 5.9%
compared with how others see us. At the time of its of boys visited proanorexia websites (Custers & Van

262 Prevention
den Bulck, 2009). Girls visiting proanorexia web- high-​risk populations may be of benefit. A focus
sites had higher drive for thinness, worse perception on indicated prevention might be the most pref-
of appearance, and elevated levels of perfection- erable approach in which adolescents with early
ism. Bardone-​Cone and Cass (2007) constructed features of AN, such as failure to gain weight
a prototypical proanorexia website, and randomly in the context of other factors such as excessive
assigned 235 female undergraduates to view either exercise or perfectionism, are targeted. The type
the proanorexia website or one of two comparison of preventive intervention is also less certain for
websites related to female fashion (using average-​ AN than for BN or BED. However, based on the
sized models) or home décor. Study participants success of family-​based interventions for adoles-
exposed to the proanorexia website endorsed more cents with AN (Keel & Haedt, 2008; Lock & le
negative affect, lower social self-​esteem, and lower Grange, 2005; Wilson, 2005), such preventive
appearance self-​efficacy following the experimental efforts might focus on parents. To examine this,
manipulation than those who viewed a compar- Jones et al. (2012) developed an online program,
ison website. In addition, these women perceived modeled after the family-​based therapy model, to
themselves to be heavier and reported an increased help parents and at-​risk students reduce their AN
likelihood of exercising, thinking about their weight risk. Twenty-​four percent of 791 girls screened
in the near future, and engaging in more image met the risk criteria for AN. Parents accessed
comparison. Like other Internet-​ based activities, the majority of the online sessions and rated the
pro-​ANA websites have evolved into pro-​ ANA program favorably. At post-​ assessment, 16 of
communities, and the messages are now available 19 participants evidenced reduced risk status.
in a variety of places including Snapchat, Twitter, Participants remained stable or increased in ideal
Facebook, Pinterest, and Tumblr. body weight and reported decreased ED attitudes
While not directly related to prevention, but and behaviors. However, surprisingly few par-
illustrating the importance of social media use in ents were willing to enroll in the program sug-
the ED population, Saffran et  al. (2016) explored gesting the program was not really practical. An
Facebook use among 415 individuals with a history alternative approach is to focus on high-​risk envi-
of receiving treatment for an ED in a group set- ronments, such as was done with Piran (1999).
ting (e.g., inpatient, residential, outpatient group). Finally, attempts to reduce media glamorization
Participants reported having an average of 10–​19 of the thin-​body ideal, as discussed above, would
Facebook friends from treatment and spending up seem important to reducing AN, although the
to 30 minutes per day interacting on Facebook with benefit of this tactic is difficult to demonstrate in
individuals from treatment or ED-​related organi- scientific studies.
zations. More comparison to treatment peers on
Facebook was associated with greater ED psycho- Dissemination/​Implementation and
pathology and ED-​related impairment. Conversely, Cost–​Benefit Issues
positive interaction with treatment peers on Most of the prevention interventions shown to
Facebook was associated with lower ED psychopa- be effective have been designed for implementation
thology and ED-​related impairment. Few partici- and dissemination. Becker et al. (2016) describe
pants (19.5%) reported that a therapist asked about the dissemination and implementation of the Body
the impact of Facebook on pathology. Whatever Project across six diverse stakeholder partnerships
impact these sites might have, it is not possible to that span academic, nonprofit, and business sec-
remove them from the Internet and as such, edu- tors at national and international levels, includ-
cators, parents, therapists and others are advised to ing a train-​the-​trainer approach (Greif, Becker, &
discuss their use with at-​risk populations. Hildebrandt, 2015). Many of the school-​based pro-
grams have been designed for implementation.
Preventing Anorexia Nervosa Much less has been written about the cost of
Although there has been impressive prog- these programs. Moessner and colleagues recently
ress in preventing ED, most of the studies have examined the costs of five school-​based dissemina-
focused on combined onset of subclinical, clini- tion strategies for an Internet-​based intervention
cal BED, BN, and/​or EDNOS. The low preva- for the prevention and early intervention of EDs
lence of AN (2%–​3% of the population) makes (Moessner et al., 2016). Three hundred ninety-​
it nearly impossible to demonstrate a popula- five schools were randomly assigned to one of five
tion-​based effect, although interventions in dissemination strategies. Strategies varied with

Taylor, Fitzsimmons-Craft, Goel 263

respect to intensity from only sending advertise- in schools, and that target multiple systems (e.g.,
ment materials and asking the school to distribute peers, media, teachers/​coaches, and families) and
them among students to organizing presentations the varied developmental needs of children and
and workshops at schools. Effects were defined as adolescents over time (i.e., onset of puberty, transi-
the number of page visits, the number of screen- tions to high school or college, increasing exposure
ings conducted, and the number of registrations to to peer pressure and media messages about appear-
the Internet-​based intervention. More expensive ance, increased pressure to be thin or lean/​muscular
strategies proved to be more cost-​effective. Cost per within competitive sports) are most needed in order
page visit ranged from €2.83 (introductory pres- to ensure that a positive social support system in
entation plus workshop) to €20.37 (dissemination which weight and eating concerns are deemphasized
by student representatives/​peers). Costs per screen- is developed and strengthened over time.
ing ranged from €3.30 (introductory presentation Of note is the programmatic research, developed
plus workshop) to €75.66 (dissemination by stu- over the past decade and informed by risk factor
dent representatives/​peers), and costs per registra- research and theories, from several research groups
tion ranged from €6.86 (introductory presentation that has led to the development of programs and
plus workshop) to €431.10 (advertisement materi- techniques that reduce risk factor scores and the
als only). The authors note that the dissemination onset of EDs, at least in older adolescent and college
of an Internet-​ based intervention for prevention women at high risk for the development of EDs.
and early intervention is challenging and expensive. Furthermore, prevention programs have not been
Kass, Balantekin, et al. (2017) estimated the costs found to cause harm. By building on the gains made
of implementing a stepped-​care model for online in ED prevention over the past two decades and
screening, prevention, and treatment among college continuing to consider important issues pertain-
students. Calculations showed that the cost to pre- ing to dissemination of materials and translation of
vent one ED case among those at highest risk is less findings to real-​world settings, it appears possible
than the cost of a “wait and treat” approach, even that researchers have the opportunity to achieve sig-
when accounting for the rate of failure. Estimates nificant reductions in risk for EDs and ED onset in
of the stepped-​ care model against standard care the coming decade.
were also estimated to yield savings. In sum, the
stepped-​care model was estimated to result in cost Future Directions
savings compared with standard care; however, the In spite of the major gains associated with sev-
authors noted that future research is needed to sys- eral of the reviewed programs, more work must be
tematically measure the costs and benefits of such a done to increase the effectiveness of ED preven-
comprehensive stepped-​care model for EDs actually tion programs. Critical next steps appear to be
implemented on the college campus. (1)  developing and evaluating programs that are
effective for younger age groups (i.e., preadoles-
Conclusion cent and early adolescent youth) in order to reduce
In conclusion, due to well-​designed studies and and ultimately prevent the emergence of early
recent meta-​analytic reviews, we now know more concerns about weight and shape and early signs
about what does and does not appear to be effec- of disordered eating that may lead to later eating
tive in regard to ED prevention programs; these pathology during adolescents; (2) enhancing indi-
conclusions provide both guidance for future pro- vidually based programs to be incorporated into
gram development (e.g., cognitive-​behavioral, dis- environmentally focused (e.g., school-​based) pro-
sonance, and/​or healthy weight regulation material grams and to engage additional individuals who
appear more effective than psychoeducation or life-​ may have important and relevant developmental
skills alone programs) as well as highlight areas in influences on youth, including parents, siblings,
which greater improvement is needed as discussed teachers, coaches, and physicians; (3) using recent
in the preceding text. It is critical for programs be mediator and moderator findings and developing
informed by both theory and risk factor research effective screening tools in order to ensure that the
and that these programs include skills that target appropriate programs are delivered to individuals
and reduce risk for overweight (e.g., healthy weight presenting with different risk factors and/​or varied
regulation material) in order to ensure public health levels of risk for ED onset; (4) ensuring that pro-
prominence for ED prevention programs. Programs grams are developed that target the issues relevant
that are delivered in high school environments or to disordered eating onset in males and in special

264 Prevention
at-​risk populations, such as athletes, adolescents Becker, C. B., Smith, L. M., & Ciao, A. C. (2006). Peer facili-
with type I diabetes, sexual minorities, and racial/​ tated eating disorders prevention: A randomized effectiveness
trial of cognitive dissonance and media advocacy. Journal of
ethnic minorities; and (5)  determining whether Counseling Psychology, 53, 550–​555.
programs that focus on a broad range of risk fac- Becker, C. B., Wilson, C., Williams, A., Kelly, M., McDaniel, L.,
tors and behaviors, including problems with affect & Elmquist, J. (2010). Peer-​facilitated cognitive dissonance
regulation, binge drinking, and excessive weight versus healthy weight eating disorders prevention:  A  ran-
concerns, can effectively reduce ED and comor- domized comparison. Body Image, 7, 280–​288.
Becker, C. B., Perez, M., Kilpela, L. S., Diedrichs, P. C., Trujillo,
bidity onset. Recent work also suggests that ED E., & Stice, E. (2016). Engaging stakeholder communi-
prevention efforts may be best paired with obesity ties as body image intervention partners: The Body Project
prevention programs; this is appealing from a pub- as a case example. Eating Behaviors, 62–67. doi: 10.1016/​
lic health perspective and also recognizes the shared j.eatbeh.2016.03.015
risk factors and overlap between disordered eating Beintner, I., Jacobi, C., & Taylor, C. B. (2012). Effects of an
Internet‐based prevention programme for eating disorders
and overweight. Addressing these issues and build- in the USA and Germany: A meta‐analytic review. European
ing on two decades of important studies might Eating Disorders Review, 20, 1–​8.
actualize the public health goal of preventing EDs. Beintner, I., Jacobi, C., & Taylor, C. B. (2014). Participant
adherence to the Internet-​ based prevention program
References StudentBodies™ for eating disorders:  A  review. Internet
Abascal, L., Bruning-​Brown, J., Winzelberg, A. J., Dev, P., & Interventions, 1, 26–​32.
Taylor, C. B. (2004). Combining universal and targeted Beintner, I., Jacobi, C., Winzelberg, A. J., & Taylor, C. B. (2008).
prevention for school-​ based eating disorder programs. Wirksamkeit eines Internet-​gestützten Präventionsprogrammes
International Journal of Eating Disorders, 35, 1–​9. in Deutschland und den USA. Unpublished manuscript.
Allen, K. L., Byrne, S. M., Crosby, R. D., & Stice, E. (2016). Bonci, C. M., Bonci, L. J., Granger, L. R., Johnson C. L.,
Testing for interactive and non-​linear effects of risk factors for Malina, R. M., Milne, L. W.,  . . .  Vanderbunt E. M. (2008).
binge eating and purging eating disorders. Behavior Research National Athletic Trainers’ Association position state-
and Therapy, 27, 40–​47. doi: 10.1016/​j.brat.2016.08.019 ment: Preventing, detecting, and managing disordered eating
Aronson, E. (1980). Persuasion via self-​ justification:  Large in athletes. Journal of Athletic Training, 43, 80–​108.
commitments for small rewards. In L. Festinger (Ed.), Bratland-​Sanda, S., & Sundgot-​Borgen, J. (2012). Eating dis-
Retrospection on social psychology (pp.  3–​21). Oxford, orders in athletes:  Overview of prevalence, risk factors and
UK: Oxford University Press. recommendations for prevention and treatment. European
Austin, S. B., Field, A. E., Wiecha, J., Peterson, K. E., & Journal of Sports Science, 13, 499–​508.
Gortmaker, S. L. (2005). The impact of a school-​based obe- Brown, T. A., & Keel, P. K. (2015). A randomized controlled trial
sity prevention trial on disordered weight-​control behaviors of a peer co-​led dissonance-​based eating disorder preven-
in early adolescent girls. Archives of Pediatrics Adolescent tion program for gay men. Behaviour Research and Therapy,
Medicine, 159, 225–​230. 74, 1–​10.
Austin, S. B., Kim, J., Wiecha, J., Troped, P. J., Feldman, H. Bruning-​Brown, J., Winzelberg, A. J., Abascal, L. B., & Taylor,
A., & Peterson, K. E. (2007). School-​ based overweight C. B. (2004). An evaluation of an Internet-​delivered eating
preventive intervention lowers incidence of disordered disorder prevention program for adolescents and their par-
weight-​control behaviors in early adolescent girls. Archives of ents. Journal of Adolescent Health, 35, 290–​296.
Pediatrics Adolescent Medicine, 161, 865–​869. Buchholz, A., Mack, H., McVey, G., Feder, S., & Barrowman, N.
Austin, S. B., Ziyadeh, N. J., Corliss, H. L., Rosario, M., Wypij, (2008). BodySense: An evaluation of a positive body image
D., Haines, J.,  . . .  Field, A. E. (2009). Sexual orientation intervention on sport climate for female athletes. Eating
disparities in purging and binge eating from early to late ado- Disorders, 16, 308–​321.
lescence. Journal of Adolescent Health, 45, 238–​245. Bulik, C. M., Kleiman, S. C., & Yilmaz, Z. (2016). Genetic
Austin, S. B. (2012). A public health approach to eating disorders epidemiology of eating disorders. Curr Opin Psychiatry, 6,
prevention: Its time for public health professionals to take a 383–388. doi: 10.1097/YCO.0000000000000275
seat at the table. BMC Public Health, 12, 843. doi: 10.1186/​ Carter, F. A., & Bulik, C. M. (2008). Childhood obesity preven-
1471-​2458-​12-​854 tion programs: How do they affect eating pathology and other
Bandura, A. (1986). Social foundations of thought and action. psychological measures? Psychosomatic Medicine, 70, 363–​371.
Englewood Cliffs, NJ: Prentice-​Hall. Carter, J. C., Stewart, D. A., Dunn, V. J., & Fairburn, C. G.
Bardone-​Cone, A. M., & Cass, K. M. (2007). What does view- (1997). Primary prevention of eating disorders:  Might it
ing a pro-​anorexia website do? An experimental examination do more harm than good? International Journal of Eating
of website exposure and moderating effects. International Disorders, 22, 167–​172.
Journal of Eating Disorders, 40, 537–​548. Cash, T. F. (1997). The body image workbook. Oakland, CA: New
Becker, C. B., Bull, S., Schaumberg, K., Cauble, A., & Franco, Harbinger.
A. (2008). Effectiveness of peer-​led eating disorders preven- Celio, A. A., Bryson, S., Killen, J. D., & Taylor, C. B. (2003). Are
tion:  A  replication trial. Journal of Consulting and Clinical adolescents harmed when asked risky weight control behav-
Psychology, 76, 347–​354. ior and attitude questions? Implications for consent proce-
Becker, C. B., Smith, L., & Ciao, A. C. (2005). Reducing eating dures. International Journal of Eating Disorders, 34, 251–​254.
disorder risk factors in sorority members: A randomized trial. Celio, A. A., Winzelberg, A. J., Wilfley, D. E., Eppstein-​Herald,
Behavior Therapy, 36, 245–​254. D., Springer, E. A., Dev, P., & Taylor, C. B. (2000). Reducing

Taylor, Fitzsimmons-Craft, Goel 265

 risk factors for eating disorders: Comparison of an Internet-​ comparison of high level dissonance, low level dissonance,
and a classroom-​ delivered psychoeducational program. and no-​treatment control. Eating Disorders, 13, 157–​170.
Journal of Consulting and Clinical Psychology, 68, 650–​657. Greif, R., Becker, C. B., & Hildebrandt, T. (2015). Reducing eat-
Ciao, A. C., Loth, K., & Neumark-​ Sztainer, D. (2014). ing disorder risk factors: A pilot effectiveness trial of a train-​
Preventing eating disorder pathology: Common and unique the-​trainer approach to dissemination and implementation.
features of successful eating disorders prevention pro- International Journal of Eating Disorders, 48, 1122–​1131.
grams. Current Psychiatry Reports, 16, 453. doi:  10.1007/​ Guarda, A. S. (2008). Treatment of anorexia nervosa:  Insights
s11920-​014-​0453-​0 and obstacles. Physiology and Behavior, 94, 113–​120.
Clarke, G. N., Hornbrook, M., Lynch, F., Polen, M., Gale, J., Haines, J., & Neumark-​Sztainer, D. (2006). Prevention of obe-
Beardslee, W., O’Connor, E., & Seeley, J. (2001). A ran- sity and eating disorders: A consideration of shared risk fac-
domized trial of a group cognitive intervention for prevent- tors. Health Education Resource, 21, 770–​782.
ing depression in adolescent offspring of depressed parents. Halliwell, E., & Diedrichs, P. C. (2014). Testing a disso-
Archives of General Psychiatry, 58, 1127–​1134. nance body image intervention among young girls. Health
Collins, L. (2008, May 12). Pixel perfect: Pascal Dangin’s virtual Psychology, 33, 201.
reality. The New Yorker. Halmi, K. A., Casper, R. C., Eckert, E. D., Goldberg, S. C., &
Coughlin, J. W., & Kalodner, C. (2006). Media literacy as a pre- Davis, J. M. (1979). Unique features associated with age of
vention intervention for college women at low-​or high-​risk onset of anorexia nervosa. Psychiatry Research, 1, 209–​215.
for eating disorders. Body Image, 3, 35–​43. doi:10.1016/​0165-​1781(79)90063-​5
Crow, S., Kendall, D., Praus, B., & Thuras, P. (2000). Binge Horowitz, J. L., & Garbe, J. (2006). The prevention of depres-
eating and other psychopathology in patients with type II sive symptoms in children and adolescents: A meta-​analytic
diabetes mellitus. International Journal of Eating Disorders, review. Journal of Consulting and Clinical Psychology, 74,
20, 222–​226. 401–​415.
Culbert K. M., Racine, S. E., & Klump, K. L. (2015). Research Ikeda, J. P., Crawford, P. B., & Woodward-​Lopez, G. (2006).
Review:  What we have learned about the causes of eating BMI screening in schools:  Helpful or harmful. Health
disorders:  A  synthesis of sociocultural, psychological, and Education Resources, 21, 761–​769.
biological research. Journal of Child Psychology and Psychiatry, Jacobi, C., Hayward, C., de Zwaan, M., Kraemer, H. C., & Agras,
56, 1141–​1164. doi: 10.1111/​jcpp.12441 W. S. (2004). Coming to terms with risk factors for eating
Custers, K., & Van den Bulck, J. (2009). Viewership of pro-​ disorders:  Application of risk terminology and suggestions
anorexia websites in seventh, ninth and eleventh grad- for a general taxonomy. Psychological Bulletin, 130, 19–​65.
ers. European Eating Disorders Review, 17, 214–​219. Jacobi, C., Abascal, L., & Taylor, C. B. (2004). Screening for eat-
doi: 10.1002/​erv.910 ing disorders and high-​risk behavior: Caution. International
Doris, E., Shekriladze, L., Javakhishvili, N., Jones, R., Treasure, Journal of Eating Disorders, 36, 280–​295.
J., & Tchanturia, K. (2015). Is cultural change associated Jacobi, C., Morris, L., Beckers, C., Bronisch-​Holtze, J., Winter,
with eating disorders? A systematic review of the literature. J., Winzelberg, A. J., & Taylor, C. B. (2007). Maintenance
Eating and Weight Disorders, 10, 149–​160. doi:  10.1007/​ of Internet-​based prevention: A randomized controlled trial.
s40519-​01500189-​9 International Journal of Eating Disorders, 40, 114–​119.
Fairburn, C. G., & Cooper, Z. (1993). The eating disorder exami- Jacobi, C., Bryson S. W., Wilfley, D., Kraemer, H. C., & Taylor,
nation (12th ed.). New York, NY: Guilford Press. C. B. (2011). Who is really at risk: Identifying risk factors
Feldman, M. B., & Meyer, I. H. (2007). Eating disorders in for subthreshold and full syndrome eating disorders in a
diverse lesbian, gay, and bisexual populations. International high-​risk sample. Psychological Medicine, 41, 1939–​1949.
Journal of Eating Disorders, 40, 218–​226. doi: 10.1017/​S0033291710002631
Festinger, L. (1957). A theory of cognitive dissonance. Palo Alto, Johnson, C., Burke, C., Brinkman, S., & Wade, T. (2016).
CA: Stanford University Press. Effectiveness of a school-​ based mindfulness program
Festinger, L. (1962). Cognitive dissonance. Scientific American, for transdiagnostic prevention in young adolescents.
207, 93–​102. Behaviour Research and Therapy, 81, 1–​11. doi: 10.1016/​
Field, A. E., Camargo, C. A., Taylor, C. B., Berkey, C. S., & j.brat.2016.03.002.
Colditz, G. A. (1999). Relation of peer and media influences Jones, J. M., Lawson, M. L., Daneman, D., Olmsted, M. P., &
to the development of purging behaviors among preadoles- Rodin, G. (2000). Eating disorders in adolescent females
cent and adolescent girls. Archives of Pediatrics Adolescent with and without type 1 diabetes:  Cross sectional study.
Medicine, 153, 1184–​1189. British Medical Journal, 320, 1563–​1566.
Frederickson, B., & Roberts, T. A. (1997). Objectification Jones, M., Luce, K. H., Osborne, M. I., Taylor, K., Cunning, D.,
theory:  Toward understanding women’s lived experiences Doyle, A. C.,  . . .  Taylor, C. B. (2008). Randomized, con-
and mental health risks. Psychology of Women Quarterly, 21, trolled trial of an Internet-​facilitated intervention for reduc-
173–​206. ing binge eating and overweight in adolescents. Pediatrics,
Goldschmidt, A. B., Aspen, V. P., Sinton, M. M., Tanofsky-​Kraff, 121, 453–​462.
M., & Wilfley, D. E. (2008). Disordered eating attitudes and Jones, M., Taylor, L. K., Kass, A. E., Burrows, A., Williams,
behaviors in overweight youth. Obesity, 16, 257–​264. J., Wilfley, D. E., & Taylor, C. B. (2014). Healthy weight
Grabe, S., Ward, L. M., & Hyde, J. S. (2008). Role of the media regulation and eating disorder prevention in high school
in body image concerns among women: A meta-​analysis of students:  a universal and targeted Web-​based intervention.
experimental and correlational studies. Psychological Bulletin, Journal of Medical Internet Research, 16, e57. doi: 10.2196/​
134, 460–​476. jmir.2995
Green, M., Scott, N., Diyankova, I., Gasser, C., & Pederson, Jones, M., Völker, U., Lock, J., Taylor, C. B., & Jacobi, C.
E. (2005). Eating disorder prevention:  An experimental (2012). Family-​ based early intervention for anorexia

266 Prevention
 nervosa. European Eating Disorders Review, 20, e137–​43. of eating disorders in college females. Healthy Psychology, 16,
doi: 10.1002/​erv.2167 215–​225.
Kass, A. E., Balantekin, K. N., Fitzsimmons-​Craft, E. E., Jacobi, Manwaring, J. L., Bryson, S. W., Goldschmidt, A. B.,
C., Wilfley, D. E., & Taylor, C. B. (2017). The economic case Winzelberg, A. J., Luce, K. H.,  . . .  Taylor, C. B. (2008). Do
for digital interventions for eating disorders among college stu- adherence variables predict outcome in an online program
dents. International Journal of Eating Disorders, 50(3), 250– for the prevention of eating disorders? Journal of Consulting
258. doi: 10.1002/eat.22680 and Clinical Psychology, 76, 341–​346.
Kass, A. E., Jones, M., Kolko, R. P., Altman, M., Fitzsimmons-​ Martinsen, M., Bahr, R., Børresen, R., Holme, I., Pensgaard, A.
Craft, E. E., Eichen, D. M., . . . Wilfley, E. E. (2017). M., & Sundgot-​Borgen, J. (2014). Preventing eating disor-
Universal prevention efforts should address eating disor- ders among young elite athletes:  A  randomized controlled
der pathology across the weight spectrum: Implications trial. Medicine Science in Sports and Exercise, 46, 435–​447.
for screening and intervention on college campuses. Eating doi: 10.1249/​MSS.0b013e3182a702fc
Behaviors, 25, 74–80. doi: 10.1016/​j.eatbeh.2016.03.019 Martinsen M., Sherman, R. T., Thompson, R. A., & Sundgot-​
Kass, A. E., Trockel, M., Safer, D. L., Sinton, M. M., Cunning, Borgen, J. (2015). Coaches’ knowledge and management
D., Rizk, M. T.,  . . .  Taylor, C. B. (2014). Internet-​based pre- of eating disorders:  A  randomized controlled trial. Medical
ventive intervention for reducing eating disorder risk: A ran- Science in Sports and Exercise, 47, 1070–​1078. doi: 10.1249/​
domized controlled trial comparing guided with unguided MSS.0000000000000489
self help. Behaviour Research and Therapy, 63, 90–​98. Martinsen, M., & Sundgot-​Borgen, J. (2013). Higher preva-
Keel, P. K., & Haedt, A. (2008). Evidence-​based psychosocial lence of eating disorders among adolescent elite athletes
treatments for eating problems and eating disorders. Journal than controls. Medicine and Science in Sports and Exercise,
of Clinical Child and Adolescent Psychology, 37, 39–​61. 45, 1188–​1197.
Killen, J. D., Taylor, C. B., Hammer, L. D., Litt, I., Wilson, D. Matusek, J. A., Wendt, S. J., & Wiseman, C. V. (2004).
M., Rich, T.,  . . .  Varady, A. (1993). An attempt to modify Dissonance thin-​ideal and didactic healthy behavior eating
unhealthful eating attitudes and weight regulation practices disorder prevention programs:  Results from a controlled
of young adolescent girls. International Journal of Eating trial. International Journal of Eating Disorders, 36, 376–​388.
Disorders, 13, 369–​384. McLean, S. A., Paxton, S. J., Wertheim, E. H. (2016). The role of
Killen, J. D., Taylor, C. B., Hayward, C., Haydel, K. F., Wilson, media literacy in body dissatisfaction and disordered eating:
D. M., Hammer, L.,  . . .  Strachowski, D. (1996). Weight A systematic review. Body Image, 19, 9–​23. doi: 10.1016/​
concerns influence the development of eating disorders: A 4-​ j.bodyim.2016.08.002
year prospective study. Journal of Consulting and Clinical McVey, G. L., Davis, R., Tweed, S., & Shaw, B. F. (2004).
Psychology, 64, 936–​940. Evaluation of a school-​based program designed to improve
Killen, J. D., Taylor, C. B., Hayward, C., Wilson, D. M., Haydel, body image satisfaction, global self-​ esteem, and eating
F., Hammer, L. D.,  . . .  Kraemer, H. (1994). Pursuit of thin- attitudes and behaviors:  A  replication study. International
ness and onset of eating disorder symptoms in a community Journal of Eating Disorders, 36, 1–​11.
sample of adolescent girls: A three-​year prospective analysis. McVey, G. L., Lieberman, M., Voorberg, N., Wardrope,
International Journal of Eating Disorders, 16, 227–​238. D., & Blackmore, E. (2003). School-​based peer support
Kraemer, H. C., Wilson, G. T., Fairburn, C. G., & Agras, W. groups: A new approach to the prevention of disordered eat-
S. (2002). Mediators and moderators of treatment effects in ing. Eating Disorders, 11, 169–​185.
randomized clinical trials. Archives of General Psychiatry, 59, McVey, G. L., Lieberman, M., Voorberg, N., Wardrope, D.,
877–​883. Blackmore, E., & Tweed, S. (2003). Replication of a peer
Leippe, M. R. (1994). Generalization of dissonance reduc- support program designed to prevent disordered eating: Is a
tion:  Decreasing prejudice through induced compliance. life skills approach sufficient for all middle school students?
Journal of Personality and Social Psychology, 67, 395–​413. Eating Disorders, 11, 187–​195.
Lock, J., & le Grange, D. (2005). Family-​based treatment of eat- McVey, G. L., Tweed, S., & Blackmore, E. (2007). Healthy
ing disorders. International Journal of Eating Disorders, 37, Schools-​Healthy Kids: A controlled evaluation of a compre-
S64–​67; discussion S87–​S69. hensive universal eating disorder prevention program. Body
Loucas, C. E., Fairburn, C. G., Whittington, C., Pennant, M. Image, 4, 115–​136.
E., Stockton, S., & Kendall, T. (2014). E-​therapy in the Meyer, I. H. (2003). Prejudice, social stress, and mental health
treatment and prevention of eating disorders:  A  systematic in lesbian, gay and bisexual populations: Conceptual issues
review and meta-​analysis. Behaviour research and therapy, 63, and research evidence. Psychological Bulletin, 129, 674–​697.
122–​131. Micali, N., Hagberg, K. W., Petersen, I., & Treasure, J. L.
Luce, K. H., Osborne, M. I., Winzelberg, A. J., Das, S., Abascal, (2013). The incidence of eating disorders in the UK
L. B., Celio, A. A.,  . . .  Taylor, C. B. (2005). Application in 2000-​ 2009:  Findings from the General Practice
of an algorithm-​driven protocol to simultaneously provide Research Database. BJC Open, 3, e002646. doi:10.1136/​
universal and targeted prevention programs. International bmjopen-​2013-​002646
Journal of Eating Disorders, 37, 220–​226. Morrison, M. A., Morrison, T. G., & Sager, C. L. (2004). Does
Lynch, F. L., Hornbrook, M., Clarke, G. N., Perrin, N., Polen, body dissatisfaction differ between gay men and lesbian
M. R., O’Connor, E., & Dickerson, J. (2005). Cost-​ women and heterosexual men and women? A meta-​analytic
effectiveness of an intervention to prevent depression in at-​ review. Body Image, 1, 127–​138.
risk teens. Archives of General Psychiatry, 62, 1241–​1248. Moessner, M., Minarik, C., Ozer, F., & Bauer, S. (2016).
Mann, T., Nolen-​ Hoeksema, S., Huang, K., Burgard, D., Effectiveness and cost-​ effectiveness of school-​ based dis-
Wright, A., & Hanson, K. (1997). Are two interventions semination strategies of an Internet-​ based program
worse than none? Joint primary and secondary prevention for the prevention and early intervention in eating

Taylor, Fitzsimmons-Craft, Goel 267

 disorders:  A  randomized trial. Preventive Science, 17, 306–​ program are similar for Asian American, Hispanic, and
313. doi: 10.1007/​s11121-​015-​0619-​y White participants. International Journal of Eating Disorders,
Musiat, P., Conrod, P., Treasure, J., Tylee, A., Williams, C., & 41, 618–​625.
Schmidt, U. (2014). Targeted prevention of common men- Roehrig, M., Thompson, J. K., Brannick, M., & van den Berg,
tal health disorders in university students:  Randomised P. (2006). Dissonance-​based eating disorder prevention pro-
controlled trial of a transdiagnostic trait-​focused web-​based gram: A preliminary dismantling investigation. International
intervention. PLoS One, 9, e93621. doi:  10.1371/​journal. Journal of Eating Disorders, 39, 1–​10.
pone.0093621 Saffran, K., Fitzsimmons-​Craft, E. E., Kass, A. E., Wilfley, D. E.,
Neumark-​ Sztainer, D. (2005). Can we simultaneously work Taylor, C. B., & Trockel, M. (2016). Facebook usage among
toward the prevention of obesity and eating disorders in those who have received treatment for an eating disorder in
children and adolescents? International Journal of Eating a group setting. International Journal of Eating Disorders, 49,
Disorders, 38, 220–​227. 764–​777. doi: 10.1002/​eat.22567
Neumark-​Sztainer, D., Sherwood, N. E., Coller, T., & Hannan, Sanci, L., Coffey, C., Olsson, C., Reid, S., Carlin, J. B., &
P. J. (2000). Primary prevention of disordered eating among Patton, G. (2008). Childhood sexual abuse and eating dis-
preadolescent girls:  Feasibility and short-​ term effect of a orders in females:  Findings from the Victorian Adolescent
community-​based intervention. Journal of the American Health Cohort Study. Archives of Pediatrics and Adolescent
Dietetic Association, 100, 1466–​1473. Medicine, 162, 261–​267.
Neumark-​Sztainer, D., Wall, M. M., Haines, J. I., Story, M. T., Shaw, H., Ramirez, L., Trost, A., Randall, P., & Stice, E.
Sherwood, N. E., & van den Berg, P. A. (2007). Shared risk (2004). Body image and eating disturbances across eth-
and protective factors for overweight and disordered eating nic groups: More similarities than differences. Psychology of
in adolescents. American Journal of Preventive Medicine, 33, Addictive Behaviors, 18, 12–​18.
359–​369. Sherwood, N. E., Harnack, L., & Story, M. (2000). Weight-​loss
O’Dea, J. A. (2005). Prevention of child obesity: “First, do no practices, nutrition beliefs, and weight-​loss program pref-
harm.” Health Education Resource, 20, 259–​265. erences of urban American Indian women. Journal of the
Ogden, C. L., Carroll, M. D., Fryar, C. D., & Flegal, K. American Dietetic Association, 100, 442–​446.
M. (2015). Prevalence of obesity among adults and Shuttlesworth, M. E., & Zotter, D. (2011). Disordered eating in
youth: United States, 2011–​2 015. NCHS Data Brief, African American and Caucasian women: The role of ethnic
219, 1–​8 . identity. Journal of Black Studies, 42, 906–​922.
Oswalt, S. B., & Wyatt, T. J. (2007). Mirror, mirror, help me Speranza, M., Atger, F., Corcos, M., Loas, G., Guilbaud, O.,
like my body:  Examining a body image media campaign. Stéphan, P.,  . . .  Jeammet, P. (2003). Depressive psychopa-
Californian Journal of Health Promotion, 5, 135–​147. thology and adverse childhood experiences in eating disor-
Perez, M., Becker, C. B., & Ramirez, A. (2010). Transportability ders. European Psychiatry, 18, 377–​383.
of an empirically supported dissonance-​ based prevention Springer, E. A., Winzelberg, A. J., Perkins, R., & Taylor, C.
program for eating disorders. Body Image, 7, 179–​186. B. (1999). Effects of a body image curriculum for college
Peterson, R. D., Grippo, K. P., & Tantleff-​Dunn, S. (2008). students on improved body image. International Journal of
Empowerment and powerlessness: A closer look at the rela- Eating Disorders, 26, 13–​20.
tionship between feminism, body image and eating distur- Stein, N. (1999). Classrooms and courtrooms:  Facing sexual
bance. Sex Roles, 58, 639–​648. harassment in K-​ 12 schools. New  York, NY:  Teachers
Pike, K. M., Hilbert, A., Wilfley, D. E., Fairburn, C. G., Dohm, College Press.
F. A., Walsh, B. T., & Striegel-​Moore, R. (2007). Toward an Stice, E., & Agras, W. S. (1998). Predicting onset and cessation
understanding of risk factors for anorexia nervosa:  A  case-​ of bulimic behaviors during adolescence:  A  longitudinal
control study. Psychological Medicine, 10, 1–​11. grouping analysis. Behavior Therapy, 29, 257–​276.
Piran, N. (1999). Eating disorders: A trial of prevention in a high Stice, E., Butryn, M. L., Rohde, P., Shaw, H., & Marti, C. N.
risk school setting. Journal of Primary Prevention, 10, 75–​90. (2013). An effectiveness trial of a new enhanced dissonance
Piran, N. (2010). A feminist perspective on risk factor research eating disorder prevention program among female college
and on the prevention of eating disorders. Journal of Treament students. Behaviour Research and Therapy, 51, 862–​871.
and Prevention, 18, 183–​198. Stice, E., Chase, A., Stormer, S., & Appel, A. (2001). A ran-
Pratt, B. M., & Woolfenden, S. R. (2002). Interventions for domized trial of a dissonance-​based eating disorder preven-
preventing eating disorders in children and adolescents. tion program. International Journal of Eating Disorders, 29,
Cochrane Database Systematic Reviews, (2):CD002891. 247–​262.
Puhl, R. M., & Heuer, C. A. (2010). Obesity stigma: Stice, E., Durant, S., Rohde, P., & Shaw, H. (2014). Effects of a
Important considerations for public health. American prototype Internet dissonance-​based eating disorder preven-
Journal of Public Health, 100, 1019–​1028. doi: 10.2105/​ tion program at 1-​and 2-​year follow-​up. Health Psychology,
AJPH.2009.159491. 33, 1558–​1567.
Record, K. L., & Austin, S. B. (2016). “Paris thin”:  A  call to Stice, E., Marti, C. N., & Rohde, P. (2013). Prevalence, inci-
regulate life-​threatening starvation of runway models in the dence, impairment, and course of the proposed DSM-​5 eat-
US fashion industry. American Journal of Public Health, 106, ing disorder diagnoses in an 8-​year prospective community
205–​206. doi: 10.2105/​AJPH.2015.302950 study of young women. Journal of Abnormal Psychology, 122,
Regan, P. C., & Cachelin, F. M. (2006). Binge eating and purging 445–​457. doi:10.1037/​a0030679
in a multi-​ethnic community sample. International Journal of Stice, E., Marti, N., Shaw, H., & O’Neil, K. (2008). General
Eating Disorders, 39, 523–​526. and program-​specific moderators of two eating disorder pre-
Rodriguez, R., Marchand, E., Ng, J., & Stice, E. (2008). Effects vention programs. International Journal of Eating Disorders,
of a cognitive dissonance-​based eating disorder prevention 41, 611–​617.

268 Prevention
Stice, E., Marti, C. N., Spoor, S., Presnell, K., & Shaw, H. Journal of Consulting and Clinical Psychology, 84, 402–​414.
(2008). Dissonance and healthy weight eating disorder pre- doi: 10.1037/​ccp0000077
vention programs:  Long-​ term effects from a randomized Tiggemann, M. (2013). Objectification theory:  Of relevance
efficacy trial. Journal of Consulting and Clinical Psychology, for eating disorder researchers and clinicians? Clinical
76, 329–​340. Psychologist, 17, 35–​45.
Stice, E., Mazotti, L., Weibel, D., & Agras, W. S. (2000). Tiggemann, M., & Kuring, J. K. (2004). The role of body objec-
Dissonance prevention program decreases thin-​ideal inter- tification in disordered eating and depressed mood. British
nalization, body dissatisfaction, dieting, negative affect, and Journal of Clinical Psychology, 43, 299–​311.
bulimic symptoms: A preliminary experiment. International Völker, U., Jacobi, C., Trockel, M. T., & Taylor, C. B. (2014).
Journal of Eating Disorders, 27, 206–​217. Moderators and mediators of outcome in Internet-​ based
Stice, E., Presnell, K., Gau, J., & Shaw, H. (2007). Testing indicated prevention for eating disorders. Behaviour Research
mediators of intervention effects in randomized controlled and Therapy, 63, 114–​121.
trials:  An evaluation of two eating disorder prevention Volpe, U., Tortorella, A., Manchia, M., Monteleone, A. M.,
programs. Journal of Consulting and Clinical Psychology, Albert, U., & Monteleone, P. (2016). Eating disorders: What
75, 20–​32. age at onset? Psychiatry Research, 238, 225–​227. doi:10.1016/​
Stice, E., Rohde, P., Durant, S., & Shaw, H. (2012). A prelimi- j.psychres.2016.02.048
nary trial of a prototype Internet dissonance-​based eating Wade, T. D., Davidson, S., & O’Dea, J. A. (2003). A prelimi-
disorder prevention program for young women with body nary controlled evaluation of a school-​based media literacy
image concerns. Journal of Consulting Clinical Psychology, 80, program and self-​esteem program for reducing eating disor-
907–​916. der risk factors. International Journal of Eating Disorders, 33,
Stice, E., Rohde, P., Gau, J., & Shaw, H. (2009). An effective- 371–​383; discussion 384–​377.
ness trial of a dissonance-​based eating disorder prevention Watson, H. J., Joyce, T., French, E., Willan, V., Kane, R. T.,
program for high-​risk adolescent girls. Journal of Consulting Tanner‐Smith, E. E., . . . Egan, S. J. (2016). Prevention
and Clinical Psychology, 77, 825. of eating disorders: A systematic review of randomized, con-
Stice, E., Rohde, P., & Shaw, H. (2012). The Body trolled trials. International Journal of Eating Disorders, 49(9),
Project:  A  dissonance-​ based eating disorder intervention 833–862. doi:10.1002/​eat.22577
(2nd ed.). Programs That Work. New  York, NY:  Oxford Wertheim, E. H., Koerner, J., & Paxton, S. J. (2001).
University Press. Longitudinal predictors of restrictive eating and bulimic
Stice, E., Rohde, P., Shaw, H., & Gau, J. (2011). An effectiveness tendencies in three different age groups of adolescent girls.
trial of a selected dissonance-​based eating disorder preven- Journal of Youth and Adolescence, 30, 69–​81.
tion program for female high school students:  Long-​term Whisenhunt, B. L., Williamson, D. A., Drab-​Hudson, D. L.,
effects. Journal of Consulting and Clinical Psychology, 79, 500. & Walden, H. (2008). Intervening with coaches to promote
Stice, E., & Shaw, H. (2004). Eating disorder prevention pro- awareness and prevention of weight pressures in cheerleaders.
grams:  A  meta-​analytic review. Psychological Bulletin, 130, Eating and Weight Disorders, 13, 102–​110.
206–​227. Wichstrom, L. (2000). Psychological and behavioral factors
Stice, E., Shaw, H., Becker, C. B., & Rohde, P. (2008). unpredictive of disordered eating:  A  prospective study of
Dissonance-​based interventions for the prevention of eating the general adolescent population in Norway. International
disorders:  Using persuasion principles to promote health. Journal of Eating Disorders, 28, 33–​42.
Preventive Science, 9, 114–​128. Wichstrom, L. (2006). Sexual orientation as a risk factor for
Stice, E., Shaw, H., Burton, E., & Wade, E. (2006). Dissonance bulimic symptoms. International Journal of Eating Disorders,
and healthy weight eating disorder prevention pro- 39, 448–​453.
grams: A randomized efficacy trial. Journal of Consulting and Wilfley, D. E., Agras, W. S., & Taylor, C. B. (2013). Reducing
Clinical Psychology, 74, 263–​275. the burden of eating disorders: A model for population‐based
Stice, E., Shaw, H., & Marti, C. N. (2007). A meta-​analytic prevention and treatment for university and college cam-
review of eating disorder prevention programs: Encouraging puses. International Journal of Eating Disorders, 46, 529–​532.
findings. Annual Review of Clinical Psychology, 3, 207–​231. Wilfley, D. E., & Cohen, L. R. (1997). Psychological treat-
Stice, E., Trost, A., & Chase, A. (2003). Healthy weight con- ment of bulimia nervosa and binge eating disorder.
trol and dissonance-​based eating disorder prevention pro- Psychopharmacology Bulletin, 33, 437–​454.
grams: Results from a controlled trial. International Journal Wilksch, S. M., Durbridge, M. R., & Wade, T. D. (2008). A
of Eating Disorders, 33, 10–​21. preliminary controlled comparison of programs designed to
Striegel-​Moore, R. H., Fairburn, C. G., Wilfley, D. E., Pike, K. reduce risk of eating disorders targeting perfectionism and
M., Dohm, F. A., & Kraemer, H. (2005). Toward an under- media literacy. Journal of the American Academy of Child and
standing of risk factors for binge-​eating disorder in black Adolescent Psychiatry, 47, 939–​947.
and white women:  A  community-​based case-​control study. Wilksch, S. M., & Wade, T. D. (2009). Reduction of shape and
Psychological Medicine, 35, 907–​917. weight concern in young adolescents: A 30-​month controlled
Taylor, C. B., Bryson, S., Luce, K. H., Cunning, D., Doyle, A. evaluation of a media literacy program. Journal of the American
C., Abascal, L. B.,  . . .  Wilfley, D. E. (2006). Prevention Academy of Child and Adolescent Psychiatry, 48, 651–​681.
of eating disorders in at-​risk college-​age women. Archives of Wilksch, S. M., Paxton, S. J., Bryne, S. M., Austin, S. B.,
General Psychiatry, 63, 881–​888. Mclean, S. A., Thompson, K. M.,  . . .  Wade, K. D. (2015).
Taylor, C. B., Kass, A., Trockel, M. E., Cunning, D., Weisman, Prevention across the spectrum:  A  randomized controlled
H., Bailey, J.,   .  .  .  Wilfley, D. E. (2016). Reducing eat- trial of three programs to reduce risk factors for both eat-
ing disorder onset in a very high risk sample with signifi- ing disorders and obesity. Psychological Medicine, 45, 1811–​
cant comorbid depression:  A  randomized controlled trial. 1823. doi:10.1017/​S003329171400289X

Taylor, Fitzsimmons-Craft, Goel 269

Wilson, G. T. (2005). Psychological treatment of eating disor- Winzelberg, A. J., Eppstein, D., Eldredge, K. L., Wilfley,
ders. Annual Review of Clinical Psychology, 1, 439–​465. D., Dasmahapatra, R., Dev, P., & Taylor, C. B. (2000).
Wilson, G. T., & Fairburn, C. G. (1993). Cognitive treatment Effectiveness of an Internet-​based program for reducing risk
for eating disorders. Journal of Consulting and Clinical factors for eating disorders. Journal of Consulting and Clinical
Psychology, 61, 261–​269. Psychology, 68, 346–​350.

270 Prevention
 CH A PT E R


Cognitive-​Behavioral Therapy
14 for Eating Disorders

G. Terence Wilson

Abstract
This chapter discusses cognitive-​behavioral therapy (CBT) as applicable to all eating disorders in adults
and adolescents. It reviews the most recent manual-​based enhanced CBT (CBT-​E), which not only
appears to be more effective than the previous protocol but also is applicable to all eating disorders
and enhances individualizing treatment even within specific diagnoses. The chapter considers the
effectiveness of CBT compared to behavior weight loss treatment, pharmacotherapy, and interpersonal
psychotherapy (IPT). It considers patient access to evidence-​based CBT and discusses effective
dissemination and implementation of competently administered CBT-​E as a research priority. It describes
and considers the effectiveness of a guided self-​help form of CBT (CBTgsh), which provides a brief, cost-​
effective, acceptable, and scalable intervention. It describes possible further development of CBTgsh as a
scalable e-​therapy (using Internet and mobile devices) given that it is a program-​based intervention that
can be widely implemented by nonspecialists.
Key Words:  cognitive-​behavioral therapy, treatment outcome, transdiagnostic, bulimia nervosa, binge
eating disorder, guided self-​help, cost-​effectiveness, treatment scalability, Internet treatment, dissemination

The present chapter focuses on cognitive-​ to maintain the disorder (Fairburn, 2008). The core
behavioral therapy (CBT) for eating disorders psychopathology is assumed to be abnormal over-
in adults and adolescents. As formulated in the evaluation of the importance of body shape and
Diagnostic and Statistical Manual of Mental Disorders, weight that then leads to dysfunctional dieting and
Fifth Edition (DSM-​5) of the American Psychiatric other extreme, unhealthy weight-​control behaviors
Association, the eating disorders are anorexia ner- such as purging. The dysfunctional dieting, in turn,
vosa (AN), bulimia nervosa (BN), and binge eating predisposes the person to binge eating. Purging is
disorder (BED). In addition CBT is also well suited primarily a function of the person trying to com-
to the treatment of what DSM-​5 labels as “Other pensate for the caloric intake involved in binge eat-
Specified Feeding or Eating Disorder” in which all ing but can also come to serve as a means of trying
the required diagnostic criteria for the disorders to reduce or cope with feelings of negative affect.
listed above are met with the exception of one or The CBT treatment derives directly from this model
two specified criteria, for example, BN or BED of and is a theory-​driven, manual-​based intervention
“low frequency and/​or limited duration” (American that is targeted at eliminating the psychopatholog-
Psychiatric Association, 2013, p. 353). ical processes that maintain the disorder, namely,
replacing dysfunctional dieting with a regular and
Bulimia Nervosa healthy pattern of eating, ceasing purging and other
Treatment Model extreme forms of weight control, and decreasing
The standard CBT model of BN spells out the overevaluation of body shape and weight (Fairburn,
psychopathological processes that are hypothesized Marcus, & Wilson, 1993; Wilson, Fairburn, &

271
Agras, 1997). The model and the derivative manual-​ alternative treatment to CBT in order to achieve
based treatment were subsequently revised and optimal results (Wilson & Schlam, 2004).
extended by Fairburn (2008) as “enhanced behavior It is important to emphasize that the Fairburn
therapy” (CBT-​E). A major change from the origi- (2008) model of the maintenance of BN—​ and
nal first-​generation treatment was the reformulation eating disorder psychopathology in general—​ is
of it as an intervention not specifically for BN but supported by the findings of scientific research
for all eating disorder psychopathology. A transdi- on the underlying psychopathology of eating dis-
agnostic treatment, CBT-​E focuses on the common orders. In their influential analysis of the efficacy
processes that maintain different forms of eating of evidence-​based CBT for mental disorders as a
disorder psychopathology as opposed to the tradi- whole, Layard and Clark (2014) reached the fol-
tional categorical diagnoses of DSM-​IV and DSM-​ lowing conclusion:  “Much of the success of CBT
5. Treatment planning is guided not by matching results from its foundation in basic psychological
therapy to different diagnoses, but by “personal- research   .  .  .  researchers ask the question ‘What
ized treatment formulations” (Fairburn, Cooper, & keeps the problem going?’ The psychological pro-
Shafran, 2008). This formulation provides an espe- cesses they identify when answering this question
cially good fit with CBT as a treatment in that it then become the main targets of therapy” (p. 136).
underscores the importance of the functional analy- Consider the example of dysfunctional dieting, or
sis of individual cases of eating disorder—​a seminal what is often referred to as “dietary restraint,” that
feature of behavior therapy from its beginnings. has been shown to be a key risk factor and main-
Fairburn (2008) described two main versions of taining mechanism in BN. It is a core component
CBT-​E: a “focused” (CBT-​Ef ) and a “broad” treat- of the CBT model and a primary and early target of
ment (CBT-​Eb). The former is very similar to the CBT strategies for eating disorders. The goal is to
earlier 1993 manual but has two main changes. reduce dietary restraint and help the patient accept
First, it details a revised strategy and methods for that eating normally and flexibly does not result in
addressing overevaluation of body weight and a loss of control. The inclusion of so-​called forbid-
shape. Second, it provides an explicit treatment den or trigger foods into a regular meal pattern is
module for what is called “mood intolerance” as planned and deliberate.
a specific trigger of binge eating and purging. The Controlled outcome research has shown that
latter, CBT-​Eb, is based on a broader model of reducing dietary restraint is a partial mediator of
the problems (comorbid disorders) that are widely change in the treatment of BN (Wilson, Fairburn,
believed to maintain eating disorders or at the very Agras, Walsh & Kraemer, 2002). This model can be
least complicate their treatment. These are perfec- contrasted with the addiction approach to eating
tionism, low self-​esteem, and interpersonal difficul- disorders. The core principle of the latter is the need
ties, and CBT-​Eb provides an expanded range of for dietary restriction—​increased restraint. Specific
strategies for treating these additional maintaining foods that are alleged to trigger binge eating and loss
mechanisms. It must be emphasized that a major of control must be avoided. It becomes irrelevant
advantage of CBE-​E is that it greatly assists the what techniques are then used to promote food
therapist in individualizing the treatment within avoidance, as it addresses the wrong problem. This
the general framework of a structured protocol so approach is diametrically opposed to the CBT for-
that it matches the patient’s problems. It has long mulation of maintaining mechanisms. No evidence
been argued—​albeit widely rebutted (e.g., Shafran exists showing the efficacy of the addiction model
et al., 2009; Wilson, 1996)—​that the use of treat- for eating disorders (Wilson, 2010).
ment manuals necessarily results in an inflexible and
uniform therapy approach with all patients. Finally, Treatment Efficacy
CBT-​E directly addresses the all-​important issue of The literature on controlled randomized trials
the patient’s motivation for addressing and working of manual-​based CBT for BN has been extensively
to overcome their eating disorder. It engages even reviewed in the past, as summarized in my chap-
the most ambivalent patients by providing a com- ter in the 1st edition of this Handbook—​Wilson
prehensible and credible account of why their eat- (2010). Arguably the most comprehensive analysis
ing problem is self-​perpetuating and what needs to of the evidence on the efficacy of treatment for eat-
be changed to overcome it. As Fairburn (2008) has ing disorders was the NICE guidelines from the
explained, competently conducted CBT-​E is inher- United Kingdom (National Institute for Clinical
ently motivating. There is no need to add some Excellence, 2004). The NICE guidelines comprise

272 Cognitive-Behavioral Therapy

a series of recommendations that are the product therapies to allegiance effects rather than fundamen-
of an interdisciplinary and rigorous process that tal effects of the treatments themselves (Luborsky
includes professional mental health organizations, et al., 1999). In this well-​designed study by Poulsen
academic institutions, and NICE itself. The rec- et al. (2014) the putative role of allegiance effects
ommendations are given a grade ranging from “A” can be dismissed. As Hollon and Wilson (2014)
(reflecting strong empirical support provided by highlight in their critical analysis of this study, the
well-​conducted RCTs) to “C” (reflecting expert psychoanalytic psychotherapy treatment enjoyed
opinion in absence of strong empirical data). The every comparative advantage such as more sessions
NICE guidelines assigned manual-​based CBT for over a much longer time period, implemented at its
BN a rarely given methodological grade of “A.” home site in Copenhagen, and the use of more expe-
Research studies published since 2010 have pro- rienced therapists. Finally, two aspects of the effi-
vided stronger support for the efficacy of CBT cacy of CBT-​E should be stressed: First, a remission
for BN. rate higher than earlier studies using the first-​gen-
eration CBT manual (e.g., Agras, Walsh, Fairburn,
Comparative Treatment Research Wilson, & Kraemer, 2000), and second, the impres-
Research prior to 2010 had shown that CBT sive maintenance of improvement in CBT-​E over
was more effective than either antidepressant medi- the 2-​year period.
cation or other psychological treatment (Wilson, The second recent comparative outcome study
2010). The focus here is on two major more recent was a transdiagnostic comparison of CBT-​E with
comparative treatment outcome studies. interpersonal psychotherapy (IPT) (Fairburn et al.,
The first was randomized controlled trial in 2015). The choice of IPT makes for a stringent
Copenhagen that compared CBT-​Ef with psychoan- comparison, given that IPT is the leading alterna-
alytic psychotherapy in the treatment of 70 patients tive to CBT as a treatment for BN. The Agras et al.
with BN (Poulsen et al., 2014). The CBT-​E com- (2000) study showed that CBT was significantly
prised 20 individual sessions over 5 months whereas more effective than IPT at post-​ treatment, but
psychoanalytic psychotherapy entailed more than IPT showed continued improvement and the dif-
70 weekly sessions over a 2-​year period. The inves- ference at a 1-​year follow-​up was no longer signifi-
tigators ensured that both treatments were imple- cant. Similarly, Fairburn, Jones, Hope, O’Connor,
mented in a highly competent fashion. Therapists & Peveler (1993) found comparable effects between
received initial training from the experts who had the two treatments although IPT was slower in
developed each of the two treatments, and the achieving its results. In addition, IPT is an effective
therapists were then closely supervised. In CBT-​E treatment for BED (discussed later).
this was off-​site via videoconferencing from Oxford. The participants in the study were 138 eating
The psychoanalytic psychotherapy therapists were disorder patients with a BMI between 17.5 and 40.
experienced in their approach (17  years on aver- Fifty-​three (40.8%) had a diagnosis of BN, 6.2%
age), whereas the CBT-​E therapists were much less had BED, and 53.1% were designated as “other
experienced—​2 years on average. Adherence ratings eating disorder” patients. Treatment consisted
conducted by independent raters were strong and of 20 50-​ minute individual sessions. The main
specific for each treatment. result at post-​treatment was that 65.5% of CBT-​
The results of this study were strikingly unam- E patients were in remission (defined as a global
biguous. After 5 months (post-​treatment for CBT-​ Eating Disorder Examination [EDE] score below
E) 42% of CBT-​E patients had completely ceased 1.74) versus 33.3% of IPT patients. Among CBT-​E
binge eating and purging versus only 6% of those patients, 44.8% reported no binge eating or purg-
in the psychoanalytic psychotherapy condition. At ing compared with only 21.7% of IPT participants.
the 2-​year point (post-​treatment for psychoanalytic At the 60-​week follow-​up the proportion of CBT-​E
psychotherapy) the respective numbers of patients patients had increased slightly to 69.4%, whereas
in remission were 44% for CBT-​E and 15% for psy- IPT participants increased to 49%. The difference
choanalytic psychotherapy. At 5 months CBT-​E had between CBT-​E and IPT on the primary outcome
also produced significantly greater improvements in measure of remission was statistically significant
other eating disorder features and general psychopa- at both post-​ treatment and follow-​ up. Fairburn
thology than psychoanalytic psychotherapy. et  al. (2015) emphasize that the CBT-​E results in
It has been commonplace in the comparative this study provide an impressive replication of the
outcome literature to attribute differences between results in an earlier study of CBT-​E (Fairburn et al.,

Wilson 273
2009—​see Wilson, 2010). In that study the rate of from the same geographical catchment area seen in
remission from both binge eating and purging in the 12 months before and after the RCT (Wales,
the BN patients was 45.6% at follow-​up, a figure Palmer, & Fairburn, 2009). The results showed that
higher than that obtained with the earlier version the three different samples of patients were strik-
of CBT. ingly similar. It is clear that controlled treatment tri-
The study has several strengths. As in the als can be designed to be representative on routine
Poulsen et  al. (2013) trial, specific steps were care patient populations.
taken to rule out any explanation in terms of alle- Three additional studies have investigated the
giance effects. Training and weekly supervision outcome of CBT-​E implemented in routine clinical
was the same for both therapies, and independent care settings. In a public-​outpatient eating disorders
ratings of treatment fidelity indicated that both program for youth and adults in Perth, Australia,
were competently implemented. Patient ratings Byrne, Fursland, Allen, and Watson (2011) analyzed
of suitability and expectancy were comparable. the treatment outcome of 125 patients including all
Finally, Fairburn et al. (2015) point out that both eating disorders. Of the total sample those with a
CBT-​E and IPT were originally developed by the BN diagnosis numbered 40. At post-​ treatment
Oxford group, who are known to be advocates of 32.5% of the full sample (using an intent-​to-​treat
both treatments. Indeed, in CBT-​Eb the treat- analysis) were in full remission, namely, no eating
ment of specific interpersonal problems is primar- disorder symptoms over the previous 28 days; 45%
ily based on the principles and procedures of IPT were in either full or partial remission. Of those
(Fairburn, 2008). who completed treatment, 50% were in full remis-
sion. Significant improvements were also evident on
Generalizability of Treatment Effects measures of depression, anxiety, stress, interpersonal
from Controlled Research to Routine problems, self-​esteem, and quality of life. Overall
Clinical Care Settings the results were very similar to those reported in the
A long-​standing criticism of randomized con- original Fairburn et  al. (2009) CBT-​E trial in the
trolled trials (RCTs) of psychological treatments UK. The single main difference between the two
has been that they have tended to exclude patients trials was the significantly higher dropout in the
with more multiple and more complex problems. Australian sample—​40% versus 22.1%.
The reason? The RCTs typically have focused on A second study evaluated the results in 272
a single diagnostic category and participants in BN and eating disorder not otherwise speci-
the study are recruited by the investigators. In fied (EDNOS) patients treated with CBT-​E in a
response it has been well documented that more National Health Service eating disorders clinic in
recent RCTs have included patients with severe Wales (Knott, Woodward, Hoefkens, & Limbert,
psychopathology and high rates of psychiatric 2015). Of the total sample, 74 (27.2%) had BN.
comorbidity. One of the most common reasons A positive treatment outcome defined in terms
for screening out potential participants in RCTs of global EDE-​Q score of less than one standard
is that the individual’s problems are not severe deviation above the community mean was achieved
enough—​do not meet strictly defined DSM-​IV or by 78% of patients who completed treatment and
DSM-​5 criteria—​to warrant inclusion. One of the 39.7% using an intent-​to-​treat analysis. The post-​
strengths of the RCTs on BN summarized here is treatment results were similar to both the Fairburn
that deliberate efforts were made to include partic- et al. (2009) and Byrne et al. (2011) findings, using
ipants who were representative of patients typically a comparable definition. Also noteworthy is that
treated in “real-​world” clinical service settings. For Knott et al. (2015) reported a 40% attrition rate—​
example, the Fairburn et al. (2009) and Fairburn comparable to Byrne et al. (2011) and much higher
et al. (2015) trials of CBT-​E used broad eligibility than the Fairburn et al. (2009) RCT. The third study
criteria so as not to be selective. Furthermore, study of CBT-​E conducted in an NHS setting in England
participants were recruited from a long-​established yielded roughly comparable results to Byrne et al.
community clinic in which the full range of eat- (2011) and Knott et al. (2015) both in terms of out-
ing disorder patients seeking help were treated. In come and attrition rate (Turner, Marshall, Stopa, &
order to determine the degree to which the RCT Waller, 2015). Collectively, these findings provide
samples were broadly representative of the “real-​ convincing evidence that the findings on CBT-​E
world” population, a comparison was made of the from RCTs can, and do, generalize to routine clini-
participants recruited for the RCT with patients cal practice.

274 Cognitive-Behavioral Therapy

Predictors and Moderators Guided Self-​Help
In general robust predictors or moderators of As detailed later in this chapter, greater adoption
treatment outcome in the treatment of BN have and implementation of CBT would be advanced if
yet to be identified. A notable exception to this pat- the intervention were to be made briefer and less
tern has been early response to treatment. Fairburn, complex than the full CBT-​E protocol. Guided self-​
Agras, Walsh, Wilson, and Stice (2004) showed that help based on the principles and procedures of CBT
what has been called an early response to CBT—​ (CBTgsh), provides such an option; CBTgsh com-
in this case a significant reduction in purging by bines a self-​help manual with a limited number of
week 4—​was a strong predictor of outcome at post-​ brief therapy sessions (Fairburn, 1995, 2008).
treatment. This finding was replicated in a second Prior reviews are consistent in showing that
large multisite study in which a 70% reduction CBTgsh can be an effective intervention for BN as
in purging by session 6 (week 4)  predicted thera- compared with a minimal control condition such
peutic success or failure at post-​treatment (Agras as a waiting list (e.g., Banasiak, Paxton, & Hay,
et al., 2000). 2005; Ljotsson et al., 2007; Sysko & Walsh, 2007).
One of the advantages of CBT-​E is that the Schmidt et  al. (2007) compared guided self-​help
protocol requires the therapist to “take stock” or based on cognitive-​behavioral principles using the
systematically evaluate the effects of treatment Schmidt and Treasure (1993) manual in the treat-
early in the course of therapy (after the first 7 ses- ment of adolescents (ages 13 to 20 years) with BN
sions). The goal is to identity nonresponders who, with the Maudsley model of family therapy. Both
the research indicates, are unlikely to show subse- treatments resulted in significant improvement
quent improvement. If the patient is progressing, in binge eating and purging at the end of treat-
no alteration of CBT-​Ef need be made. The absence ment (6  months) and a follow-​up at 12  months.
of significant improvement, however, indicates that Abstinence rates for binge eating and purging
the barriers to change must be identified and treat- combined at 12  months were 36% for CBTgsh
ment modified accordingly. For example, the focus and 41% for family therapy. The CBTgsh resulted
of treatment might be shifted to one or more of in significantly more rapid reduction in binge eat-
the additional maintaining mechanisms included ing. Moreover, CBTgsh was associated with greater
in CBT-​Eb. acceptability and lower cost than family therapy.
Cooper et al. (2016) conducted a detailed analy- As the authors point out, the absence of a control
sis of the findings from the Fairburn et al. (2015) group precludes attributing the results specifically
study comparing CBT-​E and IPT in order to iden- to the treatment, although CBTgsh was more cost-​
tify predictors and moderators of change. Two effective than the comparison therapy.
significant predictors of treatment outcome at the The Mitchell et al. (2011) multisite study is the
60-​week follow-​up assessment emerged. One was largest controlled study of CBTgsh of BN. A com-
that patients with a longer history of their eating plete program of manual-​based CBT (20 individual,
disorder were significantly less likely to benefit 50-​minute sessions) was contrasted with CBTgsh
from either treatment. This finding is consistent (eight, 20-​minute sessions using the Fairburn, 1995
with the results of a review and meta-​analysis of the book) over an 18-​week period. Patients who did not
relevant literature by Vall and Wade (2015). The show a minimum of 70% reduction in purging by
second finding was that higher levels of overevalu- session 6 were offered fluoxetine (60 mg). The rates
ation of the importance of body shape and weight of patients who received fluoxetine were 65% in the
at baseline predicted a poorer outcome in both full CBT treatment and 34% in CBTgsh. At the
CBT-​ E and IPT. This result replicates previous 18-​week post-​treatment assessment, the abstinence
research showing overevaluation of body shape and rates (cessation of both binge eating and purging)
weight to be a negative prognostic factor (Fairburn, were 15% and 11% for full CBT and CBTgsh,
Peveler, Jones, Hope, & Doll, 1993). The only sig- respectively. The remission rates (defined as no lon-
nificant moderator of outcome was that patients ger meeting DSM-​IV criteria) were 57% and 52%,
with lower self-​esteem at pretreatment were more respectively. It should be noted that the patients ran-
likely to respond successfully to CBT-​E than IPT. domized to CBTgsh conditions were significantly
Given the high prevalence of self-​esteem problems more likely to have endorsed a history of anorexia
in BN patients, this finding, assuming it is repli- nervosa, shown to be a negative prognostic indica-
cated in other studies, indicates a significant advan- tor in some studies (Agras et al., 2000). Given the
tage of CBT-​E. lack of a significant difference in outcome between

Wilson 275
the two treatment conditions, Mitchell et al. (2011) disorders, the patients in these studies typically have
concluded, “Therapist-​ assisted self-​
help was an had significant comorbid psychiatric disorders and
effective first-​level treatment” (p. 391), which is in psychosocial problems.
line with the NICE (2004) recommendation.
The Mitchell et  al. (2011) study had several CBT Versus Behavioral Weight Loss
strengths. Adequately powered and conducted by a Treatment
very experienced group of investigators, it included Early treatment outcome studies of BED com-
state-​
of-​
the-​art assessment of outcome (Eating pared CBT with behavioral weight loss treatment
Disorder Examination [EDE] interviews), manual-​ (BWL). At the time it was argued by obesity research-
based treatments, centralized training of therapists, ers that standard obesity treatment was effective for
and weekly supervision using audiorecordings of BED, that BWL was effective, and that there was
therapy sessions. Total therapist contact time was no need for specialty psychological therapies such
16 to 17 h for CBT and 2 to 3 h for CBTgsh. as CBT. Subsequent research, however, showed that
Moreover, therapists who administered CBTgsh manual-​based CBT is more effective overall than
had less experience and training in CBT and eat- BWL (Devlin, Goldfein, Petkova, Liu, & Walsh,
ing disorder treatment. A discrepant finding, as the 2007; Munsch et  al., 2007). Grilo, Masheb, and
authors point out, is the low abstinence rate in the Wilson (2011) randomly assigned 125 obese patients
full CBT condition compared with other major with BED to 16 sessions of either group CBT,
RCTs (e.g., Agras et al., 2000). BWL, or a sequential condition in which CBT was
In sum, despite the limited number of compara- administered first, followed by BWL (CBT _​BWL).
tive outcome studies, the findings on CBTgsh for Attrition rates were relatively high (24% for CBT;
BN are promising. They warrant future research and 31% for BWL; 40% for CBT _​BWL), especially in
clinical application. the combined condition. No significant differences
in remission rates emerged, although CBT produced
Binge Eating Disorder significantly greater reductions in frequency of binge
Binge eating disorder was added to the list of eating at the 6-​and 12-​month follow-​ups.
formal eating disorder diagnoses in DSM-​5 (Attia
et al., 2013). Initially the standard CBT treatment CBT Versus Pharmacotherapy
for BED was basically the Fairburn et  al. (1993) Despite some inconsistent findings, treatment
manual. Although the manual was developed pri- with antidepressant medication has been shown to
marily for treating BN, it contained modifications be superior to pill placebo (Reas & Grilo, 2008).
designed for application to binge eating in both Hence, comparison with CBT provides another
normal weight and overweight or obese patients. As comparative test of the specific efficacy of CBT. In
a transdiagnostic treatment, CBT-​E is now directly an RCT, Grilo, Masheb, and Wilson, (2005) found
applicable to BED (Fairburn, 2008). that CBT was significantly more effective than
either fluoxetine or placebo in producing remis-
Therapeutic Efficacy sion from binge eating. Grilo, Crosby, Wilson,
Manual-​based CBT has been the most inten- and Masheb (2012) subsequently reported the 12-​
sively studied form of psychological treatment of month follow-​up data of 91 of the original 108
BED. The NICE (2004) guidelines concluded study participants:  CBT plus fluoxetine and CBT
that CBT was currently the treatment of choice for plus pill-​placebo did not differ on any outcome
BED. This clinical recommendation was assigned measure, and CBT plus placebo was superior to
a methodological grade of “A,” indicating strong fluoxetine-​only on most measures. The pattern of
empirical support from RCTs. Research has con- results provides robust evidence of the longer-​term
sistently shown that manual-​based CBT produces effectiveness of CBT but not fluoxetine through
remission rates in binge eating between 50% and 12 months after treatment.
70% that are generally well maintained at follow-​ The greater efficacy of CBT compared to pharma-
up. The treatment also reliably results in reduction cological treatment is also evident in an analysis of
in specific eating disorder and general psychopa- RCTs of combined pharmacological treatments for
thology that are generally maintained at a 1-​year BED (Grilo, Reas, & Mitchell, 2016). Combining
follow-​up. Manual-​based CBT, however, does not medication with CBT produced superior outcomes
produce clinically significant improvement in body to pharmacotherapy only, but does not significantly
weight (Wilson, 2010). As is the case with all eating enhance the outcome yielded by CBT only.

276 Cognitive-Behavioral Therapy

Guided Self-​Help comprised eight sessions over a 12-​week period. The
As described above for BN, CBTgsh for BED remission rates from binge eating at the 12-​month
has typically consisted of 8 or 10 brief treatment ses- follow-​up were 64.0% and 44.6% for CBTgsh and
sions using the Fairburn (1995) manual Overcoming TAU respectively. The diagnosis of BED did not
Binge Eating. The guidance has been provided by moderate outcome. Also, CBTgsh resulted in sig-
counselors who have had different levels of training nificant improvement on other indices of eating-​
and expertise in this form of CBT. related psychopathology as well as depression and
The largest controlled treatment study of CBTgsh functional impairment. A replication and extension
for BED compared it with BWL and IPT in a large of this study in the same setting evaluated the impact
sample of 205 overweight and obese patients with of less intensive procedures for recruiting and assess-
the eating disorder (Wilson, Wilfley, Agras, & ing patients to more closely approximate the rou-
Bryson, 2010). Both BWL and IPT consisted of 20 tine service in “real-​world” settings (DeBar et al.,
sessions of individual treatment administered over 2011). The pattern of results in this sample of 160
a 6-​month period, whereas CBTgsh comprised 10 female members of the health maintenance organi-
sessions over this period, 9 of which had a maxi- zation replicated that from the Striegel-​Moore et al.
mum duration of 25 minutes. Interpersonal psy- (2010) study. At the 12-​month follow-​up, CBTgsh
chotherapy was selected as an alternative specialty resulted in significantly greater remission from
psychological therapy, as two previous RCTs had binge eating (35%) than TAU (14%). However, the
shown comparable efficacy between manual-​based magnitude of the effects of CBT was lower than in
CBT and IPT on remission from binge eating, the previous trial.
reductions in body shape and weight concerns, and
associated psychopathology (Wilfley et  al., 1993; Predictors and Moderators
Wilfley et al., 2002). As in BN, rapid response to treatment has been
Post-​treatment findings showed no significant found to be a clinically significant predictor of treat-
differences among the three treatments on remission ment outcome in BED. Grilo, Masheb, and Wilson
from binge eating (Wilson et al., 2002). The remis- (2006) found that rapid response had different
sion rates were as follows: IPT = 64%, BWL = 54%, prognostic significance and time courses across dif-
and CBTgsh = 58%. At the 2-​year follow-​up, how- ferent treatments for binge eating disorder. It pre-
ever, both CBTgsh and IPT not only successfully dicted remission rates of 73% for manual-​based
maintained their improvement, but were also signifi- CBT versus 46% for fluoxetine. Rapid response to
cantly superior to BWL in producing remission from CBT predicted improvement that was sustained or
binge eating. Behavioral weight loss therapy pro- even improved further during the remaining course
duced greater weight loss than either IPT or CBTgsh of treatment. In contrast, when rapid response
at post-​treatment, but not at follow-​up. The results occurred in pharmacotherapy, some of the improve-
provide further evidence that CBT is more effective ment tended to be lost, although it was reasonably
than BWL in eliminating binge eating in overweight maintained during the remaining treatment course.
and obese patients. Moreover, at no point did IPT Importantly, in CBT clinically important findings
differ from CBTgsh on any of the outcome meas- were observed for patients without a rapid response
ures. The design of this RCT also provided a test of to treatment. In CBT, patients without a rapid
the allegiance bias hypothesis in interpreting com- response showed a subsequent pattern of continued
parative outcome trials (Wilson, Wilfley, Agras, & improvement throughout treatment, although it
Bryson, 2011). One of the two treatment sites had did not reach the very high levels of improvement
extensive experience and expertise in IPT, the other achieved by the rapid responders.
in CBT and CBTgsh. The absence of any site × treat- A second study was an analysis of the findings
ment effect rules out an allegiance bias interpretation. from the Grilo et al. (2011) trial comparing CBT
Moreover, there were no significant individual effects and BWL. Rapid response to treatment, defined as
on any measure across both treatments. 70% or greater reduction in binge eating by week
A second study evaluated the efficacy of CBTgsh 4, was evident in 67% of CBT patients and 47%
versus treatment-​as-​usual (TAU) in a sample of 123 of BWL patients. Those treated with CBT did
patients with recurrent binge eating in a large health equally well regardless of rapid response in terms
maintenance organization in the United States of reduced binge eating, but did not show weight
(Striegel-​Moore et al., 2010). Of the full sample, loss. In patients treated with BWL, however, rapid
48% met diagnostic criteria for BED. Treatment responders were significantly more likely to achieve

Wilson 277
binge eating remission (62% vs. 13%) and greater similar binge eating frequencies to class 2 with
reductions in binge eating frequency, eating disor- lower levels of specific eating disorder psychopa-
der psychopathology, and weight loss. thology and compensatory activities. And class
An analysis of the findings from the Wilson et al. 4 was characterized by the highest average body
(2010) trial similarly showed that rapid response mass index, the most overeating episodes, fewest
predicts the outcome of CBTgsh. Defined as 70% binge eating episodes, and an absence of compen-
or more reduction of binge eating by week, rapid satory behaviors. A latent transition analysis found
response in CBTgsh but not IPT or BWL predicted a greater probability of total remission from binge
significantly greater rates of remission from binge eating among patients who received IPT in class 2
eating than nonrapid responders over the course of and CBTgsh in class 3.
the 2-​year follow-​up (Hilbert, Hildebrandt, Agras, The comparatively greater efficacy of IPT for
Wilfley, & Wilson, 2015). Based on this find- class 2 containing the patients with the most eat-
ing, Hilbert et al. (2015) suggested that CBTgsh ing disorder psychopathology is consistent with a
be used as a first-​line treatment in a stepped-​care previous moderator analysis of this trial. Although
model of treatment of BED. Failure to find rapid CBTgsh and IPT had equal effects across the full
response might then lead to implementing IPT, sample, CBTgsh did significantly less well in those
given its equal efficacy with both rapid and nonr- patients who had higher eating disorder psychopa-
apid responders. thology on the EDE. Interpersonal psychotherapy
Overvaluation of body shape and weight, had comparable effects with both the high and low
defined as undue influence of shape and weight on eating disorder psychopathology subgroups (Wilson
self-​evaluation, has been shown to be consistently et al., 2010). This finding has often been interpreted
associated with great severity of eating disorder to mean that IPT is the more robust of the two treat-
psychopathology and predictive of treatment out- ments given its apparent broader efficacy. It would
come in BED (Grilo, 2013). In an analysis of the be premature to draw this conclusion, however. The
Grilo et al. (2011) study, overvaluation was shown CBTgsh in that study was the first-​generation proto-
to be a significant predictor of nonremission from col based solely on Fairburn’s (1995) self-​help man-
binge eating and great frequency of binge eating ual. That manual explicitly eliminated any focus on
at the 12-​month follow-​up even after controlling body shape and weight concerns. Later studies of
for treatment group differences in depression and CBTgsh have included a specific module designed
self-​esteem (Grilo, White, Gueorguieva, Wilson, & to treat overvaluation (e.g., DeBar et  al., 2011;
Masheb, 2012). Striegel-​Moore et  al., 2010). Moreover, Fairburn
Analyses of the data from the Grilo et al. (2005) (2013) has since published a second edition of
comparison of CBT with fluoxetine revealed that Overcoming Binge Eating that includes specific strat-
overvaluation was both a predictor and moderator egies for addressing overvaluation of body shape and
of treatment outcome. It predicted binge eating weight among other additions to the more limited
remission. Perhaps more importantly, it moder- previous manual. Whether or not this more complex
ated remission rates by being significantly related and multifaceted form of CBTgsh enhances effec-
to a poorer outcome in participants receiving tiveness remains to be determined in relevant RCTs.
medication only (Grilo, Masheb, & Crosby, 2012).
Similarly, participants with overvaluation enjoyed Anorexia Nervosa
significantly great reductions in eating disorder psy- Particularly in adults, AN is the most difficult
chopathology and depression if treated with CBT as eating disorder to treat and to study, given its low
opposed to medication. prevalence and the difficulty in recruiting individu-
The findings from Wilson et al. (2010) also bear als with AN to participate in research studies (Agras
importantly on the moderation of treatment out- et  al., 2004). The different treatments that have
come in BED. A latent transition analysis identified been used with AN, including CBT, all received
four different classes within the diagnosis of BED a methodological grade of “C” in the NICE
(Sysko, Hildebrandt, Wilson, Wilfley, & Agras, (2004) guidelines with the single exception of the
2010). Class 1 was characterized by a lower mean Maudsley method of family therapy. The focus here
body mass index and increased physical activity. is on studies conducted since the first edition of this
Class 2 patients reported the most binge eating, Handbook.
shape and weight concerns, compensatory behav- The first study evaluated the efficacy of CBT-​E
iors, and negative affect. Class 3 patients reported in a sample of 99 adult AN patients recruited from

278 Cognitive-Behavioral Therapy

consecutive referrals to clinics in the UK and Italy, improvement in eating disorder psychopathology
respectively (Fairburn et al., 2013). All had an entry and increased BMI. More than 40% of patients
BMI of less than 17.5. The therapy consisted of 40 who completed treatment met criteria for a full
individual sessions of CBT-​Ef over 40 weeks with no response defined as a BMI > 18.5. Improvements
concurrent treatment. The therapists had experience were mostly maintained at a 12-​month follow-​up.
in treating eating disorders and received 6 months Over 80% of eligible patients agreed to the treat-
of initial training in CBT-​Ef from Fairburn and ment, and 85% completed treatment indicating
Cooper in the UK and Dalle-​Grave and Fairburn in good acceptability.
Italy. Weekly supervision sessions were conducted Collectively, the three studies summarized above
throughout the study by Fairburn (UK) and Dalle-​ encourage the further development and application
Grave (Italy). All sessions were recorded and used as for CBT-​E as a treatment for adults and adolescents
part of supervision to ensure competent implemen- with AN.
tation of CBT-​Ef. A large, multisite RCT in Germany compared
Three main findings emerged. First, 64% of CBT-​ E with manual-​ based focal psychodynamic
patients from both samples completed the full therapy and “optimized treatment as usual” (TAU)
treatment. The remaining third of patients either in the treatment of 242 adult patients with AN
dropped out or were withdrawn due to lack of (Zipfel et  al., 2014). Treatment was 10  months
progress or concern about their physical health. in duration. CBT-​E in this study is described as a
Second, completers showed substantial improve- combination of focal and broad forms (Fairburn,
ment in weight and eating disorder psychopathol- 2008). Treatment as usual included outpatient
ogy at post-​treatment. The mean weight gain was psychotherapy and structured care from a family
16.5 lbs, with over 60% meeting criteria for the doctor. Outcome was evaluated at post-​treatment
World Health Organization’s healthy BMI range. and 3-​and 12-​month follow-​ups. All three groups
Third, treatment-​induced improvements were gen- showed weight gain at post-​treatment:  0.93  kg/​m
erally well maintained over a 60-​week follow-​up for CBT-​E, 0.73  kg/​m for focal psychodynamic
period despite little exposure to further treatment. therapy, and 0.69 kg/​m for TAU. At the 12-​month
Although the lack of any control condition is a limi- follow-​up BMI values had increased further in
tation of the study, the results provide preliminary all three groups. No between-​ group differences
support for the treatment and justify further evalu- on BMI—​ the primary outcome measure of the
ation of CBT-​Ef in subsequent RCTs. study—​were statistically significant either at post-​
A second study, adopting the same design, evalu- treatment or follow-​up. At the 12-​month follow-​
ated the efficacy of CBT-​Ef in the treatment of 49 up, patients who had received focal psychodynamic
adolescents (ages 13 to 17) who met DSM-​IV crite- therapy showed a significantly higher recovery rate
ria for AN with the exception of amenorrhea (Dalle-​ than TAU. Recovery was defined using a post hoc,
Grave, Calugi, Doll, & Fairburn, 2013). Treatment combined measure of assessor ratings of patients’
consisted of 40 sessions over 40 weeks. In parallel psychiatric status and BMI; CBT-​E did not differ
findings to the Fairburn et al. (2013) study of adults from either of the other two groups.
with AN, two-​ thirds of the adolescent patients The results of this study are disappointing. The
completed the treatment. The completers showed only significant difference in favor of one of the two
significant improvement in eating disorder psycho- manual-​based specialty therapies in this trial was
pathology and weight gain. The mean BMI centile the 12-​month superiority of focal psychodynamic
increase was 26.9, with a third of patients gain- therapy over TAU on a post hoc combined outcome
ing enough weight to reach 95% of their expected measure. On all other measures TAU did compara-
weight. The results were well maintained over the bly well. And CBT-​E did not improve on TAU on
60-​week follow-​up. any measure of efficacy. In her commentary on this
A third study evaluated the efficacy of an inpa- study Bulik (2014) noted that at post-​treatment and
tient CBT-​ E treatment in 32 AN patients who the 12-​month follow-​up the mean BMI across the
had suffered from the disorder for greater than three treatment groups was still in the underweight
7 years and 34 whose duration was less than 7 years range. Over 25% of patients still suffered from the
(Calugi, El Ghoch, & Dalle-​ Grave, 2017). The full AN disorder—​29% in the focal psychodynamic
treatment was administered over a fixed period of therapy group, 26% in CBT-​E, and 27% in TAU.
20 weeks—​13 weeks inpatient followed by 7 weeks The Zipfel et al. (2014) RCT had several
in a day-​hospital. Both groups showed significant strengths. It is the largest, multisite controlled study

Wilson 279
ever of the psychological treatment of AN. Another even on this limited basis. Reviews by Bodell and
novel feature was the comparison of two promi- Devlin (2010) and Reas and Grilo (2008) of treat-
nent specialty psychological therapies with TAU. ment of BED underscored the fact that both the
Yet there are methodological features of the study longer-​term effects of pharmacotherapy and the
that demand caution in interpreting the findings. impact of discontinuation of medication are gen-
It is imperative that comparative outcome stud- erally unknown. They are not recommended in the
ies take the necessary steps to ensure that the spe- NICE (2004) guidelines. Yet the clinical reality still
cialty treatments in question are implemented in is that antidepressants and other medications are
a competent fashion by appropriately trained and widely prescribed in the United States (e.g., Grilo,
supervised therapists (Fairburn & Cooper, 2011). Crosby, Wilson et al., 2012). In early 2015 the
Studies described earlier in this chapter provide Food and Drug Administration in the United States
specific examples of how this is accomplished (e.g., approved the drug lisdexamphetamine (Vyvanse)
Fairburn et al., 2015; Poulsen et al., 2014). The for treatment of adults with BED. The data on
Zipfel et al. (2014) study does not provide the nec- which this decision was apparently based were
essary information. For example, the initial train- short-​term—​post-​treatment assessment was at 11
ing of CBT-​E therapists was limited to a 2-​day weeks (McElroy et al., 2015). Moreover, the study
workshop by Fairburn. The inadequacies of brief had very restrictive medical and psychopathological
workshop trainings are well documented. Then it inclusion criteria that do not permit generalizability
is unclear whether subsequent training and super- to routine clinical settings.
vision of study therapists met the criteria of ensur-
ing competence in the implementation of CBT-​E. Cost-​Effectiveness
As a result of these RCT design issues, it is unclear The cost-​effectiveness of treatment is increasingly
whether to attribute the Zipfel et al. (2014) findings becoming a consideration in selecting treatments,
to the treatment in question or the manner in which and CBTgsh is an example of a well-​documented
both specialty therapies were implemented. Future cost-​effective intervention. For example, in the
comparative research will need to take account of Wilson et al. (2010) study, CBTgsh and IPT were
these concerns. comparably effective in the total sample. However,
CBTgsh consisted of only 10 treatment sessions
Effectiveness and Scalability of CBT compared with 20 for IPT. Nine of the 10 sessions
The analyses and controversies surrounding the were 25 minutes or less, resulting in total thera-
effects of psychological therapies and what currently pist contact time of 4 to 5 hours versus 18 to 19
constitutes “evidence-​based treatment” have cen- hours for IPT. The CBTgsh therapists were begin-
tered almost exclusively on efficacy. Future analyses ning graduate students versus more senior doctoral
of evidence-​based treatments must necessarily take level therapists implementing IPT. Furthermore,
account of a much broader range of dimensions on their training was significantly shorter than their
which to evaluate and then select for use the diverse IPT counterparts, and, finally, weekly supervision
range of psychological and pharmacological thera- was on an “as needed” basis as opposed to required
pies. This is especially the case given the compelling individual weekly supervision meetings in IPT. In
need to improve the dissemination and implemen- sum, in several ways CBTgsh was far less costly
tation of such treatments on both local and global than IPT although not less effective overall. In the
levels (discussed later). treatment trial for adolescents in England, Schmidt
et al. (2007) reported that their CBTgsh interven-
Efficacy: Short-​and Long-​Term Results tion was less costly but not less effective than the
A major advantage of the CBT treatments sum- comparison family therapy condition.
marized above is that many of the studies support- A formal cost-​ effectiveness analysis was con-
ing their use across the different eating disorders ducted on the findings of the Striegel-​Moore et al.
have reported follow-​up findings of 1 year or more. (2010) study of CBTgsh (Lynch et al., 2010). The
The only alternative therapy with significant follow-​ results were that CBTgsh plus TAU produced signif-
up findings—​albeit less than CBT—​is IPT. icantly more binge-​free days and a lower total soci-
Consider the evidence on pharmacological etal cost over the 12  months following treatment.
treatment. The evidence is limited to short-​term The lower costs of CBTgsh were due to the reduced
treatment trials—​ with CBT proving superior use of TAU services within the health maintenance

280 Cognitive-Behavioral Therapy

organization and were obtained despite the rela- Task-​Sharing
tively high level of therapist supervision. It is well documented how people have very lim-
ited access to mental health treatment both on a
Clinical Range/​Reach global level and even in the United States (Kazdin &
The transdiagnostic nature of CBT-​ E offers Rabbitt, 2013). It is also increasingly accepted that
major advantages in extending the clinical range or one way of addressing this problem is to provide a
reach of CBT. It is the only evidence-​based treat- much greater number of trained mental health pro-
ment for all eating diagnoses. It is applicable not viders than is currently the case. However, relying
just to adults but also to adolescents who are not on trained professionals with advanced degrees, as
underweight (Dalle-​Grave et al., 2013). A transdi- is the practice in the United States, can never meet
agnostic treatment such as CBT-​E allows a men- the need for more mental health providers (Kazdin
tal health provider to be trained in one specialty & Blase, 2011). One solution is task-​ shifting/​
treatment rather than multiple different treatment sharing—​ training less-​
qualified people, including
protocols and still be versatile enough to treat all peers, to take on tasks that have previously been
eating disorders. This reduces the burden of provid- undertaken by more highly qualified individuals.
ing training across different disorders and hence is Patel, Chowdhary, Rahman, and Verdeli (2011)
more cost-​effective to boot. have underscored that “relying on mental health
Also, CBT has been shown to be applicable and professionals to deliver [evidence-​based treatments]
effective across different treatment settings such will only address a tiny fraction of the treatment gap
as specialty eating disorder clinics, general health and task-​shifting these interventions to more avail-
maintenance organizations, and community care able and affordable members of the health work-
centers in different countries. force or community is widely acknowledged to be
the only sustainable way of addressing this barrier”
Brevity (p. 524).
Brief treatment is advantageous not only Task-​sharing does not merely increase the num-
because by definition it is more cost-​effective but ber of available mental health providers. It also
also because it offers a more realistic alternative for offers an optimal means of addressing the challenge
many routine clinical care settings that do not have of serving diverse cultural, ethnic, and racial popula-
the resources or organization to permit full-​length tions. Instead of current practice of training mainly
treatments typically evaluated in RCTs. As pointed White mainstream students through contact and
out earlier in this chapter, CBT-​E has been shown supervision with diverse minority groups as is typi-
to be effective in “real-​world” clinical care settings. cally done in the United States, task-sharing allows
However, as Agras, Fitzsimmons-​Craft, and Wilfley treatment to be implemented by members of the
(2017) caution, some of these studies (e.g., Turner same minority or cultural group. These providers
et  al., 2015) have used a greater number of treat- are then familiar with specific cultural beliefs that
ment sessions than would be feasible in other rou- may influence response to treatments developed in
tine care settings. the in the United States and elsewhere, and profi-
Promoting the practical advantages of brief cient in the specific language of the patients being
treatments does not necessarily mean settling for treated (Chowdhary et al., 2014).
less than optimal treatment. First, there is evidence Task-​sharing has been documented to be effec-
that brief CBTgsh can be as effective as lengthier tive in major RCTs in the treatment of depression
manual-​based CBT treatment for BN and BED) and other disorders in places such as India and rural
(Mitchell et al., 2011; Schmidt et al., 2007; Wilson Pakistan (Patel et al., 2011; Patel et al., 2017). Yet
et al., 2010). Second, it is also important to take the currently there is minimal evidence on the use of
larger implementation context into account. Even if task-​shifting/​sharing in the treatment of eating
a treatment such as CBTgsh were less effective than disorders. What is clear even at this early stage is
full CBT-​E it might still be valuable. As Kazdin that effective treatment can be provided by a wide
and Blase (2011) have pointed out, an “interven- range of counselors. This includes nonspecialist
tion with a larger (effect size) is not invariably better “facilitators” with no formal clinical qualifications
than one with a smaller one. An intervention with a (Carter & Fairburn, 1998), clinically inexperienced
weak but reliable effect that can reach large numbers graduate students (Wilson et al., 2010; Zandberg &
with little cost would be worth having” (p. 33). Wilson, 2013), physicians (Banasiak et  al., 2005),

Wilson 281
and Masters level psychological therapists (Byrne was high. The findings of this exploratory study
et al., 2011; Striegel-​Moore et al., 2010). It should provide proof-​of-​concept for the implementation
also be noted that inexperienced undergraduate col- of culturally adapted CBTgsh for ethnic and racial
lege students with appropriate training have been minority groups.
shown to be effective in implementing an evidence-​
based eating disorder prevention program for their Scalability
peers (Kilpela et  al., 2014). Ultimately, whether A major asset of CBT is its capacity to “scale up”
nonprofessionals can provide effective treatment for treatment so as to provide greater access to treat-
individuals with eating disorders, as has been doc- ment for large numbers of people who have no
umented in other mental disorders such as severe access to mental healthcare—​both in low-​resource
depression (Patel et al., 2017), remains to be seen. countries (Patel et al., 2011) and even in a high-​
resource country such as the United States (Kazdin
Ethnic, Racial, and Cultural Considerations & Rabbitt, 2013). In order to provide greater access
Eating disorders occur among a wide range of to care Kazdin and Blase (2011) urged the adop-
diverse ethnic and racial populations in the United tion of improved means of delivering psychological
States. Clinical research on the acceptability and treatment. They called for (1) the use of nonpro-
efficacy of standard evidence-​based therapies such fessional providers (task-​ sharing) to complement
as CBT in these minority groups is sparse. In the highly trained professionals; (2) self-​help strategies;
biggest analysis to date, Thompson-​Brenner et  al. and (3) technological innovations. More than any
(2013) examined the role of race and ethnicity as other form of psychological therapy, CBT provides
possible predictors and moderators in the data from realistic means of making these innovations.
11 RCTs on BED in the United States. The main Cognitive-​behavioral therapy has lent itself well
finding was the absence of significant differences to task-​sharing and, particularly with respect to eat-
among different ethnic and racial groups in terms ing disorders, the effective use of self-​help treatment.
of treatment outcome. African Americans showed It is important to stress that guided CBTgsh is not
a greater attrition rate than other groups, however. simply a brief form of regular therapist-​adminis-
The authors caution that the overall number of tered treatment. It is a different mode of treatment.
ethnic and racial participants across the 11 RCTs A consistent finding has been that guided self-​help
was still small, possibly limiting the identification provided by a counselor is more effective than
of significant findings. “pure” self-​ help without the accompanying sup-
A qualitative study using focus group method- port from a counselor (Wilson & Zandberg, 2012).
ology was conducted to evaluate the acceptability Fairburn and Patel (2014) emphasize that the guid-
of CBTgsh for Mexican American women with ance or support need not be from a highly trained
BED, BN, or recurrent binge eating in Los Angeles, or professional therapist. In order to be scalable, the
California (Shea et al., 2012). The main finding was treatment is “program-​led” rather than “therapist-​
that these women deemed CBTgsh to be acceptable. led.” The guidance should be supportive or facilita-
They liked the CBT content and the self-​help focus, tive help that can be provided by a nonspecialist.
which was seen as empowering, and responded that Thus far, CBTgsh has been mainly delivered in the
they would recommend the treatment to family and form of Fairburn’s (1995) a self-​help book. It can
friends. The study also identified specific cultural also be provided via the Internet. The latter is more
themes that would be helpful in implementing the scalable and has the added advantage of being able
treatment:  cultural expectations regarding eating to be personalized to match the characteristics of
and body image; family dynamics; and culturally the individual patient’s specific eating problems.
specific foods and eating patterns. Internet-​based CBT treatments have been
A second study then evaluated the acceptability shown to be effective a number of different dis-
and effectiveness of CBTgsh that had been cultur- orders including anxiety and depressive disorders
ally adapted by incorporating the foregoing results (Andersson, 2009; Andrews, Cuijpers, Craske,
(Cachelin et al., 2014). The program resulted in a McEvoy, & Titov, 2010). The evidence on e-​ther-
35.5% remission rate from binge eating and 38.7% apy (Internet and mobile devices) for eating dis-
rate for diagnostic remission. Improvements in orders was reviewed by Loucas et al. (2014) using
body shape and weight concerns and self-​esteem the methodology employed by NICE in the UK.
were also obtained. Satisfaction with the treatment The authors concluded that although some positive

282 Cognitive-Behavioral Therapy

findings emerged, “the value of e-​therapy for eat- state-​of-​the-​art education in behavioral science and
ing disorders must be viewed as uncertain” (p. 122). decision-​making, but this alone will not change
The current enthusiasm for e-​therapy has outpaced clinicians’ behavior. The solution lies in establish-
the research studies of sufficient quality and rigor. ing institutional standards of accountability that
Loucas et  al. (2014) call for improved research, a require practitioners to implement empirically sup-
judgment also registered in another incisive review ported treatments where the data exist (Wilson &
of e-​therapy by Agras et al. (2017). Shafran, 2005). A move in this direction was con-
tained in the 2015 annual report entitled “Health of
Dissemination and Implementation of CBT the 51%: Women” from the chief medical officer of
There is consensus in the field of mental health the National Health System (NHS) in the United
that patients are not receiving evidence-​based treat- Kingdom. In the report she recommended that the
ments in usual clinical care (Kazdin, 2017). Even NHS England should commission services to pro-
when patients do receive these treatments there is vide CBT-​E for eating disorders.
evidence that they are often delivered in subopti- Another reason for the research–​practice gap
mal fashion (Shafran et al., 2009; Wolitzky-​Taylor, is the inadequacy of current training in evidence-​
Zimmermann, Arch, De Guzman, & Lagomasino, based interventions. We need innovative and
2015). The research–​practice gap in mental health improved training methods in evidence-​based psy-
is striking. Eating disorders are no exception chological treatments (Herschell, Kolko, Baumann,
(Lilienfeld et al., 2013). & Davis, 2010).
Suffice it to note a few examples. Wallace and von Training-​as-​usual, comprising a workshop plus
Ranson (2012) administered a Web-​based survey to a treatment manual, does not reliably increase nec-
an international sample of eating disorders prac- essary therapist skills. A major reason is that follow-​
titioners revealing that only roughly half reported up supervision is impractical or unavailable. The
using empirically supported treatments with AN, train-​the-​trainer model has been promoted as an
BN, or BED. Consistent with other surveys, most alternative. In this approach a member of a clinical
practitioners stated that they selectively folded evi- staff is trained to train and supervise other mem-
dence-​based treatments into their preferred eclectic bers of the staff. The goal is to provide continuing
approaches as opposed to implementing empirically feedback and monitoring of treatment integrity.
supported treatments in the form that had been Preliminary findings indicate that the model can
evaluated in controlled research trials. Another sur- be used in the treatment (Zandberg & Wilson,
vey of 80 clinicians treating eating disorder patients, 2013) and prevention (Kilpela et al., 2014) of eat-
and who reported using CBT, revealed that only a ing disorders.
minority actually implemented techniques that Videoconferencing technology provides another
define CBT (Waller, Stringer, & Meyer, 2012). As means of facilitating dissemination and implemen-
has been repeatedly demonstrated, a majority of the tation of CBT. For example, in the Poulsen et  al.
clinicians combined some CBT techniques with (2014) study conducted in Copenhagen described
other non-​evidence-​based methods in an eclectic above, the CBT-​ E therapists were trained and
approach. Waller et  al.’s sobering conclusion was supervised by CBT experts in Oxford using vid-
that “clinicians’ use of the label CBT is not a reli- eoconferencing technology. The most ambitious
able indicator of the therapy that is being offered” solution to training more competent therapists is
(p. 171). “Web-​centered training” (Fairburn & Patel, 2014).
Lilienfeld et  al. (2013) attributed this this evi- In this form of training, a specially designed train-
dence of a research–​practice gap to an “attitudinal ing website provides detailed instruction on the
factor” among clinicians in which they ignored treatment’s use as well as its strategies and proce-
empirically supported evidence in favor of their dures. Web-​centered training may be used alone or
subjective judgment and personal clinical expe- accompanied by guidance.
rience. The solution to this problem they argued If the latter is provided by nonspecialists, as rec-
would be recommending better education and ommended by Fairburn and Patel (2014), Web-​
training of practitioners in the superiority of actu- centered training would be scalable. Web-​centered
arial prediction versus intuitive and subjective clin- training is currently being tested to establish its
ical judgment (Kahneman, 2011; Wilson, 1996). effectiveness and scalability (Fairburn & Patel,
Certainly practitioners should receive accurate and 2014, 2017).

Wilson 283
References Cooper, Z., Allen, E., Bailey-​Straebler, S., Basden, S., Murphy,
Agras, W. S., Brandt, H. A., Bulik, C. M., Dolan-​Sewell, R., R., O’Connor, M. E., & Fairburn, C. G. (2016). Predictors
Fairburn, C. G., Halmi, K. A.,  . . .  Wilfley, D. E. (2004). and moderators of response to enhanced cognitive behaviour
Report of the National Institute of Health Workshop on therapy and interpersonal psychotherapy for the treatment of
Overcoming Barriers to Treatment Reearch in Anorexia eating disorders. Behaviour Research and Therapy, 84, 9–​13.
Nervosa. International Journal of Eating Disorders, 35, Dalle-​Grave, R., Calugi, S., Doll, H. A., & Fairburn, C. G.
509–​521. (2013). Enhanced cognitive behaviour therapy for adoles-
Agras, W. S., Crow, S. J., Halmi, K. A., Mitchell, J. E., Wilson, G. cents with anorexia nervosa:  An alternative to family ther-
T., & Kraemer, H. (2000). Outcome predictors for the cog- apy? Behaviour Research and Therapy, 51, R9–​R12.
nitive-​behavioral treatment of bulimia nervosa: Data from a DeBar, L. L., Striegel-​Moore, R. H., Wilson, G. T., Perrin, N.,
multisite study. American Journal of Psychiatry, 157, 13. Yarborough, B. J., Dickerson, J.,  . . .  Kraemer, H. (2011). Guided
Agras, W. S., Fitzsimmons-​ Craft, & Wilfley, D. E. (2017). self-​help treatment for recurrent binge eating: Replication and
Evoulation of cognitive-​behavioral therapy for eating disor- extension. Psychiatric Services, 62, 367–​373.
ders. Behaviour Research and Therapy, 88, 26–​36. Devlin, M., Goldfein, J., Petkova, E., Liu, L., & Walsh, B. T.
Agras, W. S., Walsh, B. T., Fairburn, C. G., Wilson, G. T., & (2007). Cognitive behavioral therapy and fluoxetine for
Kraemer, H. C. (2000). A multicenter comparison of cogni- binge eating disorder; two-​year follow-​up. Obesity Research,
tive-​behavioral therapy and interpersonal psychotherapy for 15, 1702–​1709.
bulimia nervosa. Archives of General Psychiatry, 57, 459–​466. Fairburn, C. G. (1995). Overcoming binge eating. New  York,
American Psychiatric Association. (2013). Diagnostic and sta- NY: Guilford Press.
tistical manual of mental disorders (5th ed.). Washington, Fairburn, C. G. (2008). Cognitive behavior therapy and eating dis-
DC: Author. orders. New York, NY: Guilford Press.
Andersson, G. (2009). Using the Internet to provide cogni- Fairburn, C. G. (2013). Overcoming binge eating (2nd ed.).
tive behavior therapy. Behaviour Research and Therapy, 47, New York, NY: Guilford Press.
175–​180. Fairburn, C. G., Agras, W. S., Walsh, B. T., Wilson, G. T., &
Andrews, G., Cuijpers, P., Craske, M. G., McEvoy, P., & Titov, Stice, E. (2004). Prediction of outcome in bulimia nervosa
N. (2010). Computer therapy for the anxiety and depres- by early change in treatment. American Journal of Psychiatry,
sive disorders in effective, acceptable and practical health 161, 2322–​2324.
care: A meta-​analysis. PLoS ONE, 5, e13196. Fairburn, C. G., Bailey-​Straebler, S., Basden, S., Doll, H. A.,
Attia, E., Becker, A., Bryant-​Waugh, R., Hoek, H., Kreipe, R., Jones, R., Murphy, R.,  . . .  Cooper, Z. (2015). A transdiag-
Marcus, M.,  . . .  Wonderlich, S. (2013). Feeding and eat- nostic comparison of enhanced cognitive behaviour therapy
ing disorders in DSM-​5. American Journal of Psychiatry, 170, (CBT-​E) and interpersonal psychotherapy in the treatment of
1237–​1239. eating disorders. Behaviour Research and Therapy, 70, 64–​71.
Banasiak, S. J., Paxton, S. J., & Hay, P. (2005). Guided self-​help Fairburn, C. G. & Cooper, Z. (2011). Therapist competence,
for bulimia nervosa in primary care:  A  randomized con- therapy quality and therapist training. Behaviour Research
trolled trial. Psychological Medicine, 35, 1283–​1294. and Therapy, 50, 373–​378.
Bodell, P., & Devlin, M. (2010). Pharmacotherapy for binge-​ Fairburn, C. G., Cooper, Z., Doll, H. A., O’Connor, M.
eating disorder. In C. Grilo & J. Mitchell (Eds.), The E., Bohn, K., Hawker, D. B.,   .  .  .  Palmer, R. L. (2009).
treatment of eating disorders (pp. 402–​414). New  York, Transdiagnostic cognitive-​ behavioral therapy for patients
NY: Guilford Press. with eating disorders: A two-​site trial with 60-​week follow-​
Bulik, C. (2014). The challenges of treating anorexia nervosa. up. American Journal of Psychiatry, 166, 311–​319.
Lancet, 383, 105–​106. Fairburn, C. G., Cooper, Z., Doll, H. A., O’Connor, M. E.,
Byrne, S. M., Fursland, A., Allen, K. L., & Watson, H. (2011). Palmer, R. L., & Dalle-​Grave, R. (2013). Enhanced cognitive
The effectiveness of enhanced cognitive behavioural therapy behaviour therapy for adults with anorexia nervosa: A UK-​
for eating disorders:  An open trial. Behaviour Research and Italy study. Behaviour Research and Therapy, 51, R2–​R8.
Therapy, 49, 219–​226. Fairburn, C. G., Cooper, Z., & Shafran, R. (2008). Enhanced
Cachelin, F. M., Shea, M., Phimphasone, P., Wilson, G. T., cognitive behavior therapy for eating disorders (“CBT-​
Thompson, D. R., & Striegel, R. H. (2014). Culturally E”): An overview. In C. G. Fairburn (Ed.), Cognitive behav-
adapted cognitive behavioral guided self-​help for binge eat- ior therapy and eating disorders (pp. 23–​34). New  York,
ing:  A  feasibility study with Mexican Americans. Cultural NY: Guilford Press.
Diversity and Ethnic Minority Psychology, 20, 449–​457. Fairburn, C. G., Jones, R., Hope, R. A., O’Connor, M., &
Calugi, S., El Ghoch, M., & Dalle-​Grave, R. (2017). Intensive Peveler, R. C. (1993). Psychotherapy and bulimia ner-
enhanced cognitive behavioural therapy for severe and vosa:  Longer-​ term effects of interpersonal psychotherapy,
enduring anorexia nervosa:  A  longitudinal outcome study. behaviour therapy, and cognitive behavior therapy. Archives
Behaviour Research and Therapy, 89, 41–​48. of General Psychiatry, 50, 419–​428.
Carter, J. C., & Fairburn, C. G. (1998). Cognitive-​behavioral Fairburn, C. G., Marcus, M. D., & Wilson. G. T. (1993).
self-​help for binge eating disorder:  A  controlled effective- Cognitive behaviour therapy for binge eating and bulimia
ness study. Journal of Consulting and Clinical Psychology, 66, nervosa:  A  comprehensive treatment manual. In C. G.
616–​623. Fairburn & G. T. Wilson (Eds.), Binge eating:  Nature,
Chowdhary, N., Jotheeswaran, A. T., Nadkarni, A., Hollon, S. assessment and treatment (pp. 361–​404). New  York,
D., King. M., Jordans, M. J. D.,  . . .  Patel, V. (2014). The NY: Guilford Press.
methods and outcomes of cultural adaptations of psychologi- Fairburn, C. G., & Patel, V. (2014). The global dissemination of
cal treatments for depressive disorders:  a systematic review. psychological treatments: A road map for research and prac-
Psychological Medicine, 44, 1131–​1146. tice. American Journal of Psychiatry, 171, 495–​498.

284 Cognitive-Behavioral Therapy

Fairburn, C. G., & Patel, V. (2017). The impact of digital tech- Kazdin, A. E., & Blase, S. (2011). Rebooting psychotherapy
nology on psychological treatments and their dissemination. research and practice to reduce the burden of mental illness.
Behaviour Research and Therapy, 88, 19–​25. Perspectives on Psychological Sciences, 6, 21–​37.
Fairburn, C. G., Peveler, R. C., Jones, R., Hope, R. A., & Doll, Kazdin, A. E., & Rabbitt, S. M. (2013). Novel models for deliv-
H. (1993). Predictors of 12-​ month outcome in bulimia ering mental health services and reducing the burdens of
nervosa and the influence of attitudes to shape and weight. mental illness. Clinical Psychological Science, 1, 170–​191.
Journal of Consulting and Clinical Psychology, 61, 696–​698. Kilpela, L., Hill, K., Kelly, M. C., Elmquist, J., Ottoson, P.,
Grilo, C. M. (2013). Why no cognitive body image feature such Keith, D.,  . . .  Becker, C. B. (2014). Reducing eating disor-
as overvaluation of shape/​weight in the binge eating disor- der risk factors: A controlled investigation of a blended task-​
der diagnosis? International Journal of Eating Disorders, 46, shifting/​train-​the-​
trainer approach to dissemination and
208–​211. implementation. Behaviour Research and Therapy, 63, 70–​82.
Grilo, C. M., Crosby, R. D., Wilson, G. T., & Masheb, R. Knott, S., Woodward, D., Hoefkens, A., & Limbert, C. (2015).
M. (2012). 12-​month follow-​up of fluoxetine and cogni- Cognitive behaviour therapy for bulimia nervosa and eating
tive behavioral therapy for binge eating disorder. Journal of disorders not otherwise specified: Translation from random-
Consulting and Clinical Psychology, 80, 1108–​1113. ized controlled trial to a clinical setting. Behavioural and
Grilo, C. M., Masheb, R. M., & Crosby, R. D. (2012). Predictors Cognitive Psychotherapy, 6, 641–​654.
and moderators of response to cognitive behavioral therapy Layard, R., & Clark, D. M. (2014). Thrive. London,
and medication for the treatment of binge eating disorder. UK: Penguin Books.
Journal of Consulting and Clinical Psychology, 80, 897–​906. Lilienfeld, S. O., Ritschel, L. A., Jay-​Lynn, S., Brown, A. P.,
Grilo, C. M., Masheb, R., & Wilson, G. T. (2005). Efficacy of Cautin, R. L., & Latzman, R. D. (2013). The research-​
cognitive behavioral therapy and fluoxetine for the treat- practice gap:  Bridging the schism between eating disorder
ment of binge eating disorder: A randomized double-​blind researchers and practitioners. International Journal of Eating
placebo-​controlled trial. Biological Psychiatry, 57, 301–​309. Disorders, 46, 386–​394.
Grilo, C. M., Masheb, R. M., & Wilson, G. T. (2006). Rapid Ljotsson, B., Lundin, C., Mitsell, K., Carlbring, P., Ramklint,
response to treatment for binge eating disorder. Journal of M., & Ghaderi, A. (2007). Remote treatment of bulimia
Consulting and Clinical Psychology, 74, 602–​613. nervosa and binge eating disorder:  A  randomized trial of
Grilo, C. M., Masheb, R. M., & Wilson, G. T. (2011). Internet assisted cognitive behavioural therapy. Behaviour
Cognitive-​behavioral therapy, behavioral weight loss, and Research and Therapy, 45, 649–​661.
sequential treatment for obese patients with binge-​eating Loucas, C. E., Fairburn, C. G., Whittington, C., Pennant, M.
disorder: A randomized controlled trial. Journal of Consulting E., Stockton, S., & Kendall, T. (2014). E-​therapy in the
and Clinical Psychology, 79, 675–​685. treatment and prevention of eating disorders:  A  systematic
Grilo, C. M., Reas, D. L., & Mitchell, J. E. (2016). Combining review and meta-​analysis. Behaviour Research and Therapy,
pharmacological and psychological treatments for binge eat- 63, 122–​131.
ing disorder:  Current status, limitations, and future direc- Luborsky, L, Diguer, L., Seligman, D. A., Rosenthal, R., Krause,
tions. Current Psychiatry Reports, 18, 55. E. D., Johnson, S.,  . . .  Schweizer, E. (1999). The research-
Grilo, C. M., White, M. A., Gueorguieva, R., Wilson, G. T., er’s own therapy allegiances: A “wild card” in comparison of
& Masheb, R. M. (2012). Predictive significance of the treatment efficacy. Clinical Psychology:  Science and Practice,
overvaluation of shape/​weight in obese patients with binge 6, 9–​106.
eating disorder:  findings from a randomized controlled Lynch, F. L., Dickerson, J., Perrin, N., DeBar, L., Wilson, G.
trial with 12-​month follow-​up. Psychological Medicine, 43, T., Kraemer, H., & Striegel-​Moore, R. (2010). Cost-​effec-
1335–​1344. tiveness of treatment for recurrent binge eating. Journal of
Grilo, C. M., White, M. A., Wilson, G. T., Gueorguieva, R., & Consulting and Clinical Psychology, 78, 322–​333.
Masheb, R. M. (2012). Rapid response predicts 12-​month McElroy, S. L., Hudson, J. I., Mitchell, J. E., Wilfley, D.,
post-​treatment outcomes in binge-​ eating disorder:  theo- Ferreira-​Cornwell, M. C., Gao, J.,  . . .  Gasior, M. (2015).
retical and clinical implications. Psychological Medicine, 42, Efficacy and safety of lisdexamfetamine for treatment of
807–​817. adults with moderate to severe binge-​eating disorder: A ran-
Herschell, A. D., Kolko, D., Baumann, B., & Davis, A. (2010). domized clinical trial. JAMA Psychiatry, 72, 235–​246.
The role of therapist training in the implementation of psy- Mitchell, J., Agras, W. S., Crow, S., Halmi, K., Fairburn, C. G.,
chosocial treatments: A review and critique with recommen- Bryson, S.,  . . .  Kraemer, H. (2011). Stepped-​care and cog-
dations. Clinical Psychology Review, 30, 448–​460. nitive behavioral therapy for bulimia nervosa:  Randomised
Hilbert, A., Hildebrandt, T., Agras, W. S., Wilfley, D. E., & trial. British Journal of Psychiatry, 19, 391–​397.
Wilson, G. T. (2015). Rapid response in psychological treat- Munsch, S., Biedert, E., Meyer, A., Michael, T., Schlup, B.,
ments for binge eating disorder. Journal of Consulting and Tuch, A.,  . . .  Margraf, J. (2007). A randomized compari-
Clinical Psychology, 83, 649–​654. son of cognitive behavioral therapy and behavioral weight
Hollon, S. D., & Wilson, G. T. (2014). Psychoanalysis or cog- loss treatment for individuals with binge eating disorder.
nitive-​behavioral therapy for bulimia nervosa: The specificity International Journal of Eating Disorders, 40, 102–​113.
of psychological treatments. American Journal of Psychiatry, National Institute for Clinical Excellence. (2004). Eating disor-
171, 13–​16. ders—​Core interventions in the treatment and management of
Kahneman, D. (2011). Thinking, fast and slow. New  York, anorexia nervosa, bulimia nervosa and related eating disorders.
NY: Farrar, Straus, & Giroux. NICE Clinical. Guideline No. 9. London, UK: NICE. www.
Kazdin, A. E. (2017). Addressing the treatment gap: A key chal- nice.org.uk.
lenge for extending evidence-​based psychosocial interven- Patel, V., Chowdhary, N., Rahman, A., & Verdeli, H. (2011).
tions. Behaviour Research and Therapy, 88, 7–​18. Improving access to psychological treatments: Lessons from

Wilson 285
 developing countries. Behaviour Research and Therapy, 49, disorder professionals. Behaviour Research and Therapy, 50,
523–​528. 215–​222.
Patel, V., Weobong, B., Weiss, H., Anand, A., Bhat, B., Katti, Waller, G., Stringer, H., & Meyer, C. (2012). What cogni-
B.,  . . .  Fairburn, C. F. (2017). The Healthy Activity Program tive behavioral techniques do therapists report using when
(HAP), a lay counselor-​delivered brief psychological treat- delivering cognitive behavioral therapy for eating disor-
ment for severe depression, in primary care in India: A ran- ders? Journal of Consulting and Clinical Psychology, 80,
domized controlled trial. Lancet, 389, 176–​185. 171–​175.
Poulsen, S., Lunn, S., Daniel, S. I. F., Folke, S., Bork-​Mathiesen, Wilfley, D. E., Agras, W. S., Telch, C. F., Rossiter, E. M.,
B., Katznelson, H., & Fairburn, C. G. (2014). A random- Schneider, J. A., Cole, A. G.,  . . .  Raeburn, S. D. (1993).
ized controlled trial of psychoanalytic psychotherapy or Group cognitive-​ behavioral therapy and group interper-
cognitive-​behavioral therapy for Bulimia Nervosa. American sonal psychotherapy for the nonpurging bulimic indi-
Journal of Psychiatry, 171, 109–​116. vidual:  A  controlled comparison. Journal of Consulting and
Reas, D. L., & Grilo, C. M. (2008). Review and meta-​analysis Clinical Psychology, 61, 296–​305.
of pharmacotherapy for binge eating disorder. Obesity, 16, Wilfley, D. E., Welch, R. R., Stein, R. I., Spurrell, E. B., Cohen,
2024–​2038. L. R., Saelens, B. E.,  . . .  Matt, G. E. (2002). A randomized
Schmidt, U., Lee, S., Beecham, J., Perkins, S., Treasure, J., Yi., comparison of group cognitive-​behavioral therapy and group
L.,  . . .  Johnson-​Sabine, E. (2007). A randomized controlled interpersonal psychotherapy for the treatment of overweight
trial of family therapy and cognitive-​behavioral guided self-​ individuals with binge eating disorder. Archives of General
care for adolescents with bulimia nervosa or related disor- Psychiatry, 59, 713–​721.
ders. American Journal of Psychiatry, 164, 591–​598. Wilson, G. T. (1996). Manual-​ based treatments:  The clini-
Schmidt, U., & Treasure, J. (1993). Getting better bit(e) by bit(e). cal application of research findings. Behaviour Research and
East Sussex, UK: Psychology Press. Therapy, 34, 295–​315.
Shafran, R., Clark, D. M., Fairburn, C. G., Arntz, A., Barlow, Wilson, G. T. (2010). Cognitive behavior therapy for eating dis-
D. H., Ehlers, A.,   .  .  .  Wilson, G. T. (2009). Mind the orders. In W. S. Agras (Ed.), Handbook of eating disorders.
gap:  Improving the dissemination of CBT. Behaviour New York, NY: Oxford University Press.
Research and Therapy, 47, 902–​907. Wilson, G. T., Fairburn, C. G., & Agras, W. S. (1997). Cognitive-​
Shea, M., Cachelin, F., Uribe, L., Striegel-​ Moore, R., H., behavioral therapy for bulimia nervosa. In D. M. Garner &
Thompson, D., & Wilson, G. T. (2012). Cultural adaptation P. Garfinkel (Eds.), Handbook of treatment for eating disorders
of a cognitive-​behavioral therapy-​guided self-​help program (pp. 67–​93). New York: Gilfoprd Press.
for Mexican American women with binge eating disorders. Wilson, G. T., Fairburn, C. G., Agras, W. S., Walsh, B. T.,
Journal of Counseling and Development, 90, 308–​312. & Kraemer, H. D. (2002). Cognitive behavior therapy
Striegel-​Moore, R. H., Wilson, G. T., DeBar, L., Perrin, N., Lynch, for bulimia nervosa:  Time course and mechanisms of
F., Rosselli, F., & Kraemer, H. C. (2010). Cognitive behavioral change. Journal of Consulting and Clinical Psychology, 70,
guided self-​help for the treatment of recurrent binge eating. 267–​274.
Journal of Consulting and Clinical Psychology, 78, 312–​321. Wilson, G. T., & Schlam, T. R. (2004). The transtheoretical
Sysko, R., Hildebrandt, T., Wilson, G. T., Wilfley, D. E., & model and motivational interviewing in the treatment of
Agras, W. S. (2010). Heterogeneity moderates treatment eating and weight disorders. Clinical Psychology Review, 24,
response among patients with binge eating disorder. Journal 361–​378.
of Consulting and Clinical Psychology, 78, 681–​690. Wilson, G. T. & Shafran, R. (2005). Eating disorders guidelines
Sysko, R., & Walsh, B. T. (2007). Guided self-​help for bulimia from NICE. Lancet, 365, 79–​81.
nervosa. In J. Latner & G. T. Wilson (Eds.), Self-​help for obe- Wilson, G. T., Wilfley, D. E., Agras, W. S., & Bryson, S. W.
sity and binge eating. New York, NY: Guilford Press. (2010). Psychological treatments for binge eating disorder.
Thompson-​Brenner, H., Franko, D. L., Thompson, D. R., Grilo, Archives of General Psychiatry, 67, 94–​101.
C. M., Boisseau, C. L., Roehrig, J. P.,  . . .  Wilson, G. T. Wilson, G. T., Wilfley, D. E., Agras, W. S., & Bryson, S. W.
(2013). Race/​ethnicity, education, and treatment parameters (2011). Allegiance bias and therapist effects:  Results of a
as moderators and predictors of outcome in binge eating randomized controlled trial of binge eating disorder. Clinical
disorder. Journal of Consulting and Clinical Psychology, 81, Psychology, 18, 119–​125.
710–​721. Wilson, G. T., & Zandberg, L. J. (2012). Cognitive-​behavioral
Turner, H., Marshall, E., Stopa, L., & Waller, G. (2015). guided self-​help for eating disorders: Effectiveness and scal-
Cognitive-​behavioral therapy for outpatients with eating dis- ability. Clinical Psychology Review, 32, 343–​357.
orders: Effectiveness for a transdiagnostic group in a routine Wolitzky-​Taylor, K., Zimmermann, M., Arch, J. J., De Guzman,
clinical setting. Behaviour Research and Therapy, 68, 70–​75. E., & Lagomasino, I. (2015). Has evidence-​based psychoso-
Vall, E., & Wade, T. (2015). Prediction of treatment outcome in cial treatment for anxiety disorders permeated usual care in
individuals with eating disorders:  A  systematic review and community mental health settings? Behaviour Research and
meta-​analysis. International Journal of Eating Disorders, 48, Therapy, 72, 9–​17.
946–​971. Zandberg, L. J., & Wilson, G. T. (2013). Train-​ the-​
Wales, J. A., Palmer, R. L., & Fairburn, C. G. (2009). Can treat- trainer:  Implementation of cognitive behavioral guided
ment trial samples be representative? Behaviour and Research self-​help for recurrent binge eating in a naturalistic setting.
Therapy, 47, 893–​896. European Eating Disorders Review, 21, 230–​237.
Wallace, L. M., & von Ranson, K. M. (2012). Perceptions and Zipfel, S., Wild, B., Grob, G., Friederich, Hans-​Christof, Teufel,
use of empirically-​supported psychotherapies among eating H.,  . . .  Herzog, W. (2014). Lancet, 383, 127–​137.

286 Cognitive-Behavioral Therapy

 CH A PT E R


Interpersonal Psychotherapy for the
15 Treatment of Eating Disorders

Natasha L. Burke, Anna M. Karam, Marian Tanofsky-​Kraff, and Denise E. Wilfley

Abstract
Interpersonal psychotherapy (IPT) is a focused, time-​limited treatment that targets interpersonal
problem(s) associated with the onset and/​or maintenance of eating disorders. It is supported
by substantial empirical evidence documenting the role of interpersonal factors in the onset and
maintenance of eating disorders. Interpersonal psychotherapy is a viable alternative to cognitive-​
behavioral therapy for the treatment of bulimia nervosa and binge eating disorder. The effectiveness
of IPT for the treatment of anorexia nervosa requires further investigation. The utility of IPT for the
prevention of obesity is promising. Future research directions include enhancing the delivery of IPT for
eating disorders, increasing the availability of IPT in routine clinical care settings through dissemination
and implementation efforts, exploring IPT adolescent and parent-​child adaptations in diverse and high-​
risk groups, and further exploring IPT for the prevention of eating and weight-​related problems that may
promote full-​syndrome eating disorders or obesity.
Key Words:  interpersonal relationships, social functioning, eating disorder, obesity, interpersonal
psychotherapy (IPT), group therapy

Introduction This chapter provides an overview of interper-

Interpersonal psychotherapy (IPT) is a brief, sonal theory and its foundation for IPT. It provides
time-​limited therapy that focuses on improving a brief review of the literature supporting the cen-
interpersonal functioning and, in turn, psychiatric tral role that interpersonal functioning plays in
symptoms, by relating symptoms to interpersonal the development, manifestation, and maintenance
problem areas and targeting strategies to improve of eating disorders. The delivery of IPT for eating
these problems (Freeman & Gil, 2004; Klerman, disorders is also explained, along with a descrip-
Weissman, Rounsaville, & Chevron, 1984). tion of the major tenets of the treatment. Empirical
Originally developed by Gerald Klerman and col- evidence supporting IPT’s efficaciousness for the
leagues (Klerman et  al., 1984) for the treatment treatment of BN and BED is reviewed, as is the
of unipolar depression, IPT is an efficacious treat- limited data on the use of IPT for AN. A  discus-
ment for bulimia nervosa (BN) (Fairburn et  al., sion of an adaptation of IPT for obesity prevention
1991; Fairburn, Peveler, Jones, Hope, & Doll, follows. Where appropriate, we provide vignettes as
1993) and binge eating disorder (BED) (Wilfley examples. Finally, more recent changes to the deliv-
et  al., 1993; Wilfley, Frank, Welch, Spurrell, & ery of IPT are described, and future directions are
Rounsaville, 1998; Wilfley et  al., 2002; Wilson, proposed.
Wilfley, Agras, & Bryson, 2010). There are lim-
ited data from randomized-​controlled trials on the Interpersonal Theory
effectiveness of IPT in the treatment of anorexia Interpersonal psychotherapy is grounded in
nervosa (AN). theories developed by Meyer, Sullivan, and Bowlby,

287
which hypothesize that interpersonal functioning is The Interpersonal Model for Eating
recognized as a critical component of psychologi- Disorders
cal adjustment and well-​being. In the 1950s, Meyer The interpersonal model for eating disorders sug-
postulated that psychopathology was rooted in mal- gests that problems with social functioning cause
adjustment to one’s social environment (Frank & difficulties with low self-​esteem and negative affect,
Spanier, 1995; Klerman et al., 1984; Meyer, 1957). which then lead to binge eating behaviors (Wilfley,
During the same time period, Sullivan (who was Pike, & Striegel-​Moore, 1997). Data support this
responsible for popularizing the term “interper- model. Eating disorders have been consistently
sonal”) theorized that a patient’s interpersonal rela- associated with poor interpersonal functioning
tionships, rather than intrapsychic processes alone, (Arcelus, Haslam, Farrow, & Meyer, 2013; Wilfley,
established the relevant focus of therapeutic atten- Stein, & Welch, 2005) including interpersonal
tion. Sullivan believed that individuals could not problem-​ solving difficulties, negative attitudes
be understood in isolation from their interpersonal toward emotional expression, and fear of intimacy
relationships and posited that enduring patterns and interpersonal distrust (Arcelus et  al., 2013).
in these relationships could either encourage self-​ Individuals with eating disorders report past difficult
esteem or result in anxiety, hopelessness, and psy- social experiences, problematic family histories, and
chopathology. Interpersonal psychotherapy is also specific interpersonal stressors more often than non-​
associated with the work of John Bowlby (1982), eating-​disordered individuals (Fairburn et al., 1998;
the originator of attachment theory. Bowlby empha- Fairburn, Welch, Doll, Davies, & O’Connor, 1997).
sized the importance of early attachment to the later Persons with bulimic symptoms tend to experience
development of interpersonal relationships and a wide range of social problems, including loneli-
emotional well-​being. He also hypothesized failures ness, lack of perceived social support, and poor
in attachment resulted in later psychopathology. The self-​esteem and social adjustment, and also often
interpersonal roles of major interest to IPT occur demonstrate difficulty with social problem-​solving
within the nuclear family (as parent, child, sibling, skills (Crow, Stewart Agras, Halmi, Mitchell, &
partner); the extended family; the friendship group; Kraemer, 2002; Ghaderi & Scott, 1999; Grissett &
the work milieu (as supervisor, supervisee, or peer); Norvell, 1992; Gual et al., 2002; Johnson, Spitzer,
and the neighborhood or community. & Williams, 2001; O’Mahony & Hollwey, 1995a;
Incorporating aspects of the theories posited by Rorty, Yager, Buckwalter, & Rossotto, 1999; Steiger,
Meyer, Sullivan, and Bowlby, IPT acknowledges a Gauvin, Jabalpurwala, Seguin, & Stotland, 1999;
two-​ way relationship between social functioning Troop, Holbrey, Trowler, & Treasure, 1994; Wilfley,
and psychopathology; disturbances in social roles Wilson, & Agras, 2003). For individuals with
can serve as antecedents for psychopathology, and BED, greater interpersonal problems are related
mental illness can produce impairments in the indi- to earlier onset of binge eating behaviors and per-
vidual’s capacity to perform social roles (Bowlby, sistent, ineffective interpersonal styles (Blomquist,
1982). Therefore, IPT is derived from a theory in Ansell, White, Masheb, & Grilo, 2012). Indeed,
which interpersonal functioning is recognized as a individuals with BED tend to lack interpersonal
critical component of psychological adjustment and problem-​solving skills (Svaldi, Dorn, & Trentowska,
well-​being. It should be noted that IPT makes no 2011), and have interpersonal hostility and dis-
assumptions about the causes of psychiatric illness; tress (Duchesne et  al., 2012) and negative marital
however, IPT does assume that the development and interactions (Whisman, Dementyeva, Baucom, &
maintenance of some psychiatric illnesses occur in a Bulik, 2012). Heightened sensitivity to interper-
social and interpersonal context and that the onset, sonal interactions appears to be a common compo-
response to treatment, and outcomes are influenced nent among individuals with symptoms of eating
by the interpersonal relations between the patient disorders (Evans & Wertheim, 1998; Humphrey,
and significant others. We describe the major tenets 1989; Steiger et  al., 1999; Tasca, Taylor, Ritchie,
of IPT for eating disorders in this chapter. However, & Balfour, 2004; Troisi, Massaroni, & Cuzzolaro,
the extensive empirical background and theoretical 2005). Laboratory paradigms suggest that inter-
foundation, as well as the strategies and techniques personal distress may trigger overeating (Steiger
of IPT, are fully described in a comprehensive book et  al., 1999; Tanofsky-​Kraff, Wilfley, & Spurrell,
by Myrna Weissman and her colleagues (Weissman, 2000) and potentially perpetuate binge eating, as
Markowitz, & Klerman, 2000). evidenced in both clinical and nonclinical samples

288 Interpersonal Psychotherapy

(Ansell, Grilo, & White, 2012; Hartmann, Zeeck, Basic Interpersonal Psychotherapy Concepts
& Barrett, 2010; Ivanova, Tasca, Proulx, & Bissada, Interpersonal psychotherapy has been adapted
2015). When assessed via ecological momentary for a range of clinical disorders (Weissman et  al.,
assessment, interpersonal problems moderate the 2000), but a number of basic concepts are common
relation between negative affect and binge eating across all adaptations of IPT, including treatment
in adults (Ambwani, Roche, Minnick, & Pincus, for eating disorders. Specifically, adaptations for
2015) and predict increases in negative affect and IPT all focus on interpersonal problem areas and
loss of control eating in youth (Ranzenhofer et al., maintain a similar treatment structure. Given the
2014). Similarly, negative affect mediates the rela- time-​limited nature of IPT, treatment success hinges
tion between social problems and loss of control on the therapist’s rapid discernment of patterns in
eating in non-​ treatment-​seeking youth (Elliott interpersonal relationships and the linking of these
et al., 2010). Further, interpersonal difficulties, low patterns to symptoms that may have precipitated
self-​esteem, and negative affect are likely intercon- and continue to maintain the disorder. Thus, in
nected in a reciprocal fashion (Fairburn et al., 1998; IPT for the treatment of eating disorders, treatment
Fairburn et al., 1997; Gual et al., 2002) and serve to centers on facilitating a patient’s awareness of the
perpetuate a cycle, with each factor exacerbating the links among their relationship interactions, negative
other and combining to precipitate and/​or main- affect, and disordered eating symptoms. Early iden-
tain dysfunctional bulimic or binge eating patterns tification of the problem area(s) and treatment goals
(Herzog, Keller, Lavori, & Ott, 1987; Lavender by the therapist and patient is crucial. Throughout
et al., 2016). Individuals with AN also report dif- every session, interpersonal functioning is continu-
ficulties with psychosocial functioning (Hartmann ously linked to the onset and maintenance of the
et al., 2010; O’Mahony & Hollwey, 1995b; Raykos, eating disorder.
McEvoy, Carter, Fursland, & Nathan, 2014;
Ruuska, Koivisto, Rantanen, & Kaltiala-​Heino, Interpersonal Problem Areas
2007) compared with controls and individuals at A primary aim of IPT is to help patients iden-
elevated risk for eating disorders (O’Mahony & tify and address current interpersonal problems. By
Hollwey, 1995b). For individuals with BED, BN, focusing on current as opposed to past relation-
AN, and subclinical eating pathology, negative ships, IPT makes no assumptions about the etiology
affect partially explains the association between of an eating disorder. Treatment focuses on the reso-
interpersonal problems and eating disorder psycho- lution of problems within four social domains that
pathology (Ivanova, Tasca, Hammond, et al., 2015; are associated with the onset and/​or maintenance
Ivanova, Tasca, Proulx, et al., 2015). Most recently, of the eating disorder:  (1)  interpersonal deficits,
Stice and colleagues (2017) report that interpersonal (2) interpersonal role disputes, (3) role transitions,
functioning and negative affect are two of the most and (4)  grief. Interpersonal deficits apply to those
robust risk factors for all types of eating disorders patients who are either socially isolated or who are
(Stice, Gau, Rohde, & Shaw, 2017). Interpersonal involved in chronically unfulfilling relationships.
functioning is also implicated in eating disorder For clients with this problem area, unsatisfying rela-
treatment outcomes with interpersonal problems at tionships and/​or inadequate social support are fre-
the start of treatment relating to poorer treatment quently the result of poor social skills. Interpersonal
outcomes (Jones, Lindekilde, Lubeck, & Clausen, role disputes refer to conflicts with a significant other
2015; Vall & Wade, 2015). Therefore, in theory, (e.g., a partner, other family member, coworker,
the use of an interpersonally focused intervention or close friend) that emerge from differences in
appears to be especially suitable for the treatment of expectations about the relationship. Role transitions
eating disorders (Rieger et al., 2010). Interpersonal include difficulties associated with a change in life
psychotherapy is designed to improve interpersonal status (e.g., graduation, leaving a job, moving, mar-
functioning and self-​esteem, reduce negative affect riage/​divorce, retirement, changes in health). The
and, in turn, decrease eating disorder symptoms. problem area of grief is identified when the onset
of the patient’s symptoms is associated with either
Interpersonal Psychotherapy for Eating the recent or past loss of a person or a relationship.
Disorders Making use of this framework for defining one or
The core tenets of IPT are maintained in the more interpersonal problem areas, IPT for eating
treatment of eating disorders. disorders focuses on identifying and changing the

Burke, Karam, Tanofsky-Kraff, Wilfley 289

maladaptive interpersonal context in which the eat- Diagnosis and Assignment of the Sick Role
ing problem has developed and been maintained. Following a psychiatric assessment, the patient
The four problem areas are discussed in detail in the is formally diagnosed with an eating disorder and
section “Intermediate Phase.” assigned what is termed the “sick role.” The assign-
ment of the sick role is theoretical and serves a prac-
Treatment Structure tical purpose. Consistent with the medical model,
Interpersonal psychotherapy for eating disorders receiving a formal diagnosis reinforces the under-
is a time-​delineated treatment that typically includes standing that the patient has a known condition
15–​20 sessions over 4–​5  months. Regardless of the that can be treated. Accurate diagnosis is essential
exact number of sessions, IPT is delivered in three to successful treatment. Providing a diagnosis also
phases. The initial phase is dedicated to identifying explicitly identifies the patient as being in need
the problem area(s) that will be the target for treat- of help. The sick role is assigned not to demean
ment. The intermediate phase is devoted to working the patient but rather to temporarily exempt the
on the target problem area(s). The termination phase is individual from other responsibilities in order to
devoted to consolidating gains made during treatment devote full attention to recovery. This is particu-
and preparing patients for future work on their own. larly important for individuals with a tendency to
set aside their own needs and desires in order to
Implementing Interpersonal Psychotherapy care for and please others. If appropriate, the IPT
for Eating Disorders therapist might explicitly highlight the patient’s
The three phases of IPT for eating disorders are excessive caretaking tendencies and encourage the
discussed in detail. patient to redirect this energy from others toward
self-​recovery.
The Initial Phase
Sessions 1–​5 typically constitute the initial phase The Interpersonal Inventory
of IPT for eating disorders. The patient’s current A primary and critical component of the ini-
eating disorder symptoms are assessed, and a his- tial phase of IPT is the interpersonal inventory.
tory of these symptoms is obtained. The clinician The interpersonal inventory involves a thorough
provides the patient with a formal diagnosis. The examination of the patient’s interpersonal history.
eating disorder diagnosis and expectations for treat- Although clinicians have historically taken up to
ment are discussed. An assignment of the “sick three sessions to complete the interpersonal inven-
role” (described in further detail below) during tory, we recommend conducting a longer (approx-
this phase serves several functions, including grant- imately 2-​hour) first session to complete the entire
ing the patient the permission to recover, delineat- interpersonal inventory. This allows for patients to
ing recovery as a responsibility of the patient, and get “on board” early in terms of their understand-
allowing the patient to be relieved of other respon- ing of IPT and how their eating disorder fits into
sibilities in order to recover. The therapist explains the IPT rationale (Tanofsky-​Kraff & Wifley, 2009;
the rationale of IPT, emphasizing that therapy will Wilfley, 2008; Wilfley, MacKenzie, Welch, Ayres,
focus on identifying and altering current dysfunc- & Weissman, 2000). The interpersonal inventory is
tional interpersonal patterns related to eating dis- essential for adequate case formulation and devel-
order symptomatology. In order to determine the opment of an optimal treatment plan. The clini-
precise focus of treatment, the clinician conducts cal importance of investing the time to conduct a
an “interpersonal inventory” with the patient and, comprehensive interpersonal inventory cannot
in doing so, develops an interpersonal formulation be overemphasized; accurate identification of the
that specifically relates to the patient’s eating disor- patient’s primary problem area(s) is often compli-
der. In the interpersonal formulation, the therapist cated and is crucial to success in treatment. Table
links the patient’s eating disorder to at least one of 15.1 illustrates the tasks that should ideally be cov-
the four interpersonal problem areas. The patient’s ered during the first session (Dounchis, Welch, &
concurrence with the clinician’s identification of Wilfley, 1999).
the problem area and agreement to work on this The interpersonal inventory involves a review of
area are essential in order to begin the intermediate the patient’s current close relationships, social func-
phase of treatment. Indeed, a collaborative effort is tioning, relationship patterns, and expectations of
promoted throughout the interpersonal inventory relationships. Interpersonal relationships—​both pat-
and all therapy sessions. terns and changes—​are explored and discussed with

290 Interpersonal Psychotherapy

Table 15.1  Tasks of the Initial Session(s) development and symptoms. This exploration pro-
vides an opportunity for the patient to clearly under-
Discuss chief complaint and eating disorder symptoms
stand the relationship between life events, social
Obtain history of symptoms functioning, and the eating disorder, and thereby
clarifies the rationale behind IPT. Upon comple-
Place patient in the sick role
tion of the interpersonal inventory, the therapist and
Establish whether or not there is a history of prior patient collaboratively identify a primary interper-
treatments for the eating disorder or other psychiatric sonal problem area. In some cases, more than one
Problems problem area may be identified. Table 15.2 illustrates
an example of a “Life Chart” (Fairburn, 1997) devel-
Assess patient’s expectations about psychotherapy
oped by an individual with BED and the therapist
Reassure patient about positive prognosis during the interpersonal inventory (Wilfley, 2008;
Wilfley et al., 2000).
Explain IPT and its basic assumptions

Complete an Interpersonal Inventory (detailed review The Interpersonal Formulation

of important relationships) Following completion of the interpersonal
inventory, the clinician will have developed an
  I. review past interpersonal functioning (e.g.,
individualized interpersonal formulation that
family, school, social)
includes the identification of the patient’s primary
  II. examine current interpersonal functioning (e.g.,
family, work, social) problem area. Although some patients may pres-
  III. identify the interpersonal precipitants of ent for treatment with difficulties in several prob-
episodes of eating disorder symptoms lem areas, the time-​limited nature of the treatment
necessitates a focused approach. Therefore, the
Translate eating disorder symptoms into interpersonal clinician should focus treatment on not only
context the problem area(s) that appears to impact the
Explain IPT techniques patient’s interpersonal functioning most but also
those most closely linked to the eating disorder.
Contract for administrative details (i.e., length The therapist, with the agreement of the patient,
of sessions, frequency, duration of treatment, should assign one, or at most two, problem area(s)
appointment times)
on which to develop a treatment plan. We rec-
Provide feedback to patient regarding general ommend that therapists put the agreed-​on goals
understanding of her interpersonal difficulties via IPT in writing and formally present this write-​up to
Problem area (i.e., define interpersonal deficits—​ patients. The presentation of documented goals
loneliness and social isolation) can be a very effective technique that serves as a
treatment “contract” (Tanofsky-​Kraff & Wifley,
Collaborate on a contract regarding the
treatment goals 2009; Wilfley et  al., 2000). The goals developed
at this stage are referenced at future sessions and
Explain tasks in working toward treatment goals guide the day-​to-​day work of the treatment. If more
Adapted from Dounchis et al., (1999). than one problem area is identified, the patient may
choose to work simultaneously on both or may
decide to first address the problem area that seems
reference to the onset and maintenance of eating dis- most likely to respond to treatment. For example,
order symptoms. For each significant relationship, when a patient has role disputes and interpersonal
the following information is assessed:  frequency of deficits, clinical attention might first be focused
contact, activities shared, satisfactory and unsatisfac- on role disputes, since interpersonal deficits reflect
tory aspects of the relationship, and ways that the long-​term patterns that may require considerably
patient wishes to change the relationship. The thera- more time and effort to change. Once the role dis-
pist obtains a chronological history of significant pute has been resolved, the therapist and patient
life events, fluctuations in mood and self-​esteem, decide how to best address the more entrenched
interpersonal relationships, and eating disorder interpersonal deficits. Once the primary problem
symptoms. Throughout this process, the therapist area(s) have been identified and the treatment goals
works collaboratively with the patient to make con- have been agreed on, the initial phase of treatment
nections between life experiences and eating disorder is considered complete.

Burke, Karam, Tanofsky-Kraff, Wilfley 291

Table 15.2  Example of a Personal Historical Timeline of a Patient with BED
Age Problems Relationships Events/​circumstances Moods

5 Normal weight Tonsils are removed

6 Begins gaining weight

14 Grandfather died Feels sad at funeral

but does not cry
because she thinks
it would be a sign of
weakness

15 Concerns about weight; Sister gets married, Perceives parents

first binge; prescribed borrows money from as being extremely
amphetamines to lose parents, and files for disappointed in sister
weight bankruptcy with her
husband

16 Less concern about weight Meets boyfriend, 23, who Does not tell parents Fearful of parents’
because “boyfriend's works at a gas station about boyfriend given disappointment;
ex-​wife was a lot heavier father’s high-​profile worries about their
than me” but began binge job and position in the finding out
eating community

18 Binge eating when alone Becomes engaged Graduates from high

school; goes to technical
school

Loses weight Tells sisters, not parents Abortion

Boyfriend breaks off the Boyfriend “steals” back

engagement the ring (seen on his new
girlfriend); throws herself
into work as a secretary; is
promoted repeatedly

More comfortable about Meets new boyfriend, who Boyfriend’s wife pickets Does not feel guilty
weight (“boyfriend's wife works as a salesman; he says her parents’ house; about the relationship
was a lot heavier than he is separated from his wife, parents do not make
me”) who is pregnant mention of this

Binge eating when alone Lies to family and friends, Moves to Minnesota with Secrecy (wanting
(“food was my only friend telling them that they got boyfriend to be “perfect &
when he was away”); never married not disappoint my
ate when with him parents”); homesick

Spouse of coworker tells her Throws boyfriend out of

he is cheating on her the house; on his way out
he takes her ring from her
jewelry box

27 Binge eating as an outlet Gets pregnant, marries the Lies to mother that she Compliant, scared
father, an alcoholic, who is got pregnant after the
“cruel and verbally abusive” wedding; birth of first
child

28 Husband occasionally shoves “I channeled my energy Hateful

her into my son”

292 Interpersonal Psychotherapy

Table 15.2 Continued
Age Problems Relationships Events/​circumstances Moods

32 “Eating a lot” Husband hits her; she stands Does not tell anyone Scared
up to husband only once, to (“Nobody had a clue
ask him to choose between that we didn’t have a
her and alcohol wonderful marriage”)

Husband no longer drinks Continuing Den Mother Emotionally distant

but continues being verbally activities; very active in (“I made it happy for
abusive church me”)

39 Has sex with husband Husband invests $20,000 Fearful husband will
approximately 2 times a year of their joint money in hit her; obedient;
real estate—​all money proud at holding
lost; patient begins saving onto her feelings;
“every penny,” sending derives esteem from
$5,000 to her sister to keeping her trouble
open a savings account; from her children and
became a workaholic others

41 Eating as a way to “hold Sexual relationship with

everything together” husband ends; although she
does not express anger, he
yells at her, saying he can do
whatever he wants with his
money

46 260 lb., highest weight Marital therapy with

ever; blood pressure clergy for 3 months
increasing with increasing
weight

47 Loses 60 lbs. Patient files for divorce

50 Meets current boyfriend Mother dies Funeral is “a lot

less stressful [than
my grandfather’s]
because I knew it
was ok to cry” Feels
satisfied with their
relationship

51 Regains 30 lbs. Moves in with current

boyfriend

52 Binge eating at night on Does not tell family Works 14+ hour days,
objectively large amounts members she is seeking not pausing to eat or rest
of food at least 3 times psychological help during the day
per week.
Begins psychotherapy
Adapted from Wilfley (2008). Interpersonal Psychotherapy for Binge Eating Disorder (BED) Therapist’s Manual and Wilfley et al. (2000).
Interpersonal Psychotherapy for Group. New York: Basic Books.

The Intermediate Phase task throughout the intermediate phase of IPT for

The intermediate phase typically contains 8–​10 eating disorders is to assist the patient in under-
sessions and constitutes the “work” stage of the standing the connection between difficulties in
treatment. As currently conceptualized, an essential interpersonal functioning and the eating disorder

Burke, Karam, Tanofsky-Kraff, Wilfley 293

behaviors and symptoms. Therapeutic strategies and encouraged to consider new ways of becoming
goals of this phase are shaped by the primary prob- more involved with others and establishing new
lem area targeted in the treatment (see Table 15.3; interests (Wilfley et al., 2005). The distribution of
Wilfley, Dounchis, & Welch, 2004). The following the IPT problem areas among individuals with eat-
sections describe the implementation of specific ing disorders has been reviewed by Wilfley and col-
treatment strategies based on the identified prob- leagues (2003). For 12% of BN patients, grief has
lem area (Wilfley et al., 2005). been identified as their primary problem area, while
approximately 6% of individuals with AN and 6%
Problem Areas with BED present with grief.

The four problem areas addressed in treatment are Role Transitions Role transition includes any dif-
grief, role transitions, interpersonal role disputes, ficulties resulting from a change in life status.
and interpersonal deficits. Common role transitions include a career change
(i.e., promotion, firing, retirement, changing
Grief  Grief is identified as the problem area when jobs), a family change (marriage, divorce, birth of
the onset of the patient’s symptoms is associated a child, child moving out), the beginning or end
with the death of a loved one, either recent or past. of an important relationship, a move, graduation,
Grief is not limited to the physical death of a loved or diagnosis of a medical illness. The goals of ther-
one. Grief can also result from the loss of a signifi- apy include mourning and accepting loss of the old
cant relationship or the loss of an important aspect role, recognizing the positive and negative aspects
of one’s identity. The goals for treating complicated of both the old and new roles, and restoring the
bereavement include facilitating mourning and patient’s self-​esteem by developing a sense of mas-
helping the patient to find new activities and rela- tery in the new role. Key strategies in achieving these
tionships to substitute for the loss. Reconstructing goals will include a thorough exploration of the
the relationship, both the positive and negative patient’s feelings related to the role change as well
aspects, is central to the assessment of not only what as encouraging the patient to develop new skills and
has been lost but also what is needed to counter the adequate social support for the new role (Wilfley
idealization that so commonly occurs. As patients et al., 2005). Thirty-​six percent of patients with BN
become less focused on the past, they should be (Wilfley et al., 2003) and 3.7% of individuals with

Table 15.3  Interpersonal Problem Areas

Main Problem Area Description IPT Strategies

Grief •  Pathological grief stemming from fears of •  Facilitate the patient’s mourning process
being unable to tolerate the painful affect •  Help the patient reestablish interest in
associated with the loss relationships to substitute for what has been
lost

Interpersonal role •  Disputes with partner, children, or other •  Identify the dispute
disputes family members, friends, or coworkers •  Choose a plan of action
•  Modify expectations and faulty
communication to bring about a satisfactory
resolution

Role transitions •  Economic or family change including •  Mourn and accept the loss of the old role(s)
children leaving for college, new job, divorce, •  Restore self-​esteem by developing a sense of
retirement, parent’s caretaker mastery regarding the demands of the new
role(s)

Interpersonal •  A long-​standing history of social isolation, •  Reduce the patient’s social isolation
deficits low self-​esteem, loneliness, and an inability to •  Encourage the formation of new
form or maintain intimate relationships relationships
Adapted from Wilfley, Dounchis, and Welch (2004). Interpersonal psychotherapy of anorexia nervosa. In K. M. Miller, & J. S. Mizes (Eds.),
Comparative treatment of eating disorders. New York: Springer Publishing Company, and Dounchis et al. (1999). Using group interpersonal
psychotherapy (IPT-​G) for the treatment of binge eating disorders. Paper presented at the Academy of Eating Disorders Annual Meeting, San Diego.

294 Interpersonal Psychotherapy

BED in Wilfley and colleagues’ (2000) trial were Therapeutic Strategies
identified with the problem area of role transitions.
Among individuals with AN, approximately 17% In addition to general therapeutic techniques, there
present with role transitions as the primary problem are several strategies therapists can use to maximize
area (Wilfley et al., 2003). IPT’s effectiveness.

Interpersonal Role Disputes  Such disputes are con- Therapeutic Stance As with most therapies, IPT
flicts with a significant other (e.g., a partner, other places importance on establishing a positive thera-
family member, employer, coworker, teacher, or peutic alliance between therapist and patient. The
close friend), which emerge from differences in IPT therapeutic stance is one of warmth, support,
expectations about the relationship. The goals of and empathy. Further, throughout all phases of the
treatment include clearly identifying the nature of treatment, the clinician is active and advocates for
the dispute and exploring options to resolve it. It the patient rather than remaining neutral. Issues and
is important to determine the stage of the dispute; discussions are framed positively so that the thera-
once the stage of the dispute becomes clear, it may pist may help the patient feel at ease throughout
be important to modify the patient’s expectations treatment. Such an approach promotes a safe and
and remedy faulty communication in order to bring supportive working environment. Confrontations
about adequate resolution. It may be particularly and clarifications are offered in a gentle and timely
helpful to explore how nonreciprocal role expecta- manner, and the clinician is careful to encourage
tions relate to the dispute. If resolution is impossi- the patient’s positive expectations of the therapeutic
ble, the therapist assists the patient in dissolving the relationship. Finally, the therapist conveys a hopeful
relationship and in mourning its loss (Wilfley et al., and optimistic attitude about the potential for the
2005). This problem area is identified in approxi- patient to recover.
mately 64% of individuals with BN and 33% of
those with AN (Wilfley et al., 2003). Interpersonal Focusing on Goals  Because IPT is a directed, goal-​
role disputes were present in 29.6% of the patients oriented therapy, therapists should maintain a
in the Wilfley et al. (2000) BED trial. focus each week on how the patient is working on
his/​her agreed-​on goals between sessions. Phrases
Interpersonal Deficits  Interpersonal deficits include such as “moving forward on your goals” and
patients who are socially isolated or who are in “making important changes” are used to encour-
chronically unfulfilling relationships. The goal is age patients to be responsible for their treatment
to reduce the patient’s social isolation by helping while also reminding them that altering interper-
enhance the quality of existing relationships and sonal patterns requires attention and persistence.
encouraging the formation of new relationships. To Sometimes during the course of therapy, unfo-
help these patients, it is necessary to determine why cused discussions arise. The therapist should sensi-
they have difficulty in forming or maintaining rela- tively, but firmly, redirect the discussion to the key
tionships. Carefully reviewing past significant rela- interpersonal issues. By explicitly addressing goals
tionships will be particularly useful in making this each week, the patient can work toward necessary
assessment. During this review, attention should changes. This goal-​oriented focus has been sup-
be given to both the positive and negative aspects ported by research on IPT maintenance treatment
of the relationships, as well as an investigation of for recurrent depression, which has demonstrated
potentially recurrent patterns in these relationships. that the clinician’s ability to maintain focus on
It may also be appropriate to examine the nature of interpersonal themes is associated with better out-
the patient–​therapist relationship, since this may be comes (Klerman et al., 1984; Markowitz, Skodol,
the patient’s only close relationship and it is present & Bleiberg, 2006; Weissman et al., 2000). In IPT
to be observed (Wilfley et  al., 2005). For patients for eating disorders, it is essential that the clinician
with BN and AN, this problem area is seen in facilitate and strengthen the recognition of connec-
approximately 16% and 33%, respectively (Wilfley tions between patients’ problematic eating and dif-
et al., 2003). Based on one study, interpersonal defi- ficulties in their interpersonal lives. The example
cits appeared to be the most commonly identified that follows illustrates how the IPT therapist initi-
problem area among individuals with BED; 60.5% ates the discussion about goals and helps a patient
of patients presented with interpersonal deficits in treatment for BN with interpersonal deficits
(Wilfley et al., 2000). work on her goals.

Burke, Karam, Tanofsky-Kraff, Wilfley 295

 Therapist: Ellie, I would like to check in with you As this dialogue between Ellie and her therapist
to see how your work on your goals is progressing. Last illustrates, a crucial component to IPT for eating
week, you mentioned that you are starting to become disorders is helping to facilitate and strengthen the
more aware of interpersonal difficulties that trigger connections patients make between their problem-
your binge eating and purging. atic eating and difficulties they have in their inter-
Ellie: I have been paying more attention to what is personal lives. Focusing on specific goals provides
happening when I binge. It seems as though there are the structure for this to be accomplished. Ellie’s
a lot of times that I feel the urge to binge, whether it ability to have insight to make links between her
is feeling put down at work, feeling angry at my hus- interactions, mood, and disordered eating is due to
band, or feeling overwhelmed about taking care of the the therapist’s persistence in emphasizing the con-
kids. I think that I am beginning to better understand nection between Ellie’s interpersonal functioning
what happens with me when I get the urge to binge. to her eating patterns throughout all phases of the
I feel overwhelmed—​I have so much going on in my treatment.
life, I do not know how I will ever overcome the desire
to binge. Making Connections During the intermediate
Therapist: I imagine that it must feel very fright- phase, it is crucial that the therapist assist patients
ening when you have so much going on—​it can seem in recognizing, and ultimately becoming more
as though gaining control over your eating might be aware of, the connections between eating difficul-
impossible. However, you have taken a very important ties and interpersonal events between sessions. As
first step, Ellie. It is great that you have begun to iden- patients learn to make these connections, the thera-
tify triggers to your binge eating. From your work, it pist should guide them to develop strategies to alter
is clear that a lot of things are playing into your desire the interpersonal context in which the disordered
to binge. How were you able to become more aware of eating symptoms occur. As a result, the cycle of the
what was happening with you when you felt the urge eating disorder is interrupted. Patients are encour-
to binge? aged to make connections between interpersonal
Ellie:  I think that instead of just binge eating as functioning and eating patterns that are positive as
soon as I  feel the urge, I  think I  have become aware well. For example, an individual may recognize that
that something changes for me when I have the urge to communication improved with a significant other
binge. Lately, I have been trying hard to stop and see and, as a result, the patient did not engage in eating
what is going on and what I am feeling before I binge. disorder behaviors. To encourage positive and nega-
Even though I have still had binge episodes this week, tive connections, clinicians should ask the patient
I think that at least once or twice I seemed to have lost about his/​her eating patterns between sessions and,
the desire to binge once I stopped and thought about if there were any changes, inquire about any rec-
what was going on in that moment. ognized links between eating patterns and interper-
Therapist: What specifically did you notice was sonal functioning.
happening with you? In the following vignette, which is also in the
Ellie: I noticed that I was feeling frustrated about intermediate phase, the therapist encourages a
my work and how angry I feel at my husband when he patient with interpersonal role disputes in treat-
makes me feel like I am an inadequate wife and par- ment for AN to talk about the connections she has
ent. I realized that I often do not stand up for myself made between her desire to restrict her food intake
and let the people in my life know what my needs and difficulties she experiences with her divorced
are. I  don’t feel like I  know how to do this and end parents:
up expressing—​or is it suppressing?—​my frustration
through food. Therapist: Ashley, have you noticed any connec-
Therapist: This is very important work you are tions between your eating and how things went with
doing, Ellie—​good job! You have identified some really your parents this past week?
important interpersonal triggers for your binge eating. Ashley: Well, my parents have been arguing a lot—​
As you continue to become more and more aware of the mostly about where I  am going to spend the summer
circumstances surrounding your urge to binge, we can vacation. I just can’t stand it. Whenever they talk badly
begin to work on helping you find more effective ways about the other to me, I’m not hungry at all and I just
to manage your feelings and relationships so that you want to starve myself. Maybe this is a way to make
are less likely to binge. them notice how their fighting gets to me. I don’t know.

296 Interpersonal Psychotherapy

 Therapist: This is terrific work Ashley! One of so frantic that I didn’t even bother to heat up—​I ate a
the things we have been working on is getting you to huge container of soup cold!
become more aware of what is happening when you feel Therapist: Last week you talked about how eating
the urge to restrict most intensely. You have just made is a way for you to relieve stress and to relax. Instead of
an important connection between your stress and dis- allowing yourself a break or sharing your feelings with
like related to your parents’ arguments and your want- family or friends, you tend to turn to food.
ing to restrict. How do think restricting your eating Terry: That is exactly what I do—​and what I did
affects your mood? this week. It was unbelievable and I was so disgusted
Ashley: Well, focusing on not eating helps me not to with how I  was shoving food in by the mouthful—​
think about all the stuff going on between my parents. cookies, chips, leftover pizza. Have you ever eaten cold
I feel kind of numb. I guess I never realized that con- pizza? It does not even taste good! I just wanted  . . . 
nection until we started working together. All I cared Therapist:  I am going to interrupt you for
about was how much I wanted to be thin. a moment, Terry, so that I  can refocus you for a
Therapist: Now that you can more clearly see that moment, back to your goals. How has your work in
connection, how would you like to start working on finding down time to take care of yourself been com-
your relationships with your mother and father? ing along?
Terry: I think it has been going somewhat better.
Redirecting Issues Related to  Eating Disorder I signed up to join a book club through my church. It
Symptoms During treatment sessions, patients hasn’t started yet, but I did go buy the book we’re sup-
with eating disorders may raise issues relating to posed to read and I  started reading a few pages. I’m
eating disorder symptoms that are distressing (for hoping that my work schedule does not conflict with
example, binge episodes, overconcern about eating, the nights that the club will be meeting.
shape, and weight) or want to engage in extended Therapist: As we talked about last week, for a long
discussion related to these behaviors. These issues time you’ve been feeling that you need to take care
are relevant insofar as they reflect the clinical status of everyone else—​make everyone else happy—​and, in
of the patient’s eating disorder. However, the thera- doing so, you’ve put yourself last and not tended to
pist must be cognizant of how these issues are being your own needs. By not taking care of yourself, you get
discussed during the sessions and vigilantly keep the very stressed and use food to cope. I wonder if as you
session focused on the patient’s treatment goals by practice identifying your own needs and addressing
gently, but firmly, redirecting discussion to work on them—​like you have in joining the book club—​you
the treatment goals. For example, a female patient will feel less exhausted and more personally cared for.
who avoids intimacy with her husband may attri- By taking time out for yourself, you will feel calmer
bute her avoidance to body dissatisfaction related to and be less likely to turn to food. Joining the book
her obesity. She may wish to discuss her body con- club—​and beginning to read—​is already one way you
cerns at great length to circumvent actual difficul- are taking care of yourself. I have noted that you only
ties in communication with her husband, or she is had two overeating episodes this past week. When we
not yet aware that her relationship difficulties with first started working together, you were having binge
her husband are an important issue and that body episodes almost nightly during similarly stressful times.
concern is what she experiences as most distress-
ing. Dialogue related to eating disorder symptoms By redirecting the patient away from the specif-
should be consistently and repeatedly linked to its ics of her binge episode, and toward her interper-
functional role with regard to the identified inter- sonal problem area, the therapist is able to keep the
personal problem area(s). patient focused on her goals.

Therapist: How are you today, Terry? General Therapeutic Techniques The IPT therapist
Terry: I have been busier than usual and that has differs from providers of other modalities in that
made me very stressed. I was asked to cover the late shift throughout the course of treatment, he/​she main-
twice this week and since we have been understaffed, tains a constant focus on the interpersonal context
I was the only nurse to cover all of the patients during of the patient’s life and its link to the eating disor-
my shift. I had no time to eat dinner, which was prob- der symptoms. Although this approach is unique to
ably good for me, but then I came home both nights IPT, a number of the therapeutic techniques used
and had huge binges. I  found myself in the kitchen in IPT are similar to those used in other therapies.
eating anything and everything that I could find. I was Such techniques include exploratory questions,

Burke, Karam, Tanofsky-Kraff, Wilfley 297

encouragement of affect, clarification, communica- However, a primary goal is helping patients to
tion analysis, and use of the therapeutic relationship. identify, understand, and acknowledge their
1.  Exploratory Questions. Use of general, open-​ feelings whether or not they choose to verbalize
ended questions often facilitates the free discus- them to others. The following is an example:
sion of material. This is especially useful during the “The therapist immediately noticed that Sara
beginning of a session. For example, the clinician was silent and withdrawn at the beginning
might open a session with, “Tell me about your of the session. Initially, she denied any
relationship with your husband.” Once this has relationship between her nonverbal behavior
generated discussion, progressively more specific and the therapist’s observation. The therapist
questioning should follow. For instance, after the was persistent and she eventually acknowledged
patient describes an interpersonal interaction with that she was feeling hurt because her father had
her husband, the therapist might follow-​up by ask- not acknowledged her son’s first birthday. She
ing, “What happened with (or what changes did spent some time clarifying and expressing her
you notice in) your eating patterns after you talked feelings of anger and rejection with regard to
with your husband?” her own relationship with her father. The issue
2.  Encouraging Affect. Interpersonal psycho- that emerged in the session was “when do you
therapy’s focus throughout the therapeutic process stop wanting something from a parent that you
involves affect evocation and exploration (Wilfley can never get from them?” Even though she
et al., 2000). This is particularly relevant for patients became aware of and expressed many painful
with eating disorders because problematic eating feelings regarding her relationship with her
often serves to regulate negative affect. The IPT father, Sara’s goal was not to go out and express
therapist should assist patients in (1)  acknowledg- these feelings to her father directly at this time.
ing and accepting painful emotions, (2) using affec- Instead, Sara and her therapist began to discuss
tive experiences to facilitate desired interpersonal how she can find herself more fulfilled and
changes, and (3)  experiencing suppressed affect satisfied by working to make other choices in
(Wilfley, 2008; Wilfley et al., 2000). terms of who to turn to for support and care”
(Wilfley, 2008).
a.  Encourage acceptance of painful
c.  Help the patient experience suppressed
affects. Patients with eating disorders are
affects. Many who struggle with eating
often emotionally constricted in situations
disorders are emotionally constricted in
when others would typically experience
situations where strong emotions are normally
strong emotions. In the case of BN and BED,
felt. An example may be the patient who is
individuals use food to cope with negative
unassertive and does not feel anger when their
affect. Therapy provides an arena to experience
rights are violated. On the other hand, they
and express these feelings versus using food
may feel anger but may lack the courage to
to cope with these feelings. As the feelings are
express it in an assertive manner. Sometimes
expressed, it is important for the IPT therapist
patients will deny being upset, when it is clear
to validate and help the patient accept them
that an upsetting interaction has just occurred.
(Wilfley, 2008).
The therapists might say, “Although you said
b.  Teach the patient how to use affect in
you were not upset, it appears to me that you
interpersonal relationships. Although the
have shut down since you talked about the
expression of strong feelings in the session
situation with your husband.” In this way, the
is seen as an important starting point for
therapist will attempt to draw out affect when it
much therapeutic work, the expression of
is suppressed (Wilfley, 2008).
feelings outside the session is not a goal in and
of itself. The goal is to help the patient act 3.  Clarification. Clarification is a useful tech-
more constructively (e.g., not binge eating or nique that can (1) increase the patient’s awareness
purging) in interpersonal relationships, and this about what she/​ he has actually communicated,
may involve either expressing or suppressing and (2) draw awareness to contradictions that may
affects, depending on the circumstances. A have occurred in the patient’s presentation of inter-
goal for the patient in IPT is to learn when actions or situations. An example might involve
her/​his needs are met by expressing affect and contradictions between the patient’s affect and
when they are better met by suppressing affect. speech: “While you were telling me how upset you

298 Interpersonal Psychotherapy

are about your father, you had a smile on your face. particularly relevant to and useful for patients with
What do you think that’s about?” interpersonal deficits and interpersonal role dis-
4.  Communication Analysis. The technique putes. Use of this technique offers the patient the
of communication analysis is used to (1)  iden- opportunity to understand the nature of his/​her
tify potential communication difficulties that the difficulties in interacting with others and provides
patient may be experiencing and (2)  assist the the patient with helpful feedback on his/​her inter-
patient in modifying ineffective communication actional style. The following is an example of using
patterns. In using communication analysis, the the therapeutic relationship.
therapist asks the patient to describe, in great detail,
a recent interaction or argument with a significant Therapist: Joe, I  know it was hard for you last
other. As the patient describes the interactions, the week to talk about how your girlfriend does not under-
therapist garners information by using probes, such stand that it is important for you to have time with
as the examples below (Mufson, Dorta, Moreau, & your friends. Did you have a chance to discuss this with
Weissman, 2004; Young & Mufson, 2003): her over the past week?
Joe: No. I  was really busy. She would have just
“What did you specifically say?” argued with me if I brought it up anyway.
“What did he/​she say in response?” Therapist: As we have talked about before, I am
“Then what happened?” wondering if you approach her about the topic differ-
“How did you feel?” ently this week, she might be more open to your point
“Do you think you might be able to tell him/​her of view.
how you felt? Joe: I doubt it. I am really busy this week, too.
“Thinking back to how the interaction turned out, Therapist: It feels to me like you do not want help
did you send the message that you wanted to convey?” with this situation, so I am feeling a little frustrated
“How do you think it made him/​her feel?” right now. I am wondering if other people in your life
might feel the same way. What do you think?
As part of communication analysis, the clinician
then assists the patient in identifying ways in which In a nonjudgmental and straightforward manner,
the interaction could have gone differently and how the therapist not only models clear communication
the different manifestations might impact the other with Joe, but also uses the therapeutic relationship
person’s feelings and reactions. Therapeutic que- to identify a potentially dysfunction communica-
ries to facilitate this process include (Mufson et al., tion pattern.
2004; Young & Mufson, 2003):
The Termination Phase
“How do you think this interaction might have By the end of the intermediate phase, patients are
manifested differently?” often acutely aware that treatment will soon be end-
“What could have been said differently by either ing. The clinician should begin to discuss termination
you or the other person?” explicitly and address any anxiety the patient may be
“How might it have changed the way that felt experiencing. In doing so, the patient should be pre-
and/​or the interaction itself?” pared for emotions that may arise with termination,
including grief related to the ending of treatment. At
The objective is for the clinician and patient to times, patients may deny any emotion with regard to
collaboratively work to identify difficulties in com- the end of treatment and appear to have little reaction to
munication that may be impacting the process and termination. Nevertheless, the therapist should clearly
outcome of the interaction and to find more effec- address termination, as the patient may be unaware of
tive strategies. or avoiding affect related to the end of treatment.
5.  Use of the Therapeutic Relationship. The prem- The termination phase typically lasts 4–​5 ses-
ise behind this technique is that all individuals sions. During this phase, the patient should be
have characteristic patterns of interacting with oth- encouraged to reflect on the progress that has
ers. The technique involves exploring the patient’s been made during therapy—​both within sessions
thoughts, feelings, expectations, and behavior in and outside of the therapeutic milieu—​ and to
the therapeutic relationship and relating these to outline goals for remaining work after the formal
the patient’s characteristic way of behaving and/​ end of treatment. Interpersonal psychotherapy
or feeling in other relationships. This technique is does not assume that the work toward changes

Burke, Karam, Tanofsky-Kraff, Wilfley 299

in interpersonal functioning is complete after the know those girls talk about me behind my back. Why
last session of the therapy. Rather, patients and were you sitting with them?”
therapist collaboratively summarize and draft the
remaining work for the patient to continue out- Therapist: That’s a good start. What do others
side of the therapeutic milieu. Patients are encour- think?
aged to identify early warning signs of relapse Becca: I guess I  would have felt bad if I  was
(e.g., binge eating, overeating and excessive dietary Christine.
restriction, negative mood) and to prepare plans of Therapist: How so?
action. Patients are reminded that eating disorder Becca: Well, Christine might have felt accused of
symptoms tend to arise in times of interpersonal doing something wrong. I guess she might have felt as
stress and are encouraged to view such symptoms though you think she is not allowed to hang around
as important early warning signals. The identifica- with whomever she wants. I think that is how I might
tion of potential strategies to cope with such situ- have taken it.
ations is designed to increase the patient’s sense Therapist: Thanks, Becca. What do you think,
of competence and security. Nevertheless, it is also Sheila?
essential to assist patients in identifying warning Sheila: I did not mean to tell her that she cannot
signs and symptoms that may indicate the need for hang out with other people. I just wanted her to know
professional intervention in the future. that it made me feel bad that she was spending time
with girls who are not nice to me.
Use of a Group Therapist: Does anyone have thoughts about how
The group setting frequently provides an optimal Sheila might better express what she really feels?
modality for conducting IPT (Wilfley et al., 2000). Lisa: I guess you could say, “Christine, I felt upset
Data from randomized trials suggest that both indi- the other day when you were sitting with  . . . ” I can’t
vidual and group milieus of IPT are equally effec- remember their names. “They talk about me behind
tive in the treatment of BN (Nevonen & Broberg, my back. When you were sitting with them, I felt like
2006) and BED (Wilfley et  al., 2002; Wilfley, you weren’t my friend.”
Wilson, & Agras, 2008). Following an individual
session to conduct a thorough interpersonal inven- The group setting allows patients to experiment
tory, the group is an ideal milieu to work on inter- with different ways of communication within the
personal skills with other patients struggling with safe confines of the group. Members can use the ses-
similar eating problems. It also offers the therapist sions to discuss problems they are having with their
an opportunity to observe and identify character- significant relationships and how these problems
istic interpersonal patterns with other individuals. relate to their eating patterns. This often allows for
Furthermore, when another group member recog- patients to recognize that they are not alone in their
nizes and verbally identifies a dysfunctional pat- difficulties, thereby helping to reduce feelings of iso-
tern of communication in a fellow patient, it can lation (Wilfley et al., 2000).
be powerful for the patient as well as the other
group members (Wilfley et al., 2000). The follow- Review of Outcome Studies and Relevant
ing vignette provides an illustration from a group of Empirical Literature
adolescents with binge eating patterns. Interpersonal Psychotherapy for
Bulimia Nervosa
Therapist: Sheila, you have done a great job of tell- Interpersonal psychotherapy has shown to
ing us what happened this week at school. It sounds be effective for the treatment of BN. Although
like it was pretty upsetting when you saw Christine, cognitive-​behavioral therapy (CBT) is currently
your best friend, sitting with people who she knows you the most extensively researched, best-​established
do not get along with. You recognized that it was not treatment for BN (Wilson & Fairburn, 2001),
the right time to talk to her, so you just walked away IPT is the only psychological treatment for BN
and sat with someone else. But it is still understandably that has demonstrated long-​term outcomes that are
upsetting to you. The rest of the group has suggested that comparable to those of CBT (Wilson & Shafran,
you tell Christine what is causing you to feel upset with 2005). Currently, all controlled studies of IPT for
her. Sheila, what might you say? BN have been compared to CBT for BN. In early
Sheila: I don’t know. I guess I would say, “Christine, studies, similar short-​ and long-​ term outcomes
the other day you were sitting with Amy and Joyce. You for binge eating reduction between CBT and IPT

300 Interpersonal Psychotherapy

were reported (Fairburn et  al., 1995; Fairburn However, given the small sample size, lack of ran-
et al., 1993). In a subsequent multisite study com- domization, and short follow-​up duration, future
paring CBT and IPT for BN, patients receiving research should experimentally examine brief IPT
CBT demonstrated higher rates of abstinence for BN to determine whether the abbreviated inter-
from binge eating and lower rates of purging in vention is as effective as conventional IPT (Arcelus
the shorter-​ term, post-​treatment (Agras, Walsh, et al., 2012).
Fairburn, Wilson, & Kraemer, 2000). By 8-​ and An emerging literature has provided some
12-​month follow-​up, however, patients in CBT insight into predictors of success with IPT for the
demonstrated maintenance or slight relapse while treatment of eating disorders. In the multicenter
IPT participants experienced slight improvement trial conducted by Agras and colleagues (2000),
such that rates of these behaviors were equivalent a follow-​ up analysis found that while patients
in both groups. The more impressive, immediate responded with higher abstinence rates when
effect of CBT compared to IPT may be explained randomized to CBT as opposed to IPT, African
in part by a relative lack of focus on eating dis- American participants showed greater reductions in
order symptomatology in the research version of binge episodes when treated with IPT compared to
individual IPT for BN that was used in this study CBT (Chui, Safer, Bryson, Agras, & Wilson, 2007).
(Tanofsky-​Kraff & Wifley, 2009; Wilfley et  al., Although further investigation is clearly necessary,
2003). Despite the relatively slower response rates, it is possible that IPT may be especially appropri-
IPT patients rated their treatment as more suitable ate for African American women with BN, which
and expected greater success than did CBT patients. speaks to the need for further study of IPT with
Therefore, a potential strength of IPT may be that different racial and ethnic groups. Researchers from
many patients with BN perceive the interpersonal this same study also examined the impact of thera-
focus of IPT as especially relevant to their eating peutic alliance on patient expectation of improve-
disorder and to their treatment needs, perhaps more ment (Constantino, Arnow, Blasey, & Agras, 2005).
so than a cognitive-​behavioral focus on distortions Expectation of improvement was positively associ-
related to weight and shape (Tanofsky-​ Kraff & ated with outcome for both CBT and IPT, empha-
Wifley, 2009; Wilfley et al., 2003). Currently, IPT sizing the important role of patient expectations in
is considered an alternative to CBT for the treat- both treatments. Lastly, in a study of postremission
ment of BN (Wilson & Shafran, 2005). Although predictors of relapse in women with BN, the find-
it has been recommended that therapists inform ing that worse psychosocial functioning was associ-
patients of the slower response time for improve- ated with a greater risk for relapse may support the
ments compared to CBT (Wilson, 2005), it is our rationale for IPT (Keel, Dorer, Franko, Jackson, &
contention that a lack of integration of BN symp- Herzog, 2005). Indeed, the authors suggested that
toms with the interpersonal focus is likely respon- their findings may partly help to explain the long-​
sible for the delayed response to IPT in the Oxford term effectiveness of IPT for BN.
trial (Fairburn et  al., 1995; Fairburn et  al., 1993)
and the less robust results in the multisite study Interpersonal Psychotherapy for Binge
(Agras et  al., 2000). Therefore, future research Eating Disorder
linking symptoms to interpersonal functioning is Based on the initial success of IPT in BN
required. (Fairburn et  al., 1991), IPT for BED was devel-
Given the increasing need for short-​term, effec- oped and tested in the early 1990s. Wilfley and col-
tive therapies by clinicians with limited resources, a leagues first adapted IPT to a group format for adult
brief, 10-​session version of IPT was pilot tested with patients with BED (Wilfley et  al., 1993; Wilfley
adult patients with BN (Arcelus, Whight, Brewin, et  al., 2000). During their work, they found that
& McGrain, 2012). When compared to conven- a number of patients presented with chronically
tional IPT (16–​20 sessions) and a wait-​list control, unfulfilling relationships that were well suited to be
the abbreviated 10-​session IPT was as effective in addressed in the group format. Therefore, new strat-
reducing eating disorder symptomatology as the egies were adapted to specifically address such inter-
traditional IPT and more effective than the wait-​ personal deficits. For example, in the current format
list control post intervention. These results provide of group IPT for BED, group members with inter-
very preliminary support that efficacy of IPT for personal deficits are strongly encouraged to use the
BN can potentially be reached more rapidly than group as an interpersonal “laboratory”; therapists
the conventionally recommended 16–​20 sessions. can observe, firsthand, patients interacting with one

Burke, Karam, Tanofsky-Kraff, Wilfley 301

another, and patients can practice improved ways those individuals with greater interpersonal prob-
of communicating within the group. As described lems were also those who had more Axis I and Axis
previously, this social milieu is designed to decrease II psychiatric disorders and lower self-​esteem than
social isolation, support the formation of new social those with less severe problems. These individuals
relationships, and serve as a model for initiating and are likely in need of augmented or extended treat-
sustaining social relationships outside of the thera- ment. Supporting this assertion, in IPT adapted
peutic context (Wilfley et al., 1998). Additionally, for individuals with borderline personality dis-
self-​stigmatization is common among patients with order, many of whom presented with comorbid
BED, and this stigmatization contributes to the depression, Markowitz and colleagues suggest that
maintenance of the disorder. By its very nature, extending IPT effectively improves the disorder
group therapy offers a radically altered social envi- (Markowitz et al., 2006). Further, in a follow-​up
ronment for these individuals, who typically main- analysis of the Wilfley et al. (2002) trial approxi-
tain shameful eating behaviors hidden from close mately 4 years after treatment cessation, individuals
others in their social network. By participating in in both treatments maintained reductions in binge
a group with others suffering from the same types eating and disordered eating cognitions (Hilbert
of psychiatric and physical issues, individuals with et al., 2012). However, between the 1-​year and 4-​
BED are offered a unique opportunity to feel both year follow-​up periods, those randomized to IPT
understood and accepted in IPT. maintained or improved eating disorder symp-
For the treatment of BED among adults, IPT has toms compared to the CBT group, whose symp-
been demonstrated to be effective in randomized-​ toms worsened. Overall, however, 52% of those in
controlled studies. Cognitive-​ behavioral therapy CBT and 76% of those in IPT reached recovery at
for BED has also been shown to have specific and this long-​term follow-​up, though groups did not
robust treatment effects (Devlin et al., 2005; Grilo, significantly differ at any time point. Eighty per-
Masheb, & Wilson, 2005; Kenardy, Mensch, cent of patients in both CBT and IPT remitted
Bowen, Green, & Walton, 2002; Nauta, Hospers, to subclinical rates of binge eating behavior, and
Kok, & Jansen, 2000; Ricca et  al., 2001; Telch, 58% saw improvements in related eating disorder
Agras, Rossiter, Wilfley, & Kenardy, 1990; Wilfley psychopathology. These data suggest evidence for
et al., 1993). In two randomized trials comparing good long-​term maintenance of change for BED
IPT with CBT, IPT had similar effects to CBT in patients treated with IPT.
the treatment and management of BED. The first Results from a multisite trial that compared
study, comparing group CBT and IPT, revealed individual IPT to behavioral weight loss treatment
that both treatments were more effective than a or CBT guided self-​help for the treatment of BED
wait-​list control group at reducing binge eating points to the importance of making a clear connec-
and had equivalent, significant reductions in binge tion between interpersonal problems and binge eat-
eating in both the short-​ and long-​term (Wilfley ing symptoms in the delivery of IPT (Wilson et al.,
et al., 1993). In a second substantially larger sam- 2010). Similar to Wilfley et al.’s 2002 trial (Wilfley
ple size, both CBT and IPT demonstrated equiva- et  al., 2002), in this multisite study the clinicians
lent 4-​ month and 1-​ year follow-​up efficacy in linked interpersonal functioning to disordered
reducing binge eating and associated specific and eating symptoms throughout the course of IPT.
general psychopathology, with approximately 60% Findings from this study revealed that IPT was most
of the patients remaining abstinent from binge acceptable to patients; the dropout rate was signifi-
eating at 1-​year follow-​up (Wilfley et al., 2002). In cantly lower in IPT compared with the other two
contrast to the literature on IPT for BN, the time interventions (Wilfley et  al., 2008). Interpersonal
course of almost all outcomes with IPT was iden- psychotherapy and CBT guided self-​help were sig-
tical to that of CBT and all participants in both nificantly more effective than behavioral weight loss
groups significantly improved from baseline. In a in eliminating binge eating after 2  years (Wilson
follow-​up analysis of treatment predictors for the et  al., 2010). Furthermore, compared to the other
2002 study, patients with a greater extent of inter- two programs, IPT produced greater binge epi-
personal problems at baseline and midtreatment sode reductions for patients with low self-​esteem
showed poorer treatment response to both treat- and greater disordered eating behaviors and cogni-
ments (Hilbert et al., 2007). An important caveat tions, while CBT guided self-​ help was generally
of this finding, however, is that not surprisingly, effective only for those with low eating disorder

302 Interpersonal Psychotherapy

psychopathology. It is notable that in this trial, as is associated with high retention across different
with the Wilfley et  al. study (Hilbert et  al., 2007; patient profiles (e.g., high negative affect, minority
Wilfley et  al., 2002), individuals with more psy- groups), and demonstrated superior outcomes to
chopathology showed greater improvements in IPT behavioral weight loss overall, and to CBT guided
than CBT guided self-​help. This is in concert with self-​help among a subset of patients with high disor-
Hilbert and colleagues’ follow-​up data suggesting dered eating psychopathology and low-​self-​esteem.
that greater disordered eating serves as a modera- Therapists and patients should consider these alter-
tor in predicting poorer outcome in CBT (Hilbert natives when deciding the best approach to treating
et al., 2007). their disorder. Finally, behavioral weight loss should
A trial comparing group psychodynamic inter- not be considered as a first choice when treating
personal psychotherapy (an attachment-​based ther- individuals with BED.
apy similar to IPT) and group CBT for treatment In summary, the literature suggests that IPT rep-
of BED were comparable in reducing binge eating, resents an efficacious treatment alternative to CBT
negative mood, total interpersonal problems (Tasca, for BED. If delivering IPT for BED in a group for-
Mikail, & Hewitt, 2005), and most interpersonal mat, as with all group therapies, developing mem-
problem areas (Tasca, Balfour, Presniak, & Bissada, ber cohesion is paramount to the achievement of
2012). Overall, group psychodynamic interpersonal treatment success.
psychotherapy was similar, but slightly better, than
CBT for those with more cold/​distant personality Interpersonal Psychotherapy for Anorexia
types and attachment avoidance. Nervosa
The Tasca trial was limited to primarily Caucasian In general, there are very few effective treatments
participants (Tasca et al., 2012); however, there are for AN (Wilson, Grilo, & Vitousek, 2007). Although
interesting findings with a more racially diverse par- behavioral family therapy is perhaps considered the
ticipant base. In general, compared to Caucasian treatment of choice for adolescents with early onset
participants, individuals of other ethnic minorities of the disorder, these data are not especially infor-
demonstrated less retention in the multisite study mative when making recommendations for adults.
by Wilfley and colleagues (Wilfley et  al., 2008). With regard to IPT, there is a relative lack of research
Although there was no treatment by ethnicity effects examining its utility for AN. Indeed, there have been
in this regard, there was very low attrition for minor- no controlled studies yet that have demonstrated the
ity participants in IPT and very high dropout rates by efficacy of IPT for AN. To date, only one group has
minorities in CBT guided self-​help. The small sam- tested IPT for AN (McIntosh et al., 2005). Fifty-​six
ple size of minority participants across sites precludes women with AN were randomized to IPT, CBT, or a
definitive conclusions. Nevertheless, this pattern is control comparison (nonspecific, supportive clinical
in concert with the finding that IPT was particularly management). In contrast to the impressive effects
helpful for African American participants in the pre- of IPT for both BN and BED, this study found
viously described multisite study for individuals with that IPT was associated with either modest or no
BN (Chui et al., 2007). It is possible that the person- improvement in AN symptoms compared to non-
alized nature of IPT (e.g., problem areas and goals are specific, supportive clinical management (McIntosh
developed based on each individual’s social environ- et al., 2005). Of the three therapies, nonspecific,
ment) is modifiable to, and thus particularly accept- supportive clinical management was the most effec-
able to, persons of various cultures and backgrounds. tive approach. Importantly, the authors posited that
A number of recommendations may be drawn their findings may be a result of the relative lack of
from the research presented. It is possible, from focus on eating disorder symptoms in their adap-
a cost-​ effectiveness viewpoint, that CBT guided tation of IPT (McIntosh, Bulik, McKenzie, Luty,
self-​help could be considered the first-​line treat- & Jordan, 2000) and suggest that future studies
ment for the majority of individuals with BED, implementing IPT for AN involve consistent con-
and that IPT is recommended for patients with nections between the interpersonal problem areas
low self-​esteem and high eating disorder psycho- and the core symptoms of the disorder (McIntosh et
pathology. Alternatively, IPT may be considered al., 2005). Particularly given the ego syntonic nature
a first-​line treatment for BED. This recommenda- of AN, the lack of focus on eating disorder symp-
tion is based on a number of factors: IPT has been toms may have blunted IPT’s impact and avoided
shown to be effective across multiple research sites, the essential work of the therapy (McIntosh et al.,

Burke, Karam, Tanofsky-Kraff, Wilfley 303

2005). The short timeframe for the IPT work, a rela- treatment modalities tend to learn IPT quickly
tive lack of symptom-​focus, and the brief length of and are often able to implement IPT with a high
follow-​up may have also contributed to the study degree of integrity despite minimal IPT-​specific
outcome. Indeed, in a long-​term follow-​up of this training (Birchall, 1999). Furthermore, some
study (on average 6.7 years after treatment), sup- therapists may consider IPT to be more acceptable
portive clinical management was no longer superior than CBT. Although not specific solely to IPT, a
to IPT or CBT. Rather, all three groups were similar naturalistic study of psychotherapy outcome in
in eating disorder symptomology, anthropometrics, which 145 clinicians provided information about
and general psychopathology (Carter et al., 2011). their eating disorder patients found that com-
Though initial post-​treatment results suggested that pared to CBT, psychodynamic approaches that
those randomized to IPT were least likely to have a included IPT produced better global outcomes
positive outcome (i.e., the best global outcome rat- (Thompson-​Brenner & Westen, 2005). Although
ing) compared with the other treatments, by long-​ there are limited data exploring the influence of
term follow-​ up, those in IPT evidenced the best therapist comfort on treatment outcome, it is pos-
global outcomes. Findings were reversed for support- sible that some clinicians are more comfortable
ive clinical management; post-​treatment, patients in administering treatments other than CBT, and this
this group had the best global outcome, but faired is reflected in the outcome of their work (Tanofsky-​
the worse at long-​term follow-​up. Those in CBT had Kraff & Wifley, 2009).
a more stable trajectory across time. These findings To date, there are more data in support
suggest that IPT may be of benefit to individuals with of the efficacy of CBT for eating disorders.
AN and, similar to other eating pathologies, some A  recent meta-​analysis comparing IPT to CBT
individuals continue to improve even after the course for eating disorders reveals a small, but signifi-
of treatment. Given the importance of interpersonal cant effect in favor of CBT in the short-​term,
functioning in etiological theories of AN (McIntosh though the number of available comparison
et al., 2000), continued exploration of IPT’s utility studies was small and the long-​ t erm effects
in treatment of the disorder is clearly warranted. In unclear (Cuijpers, Donker, Weissman, Ravitz,
particular, investigation of IPT for AN that includes & Cristea, 2016). To the latter point, although
a focus on eating disorder symptoms as they relate CBT has been shown to produce more rapid
to interpersonal problems is needed. It may be that effects for BN, IPT produces equivalent outcomes
for AN, IPT is optimally delivered in the context of over the long term for adults with this disorder.
other adjunctive treatments (e.g., pharmacological, Interpersonal psychotherapy for BED appears to
nutritional), rather than as a “stand alone” treatment. be equally as effective as CBT. Based on the evolv-
Staging of treatment may also be important; for ing literature, IPT may be well suited for patients
AN, IPT may be more suitable for the maintenance presenting with or without exacerbated difficulties
and relapse prevention stages of treatment than for in social functioning. Although greater problems
the weight regain phase (Jacobs, Welch, & Wilfley, were associated with poorer outcomes for both
2004). CBT and IPT in the Hilbert et al. (2007) study,
the moderator effect that patients presenting with
Choosing Treatment Modality greater psychopathology seem to respond well to
When determining the treatment approach IPT in the more recent multisite study (Wilfley
for patients with eating disorders, the clinician et al., 2008), suggest that IPT (or another special-
and patient should together evaluate the advan- ized treatment such as CBT) may be well suited
tages and disadvantages of using IPT, CBT, or for individuals with a broad range of disordered
another therapeutic approach, such as pharma- eating and general psychopathology. Moreover,
cologic treatment. In making this decision, it is IPT may be enhanced for individuals with exac-
crucial for therapists to explore their own comfort erbated psychological problems (Markowitz
level in terms of their expertise, theoretic knowl- et al., 2006). It is also possible that IPT may be
edge, and propensity toward administering an especially fitting for some minority groups, such
interpersonally focused treatment (Wilfley et  al., as African Americans. Finally, it is possible that
2000). IPT, like CBT, is a specialty treatment and some patients may express discomfort or difficul-
should be administered only by trained practi- ties with elements of CBT (e.g., keeping food dia-
tioners. However, it has been argued that experi- ries); IPT should be considered for these patients
enced therapists who have been trained in other as well (Tanofsky-​K raff & Wifley, 2009).

304 Interpersonal Psychotherapy

Interpersonal Psychotherapy for the rated themselves higher on the scale (Lemeshow
Prevention of Excessive Weight Gain et  al., 2008). Furthermore, overweight teens are
Interpersonal psychotherapy has been developed more likely to experience negative feelings about
for the prevention of excessive weight gain in ado- themselves, particularly regarding their body shape
lescents who report loss of control (LOC) eating and weight, compared with normal weight adoles-
patterns. Pediatric LOC refers to the sense that one cents (Fallon et al., 2005; Schwimmer, Burwinkle,
cannot control what or how much one is eating, & Varni, 2003; Striegel-​ Moore, Silberstein, &
regardless of whether the reported amount of food Rodin, 1986), perhaps because of their elevated
consumed is unambiguously large (Tanofsky-​Kraff, rates of appearance-​ related teasing, rejection,
2008). Loss of control is associated with excess and social isolation (Hayden-​ Wade et  al., 2005;
body weight and health complications (Radin Neumark-​Sztainer et al., 2002; Puhl, 2011; Strauss
et  al., 2015). Common among youth, LOC eat- & Pollack, 2003). The social isolation that over-
ing predicts excessive weight and fat gain over time weight teens report may be directly targeted by IPT.
(Sonneville et al., 2013; Tanofsky-​Kraff et al., 2006; Finally, IPT is posited to increase social support,
Tanofsky-​Kraff et  al., 2009), as well as worsening which has been demonstrated to improve weight
metabolic problems beyond the contribution of maintenance in overweight adults (Wing & Jeffery,
body weight (Tanofsky-​Kraff et al., 2012). Loss of 1999) and children (Wilfley et al., 2007). Indeed,
control eating is also related to, and predictive of, data suggest that low social problems predict bet-
increased psychological problems and the develop- ter response to weight loss treatment in children
ment of subsequent clinical eating pathology such (Wilfley et al., 2007).
as BED (Goldschmidt et al., 2015; Sonneville et al., Interpersonal psychotherapy for the prevention
2013; Tanofsky-​Kraff et al., 2011). The manual for of excessive weight gain for adolescents at high-​risk
IPT for the prevention of excess weight gain was for adult obesity, delivered in a group format, main-
adapted directly from IPT for the prevention of tains the key components of traditional IPT: (1) a
depression in adolescents (IPT Adolescent Skills focus on interpersonal problem areas that are related
Training, IPT-​ AST) (Young, Mufson, & Davies, to the target behavior (for example, LOC eating in
2006; Young, Mufson, & Schueler, 2016) and group the present adaptation); (2) the use of the interper-
IPT for BED (Wilfley et  al., 2000), and evolved sonal inventory at the outset of treatment to iden-
from the outcome data of psychotherapy trials for tify interpersonal problems that are contributing
the treatment of BED. An unexpected finding of to the targeted behavior; and (3)  the three-​staged
IPT and most psychological treatments for BED structure of the intervention (initial, middle, and
has been that individuals with BED who cease to termination). The primary activities of IPT for the
binge eat tend to maintain their body weight dur- prevention of excess weight gain are to provide psy-
ing and/​or following treatment (Agras et al., 1995; choeducation about risk factors for excessive weight
Agras, Telch, Arnow, Eldredge, & Marnell, 1997; gain and to teach general skill-​building to improve
Devlin et  al., 2005; Wilfley et  al., 1993; Wilfley interpersonal problems. For prevention of excess
et al., 2002; Wilson et al., 2010). Therefore, it has weight gain, IPT differs from other adaptations in
been hypothesized that treatment of LOC eating that it was developed to specifically address the par-
among youth may reduce excessive weight gain and ticular needs of adolescents at high risk for adult
prevent full-​syndrome eating disorders (Tanofsky-​ obesity due to their current BMI percentile and
Kraff et al., 2007). report of LOC eating behaviors.
A number of factors suggest that IPT is partic- Based on IPT-​AST, this adaptation is presented
ularly appropriate for the prevention of obesity in to teenagers as “Teen Talk” in order to be nonstig-
high-​risk adolescents with binge or LOC eating matizing. As designed by Young (Young & Mufson,
patterns. Specifically, youth frequently use peer 2003; Young et al., 2006; Young et al., 2016), this
relationships as a crucial measure of self-​evaluation preventive adaptation of IPT focuses on psychoedu-
(Mufson et al., 2004). The importance of perceived cation, communication analysis, and role-​playing.
social interactions and social standing on body Specific interpersonal communications skills are
weight gain over time was revealed in a prospective taught, including “Strike while the iron is cold,”
cohort study (Lemeshow et  al., 2008). Adolescent “Use ‘I’ statements,” “Be specific” (when talking
girls who rated themselves lower on a subjective about a problem), “Don’t give up,” and “Put your-
social-​standing scale were 69% more likely to gain self in their shoes” (Young & Mufson, 2003; Young
more weight over time compared with girls who et  al., 2006; Young et  al., 2016). To encourage

Burke, Karam, Tanofsky-Kraff, Wilfley 305

identification and acknowledgment of affect, the by Tanofsky-​Kraff and colleagues (Tanofsky-​Kraff,
skill “Use ‘I’ statements” has been changed to “Use Shomaker, Young, & Wilfley, 2016). For sample
‘I feel’ statements.” In addition, the skill “Put your- cases of adults with eating disorders across diagnos-
self in their shoes” has been amended to “Put your- tic categories, we recommend referring to a book by
self in their shoes (and communicate it to them)” Denise Wilfley and colleagues (Wilfley et al., 2000)
to facilitate development of empathic communi- as well as chapters on IPT for eating disorders (e.g.,
cation. Two additional skills have been added to Jacobs et al., 2004; Wilfley et al., 2003).
further develop effective communication abilities.
First, “What you don’t say speaks volumes,” has Presentation:  Kay is a 14-​ year-​old African
been added to teach adolescents how their body American female with a BMI at the 85th percentile
language has the ability to impact communication for her age and sex (Ogden, Carroll, Kit, & Flegal,
regardless of their words. Second, “Have a few solu- 2014). At intake, she reported engaging in an aver-
tions in mind and be willing to compromise” has age of 5–​6 episodes of LOC eating per month over
been added to facilitate assertive communication the 3 months prior to intake. She specifically recalled
and problem-​solving skills. During the interper- engaging in two such episodes in the past month, both
sonal inventory, a “closeness circle” (Mufson et al., when alone and feeling “bored.” She reported feeling
2004) is used to identify the close relationships of anger, distress, and regret following her LOC eating
the participant. Since this adaptation was designed episodes. However, she was only able to connect her epi-
for adolescents ranging from age 12 to 17, sessions sodes to feelings of boredom. Kay endorsed some distress
are geared toward the adolescents’ developmen- surrounding her shape and weight as well as feeling
tal level. For example, younger adolescents, who “fat” much of the time, but reported few attempts at
may be uncomfortable talking about themselves, dieting. She reported eating in response to a number
may respond better to hypothetical situations and of negative emotions, with a very strong desire to eat
games, whereas older teenagers may more read- when feeling down, sad, stressed out, worried, or bored.
ily discuss their own interpersonal issues from the Although she presented with few symptoms of depres-
outset. sion, Kay did experience some subclinical threshold
Based on IPT for BED, IPT for excess weight symptoms of anxiety. Although she reported feeling shy
gain prevention maintains focus throughout the when meeting new people, she also endorsed having
program on linking negative affect to LOC eating, close friendships with peers in whom she could confide.
overeating, times when individuals eat in response During her pretreatment meeting, Kay reported
to cues other than hunger, and overconcern about a number of family stressors. In the months prior to
shape and weight. Further, a timeline of personal intake, she had reluctantly returned to her mother’s
eating and weight-​related problems and life events home in the Midwest to attend the local middle school,
is discussed individually with participants prior after having attended boarding school for three years
to the group program. Similar to both IPT-​AST on the West Coast. She reported generally poor relation-
and IPT for BED, this adaptation is delivered in ships with her parents, who had divorced when she was
a group format. Twelve weeks in duration, IPT a baby. Kay indicated that because she and her mother
for preventing obesity is longer than IPT-​AST (8 were “a lot alike” and both very stubborn, they argued
sessions), but shorter than group IPT for BED frequently. Most often, their arguments concerned
(typically 16 to 20 sessions). Similar to IPT-​AST, Kay’s dislike of her stepfather, whom she referred to
group size is smaller than in IPT-​BED (5 vs. 9 as obnoxious and racist. She reported that her mother
members), enabling therapists to keep adolescents no longer “thinks for herself,” but rather just agrees
engaged. As with group IPT for BED (Wilfley with her stepfather. Fights with her mother typically
et al., 2000), participants meet individually with involved Kay saying something hurtful in the “heat of
the therapist(s) for a brief midtreatment meeting the moment” and/​or walking away without resolution.
to discuss progress made on proposed therapeutic Kay reported being left with emotions of both rage and
goals, areas that are particularly challenging, and guilt. During these times, she would often overeat and
plans for continued work through the second half experience a lack of control over how much she was eat-
of the group. ing. Following the eating episode, she reported that her
The following case example of “Kay” briefly negative feelings eventually “went away.” Despite liv-
illustrates the presentation and treatment of an ing in a nearby city, her biological father had little con-
adolescent group participant. For a more detailed tact with her following her parents’ divorce. She had
case presentation, we recommend a recent article spoken with him 1–​2 times per month for many years

306 Interpersonal Psychotherapy

and up until intake. Kay reported feeling abandoned mother and her father. The therapists encouraged
by her father and attributed his lack of availability to Kay to initiate a discussion with her parents during
his own social anxiety. Nonetheless, she wished for a which she would be specific about her frustrations
closer relationship, but never discussed her desire with while also trying to keep in mind an understanding
him for fear that he would not be receptive. of their perspectives. Initially, Kay was reluctant to
Problem area: Kay’s IPT problem area was concep- follow up on this work. She was therefore encouraged
tualized as a role dispute. Moving home from board- to examine her pattern of anger and then avoidance,
ing school—​halfway across the country—​decreased her particularly with her mother. During the seventh ses-
independence and exacerbated the typical changes ado- sion, Kay became frustrated with the therapists’ per-
lescents experience during developmentally appropriate sistence and grew sullen and angered. She returned
individuation from their parents. Kay struggled with the next session to report that she was angry, confused,
being unable to communicate successfully with either and disappointed to have learned about some of her
parent—​or her stepfather—​regarding her opinions and mother’s past behaviors of which Kay did not approve.
needs. In response to such disputes, Kay would experi- The therapists encouraged Kay to take “a leap of faith”
ence negative affect and eat to cope with her emotions. and try talking with her mother. After in-​session prac-
Goals: The therapists and Kay generated and agreed ticing, Kay approached her mother and talked about
on the following therapy goals for the 12-​week inter- her feelings in a calm manner. She was very pleased
vention. The first goal was that Kay would work on with her mother’s receptiveness to the discussion. By
gaining perspective to feel less frustrated with her par- the end of the middle phase, Kay was also spending
ents and work on remaining calm in the moment. some time with her father. Moreover, she had opened
Second, she would aim to express her feelings of being up a dialogue with him regarding some of her feelings
let down and hurt by her father. If possible, she would of disappointment with her. Much to her surprise her
consider discussing these feelings with him. Kay’s work father was more receptive to her self-​expression than
in treatment would involve clarifying her role within she had expected. Throughout this phase, Kay was fre-
the family, vis-​à-​vis her mother and stepfather, and quently queried about her eating patterns. Initially,
learning how to negotiate and express herself with them she noticed that that the frequency of her LOC eating
in a more functional manner. episodes had decreased. Then, she became more cogni-
zant of the times she would binge eat. For example,
IPT intervention after finding out that an old friend was ill, she became
Initial phase: Kay was very engaged during the ini- very upset and was able to link her feelings to LOC
tial phase. She actively participated in the role-​play eating.
exercises and was open about her frustration with mov- Termination phase: Kay reported that not only was
ing back home to live with her mother and stepfather. she sharing her feelings more often, but that she was
She also shared how she had a tendency to vacillate overeating less frequently. She was supportive of other
between speaking her mind and avoiding arguments group members in taking their own “leaps of faith” by
by walking away, particularly with her mother. She describing how overwhelmed she felt before speaking
reported overeating and feeling unable to stop, most with her parents, but how much better she felt after
often in response to avoiding arguments with her the conversations. By the final session, Kay was quite
mother and feeling angry. Specifically, she shared sad about the ending of the group. She reported that
that not only would her mother side with her step- she was going to miss the support of the therapists and
father during family arguments, but that her mother group members and realized that she still had a great
believed Kay was not intelligent. Her mother would deal of work to continue. The therapists focused on
often express her opinion about Kay’s intelligence with the work she had accomplished and assured her that
her stepfather and other friends. Kay was encouraged she had achieved the skills necessary to continue mak-
to role-​play a conversation with her mother in which ing improvements. Along with the therapists and the
she discussed her feelings. However, she remained skep- other members, Kay outlined the future work that she
tical that her mother would be responsive. While the would continue after the groups ended. Specifically, she
therapists encouraged Kay to practice such a conver- planned to continue dialogues with her mother and
sation in the group, they also recommended that she seek to develop a closer relationship with her father,
think about how to better tolerate her mother’s behav- with the recognition of his emotional limitations.
ior if, in fact, she was unresponsive. Following treatment, Kay reported no longer expe-
Middle phase:  Kay spent more time role-​playing riencing episodes of LOC while eating. Moreover, her
conversations that she might have with both her BMI percentile had decreased 5  percentage points

Burke, Karam, Tanofsky-Kraff, Wilfley 307

Table 15.4  IPT for the Prevention of Excess Weight Gain Sample Case Conceptualizations and Courses of Treatment
Example LOC eating Interpersonal Problem area Goal Initial Phase Middle Phase Termination Phase
Participant precipitant(s) functioning

Case #1 Sadness, Repeated heated Role dispute Gain perspective to Sharing feelings Discuss resistance Emphasis on improved
stress and arguments with decrease frustration of frustration to speaking with communication skills;
worry mother and remain calm when with mother; in mother; with group discussion of transferring
communicating with group role-​play of encouragement, began use of skills to other close
mother discussions with productive dialogues with interpersonal relationships;
mother mother gaining other outside supports

Case #2 Avoiding Does not express Interpersonal Become more Discuss discomfort Practice sharing Emphasis on improved
conflict and negative feelings deficits comfortable with surrounding feelings via role-​ communication skills; focus on
negative or discomfort with conflict and work on interactions playing; encouraged to future generalizing of skills to
affect conflict in multiple expressing feelings involving conflict communicate feelings several situations
relationships with less tense relationship

Case #3 Boredom and Expresses emotions/​ Role dispute Use more constructive Communication Continued role-​playing Emphasis on improved
frustration needs to family communication to analysis and in specific situations and communication skills; focus
(especially parents) express self group role-​play of trying out discussions on future sharing of deeper
in nonproductive poor interactions with siblings personal conflicts with parents
manner
Tanofsky-​Kraff, M., Wilfley, D. E., Young, J. F., Mufson, L., Yanovski, S. Z., Glasofer, D. R., & Salaita, C. G. (2007). Preventing Excessive Weight Gain in Adolescents: Interpersonal Psychotherapy for Binge Eating.
Obesity, 15(6), 1345–​1355. doi:10.1038/​oby.2007.162. Reprinted with permission of John Wiley & Sons.
to the 80th percentile for her age and sex (Ogden, social adjustment problems moderated these effects
Carroll, Kit, & Flegal, 2014). The courses of treat- (Tanofsky-​Kraff et  al., 2017). For BMIz, girls in
ment for this individual, along with two other case IPT and health education with lower self-​reported
examples, are illustrated in Table 15.4 (Tanofsky-​ anxiety at baseline and lower social adjustment
Kraff et al., 2007). problems at baseline displayed similar decreases in
In a pilot study testing IPT for preventing excess BMIz over 3  years. In contrast, girls in IPT with
weight gain compared with a standard health educa- higher self-​reported anxiety at baseline and higher
tion program (Bravender, 2005), IPT was shown to social adjustment problems at baseline evidenced a
be both feasible and acceptable to adolescent girls significantly steeper decline in BMIz over the course
with and without LOC eating who were at risk for of 3 years compared with girls in health education.
excessive weight gain (Tanofsky-​Kraff et al., 2010). Self-​reported social adjustment problems did not
Further, at 1-​year follow-​up, girls in IPT experi- moderate the 3-​year increases in adiposity by group;
enced less than expected increases in their BMI for however, self-​ reported anxiety did. Girls in IPT
their age and BMI percentile compared with girls with higher self-​reported anxiety at baseline did not
in health education (Tanofsky-​Kraff et  al., 2010). gain in adiposity at 3 years, which is in stark con-
In addition, among girls who reported LOC eating trast to girls in health education with higher anxiety
at baseline, those in IPT evidenced a significantly and girls in both groups with lower anxiety, who
greater decrease in LOC eating episodes compared all experienced significant adiposity gains over time
with those in health education (Tanofsky-​ Kraff (Tanofsky-​Kraff et al., 2017). Parent-​reported data
et al., 2010). An adequately powered randomized-​ generally mirrored these findings. Baseline depres-
controlled trial with adolescent girls who endorsed sion symptoms and baseline number of LOC eating
at least one LOC eating episode at baseline provided episodes were not significant moderators of BMIz
additional evidence on the effectiveness of IPT for or adiposity over time (Tanofsky-​Kraff et al., 2017).
the prevention of excess weight gain and insights on Overall, IPT was not significantly better than
concurrent psychological benefits (Tanofsky-​Kraff health education for BMIz or adiposity at 1-​or
et al., 2017; Tanofsky-​Kraff et al., 2014). 3-​year follow-​ up (Tanofsky-​ Kraff et  al., 2017;
Similar to the pilot study, girls in the randomized-​ Tanofsky-​Kraff et  al., 2014). As both groups were
controlled trial were assigned to either IPT or health active interventions, it is possible that, on the whole,
education groups. As expected based on their stage social support provided by both groups conferred
of development, all girls gained in BMI at 1-​year adequate interpersonal benefits to affect weight
follow-​up (Tanofsky-​Kraff et  al., 2014). However, outcomes (Tanofsky-​Kraff et  al., 2017; Tanofsky-​
girls in both groups decreased in adiposity and the Kraff et  al., 2014). However, at 3-​year follow-​up,
age-​adjusted BMI metrics of BMIz and BMI per- girls with higher anxiety or social adjustment prob-
centile with no significant between group differ- lems at baseline evidenced greater improvements
ences. Similarly, regardless of group assignment, in BMIz if they were randomized to IPT, and only
depression and anxiety symptoms significantly girls with higher anxiety randomized to IPT did not
improved at the 1-​year follow-​up (Tanofsky-​Kraff gain fat over time. Therefore, consistent with IPT
et al., 2014). Psychosocial functioning and average theory, IPT may be particularly promising for girls
number of LOC eating episodes were similar for girls in preventing excess weight gain over time for those
in both IPT and health education; however, girls in with high anxiety and social problems (Tanofsky-​
IPT had significantly fewer binge eating episodes at Kraff et al., 2017). Both IPT and health education
1-​year follow-​up compared with girls in health edu- groups benefitted from improvements in mood and
cation (Tanofsky-​Kraff et  al., 2014). Though only decreases in LOC eating episodes, though IPT for
one girl in IPT developed a probable eating disorder the prevention of excess weight gain may be par-
at 1-​year follow-​up compared with three in health ticularly beneficial for decreases in classic binge epi-
education, odds of developing a probable eating dis- sodes (Tanofsky-​Kraff et  al., 2017; Tanofsky-​Kraff
order did not differ significantly by group at 1-​year et al., 2014). In addition, IPT may be particularly
follow-​up (Tanofsky-​Kraff et al., 2014). helpful for girls from ethnic minority backgrounds,
Follow-​up at 3 years provided additional insights. as those who were in IPT had significantly greater
At 3 years, compared to baseline, girls decreased in reductions in LOC eating episodes compared with
BMIz and increased in adiposity with no significant those in health education (Tanofsky-​Kraff et  al.,
between group differences (Tanofsky-​Kraff et  al., 2014). This is consistent with the benefits IPT
2017). However, baseline anxiety and baseline seems to confer on racial and ethnic minorities

Burke, Karam, Tanofsky-Kraff, Wilfley 309

from eating pathology research (Chui et al., 2007; and excess weight gain in a military health system
Wilfley et al., 2008). clinic. In this pilot study, youth randomized to IPT
Indeed, qualitative investigations with African appeared to show improvements in mood symptoms
American youth, their caregivers, and community relative to those in health education. There were no
leaders provided support for the acceptability of between-​group differences in excess weight gain at
IPT for the prevention of excess weight gain as an 1-​year follow-​up. Overall, despite the challenges of
intervention for the prevention of excess weight gain conducting research trials with military families,
and provided guidance on culturally relevant adap- IPT was feasible within military treatment facilities
tations (Cassidy et al., 2016; Cassidy et al., 2013). and evidenced similar results as with civilian trials.
Such adaptations included integrating parents into An adequately powered randomized-​controlled trial
the intervention, adding a nutrition education com- is currently underway with adolescents from mili-
ponent, and reframing some communication skills tary families to investigate the effectiveness of this
to better reflect the communication styles favored preventative adaptation of IPT and will be the first
by the African American community (Cassidy to examine the effects of IPT in adolescent boys.
et  al., 2016; Cassidy et  al., 2013). For example,
some suggested that “Using “I” statements” may Future Directions
need changes in order to respect parental author- Several important areas require further study. An
ity and align with cultural standards of commu- important next step is to determine whether IPT
nication, yet core tenets of IPT would remain the for eating disorders can be translated from specialty
same (Cassidy et  al., 2016; Cassidy et  al., 2013). care centers to the primary care setting, schools, and
Considering the known obesity-​related health dis- other typically nonresearch clinical practice milieus.
parities in the African American population (Ogden In an effort to continually improve IPT and broaden
et al., 2014; Ogden, Carroll, Lawman, et al., 2016) its utility, we propose other research directions in
and the lack of long-​term effective interventions this section (Tanofsky-​Kraff & Wifley, 2009).
based on strictly behavioral factors, IPT may be par-
ticularly effective in reducing disordered eating and Enhancing Interpersonal Psychotherapy
improving weight outcomes for African American for Bulimia Nervosa and Binge Eating
youth (Cassidy et al., 2016; Cassidy et al., 2013). Disorder
Indeed, the design and structure of IPT is inher- As efforts to more frequently and consistently
ently adaptable for addressing concerns that may be link eating disorder symptoms to interpersonal func-
particularly salient to this population. A culturally tioning has evolved in the use of IPT for BED, clini-
adapted manualized intervention has been devel- cal researchers involved in developing IPT for BN
oped (Cassidy et al., 2016), and a pilot of the inter- should also consider stressing this link during the
vention is being developed. delivery of IPT so that it offers the utmost potency.
Interpersonal psychotherapy for preventing Since IPT does appear to have specific effects in BN
excess weight gain and eating disorders has also and good long-​term maintenance of change, it seems
been adapted for use in for adolescent dependents prudent to evaluate methods for improving its effi-
of military service members, a group with a simi- ciency and clinical effectiveness. For instance, it may
lar prevalence of obesity to civilian youth but with be that the slower and less potent effects observed
unique interpersonal stressors. Factors such as fre- in IPT as compared to CBT were due to the man-
quent moves and parental deployments put such ner in which IPT was implemented. Specifically, in
youth at unique risk for interpersonal conflicts order to minimize procedural overlap with CBT, the
and potentially maladaptive coping strategies such research application of IPT for BN has not included
as LOC eating. Indeed, military youth with LOC an ongoing focus on making links between symp-
eating may be especially vulnerable to disordered tomatology and interpersonal functioning, which
eating and poorer health outcomes (Schvey et  al., is in stark contrast to how IPT was developed and
2015). Yet, carrying out prevention research in mili- tested for depression. In future studies, the efficacy
tary communities is a challenge, in part due to fami- and efficiency of IPT may be enhanced by including
lies’ frequent relocation and the constraints of the a specific focus on the core symptoms of BN and
military health system. With military and civilian their connection with interpersonal issues through-
partners, Tanofsky-​Kraff and colleagues tested the out the course of treatment. Such refinements of the
feasibility of IPT compared to a health education content and delivery of IPT may further strengthen
group for reducing LOC eating, mood symptoms, its usefulness in the treatment of BN.

310 Interpersonal Psychotherapy

 In its current form, IPT already seamlessly incor- a point of intervention for the treatment of eat-
porates aspects of other therapeutic modalities. For ing and weight-​ related problems during middle
example, the collaborative, interpersonal formula- childhood (Tanofsky-​Kraff & Wifley, 2009), par-
tion of the eating disorder symptoms during the ticularly for children with greater eating pathology
interpersonal inventory is one of the ways in which (Goldschmidt et  al., 2014). A  pilot study testing
IPT may resemble the behavior therapies more than a family-​based adaptation of IPT for preventing
it does the supportive or psychodynamic therapies. excess weight gain is near completion.
Therefore, some aspects of CBT may enhance the
efficacy of IPT (Tanofsky-​Kraff & Wifley, 2009). For Developing Interpersonal Psychotherapy
example, IPT therapists might wish to encourage for the Prevention of Eating and Weight-​
self-​monitoring as a method for patients to become Related Problems
more aware of their negative affect surrounding eat- Given the increasingly high rates of obesity, it
ing disorder symptoms. Such an approach is already may be reasonably posited that the increases in dis-
being tested in other treatment modalities. Indeed, ordered eating will continue as well, considering that
Fairburn and colleagues have found the inclusion overweight is a significant risk factor for the devel-
of an interpersonal module useful when adminis- opment of eating pathology (Fairburn et al., 1998;
tering a recently modified version of CBT for eat- Fairburn et al., 1997). Therefore, the use of IPT to
ing disorders (Enhanced CBT for Eating Disorders; prevent obesity and full-​syndrome eating disorders
Fairburn, 2008). should be explored by targeting other behaviors that
promote both conditions (Tanofsky-​Kraff & Wifley,
Adolescent and Child/​Parent Adaptations 2009). Since not all overweight individuals report
Given the robust efficacy of IPT for adoles- binge or LOC eating, reducing emotional eating and
cents with depressive disorders, and the promise eating in the absence of hunger may also be suitable
of IPT for the prevention of excess weight gain, for IPT modalities. Studies suggest that LOC eating
future research should involve additional adoles- among youth is associated with eating in response
cent adaptations (Tanofsky-​Kraff & Wifley, 2009). to negative affect (Goossens, Braet, & Decaluwe,
Adolescence is a key developmental period for culti- 2006), including anger and frustration, depres-
vating social and interpersonal patterns, which may sion, and anxiety (Tanofsky-​Kraff et al., 2007), and
explain why adolescents appear to relate well to IPT. predictive of increases in depressive symptoms and
From its inception, Mufson and colleagues made disordered eating attitudes (Tanofsky-​Kraff et  al.,
important adolescent-​ relevant adaptations to the 2011). In studies of adolescents, emotional eating is
treatment (Mufson et al., 2004). For example, IPT significantly correlated with constructs of disturbed
for adolescent depression includes a parent com- eating (van Strien, 1996; Van Strien, Engels, Van
ponent and the assignment of a “limited sick role,” Leeuwe, & Snoek, 2005) and symptoms of depres-
since youth are required to attend school and reduc- sion and anxiety (Van Strien et al., 2005). Data also
ing their activities is likely to exacerbate their inter- suggest that emotional eating may be associated
personal difficulties. Given that this foundation has with overweight among youth (Braet & Van Strien,
been established, the use of IPT for adolescents with 1997) and overeating in cross-​sectional structural
BN and BED warrants investigation. models (Van Strien, Engels, Van Leeuwe, & Snoek,
Using IPT for younger children may also be a 2005). Considering that in controlled trials IPT for
promising approach. A pilot study of family-​based BED effectively reduces eating in response to nega-
IPT for the treatment of depressive symptoms in tive affect in adults (Wilfley et  al., 1993; Wilfley
9-​to 12-​year-​old children was found to be feasible et  al., 2002), and negative affect predicts onset of
and acceptable to families (Dietz, Mufson, Irvine, BED and other eating disorders (Stice et al., 2017),
& Brent, 2008), and a randomized controlled trial preventive adaptations targeting negative affect
provided evidence of the efficacy of family-​based (Stice et  al., 2017) and emotional eating require
interpersonal psychotherapy in the treatment of investigation.
depression for children 7–​12  years old (Dietz, Eating in the absence of hunger has been asso-
Weinberg, Brent, & Mufson, 2015). The moder- ciated with overweight (Lansigan, Emond, &
ating influence of social problems on weight loss Gilbert-​Diamond, 2015; Moens & Braet, 2007)
outcome in a family-​based program (Wilfley et al., and excessive weight gain over time (Lansigan et al.,
2007) suggests that targeting interpersonal func- 2015; Shunk & Birch, 2004), though not all have
tioning in the nuclear family milieu may serve as found support for its contribution to weight or fat

Burke, Karam, Tanofsky-Kraff, Wilfley 311

gain in adolescence (Kelly et  al., 2015). However, competency in IPT. Moreover, veterans treated
reported eating in the absence of hunger has been experienced significant reductions in depressive
shown to be associated with dietary restraint symptoms and reported improvements in quality
(Shomaker et  al., 2010), LOC eating, emotional of life (Stewart et al., 2014). The Veterans Health
eating, and elevations in general psychopathology Administration’s national training and implemen-
(Tanofsky-​Kraff et al., 2008). Of concern are data tation efforts provide support for the feasibility
indicating that eating in the absence of hunger is and effectiveness of the broad dissemination of
a stable trait throughout youth (Birch, Fisher, & IPT to clinical care settings.
Davison, 2003; Fisher & Birch, 2002). Promising Wilfley, Wilson, Agras, and colleagues are cur-
findings indicate that young children may be trained rently examining the dissemination and imple-
to better regulate food intake (Johnson, 2000), and mentation of IPT for eating disorders in college
a number of intervention studies targeting eating counseling centers across the United States in two
in the absence of hunger are currently underway. separate trials. In the first trial, Wilfley and col-
Interpersonal psychotherapy may serve as a natural leagues are examining the short-​ and long-​term
extension on this work; in particular, negative affect effectiveness of two methods of training therapists
associated with interpersonal problems might be in IPT (Wilfley et al., 2016; National Institute of
linked to eating in the absence of hunger. Then, rec- Mental Health, Washington University School of
ognition of internal physiological hunger cues may Medicine). The first training method includes a
be taught so that patients learn to differentiate true more traditional format, composed of an in-​person
hunger from when they are already sated. workshop and 12  months of expert consultation.
Finally, there has been a growing interest in and The second method involves training a counseling
awareness of the role that social and interpersonal center staff member in IPT who is then taught to
factors may play in behavioral health problems train other staff members within his/​her site (i.e.,
(Glass & McAtee, 2006). For obesity in particular, a train-​ the-​trainer approach). Patient outcomes
moving away from focusing solely on individual will then be assessed to examine implementation
behavioral changes (e.g., diet and exercise) and effects. The second trial is in development and
toward the greater social context has not been the seeks to create and evaluate the effectiveness of a
norm. Interpersonal psychotherapy may be particu- novel online training platform of IPT for eating
larly well suited for developing new approaches for disorders (Wilfley, 2015). If successful, training
the prevention of obesity and eating disorders on a via an online platform will allow accessibility and
broader social level (NIH, 2011; Tanofsky-​Kraff & flexibility, invaluable features considering practic-
Wifley, 2009). ing therapists’ limited time and resources and the
financial constraints of service settings. Other ben-
Disseminating and Implementing efits of online training include trainees’ abilities to
Interpersonal Psychotherapy reinforce learning with access to repeatedly review
Evidence-​based psychotherapies such as IPT and reference training material. An advantage is
exist; however, they are not routinely used in serv- that an online training platform can be updated
ice settings due to barriers in clinician training regularly, eliminating the barrier of incorporating
(Fairburn & Wilson, 2013). The field of men- and distributing new information. Future research
tal healthcare recognizes the important need to should continue to examine the dissemination and
reduce this gap in training that would likely lead implementation of IPT given the important impli-
to improvements in care (Ravitz et  al., 2014). cations these efforts have on training competent
Initial efforts have been made to disseminate and clinicians in the treatment of evidence–​based treat-
evaluate IPT in clinical settings. The Veterans ment of eating disorders. In addition, considering
Health Administration developed a national ini- IPT is effective in treating a range of mental health
tiative to promote the dissemination of evidence-​ problems (e.g., depression and anxiety; Cuijpers
based practices in their care settings, including et al., 2016), it is an efficient intervention to dis-
IPT for depression (Stewart et  al., 2014). The seminate (Kazdin, Fitzsimmons-​Craft, & Wilfley,
Veterans Affairs IPT training program involved 2016). The flexibility of IPT to address a range of
a 3-​day workshop and 6  months of weekly tel- psychopathology and its inherent adaptability to
ephone consultation with an IPT expert, and one’s cultural and social norms makes it well suited
results were promising with a high rate of train- for widespread dissemination and implementation
ing program completion (93%) and therapist efforts (Kazdin et al., 2016).

312 Interpersonal Psychotherapy

Conclusion area(s). The termination phase is devoted to
Interpersonal psychotherapy for eating disorders consolidating gains made during treatment and
is a focused, time-​ limited treatment that targets preparing for future work.
interpersonal problems associated with the onset
and/​ or maintenance of the eating disorder. The
Acknowledgments
interpersonal focus is highly relevant to individu-
NIDDK grant 1R01DK080906-​01A1 (MTK).
als with eating disorders, many of whom experience
USUHS grant R072IC (to MTK). NIMH grant
difficulties in interpersonal functioning. Depending
5R01MH064153-​ 06 (DEW). NIMH grant
on the individual’s primary problem area, specific
1K24MH070446 (DEW). Disclaimer:  The opin-
treatment strategies and goals are incorporated into
ions and assertions expressed herein are those of
the treatment plan. The primary problem area is
the authors and are not to be construed as reflect-
determined by conducting a thorough interpersonal
ing the views of USUHS or the US Department of
inventory, a unique aspect of IPT, and by devising
Defense.
an individualized interpersonal formulation for each
patient. Interpersonal psychotherapy has resulted in
Abbreviations
significant and well-​maintained improvements for
AN: Anorexia nervosa
the treatment of BN and BED. Data support the
BED: Binge eating disorder
utility of IPT for the prevention of excess weight gain
BN: Bulimia nervosa
in adolescent girls, particularly those with higher
CBT: Cognitive-​behavioral therapy
social problems and anxiety. Further investigation
IPT: Interpersonal psychotherapy
is required to determine whether IPT is suitable for
IPT-​AST: Interpersonal psychotherapy-​Adolescent
and effective in the treatment of AN. Adaptations
skills training
of IPT should be explored for adolescent popula-
LOC: Loss of control (over eating)
tions and the treatment of other eating-​and weight-​
related problems. Finally, an important next step is
References
to disseminate IPT into routine clinical care settings. Agras, W. S., Telch, C. F., Arnow, B., Eldredge, K., Detzer, M.
J., Henderson, J., & Marnell, M. (1995). Does interper-
sonal therapy help patients with binge eating disorder who
Glossary fail to respond to cognitive-​behavioral therapy? Journal of
Consulting and Clinical Psychology, 63, 356–​360.
Interpersonal inventory: Conducted at the Agras, W. S., Telch, C. F., Arnow, B., Eldredge, K., & Marnell,
initiation of treatment, the therapist conducts M. (1997). One-​year follow-​up of cognitive-​behavioral ther-
a thorough review of the patient’s interpersonal apy for obese individuals with binge eating disorder. Journal
history and links psychosocial functioning to the of Consulting and Clinical Psychology, 65, 343–​347.
eating disorder symptoms. Agras, W. S., Walsh, B. T., Fairburn, C. G., Wilson, G. T., &
Kraemer, H. C. (2000). A multicenter comparison of
Life chart: An outline of life events and the cognitive-​behavioral therapy and interpersonal psychother-
development and maintenance of eating disorder apy for bulimia nervosa. Archives of General Psychiatry, 57,
symptoms. 459–​466.
Ambwani, S., Roche, M. J., Minnick, A. M., & Pincus, A. L. (2015).
Problem areas: Focus of treatment is on Negative affect, interpersonal perception, and binge eating
the resolution of problems within four behavior: An experience sampling study. International Journal of
social domains that are associated with Eating Disorders, 48, 715–​726. doi:10.1002/​eat.22410
the onset and/​or maintenance of the Ansell, E. B., Grilo, C. M., & White, M. A. (2012). Examining
the interpersonal model of binge eating and loss of con-
eating disorder: (1) interpersonal deficits,
trol over eating in women. International Journal of Eating
(2) interpersonal role disputes, (3) role transitions, Disorders, 45, 43–​50. doi:10.1002/​eat.20897
and (4) grief. Arcelus, J., Haslam, M., Farrow, C., & Meyer, C. (2013). The
Sick role: Role delineated by the therapist to allow role of interpersonal functioning in the maintenance of eat-
ing psychopathology: A systematic review and testable model.
the patient to be relieved of other responsibilities Clinical Psychology Review, 33, 156–​167. doi:10.1016/​
in order to recover. j.cpr.2012.10.009
Treatment phases: Interpersonal psychotherapy Arcelus, J., Whight, D., Brewin, N., & McGrain, L. (2012). A
is divided into three phases: The initial phase is brief form of interpersonal psychotherapy for adult patients
with bulimic disorders:  A  pilot study. European Eating
dedicated to identifying the problem area(s) that Disorders Review, 20, 326–​330. doi:10.1002/​erv.2164
will be the target for treatment. The intermediate Birch, L. L., Fisher, J. O., & Davison, K. K. (2003). Learning
phase is devoted to working on the target problem to overeat:  Maternal use of restrictive feeding practices

Burke, Karam, Tanofsky-Kraff, Wilfley 313

 promotes girls’ eating in the absence of hunger. American Dietz, L. J., Weinberg, R. J., Brent, D. A., & Mufson, L.
Journal of Clinical Nutrition, 78, 215–​220. (2015). Family-​ based interpersonal psychotherapy for
Birchall, H. (1999). Interpersonal psychotherapy in the treat- depressed preadolescents:  examining efficacy and potential
ment of eating disorder. European Eating Disorders Review, treatment mechanisms. Journal of the American Academy of
7, 315–​320. Child and Adolescent Psychiatry, 54, 191–​199. doi:10.1016/​
Blomquist, K. K., Ansell, E. B., White, M. A., Masheb, j.jaac.2014.12.011
R. M., & Grilo, C. M. (2012). Interpersonal problems Dounchis, J. Z., Welch, R. R., & Wilfley, D. E. (1999). Using
and developmental trajectories of binge eating disorder. group interpersonal psychotherapy (IPT-​G) for the treatment
Comprehensive Psychiatry, 53, 1088–​1095. doi:10.1016/​ of binge eating disorders. Paper presented at the Academy of
j.comppsych.2012.05.003 Eating Disorders Annual Meeting, San Diego, CA.
Bowlby, J. (1982). Attachment and loss (2nd ed., Vol. 1). New York, Duchesne, M., de Oliveira Falcone, E. M., de Freitas, S. R.,
NY: Basic Books. D’Augustin, J. F., Marinho, V., & Appolinario, J. C. (2012).
Braet, C., & Van Strien, T. (1997). Assessment of emotional, Assessment of interpersonal skills in obese women with binge
externally induced and restrained eating behaviour in nine eating disorder. Journal of Health Psychology, 17, 1065–​1075.
to twelve-​year-​old obese and non-​obese children. Behaviour doi:10.1177/​1359105311432326
Research and Therapy, 35, 863–​873. Elliott, C. A., Tanofsky-​Kraff, M., Shomaker, L. B., Columbo,
Bravender, T. (2005). Health, education, and youth in K. M., Wolkoff, L. E., Ranzenhofer, L. M., & Yanovski,
Durham: HEY-​Durham curricular guide (2nd ed.). Durham, J. A. (2010). An examination of the interpersonal model
NC: Duke University Press. of loss of control eating in children and adolescents.
Carter, F. A., Jordan, J., McIntosh, V. V. W., Luty, S. E., McKenzie, Behaviour Research and Therapy, 48, 424–​428. doi:10.1016/​
J. M., Frampton, C. M. A.,  . . .  Joyce, P. R. (2011). The j.brat.2009.12.012
long‐term efficacy of three psychotherapies for anorexia ner- Evans, L., & Wertheim, E. H. (1998). Intimacy patterns and
vosa: A randomized, controlled trial. International Journal of relationship satisfaction of women with eating problems and
Eating Disorders, 44, 647–​654. doi:10.1002/​eat.20879 the mediating effects of depression, trait anxiety and social
Cassidy, O., Eichen, D. M., Burke, N. L., Patmore, J., Shore, anxiety. Journal of Psychosomatic Research, 44, 355–​365.
A., Radin, R. M.,  . . .  Tanofsky-​Kraff, M. (2016). Adapting Fairburn, C. G. (1997). Interpersonal psychotherapy for buli-
interpersonal psychotherapy for the prevention of excessive mia nervosa. In D. M. Garner & P. E. Garfinkel (Eds.),
weight gain:  Considerations for urban African American Handbook of treatment for eating disorders (Vol. 2, pp. 278–​
girls. Manuscript submitted for publication. 294). New York, NY: Guilford.
Cassidy, O., Sbrocco, T., Vannucci, A., Nelson, B., Jackson-​ Fairburn, C. G. (2008). Cognitive behavior therapy and eating dis-
Bowen, D., Heimdal, J.,   .  .  .  Tanofsky-​Kraff, M. (2013). orders. New York, NY: Guilford Press.
Adapting interpersonal psychotherapy for the prevention of Fairburn, C. G., Doll, H. A., Welch, S. L., Hay, P. J., Davies,
excessive weight gain in rural African American girls. Journal B. A., & O’Connor, M. E. (1998). Risk factors for binge
of Pediatric Psychology, 38, 965–​977. doi:10.1093/​jpepsy/​ eating disorder:  A  community-​ based, case-​control study.
jst029 Archives of General Psychiatry, 55, 425–​432. doi:10.1001/​
Chui, W., Safer, D. L., Bryson, S. W., Agras, W. S., & Wilson, G. archpsyc.55.5.425
T. (2007). A comparison of ethnic groups in the treatment of Fairburn, C. G., Jones, R., Peveler, R. C., Carr, S. J., Solomon,
bulimia nervosa. Eating Behaviors, 8, 458–​491. doi:10.1016/​ R. A., O’Connor, M. E.,  . . .  Hope, R. A. (1991). Three
j.eatbeh.2007.01.005 psychological treatments for bulimia nervosa: A comparative
Constantino, M. J., Arnow, B. A., Blasey, C., & Agras, W. S. trial. Archives of General Psychiatry, 48, 463–​469.
(2005). The association between patient characteristics and Fairburn, C. G., Norman, P. A., Welch, S. L., O’Connor, M. E.,
the therapeutic alliance in cognitive-​behavioral and interper- Doll, H. A., & Peveler, R. C. (1995). A prospective study
sonal therapy for bulimia nervosa. Journal of Consulting and of outcome in bulimia nervosa and the long-​term effects of
Clinical Psychology, 73, 203–​211. three psychological treatments. Archives of General Psychiatry,
Crow, S. J., Stewart Agras, W., Halmi, K., Mitchell, J. E., & 52, 304–​312.
Kraemer, H. C. (2002). Full syndromal versus subthreshold Fairburn, C. G., Peveler, R. C., Jones, R., Hope, R. A., & Doll,
anorexia nervosa, bulimia nervosa, and binge eating disor- H. A. (1993). Predictors of 12-​month outcome in bulimia
der:  A  multicenter study. International Journal of Eating nervosa and the influence of attitudes to shape and weight.
Disorders, 32, 309–​318. Journal of Consulting and Clinical Psychology, 61, 696–​698.
Cuijpers, P., Donker, T., Weissman, M. M., Ravitz, P., & Cristea, Fairburn, C. G., Welch, S. L., Doll, H. A., Davies, B. A., &
I. A. (2016). Interpersonal psychotherapy for mental O’Connor, M. E. (1997). Risk factors for bulimia ner-
health problems: A comprehensive meta-​analysis. American vosa:  A  community-​based case-​control study. Archives
Journal of Psychiatry, 173, 680–​687. doi:10.1176/​appi. of General Psychiatry, 54, 509–​517. doi:10.1001/​
ajp.2015.15091141 archpsyc.1997.01830180015003
Devlin, M. J., Goldfein, J. A., Petkova, E., Jiang, H., Raizman, Fairburn, C. G., & Wilson, G. T. (2013). The dissemination and
P. S., Wolk, S.,  . . .  Walsh, B. T. (2005). Cognitive behav- implementation of psychological treatments:  Problems and
ioral therapy and fluoxetine as adjuncts to group behav- solutions. International Journal of Eating Disorders, 46, 516–​
ioral therapy for binge eating disorder. Obesity Research, 13, 521. doi:10.1002/​eat.22110
1077–​1088. Fallon, E. M., Tanofsky-​Kraff, M., Norman, A. C., McDuffie, J.
Dietz, L. J., Mufson, L., Irvine, H., & Brent, D. A. (2008). R., Taylor, E. D., Cohen, M. L.,  . . .  Yanovski, J. A. (2005).
Family-​ based Interpersonal Psychotherapy (IPT) for Health-​related quality of life in overweight and nonover-
depressed preadolescents:  An open treatment trial. Early weight black and white adolescents. Journal of Pediatrics,
Intervention Psychiatry, 2, 154–​161. 147, 443–​450.

314 Interpersonal Psychotherapy

Fisher, J. O., & Birch, L. L. (2002). Eating in the absence of Journal of Consulting and Clinical Psychology, 75, 645–​651.
hunger and overweight in girls from 5 to 7 y of age. American doi:10.1037/​0022-​006X.75.4.645
Journal of Clinical Nutrition, 76, 226–​231. Humphrey, L. L. (1989). Observed family interactions among
Frank, E., & Spanier, C. (1995). Interpersonal psychotherapy for subtypes of eating disorders using structural analysis of social
depression: Overview, clinical efficacy, and future directions. behavior. Journal of Consulting and Clinical Psychology, 57,
Clinical Psychology, 2, 349–​369. 206–​214.
Freeman, L. M. Y., & Gil, K. M. (2004). Daily stress, coping, Ivanova, I. V., Tasca, G. A., Hammond, N., Balfour, L., Ritchie,
and dietary restraint in binge eating. International Journal of K., Koszycki, D., & Bissada, H. (2015). Negative affect
Eating Disorders, 36, 204–​212. mediates the relationship between interpersonal problems
Ghaderi, A., & Scott, B. (1999). Prevalence and psycho- and binge-​eating disorder symptoms and psychopathology in
logical correlates of eating disorders among females aged a clinical sample: A test of the interpersonal model. European
18–​30  years in the general population. Acta Psychiatrica Eating Disorders Review, 23, 133–​138. doi:10.1002/​erv.2344
Scandinavica, 99, 261–​266. Ivanova, I. V., Tasca, G. A., Proulx, G., & Bissada, H. (2015).
Glass, T. A., & McAtee, M. J. (2006). Behavioral science at the Does the interpersonal model apply across eating disorder
crossroads in public health: Extending horizons, envisioning diagnostic groups? A structural equation modeling approach.
the future. Social Science and Medicine, 62, 1650–​1671. Comprehensive Psychiatry, 63, 80–​87. doi:10.1016/​
Goldschmidt, A. B., Best, J. R., Stein, R. I., Saelens, B. E., j.comppsych.2015.08.009
Epstein, L. H., & Wilfley, D. E. (2014). Predictors of child Jacobs, M. J., Welch, R. R., & Wilfley, D. E. (2004).
weight loss and maintenance among family-​based treatment Interpersonal psychotherapy for anorexia nervosa, bulimia
completers. Journal of Consulting and Clinical Psychology, 82, nervosa, and binge eating disorder. In T. Brewerton (Ed.),
1140–​1150. doi:10.1037/​a0037169 Clinical handbook of eating disorders: An integrated approach
Goldschmidt, A. B., Loth, K. A., MacLehose, R. F., Pisetsky, (pp. 449–​472). New York, NY: Marcel Dekker.
E. M., Berge, J. M., & Neumark-​ Sztainer, D. (2015). Johnson, J. G., Spitzer, R. L., & Williams, J. B. (2001). Health prob-
Overeating with and without loss of control:  Associations lems, impairment and illnesses associated with bulimia nervosa
with weight status, weight-​related characteristics, and psy- and binge eating disorder among primary care and obstetric
chosocial health. The International Journal of Eating Disorders, gynaecology patients. Psychological Medicine, 31, 1455–​1466.
48, 1150–​1157. doi:10.1002/​eat.22465 Johnson, S. L. (2000). Improving preschoolers’ self-​regulation of
Goossens, L., Braet, C., & Decaluwe, V. (2006). Loss of con- energy intake. Pediatrics, 106, 1429–​1435.
trol over eating in obese youngsters. Behaviour Research and Jones, A., Lindekilde, N., Lubeck, M., & Clausen, L. (2015).
Therapy, 45, 1–​9. doi:10.1016/​j.brat.2006.01.006 The association between interpersonal problems and treat-
Grilo, C. M., Masheb, R. M., & Wilson, G. T. (2005). Efficacy ment outcome in the eating disorders: A systematic review.
of cognitive behavioral therapy and fluoxetine for the treat- Nordic Journal of Psychiatry, 69, 563–​573. doi:10.3109/​
ment of binge eating disorder: A randomized double-​blind 08039488.2015.1019924
placebo-​controlled comparison. Biological Psychiatry, 57, Kazdin, A. E., Fitzsimmons-​Craft, E. E., & Wilfley, D. E. (2016).
301–​309. Addressing critical gaps in the treatment of eating disorders.
Grissett, N. I., & Norvell, N. K. (1992). Perceived social support, Manuscript submitted for publication.
social skills, and quality of relationships in bulimic women. Keel, P. K., Dorer, D. J., Franko, D. L., Jackson, S. C., & Herzog,
Journal of Consulting and Clinical Psychology, 60, 293–​299. D. B. (2005). Postremission predictors of relapse in women
Gual, P., Perez-​Gaspar, M., Martinez-​Gonzalez, M. A., Lahortiga, with eating disorders. American Journal of Psychiatry, 162,
F., de Irala-​Estevez, J., & Cervera-​Enguix, S. (2002). Self-​ 2263–​2268.
esteem, personality, and eating disorders:  Baseline assess- Kelly, N. R., Shomaker, L. B., Pickworth, C. K., Brady, S. M.,
ment of a prospective population-​based cohort. International Courville, A. B., Bernstein, S.,  . . .  Yanovski, J. A. (2015).
Journal of Eating Disorders, 31, 261–​273. A prospective study of adolescent eating in the absence of
Hartmann, A., Zeeck, A., & Barrett, M. S. (2010). Interpersonal hunger and body mass and fat mass outcomes. Obesity, 23,
problems in eating disorders. International Journal of Eating 1472–​1478. doi:10.1002/​oby.21110
Disorders, 43, 619–​627. doi:10.1002/​eat.20747 Kenardy, J., Mensch, M., Bowen, K., Green, B., & Walton, J.
Hayden-​Wade, H. A., Stein, R. I., Ghaderi, A., Saelens, B. E., (2002). Group therapy for binge eating in Type 2 diabe-
Zabinski, M. F., & Wilfley, D. E. (2005). Prevalence, char- tes: A randomized trial. Diabetic Medicine, 19, 234–​239.
acteristics, and correlates of teasing experiences among over- Klerman, G. L., Weissman, M. M., Rounsaville, B. J., &
weight children vs. non-​overweight peers. Obesity Research, Chevron, E. S. (1984). Interpersonal psychotherapy of depres-
13, 1381–​1392. doi:10.1038/​oby.2005.167 sion. New York, NY: Basic Books.
Herzog, D., Keller, M., Lavori, P., & Ott, I. (1987). Social Lansigan, R. K., Emond, J. A., & Gilbert-​Diamond, D. (2015).
impairment in bulimia. International Journal of Eating Understanding eating in the absence of hunger among young
Disorders, 6, 741–​747. children: A systematic review of existing studies. Appetite, 85,
Hilbert, A., Bishop, M. E., Stein, R. I., Tanofsky-​Kraff, M., 36–​47. doi:10.1016/​j.appet.2014.10.032
Swenson, A. K., Welch, R. R., & Wilfley, D. E. (2012). Lavender, J. M., Utzinger, L. M., Cao, L., Wonderlich, S.
Long-​term efficacy of psychological treatments for binge eat- A., Engel, S. G., Mitchell, J. E., & Crosby, R. D. (2016).
ing disorder. The British Journal of Psychiatry, 200, 232–​237. Reciprocal associations between negative affect, binge eat-
doi:10.1192/​bjp.bp.110.089664 ing, and purging in the natural environment in women with
Hilbert, A., Saelens, B. E., Stein, R. I., Mockus, D. S., Welch, bulimia nervosa. Journal of Abnormal Psychology, 125, 381–​
R. R., Matt, G. E., & Wilfley, D. E. (2007). Pretreatment 386. doi:10.1037/​abn0000135
and process predictors of outcome in interpersonal and cog- Lemeshow, A. R., Fisher, L., Goodman, E., Kawachi, I., Berkey,
nitive behavioral psychotherapy for binge eating disorder. C. S., & Colditz, G. A. (2008). Subjective social status in the

Burke, Karam, Tanofsky-Kraff, Wilfley 315

 school and change in adiposity in female adolescents: find- Using ecological momentary assessment to examine inter-
ings from a prospective cohort study. Archives of Pediatrics personal and affective predictors of loss of control eating in
and Adolescent Medicine, 162, 23–​28. adolescent girls. International Journal of Eating Disorders, 47,
Markowitz, J. C., Skodol, A. E., & Bleiberg, K. (2006). 748–​757. doi:10.1002/​eat.22333
Interpersonal psychotherapy for borderline personality dis- Ravitz, P., Wondimagegn, D., Pain, C., Araya, M., Alem, A.,
order:  Possible mechanisms of change. Journal of Clinical Baheretibeb, Y.,   .  .  .  Leszcz, M. (2014). Psychotherapy
Psychology, 62, 431–​444. doi:10.1002/​jclp.20242 knowledge translation and interpersonal psychother-
McIntosh, V. V., Bulik, C. M., McKenzie, J. M., Luty, S. E., & apy:  Using best-​ education practices to transform mental
Jordan, J. (2000). Interpersonal psychotherapy for anorexia health care in Canada and Ethiopia. American Journal of
nervosa. International Journal of Eating Disorders, 27, 125–​139. Psychotherapy, 68, 463–​488.
McIntosh, V. V., Jordan, J., Carter, F. A., Luty, S. E., McKenzie, Raykos, B. C., McEvoy, P. M., Carter, O., Fursland, A., &
J. M., Bulik, C. M.,   .  .  .  Joyce, P. R. (2005). Three psy- Nathan, P. (2014). Interpersonal problems across restrictive
chotherapies for anorexia nervosa: A randomized, controlled and binge-​purge samples:  Data from a community-​based
trial. American Journal of Psychiatry, 162, 741–​747. eating disorders clinic. Eating Behaviors, 15, 449–​452.
Meyer, A. (1957). Psychobiology:  A  science of man. Springfield, doi:10.1016/​j.eatbeh.2014.06.008
IL: Thomas. Ricca, V., Mannucci, E., Mezzani, B., Moretti, S., Di Bernardo,
Moens, E., & Braet, C. (2007). Predictors of disinhibited eating M., Bertelli, M.,  . . .  Faravelli, C. (2001). Fluoxetine and
in children with and without overweight. Behaviour Research fluvoxamine combined with individual cognitive-​behaviour
and Therapy, 45, 1357–​1368. therapy in binge eating disorder: A one-​year follow-​up study.
Mufson, L., Dorta, K. P., Moreau, D., & Weissman, M. M. Psychotherapy and Psychosomatics, 70, 298–​306.
(2004). Interpersonal psychotherapy for depressed adolescents Rieger, E., Van Buren, D. J., Bishop, M., Tanofsky-​Kraff, M.,
(2nd ed.). New York, NY: Guilford Press. Welch, R., & Wilfley, D. E. (2010). An eating disorder-​specific
Nauta, H., Hospers, H., Kok, G., & Jansen, A. (2000). A com- model of interpersonal psychotherapy (IPT-​ ED):  Causal
parison between a cognitive and a behavioral treatment for pathways and treatment implications. Clinical Psychology
obese binge eaters and obese non-​ binge eaters. Behavior Review, 30, 400–​410. doi:10.1016/​j.cpr.2010.02.001
Therapy, 21, 441–​461. Rorty, M., Yager, J., Buckwalter, J. G., & Rossotto, E. (1999).
Neumark-​ Sztainer, D., Falkner, N., Story, M., Perry, C., Social support, social adjustment, and recovery status in buli-
Hannan, P. J., & Mulert, S. (2002). Weight-​teasing among mia nervosa. International Journal of Eating Disorders, 26, 1–​12.
adolescents:  Correlations with weight status and disordered Ruuska, J., Koivisto, A. M., Rantanen, P., & Kaltiala-​Heino, R.
eating behaviors. International Journal of Obesity and Related (2007). Psychosocial functioning needs attention in adoles-
Metabolic Disorders, 26, 123–​131. doi:10.1038/​sj.ijo.0801853 cent eating disorders. Nordic Journal of Psychiatry, 61, 452–​
Nevonen, L., & Broberg, A. G. (2006). A comparison of 458. doi:10.1080/​08039480701773253
sequenced individual and group psychotherapy for patients Schvey, N. A., Sbrocco, T., Stephens, M., Bryant, E. J., Ress, R.,
with bulimia nervosa. International Journal of Eating Spieker, E. A.,  . . .  Tanofsky-​Kraff, M. (2015). Comparison
Disorders, 39, 117–​127. of overweight and obese military-​dependent and civilian ado-
NIH, O. R. T. F. (2011). Strategic plan for NIH obesity research. lescent girls with loss-​of-​control eating. International Journal
Bethesda, MD:  Author. Retrieved from https://​obesi- of Eating Disorders, 48, 790–​794. doi:10.1002/​eat.22424
tyresearch.nih.gov/​about/​StrategicPlanforNIH_​Obesity_​ Schwimmer, J. B., Burwinkle, T. M., & Varni, J. W. (2003).
Research_​Full-​Report_​2011.pdf Health-​related quality of life of severely obese children and
O’Mahony, J. F., & Hollwey, S. (1995a). The correlates of binge adolescents. JAMA, 289, 1813–​1819.
eating in two nonpatient samples. Addictive Behaviors, 20, Shomaker, L. B., Tanofsky-​Kraff, M., Zocca, J. M., Courville, A.,
471–​480. Kozlosky, M., Columbo, K. M.,  . . .  Yanovski, J. A. (2010).
O’Mahony, J. F., & Hollwey, S. (1995b). Eating problems and Eating in the absence of hunger in adolescents: intake after
interpersonal functioning among several groups of women. a large-​array meal compared with that after a standardized
Journal of Clinical Psychology, 51, 345–​351. meal. American Journal of Clinical Nutrition, 92, 697–​703.
Ogden, C. L., Carroll, M. D., Kit, B. K., & Flegal, K. M. (2014). doi:10.3945/​ajcn.2010.29812
Prevalence of childhood and adult obesity in the United Shunk, J. A., & Birch, L. L. (2004). Girls at risk for overweight at
States, 2011–​2012. JAMA, 311, 806–​814. doi:10.1001/​ age 5 are at risk for dietary restraint, disinhibited overeating,
jama.2014.732 weight concerns, and greater weight gain from 5 to 9 years.
Ogden, C. L., Carroll, M. D., Lawman, H. G., Fryar, C. D., Journal of the American Dietetic Association, 104, 1120–​1126.
Kruszon-​Moran, D., Kit, B. K., & Flegal, K. M. (2016). Sonneville, K. R., Horton, N. J., Micali, N., Crosby, R.
Trends in obesity prevalence among children and adoles- D., Swanson, S. A., Solmi, F., & Field, A. E. (2013).
cents in the United States, 1988–​1994 through 2013–​2014. Longitudinal associations between binge eating and over-
JAMA, 315, 2292–​2299. doi:10.1001/​jama.2016.6361 eating and adverse outcomes among adolescents and young
Puhl, R. M. (2011). Weight stigmatization toward youth: A sig- adults:  Does loss of control matter? JAMA Pediatrics, 167,
nificant problem in need of societal solutions. Childhood 149–​155. doi:10.1001/​2013.jamapediatrics.12
Obesity, 7, 359–​363. doi:10.1089/​chi.2011.0500.pers Steiger, H., Gauvin, L., Jabalpurwala, S., Seguin, J. R., &
Radin, R. M., Tanofsky-​Kraff, M., Shomaker, L. B., Kelly, N. R., Stotland, S. (1999). Hypersensitivity to social interactions in
Pickworth, C. K., Shank, L. M.,  . . .  Yanovski, J. A. (2015). bulimic syndromes: Relationship to binge eating. Journal of
Metabolic characteristics of youth with loss of control eating. Consulting and Clinical Psychology, 67, 765–​775.
Eating Behaviors, 19, 86–​89. doi:S1471-​015300099-​9 Stewart, M. O., Raffa, S. D., Steele, J. L., Miller, S. A.,
Ranzenhofer, L. M., Engel, S. G., Crosby, R. D., Anderson, M., Clougherty, K. F., Hinrichsen, G. A., & Karlin, B. E. (2014).
Vannucci, A., Cohen, L. A.,  . . .  Tanofsky-​Kraff, M. (2014). National dissemination of interpersonal psychotherapy for

316 Interpersonal Psychotherapy

 depression in veterans: therapist and patient-​level outcomes. Tanofsky-​Kraff, M., Wilfley, D. E., & Spurrell, E. (2000). Impact
Journal of Consulting and Clinical Psychology, 82, 1201–​ of interpersonal and ego-​related stress on restrained eaters.
1206. doi:10.1037/​a0037410 International Journal of Eating Disorders, 27, 411–​418.
Stice, E., Gau, J. M., Rohde, P., & Shaw, H. (2017). Risk fac- Tanofsky-​Kraff, M., Wilfley, D. E., Young, J. F., Mufson, L.,
tors that predict future onset of each DSM-​5 eating disor- Yanovski, S. Z., Glasofer, D. R., & Salaita, C. G. (2007).
der:  Predictive specificity in high-​risk adolescent females. Preventing excessive weight gain in adolescents: Interpersonal
Journal of Abnormal Psychology, 126, 38–​51. doi:10.1037/​ psychotherapy for binge eating. Obesity, 15, 1345–​1355.
abn0000219 doi:10.1038/​oby.2007.162
Strauss, R. S., & Pollack, H. A. (2003). Social marginalization Tanofsky-​Kraff, M., Wilfley, D. E., Young, J. F., Mufson, L.,
of overweight children. Archives of Pediatrics and Adolescent Yanovski, S. Z., Glasofer, D. R.,  . . .  Schvey, N. A. (2010).
Medicine, 157, 746–​752. A pilot study of interpersonal psychotherapy for prevent-
Striegel-​Moore, R. H., Silberstein, L. R., & Rodin, J. (1986). ing excess weight gain in adolescent girls at-​risk for obe-
Toward an understanding of risk factors for bulimia. sity. International Journal of Eating Disorders, 43, 701–​706.
American Psychologist, 41, 246–​263. doi:10.1002/​eat.20773
Svaldi, J., Dorn, C., & Trentowska, M. (2011). Effectiveness for Tanofsky-​Kraff, M., Yanovski, S. Z., Schvey, N. A., Olsen, C. H.,
interpersonal problem-​solving is reduced in women with Gustafson, J., & Yanovski, J. A. (2009). A prospective study
binge eating disorder. European Eating Disorders Review, 19, of loss of control eating for body weight gain in children at
331–​341. doi:10.1002/​erv.1050 high risk for adult obesity. International Journal of Eating
Tanofsky-​Kraff, M. (2008). Binge eating among children and Disorders, 42, 26–​30. doi:10.1002/​eat.20580
adolescents. In E. Jelalian & R. Steele (Eds.), Handbook Tasca, G. A., Balfour, L., Presniak, M. D., & Bissada, H. (2012).
of child and adolescent obesity (pp. 41–​57):  New  York, NY. Outcomes of specific interpersonal problems for binge eat-
Springer. ing disorder:  Comparing group psychodynamic interper-
Tanofsky-​Kraff, M., Cohen, M. L., Yanovski, S. Z., Cox, C., sonal psychotherapy and group cognitive behavioral therapy.
Theim, K. R., Keil, M.,  . . .  Yanovski, J. A. (2006). A pro- International Journal of Group Psychotherapy, 62, 197–​218.
spective study of psychological predictors of body fat gain doi:10.1521/​ijgp.2012.62.2.197
among children at high risk for adult obesity. Pediatrics, 117, Tasca, G. A., Mikail, S., & Hewitt, P. L. (2005). Group psy-
1203–​1209. doi:117/​4/​1203 chodynamic interpersonal psychotherapy:  Summary of
Tanofsky-​Kraff, M., Ranzenhofer, L. M., Yanovski, S. Z., Schvey, a treatment model and outcomes for depressive symp-
N. A., Faith, M., Gustafson, J., & Yanovski, J. A. (2008). toms. In M. E. Abelian (Ed.), Focus on Psychotherapy
Psychometric properties of a new questionnaire to assess eat- Research (pp. 159–​188). Hauppauge, NY:  Nova Science
ing in the absence of hunger in children and adolescents. Publishers.
Appetite, 51, 148–​155. Tasca, G. A., Taylor, D., Ritchie, K., & Balfour, L. (2004).
Tanofsky-​Kraff, M., Shomaker, L. B., Olsen, C., Roza, C. A., Attachment predicts treatment completion in an eating dis-
Wolkoff, L. E., Columbo, K. M.,  . . .  Yanovski, J. A. (2011). orders partial hospital program among women with anorexia
A prospective study of pediatric loss of control eating and nervosa. Journal of Personality Assessment, 83, 201–​212.
psychological outcomes. Journal of Abnormal Psychology, 120, Telch, C. F., Agras, W. S., Rossiter, E. M., Wilfley, D. E., &
108–​118. doi:10.1037/​a0021406 Kenardy, J. (1990). Group cognitive-​behavioral treatment
Tanofsky-​Kraff, M., Shomaker, L. B., Stern, E. A., Miller, R., for the nonpurging bulimic: An initial evaluation. Journal of
Sebring, N., Dellavalle, D.,   .  .  .  Yanovski, J. A. (2012). Consulting and Clinical Psychology, 58, 629–​635.
Children’s binge eating and development of metabolic Thompson-​Brenner, H., & Westen, D. (2005). A natu-
syndrome. International Journal of Obesity, 36, 956–​962. ralistic study of psychotherapy for bulimia nervosa,
doi:10.1038/​ijo.2011.259 Part 2: Therapeutic interventions in the community. Journal
Tanofsky-​Kraff, M., Shomaker, L. B., Wilfley, D. E., Young, J. of Nervous and Mental Disease, 193, 585–​595.
F., Sbrocco, T., Stephens, M.,  . . .  Yanovski, J. A. (2017). Troisi, A., Massaroni, P., & Cuzzolaro, M. (2005). Early sepa-
Excess weight gain prevention in adolescents:  Three-​ ration anxiety and adult attachment style in women with
year outcome following a randomized-​ controlled trial. eating disorders. British Journal of Clinical Psychology, 44(Pt
Journal of Consulting and Clinical Psychology, 85, 218–​227. 1), 89–​97.
doi:10.1037/​ccp0000153 Troop, N. A., Holbrey, A., Trowler, R., & Treasure, J. L. (1994).
Tanofsky-​Kraff, M., Shomaker, L. B., Wilfley, D. E., Young, J. Ways of coping in women with eating disorders. Journal of
F., Sbrocco, T., Stephens, M.,  . . .  Yanovski, J. A. (2014). Nervous and Mental Disease, 182, 535–​540.
Targeted prevention of excess weight gain and eating disor- Vall, E., & Wade, T. D. (2015). Predictors of treatment outcome
ders in high-​risk adolescent girls:  A  randomized controlled in individuals with eating disorders: A systematic review and
trial. American Journal of Clinical Nutrition, 100, 1010–​ meta-​analysis. International Journal of Eating Disorders, 48,
1018. doi:10.3945/​ajcn.114.092536 946–​971. doi:10.1002/​eat.22411
Tanofsky-​Kraff, M., Shomaker, L. B., Young, J. F., & Wilfley, Van Strien, T. (1996). On the relationship between dieting and
D. E. (2016). Interpersonal psychotherapy for the preven- “obese” and bulimic eating patterns. International Journal of
tion of excess weight gain and eating disorders:  A  brief Eating Disorders, 19, 83–​92.
case study. Psychotherapy, 53, 188–​194. doi:10.1037/​ Van Strien, T., Engels, R. C., Van Leeuwe, J., & Snoek, H. M.
pst0000051 (2005). The Stice model of overeating: Tests in clinical and
Tanofsky-​Kraff, M., & Wifley, D. E. (2009). Interpersonal psy- non-​clinical samples. Appetite, 45, 205–​213.
chotherapy (IPT) for bulimia nervosa and binge eating disor- Weissman, M. M., Markowitz, J., & Klerman, G. L. (2000).
der. In C. Grilo & J. Mitchell (Eds.), The treatment of eating Comprehensive guide to interpersonal psychotherapy. New York,
disorders (pp. 271–​293). New York, NY: Guilford Press. NY: Basic Behavioral Science Books.

Burke, Karam, Tanofsky-Kraff, Wilfley 317

Whisman, M. A., Dementyeva, A., Baucom, D. H., & Bulik, C. Wilfley, D. E., Wilson, G. T., & Agras, W. S. (2003). The clinical
M. (2012). Marital functioning and binge eating disorder in significance of binge eating disorder. International Journal of
married women. International Journal of Eating Disorders, 45, Eating Disorders, 34, S96–​S106.
385–​389. doi:10.1002/​eat.20935 Wilfley, D. E., Wilson, G. T., & Agras, W. S. (2008). A multi-​
Wilfley, D. E. (2008). Interpersonal psychotherapy for binge site randomized controlled trial of interpersonal psychotherapy,
eating disorder (BED) therapist’s manual. Unpublished behavioral weight loss, and guided self-​help in the treatment of
manuscript. overweight individuals with binge eating disorder. Paper pre-
Wilfley, D. E. (2015). Harnessing technology for training clinicians sented at the Eating Disorders Research Society.
to deliver Interpersonal Psychotherapy (IPT). Paper presented Wilfley, D. E., Wilson, G. T., Van Buren, D. J., Welch, R. R.,
at the National Eating Disorders Association Conference, Eichen, D. M., Proctor, E. K.,   .  .  .  Agras, W. S. (2016).
San Diego, CA. Facilitating the dissemination and implementation of an
Wilfley, D. E., Agras, W. S., Telch, C. F., Rossiter, E. M., evidence-​ based psychological treatment for eating disorders
Schneider, J. A., Cole, A. G.,  . . .  Raeburn, S. D. (1993). within college counseling centers. Paper presented at the
Group cognitive-​behavioral therapy and group interpersonal Annual Meeting of the Eating Disorders Research Society,
psychotherapy for the nonpurging bulimic individual:  a New York, NY.
controlled comparison. Journal of Consulting and Clinical Wilson, G. T. (2005). Psychological treatment of eating dis-
Psychology, 61, 296–​305. orders. Annual Review of Clinical Psychology, 1, 439–​465.
Wilfley, D. E., Dounchis, J. Z., & Welch, R. R. (2004). doi:10.1146/​annurev.clinpsy.1.102803.144250
Interpersonal psychotherapy of anorexia nervosa. In K. M. Wilson, G. T., & Fairburn, C. G. (2001). Eating disorders. In
Miller & J. S. Mizes (Eds.), Comparative treatment of eating P. N.  J. Gorman (Ed.), Treatments that work. New  York,
disorders. New York, NY: Springer. NY: Oxford University Press.
Wilfley, D. E., Frank, M. A., Welch, R. R., Spurrell, E. B., & Wilson, G. T., Grilo, C. M., & Vitousek, K. M. (2007).
Rounsaville, B. J. (1998). Adapting interpersonal psycho- Psychological treatment of eating disorders. American
therapy to a group format (IPT-​G) for binge eating disor- Psychologist, 62, 199–​216. doi:10.1037/​0003-​066X.62.3.199
der:  Toward a model for adapting empirically supported Wilson, G. T., & Shafran, R. (2005). Eating disorders guide-
treatments. Psychotherapy Research, 8, 379–​391. lines from NICE. Lancet, 365(9453), 79–​81. doi:10.1016/​
Wilfley, D. E., MacKenzie, K. R., Welch, R. R., Ayres, V. E., S0140-​673617669-​1
& Weissman, M. M. (2000). Interpersonal psychotherapy for Wilson, G. T., Wilfley, D. E., Agras, W. S., & Bryson, S. W.
group. New York, NY: Basic Books. (2010). Psychological treatments of binge eating disorder.
Wilfley, D. E., Pike, K. M., & Striegel-​Moore, R. H. (1997). Archives of General Psychiatry, 67, 94–​101. doi:10.1001/​
Toward an integrated model of risk for binge eating disorder. archgenpsychiatry.2009.170
Journal of Gender, Culture, and Health, 2, 1–​3. Wing, R. R., & Jeffery, R. W. (1999). Benefits of recruiting
Wilfley, D. E., Stein, R. I., Saelens, B. E., Mockus, D. S., Matt, participants with friends and increasing social support for
G. E., Hayden-​Wade, H. A.,   .  .  .  Epstein, L. H. (2007). weight loss and maintenance. Journal of Consulting and
Efficacy of maintenance treatment approaches for childhood Clinical Psychology, 67, 132–​138.
overweight:  A  randomized controlled trial. Journal of the Young, J. F., & Mufson, L. (2003). Manual for interpersonal
American Medical Association, 298, 1661–​1673. psychotherapy-​adolescent skills training (IPT-​AST). New York,
Wilfley, D. E., Stein, R. I., & Welch, R. R. (2005). Interpersonal NY: Columbia University.
psychotherapy. In J. Treasure, U. Schmidt, & E. van Furth Young, J. F., Mufson, L., & Davies, M. (2006). Efficacy of inter-
(Eds.), The essential handbook of eating disorders (pp. 137–​ personal psychotherapy-​ adolescent skills training:  An indi-
154). West Sussex, UK: John Wiley & Sons. cated preventive intervention for depression. Journal of Child
Wilfley, D. E., Welch, R. R., Stein, R. I., Spurrell, E. B., Cohen, Psychology and Psychiatry and Allied Disciplines, 47, 1254–​1262.
L. R., Saelens, B. E.,  . . .  Matt, G. E. (2002). A randomized Young, J. F., Mufson, L., & Schueler, C. M. (2016). Preventing
comparison of group cognitive-​behavioral therapy and group adolescent depression:  Interpersonal psychotherapy—​adolescent
interpersonal psychotherapy for the treatment of overweight skills training. New York, NY: Oxford University Press.
individuals with binge-​eating disorder. Archives of General
Psychiatry, 59, 713–​721.

318 Interpersonal Psychotherapy

 CH A PT E R


Family Therapy for Eating Disorders
16
Daniel Le Grange and Renne Rienecke

Abstract
Family therapy is increasingly recommended as the treatment of choice for eating disorders among
adolescents. The shift from blaming parents for causing an ED to seeing them as a necessary part of
the recovery process was set in motion by Salvador Minuchin and colleagues, and then reinforced and
expanded on by researchers at the Maudsley Hospital in London, UK, and in the United States and
Australia. Data supporting the efficacy of family therapy for adolescent anorexia nervosa has been
solidified, while family-​based approaches in the treatment of adolescents with bulimia nervosa show
promise. Further research is needed to replicate the findings of existing studies and to further clarify
the utility of parental involvement in the treatment of older adolescents, or transition age youth, with
anorexia nervosa and bulimia nervosa.
Key Words:  adolescent, anorexia nervosa, bulimia nervosa, family-​based treatment, eating disorder

History of Family Therapy in Eating for the next several decades. This sentiment was in
Disorders vogue until the 1960s, at which time the role of
More than 125  years ago the family was first the family was revisited in a more positive light by
considered to be at the center of eating disordered Salvador Minuchin and his colleagues at the Child
behavior. Views regarding the role of parents in ano- Guidance Center in Philadelphia (Minuchin et al.,
rexia nervosa (AN) varied from the outset. On the 1975; Minuchin, Rosman, & Baker, 1978). This
one hand, the British physician William Gull (1874) group developed what is referred to as the psycho-
considered parents as “generally the worst atten- somatic family model, a model that exerted consid-
dants,” while the French physician Charles Lasegue erable influence on subsequent treatment efforts for
(1883) took a more inclusive stance in emphasizing AN. This model hypothesized that an adolescent
that the “preoccupations of relatives” are important. will develop an ED only when a very specific fam-
Another colleague in France, Jean-​Martin Charcot ily context is in place. The psychosomatic model
(1889), described the influence of parents as “par- characterizes this family context as rigid, enmeshed,
ticularly pernicious.” These early reflections suggest overinvolved, and conflict avoidant. These processes
that parents were not seen as playing a positive role fluctuate in concert with the adolescent’s sympto-
in their child’s illness. In fact, some clinicians went matic behavior. For AN to develop, the adolescent
one step further by blaming parents for the eating should also present with a situational vulnerabil-
disorder (ED). ity, such as being given the role as a go-​between
The turn of the century did not alter the out- in cross-​generational alliances. Markedly distinct
look regarding parents’ role in ED treatment and/​ from previously established ideology, Minuchin
or development. In fact, the emergence of the term and colleagues did not simply ascribe responsibil-
“parentectomy” as a popular concept in the 1940s ity or blame for the ED to the parents. Instead,
solidified the exclusion of parents from treatment the psychosomatic model highlighted the evolving,

319
interactive nature of the development of the illness. prerequisite interactive family context within which
However, the authors did believe that the psycho- the ED develops. A  modest number of studies
somatic family was a necessary component for the (e.g., Dare, Le Grange, Eisler, & Rutherford, 1994;
development of an ED, and that treatment should Humphrey, 1989) have embarked on a course to
aim to change the way the family functions. This systematically test Minuchin’s claim of a psychoso-
view still falls short of completely absolving the par- matic family. Researchers have attempted to deter-
ents of any blame. mine whether certain characteristics are specific to
Researchers at the Institute of Psychiatry and families of a child with AN, and, therefore, whether
the Maudsley Hospital in London (Dare, 1983; these families can be considered “typical” AN
Dare & Eisler, 1997) furthered this shift in think- families. These studies were unable to confirm any
ing about the role of families in EDs. Whereas the particular pattern that typifies families with eating-​
psychosomatic model described dysfunctional fam- disordered offspring. In addition, it remains unclear
ily characteristics that were thought to be necessary whether such characteristics, if they do exist, are
for the development of an ED, Dare and Eisler were present prior to the onset of the ED, or if they are
more interested in the family dynamics that arise in instead more indicative of the family’s response to
the midst of, or as a result of, an ED. Rather than the illness.
focusing on families’ missteps and transgressions, Thus, our current state of knowledge does not
this team of researchers developed a family therapy provide sufficient evidence for the existence of the
approach that considers the parents as a resource psychosomatic family. Instead, there is growing evi-
and places less emphasis on the etiology of the ED dence that families with an ED offspring are a het-
(Eisler et al., 1997; Eisler et al., 2000; Le Grange, erogeneous group with respect to sociodemographic
Eisler, Dare, & Russell, 1992; Russell, Szmukler, characteristics, the emotional climate of intrafamil-
Dare, & Eisler, 1987). The body of work put forth ial relationships, and the patterns of interactions
by the Maudsley group has changed the emphasis in within the family (Eisler, 2005). Moreover, families
treatment from pathologizing families to absolving in which there is a member suffering with an ED
them from being blamed for causing their child’s do not change or respond to the ED in predictable
ED. The approach still requires families to change, ways. Thus, there is a need for further investigation
as steps initially taken by parents to address their to identify what the specific targets of effective fam-
child’s ED may have been ineffective or may require ily interventions should be, how these targets may
revision. differ between families, and what processes accom-
pany any changes that may occur.
Theoretical Model of Family Therapy The role of family environment in the etiol-
in Adolescent Eating Disorders ogy of EDs is also unclear. However, there is little
It is fair to say that over the past 40 years family doubt that the presence of an ED has an important
therapy has gradually established itself as one of the effect on family life (Bara-​Carrill & Nielsen, 2003).
most prominent treatment approaches for adoles- As time passes, food, eating, and related concerns
cents with AN. The clinical and theoretical accounts begin to saturate family life, resulting in com-
of some of the pioneers of the family therapy field, promised family routines, coping, and problem-​
such as Minuchin and his colleagues (1975) and solving behaviors (Eisler, 2005). A  similar process
Selvini Palazzoli (1974), have been enhanced as is described for families with an alcoholic member
increasing empirical support for the efficacy of (Steinglass & Horan, 1988) and for families coping
family therapy for adolescents becomes available. with a wide range of chronic illnesses (Steinglass,
Le Grange and Eisler (2009) would argue that this 1998). According to the model of Steinglass et al.,
development has undoubtedly been one of the most families reorganize themselves in a stepwise fash-
significant changes in the treatment of EDs that the ion in response to the challenges brought about
field has witnessed in the past 10 to 15 years. by the illness. This alters the family’s routines and
Eisler (2005), however, points out that although decision-​making processes until such time that the
data for the efficacy of family therapy are mount- illness becomes the central organizing principle of
ing, quite ironically there has also been growing the family’s life. Typically, families in this position
evidence of fundamental flaws in the theoretical will attempt to minimize the impact of the illness
models on which the treatment approach is based. on either the sufferer or on other family members,
For instance, the influential psychosomatic fam- and as a consequence increasingly focus their atten-
ily model of Minuchin et  al. (1978) postulates a tion on the present moment while losing sight of

320 Family Therapy for Eating Disorders

the larger familial context. When this occurs, it Building on this RCT from Russell and his
becomes difficult for the family to meet their chang- group (1987), three subsequent studies compared
ing developmental needs. Steinglass and colleagues’ different forms of family interventions. An impor-
model can easily be applied to EDs. It is common tant difference from the original work is that the
for families dealing with an ED to comment that following studies tested outpatient family therapy
it feels as if time has come to a standstill because without prior hospitalization.
they have had to focus all their attention on the ED.
However, although there may be similarities in the Outpatient Treatment: Family Therapy
way families respond to an ED, it is quite difficult, Without Prior Hospitalization
if not impossible, to disentangle which family pro- The first of these three studies was also con-
cesses are cause or effect, or just incidental to the ducted at the Maudsley Hospital. Le Grange et al.
development of the ED. (1992) and Eisler et al. (2000) compared two forms
of outpatient family treatment—​ conjoint fam-
Family Therapy for Adolescent Anorexia ily therapy (CFT) and separated family therapy
Nervosa (SFT)—​among a total of 58 adolescents with AN.
The past 20 years have now seen about a dozen Both CFT and SFT shared the same goals, and
published randomized controlled trials (RCTs) for the treatment principles were similar to the fam-
adolescent AN. ily therapy used in the original Russell et al. (1987)
study. The two forms of treatment differed in their
The Seminal Study of Family Therapy structure:  In SFT the same therapist meets first
The first and perhaps most influential RCT of with the adolescent on her own and then meets
family therapy for EDs was conducted by Russell separately with the parents, whereas in CFT the
and colleagues at the Maudsley Hospital in London adolescent and the parents are seen together. Also,
(Russell et al., 1987). In this study, the relative effi- unlike CFT, SFT did not include a family meal
cacy of family therapy versus individual supportive as part of the treatment protocol. Overall, results
therapy was tested. Eighty female participants (ages were similar in that significant improvements were
14–​55 years) were first admitted for weight restora- reported for patients whether they were assigned to
tion to the inpatient program. Admission lasted an the conjoint or separated forms of family therapy.
average of 10 weeks, after which patients were dis- Using Morgan/​Russell outcome criteria, the major-
charged and randomly allocated to 1 year of either ity of participants (> 60%) were classified as having
family therapy or the control individual supportive a good or intermediate outcome at the end of treat-
therapy. Participants were divided into four sub- ment. The authors found that CFT was superior
groups based on diagnosis and/​or age, while out- to SFT in that significantly more change was dem-
come was defined by the Morgan/​Russell outcome onstrated in terms of individual psychological and
criteria (see Russell et al., 1987, p. 8). Findings were family functioning for participants in this treatment
inconclusive for those participants with AN whose modality (Eisler et al., 2000). One important differ-
illness had lasted more than 3 years, or for patients ence found between the treatment groups was that
with a diagnosis of bulimia nervosa (BN). However, families with high levels of parental criticism toward
findings for patients in one subgroup favored fam- their affected offspring (as defined by expressed
ily therapy. This subgroup comprised 21 adolescents emotion), did worse in CFT. Similar to the follow-​
with AN who had a relatively young age at onset up in the first RCT (Eisler et al., 1997), participants
(on or before 18 years) and a short duration of ill- in this trial continued to improve after treatment
ness (< 3 years). At 5-​year follow-​up, adolescents in ended. At 5-​year follow-​up, irrespective of type of
this same subgroup continued to do well, with 90% family therapy received, the majority of participants
of those who had received family therapy meeting had either a good (75%) or intermediate outcome
criteria for a good outcome (Eisler et  al., 1997). (15%), while only 10% failed to respond to treat-
Adolescents who had received the individual con- ment (Eisler, Simic, Russell, & Dare, 2007).
trol treatment did not do as well, with almost half
of this group still presenting with significant ED The First Family Therapy Trial Outside
symptoms after 5 years. This was the first long-​term the United Kingdom
follow-​up study to demonstrate that the benefits of The first family therapy treatment trial outside
a psychosocial treatment for AN could be main- the United Kingdom was conducted by Robin,
tained 5 years after the end of treatment. Siegal, Gilroy, Dennis, and Sikand (1999) in

Le Grange, Rienecke 321

Detroit. In this study, 38 adolescents with AN sessions over a period of 6 to 12  months. Finally,
were randomly assigned to either behavioral fam- patients in the Maudsley studies appeared to have
ily systems therapy (BFST), a treatment that been ill for longer, received more prior treatment,
shares several similarities with the Maudsley and had higher rates of comorbid depression.
conjoint family therapy, or to ego-​oriented indi-
vidual therapy (EOIT). The latter consisted of Development of a Treatment Manual
weekly individual sessions for the adolescent for Family Therapy
and bimonthly collateral sessions with the par- Other than the RCT of Robin et  al. (1999)
ents. The main goal of this treatment is to sup- for adolescent AN, family therapy treatment stud-
port the adolescent’s ability to resolve challenges ies have been limited to the Maudsley group. One
through strengthening ego development rather reason for the limited use of this helpful treatment
than resorting to self-​starvation as an option. At approach has been the absence of a treatment man-
post-​treatment, patients in both BFST and EOIT ual. The recent development of a clinician manual
demonstrated significant improvements, in that of family therapy for adolescent AN (Lock & Le
the majority (67%) reached their target weight Grange, 2013) has not only made dissemination of
and 80% regained menstruation. Patients contin- this treatment approach possible, but also allowed
ued to improve after the conclusion of treatment. for improvements in the design of subsequent treat-
At 1-​ year follow-​up, three-​quarters of patients ment studies. The authors of the manual refer to
reached their target weight, and 85% reported this form of treatment as family-​based treatment
regular periods (Robin et al., 1999). The authors for AN (FBT-​AN) and provide details of the goals
found that BFST was superior to EOIT in terms and techniques of this treatment in the clinician’s
of physiological improvements, that is, changes manual (Lock & Le Grange, 2013).
in weight and menses, at both post-​ treatment Briefly, FBT-​AN consists of three treatment
and follow-​up, but that changes were similar for phases. The first phase focuses entirely on weight
patients in BFST and EOIT in terms of psycho- restoration, and control over this process is given
logical improvements, that is, eating attitudes, to the parents. The second phase commences when
mood, and eating-​related family conflict. Robin, the patient is approaching a healthy weight and the
Siegal, and Moyes (1995) also reported results of parents feel reassured that handing control over eat-
observational ratings of family interaction in a ing back to the adolescent will not result in renewed
subsample of this study. In this investigation they weight loss. The third phase is shorter in duration
demonstrated significant reductions in maternal and consists of a brief overview of adolescent devel-
negative communication and a corresponding opmental issues and a discussion of how the ado-
increase in positive communication for families in lescent can meet these developmental challenges
BFST but not for those in EOIT. without reverting to self-​ starvation as a coping
Some differences between the Detroit and mechanism.
Maudsley family therapy studies warrant comment, This family therapy approach has changed the
as these could have had some impact on outcome. therapeutic focus from the traditional exploration
First, patients in the Detroit study were hospital- of the etiology of the disorder to exploring how
ized at the outset of treatment if percentage of ideal and where a family has become stymied by the
body weight (% IBW) was below 75 (≈ 50% of the ED. The therapist also helps the family to iden-
sample). Such patients remained in inpatient treat- tify their strengths in order to extricate themselves
ment until they reached 80% IBW. Patients in the from the problem and explore potential solutions.
Maudsley studies (Eisler et  al., 2000; Le Grange Emphasizing that the family is a resource, and part
et  al., 1992) were treated on an outpatient basis of the solution rather than the problem, is the most
from the outset and were only admitted to the inpa- crucial element of this family therapy. More tra-
tient unit if they were unresponsive to outpatient ditional therapies place the emphasis on making
efforts to gain weight (4 out of 58 patients were changes within the family. While this is not the pri-
admitted to the inpatient service during the study). mary objective of the Maudsley group’s treatment,
Second, in the Detroit study, patients received an families may indeed learn during the course of
average number of 30 treatment sessions over a therapy that there are ways in which they function
period of 12 to 18 months. The duration and inten- as a family that they want to change. This change,
sity of treatment were lower in the Maudsley stud- however, is secondary to the primary goal, which is
ies, with patients receiving an average number of 10 to help the child overcome the ED (Eisler, 2005).

322 Family Therapy for Eating Disorders

 All of the studies described in the text that fol- about one-​third of patients across treatments remit-
lows have employed this manualized version of fam- ted at the 4-​year mark (Le Grange et al., 2014).
ily therapy1.
The French Study
The Stanford Dosage Study A French group, led by Godart, set out to inves-
The first study to use FBT-​AN was conducted tigate whether adding an adjunctive family therapy
by Lock and his colleagues (2005) at Stanford to treatment us usual (TAU) post hospitalization,
University in California. These authors examined would provide superior outcomes to TAU. The
the treatment dose of FBT-​ AN and randomly TAU included individual sessions for the patient as
assigned 86 adolescents to either a 6-​month, 10-​ well as meetings with a psychiatrist for the patient
session version of this treatment, or to a 12-​month, and her parents. Sixty female adolescents were ran-
20-​session version. At the 1-​year mark there were no domized to one of these two treatments and then
differences in weight gain between these two doses of followed up 18 months later (Godard et al., 2012).
FBT-​AN. However, some moderators of treatment The main outcome for this RCT was defined as
were identified. The longer version of this treatment Good + Intermediate vs. Poor on the Morgan/​
was more efficacious for those patients who came Russell Outcome categories (see Russell et al., 1987,
from single-​parent families, and for patients who p. 8). At 18 months follow-​up, the authors demon-
presented with higher levels of eating-​related obses- strated a significant main effect favoring the adjunc-
sions and compulsions. In what is now the third tive family therapy over TAU.
long-​term follow-​up study for this patient sample,
Lock and his colleagues (2006) found that FBT-​ The Sydney Study
AN was equally effective regardless of treatment Challenging the notion that patients typically
dose 4 years after the end of the study. That is, 66% benefit from long hospital stays until weight has
of patients achieved healthy body weights (mean been restored, Madden and colleagues allocated 82
body mass index > 20.5) and had Eating Disorder medically unstable patients to one of two groups;
Examination scores within the normal range. inpatient weight restoration (to 90% IBW) fol-
lowed by outpatient FBT (WR) versus inpatient
Chicago/​Stanford Adolescent Focused stay until medically stable followed by outpatient
Therapy Versus FBT Study FBT (MS) (Madden et al., 2015). The authors’ main
Following the dose study, Lock and Le Grange hypothesis was upheld in that remission (defined in
and their respective teams at Stanford and Chicago terms of improvements in both weight and eating
conducted a study where 121 adolescents with disorder cognitions), as well as secondary clinical
AN were randomized to either FBT or adolescent outcomes, was similar across the two treatment
focused therapy (AFT) (EOIT, as noted in the groups after 12  months of outpatient FBT. These
Robin et al. study, was renamed AFT) (Lock et al., findings underscore the notion that most medically
2010). Patients were provided 12 months of outpa- unstable adolescents require relatively brief periods
tient therapy followed by 6-​and 12-​month post-​ of inpatient stays provided the hospital treatment
treatment assessments. Defining remission in terms is followed by targeted behaviorally focused family
of recovery in weight and eating disorder cogni- treatment such as FBT.
tions, FBT was statistically superior to AFT at both
the 6-​and 12-​month follow-​up. In an attempt to Six-​Site Study of Family-​Based Therapy and
tease out which treatment may work best for which Systemic Family Therapy
patients (moderators), the authors identified two In the largest RCT of family therapy for ado-
moderators of treatment outcome, namely, ED lescents, Agras and his colleagues randomized 167
psychopathology as measured by the EDE global patients across six sites in the United States and
score, and ED-​related obsessions and compulsions Canada to either FBT or systemic family therapy
as measured by the YBC-​ED total score. For both (SyFT) (Agras et al., 2014). Testing the hypothesis
moderators, FBT is favored over AFT when EDE whether involving families in treatment (SyFT)
and YBC-​ED are high, but there is no difference is sufficient as opposed to the more behaviorally
in treatment outcome across these two treatments focused FBT that aims to support parents in weight
when EDE and YBC-​ED are low (Le Grange et al., restoration of their adolescent, patients were pro-
2012). A 4-​year follow-​up shows that rates of remis- vided 9  months of outpatient treatment and fol-
sion were relatively stable for both groups, with lowed up at 12  months post treatment. Defining

Le Grange, Rienecke 323

remission only in terms of improvement in weight, and in contrast to adolescent AN, working with
the authors found no differences between the treat- families in the treatment of adolescents with BN
ments at the 12-​ month follow-​ up assessment. has been much more limited. A recent advance has
However, they demonstrated that FBT brought allowed for the development of family-​based treat-
about weight recovery more expeditiously, with ment for bulimia nervosa (FBT-​BN) (Le Grange
fewer patients in FBT hospitalized during outpa- & Lock, 2007). This treatment was adapted from
tient treatment, and overall treatment costs were less FBT-​AN (Lock & Le Grange, 2013), and like its
in FBT than was the case for SyFT. precursor, FBT-​BN was designed for adolescents.
Arguments in favor of parental involvement in
The Melbourne Study treatment for adolescents with BN are both theo-
The most recent RCT for this patient population retically and clinically persuasive. As reviewed ear-
was conducted at the Royal Children’s Hospital lier in this chapter, a convincing body of evidence
in Melbourne, Australia (Le Grange et  al., 2016). now supports mobilizing parents to take charge of
Adolescents (N  =  107) were randomized to FBT weight restoration in the treatment of adolescents
or parent focused therapy (PFT). This study repli- with AN. Further, researchers have found that the
cates the earlier Eisler et al. (2000) and Le Grange binge-​purge subtype of AN respond favorably to
et  al. (1992) studies, but amplifies the distinction family therapy. In treatment studies for adolescent
between the two treatments; FBT is delivered in AN, where the binge-​purge subtype typically con-
conjoint format, whereas PFT is designed such that stitutes about 20% of cases, family therapy has been
the adolescent is only seen by a nurse therapist (10 found to be equally effective for weight gain as for
minutes) to determine medical stability and mental curtailing binge and purge episodes (Eisler et  al.,
status, and the full therapy hour is conducted by 2000; Lock, Agras, Bryson, & Kraemer, 2005).
the therapist with only the parents in attendance. These data seem to suggest that parents are able to
Treatment was delivered in 18 sessions over a 6-​ both alleviate bulimic symptoms in their children
month period, followed by assessments at 6 and and reverse severe dieting.
12  months post treatment. In terms of remission, Although modified from the approach for ado-
defined as recovery in weight and cognitions, PFT lescents with AN, FBT-​BN shares many key char-
was statistically superior to FBT at the end of treat- acteristics with FBT-​AN. Most prominently, both
ment, and although clinically superior at 6 months treatments emphasize parents’ love and understand-
as well, the difference between the two treatment ing of their child and encourage the family to pro-
groups was no longer statistically significant at mote behavioral change around eating. While BN
follow-​up. The findings amplify the results of the in adolescence may be experienced as egodystonic,
earlier Maudsley studies (Eisler et  al., 2000; Le patients nevertheless tend to deny the alarming
Grange et al., 1992) in that delivering family ther- nature of their symptoms and are therefore mostly
apy in a separated model (PFT) is at least as effica- unable to appreciate the seriousness of BN. Unlike
cious, perhaps even more efficacious, than conjoint a sense of pride that often accompanies starvation
FBT. A  less “cumbersome” version of FBT (e.g., in AN, binge and purge symptoms in BN can lead
there is no family meal in PFT and meetings are to heightened feelings of shame and guilt. Such
only with the parents), might be easier to dissemi- feelings tend to isolate these adolescents from
nate and implement in community practice. parental support, which in turn can reinforce the
Taken together, it is increasingly clear that paren- symptomatic behavior. However, FBT-​BN regards
tal involvement in the treatment of adolescents with the parents as a resource for resolving the ED, and
AN is not only advisable, but perhaps a prerequisite, attempts to alleviate misplaced blame that may be
for recovery to occur. While a behaviorally focused directed toward either the parents or the adolescent.
form of family therapy, such as FBT, proves to be In most instances, the adolescent suffering from BN
most advantageous for this patient population, is unable to recognize or effectively manage their
more generic forms of family therapy, such as sys- dysfunctional eating behaviors. Consequently, the
temic family therapy, seem quite feasible as well. parents are encouraged to assist their adolescent in
bringing about the necessary behavioral changes that
Family Therapy for Adolescent Bulimia will lead to recovery. Robin and colleagues (1999)
Nervosa conceptualize the teenager with AN as “unable to
Until recently, treatment development for ado- take care of herself.” If the adolescent with BN is
lescents with BN had received almost no attention, defined in the same way, then the parents should be

324 Family Therapy for Eating Disorders

coached to work as a team with their offspring to whereas the emaciation experienced in AN makes it
develop ways to restore healthy eating. This collab- relatively easier to keep treatment focused on weight
orative effort between the adolescent and her parents restoration; and (4) the therapist and parents have
shows respect and regard for the adolescent’s point to confront the challenges of comorbid illnesses in
of view and experience. Because of this collaborative BN, which can more readily derail treatment than is
stance, information about ED symptoms is shared usually the case in AN.
between the parents and the adolescent in order to
address struggles around eating and to understand Studies of Family-​Based Treatment
the impact of the disorder on family relationships. for Adolescent Bulimia Nervosa
The FBT-​BN does not delve into the possible As noted earlier, data in support of treatments for
causes of BN and is instead primarily focused on adolescents with BN are sparse. Family therapy was
the ED symptoms. In other words, this treatment first applied to adolescents with BN in a small case
focuses on what can be done to resolve the disor- series that was conducted by the Maudsley group
der. Also, FBT-​ BN assumes that both parental (Dodge, Hodes, Eisler, & Dare, 1995). This study
guilt about having possibly caused the illness and demonstrated significant reductions in bulimic
anxiety about how best to address the symptomatic behaviors through educating the family about the
behavior serve to disable parents in their efforts. ED and helping the parents to disrupt binge eating
Consequently, a primary goal of treatment is to and purging episodes. Following the case series by
empower the parents and the adolescent in their Dodge and her colleagues (1995), Le Grange and
collaborative attempts to disrupt the ED behaviors. his colleagues (2003) provided a detailed descrip-
Another important goal of treatment is to exter- tion of an adolescent progressing in FBT-​BN. Both
nalize the disordered behaviors from the affected of these studies suggested that families can play a
adolescent. This separation of the adolescent from positive role in the recovery of adolescent BN, and
the disorder serves to promote parental action and that this is a promising avenue to pursue in the
decrease adolescent resistance to their assistance. treatment for this population. These preliminary
Once these goals have been accomplished, the par- findings were recently extended with the publica-
ents’ next task is to return control over eating to the tion of the first RCTs for adolescents with BN;
adolescent in a way that is age appropriate; that is, all three studies using family treatments in their
control over eating may be different for a 12-​year-​ design (Le Grange, Crosby, Rathouz, & Leventhal,
old versus an 18-​year-​old. Siblings are encouraged 2007; Le Grange, Lock, Agras, Bryson, & Jo, 2015;
to play a supportive role only, and are therefore shel- Schmidt et al., 2007).
tered from the job assigned to the parents. Once In the Le Grange et al. (2007) study, 80 patients
the ED symptoms have resolved and the patient is with DSM-​IV BN and partial BN, ranging in age
eating on her own in an age-​appropriate way, par- from 12 to 19 years (mean age = 16.1 years; mean
ents will then assist her in negotiating predictable duration of illness  =  20.6  months), were assigned
adolescent developmental tasks. The therapist aims to either FBT-​BN (n  =  41) or to individual sup-
to take a nondirective stance throughout treatment portive psychotherapy (SPT) (n = 39). Both treat-
and in doing so joins the family as a consultant and ments provided 20 therapy sessions over a 6-​month
sounding board, while decision-​making is left to period with assessments at four time points:  base-
the parents. This strategy facilitates parental own- line, mid-​treatment, end of treatment, and 6-​month
ership of decisions made in treatment and further follow-​up. There was no difference in adherence to
promotes their empowerment. treatment across FBT-​BN and SPT, with only 11%
The FBT-​BN differs from FBT-​AN in a num- of patients dropping out of therapy prematurely.
ber of key ways. In family treatment for BN, (1) the In terms of categorical outcomes, FBT-​BN dem-
emphasis is on regulating eating and curtailing onstrated a clinical and statistical advantage over
purging as opposed to weight restoration; (2) treat- SPT at the end of treatment as well as at 6-​month
ment follows an approach that supports a collab- follow-​up. At the end of treatment, significantly
orative effort between the adolescent and her/​his more patients in FBT-​ BN (39%) than in SPT
parents in addressing the ED, whereas in AN par- (18%) were binge and purge abstinent. Abstinence
ents take charge of weight restoration; (3) the secre- rates were not as high at 6-​ month follow-​ up;
tive nature, guilt, and shame typically associated however, significantly more patients in FBT-​ BN
with BN may make it more of a challenge for the (29%) were binge and purge free compared to SPT
family and therapist to remain symptom focused, (10%). Using random regression models, secondary

Le Grange, Rienecke 325

analyses of continuous outcome variables showed In the Schmidt and colleagues (2007) RCT,
greater improvements for FBT-​BN on behavioral family therapy (n = 41) was compared to cognitive-​
and attitudinal measures of ED psychopathology. behavioral therapy guided self-​care (n = 44) (CBT-​
Core bulimic symptoms also showed a more rapid GSC). Participants included adolescents and young
rate of improvement for FBT-​BN. Taken together, adults ages 12 to 20 years (mean age = 17.6 years)
these findings support the superiority of FBT-​BN meeting DSM-​IV criteria for BN or ED not oth-
over SPT in terms of the behavioral as well as attitu- erwise specified (EDNOS). In terms of categorical
dinal aspects of BN. outcomes, significantly more patients in CBT-​GSC
The Le Grange RCT also explored nonspecific were abstinent from binge eating at the end of
predictors, moderators, and mediators of outcome treatment compared with patients receiving family
(Ciao, Accurso, Fitzsimmons-​Craft, & Le Grange, therapy; however, this difference was no longer sig-
2015; Le Grange et  al., 2008; Lock, Le Grange, nificant at 6-​month follow-​up. There were no differ-
& Crosby, 2008). The clearest predictor to emerge ences in vomiting between the two treatment groups.
from these analyses was level of eating concern as Combining abstinence from binge eating and vom-
measured by the Eating Disorder Examination iting, there were no significant differences between
(EDE). That is, patients with lower scores on the family therapy (12.5%) and CBT-​GSC (19.4%)
EDE Eating Concern subscale at baseline were at the end of treatment or at 6-​month follow-​up
more likely to have remitted (abstinence from both (family therapy  =  41.4% vs. CBT-​GSC  =  36%).
binge eating and purging) at the end of treatment The only other differences reported were the direct
and at follow-​up, regardless of which treatment they cost of treatment, which was lower for CBT-​GSC.
received. Four EDE variables (Weight Concern, Schmidt and colleagues acknowledge that their
Shape Concern, Eating Concern, and Global sample size might have been too modest to detect
Score) significantly moderated the effects of treat- differences between two active treatments. Further,
ment on partial remission status (no longer meet- they state that without a waiting-​list or attention
ing study entry criteria). That is, partial remission placebo-​control group it would be difficult to rule
rates were much higher for FBT-​BN participants out that improvement was simply due to nonspe-
with low EDE scores. For participants receiving cific effects or the passage of time.
SPT, rates of partial remission were similar regard- Although no published manual is available for
less of EDE scores. As for mediators, changes in the family therapy used by Schmidt and her col-
the EDE Restraint subscale score at mid-​treatment leagues (2007), it appears to closely resemble FBT-​
may mediate outcome for FBT-​BN, but not for BN. One key difference is that “family” was defined
SPT (Lock et  al., 2008), suggesting that FBT-​BN in family therapy as any “close other,” rather than
may exert its effects in part by changing disordered restricting this definition to a parent or legal guard-
thinking. ian. Twenty-​five percent of all participants used a
A study examining predictors of psychologi- “close other” in their treatment. The rationale for
cal change (Ciao et  al., 2015) found that adoles- defining family in this way was likely due to the fact
cents taking psychotropic medications at baseline that the mean age of participants was at the upper
improved more quickly on the Eating Concerns end of adolescence (17.6 years), well above the age
subscale than adolescents not on medication. For of consent in the United Kingdom (16 years of age).
patients in SPT, older adolescents showed faster However, this definition of family might not be the
reductions on the Eating Concerns subscale than most effective way to approach family-​based treat-
younger adolescents. In addition, patients reporting ments with younger adolescents who are still legally
a higher purge frequency at baseline showed faster dependent on parents. Notwithstanding these
improvement on the Eating Concerns subscale if uncertainties, the abstinence rate for family therapy
they received FBT versus SPT. Finally, older ado- in Schmidt’s study was comparable to that achieved
lescents reported faster improvement in self-​esteem using FBT-​BN in Le Grange et al.’s 2007 study.
compared to younger adolescents. Generally, few In the largest RCT of adolescents with BN,
differences between treatments were found, suggest- 109 participants with DSM-​IV BN or partial BN
ing that FBT and SPT were both effective in bring- received either CBT adapted for adolescents (CBT-​
ing about improvement in psychological symptoms A) (n = 58) or FBT-​BN (n = 51) (Le Grange et al.,
for patients. These findings remain exploratory and 2015). Participants were between the ages of 12
a more detailed examination of these constructs and 18 with a mean age of 15.8 years and an aver-
awaits further testing in future controlled studies. age duration of illness of 18.9 months. Treatments

326 Family Therapy for Eating Disorders

consisted of 18 sessions over 6 months, and assess- adolescents with BN (Zaitsoff et al., 2008). Family
ments were conducted at baseline, end-​of-​treatment, therapy that empowers parents to play a significant
and 6-​and 12-​month follow-​up. The primary out- role in addressing their offspring’s ED is highly
come variable was abstinence from binge eating and demanding, in part because the adolescent is ini-
purging over the previous 4 weeks. Abstinence rates tially not allowed to make independent decisions
were significantly higher for patients receiving FBT-​ about her eating and weight-​ related behaviors,
BN at end of treatment (39.4%, versus 19.7% for and may be quite resistant to her parents’ efforts.
CBT-​A) and at 6-​month follow-​up (44.0%, versus Therefore, the question of how acceptable this treat-
25.4% for CBT-​A), although differences between ment is for both adolescents and parents is particu-
treatment groups were no longer significant at 12-​ larly salient.
month follow-​up (FBT = 48.5%; CBT = 32.0%). The initial report, a qualitative description of
The FBT-​BN was more effective in preventing hos- family therapy in a modest sample of adolescents
pitalization, with 2% requiring hospitalization dur- with AN (Le Grange & Gelman, 1998), supported
ing the study period, as opposed to 21% of patients the notion that this form of treatment was ulti-
receiving CBT-​A. An examination of moderators of mately acceptable for adolescents and their fami-
treatment found that participants with lower levels lies. A larger study of patient satisfaction in family
of family conflict did better in FBT-​BN than CBT-​ therapy for AN, employing both quantitative and
A. The study suggests that FBT-​BN brings about qualitative evaluations, provided additional empiri-
symptom improvement more rapidly than CBT-​A, cal support for this notion (Krautter & Lock, 2004).
although both treatments are viable options for this These authors found that adolescents and their par-
patient population. ents rated treatment effectiveness as well as thera-
Some questions require consideration when we peutic alliance quite highly. However, it should be
examine issues pertaining to the dissemination of noted that almost a third (30%) expressed a desire
family-​based treatments for this patient population. for individual therapy in addition to the family ther-
For instance, a treatment that involves the family apy they received. A  small study of FBT in Brazil
may not always be suitable, especially in older ado- found that the treatment approach was acceptable
lescents. For instance, 28% of eligible participants and feasible for families (Turkiewicz et al., 2010). In
in Schmidt’s study refused participation because adolescents with BN, therapeutic alliance and treat-
they did not want their families involved in treat- ment acceptability were high for both FBT-​BN and
ment. The CBT-​GSC appeared to present fewer SPT and did not differ between the two treatments
barriers, as fewer patients refused to participate. (Zaitsoff et al., 2008).
Further, patients fared as well in CBT-​GSC as they
did in family therapy (Schmidt’s study) or FBT-​BN Multifamily Therapy for Adolescent
(Le Grange’s studies). Delivering CBT-​GSC was Anorexia Nervosa
also more cost-​efficient than was the case for family Given the success of family-​based treatments for
therapy, which only serves to underscore the need adolescents with EDs, in conjunction with the need
for further evaluation of effective treatments given for more concentrated forms of interventions for
that treatment studies for adolescents with BN are those cases who do not respond to outpatient work,
still in their infancy. multiple-​family day treatment programs have been
Collectively these results support the use of FBT-​ developed in Dresden, Germany (Scholz & Asen,
BN as an effective intervention for adolescents who 2001) and in London, UK (Dare & Eisler, 2000).
are identified early in the course of their illness, Multiple-​family day treatment for EDs builds on
before the degree of psychopathology reaches lev- the effectiveness of treatment formats for fam-
els that might be less responsive to treatment. Thus ily intervention with other serious disorders (e.g.,
CBT is clearly a viable alternative treatment for this schizophrenia). It uses the same general principles
patient population. of parental empowerment while focusing only on
the specific problems related to AN as used in the
Acceptability of Family Therapy approach for single families described in the pre-
Three studies have examined the acceptabil- ceding text. Doing multiple-​family day treatment
ity of family therapy for adolescents with AN requires families to meet together for an extended
(Krautter & Lock, 2004; Le Grange & Gelman, weekend. During this time, a supportive commu-
1998; Turkiewicz, Pinzon, Lock, & Fleitlich-​Bilyk, nity is created that aims to absolve families of any
2010), and one study has been published regarding blame and provide opportunities to experiment with

Le Grange, Rienecke 327

behavioral change. Not only are expert consultants concerns, weight concerns, and amount of exercise
available but also this treatment format is an oppor- from pre-​to post-​treatment (Hollesen, Clausen, &
tunity to share experiences with other families that Rokkedal, 2013). In addition, participants reported
are confronted with similar challenges. Realizing improvement on the drive for thinness and intero-
that one’s struggles are quite similar to those of ceptive awareness subscales of the Eating Disorder
other families allows for an intensive learning envi- Inventory (EDI).
ronment under relatively controlled and support- Salaminiou, Campbell, Simic, Kuipers, & Eisler
ive conditions. After the initial extended weekend, (in press) conducted an open trial of 30 families
meetings over the ensuing months occur in a group participating in MFT. Patients had diagnoses of AN
format for a single day. The goal of these meetings (n = 27) or EDNOS (n = 3). Assessments were con-
is for families to help each other with the dilem- ducted at baseline and after 3 and 6  months into
mas that AN presents to their families. In practice, treatment. Patients’ weight improved significantly
single-​family sessions are also provided for families over time, as did scores on the Beck Depression
who participate in multiple-​family day treatment. Inventory, Rosenberg Self-​Esteem Scale, and sev-
Providing treatment in this way, multiple-​family eral subscales of the Eating Disorder Inventory.
day treatment may best be considered an attempt Mothers’ scores on the Beck Depression Inventory
to boost the efficacy of single-​family therapy for also improved. In addition, dropout rates were low
more resistant or challenging cases (Le Grange & (n  =  2), and adolescents reported their satisfac-
Eisler, 2009). tion with the treatment as moderate, whereas both
Preliminary results from the research groups in mothers and fathers rated their satisfaction as high.
London and Dresden are promising. As has been One barrier to receiving evidence-​based treat-
the case in the studies of family therapy for both ments for eating disorders is access to the limited
AN and BN, treatment retention has been high for number of providers and institutions that offer treat-
both sites. Feedback from parents and a majority of ments such as MFT. Intensive treatment options
patients (80%) in Dresden indicated that working offered over a short period of time may be a feasible
together with other families in a day hospital setting option for families who do not have local access to
was experienced as helpful and desirable (Scholz & specialized treatments. Rockwell, Boutelle, Trunko,
Asen, 2001). In particular, parents reported that the Jacobs, & Kaye (2011) reported on outcomes for
experience was helpful because of the collaborative 19 patients participating in a 5-​day intensive fam-
nature of the program and the opportunity to share ily therapy program (IFT). The IFT incorporates
ideas with other families about how to cope with FBT principles while also offering components of
their common predicament. A qualitative study of other forms of therapy, such as SyFT and CBT.
patients and parents receiving MFT found that the Participants successfully gained weight, and all but
majority felt it to be a positive experience (Voriadaki, one reported sustained weight gain at follow-​up,
Simic, Espie, & Eisler, 2015). Participants reported approximately 9 months after finishing IFT.
gaining better insight into the eating disorder. A larger study of 74 adolescents compared fami-
Adolescents reported enhanced motivation for lies receiving single-​family IFT (S-​IFT) (n  =  20)
recovery, and parents reported an improvement in to families receiving a multi-​family intensive fam-
self-​efficacy and family communication. ily therapy (M-​ IFT) (n  =  54) (Marzola et  al.,
Several uncontrolled studies have found 2015). Participants met criteria for AN (n  =  44)
improvements in weight and other measures of ED or EDNOS-​restricting type (n = 30) and were fol-
symptomatology. A retrospective chart review com- lowed up an average of 53.4 months after participat-
paring 25 patients who received multiple family ing in S-​IFT or 22.5 months after receiving M-​IFT.
therapy (MFT) and treatment as usual (TAU) to 25 The primary outcome variable was full remis-
patients receiving only TAU, found that patients in sion, defined as reaching at least 95% of expected
both groups successfully gained weight, but patients body weight (EBW), having scores on the Eating
receiving MFT reached a significantly higher % Disorder Examination-​Questionnaire (EDE-​Q)
IBW (99.6% IBW) compared with patients receiv- within 1 SD of community norms, and absence of
ing only TAU (95.4% IBW) (Gabel, Pinhas, Eisler, binge eating or purging behaviors in the previous
Katzman, & Heinmaa, 2014). A small study of 20 28  days. Partial remission was defined as reaching
adolescent females with AN (n = 8) or EDNOS-​AN at least 85% EBW, or being above 95% EBW but
(n = 12) receiving MFT found significant improve- with an elevated global score on the EDE-​Q and
ments in body mass index, restriction, eating the presence of binge eating or purging less than

328 Family Therapy for Eating Disorders

once per week. The majority of patients achieved patients was designed to assess the relative effec-
either full (n  =  45) or partial (n  =  20) remission, tiveness of three specific psychotherapies—​family
and 91.1% reported IFT to be useful. No differ- therapy, focal psychoanalytic psychotherapy, and
ences were found in treatment outcome for those cognitive analytic therapy—​versus routine care. At
who received S-​IFT versus M-​IFT. However, the the end of treatment, no differences in outcome
majority of patients received some form of treat- were reported for the three specific treatments.
ment after IFT, limiting the conclusions that can be However, patients in family therapy and focal psy-
drawn about the efficacy of the interventions. chotherapy showed modest symptomatic improve-
These results suggest that multiple-​ family ments that were superior to the control treatment.
day treatment is acceptable to families and a fea- Findings from this study were inconclusive perhaps
sible treatment for further study. Larger controlled in part because it was insufficiently powered to
research studies are needed to further investigate the detect differential therapeutic effects. Moreover, no
efficacy of MFT. treatment manuals were used.
Recently FBT has been adapted for use with
Family Therapy for Adults with Eating transition age youth (Chen et al., 2010; Chen et al.,
Disorders 2016; Dimitropoulos, Herschman, Toulany, &
Family-​
based treatment has recently been Steinegger, 2016; Dimitropoulos, Lock, Le Grange,
adapted for use with transition age youth (ages 17–​ & Anderson, 2015). The transition to financial inde-
26) and young adults with anorexia nervosa and pendence is taking longer than in previous genera-
bulimia nervosa. tions (Gutmann, Pullum-​Pinon, & Pullum, 2002),
leaving young persons dependent on their parents
Family Therapy for Adults with Anorexia for longer, and leaving parents to continue play-
Nervosa ing a large and influential role in their young adult
Compared to the adolescent literature, fam- children’s lives. Consequently, transition age youth
ily therapy for adults with AN has received much refers to young persons who may be experiencing
less attention. Only two published studies have significant life transitions such as individuating
tested the efficacy of family therapy for adults, from their family of origin, going away to college, or
both conducted at the Maudsley Hospital (Dare, transferring from pediatric to adult healthcare sys-
Eisler, Russell, Treasure, & Dodge, 2001; Russell tems. Thus, adapting FBT for transition age youth
et al., 1987). Russell and colleagues’ (1987) study, would appear to be a promising intervention for a
described in some detail earlier in this chapter, was patient population that is difficult to treat.
the first to investigate family therapy for adults In a case study of four patients with AN, par-
with AN. This was the first RCT of family therapy ticipants received 11–​20 treatment sessions over
involving adult AN patients (n = 36, mean age at 6 to 12  months (Chen et  al., 2010). Assessments
start of treatment  =  20.6  years). Participants were were conducted at baseline, end-​ of-​
treatment,
randomly assigned to either family therapy or a con- and 6-​ month follow-​ up. Participants generally
trol individual therapy at the time that they were showed improvement in body mass index, scores
discharged from the hospital. Unlike the findings on the EDE, global assessment of functioning, and
for adolescents with AN, family therapy showed the Beck Depression Inventory, and most found
no benefit over individual therapy for adults. In the treatment to be acceptable. An open trial
fact, in terms of weight gain, there was a trend in assessed FBT in 22 adult females with AN (n = 10)
favor of individual therapy for those patients with or EDNOS-​AN (n  =  12) between the ages of 18
an adult onset (mean age at onset = 24.6 years) as and 26 (Chen et  al., 2016). Participants received
opposed to those with an early onset (mean age at an average of 12 therapy sessions over 17 weeks,
onset = 14.3 years), although this trend had dissi- with assessments at baseline, mid-​treatment, end-​
pated at 5-​year follow-​up. However, at follow-​up of-​treatment, and 6-​and 12-​ month follow-​ up.
adult patients in individual therapy scored higher Improvements were found in body mass index,
in terms of psychological adjustment (based on EDE global score, eating-​ related obsessions and
Morgan/​Russell outcome criteria) compared with compulsions, global assessment of functioning, and
patients in family therapy (Eisler et al., 1997). presence of comorbid Axis I disorders.
The second study of family therapy for adults In another study, FBT has also been adapted
with AN was administered on an outpatient basis for use with transition age youth (FBT-​ TAY;
only (Dare et  al., 2001). This RCT of 84 adult Dimitropoulos et  al., 2015). Interviews with 15

Le Grange, Rienecke 329

transition age youths with EDs revealed that they be quite effective. Based on the data currently avail-
wanted continued support from their parents dur- able, more than two-​thirds of adolescent patients are
ing this transitional period, and expressed a desire successful in reaching a healthy weight by the end of
for that support to be of a collaborative nature treatment, and 80% will have further improved or
as they continue to grow more autonomous remained recovered 5 years later (Eisler et al., 1997;
(Dimitropoulos et al., 2016). Accordingly, adapta- Eisler et  al., 2007; Le Grange et  al., 2014; Lock,
tions made to FBT-​TAY include a longer course of Couturier, & Agras, 2006). Although data have
treatment (about 25 sessions), a more collaborative accumulated on the efficacy of FBT-​AN, the state
therapeutic approach, more individual time in ses- of research on the treatment for adolescents with
sions with the patient prior to bringing in the rest of BN has lagged behind. However, results from the
the family, and at times using more insight-​oriented first three published RCTs suggest that parents can
examples to explain the concept of externalization be helpful not only in restoring their child’s weight
of the illness from the patient. In addition, thera- but also in supporting their child in decreasing
pists direct the urgency of the situation toward binge eating and purging (Le Grange et al., 2007;
both the parents and the young adult rather than Le Grange et al., 2015; Schmidt et al., 2007). Taken
primarily toward the parents, as is done in FBT together, these results for adolescents with EDs are
with adolescents. Finally, therapists move patients encouraging, but must be interpreted cautiously
through phase 2 more quickly to provide them with as replication is needed with larger sample sizes.
more time to practice eating independently, and This task may now be more readily accomplished
spend more time on relapse prevention in phase 3, with the manualization of both FBT-​ AN (Lock
discussing with the patient how to prevent slips or & Le Grange, 2013) and FBT-​BN (Le Grange &
relapses when living independently (Dimitropoulos Lock, 2007).
et al., 2015). Another reason for cautious interpretation is the
Taking these studies together, it would be dif- fact that the same form of family therapy was not
ficult to draw a definitive conclusion about the effi- consistently used across the studies described in
cacy of family therapy for transition age youth and this chapter. Nevertheless, treatments that encour-
adults. Further studies are clearly required to estab- age parents to take an active role in helping their
lish whether family therapy can be helpful for this child recover from an ED, rather than observing
patient population, although adaptations of FBT from the sidelines, appear to be promising interven-
for transition age youth seems promising. tions for adolescents with a short duration of illness
who are medically suitable for outpatient treatment.
Family Therapy for Adults with Bulimia Treatment studies for transition age youth and
Nervosa adults with AN are sorely needed, as the disorder
Family therapy has received the least amount seems to become more resistant to treatment over
of attention among adults with BN. A few stud- time. The lack of effective treatments for older age
ies have described single cases of family therapy groups is especially alarming given the severe con-
for this age group (Madanas 1981; Roberto 1986; sequences of chronicity and the high mortality rate
Root, Fallon, & Friedrich, 1986; Wynne 1980), associated with AN (Crow et  al., 2009; Touyz &
and two larger studies have provided clear accounts Hay, 2015).
of this treatment (Russell et  al., 1987; Schwartz, Despite the still significant gaps in our knowl-
Barrett, & Saba, 1985). Findings from these edge, FBT is only one exciting example of the enor-
studies were inconclusive, and it remains unclear mous strides made in the field of ED treatment in
whether family therapy is helpful for this patient the last 20 years.
population.
Future Directions
Conclusion Enthusiasm for FBT ought to be tempered by
Despite a historical bias against the involvement the fact that there is a dearth of research on other
of parents in the treatment of adolescents with EDs treatment approaches for adolescents with an ED.
(Silverman, 1997), evidence in support of family For example, in AN there are only three published
interventions for AN has continued to mount over studies comparing either EOIT/​AFT or treatment
the past 40 years (Le Grange & Eisler, 2009). The as usual with family therapy, while cognitive, inter-
published controlled studies involving adolescents personal, and psychodynamic treatment approaches
with AN suggest that outpatient family therapy can have not been systematically evaluated.

330 Family Therapy for Eating Disorders

 The utility of individual procedural elements of family-​ based therapy for weight restoration in young
of FBT has not been sufficiently examined. adults with AN. Journal of Contemporary Psychotherapy, 40,
219–​224.
Dismantling studies can highlight which part(s) Chen, E. Y., Weissman, J. A., Zeffiro, T. A., Yiu, A., Eneva, K.
of FBT are necessary and which part(s) might be T., Arlt, J. M., & Swantek, M. J. (2016). Family-​based ther-
superfluous. For instance, the therapeutic value of apy for young adults with AN restores weight. International
the family meal that is typically implemented early Journal of Eating Disorders, 49, 701–​707.
on in this treatment has not been determined (c.f. Ciao, A. C., Accurso, E. C., Fitzsimmons-​Craft, E. E., & Le
Grange, D. (2015). Predictors and moderators of psycho-
Le Grange et al., 2016; Lock et al., 2015). Likewise, logical changes during the treatment of adolescent bulimia
the relative usefulness of phases two and three of nervosa. Behaviour Research and Therapy, 69, 48–​53.
FBT as opposed to the first phase of this treatment Couturier, J. L., & Kimber, M. S. (2015). Dissemination
is not known (Lock et al., 2005). and implementation of manualized family-​ based treat-
Matching patients and treatment modality is ment: A systematic review. Eating Disorders, 23, 281–​290.
Crow, S. J., Peterson, C. B., Swanson, S. A., Raymond, N. C.,
another challenge that requires attention. Although Specker, S., Eckert, E. D., & Mitchell, J. E. (2009). Increased
some advances in terms of treatment modera- mortality in bulimia nervosa and other eating disorders.
tors have been noted (Le Grange et  al., 2008; Le American Journal of Psychiatry, 166, 1342–​6134.
Grange et al., 2012; Le Grange et al., 2016), there Dare, C. (1983). Family therapy for families containing an ano-
is little to guide clinicians who are attempting to rectic youngster. Columbus, OH, IVth Ross Conference on
Medical Research, Ross Laboratories: 28–​37.
determine the appropriateness of FBT for one fam- Dare, C., & Eisler, I. (1997). Family therapy for AN. In D. M.
ily compared to another. Even less is known about Garner & P. Garfinkel (Eds.), Handbook of treatment for eat-
the mechanisms of change in FBT (c.f. Byrne, ing disorders (pp. 307–​324). New York, NY: Guilford Press.
Accurso, Arnow, Lock, & Le Grange, 2015; Lock Dare, C., & Eisler, I. (2000). A multi-​family group day treatment
et al., 2015). programme for adolescent eating disorders. European Eating
Disorders Review, 8, 4–​18.
Finally, the uptake and implementation of Dare, C., Eisler, I., Russell, G., Treasure, J., & Dodge, L. (2001).
manualized treatments among many clinicians in Psychological therapies for adults with AN:  Randomized
the community are often less than satisfactory (c.f. controlled trial of outpatient treatments. British Journal of
Accurso et al., under review; Couturier & Kimber, Psychiatry, 178, 216–​221.
2015). However, the development of published cli- Dare, C., Le Grange, D., Eisler, I., & Rutherford, J. (1994).
Redefining the psychosomatic family:  Family process of
nician manuals for both FBT-​AN and FBT-​BN pro- 26 eating disorder families. International Journal of Eating
vides an opportunity for the examination of effective Disorders, 16, 211–​226.
dissemination of these treatment modalities. Dimitropoulos, G., Freeman, V. E., Allemang, B., Couturier, J.,
McVey, G., Lock, J., & Le Grange, D. (2015). Family-​based
Note treatment with transition age youth with AN: A qualitative
1. The Maudsley group has recently published a detailed man- summary of application in clinical practice. Journal of Eating
ual describing family therapy for adolescent AN: http://​www. Disorders, 3. doi: 10.1186/​s40337-​015-​0037-​3
national.slam.nhs.uk/​services/​camhs/​camhs-​eatingdisorders/​ Dimitropoulos, G., Herschman, J., Toulany, A., & Steinegger,
resources/​ C. (2016). A qualitative study on the challenges associated
with accepting familial support from the perspective of
transition-​age youth with eating disorders. Eating Disorders,
References 24, 255–​270.
Agras, W. S., Lock, J., Brandt, H., Bryson, S. W., Dodge, E., Dimitropoulos, G., Lock, J., Le Grange, D., & Anderson, K.
Halmi, K. A.,  . . .  Woodside B. (2014). Comparison of 2 (2015). Family therapy for transition youth. In K. L. Loeb,
family therapies for adolescent anorexia nervosa: a random- D. Le Grange, & J. Lock (Eds.), Family therapy for adolescent
ized parallel trial. JAMA Psychiatry, 71, 1279–​1286. eating and weight disorders: New applications (pp. 230–​255).
Bara-​Carril, N., & Nielsen, S. (2003). Family, burden of care and New York, NY: Routledge.
social consequences. In J. L. Treasure, U. Schmidt, & E. van Dodge, E., Hodes, M., Eisler, I., & Dare, C. (1995). Family
Furth (Eds.), Handbook of eating disorders (pp. 191–​206). therapy for bulimia nervosa in adolescents:  An exploratory
Chichester, UK: Wiley. study. Journal of Family Therapy, 17, 59–​77.
Byrne, C. E., Accurso, E. C., Arnow, K. D., Lock, J., & Le Grange, Eisler, I. (2005). The empirical and theoretical base of fam-
D. (2015). An exploratory examination of patient and paren- ily therapy and multiple family day therapy for adolescent
tal self-​efficacy as predictors of weight gain in adolescent AN. anorexia nervosa. Journal of Family Therapy, 27, 104–​131.
International Journal of Eating Disorders, 48, 883–​888. Eisler, I., Dare, C., Hodes, M., Russell, G., Dodge, E., & Le
Charcot, J. M., (1889). Lecture XVII: Isolation in the treatment of Grange, D. (2000). Family therapy for adolescent anorexia
hysteria. Clinical lectures on the diseases of the nervous system deliv- nervosa: The results of a controlled comparison of two family
ered at the infirmary of la Salpetriere by Professor J. M. Charcot. interventions. Journal of Child Psychology and Psychiatry, 41,
(T. Savill, Trans.). London, UK: New Sydenham Society. 727–​736.
Chen, E. Y., Le Grange, D., Doyle, A. C., Zaitsoff, S., Doyle, Eisler, I., Dare, C., Russell, G. F. M., Szmukler, G. I., Le Grange,
P., Roehrig, J. P., & Washington, B. (2010). A case series D., & Dodge, E. (1997). Family and individual therapy in

Le Grange, Rienecke 331

 anorexia nervosa:  A  five-​year follow-​up. Archives of General Le Grange, D., Lock, J., Accurso, E., Agras, S., Darcy, A.,
Psychiatry, 54, 1025–​1030. Forsberg, S., & Bryson, S. (2014). Relapse from remission
Eisler, I., Simic, M., Russell, G. F. M., & Dare, C. (2007). A at four-​year follow-​up in two treatments for adolescent AN.
randomized controlled treatment trial of two forms of Journal of the American Academy for Child and Adolescent
family therapy in adolescent anorexia nervosa:  A  five-​year Psychiatry, 53, 1162–​1167.
follow-​up. Journal of Child Psychology and Psychiatry, 48, Le Grange, D., Lock, J., Agras, W. S., Bryson, S. W., & Jo, B.
552–​560. (2015). Randomized clinical trial of family-​based treatment
Gabel, K., Pinhas, L., Eisler, I., Katzman, D., & Heinmaa, M. and cognitive-​behavioral therapy for adolescent bulimia ner-
(2014). The effect of multiple family therapy on weight gain vosa. Journal of the American Academy of Child and Adolescent
in adolescents with anorexia nervosa:  Pilot data. Journal of Psychiatry, 54, 886–​894.
the Canadian Academy of Child and Adolescent Psychiatry, 23, Le Grange, D., Lock, J., Accurso, E. C., Agras, W. S., Darcy, A.,
196–​199. Forsberg, S., & Bryson, S. W. (2014). Relapse from remission
Godart, N., Berthoz, S., Curt, F., Perdereau, F., Rein, Z., Wallier, at two-​to four-​year follow-​up in two treatments for ado-
J.,  . . .  Jeammet, P. (2012). A randomized controlled trial of lescent anorexia nervosa. Journal of the American Academy of
adjunctive family therapy and treatment as usual following Child and Adolescent Psychiatry, 53, 1162–​1167.
inpatient treatment for anorexia nervosa adolescents. PloS Le Grange, D., Lock, J., Agras, W. S., Moye, A., Bryson, S., Jo,
ONE, 7, e28249. doi:10.1371/​journal.pone.0028249 B., & Kraemer, H. (2012). Moderators and mediators of
Gull, W. (1874). Anorexia nervosa (apepsia hysterica, anorexia remission in family-​based treatment and adolescent focused
hysterica). Transactions of the Clinical Society of London, 7, therapy for anorexia nervosa. Behavior Research and Therapy,
222–​228. 50, 85–​92.
Gutmann, M. P., Pullum-​Pinon, S. M., & Pullum, T. W. (2002). Le Grange, D., Lock, J., & Dymek, M. (2003). Family-​based
Three eras of young adults home leaving in twentieth-​century therapy for adolescents with bulimia nervosa. American
America. Journal of Social History, 35, 533–​576. Journal of Psychotherapy, 67, 237–​251.
Hollesen, A., Clausen, L., & Rokkedal, K. (2013). Multiple fam- Lock, J., Agras, W. S., Bryson, S., & Kraemer, H. C. (2005). A
ily therapy for adolescents with anorexia nervosa:  A  pilot comparison of short-​and long-​term family therapy for ado-
study of eating disorder symptoms and interpersonal func- lescent anorexia nervosa. Journal of the American Academy of
tioning. Journal of Family Therapy, 35, 53–​67. Child and Adolescent Psychiatry, 44, 632–​639.
Humphrey, L. (1989). Observed family interactions among sub- Lock, J., Couturier, J., & Agras, W. S. (2006). Comparison of
types of eating disorders using structural analysis of social long term outcomes in adolescents treated with family ther-
behavior. Journal of Consulting and Clinical Psychology, 57, apy. Journal of the American Academy of Child and Adolescent
206–​214. Psychiatry, 45, 666–​672.
Krautter, T., & Lock, J. (2004). Is manualized family-​based treat- Lock, J., & Le Grange, D. (2013). Treatment manual for anorexia
ment for adolescent anorexia nervosa acceptable to patients? nervosa:  A  family-​based approach (2nd ed.). New  York,
Patient satisfaction at end of treatment. Journal of Family NY: Guilford Press.
Therapy, 26, 65–​81. Lock, J., Le Grange, D., Agras, W. S., Moye, A., Bryson, S. W.,
Lasegue, C. (1883). De l’anorexie hysterique. Archives Generales & Jo, B. (2010). Randomized clinical trial comparing family-​
de Medecine, 21, 384–​403. based treatment with adolescent-​focused individual therapy
Le Grange, D., Crosby, R., & Lock, J. (2008). Predictors and for adolescents with anorexia nervosa. Archives of General
moderators of outcome in family-​based treatment for ado- Psychiatry, 67, 1025–​1032.
lescent bulimia nervosa. Journal of the American Academy of Lock, J., Le Grange, D., & Crosby, R. (2008). Exploring mecha-
Child and Adolescent Psychiatry, 47, 464–​470. nisms of change in family-​based treatment for adolescent
Le Grange, D., Crosby, R. D., Rathouz, P. J., & Leventhal, B. bulimia nervosa. Journal of Family Therapy, 30, 260–​271.
L. (2007). A randomized controlled comparison of family-​ Lock, J., Le Grange, D., Agras, W. S., Fitzpatrick, K. K., Jo, B.,
based treatment and supportive psychotherapy for ado- Accurso, E.,   .  .  .  Stainer, M. (2015). Can adaptive treat-
lescent bulimia nervosa. Archives of General Psychiatry, 64, ment improve outcomes in family-​based therapy for adoles-
1049–​1056. cent anorexia nervosa? Feasibility and treatment effects of a
Le Grange, D., & Eisler, I. (2009). Family interventions in ado- multi-​site treatment study. Behavior Research and Therapy,
lescent anorexia nervosa. Child and Adolescent Psychiatric 73, 90–​95.
Clinics of North America, 18, 159–​173. Madanas, C. (1981). Strategic family therapy. San Francisco,
Le Grange, D., Eisler, I., Dare, C., & Russell, G. F. (1992). CA: Jossey-​Bass.
Evaluation of family treatments in adolescent anorexia ner- Madden S., Miskovic-​Wheatley, J., Wallis, A., Kohn, M., Lock,
vosa: A pilot study. International Journal of Eating Disorders, J., Le Grange, D.,  . . .  Touyz, S. (2015). A randomized con-
12, 347–​357. trolled trial of in-​patient treatment for anorexia nervosa in
Le Grange, D., & Gelman, T. (1998). The patient’s perspective medically unstable adolescents. Psychological Medicine, 45,
of treatment in eating disorders: A preliminary study. South 415–​427.
African Journal of Psychology, 28, 182–​186. Marzola, E., Knatz, S., Murray, S. B., Rockwell, R., Boutelle, K.,
Le Grange, D., Hughes, E. K., Court, A., Yeo, M., Crosby, R. Eisler, I., & Kaye, W. H. (2015). Short-​term intensive family
D., & Sawyer, S. M. (2016). Randomized clinical trial of therapy for adolescent eating disorders: 30-​month outcome.
parent-​ focused treatment and family-​ based treatment for European Eating Disorders Review, 23, 210–​218.
adolescent anorexia nervosa. Journal of the American Academy Minuchin, S., Baker, B. L., Rosman, B. L., Liebman, R., Milman,
of Child & Adolescent Psychiatry, 55, 683–​692. L., & Todd, T. C. (1975). A conceptual model of psycho-
Le Grange, D., & Lock, J. (2007). Treating bulimia in adoles- somatic illness in children:  Family organization and family
cents: A family-​based approach. New York, NY: Guilford Press. therapy. Archives of General Psychiatry, 32, 1031–​1038.

332 Family Therapy for Eating Disorders

Minuchin, S., Rosman, B. L., & Baker, B. L. (1978). nervosa and bulimia. New  York, NY:  Guilford Press,
Psychosomatic families:  Anorexia nervosa in context. 280–​310.
Cambridge, MA: Harvard University Press. Selvini Palazzoli, M. (1974). Self-​starvation: From the intrapsychic
Roberto, L. (1986). Bulimia: The transgenerational view. Journal to the transpersonal approach. Oxford, UK: Chaucer.
of Marital and Family Therapy, 12, 231–​240. Silverman, J. (1997). Anorexia nervosa: Historical perspective on
Robin, A., Siegal, P., & Moye, A. (1995). Family versus individual treatment. In D. Garner & P. Garfinkel (Eds.), Handbook of
therapy for anorexia: Impact on family conflict. International treatment for eating disorders (2nd ed., pp. 3–​10). New York,
Journal of Eating Disorders, 17, 313–​322. NY: Guilford Press.
Robin, A., Siegal, P., Gilroy, M., Dennis, A., & Sikand, A. (1999). Steinglass, P. (1998). Multiple family discussion groups for
A controlled comparison of family versus individual therapy patients with chronic medical illness. Families, Systems, and
for adolescents with anorexia nervosa. Journal of the American Health, 16, 55–​70.
Academy of Child and Adolescent Psychiatry, 38, 1482–​1489. Steinglass, P., & Horan, M. (1988). Families and chronic
Rockwell, R. E., Boutelle, K., Trunko, M. E., Jacobs, M. J., & medical illness. Journal of Psychotherapy and the Family, 3,
Kaye, W. H. (2011). An innovative short-​term, intensive, 127–​142.
family-​based treatment for adolescent anorexia nervosa: Case Touyz, S., & Hay, P. (2015). Severe and enduring anorexia ner-
series. European Eating Disorders Review, 19, 362–​367. vosa (SE-​AN): In search of a new paradigm. Journal of Eating
Root, M. P. P., Fallon, P., & Friedrich, W. N. (1986). Bulimia: A sys- Disorders, 3, 26.
tems approach to treatment. New York, NY: Norton. Turkiewicz, G., Pinzon, V., Lock, J., & Fleitlich-​Bilyk, B. (2010).
Russell, G. F., Szmukler, G. I., Dare, C., & Eisler, I. (1987). An Feasibility, acceptability, and effectiveness of family-​based
evaluation of family therapy in anorexia nervosa and buli- treatment for adolescent anorexia nervosa: An observational
mia nervosa. Archives of General Psychiatry, 44, 1047–​1056. study conducted in Brazil. Revista Brasileira de Psiquiatria,
Salaminiou, E., Campbell, M., Simic, M., Kuipers, E., & Eisler, 32, 169–​172.
I. (in press). Intensive multi-​family therapy for adolescent Voriadaki, T., Simic, M., Espie, J., & Eisler, I. (2015).
anorexia nervosa:  An open study of 30 families. Journal of Intensive multi-​family therapy for adolescent anorexia ner-
Family Therapy. vosa:  Adolescents’ and parents’ day-​to-​day experiences.
Schmidt, U., Lee, S., Beecham, J., Perkins, S., Treasure, J., Yi, Journal of Family Therapy, 37, 5–​23.
I.,  . . .  Eisler, I. (2007). A randomized controlled trial of Wynne, L. (1980). Paradoxical interventions:  Leverage for
family therapy and cognitive behavior therapy guided self-​ therapeutic change in individual and family systems. In
care for adolescents with bulimia nervosa and related disor- T. Strauss, S. Bowers, S. Downey, S. Fleck, & I. Levin
ders. American Journal of Psychiatry, 164, 591–​598. (Eds.), The Psychotherapy of Schizophrenia. New  York,
Scholz, M., & Asen, K. E. (2001). Multiple family therapy with NY: Plenum Press.
eating disordered adolescents. European Eating Disorders Zaitsoff, S., Doyle, A. C., Hoste, R. R., Le Grange, D.
Review, 9, 33–​42. (2008). How do adolescents with bulimia nervosa rate the
Schwartz, R., Barrett, M., & Saba, G. (1985). Family ther- acceptability and therapeutic relationship in family-​based
apy for bulimia. Handbook of psychotherapy for anorexia treatment? International Journal of Eating Disorders, 41,
390–​398.

Le Grange, Rienecke 333

 C H A PT E R


Dialectical Behavior Therapy and
17 Emotion-​Focused Therapies for
Eating Disorders

Eunice Y. Chen, Angelina Yiu, and Debra L. Safer

Abstract
This chapter provides a description and review of the research evidence for the adaptation of dialectical
behavior therapy (DBT) and emotion-​focused therapies for eating disorders (EDs). First, the chapter
briefly describes the standard DBT program as originally developed for women with borderline
personality disorder (BPD) and the evidence for standard DBT with BPD. Second, the rationale for
the adaptation of DBT for EDs is given and preliminary evidence for DBT and other emotion-​focused
treatments is described. Finally, the Stanford DBT program for EDs is outlined. Given the promise
of alternative treatments for eating disorders, further development, adaptation, and testing of
transdiagnostic emotion regulation treatments is warranted.
Key Words:  borderline personality disorder, dialectical behavior therapy, eating disorder, trial, evidence

Introduction the randomized controlled trial evidence of DBT for

Standard dialectical behavior therapy (DBT) is BPD; (3) present the rationale for the adaptation of
an outpatient cognitive-​ behavioral therapy origi- DBT for EDs; (4) review the randomized controlled
nally developed for women with borderline person- trial evidence for DBT and other emotion-​focused
ality disorder (BPD), who frequently struggle with therapies for EDs; (5) describe the DBT model of
emotion dysregulation and recurrent suicidal behav- maintenance and etiology of EDs; (6)  outline the
ior. Dialectical behavior therapy is a comprehensive DBT program as adapted for binge eating disor-
skills-​based treatment and emphasizes the integra- der (BED) and bulimia nervosa (BN); and (7) offer
tion of opposing principles (the dialectic) such as conclusions and future research directions.
behavior change (problem solving, contingency
management, exposure-​based procedures, cognitive Standard Dialectical Behavior Therapy
modification) with acceptance-​based practices such Treatment
as Zen and contemplation practice (such as mind- According to standard DBT, BPD is conceptu-
fulness and validation). The dialectical framework alized as a disorder of pervasive emotion dysregu-
within DBT stresses wholeness, interrelatedness, lation. The biosocial theory is used to describe its
and process. Over time, standard DBT has been etiology and maintenance, which postulates that
adapted to address a variety of problem behaviors BPD behaviors (e.g., intense and labile emotions,
associated with emotion dysregulation, including nonsuicidal self-​injury, self-​
damaging behaviors,
eating disorders (EDs). suicidal behaviors) develop due to the transaction
The aims of this chapter are to (1) briefly review over time between a biological vulnerability to
standard DBT as originally developed for women emotion dysregulation and an environment expe-
with BPD, including its biosocial theory; (2) review rienced as emotionally invalidating. Emotional

334
vulnerability refers to high negative affect at base- agenda, dictates the content of individual sessions.
line, heightened sensitivity to emotional stimuli, For BPD, the Stage I target hierarchy is to (1) cease
intense emotional responses, and a slow return life-​threatening behaviors (e.g., suicide attempts,
to emotional baseline. An environment is labeled increase in suicide ideation, nonsuicidal self-​injurious
invalidating if it (1) indiscriminately rejects an indi- behaviors, homicidal threats and behaviors); (2) cease
vidual’s communication of personal experience, therapy-​interfering behaviors (e.g., missed sessions);
particularly emotions; (2) intermittently reinforces (3)  cease quality-​of-​life-​interfering behaviors (e.g.,
an escalation of emotions; and (3) oversimplifies EDs or other Axis I  disorders, homelessness); and
problem solving and meeting goals. These charac- (4) increase behavioral skills.
teristics lead individuals to have difficulties validat- Standard DBT functions to enhance a client’s
ing their own internal experiences and to search use of skillful behavior both within and outside of
the environment for ways to respond. As such, therapy sessions, as well as both client and therapist’s
individuals may often react with extreme oscilla- motivation to engage with and deliver the treatment.
tions between emotional inhibition and intense This is accomplished by reducing reinforcement for
responses in order to communicate private experi- dysfunctional or ineffective client behavior (which
ence, as well as forming unrealistic goals. The per- may include restructuring the environment to sup-
vasive emotion dysregulation that results includes port progress and adaptive change), and generaliz-
several difficulties: inhibiting mood-​ dependent ing behavior from the therapy setting to the natural
behaviors that may be inappropriate or impulsive; environment. To fulfill these functions, the modes
organizing behavior in the service of goals, inde- of treatment in standard DBT involve (1)  weekly
pendent of current mood; up-​or down-​regulating individual psychotherapy; (2)  weekly group skills
physiological arousal as needed; diverting atten- training; (3)  24-​hour telephone consultation; and
tion away from emotionally evocative stimuli; and/​ (4) a weekly therapist consultation team.
or experiencing emotion without avoidance or an Each mode involves a different hierarchy of
extreme secondary negative emotion. targets. The individual psychotherapist is respon-
The affect dysregulation model within the bio- sible for the assessment and problem solving of
social theory posits that suicidal and self-​injurious skill deficit and motivational problems, as well as
behaviors function to reduce painful, intolerable organizing other treatment modalities in service
emotional states in individuals who lack adaptive of these goals. Group skills training targets the
skills to modulate emotions. Engagement in sui- acquisition of new behavioral skills in a structured
cidal and self-​injurious behavior may bring tem- psychoeducational format and includes four mod-
porary relief, but such behavior typically leads to ules:  Mindfulness, Distress Tolerance, Emotion
more distressing emotions or other negative con- Regulation, and Interpersonal Effectiveness skills.
sequences. For example, secondary emotions, such Clients are instructed to use telephone consultation
as shame, can arise from engagement in suicidal or for coaching to generalize the use of skills to the
self-​injurious behavior. Intolerance of painful emo- natural environment, decrease suicidal behaviors
tions then causes the cycle of maladaptive emotion and nonsuicidal self-​injury, and to repair possible
regulation behaviors to repeat itself. ruptures to the client–​therapist relationship. Finally,
Standard DBT treatment (Linehan, 1993a; the therapist consultation team enhances therapist
Linehan, 1993b; Linehan, 2014) is organized motivation and skills and manages problems that
around the client’s level of severity and chronic- arise in the delivery of DBT.
ity, with different treatment stages associated with In addition to the focus on treatment hierarchies
particular treatment goals. The DBT stages involve that dictates the structure of individual sessions,
(1)  Pretreatment, orientation and commitment to dialectical strategies emphasize the balance between
treatment; (2) Stage I, stopping out-​of-​control behav- change and acceptance for each treatment strategy
iors; (3) Stage II, replacing “quiet desperation” with in standard DBT (e.g., core strategies, stylistic strat-
nontraumatic emotional experiencing; (4) Stage III, egies, and case management strategies). Dialectical
reducing ongoing disorders and problems in living; strategies highlight dichotomous relationships, such
and (5) Stage IV, resolving a sense of incompleteness as feelings/​beliefs versus wise mind and good ver-
to achieve freedom. Each stage of treatment is asso- sus bad, and assist clients in finding balanced and
ciated with a target hierarchy. Dialectical behavior synthesized responses. Dialectical strategies include
therapy is unique from other therapies, as an over- the use of metaphors, stories, paradox, playing
all target hierarchy, rather than a prescribed session devil’s advocate, fluctuating between ambiguity and

Chen, Yiu, Safer 335

certainty, using cognitive restructuring, highlight- powerless, with the therapist acting directly on the
ing continual change, and validating a client’s intui- client’s behalf. Consultation-​to-​the-​therapist strat-
tive wisdom. egy involves individual therapists seeking consulta-
The core strategies balance acceptance (e.g., valida- tion within their team for support and assistance in
tion) and behavioral change (e.g., problem-​solving) delivering DBT effectively.
strategies. The latter include chain analyses, which The preceding is a brief description of the stan-
meticulously examine the topography, intensity, fre- dard DBT program for women with BPD. A more
quency, duration, situation, antecedents, and con- exhaustive description of Stage I DBT for BPD can
sequences of a problem behavior. In standard DBT, be found in Linehan’s manuals (Linehan, 1993a,
chains are typically conducted during individual 1993b, 2014). As discussed later, standard DBT
psychotherapy. Repeated chain analyses of a prob- has been adapted to address a variety of problem
lem behavior provide the client and therapist with behaviors associated with emotion dysregulation,
information on the cues, maintaining factors and including ED behavior, as described in the text that
function of a problematic behavior. This enables the follows.
client and therapist to identify what prevented the
client from being effective in a situation, and then Efficacy of Dialectical Behavior Therapy
to teach, role-​play, rehearse new skills (e.g., emotion with Borderline Personality Disorder
regulation skills) or use cognitive modification or Standard DBT is currently the most empirically
mindfulness to address faulty cognitions or clarify validated treatment for BPD and is generally con-
environmental contingencies. Having established a sidered the treatment-​of-​choice for individuals with
solution with the client and a plan to prevent future BPD (Lieb, Zanarini, Schmahl, Linehan, & Bohus,
problem behavior, the therapist assesses the client’s 2004). Nine randomized controlled trials provide
commitment using a variety of commitment strate- evidence for its efficacy ([1]‌Linehan, Armstrong,
gies such as evaluating pros and cons, playing devil’s Suarez, Allmon, & Heard, 1991, with follow-​up
advocate, and using “foot in the door/​door in the reported in Linehan, Heard, & Armstrong, 1993;
face” strategies. The client and therapist then trou- Linehan, Tutek, Heard, & Armstrong 1994; [2]
bleshoot this plan, and search for commitment to Linehan et al., 1999; [3] Linehan et al., 2002; [4]
the revised plan. Turner, 2000; [5] Koons et  al., 2001; [6] Verheul
In every DBT encounter with a client, change et al., 2003, van den Bosch, Verheul, Schippers, &
strategies are balanced with acceptance strategies van den Brink, 2002; [7] Linehan et al., 2006; [8]
(e.g., validation) to build and maintain a strong McMain et  al., 2009; [9] Carter, Willcox, Lewin,
therapeutic relationship. Validation strategies range Conrad, & Bendit, 2010). Apart from trials con-
on a continuum from listening in an interested ducted by the developer, five independent research
fashion to being radically genuine, that is, treating teams conducted these trials (Carter et  al., 2010;
the client as one would treat an equal, a sister, or a Koons et  al., 2001; McMain et  al., 2009; Turner,
friend. 2000; van den Bosch et al., 2002) in three countries
Stylistic strategies also balance acceptance and other than the United States—​ the Netherlands,
change. Reciprocal communication, on the one Canada, and Australia ( Carter et al., 2010 McMain
hand, involves interpersonal warmth, responsiveness et al., 2009; van den Bosch et al., 2002).
to a client’s concerns and strategic self-​disclosure, Standard DBT has been compared to treatment-​
which models skills application as well as nonjudg- as-​usual (Carter et  al., 2010; Koons et  al, 2001;
mental sharing of one’s own vulnerabilities. On the Linehan et  al., 1991, 1999; Verheul et  al., 2003),
other hand, irreverent communication involves an treatment-​by-​experts (TBE; Linehan et  al., 2006;
outrageous, humorous, or blunt style and is used McMain et  al., 2009), comprehensive validation
when therapy becomes polarized and a therapist (Linehan et al., 2002), and client-​centered therapy
and client become deadlocked. (Turner, 2000). Compared to its control condi-
Case management strategies include “consultation-​ tion, DBT showed significantly greater reductions
to-​the-​client,” “environmental interventions,” and in suicide attempts, intentional self-​injury, and sui-
“consultation-​to-​the-​therapist.” The first teaches cidal ideation (Koons et  al., 2001; Linehan et  al.,
clients how to skillfully interact with the environ- 1991, 1999, 2002, 2006; Turner, 2000; Verheul
ment rather than organizing the environment to et  al., 2003). Standard DBT also resulted in sig-
meet their needs. Environmental interventions are nificantly less treatment dropout than control treat-
used when a client is in immediate danger or is ments (Linehan et al., 1991, 1999, 2006; Verheul

336 Dialectical Behavior Therapy

et  al., 2003) and clients receiving DBT were less 2003; Lynch et  al., 2007), treatment-​ resistant
likely to use inpatient (Linehan et al., 1991, 2006; depressed individuals (Harley, Sprich, Safren,
Turner, 2000) or emergency room services (Linehan Jacobo, & Fava, 2008), adults with bipolar disorder
et  al., 2006). Treatment with DBT demonstrated I or II (van Dijk, Jeffrey, & Katz, 2013). Controlled
improvements in secondary outcome variables trials using DBT have also been conducted, such as
(Linehan et  al., 2002), such as depressed mood with suicidal adolescents with BPD features (Rathus
(Linehan et al., 2006), with some studies showing & Miller, 2002).
that DBT significantly improved these variables
compared to control conditions (Koons et al., 2001; Why Adapt Dialectical Behavior Therapy
Linehan et al., 1991, 1999; Turner, 2000). for Eating Disorders?
Two large randomized controlled trials com- Dialectical behavior therapy offers an alterna-
pared DBT to TBE or general psychiatric manage- tive for difficult-​to-​treat clients for whom existing
ment among suicidal BPD participants (Linehan treatments have failed. Currently the most empiri-
et  al., 2006; McMain et  al., 2009). The first trial cally tested treatments for EDs, cognitive-​behavior
(Linehan et al., 2006) controlled for therapist avail- therapy (CBT) and interpersonal psychotherapy
ability, expertise, allegiance, gender, training and (IPT), result in about 50% of BN and BED patients
experience, consultation availability, and institu- (Fairburn & Brownell, 2001; Keel & Brown, 2010)
tional prestige between treatment arms. After treat- remaining symptomatic after treatment. Predictors
ment and 1 year of follow-​up, DBT clients were half of poor outcome in CBT for EDs may include co-​
as likely to make a suicide attempt (23% vs. 46% of occurring Axis I and II disorders or symptoms such
clients), were less likely to be to be hospitalized for as BPD and depressive symptoms (Grilo, Masheb,
suicidal ideation, were less likely to drop out (25% & Wilson, 2001; Stice & Agras, 1999; Wilfley
vs. 59%), and had lower medical risk across all sui- et al., 2000). Standard DBT thus represents a viable
cide attempts and self-​injurious acts. Overall, DBT option for ED clients with co-​occurring psychopa-
resulted in significant reductions of suicidal and thology or for whom existing treatments have failed,
self-​injurious behaviors compared to the control due to its efficacy for individuals with multiple and
condition. However, in a second trial conducted “difficult-​to-​treat” disorders.
in Canada that compared DBT with general psy- As described briefly, DBT is uniquely based on
chiatric management, it was found that both treat- a broad affect regulation model. In other words,
ments were equivalent in reducing suicidal and the precursors of binge eating are understood to
self-​injurious behaviors, use of healthcare services, be the result of affect rather than due directly to
and improvements of BPD symptoms, with no sig- dietary restraint and weight and shape concerns (as
nificant differences in dropout rates (McMain et al., in CBT) or as a result of difficulties with resolving
2009). Treatment in both treatment arms was deliv- interpersonal problems (as in IPT). Although this
ered by clinicians with expertise in treating BPD affect regulation model does not preclude the role of
and clinicians delivering DBT were adherent to weight and shape concerns, interpersonal difficul-
the treatment. Findings from both large, random- ties, thoughts about food or the self, or perfectionis-
ized controlled trials suggest that standard DBT tic thinking in triggering affective states, this model
is efficacious for individuals with suicidal BPD in parsimoniously focuses on affect as an important
reducing suicidal ideation, suicide attempts, and trigger leading to problematic ED behavior. The
use of healthcare services and retaining individu- affect regulation model views binge eating and other
als in treatment (Linehan et  al., 2006; McMain types of ED behavior (e.g., vomiting) as the means
et al., 2009). (albeit maladaptive) by which individuals regulate
In addition to the efficacy of DBT for BPD, emotions. Because standard DBT treatment is spe-
evidence suggests that DBT is efficacious for BPD cifically designed to teach adaptive affect regulation
with co-​ occurring disorders, such as substance skills and to target behaviors resulting from emo-
abuse (Linehan, Schmidt, Dimeff, Craft, Kanter, & tion dysregulation, a theoretical rationale exists for
Comtois, 1999; van den Bosch, Verheul, Schippers, applying DBT to treating EDs.
& van den Brink, 2002), and DBT has also been By using the biosocial theory, DBT also offers
adapted for a variety of populations not specifically an etiological theory for the development of ED
selected for BPD (Dimeff & Koerner, 2007). These behaviors over time. This theory describes the areas
include randomized controlled trials for depressed needed for change but also encourages an attitude
older adults (Lynch, Morse, Mendelson, & Robins, of effective compassion in the therapist. This theory

Chen, Yiu, Safer 337

is also validating to clients and serves to reduce Evidence for Dialectical Behavior Therapy
some of the self-​judgment and blame these indi- Adaptations
viduals possess with regard to their disorder. Eating Currently, DBT adaptations for EDs, compared
disorder behaviors are viewed as a development over to other emotion-​focused therapies for EDs has the
time from a transaction between a biologically emo- largest evidence base. In particular, evidence from
tionally vulnerable individual poorly matched with case series demonstrates that women with BPD
an environment experienced as invalidating. and ED show improvements in ED symptoms
As noted, DBT is a protocol-​driven treatment with minimally adapted DBT in an outpatient
in which sessions are organized by the highest setting (Chen, Matthews, Allen, Kuo, & Linehan,
ranking (or highest risk) target behavior that took 2008; Palmer et al., 2003). Furthermore, evidence
place that day or over the last week. In addition, suggests that a DBT-​informed treatment in partial
there are special protocols designed for addressing hospitalization and intensive outpatient programs
life-​threatening and therapy-​interfering behavior. for individuals with EDs with multiple diagnoses,
Dialectical behavior therapy provides detailed including BPD, shows promise in reducing ED
guidance for therapists in managing crisis behav- symptoms, as well as improving affect regulation
iors, which may be particularly, as individuals with among those with co-​occurring ED and BPD (Ben-​
EDs often engage in suicidal and nonsuicidal self-​ Porath, Wisniewski, & Warren, 2009; Federici &
injurious behaviors (e.g., Claes & Muehlenkamp, Wisniewski, 2013). Additionally, a stepped-​ care
2014; Franko & Keel, 2006; Svirko & Hawton, trial for early weak responders to guided self-​help
2007). In addition, DBT uniquely targets therapy-​ CBT with BED and BN compared minimally
interfering behavior (e.g., the client/​ therapist adapted standard DBT with group and individual
missing sessions or being late to sessions, home- CBT and found that both intensive treatments
work incompletion). The protocol-​driven nature resulted in similar binge frequency reduction at
of DBT offers flexibility within a session, thus end of treatment and at 1-​year follow up (Chen
allowing multiple problems besides the ED to be et al., 2016).
addressed. Although there is evidence for the effective-
The fusion of behavior change strategies with ness of different adaptations of DBT for EDs, the
novel acceptance-​based strategies, such as mindful- only adaptation that has been supported through
ness skills, makes DBT unique among ED treat- randomized controlled trials is the Stanford DBT
ments. Mindfulness-​based approaches to thoughts model (Safer, Telch, & Chen, 2009; Telch, 1997a)
and emotions contrast with older-​style cognitive for BED and BN. The Stanford DBT model for
restructuring techniques. Clients are taught to BED and BN differs from the standard DBT
observe and describe their thoughts and feelings, to program for BPD in notable ways, which affects
watch them come and go, and to note this constant treatment structure and content. For example, the
passage. In particular, clients learn that thoughts adapted Stanford version employs a single modality
and feelings are simply thoughts and feelings, of treatment delivery (weekly 2-​hour group DBT
which can be sat with and observed without being for BED and or weekly 50-​to 60-​minute individual
acted on. DBT for BN) versus standard DBT’s weekly indi-
The nonjudgmental component of mindful- vidual therapy (50–​60 minutes) plus 2-​hour group
ness is important for both clients and therapists. skills training.
For example, clients with EDs are often judgmental The Stanford DBT Model for BED and BN was
about their appearance and themselves. Therapists originally developed for adult women 18 to 65 years
may also become judgmental of client’s thoughts old. Exclusion criteria included (1)  current use of
and behaviors, as ED behaviors may be trivialized psychotropic medications, (2) psychotic or bipolar
or judged to be scheming, deceitful, or superficial. affective disorders diagnoses, (3)  current involve-
This can lead to invalidation of clients, a lack of ment in psychotherapy or weight loss treatments,
motivation for treatment, and burnout of both cli- (4) current suicidality, (5) current substance abuse
ents and therapists. The nonjudgmental standpoint or dependence, or (6) pregnancy. Clients with BPD
of DBT allows ED behaviors to be viewed non- were not specifically excluded.
pejoratively and thus usefully defined as responses The Stanford DBT model (has been researched
that are within a client’s current skill repertoire but in one randomized-​controlled trial in a compari-
that can be replaced by more helpful or adaptive son with active comparison group therapy (ACGT)
responses. for BED (Safer, Robinson, & Jo, 2010), in two

338 Dialectical Behavior Therapy

randomized-​ controlled trials using wait-​ list con- anxious, or depressed. In addition, the Positive and
trols, one for BED (Telch, Agras, & Linehan, 2001) Negative Affect Scale (Watson, Clark, & Tellegen,
and one for BN (Safer, Telch, & Agras, 2001a), as 1988) showed significant decreases in participant’s
well as in an uncontrolled trial (Telch, Agras, & experience of negative affect. At post-​treatment the
Linehan, 2000) for BED and two case reports (Safer dropout rate in DBT was 0%, compared with 6.7%
et al., 2001b; Telch, 1997b). in the wait-​list control group.
These results for the Stanford DBT model are Preliminary evidence on other DBT adaptations
supportive of its use. For example, in the first ran- for EDs includes DBT-​guided self-​help (DBTgsh),
domized controlled trial of DBT for BED, 16 of the radically open-​ DBT (RO-​ DBT), and appetite-​
18 women (89%) who received DBT were abstinent focused DBT. The DBTgsh protocol was developed
from binge eating at the end of the 20-​week treat- to encourage the dissemination of DBT for BED
ment compared to 2 of 16 (12.5%) wait-​list controls in a cost-​effective way. In a randomized controlled
(Telch et al., 2001). The dropout rate was low, with trial that compared DBTgsh with a wait-​list con-
only 9% (2 of 22) of the sample dropping out after trol, it was found that those who received DBTgsh
initially beginning DBT. At post-​treatment, clients reported significantly fewer past-​month binge eat-
in DBT reported significantly improved weight and ing episodes at end of treatment and greater rates of
shape concerns and eating concerns and demon- abstinence at 6-​month follow-​up than those in the
strated reduced urges to eat on the Emotional Eating control condition (Masson, von Ranson, Wallace,
Scale (Arnow, Kenardy, & Agras, 1995), especially & Safer, 2013).
when angry. At the 3-​month follow-​up, 67% of the In contrast to DBTgsh for BED, RO-​DBT was
18 participants in DBT were abstinent from binge developed for anorexia nervosa (AN)-​ restricting
eating and 56% of the 18 at the 6-​month follow-​ type and focuses on skills to reduce emotional
up. The DBT participants reported practicing on overcontrol. Two preliminary trials of RO-​ DBT
average 3.6 different skills per week an average of for AN found that standard DBT skills plus skills
4  days per week at the final assessment. The high that address overcontrol resulted in increased BMI
abstinence rates were consistent with those of the in inpatient (Lynch et  al., 2013) and outpatient
smaller uncontrolled trial of DBT for BED, where (Chen et al., 2015) settings, which was maintained
82% of the participants were abstinent from binge at 1-​year follow-​up (Chen et al., 2015), as well as
eating after 20 group sessions, with none dropping improvements in ED-​related symptoms at the end
out after commencing treatment (Telch et al. 2000). of treatment (Chen et al., 2015; Lynch et al., 2013).
Similar findings were found as part of a replication/​ Notably, Lynch and colleagues (2013) found that,
extension study of the Stanford DBT model for within an inpatient setting, 90% of treatment com-
BED, in which the client population was expanded pleters achieved either full or partial remission of
to include both men and women and individuals on AN symptoms. Moreover, participants from both
stable psychotropic medication (Safer et al., 2010). RO-​ DBT trials found the treatment acceptable,
Sixty-​four percent of those who received DBT for resulting in retention of 8/​9 patients (Chen et al.,
BED were binge abstinent after 20 sessions, com- 2015) and 34/​ 47 patients (Lynch et  al., 2013),
pared with 36% of those in the comparison group. respectively.
These rates are similar to abstinence rates found in In addition to trials on DBTgsh for BED and
CBT and IPT for BED (Wilfley et al., 1993, 2002). RO-​DBT for AN, appetite-​focused DBT has been
Findings from a randomized controlled trial examined in a preliminary trial with BN. In a com-
of group DBT for BN were also promising. In 20 parison of 12 weeks of appetite-​focused DBT with a
weeks of individually delivered DBT for bulimic 6-​week delayed treatment control (Hill, Craighead,
symptoms, abstinence from binge eating/​purging & Safer, 2011), it was found that BN symptoms
behaviors at the end of 20 weeks of treatment was were significantly reduced among participants
28.6% (4 of 14) for DBT and 0% (0 of 15) for the who received appetite-​ focused DBT, relative to
wait-​list control (Safer et  al., 2001a). These find- the control condition at 6 weeks. Furthermore, at
ings were similar to post-​treatment abstinence rates post-​treatment, 26.9% of individuals who received
from the largest multisite CBT for BN trial (Agras, appetite-​focused DBT were binge/​purge abstinent
Walsh, Fairburn, Wilson, & Kraemer, 2000). DBT for the past month, and 61.5% of individuals no
resulted in moderate to large effect-​size changes on longer met full or subthreshold BN criteria. Taken
the Emotional Eating Scale (Arnow et  al., 1995), together, preliminary evidence DBT adaptations
reducing urges to eat when angry or frustrated, for AN and BN suggest that skills training targeting

Chen, Yiu, Safer 339

rigidity in AN and skills training targeting mindful (emotional eating) than non-​binge-​eaters, indepen-
eating in BN warrant further consideration. dent of degree of overweight (Eldredge & Agras,
Finally, preliminary data from open trials, case 1996). Specifically, the most frequently cited emo-
series, and one randomized controlled trial (Parling, tion triggering binge eating is anxiety, followed by
Cernvall, Ramklint, Holmgren, & Ghaderi, 2016) sadness, loneliness, and anger, with happiness being
on other novel emotion-​ focused treatments for the least frequently cited (Masheb & Grilo, 2006).
EDs are promising. These treatments include emo- Furthermore, self-​defined binges in obese women
tion acceptance behavior therapy for AN (Wildes with BED were significantly more associated with
et  al., 2014), emotion-​focused group therapy for negative mood, relative to caloric deprivation (Agras
inpatient AN (Tchanturia, 2011), emotion-​focused & Telch, 1998). In an experience sampling study,
group therapy for BN (Wnuk, Greenberg, & binge eating was preceded by more aversive mood
Dolhanty, 2015) as well as acceptance and commit- states (Greeno, Wing, & Shiffman, 2000). Using
ment therapy for AN (Berman, Boutelle, & Crow, a similar methodology, greater diurnal fluctuation
2009; Parling et al., 2016), inpatient EDs (Juarascio in depression and anxiety and higher frequency of
et al., 2013) and BED (Juarascio et al., 2017; Hill, these moods during eating and binge eating was
Masuda, Melcher, Morgan, & Twohig, 2015). reported in BED subjects compared with non-​BED
subjects (Lingswiler, Crowther, & Stephens, 1987).
Dialectical Behavior Therapy for Eating In addition to eating in response to negative moods,
Disorders individuals who binge eat also appear to evaluate
We first describe the model of the etiology and situations as more stressful than non-​ED individu-
maintenance of eating disorders in DBT for eating als (Hansel & Wittrock, 1997). This may be due to
disorders. difficulties in identifying and making sense of emo-
tional states, along with limited access to emotion
Affect Regulation Model for Eating regulation strategies (Whiteside et al., 2007).
Disorders
Stress and negative mood are the most frequently Adaptation of the Biosocial Theory
cited precipitants of binge eating (Polivy & Herman, for Eating Disorders
1993). In DBT for EDs, binge eating is viewed as In DBT, ED behaviors, just like BPD behaviors,
analogous to that of self-​injury in standard DBT are understood as resulting from a transaction over
treatment for individuals with BPD. Binge eating time between an individual biologically predisposed
or bulimic behaviors are understood as the result of to be more emotionally vulnerable and a mismatch
attempts to escape from primary or secondary aver- with an environment experienced (but not neces-
sive emotions that may be triggered by thoughts sarily intended) as invalidating. This invalidating
regarding food, body image, perfectionism, nega- environment may punish emotional displays, lead-
tive thoughts about the self, or interpersonal situ- ing individuals to engage in ED behaviors to man-
ations (Linehan & Chen, 2005). Binge eating and age their emotions and the secondary emotion of
bulimic behavior function to quickly narrow atten- shame that may result. Sometimes, the invalidation
tion and cognitive focus from these thoughts and may be specific to ED behaviors (e.g., “Why can’t
to provide immediate escape from physiological you just stop eating?”), or take the form of weight-​
responses and feelings. Over time, binge eating as related teasing or overconcern with weight by peers
an escape behavior becomes reinforced (Heatherton and family. The invalidating environment may
& Baumeister, 1991), particularly if there is a lack also be broader, including typical Western societal
of adaptive emotion regulation skills. Disordered messages idealizing thinness and disparaging over-
eating behaviors may then be further reinforced by weight, such that each is associated with polar moral
secondary emotions, such as shame (e.g., Sanftner values. Further invalidation may be introduced by
& Crowther, 1998), with binge eating as a long-​ the media with the notion that weight loss should
term consequence of an overlearned dysfunctional not be difficult (e.g., “lose 10 pounds in 10 days”
response to dysregulated emotions. advertisements).
Research evidence supports the role of negative Over time, the results of these transactions may
emotions and emotion regulation in binge eat- include (1)  difficulties in identifying and regulat-
ing. Self-​report studies of weight loss participants ing emotional arousal; (2)  difficulties in tolerat-
show that individuals with BED report higher ing emotional distress without engaging in ED
urges to binge eat in response to negative emotions behavior; (3)  difficulties in trusting one’s own

340 Dialectical Behavior Therapy

emotional responses as valid, that is, engaging in potential theoretical overlap with IPT. For clinicians
self-​invalidation; and (4)  formation of unrealistic and programs that are not limited by the constraints
goals and expectations due to oversimplification of of time, resources, or research, there is no research-​
problem solving and goal setting by the invalidating based reason not to include the Interpersonal
environment. Self-​invalidation may make individu- Effectiveness module—​particularly given the data
als particularly vulnerable to turning to body-​image-​ on IPT’s efficacy with BED.
focused environments as sources of information
about what the self “should” look like. This may Structure of Treatment (Session Structure,
increase the likelihood of establishing unrealistic Number of Sessions, Modules Covered)
expectations among overweight or normal-​weight The DBT for BED/​BN combines elements of
clients regarding weight loss. individual psychotherapy and group skills training
from standard DBT. Dialectical behavior therapy
The Stanford Model for BED/​BN incorporates chain analysis strategies,
The Stanford model of DBT was developed to typically conducted in individual psychotherapy in
target clients whose primary focus of treatment is standard DBT, with skills training, typically offered
BED or BN symptoms that interfere with their in a group format in standard DBT.
quality of life. It was not intended for individu- The format of each session is divided evenly. The
als with active suicidal or self-​injurious behaviors, first half, consisting of 50 to 60 minutes if treatment
who should be offered the full program of stan- is carried out in a 2-​hour group format or 25 to 30
dard DBT. In this section, the Stanford model of minutes if treatment is carried out in a 50-​to 60-​
DBT as adapted for BED (in a group format) and minute individual format, is devoted to homework
for BN (in an individual format) is described. As review and includes discussion of client diary cards
briefly noted above, the Stanford adaptation of and chain analyses (motivation and skills strength-
DBT for BED or BN differs in three important ening). During this review, each group member has
ways from standard DBT for BPD. First, it differs between 5 and 10 minutes to report on his or her
in its structure. The Stanford DBT model combines use of new skills in the past week and to describe
two modalities in standard DBT (individual treat- specific successes or difficulties in applying the skills
ment and group skills treatment) into one modality, to replace the targeted problem eating behaviors.
either a 2-​hour group treatment for BED or a 50-​to The length of time each member has for home-
60-​minute individual session for BN. As opposed to work review varies given the total duration of the
what is typically a year-​long treatment in standard session (which may be 2 hours or 2½ hours) and
DBT, the Stanford DBT model is briefer and uses the number in attendance so that sufficient time is
20 sessions of treatment covering three (e.g., Core available for everyone to share. Group members are
Mindfulness, Emotion Regulation, and Distress encouraged to help one another identify solutions
Tolerance) as opposed to four skills-​training mod- to problems encountered in using the skills and
ules. Second, the Stanford DBT model differs from to “cheerlead” efforts made. Separated by a 5-​to
standard DBT in the use of specific ED behavior 10-​minute break when using a group format, the
targets, resulting in adaptations to the treatment second half of each session is devoted to teaching
hierarchy, diary card, and behavioral chain analysis. new content and practice of new skills (e.g., skills
Third, the Stanford model involves the addition of acquisition).
particular concepts and skills specific to ED clients Like skills-​training groups in standard DBT,
(e.g., the concept of dialectical abstinence, use of groups in the Stanford DBT model for BED are
skills such as mindful eating, urge surfing, alternate composed of 8 to 10 members and are taught
rebellion, etc.). by 2 skills trainers—​a leader and a coleader. As
Adaptations made for the Stanford DBT model described, the Stanford DBT model for BED used
were made primarily for research purposes. For a group format, and the Stanford DBT model for
instance, other efficacious treatments for BED and BN/​partial BN was carried out in an individual
BN, such as CBT and IPT, have been researched format. In the Stanford model, DBT for BN was
using 20 sessions. Thus, for the Stanford DBT administered individually due to difficulties in
model to be compared with other efficacious treat- recruiting sufficient numbers to start a group in a
ments, a similar number of sessions are required. timely fashion (as needed in a research study) and
The Interpersonal Effectiveness module was because most treatments for BN are individually
removed due to time constraints and because of a delivered.

Chen, Yiu, Safer 341

Sequence of Treatment Adaptations to Treatment Hierarchy,
The pretreatment stage of the Stanford model Diary Card, and Behavioral Chain
comprises a pretreatment interview as well as ses- Analysis
sions 1 and 2. During the pretreatment interview, Adaptations from standard DBT were made
each participant meets individually with one of the to the treatment hierarchy and related tools to
cotherapists (or, for BN, the individual therapist) reflect the needs of the population for whom the
for 30 to 45 minutes before beginning therapy. The Stanford DBT model was developed—​ clients
major goals of this pretreatment visit involves ori- whose ED symptoms (e.g., binge eating and/​or
enting the participant to the DBT program (e.g., purging) are the primary treatment focus. For
dates of treatment), assessing prior group experi- most clients with primary BN or BED, the high-
ence, introducing the affect regulation model of est treatment target is to prevent any behavior(s)
binge eating and the targets of treatment, describ- that interfere with the successful delivery of treat-
ing the expectations for participants (e.g., regular ment (e.g., life-​ threatening behaviors, therapy-​
timely attendance, listening to tapes of any missed interfering behaviors), followed by those outlined
sessions, completing homework assignments) and in the Path to Mindful Eating—​an additional tar-
therapists, and orienting individuals to the goals of get hierarchy adapted specifically for individuals
treatment. with BED or BN (see Table 17.1).
In session 1, a major goal is to obtain a com- The Path to Mindful Eating hierarchy addresses
mitment from each group member to cease mindless eating, which refers to inattention when
binge eating (or for BN, the individual client’s eating, such as when eating popcorn while watching
commitment to cease binge eating and purging). television, only to discover the popcorn is finished
Standard DBT commitment strategies are used, without awareness of how this happened. Mindless
such as having therapists play devil’s advocate eating, in contrast to binge eating, is defined as
in order to have clients consider and argue why occurring without the experience of a sense of
they cannot continue binge eating and have the loss of control. Food preoccupation is when one’s
quality of life that they desire. Other tasks of thoughts or attention are absorbed or focused on
this session are to explain the biosocial theory, food to the point of interference with functioning
review and sign formal therapist and client (e.g., work). Capitulating involves giving up on
agreements, and introduce the diary card and one’s goals to stop binge eating and in using skillful
chain analysis. behavior to cope with emotions. When capitulat-
In session 2, after conducting homework review ing, one acts as if there is no other option or way to
in the first half of the session, therapists introduce cope with aversive emotions than with food. Finally,
clients to the concept of dialectical abstinence, apparently irrelevant behaviors are those that do not
a concept originally developed in DBT for sub- seem relevant to binge eating and purging or that a
stance use disorders (Linehan & Dimeff, 1997). client insists do not matter but are actually impor-
This concept is described in greater detail in what tant given information from chain analyses. For
follows. example, purchasing extra dessert “for company”
After the introductory sessions, the modules cov- may not seem to matter, but the presence of extra
ered, in sequence, are: the Core Mindfulness mod- dessert in the home may actually be linked to a cli-
ule (sessions 3–​5), the Emotion Regulation module ent’s binge eating.
(sessions 6–​12), and the Distress Tolerance module The Stanford model of DBT for BED/​BN also
(sessions 14–​18). involves adaptations to the standard DBT diary card
Sessions 19 and 20 are devoted to a review of
the different skills modules and relapse prevention.
Clients are asked to detail their plans for continuing Table 17.1  Path to Mindful Eating
to practice the skills taught, to outline their specific 1. Stop binge eating (and purging—​for BN clients).
plans for skillfully managing emotions that may
trigger binge eating in the future, and to consider 2. Eliminate mindless eating.
what they need to do next to continue to build a 3. Decrease cravings, urges, preoccupation with food.
satisfying and rewarding quality of life. Clients say
their final goodbyes and perhaps conduct a good- 4. Decrease capitulating.
bye ritual (e.g., writing cards) to mark the ending 5. Decrease apparently irrelevant behaviors.
of treatment.

342 Dialectical Behavior Therapy

to allow clients to track any dysfunctional eating bridges were adapted for binge eating and/​or purg-
behaviors as outlined on the Path to Mindful Eating ing, and taught over sessions 3 to 5.
(Table 17.1). A sample diary card can be found in Mindful eating, as opposed to mindless eating,
Table 17.2. Finally, the chain analysis used in DBT is the experience of full participation in eating,
for BED or BN is the same as that of standard DBT. which involves observing and describing the experi-
Clients are directed to fill them out each week on the ence in one’s mind. It is eating with full awareness
highest disordered eating behavior since the previous and attention (one-​ mindfully) but without self-​
session as listed in the Path to Mindful Eating. consciousness or judgment.
Urge surfing involves mindful, nonattached
Dialectical Behavior Therapy for Binge observation of urges to binge or to eat mindlessly.
Eating and/​or Purging Clients are educated about how urges and cravings
A number of DBT concepts and skills were are classically conditioned responses that have been
added in DBT for BED/​BN to specifically address associated with a particular cue. Mindful urge surf-
the needs of clients who binge eat and/​or purge. As ing involves awareness without engaging in impul-
noted, many of these were originally developed for sive mood-​dependent behavior. One learns to “let
DBT for substance abuse (see Linehan & Dimeff, go” or “detach” from the object of the urge, being
1997; Marlatt & Gordon, 1985). Included are fully in the moment, “riding the wave” of the urge
the concept of dialectical abstinence as well as ED and noticing its ebb and flow. Although similar to
specific skills such as mindful eating, urge surfing, mindfulness of the current emotion, urge surfing
alternate rebellion, and burning bridges. Each is is a mindfulness skill that involves nonjudgmental
reviewed in turn. observing and describing of urges, cravings, and
Dialectical abstinence, introduced in session 2, food preoccupation.
is a synthesis of a 100% commitment to abstinence Alternate rebellion involves being effective in sat-
and a 100% commitment to relapse prevention isfying a wish to rebel without destroying one’s over-
strategies. Before a client engages in binge eating, riding objective to stop binge eating. The purpose
there is an unrelenting insistence on total absti- is not to suppress or judge the rebellion but to find
nence. After a client has engaged in binge eating, creative ways to rebel that do not involve “cutting
however, the emphasis is on radical acceptance, off your nose to spite your face.” Many clients with
nonjudgmental problem solving and effective BED have described the desire to “get back” at soci-
relapse prevention, followed by a speedy return to ety, friends, and/​or family perceived to be judgmental
an unrelenting insistence on abstinence. Therapists about their weight. For these clients, “getting back”
liken clients to Olympic athletes (Safer et al., 2009; often involves rebelling by consuming even more
Telch, 1997a) with therapists as their coaches. As food. However, this is not effective, as it runs con-
an Olympian, one only focuses on “going for the trary to achieving the goal of binge abstinence and
gold” as opposed to focusing on what might happen an improved quality of life. Therapists can encourage
if one were to fall or telling oneself before the race clients to observe the need to rebel, label it as such,
“Maybe a bronze would be okay.” Similarly, clients and then, if the decision is to act on the wish, to do
must focus only on binge abstinence. Yet athletes so effectively. Clients can be creative in thinking up
and clients must be prepared for the possibility of alternate rebellion strategies. For example, a client
failure. The key is to be prepared to fail well. The who feels judged by society for being obese might
dialectical dilemma is that both success and failure “rebel” by buying attractive lingerie that makes her
exist. The dialectical abstinence solution involves feel beautiful or mindfully sitting in a well-​regarded
100% certainty that binge eating is out of the ques- restaurant and openly, unselfconsciously treating her-
tion and 100% confidence that one will never binge self to a healthy and delicious bowl of soup.
again. However, simultaneously, one keeps in mind Burning bridges is a radical acceptance skill that
(“Way, way back in the very farthest part so that it involves accepting at the deepest and most radical
never interferes with your resolve”) that if one slips, level the idea that one is really not going to binge
one will deal with it effectively by accepting the eat, mindlessly eat, or abuse oneself with food ever
behavior nonjudgmentally and picking oneself back again, thus burning the bridge to those behaviors.
up, knowing one will never slip again. One accepts that they will no longer block, deny, or
In addition to the concept of dialectical absti- avoid reality with binge eating. Instead, one makes a
nence, ED specific mindfulness skills of mindful commitment from deep within to accept reality and
eating, urge surfing, alternate rebellion, and burning one’s experiences as they are.

Chen, Yiu, Safer 343

Table 17.2a.  Stanford DBT Model Diary Card
Instructions for Completing Your Diary Card
Urge to Binge: Refer to the legend and choose the number from the scale (0–​6) that best represents your highest rating for the day. The key characteristics to consider when making
your rating are intensity (how strongly you felt the urge) and duration (how long the urge lasted).
Binge Episodes: Write the number of binge episodes you had each day. A binge refers to an eating episode in which you felt a loss of control during the eating.
Mindless eating: Write in the number of “mindless” eating episodes that you had each day. Mindless eating refers to not paying attention to what you are eating, although you do not
feel the sense of loss of control that you do during binge episodes. A typical example of mindless eating would be sitting in front of the TV and eating a bag of microwave popcorn
without any awareness of the eating (i.e., somehow, the popcorn was gone and you were only vaguely aware of having eaten it). Again, however, you didn’t feel a sense of being out of
control during the eating.
Apparently Irrelevant Behaviors (AIB): Circle either “yes” or “no” depending on whether you did or did not have any AIBs that day. If you did, briefly describe the AIB in the
place provided or on another sheet of paper. An AIB refers to behaviors that, on first glance, do not seem relevant to binge eating and purging but that actually are important in the
behavior chain leading to these behaviors. You may convince yourself that the behavior doesn’t matter or really won’t affect your goal to stop binging and purging when, in fact, the
behavior matters a great deal. A typical AIB might be buying several boxes of your favorite Girl Scout cookies because you wanted to help out a neighbor’s daughter (of course, you
could buy the cookies and donate them to the neighbor, rather than taking them home).
Capitulating: Refer to the legend and choose the number from the scale (0–​6) that best represents your highest rating for the day. The key characteristics to consider when making
your rating are intensity (strength of the capitulating) and duration (how long it lasted). Capitulating refers to giving up on your goals to stop binge eating and to skillfully cope with
emotions. Instead, you capitulate or surrender to binge eating, acting as if there is no other option or way to cope than with food.
Food Preoccupation: Refer to the legend and choose the number from the scale (0–​6) that best represents your highest rating for the day. Food preoccupation refers to your thoughts
or attention being absorbed or focused on food. For example, your thoughts of a dinner party and the presence of your favorite foods may absorb your attention so much that you
have trouble concentrating at work.
Emotion Columns: Refer to the legend and choose the number from the scale (0–​6) that best represents your highest rating for the day. The key characteristics to consider when
making your rating are intensity (strength of the emotion) and duration (how long it lasted).
Used Skills: Refer to the legend and choose the number from the scale (0–​6) that best represents your attempts to use the skills each day. When making your rating, consider whether
or not you thought about using any of the skills that day, whether or not you actually used any of the skills, and whether or not the skills helped.
Weight: Weigh yourself once each week and record your weight in pounds in the space provided. Please write in the date you weighed. It is best if you choose the same day each week
to weigh. Many women find that arriving a few minutes early to the session and weighing at the clinic is a good way to remember to weigh.
Urge to Quit Therapy: Indicate your urge to quit therapy before the session and after the session each week. Both of these ratings should be made for the same session as the one in
which you received the diary card. It is best to make both of these ratings as soon as possible following that day’s session. Use a 0 to 6 scale of intensity of the urge, with 0 indicating
no urge to quit and a 6 indicating the strongest urge to quit.
Completing the Skills Side of the Diary Card:
How Often Did You Fill Out This Side? Place a check mark to indicate how frequently you filled out the skills side of the diary card during the week.
Skills Practice: Go down the column for each day of the week and circle each skill that you practiced/​used that day. If you did not practice or use any of the skills that particular day,
then circle that day on the last line, which states, “Did not practice/​use any skills.”
b.  Symptoms and Behaviors Diary Card
Diary Initials ID How often did you fill out this side? _​_​_​Daily _​_​_​_​4–​6 × _​_​_​_​ 2–​3 × _​_​
Card _​ Once
Day Urgea to Binge Mindless AIBb Capitulatinga Fooda Fooda Angera Sadnessa Feara Shamea Pridea Happinessa Usedc
And Binge Episodes Eating Craving Preoccupation Skills
Date (0–​6)
# OBE lg # SBE sm # episodes Circle one (0–​6) (0–​6) (0–​6) (0–​6) (0–​6) (0–​6) (0–​6) (0–​6) (0–​6) (0-​7)
Mon yes/​no
Tues yes/​no
Wed yes/​no
Thurs yes/​no
Fri yes/​no
Sat yes/​no
Sun yes/​no
a
Use the following scale to indicate the highest rating for the day:
0 = urge/​thought/​feeling not experienced
1 = urge/​thought/​feeling experienced slightly and briefly
2 = urge/​thought/​feeling experienced moderately and briefly
3 = urge/​thought/​feeling experienced intensely and briefly
4 = urge/​thought/​feeling experienced slightly and endured
5 = urge/​thought/​feeling experienced moderately and endured
6 = urge/​thought/​feeling experienced intensely and endured
b
Describe Apparently Irrelevant Behaviors (AIB):
c
Used Skills:
0 = Not thought about or used
1 = Thought about, not used, didn’t want to
2 = Thought about, not used, wanted to
3 = Tried but couldn’t use them
4 = Tried, could do them, but they didn’t help
5 = Tried, could use them, helped
6 = Didn’t try, used them, didn’t help
7 = Didn’t try, used them, helped
Weight _​_​_​_​_​_​_​_​ Date Weighed _​_​_​_​_​_​_​_​_​
Urge to quit therapy (0–​5): Before therapy session: _​_​_​_​_​ After therapy session: _​_​_​_​
NIMH 1997–​2000
ER BED TELCH
c.  The Skills Use Weekly Diary Card
Skills diary card Instructions: Circle the days you worked on each skill. How often did you fill out this side? _​_​_​Daily _​_​_​4–​6 × _​_​_​ 2–​3
× _​_​_​_​ Once
1. Diaphragmatic Breathing Mon Tues Wed Thurs Fri Sat Sun
2. Wise mind Mon Tues Wed Thurs Fri Sat Sun
3. Observe: just notice Mon Tues Wed Thurs Fri Sat Sun
4. Describe: put words on Mon Tues Wed Thurs Fri Sat Sun
5. Participate: enter into the experience Mon Tues Wed Thurs Fri Sat Sun
6. Mindful eating Mon Tues Wed Thurs Fri Sat Sun
7. Nonjudgmental stance Mon Tues Wed Thurs Fri Sat Sun
8. One-​mindfully: in-​the-​moment Mon Tues Wed Thurs Fri Sat Sun
9. Effectiveness: focus on what works Mon Tues Wed Thurs Fri Sat Sun
10. Urge surfing Mon Tues Wed Thurs Fri Sat Sun
11. Alternate rebellion Mon Tues Wed Thurs Fri Sat Sun
12. Mindful of current emotion Mon Tues Wed Thurs Fri Sat Sun
13. Loving your emotions Mon Tues Wed Thurs Fri Sat Sun
14. Reduce vulnerability: PLEASE Mon Tues Wed Thurs Fri Sat Sun
15. Build MASTERy Mon Tues Wed Thurs Fri Sat Sun
16. Build positive experiences Mon Tues Wed Thurs Fri Sat Sun
17. Mindful of positive experiences Mon Tues Wed Thurs Fri Sat Sun
18. Opposite-​to-​emotion action Mon Tues Wed Thurs Fri Sat Sun
19. Observing-​your-​breath Mon Tues Wed Thurs Fri Sat Sun
20. Half-​smiling Mon Tues Wed Thurs Fri Sat Sun
21. Awareness exercises Mon Tues Wed Thurs Fri Sat Sun
22. Radical acceptance Mon Tues Wed Thurs Fri Sat Sun
23. Turning the mind Mon Tues Wed Thurs Fri Sat Sun
24. Willingness Mon Tues Wed Thurs Fri Sat Sun
25. Burning your bridges Mon Tues Wed Thurs Fri Sat Sun
26. Distract Mon Tues Wed Thurs Fri Sat Sun
27. Self-​soothe Mon Tues Wed Thurs Fri Sat Sun
28. Improve the moment Mon Tues Wed Thurs Fri Sat Sun
29. Pros and cons Mon Tues Wed Thurs Fri Sat Sun
30. Commitment Mon Tues Wed Thurs Fri Sat Sun
30. Did not practice any skills Mon Tues Wed Thurs Fri Sat Sun
NIMH 1997–​2000 ER BED TELCH
Conclusion •  What are the mechanisms of action for DBT,
In summary, this chapter briefly describes stan- and how are they distinguished from mechanisms
dard DBT and adaptations that have been made of action for CBT and IPT?
to offer an alternative treatment approach for EDs •  Are there ways to combine elements of DBT
that have failed to respond to existing therapies. with CBT and/​or IPT?
Dialectical behavior therapy is uniquely based on an •  How can maintenance of treatment gains
affect regulation model, and fuses behavior change in DBT be improved for individuals with
strategies with novel acceptance-​ based strategies BED or BN?
such as mindfulness. Dialectical behavior therapy •  What client or therapist or combinations of
also includes distinctive protocols for addressing sui- these features are associated with better or worse
cidal behavior, nonsuicidal self-​injury, and therapy-​ outcome in DBT for BED or BN?
interfering behaviors, such as drop out, homework •  What are the mechanisms of action for
incompletion, and lack of attendance. the Stanford DBT model? Which components
Standard DBT is an efficacious evidence-​based are particularly helpful for individuals with
treatment for BPD, with promising treatment BED or BN?
development data for use with individuals with •  Can the Stanford DBT model be efficacious
BPD and EDs. For individuals with BED or BN for adolescents with BED/​BN?
that primarily affects quality-​of-​life, an adaptation •  Does DBT specifically change vulnerability to
of standard DBT for individuals with BED and emotions in individuals with BED or BN?
BN by researchers at Stanford University has been •  Further development and testing of DBT for
found to be efficacious in randomized controlled adults with AN and BPD is needed.
trials. The Stanford University DBT model involves •  Further testing of self-​help DBT for EDs is
20 group (for BED) or individual (for BN) sessions. needed.
These sessions teach mindfulness, emotion regula-
tion and distress tolerance skills, integrating these
References
with the use of chain analyses (careful step-​by-​step Agras, W. S., & Telch, C. F. (1998). The effects of caloric depriva-
behavioral analyses) for problem behaviors. This tion and negative affect on binge eating in obese binge-​eating
model uses a unique hierarchy of ED behaviors disordered women. Behavior Therapy, 29, 491–​503.
(the “Path to Mindful Eating”) to target, including Agras, W. S., Walsh, T., Fairburn, C. G., Wilson, G. T., &
Kraemer, H. C. (2000). A multicenter comparison of
(1) stopping binge eating (and purging, for clients
cognitive-​behavioral therapy and interpersonal psychother-
with BN); (2)  eliminating mindless eating (eating apy for bulimia nervosa. Archives of General Psychiatry, 57,
without awareness); (3) decreasing cravings, urges, 459–​466.
preoccupation with food; (4) decreasing capitulat- Arnow, B., Kenardy, J., & Agras, W. S. (1995). The Emotional
ing (i.e., giving in to binge eating); and (5) decreas- Eating Scale: The development of a measure to assess coping
with negative affect by eating. International Journal of Eating
ing apparently irrelevant behaviors (i.e., behaviors
Disorders, 18, 79–​90.
that appear not to matter but play a role in lead- Porath, D. D., Wisniewski, L., & Warren, M. (2009).
Ben-​
ing to a binge). Finally, the Stanford DBT model Differential treatment response for eating disordered patients
for BED and BN adds ED-​specific concepts and with and without a comorbid borderline personality diag-
skills to the standard treatment. The ED-​specific nosis using a dialectical behavior therapy (DBT)-​informed
approach. Eating Disorders, 17, 225–​241. doi:10.1080/​
DBT concepts include dialectical abstinence, which
10640260902848576
teaches clients to focus on attaining binge absti- Berman, M. I., Boutelle, K. N., & Crow, S. J. (2009). A case series
nence while simultaneously being aware that if one investigating acceptance and commitment therapy as a treat-
slips, this can be addressed effectively. Additional ment for previously treated, unremitted patients with anorexia
ED-​ specific DBT skills include mindful eating, nervosa. European Eating Disorders Review, 17, 426–​434.
Carter, G. L., Willcox, C. H., Lewin, T. J., Conrad, A. M., &
urge surfing (i.e., surfing urges to binge eat), alter-
Bendit, N. (2010). Hunter DBT project: Randomized con-
native rebellion (effectively managing urges to rebel trolled trial of dialectical behaviour therapy in women with
without binge eating), and burning bridges (radical borderline personality disorder. Australian and New Zealand
acceptance of a commitment to cease ED behavior). Journal of Psychiatry, 44, 162–​173.
Chen, E. Y., Cacioppo, J., Fettich K., Gallop, R., McCloskey
M. S., Olino T., & Zeffiro, T. A. (2016). An adaptive ran-
Future Directions domized trial of dialectical behavior therapy and cognitive
•  How does DBT fare compared with other behavior therapy for binge-​ eating. Psychological Medicine.
Nov 17:1–​15. [Epub ahead of print]
emotion-​focused therapies for BED and BN?

Chen, Yiu, Safer 347

Chen, E. Y., Matthews, L., Allen, C., Kuo, J. R., & Linehan, pilot study of an acceptance-​based behavioral treatment
M. M. (2008). Dialectical behavior therapy for clients with for binge eating disorder. Journal of Contextual Behavioral
binge-​eating disorder or bulimia nervosa and borderline per- Science.
sonality disorder. International Journal of Eating Disorders, Keel, P. K., & Brown, T. A. (2010). Update on course and out-
41, 505–​512. doi:10.1002/​eat.20522 come in eating disorders. International Journal of Eating
Chen, E. Y., Segal, K., Weissman, J., Zeffiro, T. A., Gallop, R., Disorders, 43, 195–​204.
Linehan, M. M.,  . . .  Lynch, T. R. (2015). Adapting dialec- Koons, C. R., Robins, C. J., Tweed, J. L., Lynch, T. R.,
tical behavior therapy for outpatient adult anorexia nervosa—​ Gonzalez, A. M., Morse, J. Q.,  . . .  Bastian, L. A. (2001).
A pilot study. International Journal of Eating Disorders, 48, Efficacy of dialectical behavior therapy in women veterans
123–​132. with borderline personality disorder. Behavior Therapy, 32,
Claes, L., & Muehlenkamp, J. J. (2014). Non-​suicidal self-​injury 371–​390.
in eating disorders. Heidelberg, Germany and New  York, Lieb, K., Zanarini, M. C., Schmahl, C., Linehan, M. M., &
NY: Springer. Bohus, M. (2004). Borderline personality disorder. Lancet,
Dimeff, L. A., & Koerner, K. (2007). Dialectical behavior ther- 364, 453–​461.
apy in clinical practice:  Applications across disorders and set- Linehan, M. M. (1993a). Cognitive-​behavioral treatment of bor-
tings: New York, NY: Guilford Press. derline personality disorder. New York, NY: Guilford Press.
Eldredge, K. L., & Agras, W. S. (1996). Weight and shape over-​ Linehan, M. M. (1993b). Skills training manual for treating bor-
concern and emotional eating in binge eating disorder. derline personality disorder. New York, NY: Guilford Press.
International Journal of Eating Disorders, 19, 73–​82. Linehan, M. M. (2014). DBT® skills training manual. Second
Fairburn, G., & Brownell, K. (2001). Eating disorders and obesity. edition. New York, London: Guilford Press.
New York, NY: Guilford Press. Linehan, M. M., Armstrong, H. E., Suarez, A., Allmon, D.,
Federici, A., & Wisniewski, L. (2013). An intensive DBT pro- & Heard, H. L. (1991). Cognitive-​ behavioral treatment
gram for patients with multidiagnostic eating disorder pre- of chronically parasuicidal borderline patients. Archives of
sentations:  A  case series analysis. International Journal of General Psychiatry, 48, 1060–​1064.
Eating Disorders, 46, 322–​331 [serial online]. Available from Linehan, M. M., & Chen, E. Y. (2005). Dialectical behavior
PsycINFO, Ipswich, MA. Accessed October 4, 2016. therapy for eating disorders. In S. Felgoise, A. M. Nezu,
Franko, D. L., & Keel, P. K. (2006). Suicidality in eating dis- C. M. Nezu, M. A. Reinecke, & A. Freeman (Eds.), The
orders:  Occurrence, correlates, and clinical implications. encyclopedia for cognitive-​ behavioral therapy. New  York,
Clinical Psychology Review, 26, 769–​782. NY: Springer.
Greeno, C. G., Wing, R. R., & Shiffman, S. (2000). Binge Linehan, M. M., Comtois, K. A., Murray, A. M., Brown, M.
antecedents in obese women with and without binge eat- Z., Gallop, R. J., Heard, H. L.,  . . .  Lindenboim N. (2006).
ing disorder. Journal of Consulting and Clinical Psychology, Two-​year randomized controlled trial and follow-​up of dia-
68, 95–​102. lectical behavior therapy vs. therapy by experts for suicidal
Grilo, C. M., Masheb, R. M., & Wilson, G. (2001). Subtyping behaviors and borderline personality disorder. Archives of
binge eating disorder. Journal of Consulting and Clinical General Psychiatry, 63, 757–​766.
Psychology, 69, 1066–​1072. Linehan, M. M., & Dimeff, L. A. (1997). Dialectical behavior
Hansel, S. L., & Wittrock, D. A. (1997). Appraisal and coping therapy manual of treatment interventions for drug abusers
strategies in stressful situations: A comparison of individuals with borderline personality disorder. Unpublished manuscript.
who binge eat and controls. International Journal of Eating University of Washington Seattle, WA.
Disorders, 21, 89–​93. Linehan, M. M., Dimeff, L. A., Reynolds, S. K., Comtois, K.
Harley, R., Sprich, S., Safren, S., Jacobo, M., & Fava, M. (2008). A., Welch, S. S., Heagerty, P., & Kivlahan, D. R. (2002).
Adaptation of dialectical behavior therapy skills training Dialectical behavior therapy versus comprehensive validation
group for treatment-​resistant depression. Journal of Nervous therapy plus 12-​step for the treatment of opioid dependent
and Mental Disorders, 196, 136–​143. women meeting criteria for borderline personality disorder.
Heatherton, T. F., & Baumeister, R. F. (1991). Binge eating as Drug and Alcohol Dependence, 67, 13–​26.
escape from self-​awareness. Psychological Bulletin, 110, 86–​108. Linehan, M. M., Heard, H. L., & Armstrong, H. E. (1993).
Hill, D. M., Craighead, L. W., & Safer, D. L. (2011). Appetite-​ Naturalistic follow-​up of a behavioral treatment for chroni-
focused dialectical behavior therapy for the treatment of cally parasuicidal borderline patients. Archives of General
binge eating with purging: A preliminary trial. International Psychiatry, 50, 971–​974.
Journal of Eating Disorders, 44, 249–​261. Linehan, M. M., Schmidt, H., III, Dimeff, L. A., Craft, J. C.,
Hill, M. L., Masuda, A., Melcher, H., Morgan, J. R., & Twohig, Kanter, J., & Comtois, K. A. (1999). Dialectical behavior
M. P. (2015). Acceptance and commitment therapy for therapy for patients with borderline personality disorder
women diagnosed with binge eating disorder: A case-​series and drug-​dependence. American Journal of Addiction, 8,
study. Cognitive and Behavioral Practice, 22(3), 367–​378. 279–​292.
Juarascio A., Shaw J., Forman E., Timko C. A., Herbert J., Linehan, M. M., Tutek, D. A., Heard, H. L., & Armstrong,
Butryn M., Bunnell D., Matteucci A., Lowe M. (2013). H. E. (1994). Interpersonal outcome of cognitive behav-
Acceptance and commitment therapy as a novel treatment ioral treatment for chronically suicidal borderline patients.
for eating disorders: an initial test of efficacy and mediation. American Journal of Psychiatry, 151, 1771–​1776.
Behavior Modification, 37, 459–​489. Lingswiler, V. M., Crowther, J. H., & Stephens, M. A. (1987).
Juarascio, A. S., Manasse, S. M., Espel, H. M., Schumacher, Emotional reactivity and eating in binge eating and obesity.
L. M., Kerrigan, S., & Forman, E. M. (2017, in press). A Journal of Behavioral Medicine, 10, 287–​299.

348 Dialectical Behavior Therapy

Lynch, T. R., Cheavens, J. S., Cukrowicz, K. C., Thorp, S. R., Sanftner, J. L., & Crowther, J. H. (1998). Variability in self-​
Bronner, L., & Beyer, J. (2007). Treatment of older adults esteem, moods, shame, and guilt in women who binge.
with comorbid personality disorder and depression:  A  dia- International Journal of Eating Disorders, 23, 391–​397.
lectical behavior therapy approach. International Journal of Stice, E., & Agras, W. S. (1999). Subtyping bulimic women along
Geriatric Psychiatry, 22, 131–​143. dietary restraint and negative affect dimensions. Journal of
Lynch, T. R., Gray, K. H., Hempel, R. J., Titley, M., Chen, E. Consulting and Clinical Psychology, 67, 460–​469.
Y., & O’Mahen, H. A. (2013). Radically open-​dialectical Svirko, E., & Hawton, K. (2007). Self-​injurious behavior and
behavior therapy for adult anorexia nervosa: Feasibility and eating disorders:  The extent and nature of the association.
outcomes from an inpatient program. BMC Psychiatry, 13, Suicide and Life-​Threatening Behavior, 37, 409–​421.
293. doi:10.1186/​1471-​244X-​13-​293 Tchanturia, K. (2011). A brief emotion focused intervention for
Lynch, T. R., Morse, J. Q., Mendelson, T., & Robins, C. J. inpatients with anorexia nervosa: a qualitative study. Journal
(2003). Dialectical behavior therapy for depressed older of health psychology, 16, 947–​958.
adults:  A  randomized pilot study. American Journal of Telch, C. F. (1997a). Emotion regulation skills training treatment
Geriatric Psychiatry, 11, 33–​45. for binge eating disorder:  Therapist manual. Unpublished
Marlatt, G., & Gordon, J. R. (1985). Relapse prevention and manuscript. Stanford University.
maintenance strategies in the treatment of addictive behaviours. Telch, C. F. (1997b). Skills training treatment for adaptive
New York, NY: Guilford Press. affect regulation in a woman with binge-​eating disorder.
Masheb, R. M., & Grilo, C. M. (2006). Emotional overeat- International Journal of Eating Disorders, 22, 77–​81.
ing and its associations with eating disorder psychopathol- Telch, C. F., Agras, W. S., & Linehan, M. M. (2000). Group
ogy among overweight patients with binge eating disorder. dialectical behavior therapy for binge-​eating disorder: A pre-
International Journal of Eating Disorders, 39, 141–​146. liminary, uncontrolled trial. Behavior Therapy, 31, 569–​582.
Masson, P. C., von Ranson, K. M., Wallace, L. M., & Safer, D. Telch, C. F., Agras, W. S., & Linehan, M. M. (2001). Dialectical
L. (2013). A randomized wait-​list controlled pilot study of behavior therapy for binge eating disorder. Journal of
dialectical behaviour therapy guided self-​help for binge eat- Consulting and Clinical Psychology, 69, 1061–​1065.
ing disorder. Behaviour Research and Therapy, 51, 723–​728. Turner, R. M. (2000). Naturalistic evaluation of dialectical behav-
doi:10.1016/​j.brat.2013.08.001 ior therapy-​ oriented treatment for borderline-​ personality
McMain, S. F., Links, P. S., Gnam, W. H., Guimond, T., Cardish, disorder. Cognitive Behavioral Practice, 7, 413–​419.
R. J., Korman, L., & Streiner, D. L. (2009). A randomized van den Bosch, L. M., Verheul, R., Schippers, G. M., & van
trial of dialectical behavior therapy versus general psychiatric den Brink, W. (2002). Dialectical behavior therapy of
management for borderline personality disorder. American borderline patients with and without substance use prob-
Journal of Psychiatry, 166, 1365–​1374. lems:  Implementation and long-​ term effects. Addictive
Palmer, R. L., Birchall, H., Damani, S., Gatward, N., McGrain, Behaviors, 27, 911–​923.
L., & Parker, L. (2003). A dialectical behavior therapy pro- Van Dijk, S., Jeffrey, J., & Katz, M. R. (2013). A randomized,
gram for people with an eating disorder and borderline per- controlled, pilot study of dialectical behavior therapy skills in
sonality disorder:  Description and outcome. International a psychoeducational group for individuals with bipolar disor-
Journal of Eating Disorders, 33, 281–​286. der. Journal of Affective Disorders, 145, 386–​393.
Parling, T., Cernvall, M., Ramklint, M., Holmgren, S., & Ghaderi, Verheul, R., Van Den Bosch, L. M., Koeter, M. W., De Ridder,
A. (2016). A randomised trial of acceptance and commitment M. A., Stijnen, T., & Van Den Brink, W. (2003). Dialectical
therapy for anorexia nervosa after daycare treatment, includ- behavior therapy for women with borderline personal-
ing five-​year follow-​up. BMC Psychiatry, 16, 272. ity disorder:  12-​month, randomised clinical trial in The
Polivy, J., & Herman, C. P. (1993). Etiology of binge eat- Netherlands. British Journal of Psychiatry, 182, 135–​140.
ing: Psychological mechanisms. In C. G. Fairburn & G. T. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development
Wilson (Eds.), Binge eating: Nature, assessment, and treatment and validation of brief measures of positive and negative
(pp. 173–​205). New York, NY: Guilford Press. affect:  The PANAS scales. Journal of Personality and Social
Rathus, J. H., & Miller, A. L. (2002). Dialectical behavior Psychology, 54, 1063–​1070.
therapy adapted for suicidal adolescents. Suicide and Life Whiteside, U., Chen, E. Y., Neighbors, C., Hunter, D., Lo, T.,
Threatening Behaviors, 32, 146–​157. & Larimer, M. E. (2007). Binge eating and emotion regula-
Safer, D. L., Robinson, A. H., & Jo, B. (2010). Outcome from a tion: Do binge eaters have fewer skills to modulate and toler-
randomized controlled trial of group therapy for binge eating ate negative affect? Eating Behaviors, 8, 162–​169.
disorder: Comparing dialectical behavior therapy adapted for Wildes, J. E., Marcus, M. D., Cheng, Y., McCabe, E. B., &
binge eating to an active comparison group therapy. Behavior Gaskill, J. A. (2014). Emotion acceptance behavior therapy
Therapy, 41, 106–​120. for anorexia nervosa: A pilot study. International Journal of
Safer, D. L., Telch, C. F., & Agras, W. S. (2001a). Dialectical Eating Disorders, 47, 870–​873.
behavior therapy for bulimia nervosa. American Journal of Wilfley, D. E., Agras, W., Telch, C. F., Rossiter, E. M., Schneider,
Psychiatry, 158, 632–​634. J. A., Cole, A. G.,   .  .  .  Raeburn, S. D. (1993). Group
Safer, D. L., Telch, C. F., & Agras, W. (2001b). Dialectical behav- cognitive-​behavioral therapy and group interpersonal psy-
ior therapy adapted for bulimia: A case report. International chotherapy for the nonpurging bulimic individual:  A  con-
Journal of Eating Disorders, 30, 101–​106. trolled comparison. Journal of Consulting and Clinical
Safer, D. L., Telch, C. F., & Chen, E. Y. (2009). Dialectical behav- Psychology, 61, 296–​305.
ior therapy as adapted for binge eating disorder and bulimia. Wilfley, D. E., Friedman, M. A., Dounchis, J. Z., Stein,
New York, NY: Guilford Press. R. I., Welch, R. R., & Ball, S. A. (2000). Comorbid

Chen, Yiu, Safer 349

 psychopathology in binge eating disorder: Relation to eating interpersonal psychotherapy for the treatment of overweight
disorder severity at baseline and following treatment. Journal individuals with binge-​eating disorder. Archives of General
of Consulting and Clinical Psychology, 68, 641–​649. Psychiatry, 59, 713–​721.
Wilfley, D. E., Welch, R., Stein, R. I., Spurrell, E. B., Cohen, L. Wnuk, S. M., Greenberg, L., & Dolhanty, J. (2015). Emotion-​
R., Saelens, B. E.,  . . .  Matt, G. E. (2002). A randomized focused group therapy for women with symptoms of bulimia
comparison of group cognitive-​behavioral therapy and group nervosa. Eating Disorders, 23, 253–​261.

350 Dialectical Behavior Therapy

 CH A PT E R


Self-​Help and Stepped Care Treatments for
18 Eating Disorders

Carol B. Peterson, Emily M. Pisetsky, and Caroline E. Haut

Abstract
This chapter provides an overview of self-​help and guided self-​help treatments for eating disorders as
well as stepped care models for treatment delivery. Empirical evidence suggests that although guided
self-​help approaches may have relatively higher efficacy and retention rates than self-​help treatment, data
from comparison trials are inconsistent. Robust treatment predictors, moderators, and mediators have
not been identified other than rapid response as a predictor of outcome for cognitive-​behavioral guided
self-​help, which may be useful in informing stepped care treatment. Stepped care models have received
some empirical support and, in addition to potentially reducing treatment costs, may enhance efficacy
by providing individuals who are not responsive to initial treatments with alternative or adjunctive
interventions. Research using adaptive and tailored designs for treatment is needed to improve treatment
efficacy and dissemination. Further research is needed in cost-​efficacy, implementation, clinician training
models, and patient preferences and acceptability.
Key Words:  self-​help, guided self-​help, stepped care, early response, treatment scalability

Outpatient treatment for eating disorders has including binge eating and purging behaviors,
historically involved traditionally delivered psy- and are included in the treatment recommenda-
chotherapy and/​ or pharmacotherapy in clinical tions provided by the National Institute of Clinical
settings. However, stepped care models were pro- Excellence (NICE, 2004).
posed given the difficulty in determining at the Paralleling the accumulating body of research
start of treatment which patients will benefit the supporting SH and GSH as well as stepped care
most from which type of treatment (Fairburn & approaches, an increasing focus on identifying
Peveler, 1990; Wilson, Vitousek, & Loeb, 2000). stepped care models of eating disorder treatment
The potential efficacy of self-​help (SH) and guided has been driven by the need for cost-​ effective
self-​
help (GSH) interventions as components of delivery of evidence-​based treatments in the con-
stepped care treatment as well as stand-​alone inter- text of increasing healthcare costs. Stepped care
ventions for eating disorders has paralleled findings models that do not involve SH/​GSH but follow
in other areas of psychopathology treatment includ- algorithms based on predicted response/​ nonre-
ing depression and anxiety (Cuijpers, Donker, van sponse in directing alternative and adjunctive
Straten, Li, & Andersson, 2010). As the eating dis- treatments (e.g., combining psychotherapy and
order treatment field has evolved, SH and GSH medication using specific sequences) may also
interventions have accumulated growing empirical improve treatment efficacy and reduce attrition.
support, demonstrating their potential efficacy in Self-​help and GSH treatments, whether admin-
treating certain types of eating disorder symptoms istered in the context of stepped care treatment

351
or independently, have the potential advantage strategies including monitoring eating behavior,
of increasing treatment accessibility to eating dis- planning meals and snacks, identifying behavioral
order patients because of scalability and reduced alternatives to binge eating, reducing dietary res-
treatment costs. As summarized by Perkins, traint, and using problem-​ solving strategies (of
Murphy, Schmidt, and Williams (2006), these note, an updated version of this manual includes
approaches may also increase self-​efficacy, reduce additional information about weight and shape con-
treatment delays, minimize barriers to treatment cerns; Fairburn, 2013). Earlier research using the
associated with shame related to eating disorder Fairburn (1995) manual compared “pure” SH with
symptoms, reinforce and consolidate learning, CBTgsh among participants with bulimia nervosa
and allow the patient to access treatment resources (BN), binge eating disorder (BED), or mixed diag-
in the case of symptom lapse or relapse (Perkins nostic samples (Carter & Fairburn, 1998; Ghaderi &
et al., 2006). Alternatively, limitations of SH and Scott, 2003; Ghaderi, 2006; Loeb, Wilson, Gilbert,
GSH include the risk of medical instability and/​or & Labouvie, 2000; Palmer, Birchall, McGrain, &
suicidality with intermittent or minimal clinical Sullivan, 2002). Although no significant differences
interaction; the potential inaccessibility of reading between groups were observed in several of these
materials to individuals with learning disabilities, studies (see treatment review section below), Loeb
cognitive impairments, or language barriers; and et al. (2000) observed that GSH had a significantly
patient preference for treatments that incorporate greater reduction in binge eating frequency in com-
more frequent clinical contact. parison with SH, and two studies found higher
attrition in SH compared with GSH (Carter, 1998;
Palmer et al., 2002).
Self-​Help and Guided Self-​Help Treatment
Randomized studies comparing GSH using the
Although emerging interventions incorporate
Fairburn (1995; CBTgsh) manual with other treat-
technology-​ based sources of treatment delivery
ments have found that CBTgsh is generally com-
(see c­ hapters 27 and 28), the majority of research
parable or superior to other treatment conditions.
investigating SH and GSH for eating disorders
Striegel-​Moore et  al. (2010) found that CBTgsh
treatment has relied on written information in the
was associated with better binge eating outcome
form of books and has been administered indi-
compared with treatment as usual in a health main-
vidually. In most SH studies, participants are pro-
tenance organization setting. Grilo and Masheb
vided with treatment materials that they read and
(2005) observed that CBTgsh produced signifi-
review on their own. In contrast, GSH treatment
cant improvements in binge eating remission com-
research has typically involved a nonspecialist cli-
pared with SH behavioral weight loss and also had
nician meeting briefly with the participant to pro-
fewer dropouts. Wilson, Wilfley, Agras, and Bryson
vide support, encouragement, and information as
(2010) found that CBTgsh produced comparable
an adjunct to the written SH materials. Several
reductions in binge eating compared with interper-
research trials have incorporated SH and GSH
sonal therapy (IPT) at 2-​year follow-​up and that
into stepped care designs. Each of these research
both of these treatments were superior to behavioral
areas is described in what follows, along with rec-
weight loss (Wilson et al., 2010).
ommendations for future directions and clinical
In a primary care study, 104 obese patients with
implications.
BED were randomized to sibutramine, placebo, SH
using the Fairburn (1995) manual (shCBT) plus
Overcoming Binge Eating
placebo, or shCBT plus sibutramine (Grilo et  al.,
The SH treatment manual that has received the
2014). Although the shCBT group had lower binge
most empirical support is Overcoming Binge Eating
eating frequency at the 6-​month assessment, no
(Fairburn, 1995), which can be delivered either as
other treatment differences were observed and treat-
pure SH or GSH (typically referred to as CBTgsh
ment effects were modest. These findings replicated
when GSH used with this particular manual) with
an earlier study showing modest effects of shCBT
a trained practitioner reviewing progress and out-
for binge eating in primary care (Grilo, White,
comes. Adapted from cognitive-​behavioral therapy
Gueorguieva, Barnes, & Masheb, 2013). Similarly,
(CBT; Fairburn, Marcus, & Wilson, 1993), the self-​
in another primary care study involving participants
help manual includes two sections. The first pro-
with BN that compared fluoxetine or placebo with
vides psychoeducational information about binge
or without CBTgsh, attrition for the CBTgsh con-
eating, and the second specifies cognitive-​behavioral
dition was extremely high (71%; Walsh, Fairburn,

352 Self-Help and Stepped Care Treatments

Mickley, Sysko, & Parides, 2004). Although out- At end of treatment, therapist-​led and therapist-​
come for fluoxetine was better than placebo, assisted conditions were associated with higher
CBTgsh did not appear to provide any additive binge eating abstinence rates than the SH con-
effect to medication for treatment outcome. dition, and all three groups were superior to the
wait-​list control condition; however, there were
Getting Better Bit(e) by Bit(e) 1993 no differences between the treatment conditions
Several earlier trials compared a different SH at 6-​month or 12-​month follow-​up. In addition,
treatment manual, Getting Better Bit(e) by Bit(e) the therapist-​led group had a significantly higher
(Schmidt & Treasure, 1993) in randomized designs. proportion of participants complete the treatment
Treasure and colleagues (1994) compared this SH compared with therapist-​assisted and SH groups
manual to CBT and wait-​list control in a sample (Peterson et al., 2009). Although the inclusion of
of 81 adults with symptoms of BN. The CBT a therapist produced better treatment response ini-
condition was associated with more reductions in tially, this difference was not sustained at follow-​
self-​induced vomiting than the SH condition, but up, and abstinence rates were modest.
both treatments showed comparable reductions in
binge eating and other eating disorder symptoms as Systematic Reviews and Meta-​Analyses
well as remission rates (22%–​24% compared with Perkins et  al. (2006) conducted a review of 15
11% of the wait-​list group; Treasure et al., 1994). studies using SH and GSH among adults with
In a follow-​up sequential treatment study for BN BED, BN, and eating disorder not otherwise speci-
(Treasure et al., 1996), the SH treatment followed fied (EDNOS). The authors found that when ana-
by an abridged CBT (if needed) was compared with lyzed as a group, these treatments did not differ
a standard CBT treatment. No differences were significantly from wait-​list conditions in symptom
observed in outcome between the two conditions, abstinence at end of treatment. However, SH and
and remission rates were comparable (30%). This GSH were associated with significant reductions
treatment manual was also used in a randomized in eating disorder symptoms, interpersonal func-
trial for adolescent BN comparing GSH with family tioning, and some comorbid psychiatric symptoms
therapy based on the Maudsley approach (Schmidt (not including depression). The authors concluded
et al., 2007). Although GSH was preferred by the that, compared with other treatments including
adolescent participants and associated with better therapist-​delivered psychotherapies, SH and GSH
outcome at the 6-​month assessment, no differences did not differ significantly at end of treatment or
between the two treatments were observed at 1-​ follow-​up on any outcome measure including attri-
year follow-​up, with abstinent rates ranging from tion (Perkins, et al., 2006). The same year, Stefano,
36% to 41%; direct costs of treatment were lower Bacaltchuk, Blay, and Hay (2006) published a
for GSH. meta-​analysis of 9 SH and GSH studies for BN
and BED that met inclusion criteria for relevance
Psychoeducational Videotapes and quality. In contrast to the Perkins et al. (2006)
Although most SH and GSH studies have review, the authors concluded that participants
used individual delivery models, a group design who received any type of SH or GSH had better
was used in a comparison study conducted by outcomes compared with wait-​list control condi-
Peterson, Mitchell, Crow, Crosby, and Wonderlich tions (Stefano et al., 2006). However, GSH was not
(2009) based on the rationale that groups have clearly superior to SH.
the advantages of interpersonal support and fur- Both the Perkins et al. (2006) and Stefano et al.
ther cost reduction. Participants with BED were (2006) reviews cited significant limitations in the
randomized to SH, therapist-​ led, or therapist-​ research literature. Although these reviews were
assisted group CBT (or wait-​list). The therapist-​ conducted a decade ago, many of these criticisms
led condition began with the therapist providing remain relevant, including the exclusion of males
psychoeducation during the first half of each ses- from research trials, the inconsistencies in clinician
sion followed by homework review and discussion. training, and the fact that many of these studies
The therapist-​ assisted and SH groups watched (particularly the earlier ones) are underpowered. In
a psychoeducational video during the first half addition, the inconsistencies in the findings between
and participated in either a therapist-​led (for the the two reviews may be the result of combining
therapist-​
assisted condition) or self-​
led (for the studies using different types of self-​help manuals,
SH condition) discussion and homework review. treatment designs, and assessment measures.

Peterson, Pisetsky, Haut 353

Treatment Predictors, Moderators, and which any group differences are enduring is unclear.
Mediators Similarly, although self-​help approaches appear to
The identification of treatment moderators be comparable to other treatments when compared
and predictors of outcome has the potential of in randomized trials, these findings are not clearly
increasing efficacy by pairing specific treatments robust. An additional uncertainty is the extent to
with individuals who are most likely to respond to which SH and GSH are acceptable to patients with
them. Identifying treatment mediators may facili- eating disorders and if they are associated with
tate enhanced efficacy through the understanding higher rates of attrition.
and targeting of mechanisms of action. Recently, Nonetheless, overall, the research literature
Hilbert, Hildebrandt, Agras, Wilfley, and Wilson suggests that SH and GSH are promising for the
(2015) demonstrated that in contrast to behavioral outpatient treatment of binge eating and bulimic
weight loss and IPT in a randomized trial, rapid symptoms among adults and, possibly, adolescents.
response, defined as greater than or equal to 70% One clearly robust finding is that research partici-
reduction in binge eating by week 4, among partici- pants who do not respond to CBT or CBTgsh as
pants receiving CBTgsh predicted short-​and long-​ evidenced by a significant reduction in eating dis-
term treatment outcome in BED (Hilbert et  al., order behaviors (i.e., binge eating and/​or purging)
2015). Similarly, rapid response was the only signif- by the end of the first month in treatment should
icant predictor of treatment outcome among adult be offered alternative or adjunctive treatments, as
females with bulimic symptoms receiving CBTgsh described below in the stepped care section. Other
(Vaz, Conceição, & Machado, 2014). Masheb and robust predictors and moderators for SH and GSH
Grilo (2007) observed that rapid response was a have not been identified. Another notable finding is
significant predictor of outcome for CBTgsh and that a minority and, perhaps, a sizable minority of
behavioral weight loss in terms of eating disorder outpatients with eating disorder symptoms charac-
pathology and binge eating remission; however, terized by binge eating (with or without purging)
binge eating frequency was comparable between may remit with SH or GSH without the need for
rapid responders and nonrapid responders in the additional care.
CBTgsh condition. In a follow-​up analyses of their Finally, the findings of the Wilson et al. (2010)
2014 primary care study, Grilo, White, Masheb, study are particularly important in highlighting
and Gueorguieva (2015) found that rapid response that the 10-​session CBTgsh treatment produced
was a robust predictor of outcome and suggested comparable overall findings to a longer and more
that these findings support the use of self-​help and time intensive psychotherapy treatment, IPT, and
medication treatment in the context of stepped that these effects endured through follow-​up. The
care designs, in spite of their modest outcome in implications are particularly important for cost-​
the main treatment outcome analyses (Grilo et al., effectiveness as well as treatment access and poten-
2015). In their treatment comparison trial, Wilson tial scalability.
et al. (2010) found that CBTgsh was less efficacious
for participants with greater eating disorder symp- Stepped Care Treatment
toms combined with low self-​esteem at baseline, The most rigorous BN stepped care study con-
who responded better to IPT. Using the same data ducted in the eating disorders field is a multisite
set, Sysko, Hildebrandt, Wilson, Wilfley, and Agras randomized trial in which participants were ran-
(2010) conducted an empirical classification using domized to one of two conditions (Mitchell et al.,
latent class analysis and observed that CBTgsh had 2011). In the CBT condition, participants received
the highest efficacy for the class characterized by individual CBT (Fairburn et al., 1993) followed by
a greater frequency of binge eating episodes along the offer of adjunctive fluoxetine if they were still
with lower frequency of other eating disorder symp- symptomatic after 6 sessions. In the stepped-​care
toms including purging (Sysko et al., 2010). condition, CBTgsh was followed by adjunctive
fluoxetine and, if needed, individual CBT. Although
Summary remission rates were comparable at 1-​year follow-​
In summary, the results of the SH and GSH up for CBT (44%) and stepped care (32%), the
studies for the treatment of bulimic and binge eat- stepped care condition had more significant reduc-
ing symptoms are not straightforward. Although lit- tions in binge eating and compensatory behavior as
erature reviews often conclude that GSH is superior well as comorbid symptoms including depression.
to SH, the data are inconsistent and the extent to In addition, participants classified as treatment

354 Self-Help and Stepped Care Treatments

nonresponders at week 4 had higher abstinence therapy. Additional proposed alternative/​adjunctive
rates in the stepped care (25%) than the CBT treatments include mindfulness, appetite aware-
(4%) condition. The stepped care treatment was ness training, and exercise as well as behavioral
also more cost-​ effective (Crow, Agras, Fairburn, weight loss following remission from binge eat-
Mitchell, & Nyman, 2013). This study highlights ing (Iacovino, Gredysa, Altman, & Wilfley, 2012).
the potential value of stepped care treatment for Other potential models include the use of the
BN, especially the importance of providing alterna- overvaluation of shape and weight as a specifier for
tive and/​or adjunctive treatment after 1 month for BED (Grilo et al., 2008) and the overvaluation of
individuals receiving cognitive-​behavioral interven- shape and weight combined with low self-​esteem
tions who are not exhibiting significant reductions (Iacovino et al., 2012).
in symptoms.
These findings are consistent with earlier work Summary
demonstrating that sequencing full IPT or medica- For over three decades, researchers and clinicians
tion trials for individuals who were not responsive in the field of eating disorders have been emphasiz-
to a full sequence of CBT did not yield significant ing the potential utility of stepped care approaches
improvement in treatment outcome and was asso- for treatment. The rationale for the use of stepped
ciated with high rates of attrition (Mitchell et  al., care treatment is well founded, considering the
2002). Most other stepped care treatment research potential for improved treatment outcome, reduced
in eating disorders has not involved randomized clinical burden, and increased cost-​effectiveness. In
designs. In a description of a clinically based stepped addition, the robust finding that nonresponse to
care model for bulimic symptoms in a specialized CBT and CBTgsh after the first month of treatment
clinic, for example, Ramklint and colleagues observed is predictive of short-​and longer-​term outcome is
that the majority of the participants were deemed particularly useful for stepped care designs in which
suitable upon screening to receive CBTgsh initially cognitive-​behavioral interventions are provided as
and that the overall treatment outcome effect sizes the first step. Notably, however, limited data using
were relatively larger for those who completed more randomized designs are available to support the use
CBTgsh sessions (Ramklint, Jeansson, Holmgren, & of specific stepped care models for eating disorders,
Ghaderi, 2012). Although attrition was significant, and most of the proposed stepped care models have
the authors suggest that CBTgsh may be helpful for been speculative or based on preliminary clinical
approximately 30% of their sample. data. Clearly, more efficacy and effectiveness data
Emerging data support stepped care approaches are needed to determine optimal stepped care inter-
for BED treatment (Amianto, Ottone, Daga, & ventions for eating disorder treatment.
Fassino, 2015). As noted by Hilbert et al. (2015),
CBTgsh followed by interpersonal therapy for indi- Future Directions
viduals who do not show rapid response (as defined Based on current research and clinical care, future
by 70% or greater reduction in binge eating by directions for eating disorders treatment using SH,
week 4) may be an optimal stepped care design for GSH, and stepped care include five categories. The
the treatment of binge eating (Hilbert et al., 2015). first research and clinical priority is improving out-
come by advancing individualized care. Greater
Specific Stepped Care Models access to “big data” as well as other data resources
Most proposals of stepped care models for eat- will optimize the identification of algorithms to
ing disorders advocate for CBT/​ CBTgsh as the determine potential sequencing models for indi-
first step in treatment (e.g., Dalle Grave, Ricca, viduals with eating disorders in clinical settings. In
& Todesco, 2001) and suggest incorporating the addition, state-​of-​the-​art research designs for adap-
prediction of response or nonresponse to CBT or tive treatment include sequential multiple assign-
CBTgsh after the first month of treatment to guide ment for randomized treatment (SMART) and
treatment selection, with adjunctive or alternative other tailored treatment models that are responsive
treatment provided based on degree of symptom to both baseline patient characteristics and initial
response (e.g., +/​-​65%–​70% reduction in binge treatment response can potentially increase efficacy
eating). Consistent with Mitchell and colleagues while reducing costs (Norcross & Wampold, 2011).
(2011), adjunctive treatment may include medi- Incorporating SH/​ GSH interventions into these
cation. Alternative treatments after CBT/​CBTgsh designs, along with technology-​based components,
nonresponse include IPT or dialectical behavior may be especially effective and is clearly needed

Peterson, Pisetsky, Haut 355

to advance eating disorder treatment. In addition, predicted nonresponse after the first month of treat-
advances in the use of neurobiological variables and ment) that do not include self-​help components,
genotypes, along with behavioral phenotypes, in which may not be appropriate for individuals who
determining individual patient characteristics in tai- are underweight and potentially medically com-
loring treatment may result in substantial improve- promised. Finally, research is needed to determine
ments in treatment effectiveness. Notably, these optimal training models for clinicians and parapro-
identifiers are likely to be “transdiagnostic,” given fessionals to deliver these types of treatments, as well
the limitations in the current psychiatric nosology. as determining the scalability of these types of train-
The second priority is scalability and dissemination ing programs (Wilson, 2011). Zandberg and Wilson
(Wilson & Zandberg, 2012). The use of SH and (2013) describe optimal training for GSH clini-
GSH interventions bypasses a number of barriers cians, as well as the possibility that this treatment
to mental health treatment in general and eating can be delivered effectively by individuals without
disorder treatment in particular. Applying princi- advanced degrees. Clearly, training models need to
ples of implementation science (e.g., Betancourt & be examined empirically, along with optimal char-
Chambers, 2016) will be essential for maximizing acteristics of GSH clinicians. Within the context of
and optimizing dissemination in the context of SH/​ training and scalability, risk assessment for medical
GSH as well as stepped care models. Evidence-​based instability and suicidality is crucial.
approaches to dissemination can be used to guide
public health policy treatment and prevention of Conclusions
eating disorders and comorbid conditions (Becker, The literature suggests that both SH and GSH
Plasencia, Kilpela, Briggs, & Stewart, 2014; Paxton, are promising for the outpatient treatment of binge
2013). In addition, overlapping with the first two eating and bulimic symptoms and some patients
categories, cost effectiveness (see ­chapter  22) is a may remit with SH or GSH without the need for
critical consideration in the advancement of SH/​ additional treatment. Thus, SH and GSH provide
GSH and stepped care treatment. Although cost treatment options for individuals with binge eating
savings is presumed in these types of treatment, rig- that appear to be cost-​effective and highly scalable,
orous data are needed to demonstrate the impact of which is crucial given the difficulties many patients
these types of interventions on healthcare costs and experience accessing specialist care. However, more
to build a foundation for adoption of these treat- work is needed to increase patient acceptability and
ments into clinical settings. In addition to health- retention in these types of treatments. Additionally,
care costs, research on the use of SH/​ GSH and the research on CBT-​ based approaches indicates
stepped care treatment for eating disorders should that patients who do not exhibit a significant reduc-
use a broader range of outcome variables beyond tion in eating disorder behaviors by the end of the
eating disorder symptoms including quality of life, first month should be offered alternative or adjunc-
health outcomes, clinician preferences, and patient tive treatments such as psychopharmacology or a
acceptability (Peterson, Becker, Treasure, Shafran, & different psychotherapeutic modality. More research
Bryant-​Waugh, 2016). The issue of patient accep- in adaptive and tailored treatment designs is needed
tance is particularly important given the lower rat- to determine the optimal interventions and to pro-
ings of CBTgsh compared to IPT in a randomized vide individualized treatment in a stepped-​ care
trial (Wilson, et  al., 2010), although other studies approach. Overall, this body of research indicates
have found higher levels of participant acceptability that SH, GSH, and stepped-​care approaches are
of GSH (e.g., Mitchell, et al., 2011; Striegel-​Moore, empirically supported alternatives to the traditional
et al., 2010). An additional limitation in the current full course of psychotherapy and pharmacotherapy
research is the paucity of stepped care studies exam- delivered in specialist clinical settings, which will
ining transdiagnostic and anorexia nervosa patients. allow more patients to access treatment and provid-
One study that was conducted with anorexia ner- ers to determine effective treatments more quickly.
vosa indicates that stepped care approaches may be
used based on patient preference to increase partici- References
pation in intensive treatment (Tasca et  al., 2012). Amianto, F., Ottone, L., Daga, G. A., & Fassino, S. (2015).
Similarly, the use of evidence-​based approaches such Binge-​eating disorder diagnosis and treatment:  A  recap in
as family-​based treatment for anorexia nervosa may front of DSM-​5. BMC Psychiatry, 15, 1.
be appropriate for adaptation to stepped care models Becker, C. B., Plasencia, M., Kilpela, L. S., Briggs, M., &
Stewart, T. (2014). Changing the course of comorbid eating
(e.g., adjunctive treatment could be considered with

356 Self-Help and Stepped Care Treatments

 disorders and depression: What is the role of public health care:  The significance of early rapid response. Journal of
interventions in targeting shared risk factors? Journal of Consulting and Clinical Psychology, 83, 387.
Eating Disorders, 2, 1. Hilbert, A., Hildebrandt, T., Agras, W. S., Wilfley, D. E., &
Betancourt, T. S., & Chambers, D. A. (2016). Optimizing an Wilson, G. T. (2015). Rapid response in psychological treat-
era of global mental health implementation science. JAMA ments for binge eating disorder. Journal of Consulting and
Psychiatry, 73, 99–​100. Clinical Psychology, 83, 649.
Carter, J. C., & Fairburn, C. G. (1998). Cognitive–​behavioral Iacovino, J. M., Gredysa, D. M., Altman, M., & Wilfley, D. E.
self-​help for binge eating disorder: A controlled effectiveness (2012). Psychological treatments for binge eating disorder.
study. Journal of Consulting and Clinical Psychology, 66, 616. Current Psychiatry Reports, 14, 432–​446.
Crow, S. J., Agras, W. S., Fairburn, C. G., Mitchell, J. E., & Loeb, K. L., Wilson, G. T., Gilbert, J. S., & Labouvie, E. (2000).
Nyman, J. A. (2013). A cost effectiveness analysis of stepped Guided and unguided self-​help for binge eating. Behaviour
care treatment for bulimia nervosa. International Journal of Research and therapy, 38, 259–​272.
Eating Disorders, 46, 302–​307. Masheb, R. M., & Grilo, C. M. (2007). Rapid response predicts
Cuijpers, P., Donker, T., van Straten, A., Li, J., & Andersson, G. treatment outcomes in binge eating disorder: Implications
(2010). Is guided self-​help as effective as face-​to-​face psycho- for stepped care. Journal of Consulting and Clinical Psychology,
therapy for depression and anxiety disorders? A  systematic 75, 639–644.
review and meta-​analysis of comparative outcome studies. Mitchell, J. E., Agras, S., Crow, S., Halmi, K., Fairburn, C. G.,
Psychological Medicine, 40, 1943–​1957. Bryson, S., & Kraemer, H. (2011). Stepped care and cognitive-​
Dalle Grave, R., Ricca, V., & Todesco, T. (2001). The behavioural therapy for bulimia nervosa: Randomised trial.
stepped-​ care approach in anorexia nervosa and bulimia British Journal of Psychiatry, 198, 391–​397.
nervosa: Progress and problems. Eating and Weight Disorders-​ Mitchell, J. E., Halmi, K., Wilson, G. T., Agras, W. S., Kraemer,
Studies on Anorexia, Bulimia and Obesity, 6, 81–​89. H., & Crow, S. (2002). A randomized secondary treatment
Fairburn, C. G. (1995). Overcoming binge eating. New  York, study of women with bulimia nervosa who fail to respond to
NY: Guilford Press. CBT. International Journal of Eating Disorders, 32, 271–​281.
Fairburn, C. G. (2013). Overcoming binge eating: The proven pro- National Institute for Clinical Excellence (NICE). (2004).
gram to learn why you binge and how you can stop. New York. Eating disorders: Core interventions in the treatment and man-
NY: Guilford Press. agement of anorexia nervosa, bulimia nervosa and related eating
Fairburn, C. G., Marcus, M. D., & Wilson, G. T. (1993). disorders. Leicester UK: British Psychological Society & Royal
Cognitive-​behavioral therapy for binge-​eating:  A  compre- College of Psychiatrists.
hensive treatment manual. In G. T. Wilson & C. G. Fairburn Norcross, J. C., & Wampold, B. E. (2011). What works for
(Eds.), Binge eating:  Nature, assessment, and treatment (pp. whom:  Tailoring psychotherapy to the person. Journal of
361–​404). New York, NY: Guilford Press. Clinical Psychology, 67, 127–​132.
Fairburn, C. G., & Peveler, R. C. (1990). Bulimia nervosa and a Palmer, R. L., Birchall, H., McGrain, L., & Sullivan, V. (2002).
stepped care approach to management. Gut, 31, 1220–​1222. Self-​help for bulimic disorders: A randomised controlled trial
Ghaderi, A. (2006). Attrition and outcome in self-​help treatment comparing minimal guidance with face-​to-​face or telephone
for bulimia nervosa and binge eating disorder: A construc- guidance. British Journal of Psychiatry, 181, 230–​235.
tive replication. Eating Behaviors, 7, 300–​308. Paxton, S. J. (2013). Dissemination in the Internet age: Taming
Ghaderi, A., & Scott, B. (2003). Pure and guided self‐help for a wild thing. International Journal of Eating Disorders, 46,
full and sub‐threshold bulimia nervosa and binge eating dis- 525–​528.
order. British Journal of Clinical Psychology, 42, 257–​269. Perkins, S. S., Murphy, R. R., Schmidt, U. U., & Williams, C.
Grilo, C. M., Hrabosky, J. I., White, M. A., Allison, K. C., (2006). Self‐help and guided self‐help for eating disorders.
Stunkard, A. J., & Masheb, R. M. (2008). Overvaluation of Cochrane Database of Systematic Reviews, 3, CD004191.
shape and weight in binge eating disorder and overweight Peterson, C. B., Becker, C. B., Treasure, J., Shafran, R., & Bryant-​
controls:  Refinement of a diagnostic construct. Journal of Waugh, R. (2016). The three-​legged stool of evidence-​based
Abnormal Psychology, 117, 414. practice in eating disorder treatment: research, clinical, and
Grilo, C. M., & Masheb, R. M. (2005). A randomized con- patient perspectives. BMC Medicine, 14, 1.
trolled comparison of guided self-​help cognitive behavioral Peterson, C. B., Mitchell, J. E., Crow, S. J., Crosby, R. D., &
therapy and behavioral weight loss for binge eating disorder. Wonderlich, S. A. (2009). The efficacy of self-​help group
Behaviour Research and Therapy, 43, 1509–​1525. treatment and therapist-​led group treatment for binge eating
Grilo, C. M., Masheb, R. M., White, M. A., Gueorguieva, disorder. American Journal of Psychiatry, 166, 1347–​1354.
R., Barnes, R. D., Walsh, B. T.,   .  .  .  Garcia, R. (2014). Ramklint, M., Jeansson, M., Holmgren, S., & Ghaderi, A.
Treatment of binge eating disorder in racially and ethni- (2012). Guided self‐help as the first step for bulimic symp-
cally diverse obese patients in primary care:  Randomized toms: Implementation of a stepped‐care model within spe-
placebo-​controlled clinical trial of self-​help and medication. cialized psychiatry. International Journal of Eating Disorders,
Behaviour Research and Therapy, 58, 1–​9. 45, 70–​78.
Grilo, C. M., White, M. A., Gueorguieva, R., Barnes, R. D., & Schmidt, U., Lee, S., Beecham, J., Perkins, S., Treasure, J., Yi,
Masheb, R. M. (2013). Self-​help for binge eating disorder in I.,  . . .  Johnson-​Sabine, E. (2007). A randomized controlled
primary care: A randomized controlled trial with ethnically trial of family therapy and cognitive behavior therapy guided
and racially diverse obese patients. Behaviour Research and self-​care for adolescents with bulimia nervosa and related dis-
Therapy, 51, 855–​861. orders. American Journal of Psychiatry, 164, 591–​598.
Grilo, C. M., White, M. A., Masheb, R. M., & Gueorguieva, Schmidt, U, & Treasure, J. (1993). Getting better bit(e) by bit(e).
R. (2015). Predicting meaningful outcomes to medication A survival kit for sufferers of bulimia nervosa and binge eating
and self-​help treatments for binge-​eating disorder in primary disorder. Hove, UK: Brunner-​Routledge.

Peterson, Pisetsky, Haut 357

Stefano, S. C., Bacaltchuk, J., Blay, S. L., & Hay, P. (2006). nervosa incorporating a self-​care manual. British Journal of
Self‐help treatments for disorders of recurrent binge eat- Psychiatry, 168, 94–​98.
ing: A systematic review. Acta Psychiatrica Scandinavica, 113, Vaz, A. R., Conceição, E., & Machado, P. P. (2014). Early
452–​459. response as a predictor of success in guided self‐help treat-
Striegel-​Moore, R. H., Wilson, G. T., DeBar, L., Perrin, N., ment for bulimic disorders. European Eating Disorders
Lynch, F., Rosselli, F., & Kraemer, H. C. (2010). Cognitive Review, 22, 59–​65.
behavioral guided self-​help for the treatment of recurrent Walsh, B. T., Fairburn, C. G., Mickley, D., Sysko, R., &
binge eating. Journal of Consulting and Clinical Psychology, Parides, M. K. (2004). Treatment of bulimia nervosa in a
78, 312. primary care setting. American Journal of Psychiatry, 161,
Sysko, R., Hildebrandt, T., Wilson, G. T., Wilfley, D. E., & 556–​561.
Agras, W. S. (2010). Heterogeneity moderates treatment Wilson, G. T. (2011). Treatment of binge eating disorder.
response among patients with binge eating disorder. Journal Psychiatric Clinics of North America, 34, 773–​783.
of Consulting and Clinical Psychology, 78, 681. Wilson, G. T., Vitousek, K. M., & Loeb, K. L. (2000). Stepped
Tasca, G. A., Keating, L., Maxwell, H., Hares, S., Trinneer, A., care treatment for eating disorders. Journal of Consulting and
Barber, A. M.,  . . .  Bissada, H. (2012). Predictors of treat- Clinical Psychology, 68, 564.
ment acceptance and of participation in a randomized con- Wilson, G. T., Wilfley, D. E., Agras, W. S., & Bryson, S. W.
trolled trial among women with anorexia nervosa. European (2010). Psychological treatments of binge eating disorder.
Eating Disorders Review, 20, 155–​161. Archives of General Psychiatry, 67, 94–​101.
Treasure, J., Schmidt, U., Troop, N., Tiller, J., Todd, G., Keilen, Wilson, G. T., & Zandberg, L. J. (2012). Cognitive-​behavioral
M., & Dodge, E. (1994). First step in managing bulimia guided self-​help for eating disorders: Effectiveness and scal-
nervosa: Controlled trial of therapeutic manual. BMJ, 308, ability. Clinical Psychology Review, 32, 343–​357.
686–​689. Zandberg, L. J., & Wilson, G. T. (2013). Train-​ the-​
Treasure, J., Schmidt, U., Troop, N., Tiller, J., Todd, G., & trainer:  Implementation of cognitive-​ behavioural guided
Turnbull, S. (1996). Sequential treatment for bulimia self-​help for recurrent binge eating in a naturalistic setting.
European Eating Disorders Review, 21, 230–​237.

358 Self-Help and Stepped Care Treatments

 CH A PT E R


Pharmacotherapy for Eating Disorders
19
Susan L. McElroy, Anna I. Guerdjikova, Nicole Mori, and Paul E. Keck Jr.

Abstract
This chapter addresses the pharmacotherapy of the eating disorders (EDs). Many persons with EDs
receive pharmacotherapy, but pharmacotherapy research for EDs has lagged behind that for other major
mental disorders. This chapter first provides a brief rationale for using medications in the treatment of
EDs. It then reviews the data supporting the effectiveness of specific medications or medication classes
in treating patients with anorexia nervosa (AN), bulimia nervosa (BN), binge eating disorder (BED), and
other potentially important EDs, such as night eating syndrome (NES). It concludes by summarizing these
data and suggesting future areas for research in the pharmacotherapy of EDs.
Key Words:  antidepressant, antiepileptic, drug, antipsychotic, eating disorder, mood stabilizer,
pharmacotherapy, stimulant, weight, loss

Introduction (NES). We then summarize these data and suggest

Many persons with eating disorders (EDs) future areas for research.
receive pharmacotherapy (Fazeli et  al., 2012;
Grigoriadis, Kaplan, Carter, & Woodside, 2001; Rationale for Using Pharmacotherapy
Mond, Hay, Rodgers, & Owen, 2007; Walsh et al., in Treating Eating Disorders
2006). An increasing number of reviews (Yager & There are several rationales for using pharmaco-
Powers, 2007) and guidelines (Aigner et al., 2011; therapy to treat EDs. First, EDs are major mental
American Psychiatric Association [APA], 2006; disorders with genetic etiologic contributions and
Hay et  al., 2014; National Institute for Clinical neurobiological abnormalities that do not always
Excellence, 2004) summarize the role pharmaco- respond adequately to available psychological inter-
therapy might play in treating patients with EDs. ventions (Bulik et al., 2007; Slof-​Op’t Landt et al.,
However, only two drugs have regulatory approval 2005). Some patients, including those with chronic
for the treatment of an ED (fluoxetine for bulimia or intractable illness, may need medication for
nervosa [BN] and lisdexamfetamine dimesylate optimal outcomes (Yager, 2007). Second, medica-
[LDX] for binge eating disorder [BED]), and phar- tions that are effective in conditions related to EDs
macotherapy research for EDs lags behind that for might be effective in EDs themselves, or might be
other major mental disorders. Indeed, no drug has useful in managing co-​occurring conditions in ED
regulatory approval for the treatment of anorexia patients (Woodside & Staab, 2006). Third, many
nervosa (AN), and no drug has yet been specifically medications have effects on appetite and weight,
developed to treat an ED. as well as the central or peripheral systems impor-
In this chapter, we review research with spe- tant in regulating eating behavior and weight con-
cific medications or medication classes in treating trol. These include drugs being developed for mood,
patients with AN, BN, BED, and other potentially psychotic, and other mental disorders; cachexia and
important EDs, such as night eating syndrome obesity; and epilepsy. Some of these agents may

359
prove to have beneficial psychotropic properties & Cohen, 1986). Seventeen (74%) amitriptyline-​
and/​or useful effects on appetite or weight regula- treated patients and 16 (64%) placebo-​ treated
tion, and might therefore have therapeutic effects in patients achieved target weight. Among patients
ED patients, including the substantial portion who who achieved target weight, excluding noncom-
are inadequately responsive to current treatments. pleters, the daily rate of weight increase was numeri-
cally, but not statistically significantly, higher in the
Pharmacotherapy of Anorexia Nervosa amitriptyline group. Significantly fewer days were
Two primary randomized controlled trial (RCT) needed to achieve target weight with amitriptyline
designs have been used to evaluate medications in than placebo. Attrition was 30% for amitriptyline
AN:  studies aimed to restore weight in acutely ill and 20% for placebo. In the third trial, clomip-
underweight patients (weight restoration trials) and ramine 50 mg/​day was not associated with greater
those aimed to maintain weight gain in patients weight gain than placebo in 16 female inpatients
whose weight has been restored to some degree with AN after an 11-​week acute phase of treatment,
(relapse prevention or weight maintenance trials). after which medication was discontinued, or at 1-​
The primary outcome in the former has usually year and 4-​year follow-​up evaluations (Crisp, Lacey,
been a measure of weight gain or time to a certain & Crutchfield, 1987). In the fourth trial, 31 female
amount of weight gain; in the latter, it often has inpatients with AN (ages 16–​45  years) who had
been time to or rate of relapse. However, varied pri- achieved 65% of ideal body weight were randomly
mary outcome measures have been used in both assigned to fluoxetine, up to 60 mg/​day, or pla-
types of trials. Secondary outcomes have included cebo for 7 weeks on a clinical research unit (Attia,
measures of ED psychopathology, depressive and/​ Haiman, Walsh, & Flater, 1998). Average (SD) body
or anxiety symptoms, overall clinical improvement, mass index (BMI) at randomization was 15 mg/​kg
and treatment acceptability. Various biomarkers in (4.2). Mean (SD) fluoxetine dose at treatment end-
addition to weight have also been assessed, such as point was 56.0 (11.2) mg/​day. There were no sig-
vital signs, endocrine parameters, gastric emptying, nificant differences between fluoxetine and placebo
menstrual function, and bone density. Importantly, on weight gain, ED psychopathology, obsessive-​
in most studies, pharmacotherapy was given in con- compulsive symptoms, measures of depression or
junction with inpatient, supportive, and/​ or spe- anxiety, or clinical global improvement.
cialist psychotherapeutic treatment, which could In 2006, Claudino et al. published a Cochrane
mitigate any efficacy signal from medication. review evaluating evidence of efficacy and accept-
The medications most commonly evaluated in ability of antidepressant treatment for weight
AN for weight restoration in RCTs have been anti- restoration in AN from seven RCTs: the four above-​
depressants, antipsychotics, and appetite stimulants noted studies that compared antidepressants to
(Table 19.1). The only medications evaluated in AN placebo (Attia et al., 1998; Biederman et al., 1985;
for weight maintenance in RCTs have been fluox- Halmi et al., 1986; Lacey & Crisp, 1980) and three
etine and recombinant human growth hormone studies that compared different antidepressant drugs
(rhGH). Other medications evaluated in RCTs with one another (Brambilla, Draisci, Peirone, &
include prokinetics, zinc, hormonal agents, ben- Brunetta, 1995a, 1995b; Ruggiero et  al., 2001).
zodiazepines, relamorelin, lithium, opioid antago- Due to methodological limitations, aggregation of
nists, and d-​cycloserine. data for meta-​analysis was not possible for most out-
comes. However, it was concluded that the studies
Antidepressants were not able to show any effect of antidepressants
At least four randomized, placebo-​ controlled compared to placebo in the majority of outcomes,
trials of antidepressants in acutely ill, underweight including weight gain, ED symptoms, associated
patients with AN have been published (Claudino anxious and depressive symptoms, or clinical global
et  al., 2006). In the first trial, amitriptyline, up improvement. In the three comparative studies,
to 175 mg/​day, did not differ from placebo in 25 the only findings were a greater effect for aminep-
youth, ages 11 to 17  years, on weight, eating, or tine (an atypical tricyclic that selectively inhibits
mood outcomes (Biederman et  al., 1985). There dopamine and, to a lesser extent, norepinephrine
were no dropouts. In the second trial, 72 female reuptake) compared to fluoxetine in reducing end-​
inpatients with AN, ages 13 to 26 years, were ran- of-​
treatment Eating Disorder Inventory (EDI)
domized to amitriptyline (n = 23), cyproheptadine scores, and a greater effect of nortriptyline com-
(n = 24), or placebo (n = 25) (Halmi, Eckert, LaDu, pared to fluoxetine in decreasing mean Hamilton

360 Pharmacotherapy
Table 19.1  Medications Studied for Anorexia Nervosa in Randomized, Placebo-​Controlled Trials: Qualitative
Results
Medication Maximum Dosage Effect on Weight Effects on Effect on Weight
Studied (mg/​day) Restoration Psychological Maintenance
Symptoms

Antidepressants

 Amitriptyline 175 − − NDA

 Clomipramine 50 − − NDA

 Fluoxetine 60 − − +/​−

Antipsychotics

 Olanzapine 15 ++ +/​− NDA

 Pimozide 6 +/​− NDA NDA

 Quetiapine 178a − − NDA

 Risperidone 4 − − −

 Sulpiride 400 − NDA NDA

Appetite Stimulants

 Cyproheptadine 32 +/​− NDA NDA

Cannabinoids

 Tetrahydrocannabinol 30 − NDA NDA

 Dronabinol 5 + − NDA

Prokinetics

 Cisapride 30 +/​− +/​− NDA

Hormonal Agents

 Dehydroepiandrosterone 100 mg − − NDA

 Estradiol 100 mcg twice weekly − − NDA

 Oxytocin 40 IU NDA + NDA

 rhGH 15 mg/​kg − NDA NDA

 Testosterone 300μg − +/​− NDA

Other Agents

 Alprazolam .75 NDA − NDA

 Lithium NDAb + +/​− NDA

 Relamorelin 100μg +/​− NDA NDA

 Zinc 100 + NDA NDA

Key: ++ = ≥ 2 positive RCTs; + = ≥1 positive RCT; +/​− = mixed results; − = ≥1 negative RCTs and no positive RCTs; NDA = no data
available; rhGH = recombinant human growth hormone.
a
Mean dose.
b
Mean plasma lithium level = 1.0 mEq/​L.
Anxiety Scale (HAM-​A) scores. These isolated find- In an uncontrolled but randomized and pro-
ings were of unclear significance. spective 1-​year study evaluating treatment accep-
Two randomized, placebo-​ controlled relapse tance of medication versus psychotherapy, 122
prevention studies have been conducted with the outpatients with AN who were within 75% of
selective serotonin-​reuptake inhibitor (SSRI) flu- their target weight received fluoxetine 60 mg/​day
oxetine in AN. In the first trial, 35 patients (34 alone, CBT alone, or the combination (Halmi
inpatients) with restricting-​type AN with or with- et  al., 2005). Similar percentages (17%–​18%)
out purging behavior (none had displayed binge of patients were withdrawn (primarily for treat-
eating during their lifetime) who had been restored ment failure) from the three conditions. However,
to 76% to 100% of average body weight (with most among the remaining patients, there were more
above 90%) were randomized prior to hospital dis- noncompleters in the fluoxetine alone group
charge to fluoxetine (n  =  16) or placebo (n  =  19) (56%) than the CBT alone (40%) or the combi-
for 52 weeks (Kaye et  al., 2001). Only patients nation (41%) groups. The authors concluded that
with restricting-​type AN were included because of fluoxetine given alone had a very low treatment
prior open-​label data suggesting this subtype might acceptance rate.
respond better to fluoxetine than bulimic-​type AN In sum, taken together, these studies have led
(Kaye, Weltzin, Hsu, & Bulik, 1991). Fluoxetine some to conclude that antidepressants are inef-
was initiated at 20 mg/​day and adjusted to a maxi- fective for promotion or maintenance of weight
mum of 60 mg/​day. Adjunctive outpatient psycho- gain in patients with AN. However, it has also
therapy was allowed but not required; the number been noted that these trials have methodological
of participants who received psychotherapy was not shortcomings that limit their interpretation. These
provided. Fluoxetine-​ recipients were more likely include the use of small sample sizes leading to
to complete the trial: 10 (63%) patients remained inadequate power to detect potential differences in
on fluoxetine for 1 year, whereas only three (16%) effects; use of patients in various stages of illness;
patients remained on placebo for that period of use of narrow entry criteria that limit generaliz-
time (p = .006). Four groups were compared: fluox- ability of findings; and use of antidepressants in
etine completers (n = 10), fluoxetine noncompleters combination with other interventions designed
(n  =  6), placebo completers (n  =  3), and placebo specifically to promote weight gain or prevent
noncompleters (n =16). These four groups did not weight loss (Walsh et al., 2006). Thus, the possi-
show differences in weight, obsessive-​ compulsive bility that fluoxetine might be effective for relapse
ED symptoms, anxiety, or depression. However, by prevention if given at a different stage of illness
paired t test, only patients remaining on fluoxetine (i.e., after a longer period of weight restoration)
for 1  year showed a significant increase in weight or as a sole intervention (i.e., not as an adjunct
and reduction in ED psychopathology and mood to a structured psychotherapy designed to prevent
symptoms. relapse) cannot be excluded.
In the second trial, 93 patients with AN who had Another consideration is that findings with one
regained weight to a minimum BMI of 19.0 mg/​ class of antidepressant (e.g., SSRIs) may not gener-
kg2 after intensive inpatient or day program treat- alize to other classes. For example, there are reports
ment were randomized to outpatient fluoxetine of AN patients responding to the unique antide-
(n = 49) or placebo (n = 44) for up to 1 year (Walsh pressant mirtazapine, which enhances norepi-
et  al., 2006). All patients also received individual nephrine and serotonin (5-​HT) release; stimulates
cognitive-​ behavioral therapy (CBT) specifically 5-​HT1 receptors; blocks 5-​HT2, 5-​HT3, and hista-
designed to prevent relapse. Similar percentages of mine H1 receptors; and is associated with increased
fluoxetine recipients and placebo recipients main- appetite and weight gain (Hrdlicka, Beranova,
tained a BMI ≥ 18.5 mg/​kg2 and stayed in the study Zamecnikova, & Urbanek, 2008; Jaafar, Daud,
for 1  year (fluoxetine, 26.5%; placebo, 31.5%; Rahman, & Baharudin, 2007). There are also reports
p = .57). In addition, there was no significant dif- of patients with treatment-​resistant AN respond-
ference between fluoxetine and placebo in time-​to-​ ing to antidepressants when given in conjunction
relapse (hazard ratio [HR] 1.12; 95% confidence with other agents, such as antipsychotics, mood
interval [CI] 0.65, 2.01; p = .64). The authors con- stabilizers, and other antidepressants (Fountoulakis,
cluded that their study failed to demonstrate any Iacovides, Siamouli, Koumaris, & Kaprinis, 2006;
benefit from fluoxetine in the treatment of patients Newman-​Toker, 2000; Reilly, 1977; Wang, Chou,
with AN after weight restoration. & Shiah, 2006) (see the sections that follow).

362 Pharmacotherapy
Antipsychotics with restricting-​type AN and 12 with binge eating-​
Two randomized, placebo-​controlled, crossover purging type AN) received olanzapine (2.5 mg/​day
trials of first-​generation antipsychotics for weight for 1 month, then 5 mg/​day for 2 months) or pla-
restoration in patients with AN have been con- cebo, in addition to CBT (Brambilla et al., 2007).
ducted. In the first, 18 female inpatients with AN Body mass index increased significantly but simi-
based on criteria of the Diagnostic and Statistical larly in both groups. There were also no differences
Manual, 3rd Edition (DSM-​III) were randomized to between groups in improvements in Eating Disorder
receive a single dose of pimozide (4 or 6 mg) or pla- Inventory-​2 (EDI-​2) individual item or total val-
cebo in alternating 3-​week periods (Vandereycken ues, Yale-​ Brown-​ Cornell Eating Disorder Scale
& Pierloot, 1982). All patients received concomi- (YBC-​EDS) obsessiveness or total values, or Buss-​
tant behavior therapy. Mean changes in weight Durkee Scale (BDS) total values for aggressiveness.
were positive with pimozide but negative with pla- However, measures of rituals (on the YBC-​EDS),
cebo. A  crossover analysis showed a trend for the direct aggressiveness (on the BDS), depression, and
pimozide group to be associated with more weight persistence (on the Temperament and Character
gain (p  =  .067). After the first 3-​week period, for Inventory) improved significantly with olanzapine
example, patients receiving pimozide (n  =  8) had compared with placebo. Olanzapine was well tol-
a mean daily weight gain of 135 grams, whereas erated, with mild sleepiness as the only side effect.
those receiving placebo (n = 10) had a mean daily When stratifying for AN subtype, changes in BMI,
weight gain of 80 grams. In the second study, 18 depression, and direct aggression were significant
female inpatients with DSM-​III AN were random- among binge eating-​purging type patients, whereas
ized to sulpiride (300 or 400 mg/​day) or placebo in change in persistence was significant among
alternating 3-​week periods (Vandereycken, 1984). restricting-​
type patients. It was concluded that
Crossover analyses showed no direct effects of sul- olanzapine might improve different symptoms in
piride on weight change, clinical scales, or self-​ different subtypes of AN.
report questionnaires. However, individual analysis In the second study, 34 patients with AN receiv-
of the data showed numerically greater weight gain ing day treatment were randomized to receive flex-
in both periods with sulpiride than placebo, suggest- ible dose olanzapine (n  =  16) or placebo (n  =  18)
ing that, as in the first trial, negative findings could for 10 weeks (Bissada, Tasca, Barber, & Bradwejn,
be due to small sample size and inadequate power. 2008). Twenty-​eight patients (14 in each group)
Second-​generation (atypical) antipsychotics have completed the trial. Compared with placebo, olan-
been reported effective for AN in open-​label reports zapine was associated with a greater rate of increase
in children, adolescents, and treatment-​ resistant in BMI (p = .03) and a greater rate of decrease in
patients (Dunican & DelDotto, 2007; Duvvuri, obsessive symptoms (p = .02). Of the total sample,
Cromley, Klabunde, Boutelle, & Kaye, 2012; 87.5% of olanzapine recipients achieved weight
Mehler-​ Wex, Romanos, Kirchheiner, & Schulze, restoration, compared with 55.6% of placebo
2008). Olanzapine has been the most commonly recipients (p  =  .02). There were no differences in
used drug, but positive reports of aripiprazole, reductions between the groups on measures of anxi-
quetiapine, and risperidone have also appeared ety, depression, or compulsions. The mean (SD)
(Aragona, 2007; Court et  al., 2010; Frank, 2016; olanzapine dose over the 10-​week treatment period
Newman-​Toker, 2000; Powers, Bannon, Eubanks, for study completers was 6.61 (2.32) mg/​day. No
& McCormick, 2007; Umehara, Iga, & Ohmori, differences in adverse effects were observed between
2014). These drugs have been described as helpful the two treatment conditions. There were no serious
for weight restoration; for many of the core psy- adverse effects.
chological symptoms of AN, such as fear of fatness, In the third RCT, 23 outpatients with AN who
difficulty eating, distorted body image, obsessive-​ were 16  years of age or older were randomized to
compulsive features, and poor insight; and for many receive olanzapine or placebo (Attia et  al., 2011).
of the associated symptoms of AN, including binge Study drug was administered with medication man-
eating, purging, hyperactivity, delusionality, depres- agement sessions aimed to enhance compliance.
sion, anxiety, and mood instability. Seventeen (74%) patients completed the trial. End-​
Six randomized, placebo-​ controlled trials of of-​treatment BMI, with initial BMI as covariate,
second-​ generation antipsychotics for weight res- was significantly greater in patients receiving olan-
toration have been conducted in AN. In the first zapine. Psychological symptoms improved in both
RCT, 30 female outpatients with DSM-​IV AN (18 groups with no significant differences. Olanzapine

McElroy, Guerdjikova, Mori, Keck Jr. 363

was well tolerated with no adverse metabolic effects; from baseline (15.7 mg/​kg2 ± 1.9) to endpoint
sedation was the most common side effect. (18.1 mg/​kg2 ± 2.5; p = .03). Significant improve-
In the fourth RCT, 20 adolescent females with ment was also observed on the EDI (p  =  .02),
AN receiving comprehensive eating disorder treat- Eating Attitude Test (EAT; p = .009), and CGI-​I
ment were randomized to receive olanzapine or scale (p  =  .001) scores. There are also reports of
placebo (Kafantaris et  al., 2011). Fifteen patients antipsychotics being helpful in AN patients with
completed the study. Mean olanzapine dose was serious comorbid neuropsychiatric and medical
8.5 mg/​day at week 10. Change in percent median disorders, including schizotypal personality dis-
body weight improved similarly in both treatment order (Nagata, Ono, & Nakayama, 2007), autism
groups at midpoint and endpoint. Olanzapine and (Fisman, Steele, Short, Byrne, & Lavallee, 1996),
placebo also produced similar improvements in eat- and epilepsy with chronic renal failure (Aragona,
ing pathology, psychological functioning, and rest- 2007). Of note, because second-​generation anti-
ing energy expenditure. Olanzapine was associated psychotics may cause glucose intolerance in AN
with a trend of increasing fasting insulin and glu- (possibly by inducing insulin resistance), it has been
cose levels at 10 weeks. recommended that AN patients receiving treatment
In the fifth study, 40 female AN patients 12 to with olanzapine have their glucose metabolism
21 years of age who were receiving treatment in a closely monitored (Yasuhara, Nakahara, Harada, &
specialized eating disorder program were random- Inui, 2007).
ized to receive risperidone or placebo for 11 weeks All RCTs of antipsychotics in the treatment of
(Hagman et al., 2011). Patients had to be actively AN are limited by small sample size and, hence,
engaged in the specialized eating disorder treatment inadequate power to detect potential clinically sig-
program to be enrolled in the study. Risperidone nificant differences. Negative trials (Hagman et al.,
(mean dose 2.5 mg/​day) was associated with a sig- 2011; Kafantaris et al., 2011; Powers et al., 2012)
nificantly greater reduction of drive for thinness on could therefore represent failed trials. Some nega-
the Eating Disorder Inventory-​2 (EDI-​2) (Garner, tive trials are further limited by requiring partici-
1991) over the first 7 weeks, but this difference was pants to be receiving comprehensive psychosocial
not sustained at week 11. Risperidone was also asso- treatment (Hagman et  al., 2011; Kafantaris et  al.,
ciated with a significantly greater decrease on the 2011); with small sample sizes, it may be difficult to
EDI-​2 Interpersonal Distrust Subscale. There were show medication effects beyond those of psychoso-
no drug-​placebo differences on any other measure cial treatment. Indeed, the mixed findings have led
of psychological symptoms. There were no changes to different conclusions on the usefulness of anti-
between risperidone and placebo for change in ideal psychotics in AN. Four groups, conducting meta-​
body weight or BMI: 33% of risperidone patients analyses of RCTs of antipsychotics in AN, have
and 45% of placebo patients reached target weight concluded that antipsychotics are not efficacious
and maintained it for 4 weeks. Side effects of ris- for weight gain or psychological symptoms in AN
peridone were fatigue and dizziness. Prolactin levels (Dold, Aigner, Klabunde, Treasure, & Kasper, 2015;
were significantly increased in risperidone-​treated Kishi, Kafantaris, Sunday, Sheridan, & Correll,
patients at weeks 7 and 11. 2012; Lebow, Sim, Erwin, & Murad, 2013; de Vos
In the sixth study, 15 participants with AN et al., 2014). Other experts, however, have argued
were randomized to receive quetiapine (n  =  6) or that olanzapine in particular may be efficacious,
placebo (n  =  9) for 8 weeks (Powers, Klabunde, especially if it has to be administered alone with-
& Kaye, 2012). Ten patients completed the study. out adjunctive psychosocial treatment (Brewerton,
Quetiapine (mean daily dose 177.7 mg) was not 2012). Indeed, authors of negative trials have
superior to placebo in reducing core ED, depressive, qualified their findings by reporting that some AN
or obsessional symptoms. Additionally, there was no patients do respond well to second-​generation anti-
difference in change in BMI between study groups. psychotics (Dold et al., 2015; Powers et al., 2012).
No controlled studies of antipsychotics for
weight maintenance in AN have yet been pub- Antidepressant–​Antipsychotic Combinations
lished, but there are open-​label reports. For exam- Anorexia nervosa has similarities with psychotic
ple, Cassano et  al. (2003) treated 13 outpatients depression including depressive symptoms, delu-
with treatment-​ resistant restricting-​
type AN for sional thinking, and hypercortisolism (Monteleone
6 months with haloperidol in addition to standard et  al., 2001; Parsons & Sapse, 1985; Steinglass,
therapy. Body mass index increased significantly Eisen, Attia, Mayer, & Walsh, 2007). Psychotic

364 Pharmacotherapy
depression is characterized by better response to Cannabinoids
antidepressant–​antipsychotic combination therapy Two RCTs have examined cannabinoids in AN.
than to antipsychotic monotherapy. No controlled In the first, a 4-​ week, double-​ blind, diazepam-​
studies of antidepressant–​antipsychotic combina- controlled, crossover study in 11 female patients,
tions have been published in AN, but there are tetrahydrocannabinol and diazepam were associated
open-​label reports of patients reporting to such with comparable amounts of weight gain (Gross
regimens. These include descriptions of treatment-​ et  al., 1983). Patients had greater somatization,
resistant AN patients responding to the addition interpersonal sensitivity, and sleep disturbance dur-
of a second-​generation antipsychotic to an anti- ing tetrahydrocannabinol treatment. Also, three
depressant (Marzola et al., 2015; Newman-​Toker, patients had severe dysphoric reactions (includ-
2000). There are also reports of AN patients with ing paranoia and feelings of loss of control) with
depressive symptoms responding to the combina- tetrahydrocannabinol.
tion of olanzapine and mirtazapine (Fountoulakis In the second RCT, a crossover study in 24
et al., 2006; Wang et al., 2006). Randomized con- patients with enduring AN, adjunctive dronabinol,
trolled trials of combination antipsychotic and given at 2.5 mg BID for 4 weeks, was superior to
antidepressant therapy in the treatment of AN are placebo for weight gain (Andries, Frystyk, Flyvbjerg,
greatly needed. & Støving, 2014). Specifically, participants gained
0.73  kg (p < .01) during dronabinol treatment
Appetite Stimulants above that gained during placebo treatment. Also
Two randomized, placebo-​ controlled trials during dronabinol treatment, physical activity
of cyproheptadine, an antiallergy and appetite-​ intensity increased by 20% (p = .01), resulting in an
stimulating drug with high affinity for various increased energy expenditure of 68.2 kcal/​day above
serotonin, histamine, dopamine, adrenergic, and placebo (p = .01); there was a transient increase in
muscarinic receptors (Goudie, Cooper, Cole, & leptin levels at 3 weeks; and urinary free cortisol lev-
Sumnall, 2007), have been conducted. In the first els were decreased (Andries, Frystyk, Flyvbjerg, &
study, 81 female inpatients with AN were random- Støving, 2015; Andries, Gram, & Støving, 2015).
ized to one of four treatment combinations of Changes in EDI-​2 scores during treatment with
cyproheptadine (12–​32 mg/​day) or placebo with dronabinol or placebo were minimal, and there
or without behavioral therapy (Goldberg, Halmi, were no statistically significant differences on EDI-​
Eckert, Casper, & Davis, 1979). Mean weight 2 scores between treatment periods. The drug was
gain did not differ between the groups receiving well tolerated without adverse psychiatric effects.
cyproheptadine (5.11 kg) versus placebo (4.32 kg).
However, in a subgroup analysis, cyproheptadine Prokinetics
was superior to placebo for weight gain in patients Two randomized, placebo-​controlled trials of the
with a history of two or more birth delivery compli- prokinetic drug cisapride have been done in AN.
cations compared to those with none. In the first, 12 outpatients with DSM-​III-​R AN
In the second study, 72 female inpatients with received cisapride 10 mg administered orally three
AN were randomized to cyproheptadine (n = 24), times daily (n = 6) or placebo (n = 6) for 6 weeks after
amitriptyline (n = 23), or placebo (n = 25) (Halmi completing an 8-​week inpatient program (Stacher,
et  al., 1986). Eighty-​three percent of cyprohepta- Abatzi-​Wenzel, et  al., 1993). All patients then
dine recipients versus 64% of placebo recipients received 6 weeks of open-​label cisapride. Gastric
achieved their target weights. Among these patients, emptying was accelerated in all six patients receiving
significantly fewer days were required to achieve tar- cisapride; gastric emptying was accelerated in three
get weight with cyproheptadine (mean [SD] = 36.5 patients receiving placebo and slowed in the other
[19.5]) than placebo (mean [SD]  =  45.0[18.3]; p three. Gastric retention symptoms and constipation
< .05). In addition, cyproheptadine significantly were numerically improved in the cisapride group
increased treatment efficiency in the nonbulimic versus the placebo group. Mean weight (SD) gain
patients and significantly decreased treatment effi- with cisapride was greater than with placebo (7.3%
ciency in the bulimic patients. (Treatment efficacy [7.1] versus 1.7% [3.1] of ideal body weight, respec-
was the reciprocal of the number of days to target tively). After the 6 weeks of cisapride treatment dur-
weight times the constant 90, the maximal length of ing the second period, gastric retention symptoms
treatment.) Attrition was 25% for cyproheptadine stayed reduced in the cisapride-​first patients and
and 20% for placebo. decreased in the placebo-​first patients. Five of the

McElroy, Guerdjikova, Mori, Keck Jr. 365

six placebo-​first patients gained weight. However, 100 mg/​day (n = 16) or placebo (n = 19) until they
after their second 6 weeks of cisapride, only one of achieved a 10% increase in BMI (Birmingham,
the six cisapride-​first patients gained more weight; Goldner, & Bakan, 1994). The rate of BMI increase
the other five lost weight (mean change, –​3.0%). in the zinc-​supplemented group was twice that of
In the second study, 29 inpatients with AN were the placebo group (p = .03). However, 19 patients
randomized to cisapride 10 mg three times daily who did not complete the study (n = 10 receiving
or placebo for 8 weeks (Szmukler, Young, Miller, zinc) did not appear to be included in the efficacy
Lichtenstein, & Binns, 1995). Both gastric empty- analysis. In addition, antidepressants and major
ing and weight improved significantly but equally in tranquilizers were used but their use was not quan-
both groups. However, patients receiving cisapride titatively presented. Zinc was not associated with
rated themselves as hungrier (p  =  .02) and more any harmful effects.
improved on a global measure of symptom change Another study was a 12-​week crossover trial with
(p = .02). The correlation between gastric emptying 6 weeks of oral zinc sulphate 50 mg/​day alternat-
and weight gain was modest (r = .30; p = .11). There ing with 6 weeks of placebo to be given to children
were no correlations between gastric emptying and with AN (Lask, Fosson, Rolfe, & Thomas, 1993).
symptomatic measures. Though described as double-​blind, this trial was
Cisapride’s access has since been restricted unlikely to be randomized because of 26 patients
because of an association with potentially fatal car- enrolled, seven children received zinc supplemen-
diac arrhythmias (Wysowski, Corken, Gallo-​Torres, tation, whereas 19 received standard treatment.
Talarico, & Rodriguez, 2001). Except for one small Moreover, only three of seven trials of zinc supple-
placebo-​controlled crossover study showing eryth- mentation were completed. Thus, no conclusions
romycin 200 mg accelerated gastric emptying in 10 about zinc supplementation could be drawn from
patients with AN (Stacher, Peters, et al., 1993), no this study.
other controlled studies of prokinetic agents in AN Based on the above preliminary data, along
have been done. Open reports describe successful use with zinc’s low cost and benign side-​effect profile,
of domperidone and metoclopramide to decrease some have argued that oral zinc administration dur-
symptoms and promote weight gain (Russell et al., ing weight restoration should be routinely consid-
1983; Saleh & Lebwohl, 1980). In one case, how- ered (Birmingham & Gritzner, 2006). The World
ever, excessive weight gain was described (Sansone Federation of Societies of Biological Psychiatry
& Sansone, 2003). Of note, if either domperidone (WFSBP) pharmacotherapy of eating disorder
or metoclopramide are used, patients should be guidelines also concluded there was evidence of effi-
monitored for extrapyramidal side effects. cacy of zinc in the treatment of AN (Aigner et al.,
2011). Other guidelines, however, do not recom-
Zinc mend routine zinc supplementation (Hay et  al.,
Two randomized, placebo-​ controlled trials 2014). Further randomized, placebo-​ controlled
of zinc supplementation have been conducted studies are needed to better establish zinc’s effi-
in patients with AN (Su & Birmingham, 2002). cacy, adverse event profile, and optimal dosing and
In the first study, six adolescents with AN who treatment duration in the management of patients
received elemental zinc 50 mg/​day for 6  months with AN.
showed decreased depression and anxiety on the
Zung Depression Scale (p < .05) and the Stale-​ Hormonal Treatments
Trait Anxiety Inventory (p < .05) compared with Miller, Grieco, and Klibanski (2005) random-
seven adolescents who received placebo (Katz et al., ized 33 women with AN and relative testosterone
1987). The zinc-​supported group also showed a deficiency to transdermal testosterone (150 or 300
greater weight gain and increase in height, improved μg) or placebo for 3 weeks. Serum total and free
taste function, greater advancement in sexual matu- testosterone levels increased significantly in patients
ration, and better resolution of skin abnormalities, receiving testosterone. Significant improvement in
but these differences did not reach statistical signifi- depression was seen in depressed patients receiv-
cance. A fourteenth patient who dropped out of the ing testosterone, whereas there was no change in
trial was excluded from analysis. In the second trial, depressed patients receiving placebo (p  =  .02).
35 female inpatients with AN who were considered Testosterone recipients showed improved spatial
completers had been randomized to zinc gluconate cognition (p  =  .0015). Weight, however, did not

366 Pharmacotherapy
change over the 3 weeks in either group. Attrition trial conducted in a laboratory setting (Kim, Kim,
rate was 13%. Cardi, et al., 2014; Kim, Kim, Park, Pyo, & Treasure,
Two randomized, placebo-​controlled studies of 2014). Patients with AN showed significant reduc-
rhGH have been conducted in patients with AN. tions in attentional biases toward eating-​ related
In the first, 15 inpatients with AN, ages 12 to stimuli and negative shape stimuli, and the effect
18 years, received rhGH .05 mg/​kg subcutaneously of oxytocin was correlated with autistic spectrum
(n = 8) or an equivalent volume of placebo (n = 7) traits. However, oxytocin had no effect on amount
daily for 28 days in addition to a standard refeeding of fruit juice consumed.
protocol (Hill et  al., 2000). Patients given rhGH
reached medical/​cardiovascular stability (defined as Ghrelin Agonists
normal orthostatic heart rate response to a standing In a phase 2 proof-​of-​concept study of the ghre-
challenge) significantly sooner than those receiv- lin agonist relamorelin in 22 females with AN,
ing placebo (median 17 versus 37  days, p  =  .02). participants who received relamorelin 100 µg/​day
Numerical improvements were also seen in weight for 4 weeks had significantly shorter gastric empty-
gain and hospitalization length in the rhGH group. ing time (p = .03) and greater weight gain (.86 kg
There were no dropouts. In the second study, 21 versus .04  kg, respectively, p  =  .12) as compared
women with AN were randomized to receive sup- with participants who received placebo (Fazeli
raphysiological doses of rhGH (15 µg/​kg daily by et al., 2016).
subcutaneous injection) or placebo for 12 weeks
(Fazeli et  al., 2010). Mean weight gain between Lithium
groups (0.3 kg for rhGH versus 0.85 kg for placebo) A 4-​week randomized, placebo-​controlled trial of
was not significantly different. Total fat mass and lithium was conducted in 16 female inpatients with
percentage fat mass, however, decreased with rhGH AN (Gross et al., 1981). All patients received behav-
and increased with placebo (p = .004). ior modification therapy, which included weekly
Bloch, Ish-​Shalom, Greenman, Klein, & Latzer group therapy, twice weekly individual therapy, and
(2012) randomized 26 premenopausal female occasional tube findings. The eight patients receiv-
patients with AN to dehydroepiandrosterone ing lithium showed significantly greater weight gain
(DHEA), a hormone produced by the adrenal gland after 3 (p = .04) and 4 weeks (p = .03) of treatment
and brain that enhances production of androgens than the eight patients receiving placebo. After 4
and estrogens, 50 mg twice daily or placebo in a weeks, lithium-​treated patients also showed signifi-
3:2 ratio for 6 months. All patients received psycho- cantly more improvement on an item measuring
therapy, weekly nutritional assessments, and daily “denial and minimization of illness” and ingested
calcium carbonate 600 mg and vitamin D3 200 IU. more fat per day. The mean (SD) plasma lithium
Body mass index in DHEA-​treated patients was level over the 4 weeks of treatment was 1.0 (0.1)
significantly increased at 4 months compared with mEq/​L. Lithium was well tolerated, and there were
placebo-​treated patients. However, the difference no serious adverse events.
in BMI increase was not statistically significantly Many authorities are extremely reluctant to
different across the 6 months of the study. There consider lithium for AN, given the drug’s low
were also no drug-​placebo differences in depressive therapeutic index and need for monitoring, and
symptoms, bone mineral density, or bone mineral AN’s association with dehydration, electrolyte
content. DHEA was well tolerated. abnormalities, and cardiac arrhythmias (Kolata,
Misra et  al. (2013) randomized 72 adolescent 1980). It should be noted, however, that case
girls with AN to receive transdermal estradiol (100 reports describe patients with AN and comorbid
mcg twice weekly) with cyclic progesterone or pla- bipolar disorder responding to lithium alone or in
cebo patches and pills for 18 months. Thirty-​seven combination with carbamazepine (Hudson, Pope,
patients completed the trial. Changes in BMI did Jonas, & Yurgelun-​ Todd, 1985). There are also
not differ between treatment groups. Estrogen case reports of patients with treatment-​resistant AN
replacement produced a decrease in trait anxiety responding to lithium (Barcai, 1977; Reilly, 1977;
scores but had no effect on state anxiety, eating atti- Stein, Hartshorn, Jones, & Steinberg, 1982). Use
tudes, or body shape perception. of lithium in AN requires careful monitoring of
Intranasal oxytocin was evaluated in 64 patients patients’ fluid, electrolyte, renal, cardiac, and thy-
with AN in a single dose, within-​subject, crossover roid status.

McElroy, Guerdjikova, Mori, Keck Jr. 367

Opioid Antagonists and Opioids reduce anxiety as compared with placebo, but did
One placebo-​ controlled study of an opioid increase fatigue. The authors concluded that short-​
antagonist has been conducted in AN. Six AN acting benzodiazepines may have a limited thera-
patients treated with naltrexone, up to 200 mg/​day, peutic role in AN.
or placebo in individual randomized crossover tri-
als, showed significant reductions in binge eating Antiepileptic Drugs
and purging episodes, but not a significant increase There are no randomized, placebo-​controlled tri-
in daily food intake, on naltrexone compared with als of an antiepileptic drug in AN, but open-​label
placebo (Marrazzi, Bacon, Kinzie, & Luby, 1995). reports have been published. A 16-​year-​old girl with
Although quantitative weight data were not pro- “classical” AN beginning simultaneously with par-
vided, no patients lost weight on naltrexone. tial complex seizures showed both weight gain and
Naltrexone 25 to 75 mg/​day has also been used for seizure control with phenytoin treatment (Szyper
successful weight restoration in hospitalized patients & Mann, 1978). A  13-​year-​old girl with AN and
with chronic AN (Luby, Marrazzi, & Kinzie, 1987). epilepsy responded to the combination of valpro-
In an open-​label trial, 12 inpatients with AN receiv- ate and clonazepam (Tachibana, Sugita, Teshima,
ing behavioral treatment and antidepressants gained & Hishikawa, 1989). Case reports of topiramate in
significantly more weight during constant nalox- AN are mixed:  Topiramate improved the concur-
one infusion, up to 3.2 to 6.4 mg/​day for 5 weeks, rent AN of a patient with bipolar disorder (Guille &
compared with the 4 weeks post-​infusion (p < .01) Sachs, 2002), but reports of development or exac-
(Moore, Mills, & Forster, 1981). Finally, tramadol, erbation of AN after topiramate initiation for epi-
a synthetic opioid that binds μ-​opioid receptors and lepsy or migraine have also been described (Lebow
weakly inhibits the uptake of 5-​HT and norepi- et al., 2015; Rosenow, Knake, & Hebebrand, 2002).
nephrine, was reported to be helpful for a patient Additionally, ED patients have misused topira-
with intractable AN (Mendelson, 2001). mate to lose weight (Chung & Reed, 2004; Colom
et al., 2001).
Anxiolytic Medications
D-​cycloserine is a partial agonist at the N-​ Nutritional Supplements
methyl-​D-​aspartate (NMDA) glutamatergic recep- A study of nutritional supplementation of fluox-
tor that may facilitate extinction of conditioned fear etine aimed to enhance serotonergic neurotransmis-
and may be helpful as a short-​term adjunctive inter- sion was ineffective in promoting weight gain in AN
vention to exposure therapy for anxiety disorders, (Barbarich et  al., 2004). Twenty-​six patients with
including phobias (Hofmann et al., 2006). Eleven AN receiving fluoxetine were randomized to receive
patients with AN were randomly assigned to receive a nutritional supplement containing tryptophan,
either D-​cycloserine 50 mg or placebo before each vitamins, minerals, and essential fatty acids, or a
of four exposure therapy interventions (e.g., train- nutritional placebo. There were no significant dif-
ing meals) aimed to increase meal intake (Steinglass ferences in weight gain or in mean changes in anxi-
et al., 2007). A trend (p = .06) with a large effect ety or obsessive and compulsive symptoms between
size (d = 1.33) was seen for d-​cycloserine recipients the two groups.
to experience a greater decrease in self-​ reported
depressive symptoms than placebo recipients from Other Medications
the baseline test meal to the final test meal (a mean In a long-​term randomized, placebo-​controlled,
of 28 days), but there were no other outcome dif- crossover trial in four patients with AN, the alpha 2-​
ferences between the groups. There were no adverse adrenergic agonist clonidine had no effect on rate of
events with d-​cycloserine. This study was limited weight gain, but was associated with lowered blood
by small sample size. Also, patients assigned to d-​ pressure, reduced pulse rate, and sedation (Casper,
cycloserine had significantly lower postmeal anxiety Schlemmer, & Javaid, 1987).
than those assigned to placebo at the baseline meal.
In a randomized, placebo-​controlled, crossover Pharmacotherapy of Bulimia Nervosa
study, one dose of alprazolam 0.75 mg was adminis- Two primary pharmacotherapy RCT designs
tered to 17 inpatients with AN (Steinglass, Kaplan, have been used to evaluate medications in BN: short-​
Liu, Wang, & Walsh, 2014). Within-​subject com- term acute studies of patients who are actively binge
parisons showed that alprazolam did not increase eating and purging and long-​ term maintenance
caloric intake during a laboratory test meal or studies of patients whose bulimic symptoms have

368 Pharmacotherapy
responded to an acute intervention. In addition, three with atypical drugs (mianserin, trazodone,
pharmacotherapy studies in BN have been done and bupropion; Horne et  al., 1988; Pope, Keck,
as monotherapy trials, in which medication alone McElroy, & Hudson, 1989; Sabine, Yonace,
is compared with placebo, another medication, Farrington, Barrant, & Wakeling, 1983). Study
or a psychological treatment, and as combination durations ranged from 6 to 16 weeks. Meta-​analysis
therapy trials, where medication plus a psycho- showed that the pooled relative risk for remission
logical treatment is compared with the psychologi- of binge eating episodes was 0.87 (95% CI = 0.81,
cal treatment alone and/​or the medication alone. 0.93; p < .001), favoring antidepressants. The num-
Primary outcomes in the acute trials have usually ber needed to treat (NNT) for a mean treatment
been measures of the frequency of binge eating epi- duration of 8 weeks, taking the 92% nonremis-
sodes and/​or inappropriate compensatory behaviors sion rate in the placebo controls as a measure of the
(e.g., vomiting), or rates of response or remission baseline risk, was 9 (95% CI = 6, 16). The relative
of bulimic symptoms. Primary outcomes in the risk for clinical improvement, defined as a 50% or
maintenance trials have usually been time to relapse greater reduction in binge episodes, was 0.63 (95%
or rate of relapse. Secondary outcomes have been CI = 0.55, 0.74). The NNT for a mean duration of
measures of ED psychopathology, mood symptoms, 9 weeks was 4 (95% CI = 3, 6), with 67% unim-
global clinical improvement, and treatment adher- proved in the placebo group. There was no evidence
ence. The two major classes of drugs studied thus of statistically significant differences in efficacy
far in BN in randomized, placebo-​controlled trials among the different classes of antidepressants.
have been antidepressants and antiepileptics (Table However, remission rates were low and a consider-
19.2). Drug classes that have received less study able fraction of patients did not show a reduction of
include 5-​HT3 antagonists, opioid antagonists, hor- at least 50% in bulimic symptoms. Patients receiv-
monal agents, and stimulants. ing tricyclics dropped out due to any cause more
frequently than those receiving placebo, though the
Antidepressants opposite was found for fluoxetine. The authors con-
Many different antidepressant classes have been cluded that, in general, a single antidepressant agent
evaluated in BN in randomized, placebo-​controlled is clinically effective for the treatment of BN when
trials; these include SSRIs, tricyclics, monoamine compared with placebo, but the effect is modest.
oxidase inhibitors, and atypical agents. Drugs from Importantly, among the studies reviewed was the
each of these classes have been shown superior to first pivotal RCT of fluoxetine in 387 women with
placebo for reducing the frequency of both binge BN in which 60 mg/​day was shown to be superior
eating and purging episodes in BN (Shapiro et al., to placebo for reducing binge eating and vomiting
2007; Yager & Powers, 2007) (Table 19.2). As episodes, while 20 mg/​day was shown to have an
noted earlier, an antidepressant, fluoxetine, is the intermediate effect (FBNCSG, 1992). Fluoxetine
only medication with regulatory approval for the 60 mg/​day was also superior to placebo for reducing
treatment of BN. depression, carbohydrate craving, and pathological
In 2003, Bacaltchuk and Hay published eating attitudes and behaviors. Also reviewed was
a Cochrane Review of randomized, placebo-​ the single study of bupropion showing that though
controlled trials of antidepressants in patients with efficacious for reducing binge eating and purging,
BN. Nineteen studies were included: six trials with this agent was associated with an increased risk of
tricyclics (imipramine, desipramine, and amitripty- seizures (Horne et al., 1988). It is therefore contra-
line; Agras, Dorian, Kirkley, Arnow, & Bachman, indicated for the treatment of BN and AN.
1987; McCann & Agras, 1990; Mitchell & Groat, Antidepressants have been studied both against
1984; Mitchell et al., 1990; Pope, Hudson, Jonas, and in combination with a variety of psychologi-
& Yurgelun-​Todd, 1983; Walsh, Hadigan, Devlin, cal interventions in BN, most commonly CBT but
Gladis, & Roose, 1991); five with monoamine-​ also intensive inpatient psychotherapy and nutri-
oxidase inhibitors (phenelzine, isocarboxazid, tional counseling (Bacaltchuk, Hay, & Trefiglio,
moclobemide, and brofaromine; Carruba et  al., 2001; Shapiro et  al., 2007). Designs and results
2001; Kennedy et al., 1993, 1988; Rothschild et al., have varied, making firm conclusions difficult
1994; Walsh, Stewart, Roose, Gladis, & Glassman, to make. In 2001, Bacaltchuk et  al. published a
1984), five with the SSRI fluoxetine (FBNCSG, Cochrane Review of RCTs in which antidepres-
1992; Kanerva et  al., 1994; Mitchell et  al., 2001; sants were compared with psychological treat-
Walsh et  al., 2000; Wheadon et  al., 1992), and ments or the combination of antidepressants

McElroy, Guerdjikova, Mori, Keck Jr. 369

Table 19.2  Medications Studied for Bulimia Nervosa in Randomized, Placebo-​Controlled trials:
Qualitative Results
Medication Maximum Dosage Reduction in Reduction in
Studied (mg/​day) Binge Eating Purging

Tricyclic Antidepressants

 Amitriptyline 150 ++ ++

 Desipramine 300 ++ ++

 Imipramine 300 ++ ++

Monamine Oxidase Inhibitors

 Brofaromine 200 − ++

 Isocarboxazid 60 ++ ++

 Moclobemide 600 − −

 Phenelzine 90 +++ +++

SSRIs

 Fluoxetine 60 ++++ ++++

 Fluvoxamine 300 ++ ++

Other Antidepressants

 Bupropion 450 ++ ++

 Mianserin 60 − −

 Trazodone 400 ++ ++

Antiepileptics

 Carbamazepine NDAa − −

 Topiramate 400 +++ +++

Other Agents

 Dexfenfluraminec 60 ++ ++

 Flutamide 500 ++ −

 Lithium 600–​1200b − −

 Naltrexone 200 +/​-​ +/​-​

 Ondansetron 24 ++ ++

 Spironolactone 150 − −

Key: ++++ = ≥2 positive RCTs and evidence for maintenance of efficacy; +++ = ≥2 positive RCTs; ++ = ≥1 positive RCT; +/​
-​ = mixed data; − = only negative data; NDA = no data available; SSRI = selective serotonin reuptake inhibitor.
a
Plasma carbamazepine levels = 6–​10 μg/​mL.
b
Mean plasma lithium level = .62 mEq/​L.
c
Removed from the market for safety concerns.
with psychological treatments was compared to in BN. In the first, 72 patients with BN success-
each treatment alone for reducing symptoms in fully treated with intensive inpatient psychotherapy
BN. The main efficacy outcome was remission were randomized to receive fluvoxamine (n  =  33)
of bulimic symptoms. Three comparisons were or placebo (n  =  39) as outpatients for 12 weeks
made. In the first, which included five trials and (Fichter, Krüger, Rief, Holland, & Döhne, 1996).
237 patients, antidepressants alone were compared Fluvoxamine was begun 3 weeks before hospital
with psychological treatments alone (Agras et  al., discharge, for a total of 15 weeks of treatment.
1992; Goldbloom et  al., 1997; Leitenberg et  al., The relapse rate was significantly lower for fluvox-
1994; Mitchell et  al., 1990; Walsh et  al., 1997). amine than placebo, as shown by (1)  10% versus
In the second, which included five trials and 247 46% deterioration on the Psychiatric Status Rating
patients, antidepressants alone were compared with Scale for Bulimia Nervosa, (2) 111% versus 270%
antidepressant–​psychological treatment combina- increase in self-​reported binge eating episodes in the
tions (Agras et al., 1992; Goldbloom et al., 1997; last week, and (3)  50% versus 175% increase on
Leitenberg et al., 1994; Mitchell et al., 1990; Walsh the Structured Interview for Anorexia and Bulimia
et  al., 1997). In the third, which included seven Nervosa (SIAB) subscale of bulimic behavior. In
trials and 343 patients, psychological treatments addition, at the end of relapse-​prevention, the flu-
alone were compared with combination treatment voxamine group had significantly more patients
(Agras et al., 1992; Beaumont et al., 1997; Fichter reporting no binge eating episodes in the past
et  al., 1991; Goldbloom et  al., 1997; Leitenberg week than the placebo group (p < .05). However,
et  al., 1994; Mitchell et  al., 1990; Walsh et  al., the dropout rate was high (33%), with 14 (38%)
1997). The first comparison found remission rates fluvoxamine recipients stopping drug prematurely
for antidepressant treatment alone were 20% ver- compared with five (14%) placebo recipients.
sus 39% for psychological treatment alone (relative In the second study, 232 outpatients with DSM-​
risk = 1.28; 95% CI = 0.98, 1.67). Dropout rates IV BN, purging type, received single-​blind treat-
were higher for antidepressants alone than psy- ment with fluoxetine 60 mg/​day for 8 weeks; 150
chological treatments alone (relative risk  =  2.18; (65%) met response criteria (a decrease greater than
95% CI  =  1.09, 4.35). The number needed to or equal to 50% from baseline in vomiting episode
harm (NNH) for a mean treatment duration of frequency during one of the two preceding weeks)
17.5 weeks was 4 (95% CI  =  3, 11). The second and were randomly assigned to continue fluoxetine
comparison found remission rates for the combina- 60 mg/​day (n = 76) or switch to placebo (n = 74)
tion of 42% versus 23% for antidepressants alone for 52 weeks (Romano, Halmi, Sarkar, Koke, & Lee,
(relative risk  =  1.38; 95% CI  =  0.98, 1.93). The 2002). Fluoxetine-​treated patients exhibited a sig-
third comparison found remission rates of 36% for nificantly longer time to relapse (defined as a return
psychological treatments alone versus 49% for the to baseline vomiting frequency that persisted for
combination (relative risk = 1.21; 95% CI = 1.02, 2 weeks) than placebo-​treated patients (χ2  =  5.79,
1.45). Dropout rates were higher for the combina- f  =  1, p < .02). Endpoint analysis showed statisti-
tion compared with psychological treatments alone cally significant differences favoring fluoxetine for
(relative risk =.57; 95% CI =.38, .88). The NNH vomiting episodes, binge eating episodes, obsessive-​
was 7 (95% CI  =  4, 21). Using a conservative compulsive symptoms, and clinical global outcome.
approach, the only statistically significant differ- However, relapse rates and symptom measures
ence between groups was that combination therapy increased over the trial in both treatment groups. In
was superior to psychological treatment alone. The addition, the attrition rate was very high, with 63
authors concluded that the effectiveness of com- (83%) fluoxetine recipients and 68 (92%) placebo
bined antidepressant–​psychological approaches recipients discontinuing the study prematurely.
was superior to psychotherapy alone, but that the Of note, several antidepressant classes have
number of trials might be insufficient to show com- not yet been evaluated in randomized, placebo-​
bination therapy or psychotherapy alone superior controlled trials in BN. These include serotonin
to antidepressants alone. They also concluded that norepinephrine reuptake inhibitors (SNRIs; e.g.,
psychotherapy was more acceptable to patients and desvenlafaxine, duloxetine, milnacipran, and venla-
that the addition of antidepressants to psychother- faxine), norepinephrine reuptake inhibitors (NRIs;
apy reduced its acceptability. e.g., reboxetine), and novel agents such as vilazo-
At least two randomized, controlled relapse-​ done and vortioxetine. Open-​label data, however,
prevention trials have been done with antidepressants suggest milnacipran, reboxetine, and, to a lesser

McElroy, Guerdjikova, Mori, Keck Jr. 371

extent, duloxetine, may be effective in BN, includ- eating and/​or purge days was significantly greater
ing in treatment-​resistant cases (El-​Giamal et  al., for the topiramate group (52%) than the placebo
2000, 2003; Fassino, Daga, Boggio, Garzaro, & group (24%; p = .012). Remission rates from binge
Piero, 2004; Hazen & Fava, 2006; Noma, Uwatoko, eating and purging were numerically, but not signif-
Yamamoto, & Hayashi, 2008). icantly, higher for topiramate (23%) than placebo
(6%); attrition rates were numerically lower for
Antiepileptic Drugs topiramate (34%) than placebo (47%). One patient
Three randomized, placebo-​controlled studies of discontinued topiramate for an adverse event (nau-
antiepileptic drugs have been conducted in BN—​ sea). The most common side effects associated with
one with carbamazepine and two with topiramate. topiramate were fatigue, flu-​ like symptoms, and
In the first RCT of an AED in BN, 16 patients paresthesias.
with DSM-​III BN and at least one binge episode In the second study, 60 patients who had BN
per week and no binge-​free internals of longer than for at least 12 months received topiramate (n = 30;
3 weeks during the previous year received carba- titrated to 250 mg/​day by the sixth week with the
mazepine in a crossover design (Hudson & Pope, dosage, then held constant) or placebo (n = 30) for
1988; Kaplan, 1987; Kaplan, Garfinkel, Darby, 10 weeks (Nickel et al., 2005). Topiramate was asso-
& Garner, 1983). The first six patients received 6-​ ciated with significant decreases in the frequency of
week intervals of placebo–​carbamazepine–​placebo binge eating/​purging (defined as a > 50% reduc-
or carbamazepine–​placebo–​carbamazepine over tion; 37% for topiramate and 3% for placebo);
18 weeks. The next 10 patients received two 6-​ body weight (difference in weight loss between the
week intervals of placebo–​carbamazepine or two groups = 3.8 kg); and all scales on the SF 36
carbamazepine–​placebo over 12 weeks. There was Health Survey (all p’s < .001). Five (17%) patients
no significant difference in response between carba- on topiramate and six (20%) patients on placebo
mazepine and placebo. One patient had a complete were considered dropouts. All patients tolerated
remission of binge eating, one patient had a marked topiramate well.
response, and three additional patients improved In addition, topiramate has been reported to
on carbamazepine compared with baseline but reduce binge eating and/​or purging in BN patients
did not show a difference on drug compared with with treatment-​resistant illness, those with comor-
placebo. Of note, the patient who had a remission bid mood or personality disorders, and those receiv-
had comorbid cyclothymic disorder; she showed ing the drug as adjunctive therapy in combination
marked improvement in both her mood and her with antidepressants, mood stabilizers, and/​ or
bulimic symptoms while receiving carbamazepine. antipsychotics (Barbee, 2003; Bruno, Riganello, &
The first study of topiramate in BN was a 10-​ Marino, 2009; Felstrom & Blackshaw, 2002). There
week trial in 69 acutely ill patients (Hedges et al., is also a report of topiramate decreasing binge eat-
2003; Hoopes et al., 2003). Twenty-​two (63%) of ing in a woman with BN and epilepsy (Knable,
35 topiramate recipients and 18 (53%) of 34 pla- 2001). Her BN antedated her epilepsy and had not
cebo recipients completed the trial. Topiramate responded to 5 years of treatment with phenytoin,
(median dose 100 mg/​day; range 25 to 400 mg/​day) which had been effective in preventing her seizures.
was superior to placebo in reducing the frequency of Regarding other antiepileptics in BN, an open-​
binge eating and purge days (days during which at label trial in 12 patients found zonisamide reduced
least one binge eating or purging episode occurred; binge eating and purging symptoms but was associ-
p  =  .004); the bulimia/​uncontrollable overeating ated with a high dropout rate (Guerdjikova, Blom,
(p = .005), body dissatisfaction (p = .007), and drive Martens, Keck, & McElroy, 2013). There was no
for thinness (p  =  .002) subscales of the EDI; the change in body weight. Valproate was effective in
bulimia/​food preoccupation (p = .019) and dieting three hospitalized women with BN and comorbid
(p = .031) subscales of the EAT; the mean HAM-​A rapid-​cycling bipolar disorder who were previously
score (p = .046); and body weight (mean decrease inadequately unresponsive to lithium and antipsy-
of 1.8  kg for topiramate versus 0.2  kg increase chotics (Herridge & Pope, 1985; Hudson & Pope,
for placebo; p  =  .004). Significantly more topira- 1988; McElroy, Keck, & Pope, 1987). Two patients
mate recipients than placebo recipients reported received valproate alone, and one received valpro-
improvement on the Patient Global Improvement ate in combination with lithium. All three patients
scale (p  =  .004). The percentage of patients who showed marked improvement of both bulimic and
achieved ≥ 50% reduction in the number of binge mood symptoms. As noted earlier, carbamazepine

372 Pharmacotherapy
was reported to be effective in a patient with BN In an open-​label study, the effects of an antian-
and comorbid cyclothymic disorder (Hudson & drogenic oral contraceptive (30 μg ethinyl estradiol
Pope, 1988; Kaplan, 1987; Kaplan et  al., 1983). plus 3 mg drospirenone; Yasmin) were evaluated in
Finally, lamotrigine has been reported to be help- 21 women with BN and 17 age-​and BMI-​matched
ful in patients with BN and co-​occurring affective controls (Naessén, Carlström, Byström, Pierre, &
dysregulation (Trunko, Schwartz, Marzola, Klein, Hirschberg, 2007). Before treatment, women with
& Kaye, 2014). BN had a higher frequency of menstrual distur-
bances, higher acne and hirsutism scores, and higher
5-​HT3 Receptor Antagonists levels of testosterone, but lower meal-​related cho-
Faris et al. (2000) conducted a 4-​week random- lecystokinin (CCK) secretion than controls. After
ized, placebo-​ controlled trial of ondansetron, a 3  months of treatment, meal-​related hunger and
potent and selective antagonist of the 5-​HT3 recep- gastric distention were decreased in BN women.
tor, in 26 women with severe BN. To be enrolled, Meal-​ related CCK secretion was unchanged in
patients had a minimum frequency of seven cou- BN women but decreased in control women.
pled episodes of binge eating followed by self-​ Testosterone and free testosterone were decreased
induced vomiting per week for at least 6  months. in patients and controls. Frequency of self-​induced
Ondansetron (n = 14), which was self-​administered vomiting decreased during treatment (p < .05), but
in 4-​mg capsules up to six per day upon the urge binge eating and weight phobia were not signifi-
to binge eat or vomit, was associated with a signifi- cantly changed. Compared with nonresponders, the
cantly greater decrease in frequency of binge eating/​ six (29%) responders had significantly higher levels
vomiting episodes (p < .001) and with a significant of total and free testosterone, binge eating, and self-​
increase in normal meals consumed (p < .03) com- induced vomiting at baseline, but lower levels of
pared with placebo (n  =  12). The drug was also weight phobia. Reduced frequency of vomiting cor-
associated with significant improvement in the time related with reduced testosterone levels (r’s = .50, p
spent engaging in bulimic behaviors (p < .05). There < .05). The authors concluded that antiandrogenic
was no difference in weight change between groups. oral contraceptives might be a treatment strategy for
One patient receiving ondansetron discontinued women with BN and hyperandrogenic symptoms.
due to accidental injury. There is one RCT of oxytocin in BN (Kim,
Eom, Yang, Kang, & Treasure, 2015). In a single
Hormonal Treatments dose, placebo-​controlled, crossover study, 34 BN
One RCT of an antiandrogenic compound patients, along with 35 AN patients and 33 healthy
has been conducted in women with BN. Forty-​ controls, received oxytocin 40 IU intravenously fol-
six women meeting the DSM-​ IV criteria for lowed by an emotion recognition task and an apple
BN, purging type, were randomized to flutamide juice drink. The BN patients showed a decrease in
(n  =  9), citalopram (n  =  15), flutamide plus cita- 24-​hour caloric consumption and enhanced emo-
lopram (n = 10), or placebo (n = 12) for 3 months tion recognition, while AN patients showed no
(Sundblad, Landén, Eriksson, Bergman, & response on either outcome. In healthy controls,
Eriksson, 2005). Final flutamide and citalopram oxytocin produced enhanced emotional sensitivity
doses were 500 mg/​day and 40 mg/​day, respectively. but no impact on calorie consumption.
Ten patients did not complete the trial. On a self-​
rated global assessment of symptom intensity, all Opioid Antagonists
three active treatment groups were superior to pla- Four RCTs of opioid antagonists in the treatment
cebo. A comparison of all flutamide-​recipients ver- of BN have been conducted. In the first, 16 nor-
sus placebo-​recipients showed significant reductions mal weight women with DSM-​III BN completed
in global ratings (p = .03) and binge eating (p = .02) a 6-​week, placebo-​controlled, crossover trial of nal-
but not vomiting. Binge eating was significantly trexone (50 mg/​day) (Mitchell et  al., 1989). No
reduced only in the group given the combination significant differences in frequency of binge-​eating
of flutamide and citalopram (p  =  .04); vomiting or vomiting episodes between active drug and pla-
was not significantly decreased in any group. Dry cebo were found. In a subsequent study, 28 women
skin was the most common side effect with flu- with DSM-​III-​R BN and 41 obese patients with
tamide. Two patients discontinued flutamide for binge eating were randomized to receive naltrex-
moderate but reversible increases in serum trans- one (100–​150 mg/​d), imipramine, or placebo for
aminase levels. 8 weeks (Alger, Schwalberg, Bigaouette, Michalek,

McElroy, Guerdjikova, Mori, Keck Jr. 373

& Howard, 1991). Among all patients, there was Lithium
no change in frequency or duration of binge eating. In a RCT in 91 patients with BN, lithium (mean
Among the 22 BN patients who completed, nal- level .62 mEq/​L) was not superior to placebo in
trexone was associated with a significant reduction decreasing binge eating episodes, except possibly
in binge-​eating duration (p = .02), but not binge-​ in depressed patients (Hsu, Clement, Sandhouse,
eating frequency. In a study of 13 patients with BN & Ju, 1991). Importantly, there were no serious
receiving naltrexone up to 200 mg/​day or placebo adverse events.
in individualized crossover 6-​week trials, significant Case reports have described the successful
reductions in binge eating, purging, urges to binge treatment with lithium of patients with BN and
eat, and urges to purge were seen with active drug comorbid bipolar disorder (McElroy, Kotwal,
(Marrazzi, Bacon, et al., 1995). In a fourth study of Keck, & Akiskal, 2005). Lithium was effec-
41 women, intravenous administration of the opi- tive for eating and mood disorder symptoms in a
oid antagonist naloxone selectively suppressed the woman with BN who became manic on imipra-
consumption of sweet high-​fat foods in obese and mine (Shisslak, Perse, & Crago, 1991). In another
lean subjects with BN (n = 20), but not in controls report, one of two women with BN and rapid
(n = 21; Drewnowski, Krahn, Demitrack, Nairn, & cycling bipolar disorder responded to lithium with
Gosnell, 1995). imipramine; the other woman failed lithium alone
Open studies suggest some BN patients, includ- (Leyba & Gold, 1988). In yet another report, two
ing those resistant to antidepressants and psycho- of three men with BN and bipolar (n = 2) or cyclo-
therapy and those with type 1 diabetes, may respond thymic (n  =  1) disorders responded to lithium
when treated with doses of naltrexone up to 400 alone (n = 1) or lithium plus imipramine (n = 1);
mg/​day (Raingeard, Courtet, Renard, & Bringer, one patient failed lithium alone (Pope, Hudson,
2004). In a comparison of standard-​dose (50–​100 & Jonas, 1986).
mg/​day) versus high-​dose (200–​300 mg/​day) nal-
trexone in 16 patients with antidepressant-​resistant Weight-​Loss Drugs
BN, participants in the standard-​dose group had no Three RCTs of fenfluramine or its isomer
significant change in frequency of binge eating or dexfenfluramine in BN were conducted before
purging after 6 weeks of treatment, whereas partici- the drug was removed from the worldwide mar-
pants in the high-​dose group had significant reduc- ket in 1997 because of concerns it caused cardiac
tions in both behaviors (Jonas & Gold, 1988). valvular abnormalities (Colman, 2005; Gardin
et al., 2000). In one study, fenfluramine was com-
Drugs for Attention Deficit Hyperactivity pared with desipramine in a 15-​week randomized,
Disorder placebo-​controlled crossover trial in 22 outpatients
In a double-​blind, randomized, crossover trial, with BN (Blouin et al., 1988). Both drugs reduced
eight patients with BN were given methylamphet- binge eating and vomiting frequency, but a greater
amine or placebo intravenously followed by a test proportion of patients responded to fenfluramine
meal and separated by a 1-​ week interval (Ong, than to desipramine. In another study, 42 patients
Checkley, & Russell, 1983). Significantly fewer mean with BN were randomized to dexfenfluramine or
(SD) calories were consumed after methylamphet- placebo for 12 weeks (Russell, Checkley, Feldman,
amine (224 [111]) than after placebo (943 [222]; p & Eisler, 1988). Dexfenfluramine was associated
< .02). In addition, “the frequency of bulimia” was with significantly greater decreases in binge eating
significantly lower after methylamphetamine (zero and self-​induced vomiting, but not in measures of
of eight patients) than after placebo (four of eight depressive symptoms. In the third trial, however,
patients; p < .05). Importantly, a growing number of dexfenfluramine plus CBT was not superior to pla-
case reports have described the successful use of meth- cebo plus CBT in reducing bulimic or depressive
ylphenidate in treating patients with BN, including symptoms in 43 women with BN (Fahy, Eisler, &
those resistant to psychotherapy and antidepressants Russell, 1993).
and those with comorbid cluster B personality dis- No other RCTs of weight-​loss drugs in BN have
orders or ADHD (Drimmer, 2003; Dukarm, 2005; been published. However, misuse of the lipase inhib-
Guerdjikova & McElroy, 2013; Keshen & Ivanova, itor orlistat by patients with BN has been reported
2013; Schweickert, Strober, & Moskowitz, 1997; (Deb, Gupta, & Varshney, 2014; Fernández-​Aranda
Sokol, Gray, Goldstein, & Kaye, 1999). et al., 2001; Malhotra & McElroy, 2002).

374 Pharmacotherapy
Prokinetic Agents most BED trials have been short term (6–​21 weeks)
In the only RCT with a prokinetic agent, 29 and conducted in patients who were actively binge
patients with BN were randomized to receive eryth- eating. The primary outcome has usually been a
romycin up to 500 mg three times daily or placebo measure of binge eating episode or binge eating day
for 6 weeks (Devlin et al., 2012). Thirteen patients frequency, rate of response or cessation of binge eat-
in each group completed the trial. Treatment ing behavior, or change in Binge Eating Scale (BES)
with erythromycin showed no beneficial clinical scores (Gormally, Black, Dasson, & Rardin, 1982).
effect:  Patients receiving erythromycin had weekly Secondary outcomes have included measures of ED
mean (SD) binge eating/​vomit frequencies of 11.4 psychopathology (often assessed with the Three
(10.7)/​11.3 (10.9), while those receiving placebo Factor Eating Questionnaire) (Stunkard & Messick,
had weekly binge eating/​vomit frequencies of 7.2 1985); mood, anxiety, obsessive-​ compulsive, and
(4.1)/​7.6 (4.4). impulsive symptoms; global clinical improvement;
weight, BMI, and other metabolic parameters; and
Antipsychotics adherence. Only one placebo-​ controlled mainte-
There are no published RCTs of antipsychotics nance monotherapy trial in a group of BED patients
in BN. Case reports of the successful use of aripip- whose binge eating had responded to an acute inter-
razole in the treatment of patients with BN have vention has been conducted (Hudson, McElroy,
been described (Takaki & Okabe, 2015; Trunko, Ferreira-​Cornwell, Radewonuk, & Gaisor, 2017).
Schwartz, Duvvuri, & Kaye, 2011). However, there
are also reports of second-​generation antipsychotics Drugs for Attention Deficit Hyperactivity
inducing or exacerbating BN symptoms in patients Disorder
receiving the drugs (Brewerton & Shannon, 1992; A prodrug of d-​amphetamine, LDX is approved
Crockford, Fisher, & Barker, 1997; Gebhardt for treatment of children and adults with ADHD;
et al, 2007). it is the only drug that has regulatory approval for
the treatment of BED. Specifically, it is approved
Other Medications in the United States for adults with moderate-​to-​
A randomized, placebo-​ controlled trial in 93 severe BED. This approval was based on two 12-​
women with DSM-​IV BN found no effect with week phase 3 studies that each found that LDX,
spironolactone, a diuretic with mineralocorti- titrated to 50 mg/​day or 70 mg/​day, was efficacious
coid and aldosterone antagonistic properties (von for reducing binge eating (McElroy et al., 2016). An
Wietersheim et al., 2008). By contrast, in an open-​ earlier phase 2 study found that LDX at 50 mg/​day
label trial, the gamma-​aminobutyric acid (GABA) B and 70 mg/​day, but not 30 mg/​day, was efficacious
agonist baclofen, given at 60 mg/​day for 10 weeks, for reducing binge eating (McElroy et al., 2015). All
reduced binge eating in two of three patients with three studies also found that LDX was superior to
BN (Broft et al., 2007). However, a case of a woman placebo for inducing cessation of binge eating and
with an alcohol use disorder and BN reported that for reducing ED psychopathology, obsessive-​com-
alcohol craving but not food craving responded to pulsive features of binge eating, and body weight.
high-​dose baclofen (Weibel, Lalanne, Riegert, & In the phase 2 study, LDX was also superior to pla-
Bertschy, 2015). Finally, in a 12-​week, open-​label cebo for reducing trait impulsivity (McElroy et al.,
trial of a N-​acetylcysteine (NAC) in eight patients 2017). The tolerability and safety profile of LDX
with BN, no patient improved and NAC was associ- was consistent with previous findings in adults with
ated with a high discontinuation rate (Guerdjikova, ADHD.
Blom, Mori, & McElroy, 2013). The anti–​binge eating effect of LDX was shown
to persist for 6 months in a double-​blind, placebo-​
Pharmacotherapy of Binge Eating Disorder controlled, randomized withdrawal study (Hudson
Randomized, placebo-​ controlled studies of et al., 2015). Patients who responded to 12 weeks
BED have been conducted primarily with four of open-​label treatment with LDX 50 or 70 mg/​
drug classes: ADHD medications, antidepressants, day (n = 275) were randomized to receive continued
weight-​loss agents, and antiepileptics (Table 19.3). LDX or switched to placebo for 6  months. Based
In most studies, medication was used as monother- on a Cox proportional hazards model, the estimated
apy; in a few studies medication was used adjunc- hazard for relapse with LDX was 11 times less than
tively with psychological interventions. To date, placebo (hazard ratio  =  0.09). Specifically, 5% of

McElroy, Guerdjikova, Mori, Keck Jr. 375

Table 19.3  Medications Studied for Binge Eating Disorder in Randomized, Placebo-​Controlled
Trials: Qualitative Results
Medication Maximum Dosage Reduction of Reduction of
Studied (mg/​day) Binge Eating Body Weight

ADHD Medications

 Atomoxetine 120 + +

 Lisdexamfetamine 70 +++ +++

Tricyclic Antidepressants

 Imipramine 200 − +

SSRI Antidepressants

 Citalopram 60 + +

 Escitalopram 30 + +

 Fluoxetine 80 + +

 Fluvoxamine 300 + +

 Sertraline 200 + +

SNRI Antidepressants

 Duloxetine 120 + +

Other Antidepressants

 Bupropion 300 − +

Weight-​loss Drugs

 Dexfenfluraminea 30 +++ −

 Orlistat 360 + ++

 Sibutramine a
15 ++ ++

Antiepileptics

 Lamotrigine 400 − +/​−

 Topiramate 400 ++ ++

 Zonisamide 600 + +

Key: +++ = ≥2 positive RCTs and evidence from ≥1 RCT for maintenance of efficacy; ++ = ≥ 2 positive RCTs; + = ≥1 positive RCT;
+/​− = mixed data; − = only negative data; ADHD = attention deficit hyperactivity disorder; SSRI = selective serotonin reuptake inhibitor.
a
Removed from the market for safety concerns.

LDX-​treated patients relapsed as compared with There were minor increases in pulse and blood pres-
34.4% of placebo-​treated patients. sure and a modest decrease in body weight. There
In a 12-​month open-​label safety and tolerability was no evidence of new safety concerns, including
study in 604 BED patients, the safety and tolerabil- after abrupt discontinuation of LDX.
ity of LDX was consistent with its safety profile for In the only RCT of a nonstimulant ADHD drug
treatment of ADHD (Gaisor et al., 2017). The most in BED, 40 patients were randomized to receive
common adverse events were dry mouth, headache, atomoxetine, a highly selective NRI, or placebo
insomnia, and upper respiratory tract infection. for 10 weeks (McElroy, Guerdjikova, et al., 2007).

376 Pharmacotherapy
Atomoxetine was flexibly dosed from 40 to 120 bupropion, 61 overweight or obese patients received
mg/​day; the mean (SD) dose at endpoint evalua- the drug or placebo for 8 weeks (White & Grilo,
tion was 106(21) mg/​day. Compared with placebo-​ 2013). Bupropion was similar to placebo in reduc-
treated patients (n  =  20), atomoxetine-​treated ing binge eating but associated with greater weight
patients (n = 20) showed a significantly greater rate loss. Moreover, bupropion was well tolerated and
of reduction in binge eating episode frequency (the there were no seizures. In the only study of BED
primary outcome measure), as well as decreases in with a co-​occurring depressive disorder, duloxetine
binge day frequency, weight, BMI, and measures (mean dose 78.7 mg/​day) was superior to placebo
of global severity, obsessive features of binge eating, in reducing binge eating, global severity of BED
and hunger. Fifteen patients (six receiving atomox- and depressive symptoms, and body weight after
etine) did not complete the 10-​week trial. The most 12 weeks of treatment (Guerdjikova et  al., 2012).
common side effects associated with atomoxetine Similarly, a retrospective chart review found that the
were dry mouth, nausea, nervousness, insomnia, SNRI venlafaxine was effective for reducing binge
headache, constipation, and sweating. Three atom- eating and body weight in 33 patients with BED
oxetine recipients discontinued because of increased and obesity (Malhotra, King, Welge, Brunsman-​
depressive symptoms, constipation, and nervous- Lovins, & McElroy, 2002).
ness (n = 1 each, respectively). A meta-​analysis of seven of these studies (one
The only other selective NRI studied in BED with a tricyclic, six with SSRIs) showed signifi-
has been reboxetine. In a 12-​week open-​label trial cantly higher binge eating remission rates for the
in nine patients, significant reductions were seen in antidepressant group compared with the placebo
binge eating frequency and BMI (Silveira, Zanatto, group:  40.5% versus 22.2% (relative risk  =  0.77
Appolinário, & Kapezinski, 2005). [95% CI  =  0.65, 0.92; p  =  .003]) (Stefano,
Bacaltchuk, Blay, & Appolinario, 2008). Evaluating
Antidepressants studies that used the Hamilton Depression Rating
To date, at least 10 randomized, placebo-​ Scale to evaluate change in depressive symptoms,
controlled studies of antidepressants have been a statistically significant difference between groups
published in patients with BED as defined by all or was also seen favoring antidepressants (SMD  =  –​
most of the DSM-​IV criteria (Arnold et al., 2002; 0.38 [95% CI = –​74, –​0.03; p = .03]). However,
Grilo, Masheb, & Wilson, 2005; Guerdjikova et al., no differences between groups were found in the
2008; Guerdjikova et  al., 2012; Hudson et  al., mean frequency of binge eating episodes at the
1998; Laederach-​Hofmann et  al., 1999; McElroy end of treatment (SMD = –​0.36 [95% CI = –​0.74,
et al., 2000, 2003; Pearlstein et al., 2003; White & 0.01; p = 0.06]), in BMI (SMD = 0.03 [95% CI = –​
Grilo, 2013). Eight studies evaluated antidepressant 0.49, 0.55]), or in treatment discontinuation for
monotherapy, and two compared antidepressants in any reason (relative risk  =  1.35 [95% CI  =  0.61,
combination with weight loss therapy (Laederach-​ 3.00]). In light of the frequent chronicity of BED,
Hofmann et al., 1999) or CBT (Grilo, Masheb, & it was concluded that the data were not sufficient
Wilson, 2005). All studies were small, consisting of to formally recommend antidepressants as a single
15 to 85 patients; of short duration, ranging from first-​line therapy for short-​term remission of binge
6 to 16 weeks; and had substantial dropout rates. eating episodes and weight reduction in BED
Antidepressant families studied included SSRIs patients.
(n = 7), tricyclics (n = 1), SNRIs (n = 1), and bupro- Controlled combination therapy studies have
pion (n  =  1). The SSRI doses were comparable to had contrasting results. In one, diet counseling with
those used in BN. A study of fluoxetine used doses psychological support plus imipramine was superior
up to 80 mg/​day (Arnold et al., 2002) and a study to diet counseling and psychological support plus
of escitalopram used supratherapeutic doses (up to placebo for 8 weeks in decreasing binge eating (p
30 mg/​day; Guerdjikova et al., 2008). < .01) and weight (p < .001; Laederach-​Hofmann
When viewed collectively, SSRIs led to greater et  al., 1999). In the other, a 16-​week trial, CBT
reductions in binge eating, ED psychopathology, with placebo and CBT with fluoxetine were both
and body weight than placebo, but were associ- superior to fluoxetine alone and placebo alone for
ated with substantial dropout rates (16%–​57%; decreasing binge eating (Grilo, Masheb, & Wilson,
Brownley, Berkman, Sedway, Lohr, & Bulik, 2007). 2005). There was no difference between fluoxe-
Also, most weight reductions would not be con- tine and placebo. No treatment was effective for
sidered clinically significant. In the only study of weight loss.

McElroy, Guerdjikova, Mori, Keck Jr. 377

 Open-​ label data have suggested some BED pressure, and insulin level were also significantly
patients who initially respond to an antidepressant improved with orlistat. Effectiveness of orlistat for
with decreased binge eating and weight loss may binge eating behavior was less clear:  at 24 weeks,
maintain these beneficial effects for up to 6 months the mean number of weekly binge eating episodes
with continuation of the antidepressant (Leombruni was numerically but not significantly decreased
et al., 2006; Leombruni, Pierò, et al., 2008). The (1.0 for orlistat-​treated patients vs. 1.7 for placebo-​
only randomized, placebo-​controlled maintenance treated patients). Also, similar rates of patients in
trial of an antidepressant published in BED, how- both groups who completed the study continued
ever, suggested that benefits may not be maintained to suffer from BED (23% for orlistat vs. 29% for
over longer periods of antidepressant treatment. placebo). However, the Eating Disorder Inventory
In that study, 116 BED patients who achieved a 12 score at week 24 was significantly lower for orli-
75% or greater reduction in binge frequency after stat than placebo (p = .011). In addition, fat intake
a 5-​month initial phase of group behavioral weight was significantly lower in orlistat-​treated patients at
control therapy received maintenance weight con- week 12; total caloric intake was significantly lower
trol treatment for up to 24 months. Patients were at week 24. Eighteen patients discontinued the
randomized twice: to fluoxetine or placebo, and study prematurely: five (11%) in the orlistat group
to CBT or no CBT. Results showed fluoxetine and 13 (29%) in the placebo group. No patient dis-
appeared to be effective for depressive symptoms, continued orlistat because of an adverse event. Data
but not for binge eating or weight reduction on side effects were otherwise not reported.
(Devlin et al., 2005; Devlin, Goldfein, Petkova, In the third study, 79 obese Spanish-​speaking-​
Liu, & Walsh, 2007). only Latino/​as with BED (n = 40) or without BED
Of note, several promising types of antidepres- (n = 39) at a community mental health center were
sants have not yet been studied in BED in random- randomly assigned to 4 weeks of orlistat 120 mg
ized, placebo-​controlled trials. These include several three times daily plus behavioral weight loss ther-
SNRIs, vilozodone, and vortioxitine. apy (BWLT) or placebo plus BWLT (Grilo &
White, 2013). Seventy-​ eight percent of patients
Weight-​Loss Agents completed the trial. Among BED patients, remis-
A number of weight-​loss agents have received at sion of binge eating did not differ between orlistat
least some systematic study in obese patients with (60%) and placebo (70%), and there was no differ-
BED. Three randomized, placebo-​controlled stud- ence in improvement in ED psychopathology, BMI,
ies have been conducted with the lipase inhibitor or depressive symptoms. Orlistat plus BWLT pro-
orlistat. In the first study, 50 patients with BED duced greater weight loss than placebo plus BWLT
and obesity were randomized to 12 weeks of guided in patients with obesity alone, but not in patients
self help CBT (CBTgsh) with orlistat 120 mg three with BED. Of note, there are reports of patients
times daily or placebo (Grilo, Masheb, & Salant, with binge eating who misuse orlistat (Cochrane
2005). Binge eating remission rates were signifi- & Malcolm, 2002; Deb et  al., 2014; Malhotra &
cantly higher for orlistat plus CBTgsh (64% vs. McElroy, 2002).
36%) at post-​treatment but not at 3-​month follow-​ The glucagon-​like peptide-​1 (GLP-​1) receptor
up (52% in both groups). Rates for achieving at agonist liraglutide 1.8 mg daily sc injection was
least 5% weight loss were significantly higher for tested in 44 nondiabetic obese participants with
orlistat plus CBTgsh than placebo plus CBTgsh at subclinical binge eating behavior, defined as a BES
both post-​treatment (36% vs. 8%) and 3-​month score ≥ 18, in a 12-​week trial (Robert et al., 2015).
follow-​up (32% vs. 8%). In the second study, 89 Participants were randomized to receive liraglutide
patients with BED and obesity were randomized to (n = 21) plus diet and exercise or to diet and exer-
orlistat 120 mg three times daily (n = 44) or placebo cise alone (n  =  21). Participants receiving liraglu-
(n  =  45), in combination with a mildly reduced tide showed significant reductions in binge eating
calorie diet, for 24 weeks (Golay et  al., 2005). At behavior, body weight, BMI, waist circumference,
endpoint, the mean percentage weight loss (the pri- systolic blood pressure, fasting glucose, and total
mary outcome measure) for orlistat-​treated patients cholesterol. This study was limited by the lack of
was significantly greater than for placebo-​treated use of placebo. Of note, we found no studies of lira-
patients (–​7.4% vs. –​2.3%; p = .0001). Waist cir- glutide 3 mg daily sc injection in BED, the dose
cumference, hip circumference, total percentage of liraglutide approved for the treatment of obesity
body fat, total cholesterol level, diastolic blood (Bray, Frühbeck, Ryan, & Wilding, 2016).

378 Pharmacotherapy
 In another open-​label study, phentermine, up obese patients with BED were randomized to active
to 30 mg/​day, in combination with fluoxetine, up drug or placebo for 6 months (Pataky et al., 2013).
to 60 mg/​day, and CBT was assessed in 16 obese All patients were also prescribed a mild hypocaloric
women, 14 of whom met DSM-​ IV criteria for diet. Rimonabant-​treated patients showed a greater
BED (Devlin, Goldfein, Carino, & Wolk, 2000). change in body weight (the primary outcome)
During the 20-​week active treatment phase, mean compared with placebo-​ treated patients:  Patients
weekly binge eating frequency declined by 95% receiving rimonabant lost 4.7% of their initial body
from baseline, with 12 (75%) patients showing weight while patients receiving placebo lost 0.4%.
complete remission. In addition, mean body weight Rimonabant-​ recipients also showed significant
and BMI declined by 8.6% and 8.7%, respectively, reductions in BES scores.
from baseline. After 6 months of maintenance treat-
ment, 10 patients were still taking both medications Antiepileptic Drugs
(two patients were taking fluoxetine alone); though Five randomized, placebo-​controlled studies of
binge eating frequency was 63% lower than at the antiepileptic drugs (three for topiramate, one for
start of treatment, only five (42%) of 12 patients zonisamide, and one for lamotrigine) in DSM-​
were free of binge eating. Only six patients com- IV–​defined BED and two small crossover studies
pleted 18  months of maintenance therapy; two of phenytoin in the similar condition of compul-
were taking both medications and four were taking sive or binge eating have been conducted (McElroy
fluoxetine alone. Though binge eating remained et al., 2009).
improved, patients had regained most of the weight The three RCTs of topiramate each found that
they had lost at the end of active treatment. topiramate reduced binge eating and excessive body
In a 24-​week, open-​label trial in 23 obese or weight in BED. In the first study, 61 BED patients
overweight women with current major depres- with obesity received topiramate (n  =  30) or pla-
sive disorder, most of whom had clinically signifi- cebo (n = 31) for 14 weeks (McElroy, Arnold, et al.,
cant binge eating, the combination of naltrexone 2003). Topiramate (median dose 212 mg daily) was
and bupropion (as sustained release formula- significantly superior to placebo in reducing binge
tions) reduced binge eating behavior, depressive eating frequency, as well as obsessive-​compulsive
symptoms, and body weight at 16 and 24 weeks features of binge eating, global illness severity, body
(McElroy, Guerdjikova, Kim, et  al., 2013). Mean weight, and BMI. Topiramate-​treated patients expe-
(SD) percent weight loss was –​4.0 (4.6) at week 12 rienced a 94% reduction in binge frequency and a
and –​5.3 (6.5) at week 24. Finally, two patients with mean weight loss of 5.9 kg, whereas placebo-​treated
BED and obesity have been described who had ces- patients experienced a 46% reduction in binge fre-
sation of binge eating behavior and clinically sig- quency and a mean weight loss of 1.2 kg. Dropout
nificant weight loss in response to treatment with rate, however, was high:  14 (47%) topiramate
the combination of phentermine and topiramate recipients and 12 (39%) placebo recipients failed to
(Guerdjikova, Fitch, & McElroy, 2015). complete the trial. The most common side effects
Three weight-​loss agents that have been removed associated with topiramate were paresthesias, dry
from the market because of safety concerns have mouth, headache, and dyspepsia.
also been evaluated in RCTs in BED. Thus, sibutra- The second controlled study was a multicenter
mine, a reuptake inhibitor of norepinephrine, 5-​ trial in which 407 patients with BED and ≥ three
HT, and, to a lesser extent, dopamine, was shown to binge eating days/​week, a BMI between 30 and
reduce both binge eating and excessive body weight 52  kg/​m2, and no current psychiatric disorders
in BED in three placebo-​controlled, randomized or substance abuse were randomized to receive
trials (Appolinario et al., 2003; Milano et al., 2005; topiramate or placebo for 16 weeks (McElroy,
Wilfley et al., 2008). In the only RCT of dexfenflu- Hudson, et  al., 2007). Compared with placebo,
ramine, 28 women with BED and obesity received topiramate (median final dose 300 mg daily) sig-
active drug, up to 30 mg/​day, or placebo for 8 weeks nificantly reduced binge eating days/​week, binge
(Stunkard, Berkowitz, Tanrikut, Reis, & Young, eating episodes/​week, body weight, and BMI (all
1996). The rate of binge eating fell three times more p’s < .001). Topiramate also significantly decreased
rapidly in the dexfenfluramine group than the pla- obsessive-​compulsive features of binge eating, dis-
cebo group, but no significant weight changes were inhibited eating, and hunger; trait impulsivity; and
observed. In a study of the selective endocannabi- measures of overall, social, and family life disabil-
noid CB1 receptor inverse agonist rimonabant, 289 ity. Significantly more topiramate-​treated subjects

McElroy, Guerdjikova, Mori, Keck Jr. 379

(58%) achieved remission compared with placebo-​ also with a high discontinuation rate (McElroy,
treated subjects (29%; p < .001). Discontinuation Kotwal, Hudson, Nelson, & Keck, 2004). It is also
rates were 30% in each group; adverse events were consistent with an open-​label study of zonisamide
the most common reason for topiramate discon- given in conjunction with CBT in patients with
tinuation (16%; placebo, 8%). Paresthesias, upper threshold and subthreshold BED (Ricca, Castellini,
respiratory tract infection, somnolence, and nausea LoSauro, Rotella, & Faravelli, 2009). In this study,
were the most frequent topiramate side effects. 28 patients received zonisamide while 24 received
The third controlled study of topiramate in BED CBT alone for 6  months. At end of treatment,
was another multicenter trial in which 73 patients zonisamide recipients showed greater reductions in
with BED and obesity were randomized to 19 BMI and BES scale scores.
sessions of CBT in conjunction with topiramate In the fifth RCT of an antiepileptic, 51 patients
(n = 37) or placebo (n = 36) for 21 weeks (Claudino with BED were randomized to receive lamotrigine
et al., 2007). Compared with patients receiving pla- or placebo for 16 weeks (Guerdjikova et al., 2009).
cebo, patients receiving topiramate showed a signif- Lamotrigine (mean dose 236 mg/​day) and placebo
icantly greater rate of reduction in body weight, the had similar rates of reduction of weekly frequency
primary outcome measure (p < .001). Topiramate of binge eating episodes and binge eating days,
recipients lost a mean of 6.8  kg While placebo weight, and BMI; and measures of eating pathol-
recipients lost 0.9  kg. Rates of reduction of binge ogy, obsessive-​compulsive symptoms, impulsivity,
eating frequencies and BES and depression scores and global severity of illness. However, lamotrigine
did not differ between the groups, but a greater was associated with a numerically greater amount of
percentage of topiramate-​ treated patients (84%) weight loss (1.17 vs. 0.15 kg) and significant reduc-
attained remission of binge eating as compared with tions in fasting levels of glucose, insulin, and triglyc-
placebo-​treated patients (61%; p = .03). There was erides. It was also well tolerated and associated with
no difference between groups in completion rates, no serious adverse events.
though one topiramate recipient withdrew for an The two small placebo-​controlled crossover trials
adverse effect. Paresthesias and taste perversion were of phenytoin in patients with compulsive or binge
more frequent with topiramate, whereas insomnia eating had contrasting results (Hudson & Pope,
was more frequent with placebo. 1988). In the negative trial, four obese patients with
In the only RCT of zonisamide, 60 outpatients compulsive eating showed no significant differences
with DSM-​ IV BED and obesity received flexi- between phenytoin and placebo on any outcome
bly dosed (100–​600 mg/​day) zonisamide (n  =  30) measure and no patient had a marked response to
or placebo (n  =  30) for 16 weeks (McElroy et  al., phenytoin (Greenway, Dahms, & Bray, 1977). In
2006). Compared with placebo, zonisamide was the positive study, 19 of 20 women with “binge eat-
associated with a significantly greater rate of reduc- ing syndrome” completed 12 weeks wherein they
tion in binge eating episode frequency (p  =  .021), received phenytoin and placebo for 6 weeks each in a
body weight (p < .0001), BMI (p  =  .0001), and counterbalanced design (Wermuth, Davis, Hollister,
measures of global severity, obsessive-​ compulsive & Stunkard, 1977). Twelve patients had final serum
features, and disinhibited eating (all p’s < .0001). phenytoin levels of 10 to 20 μg/​mL; five had levels of
The mean (SD) zonisamide daily dose at endpoint 5 to 10 μg/​mL. Patients given phenytoin first expe-
evaluation was 436 (159) mg. Attrition rate, how- rienced a 37% decrease in binge eating frequency (p
ever, was high, with 18 (60%) zonisamide patients < .01), but when administered placebo, binge eat-
and 12 (40%) placebo patients not completing ing frequency was unchanged. Patients given pla-
the 12-​ week treatment period. Eight zonisamide cebo first experienced no change in binge eating
recipients discontinued for adverse events. The most frequency, and they did experience a 39% decrease
common reasons for stopping zonisamide were cog- after switching to phenytoin (p < .01). Eight (42%)
nitive complaints (n = 2), psychological complaints patients displayed a moderate or better response (≥
(n = 2), and bone fracture (n = 2). The most com- 50% reduction in binge eating episode frequency)
mon side effects associated with zonisamide were dry on phenytoin, but only one patient experienced a
mouth, somnolence, headache, nausea, nervousness, remission of binge eating. When the two groups
and altered taste. were compared, there were significantly fewer eating
This trial was consistent with an earlier open-​ binges in the phenytoin–​placebo group than in the
label study in which zonisamide was associated placebo–​phenytoin group (p < .02), indicating a car-
with reduced binge eating and body weight but ryover effect for the phenytoin-​first sequence.

380 Pharmacotherapy
 There are no controlled relapse prevention stud- Gastaldi et  al., 2008). Four patients showed a
ies of antiepileptics in BED, but an open-​ label decrease in binge eating and three lost weight,
extension trial has suggested that the anti–​binge but three reported no change in binge eating, two
eating and weight loss effects of topiramate may be showed no weight change, and five gained weight.
maintained over the long term. The BED patients Five patients discontinued the drug, and seven
who completed the first RCT of topiramate (n = 35) reported side effects.
were offered participation in a 42-​week open-​label
extension study of the drug (McElroy, Shapira,
et al., 2004). Forty-​four patients (31 who received Opioid Antagonists
topiramate in the open-​ label trial plus 13 who Two RCTs of opioid antagonists have been pub-
received topiramate in the double-​blind study only) lished in BED. In the first, 33 obese binge eaters
received at least one dose of topiramate; 43 patients and 22 normal weight bulimics were treated for 8
provided outcome measures at a median final dose weeks with naltrexone (100–​150 mg/​day), imipra-
of 250 mg/​day. Mean weekly binge eating frequency mine, or placebo (Alger et  al., 1991). Naltrexone
declined significantly from baseline to final visit for did not significantly reduce binge eating frequency
all 43 patients (–​3.2; p < .001), for the 15 patients or duration in the obese binge eaters. In the second
who received topiramate during the controlled and trial, 62 patients with BED were randomized to the
open-​label studies (–​4.0; p < .001), and for the novel opioid antagonist ALKS-​33 (now called sami-
15 patients who received topiramate only during dorphan; n = 26) or placebo (n = 36) for 6 weeks
the open-​label trial (–​2.5; p  =  .044). Patients also (McElroy, Guerdjikova, Blom, et  al., 2013). Both
exhibited a statistically significant reduction in body drug and placebo produced similar large reductions
weight. However, only 10 (32%) of the 31 patients in binge eating episode frequency, raising the possi-
entering the extension trial completed the 42 weeks bility of a failed rather than negative trial. However,
of open-​label treatment; the most common reasons there were also no differences between drug and
for topiramate discontinuation were protocol non- placebo in other measures of binge eating, eating
adherence (n = 11) and adverse events (n = 8). psychopathology, or body weight.
There are also open-​label descriptions of anti- In contrast, two favorable case reports of naltrex-
epileptics being helpful in difficult-​to-​treat patients one in BED have been published. One was a posi-
with BED. Thus, topiramate has been reported tive on-​off-​on case of naltrexone monotherapy using
to reduce binge eating and/​or overweight in BED doses of 200 and 400 mg/​day (Marrazzi, Markham,
patients with treatment-​resistant illness, comorbid Kinzie, & Luby, 1995). The other was the success-
depressive or bipolar disorders, traumatic brain ful augmentation of fluoxetine using naltrexone 100
injury, and those receiving the drug adjunctively mg/​day (Neumeister, Winkler, & Wöber-​Bingöl,
with antidepressants and/​or mood stabilizers (De 1999). Additionally, a Phase 2 placebo-​controlled
Bernardi, Ferraris, D’Innella, Do, & Torre, 2005; study of intranasal naloxone spray has been reported
Dolberg, Barkai, Gross, & Schreiber, 2005; Kotwal, in abstract form but not yet published (Alho et al.,
Guerdjikova, McElroy, & Keck, 2006; Schmidt 2013). Specifically, 127 participants with BED were
do Prado-​Lima & Bacaltchuck 2002; Shapira, randomized to intranasal naloxone spray or intrana-
Goldsmith, & McElroy, 2000). Topiramate has been sal placebo spray for 24 weeks. Naloxone 2 mg was
successfully used to reduce binge eating and weight administered before each binge eating episode up to
gain after adjustable gastric banding or gastric a maximum of 4 mg/​day; 81% of participants com-
bypass surgery (Guerdjikova, Kotwal, & McElroy, pleted the trial. Naloxone produced a significantly
2005; Zilberstein et al., 2004). There is also a suc- greater reduction than placebo in time spent binge
cessful case of lamotrigine in the treatment of a eating. Also, BMI decreased significantly from week
61-​year-​old woman with BED, bipolar depression, 12 to week 24 among naloxone recipients but not
and treatment-​resistant type 2 diabetes (Yamamoto, placebo recipients. There were no serious adverse
Kanahara, Hirai, Watanabe, & Iyo, 2013). events.
By contrast, valproate has been reported to
worsen binge eating and enhance weight gain in Baclofen
patients with BED and comorbid bipolar disorder In one small crossover RCT in 12 individuals
(Shapira et al., 2000). In addition, a case series of with binge eating, participants were randomized to
nine obese patients with BED treated with oxcar- receive baclofen (titrated to 60 mg/​d) for 48  days
bazepine had inconsistent findings (Leombruni, followed by placebo for 48  days, or the reverse

McElroy, Guerdjikova, Mori, Keck Jr. 381

(Corwin, Boan, Peters, & Ulbrecht, 2012). Relative marketed for the treatment of excessive daytime
to the placebo phase, baclofen produced a slight sleepiness and cataplexy in patients with narcolepsy
but statistically significant reduction in binge eating (McElroy et  al., 2011). Sodium oxybate was asso-
frequency. The BES and food craving scores were ciated with significant reductions in binge eating
decreased similarly during baclofen and placebo frequency, obsessive-​compulsive features of binge
phases. By contrast, baclofen produced a small but eating, BMI, and body weight. However, only five
statistically significant increase in depressive symp- patients completed the 16-​week trial. Finally, in 12
toms. The most commonly reported side effects obese patients with BED and bipolar disorder, lith-
were tiredness, fatigue, and upset stomach. Of note, ium augmentation of topiramate improved mood
patients with BED who do not respond to 60 mg and reduced both binge eating and body weight
daily may respond to higher baclofen doses (up to (Kotwal et al., 2006).
180 mg daily) (De Beaurepaire, Joussaume, Rapp,
& Jaury, 2015). Pharmacotherapy of Night Eating
Syndrome
Chromium Two RCTs of pharmacotherapy have been pub-
Chromium is an essential nutrient that may lished in NES. In the first, 34 outpatients with
improve mood, appetite, and glucose regulation. NES were randomly assigned to receive sertraline
In one small RCT, 24 overweight or obese patients (n = 17) or placebo (n = 17) for 8 weeks (O’Reardon
with BED were randomized to one of three treat- et  al., 2006). Sertraline was flexibly dosed at 50
ments for 6  months each:  high-​dose (1,000 mcg to 200 mg/​day. On the primary outcome meas-
daily) chromium (n  =  8), moderate-​ dose (600 ure, the Clinical Global Impression-​Improvement
mcg daily) chromium (n  =  9), or placebo (n  =  7) scale, sertraline was significantly superior to pla-
(Brownley et  al., 2013). Numerically greater cebo:  12 (71%) sertraline-​ treated patients were
reductions in binge frequency, body weight, and classified as improved compared with three (18%)
depressive symptoms were observed in chromium placebo-​treated patients. There was also significant
recipients compared with placebo recipients, but improvement with sertraline in night eating symp-
reductions were not statistically significant. Fasting toms, global severity ratings, frequency of nocturnal
glucose was significantly reduced in both chromium ingestions and awakenings, caloric intake after the
groups, with larger effects noticed with the higher evening meal, and quality of life ratings. In addi-
dose of chromium. Chromium was well tolerated. tion, overweight and obese patients receiving ser-
traline lost a significant amount of weight by week
Antipsychotics 8 (mean = –​2.9 kg) compared with overweight and
There are no published RCTs of antipsychotics obese patients receiving placebo (mean = –​0.3 kg).
in the treatment of BED. Indeed, second-​generation In the other RCT, 40 patients with NES were ran-
antipsychotics have been reported to induce or exacer- domly assigned to receive escitalopram (n = 20) or
bate binge eating, including BED, in patients receiving placebo (n  =  20) for 12 weeks (Vanderwal, Gang,
the drugs for psychotic disorders (Theisen et al., 2003). Griffing, & Gadde, 2012). At study endpoint, there
was no difference in mean change of the Night
Other Agents Eating Questionnaire total score (the primary
The glutamate modulating agent memantine outcome measure). Similarly, there were no drug-​
has been reported to reduce binge eating in BED in placebo differences for changes in body weight or
two open-​label trials. In the first, five women with mood ratings.
BED and obesity received memantine 10 mg in Finally, there are open-​label reports of agomela-
the morning and 10 to 20 mg in the late afternoon tine and of topiramate successfully reducing night-
(Hermanussen & Tresguerres, 2005). As a group, time eating in patients with NES (Kucukgoncu,
the women lost weight; on average, 1.2 kg per week. Midura, & Tek, 2015). Improvements in weight
In the second trial, 16 patients received a mean end- and sleep were also described.
point memantine dose of 18.3 mg/​day. Although
they did not lose weight as a group, weight loss was Conclusions
seen in the four patients who had a remission of Research into the pharmacotherapeutic treat-
binge eating (Brennan et al., 2008). ment of EDs has lagged behind that into most
In another open-​label study, 12 patients with other serious mental disorders. Only two medica-
BED were treated with sodium oxybate, which is tions have regulatory approval for use in an ED

382 Pharmacotherapy
(fluoxetine in BN and LDX in BED). No drug has 2003). Although the therapeutic effects of antide-
been specifically developed to treat an ED. Many pressants in general on binge eating and purging
of the available pharmacotherapy studies in EDs are moderate, RCTs have shown fluoxetine may be
are plagued by limitations, such as small sample useful in the primary care setting, may be effective
size and inadequate power to detect effects, and when psychotherapy is inadequate, and may work
unclear generalizability of findings to real-​world over the long term (Romano et  al., 2002; Walsh
clinical situations. Moreover, some treatments et  al., 2000; Walsh, Fairburn, Mickley, Sysko, &
may have been prematurely dismissed as ineffec- Parides, 2004). Other available treatments that
tive or unsafe despite studies having these limi- show promise for BN and merit further study are
tations (e.g., antipsychotics and mood stabilizers the antiepileptic topiramate and the 5-​HT3 receptor
in AN). antagonist ondansetron (Faris et al., 2000; Hedges
Some preliminary conclusions about the phar- et  al., 2003; Nickel et  al., 2005). Antiandrogen
macotherapy of EDs can nonetheless be made. agents in women with hyperandrogenism, nal-
Regarding AN, neither tricyclics nor the SSRI fluox- trexone (at doses higher than used for substance
etine appear to be effective in promoting weight gain abuse), and ADHD medications also warrant fur-
in underweight patients when used adjunctively ther evaluation.
with hospital care (Claudino et al., 2006). In addi- A considerate amount of double-​blind, placebo-​
tion, one well-​designed study showed that fluox- controlled data show that LDX is efficacious for
etine does not appear to be effective in maintaining reducing binge eating in BED, as well as obsessive-​
weight gain in weight-​restored patients with AN compulsive features of binge eating and body
when used in conjunction with CBT (Walsh et al., weight. These effects may be maintained up to
2006). However, a smaller study showed fluoxetine 6  months. Similar to BN, antidepressants appear
might be helpful for weight maintenance in restrict- to have a modest beneficial effect on binge eat-
ing AN when CBT is not a required treatment com- ing in BED (Stefano et  al., 2008). They do not,
ponent (Kaye et al., 2001). Whether these findings however, appear to have clinically significant ben-
can be generalized to antidepressants from other efits on body weight, and their long-​term efficacy
classes is presently unknown. Questions also remain is unknown. Also similar to BN, a considerable
as to the efficacy of antidepressants in weight resto- amount of double-​blind, placebo-​controlled data
ration and weight maintenance in AN when used in show that topiramate is effective for binge eating
combination with other classes of compounds (e.g., in BED with obesity; these data further show that
antipsychotics; see below); in AN with comorbid topiramate is effective for weight loss in this patient
depressive, anxiety, or obsessive-​compulsive disor- population (McElroy, Hudson, et al., 2007). One
ders; and/​or in treatment-​refractory, intractable, or small open-​label study suggests topiramate’s anti–​
chronic AN. binge eating and weight loss effects in BED may
By contrast, emerging placebo-​ controlled persist for up to 1 year, but that drug discontinu-
evidence shows olanzapine may be effective for ation rates are high, in part due to adverse events
weight restoration in AN, as well as some of the (McElroy, Shapira, et  al., 2004). Orlistat, when
core and associated symptoms of AN (Attia et al., used in combination with CBT or dietary ther-
2011; Bissada, Tasca, Barber, & Bradwejn, 2008; apy, may lead to weight loss and possibly reduced
Brambilla, et al., 2007). Further controlled trials of binge eating (Golay et  al., 2005; Grilo, Masheb,
olanzapine and other antipsychotics, both first and & Salant, 2005). Other available compounds that
second generation, for weight restoration in AN show promise for BED and merit further evalua-
are needed, as are RCTs of these agents for weight tion are zonisamide, naltrexone (especially at higher
maintenance. An adequately sized randomized, doses than used for substance abuse and in com-
placebo-​controlled trial comparing olanzapine with bination with antidepressants), stimulants other
another second-​generation antipsychotic would be than LDX, and glutamate-​modulating agents (e.g.,
particularly informative. Other compounds that memantine).
hold at least some promise for AN and need further Pharmacotherapy research into NES has been
study include cyproheptadine, dronabinol, relamo- extremely limited. One small controlled study
relin, rhGH, prokinetics, zinc, and d-​cycloserine. has provided support for sertraline in NES, while
In contrast to AN, substantial evidence indi- another with escitalopram was negative. Open-​label
cates SSRIs and antidepressants from several other data suggests topiramate may be helpful for this
classes are efficacious for BN (Bacaltchuk & Hay, condition and merits further study.

McElroy, Guerdjikova, Mori, Keck Jr. 383

Future Directions sc injection, need to be evaluated in obese patients
Further pharmacotherapy research into AN, with BED or BN (Bray, Frühbeck, Ryan, & Wilding,
BN, BED, and other EDs is greatly needed at many 2016). Randomized, placebo-​ controlled trials of
levels. Further trials need to be conducted for olan- topiramate are probably warranted for NES.
zapine and zinc in AN, ondansetron in BN, and Studies of new compounds are also greatly needed
zonisamide in BED to determine whether initial in the treatment of EDs. Novel drugs in develop-
positive findings can be replicated. Randomized, ment for psychotic and bipolar disorders and dis-
placebo-​controlled maintenance trials would be eases characterized by cachexia might be considered
useful for olanzapine in AN and for topiramate in as potential candidates for study in the treatment
both BN and BED. Adequately sized RCTs of the of AN. Because hypercortisolemia usually accom-
ghrelin agonist relamorelin and the hormone oxy- panies AN, antiglucocorticoid agents might also
tocin in patients with AN would be informative. represent potential therapeutic candidates (Parsons
Also, RCTs in which a novel medication strategy & Sapse, 1985). Novel compounds that might hold
(e.g., maintenance olanzapine treatment of AN or promise for both BN and BED include some of
maintenance topiramate treatment of BN or BED) those in development for ADHD, mood disorders,
is combined with a specific psychological treatment addiction, and obesity. Thus, the norepinephrine-​
would be useful. dopamine reuptake inhibitor dasotraline is currently
There have been extremely few RCTs devoted being evaluated in patients in ADHD and those
to ED patients who have had partial or inade- with BED (Childress & Tran, 2016; Koblan et al.,
quate responses to pharmacotherapy or who have 2015). Examples of such agents being developed
treatment-​ resistant illness. As has been done in for depression include serotonin-​ norepinephrine-​
other major mental disorders, trials are needed that dopamine (triple) reuptake inhibitors, serotonin-​
explore strategies where medications are optimized, melatonin agents, and corticotropin-​ releasing
switched, augmented, or combined. For example, hormone antagonists (Connolly & Thase, 2012).
antipsychotics in combination with other agents, Examples of such compounds being developed
such as antidepressants, zinc, rhGH, or relamor- for obesity include triple reuptake inhibitors (e.g.,
elin, should be explored in randomized, placebo-​ tesofensine), GLP-​1 receptor agonists beyond lira-
controlled trials in treatment-​resistant or chronic glutide, lipase inhibitors beyond orlistat (cetilistat),
AN. Similarly, studies of topiramate in combination melanocortin-​4 receptor agonists (setmelanotide),
with antidepressants or stimulants would be impor- neuropeptide Y5 receptor inhibitors, leptin recep-
tant in patients with treatment-​resistant or chronic tor agonists (metreleptin), cannabinoid receptor
forms of BN and BED (Brambilla et al., 2009). In antagonists beyond rimonabant; and the combi-
addition, studies are needed in ED patients who nation of bupropion with zonisamide (Kakkar &
have clinically important comorbidities, such as Dahiya, 2015; Kühnen et  al., 2016). Drugs with
major depressive disorder, bipolar disorder, anxiety potential for both depression and obesity (e.g., tri-
disorders, substance use disorders, borderline per- ple reuptake inhibitors) might be especially promis-
sonality disorder, and diabetes (Woodside & Staab, ing candidates for BED. Some of the many drugs in
2006). For BED, further trials in patients with development for epilepsy might also be considered
co-​occurring obesity and obesity-​ related medical as potential therapeutic agents for EDs, especially if
conditions (such as diabetes and hypertension) are they are associated with changes in appetite or body
needed. weight (French, Schachter, Sirven, & Porter, 2015).
Available drugs that may hold promise for BN Pharmacotherapy research for EDs will need to
and BED and merit evaluation in RCTs include be informed by advances in both clinical trial design
stimulants beyond LDX; nonstimulant medications and molecular genetics. Regarding the former, there
for ADHD; the SNRIs desvenlafaxine, duloxetine, is presently a lack of consensus about what consti-
milnacipran, and venlafaxine; the novel antidepres- tutes ideal clinical trial design in ED pharmaco-
sants vilozidone and vortioxetine; 5-​HT3 receptor therapy research. This includes lack of agreement
antagonists; and memantine and other glutamate-​ on assessment instruments; definitions of primary
modulating agents. Also, RCTs of atomoxetine and outcome and of response, remission, recovery, and
zonisamide in BN would be informative. Recently relapse; stages of illness to be studied; metrics for
approved weight-​loss drugs, including lorcaserin, reporting outcome; and how best to manage low
naltrexone-​bupropion combination, phentermine-​ completion rates (Brownley et  al., 2007; Bulik
topiramate combination, and liraglutide 3 mg daily et al., 2007; Halmi et al., 2005; Shapiro et al., 2007;

384 Pharmacotherapy
Walsh et al., 2006). Regarding the latter, intensive Alger, S. A., Schwalberg, M. D., Bigaouette, J. M., Michalek,
research is needed to identify genes and endophe- A. V., & Howard, L. J. (1991). Effect of a tricyclic antide-
pressant and opiate antagonist on binge eating behavior in
notypes that will predict response to treatment normo-​ weight bulimic and obese, binge-​ eating subjects.
and facilitate novel drug discovery (Bulik et  al., American Journal of Clinical Nutrition, 53, 865–​871.
2007; Ramoz, Versini, & Gorwood, 2007; Yilmaz, Alho, H., Lahti, T., Appelberg, B., et al. (2013). Opioid antago-
Hardaway, & Bulik, 2015). nist naloxone nasal spray treatment for patients with binge eat-
Finally, for pharmacotherapy research in EDs to ing disorder: a randomized controlled trial. Poster # NR4-​29.
Presented at American Psychiatric Association 166th Annual
truly advance, it will need to be made a national Meeting, May 18–​22, San Francisco, CA.
priority. Such an advance will require collabora- American Psychiatric Association. (1980). Diagnostic and sta-
tions among academia, the pharmaceutical indus- tistical manual of mental disorders, 3rd ed. Washington,
try, and governmental agencies. For example, one DC: American Psychiatric Association.
goal would be to form a network of sites devoted American Psychiatric Association. (1994). Diagnostic and sta-
tistical manual of mental disorders, 4th ed. Washington,
to conducting RCTs in EDs similar to the networks DC: American Psychiatric Association.
that have been successful in conducting RCTs in American Psychiatric Association. (2006). Treatment of patients
mood and psychotic disorders (Bowden et al., 2012; with eating disorders. (3rd ed.). American Psychiatric
Lieberman et al., 2005; Rush, 2011). Another goal Association. American Journal of Psychiatry, 163(Suppl
would be to foster public-​private partnership pro- 7), 4–​54.
Andries, A., Frystyk, J., Flyvbjerg, A., & Støving, R. K. (2014).
grams devoted to developing potential therapeutic Dronabinol in severe, enduring anorexia nervosa: A random-
compounds specifically targeting EDs, as has been ized controlled trial. International Journal of Eating Disorders,
done by the National Institute of Mental Health 47, 18–​23.
for mood, anxiety, and psychotic disorders (Brady, Andries, A., Frystyk, J., Flyvbjerg, A., & Støving, R. K. (2015).
Winsky, Goodman, Oliveri, & Stover, 2009. Changes in IGF-​I, urinary free cortisol and adipokines dur-
ing dronabinol therapy in anorexia nervosa: Results from a
In sum, pharmacotherapy has an important randomised, controlled trial. Growth Hormone IGF Research,
role in the management of EDs, especially in 25, 247–​252.
patients who refuse or are unresponsive to psycho- Andries, A., Gram, B., & Støving, R. K. (2015). Effect of
therapy, patients with comorbid mental or medi- dronabinol therapy on physical activity in anorexia ner-
cal disorders, and those with chronic or intractable vosa:  a randomised, controlled trial. Eating and Weight
Disorders, 20, 13–​21.
EDs. However, the available pharmacotherapeutic Appolinario, J. C., Bacaltchuk, J., Sichieri, R., Claudino, A. M.,
armamentarium and its supporting evidence base Godoy-​Matos, A., Morgan, C.,  . . .  Coutinho, W. (2003).
for EDs is still far from adequate, and further study A randomized, double-​ blind, placebo-​ controlled study
is needed to clarify which specific agents might be of sibutramine in the treatment of binge-​eating disorder.
most useful for which patient subgroups. Novel Archives of General Psychiatry, 60, 1109–​1116.
Aragona, M. (2007). Tolerability and efficacy of aripirazole in
medical treatments for EDs are greatly needed. a case of psychotic anorexia nervosa comorbid with epi-
In particular, rational drug discovery devoted to lepsy and chronic renal failure. Eating Weight Disorders, 12,
EDs needs to occur. In the meantime, current e54–​e57.
and future medications with psychotropic ben- Arnold, L. M., McElroy, S. L., Hudson, J. E., Welge, J. A.,
efits and/​or effects on appetite and weight might Bennett, A. J., & Keck, P. E. (2002). A placebo-​controlled,
randomized trial of fluoxetine in the treatment of binge-​
be considered as potential therapeutic agents for eating disorder. Journal of Clinical Psychiatry, 63, 1028–​1033.
these conditions. Attia, E., Haiman, C., Walsh, B. T., & Flater, S. R. (1998). Does
fluoxetine augment the inpatient treatment of anorexia ner-
vosa? American Journal of Psychiatry, 155, 548–​551.
References
Attia, E., Kaplan, A. S., Walsh, B. T., Gershkovich, M., Yilmaz,
Agras, W. S., Dorian, B., Kirkley, B. G., Arnow, B., & Bachman,
Z., Musante, D.,  . . .  Wang, Y. (2011). Olanzapine versus
J. (1987). Imipramine in the treatment of bulimia: A double-​
placebo for out-​patients with anorexia nervosa. Psychological
blind controlled study. International Journal of Eating
Medicine, 41, 2177–​2182.
Disorders, 6, 29–​38.
Bacaltchuk, J., & Hay, P. (2003). Antidepressants versus pla-
Agras, W. S., Rossiter, E. M., Arnow, B., Schneider, J. A.,
cebo for people with bulimia nervosa. Cochrane Database of
Telch, C. F., Raeburn, S. D.,   .  .  .  Loran, L. M. (1992).
Systematic Reviews, Issue 4, CD003391.
Pharmacologic and cognitive-​behavioral treatment for buli-
Bacaltchuk, J., Hay, P., & Trefiglio, R. (2001). Antidepressants
mia nervosa: A controlled comparison. American Journal of
versus psychological treatments and their combination for
Psychiatry, 149, 82–​87.
bulimia nervosa. Cochrane Database of Systematic Reviews,
Aigner, M., Treasure, J., Kaye, W., Kasper, S., WFSBP Task Force
Issue 4, CD003385.
on Eating Disorders. (2011). World Federation of Societies
Barbarich, N. C., McConaha, C. W., Halmi, K. A., Gendall, K.,
of Biological Psychiatry (WFSBP) guidelines for the phar-
Sunday, S. R., Gaskill, J.,  . . .  Kaye, W. H. (2004). Use of
macological treatment of eating disorders. World Journal
nutritional supplements to increase the efficacy of fluoxetine
Biological Psychiatry, 12, 400–​443.

McElroy, Guerdjikova, Mori, Keck Jr. 385

 in the treatment of anorexia nervosa. International Journal of and pharmacological treatment. International Clinical
Eating Disorders, 35, 10–​15. Psychopharmacology, 24, 312–​317.
Barbee, J. G. (2003). Topiramate in the treatment of severe buli- Bray, G. A., Frühbeck, G., Ryan, D. H., & Wilding, J. P. (2016).
mia nervosa with comorbid mood disorders:  A  case series. Management of obesity. Lancet, 387, 1947–​1956.
International Journal of Eating Disorders, 33, 456–​472. Brennan, B. P., Roberts, J. I., Fogarty, K. V., Reynolds, K. A.,
Barcai, A. (1977). Lithium in adult anorexia nervosa: A pilot report Jonas, J. M., & Hudson, J. I. (2008). Memantine in the treat-
on two patients. Acta Psychiatrica Scandinavica, 55, 97–​101. ment of binge eating disorder:  An open-​label, prospective
Beaumont, P. J., Russell, J. D., Touyz, S. W., Buckley, C., trial. International Journal of Eating Disorders, 41, 520–​526.
Lowinger, K., Talbot, P., & Johnson, G. F. (1997). Intensive Brewerton, T. D. (2012). Antipsychotic agents in the treatment
nutritional counselling in bulimia nervosa:  A  role for sup- of anorexia nervosa:  Neuropsychopharmacologic ratio-
plementation with fluoxetine? Australian and New Zealand nale and evidence from controlled trials. Current Psychiatry
Journal of Psychiatry, 31, 514–​524. Reports, 14, 398–​405.
Biederman, J., Herzog, D. B., Rivinus, T. M., Harper, G. Brewerton, T. D., & Shannon, M. (1992). Possible clozap-
P., Ferber, R. A., Rosenbaum, J. F.,  . . .  Schildkraut, J. J. ine exacerbation of bulimia nervosa. American Journal of
(1985). Amitriptyline in the treatment of anorexia ner- Psychiatry, 149, 1408–​1409.
vosa:  A  double-​blind, placebo-​controlled study. Journal of Broft, A. I., Spanos, A., Corwin, R. L., Mayer, L., Steinglass, J.,
Clinical Psychopharmacology, 5, 10–​16. Devlin, M. J,  . . .  Walsh, B. T. (2007). Baclofen for binge
Birmingham, C. L., Goldner, E. M., & Bakan, R. (1994). eating:  An open-​label trial. International Journal of Eating
Controlled trial of zinc supplementation in anorexia nervosa. Disorders, 40, 687–​691.
International Journal of Eating Disorders, 15, 251–​255. Brownley, K. A., Berkman, N. D., Sedway, J. A., Lohr, K., &
Birmingham, C. L., & Gritzner, S. (2006). How does zinc sup- Bulik, C. M. (2007). Binge eating disorder treatment: A sys-
plementation benefit anorexia nervosa? Eating and Weight tematic review of randomized controlled trials. International
Disorders, 11, 109–​111. Journal of Eating Disorders, 40, 337–​348.
Bissada, H., Tasca, G. A., Barber, A. M., & Bradwejn, J. (2008). Brownley, K. A., Von Holle, A., Hamer, R. M., La Via, M., &
Olanzapine in the treatment of low body weight and obses- Bulik, C. M. (2013). A double-​blind, randomized pilot trial
sive thinking in women with anorexia nervosa: A random- of chromium picolinate for binge eating disorder: Results of
ized, double-​blind, placebo-​controlled trial. American Journal the Binge Eating and Chromium (BEACh) study. Journal of
of Psychiatry, 165, 1281–​1288. Psychosomatic Research, 75, 36–​42.
Bloch, M., Ish-​Shalom, S., Greenman, Y., Klein, E., & Latzer, Bruno, A., Riganello, D., & Marino, A. (2009). Treatment with
Y. (2012). Dehydroepiandrosterone treatment effects on aripiprazole and topiramate in an obese subject with bord-
weight, bone density, bone metabolism and mood in women erline personality disorder, obsessive-​compulsive symptoms
suffering from anorexia nervosa-​ a pilot study. Psychiatry and bulimia nervosa: A case report. Cases Journal, 2, 7288.
Research, 200, 544–​549. Bulik, C. M., Hebebrand, J., Keski-​Rahkonen, A., Klump, K.
Blouin, A. G., Blouin, J. H., Perez, E. L., Bushnik, T., Zuro, C., & L., Reichborn-​Kjeennerud, T., Mazzeo, S. E.,   .  .  .  Wade,
Mulder, E. (1988). Treatment of bulimia with fenfluramine T. D. (2007). Genetic epidemiology, endophenotypes, and
and desipramine. Journal of Clinical Psychopharmacology, 8, eating disorder classification. International Journal of Eating
261–​269. Disorders, 40, S52–​S60.
Bowden, C. L., Perlis, R. H., Thase, M. E., Ketter, T. A., Ostacher, Carruba, M. O., Cuzzolaro, M., Riva, L., Bosello, O., Liberti, S.,
M. M., Calabrese, J. R.,  . . .  Sachs, G. S. (2012). Aims and Castra, R.,  . . .  Nisoli, E. (2001). Efficacy and tolerability of
results of the NIMH systematic treatment enhancement pro- moclobemide in bulimia nervosa: A placebo-​controlled trial.
gram for bipolar disorder (STEP-​BD). CNS Neuroscience and International Journal of Psychopharmacology, 16, 27–​32.
Therapeutics, 18, 243–​249. Casper, R. C., Schlemmer, R. F., & Javaid, J. I. (1987). A
Brady, L. S., Winsky, L., Goodman, W., Oliveri, M. E., & Stover, placebo-​controlled crossover study of oral clonidine in acute
E. (2009). NIMH initiatives to facilitate collaborations anorexia nervosa. Psychiatry Research, 20, 249–​260.
among industry, academia, and government for the discovery Cassano, G. B., Miniati, M., Pini, S., Rotondo, A., Banti, S.,
and clinical testing of novel models and drugs for psychiatric Borri, C.,  . . .  Mauri, M. (2003). Six-​month open trial of
disorders. Neuropsychopharmacology, 34, 229–​243. haloperidol as an adjunctive treatment for anorexia ner-
Brambilla F., Draisci, A., Peirone, A., & Brunetta, M. (1995a). vosa:  A  preliminary report. International Journal of Eating
Combined cognitive-​ behavioral, psychopharmacologi- Disorders, 33, 172–​177.
cal and nutritional therapy in eating disorders. 1. Anorexia Childress, A., & Tran, C. (2016). Current investigational drugs
nervosa—​restricted type. Neuropsychobiology, 32, 59–​63. for the treatment of attention-​deficit/​hyperactivity disorder.
Brambilla F., Draisci, A., Peirone, A., & Brunetta, M. (1995b). Expert Opinion on Investigational Drugs, 25, 463–​474.
Combined cognitive-​behavioral, psychopharmacological and Chung, A. M., & Reed, M. D. (2004). Intentional topiramate
nutritional therapy in eating disorders. 2. Anorexia nervosa—​ ingestion in an adolescent female. Annals of Pharmacotherapy,
binge-​eating/​purging type. Neuropsychobiology, 32, 64–​67. 38, 1439–​1442.
Brambilla, F., Garcia, C. S., Fassino, S., Daga, G. A., Favara, A., Claudino, A. M., de Oliveira, I. R., Appolinario, J. C., Cordás,
Santonastaso, P.,   .  .  .  Monteleone, P. (2007). Olanzapine T. A., Duchesne, M., Sichiere, R.,  . . .  Bacaltchuk, J. (2007).
therapy in anorexia nervosa:  Psychobiological effects. Double-​blind, randomized, placebo-​controlled trial of topi-
International Clinical Psychopharmacology, 22, 197–​204. ramate plus cognitive-​behavior therapy in binge-​eating disor-
Brambilla, F., Samek, L., Company, M., Lovo, F., Cioni, L., & der. Journal of Clinical Psychiatry, 68, 1324–​1332.
Mellado, C. (2009). Multivariate therapeutic approach to Claudino, A. M., Hay, P., Lima, M. S., Bacaltchuk, J., Schmidt,
binge-​eating disorder: Combined nutritional, psychological U., & Treasure, J. (2006). Antidepressants for anorexia

386 Pharmacotherapy
 nervosa. Cochrane Database Systematic Reviews, Issue 1, Dolberg, O. T., Barkai, G., Gross, Y., & Schreiber, S. (2005).
CD004365. Differential effects of topiramate in patients with traumatic
Cochrane, C., & Malcolm, R. (2002). Case report of abuse of brain injury and obesity:  A  case series. Psychopharmacology
orlistat. Eating Behavior, 3, 167–​169. (Berl.), 179, 838–​845.
Colman, E. (2005). Anorectics on trial: A half century of federal Dold, M., Aigner, M., Klabunde, M., Treasure, J., & Kasper, S.
regulation of prescription appetite suppressants. Annals of (2015). Second-​generation antipsychotic drugs in anorexia
Internal Medicine, 143, 380–​385. nervosa:  A  meta-​analysis of randomized controlled trials.
Colom, F., Vieta, E., Benabarre, A., Martinez-​Arán, A., Reinares, Psychotherapy and Psychosomatics, 84, 110–​116.
M., Corbella, B.,  . . .  Gastó, C. (2001). Topiramate abuse in Drewnowski, A., Krahn, D. D., Demitrack, M. A., Nairn, K., &
a bipolar patient with an eating disorder. Journal of Clinical Gosnell, B. A. (1995). Naloxone, an opiate blocker, reduces
Psychiatry, 62, 475–​476. the consumption of sweet high-​fat foods in obese and lean
Connolly, K. R., & Thase, M. E. (2012). Emerging drugs for female binge eaters. American Journal of Clinical Nutrition,
major depressive disorder. Expert Opinion of Emerging Drugs, 61, 1206–​1212.
17, 105–​126. Drimmer, E. J. (2003). Stimulant treatment of bulimia ner-
Corwin, R. L., Boan, J., Peters, K. F., & Ulbrecht, J. S. (2012). vosa with and without attention-​deficit disorder: Three case
Baclofen reduces binge eating in a double-​blind, placebo-​ reports. Nutrition, 19, 76–​77.
controlled, crossover study. Behavioral Pharmacology, 23, Dukarm, C. P. (2005). Bulimia nervosa and attention deficit
616–​625. hyperactivity disorder: A possible role for stimulant medica-
Court, A., Mulder, C., Kerr, M., Yuen, H. P., Boasman, M., tion. Journal of Women’s Health, 14, 345–​350.
Goldstone, S.,  . . .  Berger, G. (2010). Investigating the effec- Dunican, K. C., & DelDotto, D. (2007). The role of olan-
tiveness, safety and tolerability of quetiapine in the treatment zapine in the treatment of anorexia nervosa. Annals of
of anorexia nervosa in young people: A pilot study. Journal of Pharmacotherapy, 41, 111–​115.
Psychiatric Research, 44, 1027–​1034. Duvvuri, V., Cromley, T., Klabunde, M., Boutelle, K., & Kaye,
Crisp, A. H., Lacey, J. H., & Crutchfield, M. (1987). W. H. (2012). Differential weight restoration on olanzapine
Clomipramine and “drive” in people with anorexia ner- versus fluoxetine in identical twins with anorexia nervosa.
vosa:  An inpatient study. British Journal of Psychiatry, 150, International Journal of Eating Disorders, 45, 294–​297.
355–​358. El-​Giamal, N., deZwaan, M., Bailer, U., Lennkh, C., Schussler,
Crockford, D. N., Fisher, G., & Barker, P. (1997). Risperidone, P., Strnad, A.,   .  .  .  Kasper, S. (2000). Reboxetine in the
weight gain, and bulimia nervosa. Canadian Journal of treatment of bulimia nervosa:  A  report of seven cases.
Psychiatry, 42, 326–​327. International Clinical Psychopharmacology, 15, 351–​356.
De Beaurepaire, R., Joussaume, B., Rapp, A., & Jaury, P. El-​Giamal, N., deZwaan, M., Bailer, U., Strnad, A., Schüssler,
(2015). Treatment of binge eating disorder with high-​dose P., & Kasper, S. (2003). Milnacipran in the treat-
baclofen: A case series. Journal of Clinical Psychopharmacology, ment of bulimia nervosa:  A  report of 16 cases. European
35, 357–​359. Neuropsychopharmacology, 123, 73–​79.
De Bernardi, C., Ferraris, S., D’Innella, P., Do, F., & Torre, E. Fahy, T. A., Eisler, I., & Russell, G. F. (1993). A placebo-​
(2005). Topiramate for binge eating disorder. Progress in controlled trial of d-​fenfluramine in bulimia nervosa. British
Neuro-​ Psychopharmacology and Biological Psychiatry, 29, Journal of Psychiatry, 162, 597–​603.
339–​341. Faris, P. L., Kim, S. W., Meller, W. H., Goodale, R. L., Oakman,
Deb, K. S., Gupta, R., & Varshney, M. (2014). Orlistat abuse S. A., Hofbauer, R. D.,  . . .  Hartman, B. K. (2000). Effect of
in a case of bulimia nervosa:  The changing Indian society. decreasing afferent vagal activity with ondansetron on symp-
General Hospital Psychiatry, 36, e3–​e4. toms of bulimia nervosa: A randomized, double-​blind trial.
Devlin, J. J., Goldfein, J. A., Carino, J. S., & Wolk, S. L. (2000). Lancet, 35, 792–​797.
Open treatment of overweight binge eaters with phentermine Fassino, S., Daga, G. A., Boggio, S., Garzaro, L., & Piero, A.
and fluoxetine as an adjunct to cognitive-​behavioral therapy. (2004). Use of reboxetine in bulimia nervosa: A pilot study.
International Journal of Eating Disorders, 28, 325–​332. Journal of Psychopharmacology, 18, 423–​428.
Devlin, M. J., Goldfein, J. A., Petkova, E., Jiang, H., Raizman, Fazeli, P. K., Calder, G. L., Miller, K. K., Misra, M., Lawson, E.
P. S., Wolk, S.,  . . .  Walsh, B. T. (2005). Cognitive behav- A., Meenaghan, E.,  . . .  Klibanski, A. (2012). Psychotropic
ioral therapy and fluoxetine as adjuncts to group behav- medication use in anorexia nervosa between 1997 and 2009.
ioral therapy for binge eating disorder. Obesity Research, 13, International Journal of Eating Disorders, 45, 970–​976.
1077–​1088. Fazeli, P., Lawson, E. A., Faje, A. T., Eddy, K. T., Gaal, I.,
Devlin, M. J., Goldfein, J. A., Petkova, E., Liu, L., & Walsh, B. DeSanti, R.,  . . .  Klibanski, A. (2016). Short-​term treatment
T. (2007). Cognitive behavioral therapy and fluoxetine for with a ghrelin agonist significantly improves gastric emptying in
binge eating disorder:  Two-​year follow-​up. Obesity (Silver anorexia nervosa. Presented at The Endocrine Society Annual
Spring), 15, 1702–​1709. Meeting; April 1–​4, 2016; Boston, MA.
Devlin, M. J., Kissileff, H. R., Zimmerli, E. J., Samuels, F., Fazeli, P., Lawson, E. A., Prabhakaran, R., Miller, K. K., Donoho,
Chen, B. E., Brown, A. J.,  . . .  Walsh, B. T. (2012). Gastric D. A., Clemmons, D. R.,  . . .  Klibanski, A. (2010). Effects
emptying and symptoms of bulimia nervosa: Effect of a pro- of recombinant human growth hormone in anorexia ner-
kinetic agent. Physiology and Behavior, 106, 238–​242. vosa:  A  randomized, placebo-​ controlled study. Journal
de Vos, J., Houtzager, L., Katsaragaki, G., van de Berg, E., Clinical Endrocrinology Metabolism, 95, 4889–​4897.
Cuijpers, P., & Dekker J. (2014). Meta analysis on the effi- Felstrom, A., & Blackshaw, S. (2002). Topiramate for buli-
cacy of pharmacotherapy versus placebo on anorexia nervosa. mia nervosa with bipolar II disorders. American Journal of
Journal of Eating Disorders, 2, 27. Psychiatry, 159, 1246–​1247.

McElroy, Guerdjikova, Mori, Keck Jr. 387

Fernández-​Aranda, F., Amor, A., Jiménez-​Murcia, S., Giménez-​ Goudie, A. J., Cooper, G. D., Cole, J. C., & Sumnall, H. R.
Martínez, L., Turón-​Gil, V., & Vallejo-​Ruiloba, J. (2001). (2007). Cyproheptadine resembles clozapine in vivo follow-
Bulimia nervosa and misuse of orlistat:  Two case reports. ing both acute and chronic administration in rats. Journal of
International Journal of Eating Disorders, 30, 458–​461. Psychopharmacology, 21, 179–​190.
Fichter, M. M., Krüger, R., Rief, W., Holland, R., & Döhne, Greenway, F. L., Dahms, W. T., & Bray, G. A. (1977). Phenytoin
J. (1996). Fluvoxamine in prevention of relapse in buli- as a treatment of obesity associated with compulsive eating.
mia nervosa:  Effects on eating-​ specific psychopathology. Current Therapeutic Research, 21, 338–​342.
Pharmacopsychiatry, 30, 85–​92. Grigoriadis, S., Kaplan, A., Carter, J., & Woodside, B. (2001).
Fichter, M. M., Leibl, K., Rief, W., Brunner, E., Schmidt-​ What treatments patients seek after inpatient care:  A  fol-
Auberger, S., & Engel, R. R. (1991). Fluoxetine versus lowup of 24 patients with anorexia nervosa. Eating and
placebo: A double-​blind study with bulimic inpatients under- Weight Disorders, 6, 115–​120.
going intensive psychotherapy. Pharmacopsychiatry, 24, 1–​7. Grilo, C. M., Masheb, R. M., & Salant, S. L. (2005). Cognitive
Fisman, S., Steele, M., Short, J., Byrne, T., & Lavallee, C. (1996). behavioral therapy guided self-​ help and orlistat for the
Case study: Anorexia nervosa and autistic disorder in an ado- treatment of binge eating disorder:  A  randomized, double-​
lescent girl. Journal of the American Academy of Child and blind, placebo-​ controlled trial. Biological Psychiatry, 57,
Adolescent Psychiatry, 35, 937–​940. 1193–​1201.
Fluoxetine Bulimia Nervosa Collaborative Study Group Grilo, C. M., Masheb, R. M., & Wilson, G. T. (2005). Efficacy
(FBNCSG). (1992). Fluoxetine in the treatment of bulimia of cognitive behavioral therapy and fluoxetine for the treat-
nervosa:  A  multicenter, placebo-​controlled, double-​blind ment of binge eating disorder: A randomized double-​blind
trial. Archives of General Psychiatry, 49, 139–​147. placebo-​controlled comparison. Biological Psychiatry, 57,
Fountoulakis, K. N., Iacovides, A., Siamouli, M., Koumaris, V., 301–​309.
& Kaprinis, G. S. (2006). Successful treatment of anorexia Grilo, C. M., & White, M. A. (2013). Orlistat with behavioral
with a combination of high-​dose olanzapine, fluoxetine, and weight loss for obesity with versus without binge eating dis-
mirtazapine. International Journal of Clinical Pharmacology order: Randomized placebo-​controlled trial at a community
Therapy, 44, 452–​453. mental health center serving educationally and economi-
Frank, G. K. (2016). Aripiprazole, a partial dopamine ago- cally disadvantaged Latino/​as. Behavior Research Therapy, 51,
nist to improve adolescent anorexia nervosa-​A case series. 167–​175.
International Journal of Eating Disorders, 49, 529–​533. Gross, H., Ebert, M. H., Faden, V. B., Goldberg, S. C., Kaye, W.
French, J. A., Schachter, S. C., Sirven, J., & Porter, R. (2015). H., Caine, E. D.,  . . .  Zinberg, N. (1983). A double-​blind
The Epilepsy Foundation’s 4th biennial epilepsy pipeline trial of delta 9-​tetrahydrocannabinol in primary anorexia
update conference. Epilepsy Behavior, 46, 34–​50. nervosa. Journal of Clinical Psychopharmacology, 3, 165–​171.
Gaisor, M., Hudson, J., Quintero, J., Ferreira-​Cornwell, M. C., Gross, H., Ebert, M. H., Faden, V. B., Goldberg, S. C., Kaye, W.
Radewonak, J., & McElroy, S. L. (2017). A phase 3, multi- H., Caine, E. D.,  . . .  Kaye, W. H. (1981). A double-​blind
center, open-​label, 12-​month extension safety and tolerability controlled study of lithium carbonate in primary anorexia
trial of lisdexamfetamine dimesylate in adults with binge eating nervosa. Journal of Clinical Psychopharmacology, 1, 376–​381.
disorder. Journal of Clinical Psychopharmacology, 50, 315–322. Guerdjikova, A. I., Blom, T. J., Martens, B. E., Keck, P. E., Jr.,
Gardin, J. M., Schumacher, D., Constantine, G., Davis, K. D., & McElroy, S. L. (2013). Zonisamide in the treatment of
Leung, C., & Reid, C. L. (2000). Valvular abnormalities and bulimia nervosa:  An open-​ label, pilot, prospective study.
cardiovascular status following exposure to dexfenfluramine International Journal of Eating Disorders, 46, 747–​750.
or phentermine/​fenfluramine. JAMA, 283, 1703–​1709. Guerdjikova, A. I., Blom, T. J., Mori, N., & McElroy, S. L.
Garner, D. M. (1991). Eating disorder inventory-​2:  Professional (2013). N-​ acetylcysteine in bulimia nervosa:  Open-​ label
manual. Odessa, FL: Psychological Assessment Resources. trial. Eating Behavior, 14, 87–​89.
Gebhardt, S., Haberhausen, M., Krieg, J. C., Remschmidt, H., Guerdjikova, A. I., Fitch, A., & McElroy, S. L. (2015). Successful
Heinzel-​Gutenbrunner, M., Hebebrand, J.,   .  .  .  Theisen, treatment of binge eating disorder with combination
F. M. (2007). Clozapine/​olanzapine-​induced recurrence or phentermine/​ topiramate extended release. Primary Care
deterioration of binge eating-​related eating disorders. Journal Companion for CNS Disorders, 17.
of Neural Transmission, 114, 1091–​1095. Guerdjikova, A. I., Kotwal, R., & McElroy, S. L. (2005).
Golay, A., Laurent-​ Jaccard, A., Habicht, F., Gachoud, J. P., Response of recurrent binge eating and weight gain to topi-
Chabloz, M., Kammer, A.,  . . .  Schutz, Y. (2005). Effect of ramate in patients with binge eating disorder after bariatric
orlistat in obese patients with binge eating disorder. Obesity surgery. Obesity Surgery, 15, 273–​277.
Research, 13, 1701–​1708. Guerdjikova, A. I., & McElroy, S. L. (2013). Adjunctive methyl-
Goldberg, S. C., Halmi, K. A., Eckert, E. D., Casper, R. C., & phenidate in the treatment of bulimia nervosa co-​occurring
Davis, J. M. (1979). Cyproheptadine in anorexia nervosa. with bipolar disorder and substance dependence. Innovative
British Journal of Psychiatry, 134, 67–​70. Clinical Neuroscience, 10, 30–​33.
Goldbloom, D. S., Olmsted, M., Davis, R., Clewes, J., Heinmaa, Guerdjikova, A. I., McElroy, S. L., Kotwal, R., Welge, J. A.,
M., Rockert, W.,  . . .  Shaw, B. (1997). A randomized con- Nelson, E., Lake, K.,  . . .  Hudson, J. I. (2008). High-​dose
trolled trial of fluoxetine and cognitive behavioral therapy for escitalopram in the treatment of binge-​eating disorder with
bulimia nervosa:  Short-​term outcome. Behavioral Research obesity:  A  placebo-​controlled monotherapy trial. Human
and Therapy, 35, 803–​811. Psychopharmacology, 23, 1–​11.
Gormally, J., Black, S., Daston, S., & Rardin, D. (1982). The Guerdjikova, A. I., McElroy, S. L., Welge, J. A., Nelson, E.,
assessment of binge eating severity among obese persons. Keck, P. E., & Hudson, J. I. (2009). Lamotrigine in the treat-
Addictive Behavior, 7, 47–​55. ment of binge-​eating disorder with obesity: A randomized,

388 Pharmacotherapy
 placebo-​controlled monotherapy trial. International Clinical binge and purge measures. Journal of Clinical Psychiatry, 64,
Psychopharmacology, 24, 150–​158. 1335–​1341.
Guerdjikova, A. I., McElroy, S. L., Winstanley, E. L., Nelson, Horne, R. L., Ferguson, J. M., Pope, H. G., Jr, Hudson, J. I.,
E. B., Mori, N., & McCoy, J. (2012) Duloxetine in the Lineberry, C. G., Ascher, J.,  . . .  Cato, A. (1988). Treatment
treatment of binge eating disorder with depressive disor- of bulimia with bupropion:  A  multicenter controlled trial.
ders:  A  placebo-​controlled trial. International Journal of Journal of Clinical Psychiatry, 49, 262–​266.
Eating Disorders, 45, 281–​289. Hrdlicka, M., Beranova, I., Zamecnikova, R., & Urbanek, T.
Guille, C., & Sachs, G. (2002). Clinical outcome of adjunc- (2008). Mirtazapine in the treatment of adolescent anorexia
tive topiramate treatment in a sample of refractory nervosa. Case-​control study. European Child and Adolescent
bipolar patients with comorbid conditions. Progress in Psychiatry, 17, 187–​189.
Neuro -​ Psychopharmacology and Biological Psychiatry, 26, Hsu, L. K., Clement, L., Sandhouse, R., & Ju, E. S. (1991).
1035–​1039. Treatment of bulimia nervosa with lithium carbonate: A con-
Hagman, J., Gralla, J., Sigel, E., Ellert, S., Dodge, M., Gardner, trolled study. Journal of Mental and Nervous Disease, 179,
R.,  . . .  Wamboldt, M. Z. (2011). A double-​blind, placebo-​ 351–​355.
controlled study of risperidone for the treatment of ado- Hudson, J. I., McElroy, S. L., Ferreira-​ Cornwell, M. C.,
lescents and young adults with anorexia nervosa:  A  pilot Radewonuk, J., & Gaisor, M. (2017). Efficacy of lisdex-
study. Journal of the American Academy of Child Adolescent amfetamine in adults with moderate to severe binge eating
Psychiatry, 50, 915–​924. disorder: A randomized clinical trial. JAMA Psychiatry, 74,
Halmi, K. A., Agras, W. S., Crow, S., Mitchell, J., Wilson, G. 903–910.
T., Bryson, S. W.,  . . .  Kraemer, H. C. (2005). Predictors Hudson, J. I., McElroy, S. L., Raymond, N. C., Crow, S., Keck,
of treatment acceptance and completion in anorexia ner- P. E., Jr., Carter, W. P.,  . . .  Jonas, J. M. (1998). Fluvoxamine
vosa:  Implications for future study designs. Archives of in the treatment of binge-​eating disorder:  A  multicenter
General Psychiatry, 62, 776–​781. placebo-​controlled, double-​blind trial. American Journal of
Halmi, K. A., Eckert, E., LaDu, T. J., & Cohen, J. (1986). Psychiatry, 155, 1756–​1762.
Anorexia nervosa: Treatment efficacy of cyproheptadine and Hudson, J. I., Pope, H. G., Jr., Jonas, J. M., & Yurgelun-​
amitriptyline. Archives of General Psychiatry, 43, 177–​181. Todd, D. (1985). Treatment of anorexia nervosa with
Hay, P., Chinn, D., Forbes, D., Madden, S., Newton, R., antidepressants. Journal of Clinical Psychopharmacology,
Sugenor, L.,   .  .  .  Ward, W. (2014). Royal Australian and 5, 17–​23.
New Zealand College of Psychiatrists clinical practice guide- Hudson, J. I., & Pope, H. G., Jr. (1988). The role of anticon-
lines for the treatment of eating disorders. Australian and vulsants in the treatment of bulimia. In S. L. McElroy, & H.
New Zealand Journal of Psychiatry, 48, 977–​1008. G. Pope Jr. (Eds.), Use of anticonvulsants in psychiatry (pp.
Hazen, E., & Fava, M. (2006). Successful treatment with dulox- 141–​153). Clifton, NJ: Oxford Health Care.
etine in a case of treatment refractory bulimia nervosa: A case Jaafar, N. R., Daud, T. I., Rahman, F. N., & Baharudin, A.
report. Journal of Psychopharmacology, 20, 723–​724. (2007). Mirtazapine for anorexia nervosa with depres-
Hedges, D. W., Reimherr, F. W., Hoopes, S. P., Rosenthal, N. sion. Australia and New Zealand Journal of Psychiatry, 41,
R., Kamin, R., & Capace, J. A. (2003). Treatment of buli- 768–​769.
mia nervosa with topiramate in a randomized, double-​blind, Jonas, J. M., & Gold, M. S. (1988). The use of opiate antagonists
placebo-​controlled trial, part 2: Improvement in psychiatric in treating bulimia:  A  study of low dose versus high dose
measures. Journal of Clinical Psychiatry, 64, 1449–​1454. naltrexone. Psychiatry Research, 24, 195–​199.
Hermanussen, M., & Tresguerres, J. A. (2005). A new anti-​ Kafantaris, V., Leigh, E., Hertz, S., Berest, A., Schebendach, J.,
obesity drug treatment: First clinical evidence that, antago- Sterling, W. M.,.  . . .  Malhotra, A. K. (2011). A placebo-​
nizing glutamate-​gated Ca2 + ion channels with memantine controlled pilot study of adjunctive olanzapine for adoles-
normalizes binge-​ eating disorders. Economics and Human cents with anorexia nervosa. Journal of Child and Adolescent
Biology, 3, 329–​337. Psychopharmacology, 21, 207–​212.
Herridge, P. L., & Pope, H. G., Jr. (1985). Treatment of buli- Kakkar, A. K. & Dahiya, N. (2015). Drug treatment of obe-
mia and rapid-​cycling bipolar disorder with sodium valpro- sity: Current status and future prospects. European Journal of
ate: A case report. Journal of Clinical Psychopharmacology, 5, Internal Medicine, 26, 89–​94.
229–​230. Kanerva, R., Rissanen, A, & Sarna, S. (1994). Fluoxetine in
Hill, K., Bucuvalas, J., McClain, C., Kryscio, R., Martini, R. the treatment of anxiety, depressive symptoms, and eating-​
T., Alfaro, M. P.,  . . .  Maloney, M. (2000). Pilot study of related symptoms in bulimia nervosa. Nordic Journal of
growth hormone administration during the refeeding of Psychiatry, 49, 237–​242.
malnourished anorexia nervosa patients. Journal of Child and Kaplan, A. S. (1987). Anticonvulsant treatment of eating dis-
Adolescent Psychopharmacology, 10, 3–​8. orders. In P. E. Garfinkel & D. M. Garner (Eds.), The
Hofmann, S. G., Meuret, A. E., Smits, J. A., Simon, N. M., role of drug treatments for eating disorders (pp. 96–​123).
Pollack, M. H., Eisenmenger, K.,.  . . .  Pollack, M. H. (2006). New York: Brunner/​Mazel.
Augmentation of exposure therapy with D-​cycloserine for Kaplan, A. S., Garfinkel, P. E., Darby, P. L., & Garner, D.
social anxiety disorder. Archives of General Psychiatry, 63, M. (1983). Carbamazepine in the treatment of bulimia.
298–​304. American Journal of Psychiatry, 140, 1225–​1226.
Hoopes, S. P., Reimherr, F. W., Hedges, D. W., Rosenthal, N. R., Katz, R. L., Keen, C. L., Litt, I. F., Hurley, L. S., Kellams-​
Kamin, R., Karim, R.,  . . .  Karvois, D. (2003). Treatment of Harrison, K. M., & Glader, L. J. (1987). Zinc deficiency
bulimia nervosa with topiramate in a randomized, double-​ in anorexia nervosa. Journal of Adolescent Health Care, 8,
blind, placebo-​ controlled trial, part  1:  Improvement in 400–​406.

McElroy, Guerdjikova, Mori, Keck Jr. 389

Kaye, W. H., Nagata, T., Weltzin, T. E., Hsu, L. K., Sokol, M. Laederach-​ Hofmann, K., Graf, C., Horber, F., Lippuner,
S., McConaha, C.,   .  .  .  Deep, D. (2001). Double-​blind K., Lederer, S., Michel, R.,   .  .  .  Schneider, M. (1999).
placebo-​controlled administration of fluoxetine in restrict- Imipramine and diet counseling with psychological sup-
ing and restricting-​purging-​type anorexia nervosa. Biological port in the treatment of obese binge eaters: A randomized,
Psychiatry, 49, 644–​652. placebo-​controlled double-​blind study. International Journal
Kaye, W. H., Weltzin, T. E., Hsu, L. K., & Bulik, C. M. (1991). of Eating Disorders, 26, 231–​244.
An open trial of fluoxetine in patients with anorexia nervosa. Lask, B., Fosson, A., Rolfe, U., & Thomas, S. (1993). Zinc
Journal of Clinical Psychiatry, 52, 464–​471. deficiency and childhood-​onset anorexia nervosa. Journal of
Kennedy, S. H., Goldbloom, D. S., Ralevski, E., Davis, C., Clinical Psychiatry, 54, 63–​66.
Disouza, J. D., & Lofchy, J. (1993). Is there a role for selec- Lebow, J., Chuy, J. A., Cedermark, K., Cook, K., & Sim, L. A.
tive monoamine oxidase inhibitor therapy in bulimia ner- (2015). The development or exacerbation of eating disor-
vosa? A  placebo-​controlled trial of brofaromine. Journal of der symptoms after topiramate initiation. Pediatrics, 135,
Clinical Psychopharmacology, 13, 415–​422. e1312–​e1316.
Kennedy, S. H., Piran, N., Warsh, J. J., Prendergast, P., Lebow, J., Sim, L. A., Erwin, P. J., & Murad, M. H. (2013). The
Mainprize, E., Whynot, C.,  . . .  Lofchy, J. (1988). A trial of effect of atypical antipsychotic medications in individuals
isocarboxazid in the treatment of bulimia nervosa. Journal of with anorexia nervosa: A systematic review and meta-​analysis.
Clinical Psychopharmacology, 8, 391–​396. International Journal of Eating Disorders, 46, 332–​339.
Keshen, A., & Ivanova, I. (2013). Reduction of bulimia nervosa Leitenberg, H., Rosen, J. C., Wolf, J., Vara, L. S., Detzer, M.
symptoms after psychostimulant initiation in patients with J., & Srebnik, D. (1994). Comparison of cognitive-​behavior
comorbid ADHD:  Five case reports. Eating Disorders, 21, therapy and desipramine in the treatment of bulimia ner-
360–​369. vosa. Behavior Research Therapy, 32, 37–​45.
Kim, Y. R., Eom, J. S., Yang, J. W., Kang, J., & Treasure, J. Leombruni, P., Gastaldi, F., Lavagnino, L., & Fassino, S. (2008).
(2015). The impact of oxytocin on food intake and emotion Oxcarbazepine for the treatment of binge eating disor-
recognition in patients with eating disorders: A double blind der: A case series. Advances in Therapy, 25, 718–​724.
single dose within-​subject cross-​over design. PLos One, 10, Leombruni, P., Pierò, A., Brustolin, A., Mondelli, V., Levi, M.,
e0137514. Campisi, S.,  . . .  Fassino, S. (2006). A 12 to 24 weeks pilot
Kim, Y. R., Kim, C. H., Cardi, V., Eom, J. S., Seong, Y., & study of sertraline treatment in obese women binge eaters.
Treasure, J. (2014). Intranasal oxytocin attenuates atten- Human Psychopharmacology, 21, 181–​188.
tional bias for eating and fat shape stimuli in patients with Leombruni, P., Pierò, A., Lavagnino, L., Brustolin, A., Campisi,
anorexia nervosa. Psychoneuroendocrinology, 44, 133–​142. S., & Fassino, S. (2008). A randomized, double-​ blind
Kim, Y. R., Kim, C. H., Park, J. H., Pyo, J., & Treasure, J. trial comparing sertraline and fluoxetine 6-​ month treat-
(2014). The impact of intranasal oxytocin on attention ment in obese patients with binge eating disorder. Progress
to social emotional stimuli in patients with anorexia ner- in Neuropsychopharmacology and Biological Psychiatry, 32,
vosa:  A  double blind within-​subject cross-​over experiment. 1599–​1605.
PLoS One, 9, e90721. Leyba, C. M., & Gold, D. D. (1988). The relation between
Kishi, T., Kafantaris, V., Sunday, S., Sheridan, E. M., & Correll, rapid-​cycling cyclothymia and bulimia; Case reports of two
C. U. (2012). Are antipsychotics effective for the treatment women. South Dakota Journal of Medicine, 41, 21–​22.
of anorexia nervosa? Results from a systematic review and Lieberman, J. A., Stroup, T. S., McEvoy, J. P., Swartz, M. S.,
meta-​analysis. Journal of Clinical Psychiatry, 73, e757–​e766. Rosenheck, R. A., Perkins, D. O.,  . . .  Hsiao, J. K. (2005).
Knable, M. (2001). Topiramate for bulimia nervosa in epilepsy. Effectiveness of antipsychotic drugs in patients with chronic
American Journal of Psychiatry, 158, 322–​323. schizophrenia. New England Journal of Medicine, 353,
Koblan, K. S., Hopkins, S. C., Sarma, K., Jin, F., Goldman, R., 1209–​1223.
Kollins, S. H.,  . . .  Loebel, A. (2015). Dasotraline for the Luby, E. D., Marrazzi, M. A., & Kinzie, J. (1987). Treatment
treatment of attention-​deficit/​hyperactivity disorder: A ran- of chronic anorexia nervosa with opiate blockade. Journal of
domized, double-​blind, placebo-​controlled, proof-​of-​concept Clinical Psychopharmacology, 7, 52–​53.
trial in adults. Neuropsychopharmacology, 40, 2745–​2752. Malhotra, S., King, K. H., Welge, J. A., Brunsman-​Lovins, L., &
Kolata, G. G. (1980). NIH shaken by death of research volun- McElroy, S. L. (2002). Venlafaxine treatment of binge-​eating
teer. Science, 209, 475–​476, 478–​479. disorder associated with obesity:  A  series of 35 patients.
Kotwal, R., Guerdjikova, A., McElroy, S. L., & Keck, P. E., Jr. Journal of Clinical Psychiatry, 63, 802–​806.
(2006). Lithium augmentation of topiramate for bipolar Malhotra, S., & McElroy, S. L. (2002). Orlistat misuse in bulimia
disorder with comorbid binge eating disorder and obesity. nervosa. American Journal of Psychiatry, 159, 492–​493.
Human Psychopharmacology, 21, 425–​431. Marrazzi, M. A., Bacon, J. P., Kinzie, J., & Luby, E. D. (1995).
Kucukgoncu, S., Midura, M., & Tek, C. (2015). Optimal man- Naltrexone use in the treatment of anorexia nervosa and
agement of night eating syndrome: Challenges and solutions. bulimia nervosa. International Clinical Psychopharmacology,
Neuropsychiatric Disease and Treatment, 11, 751–​760. 10, 163–​172.
Kühnen, P., Clément, K., Wiegand, S., Blankenstein, O., Marrazzi, M. A., Markham, K. M., Kinzie, J., & Luby, E. D.
Gottesdiener, K., Martini, L. L.,   .  .  .  Krude, H. (2016). (1995). Binge eating disorder:  Response to naltrexone.
Proopiomelanocortin deficiency treated with a melanocortin-​ International Journal of Obesity Related Metabolic Disorders,
4 receptor agonist. New England Journal of Medicine, 375, 19, 143–​145.
240–​246. Marzola, E., Desedime, N., Giovannone, C., Amianto, F.,
Lacey, J. H., & Crisp, A. H. (1980). Hunger, food intake and Fassino, S., & Abbate-​Daga, G. (2015). Atypical antipsy-
weight: The impact of clomipramine on a refeeding anorexia chotics as augmentation therapy in anorexia nervosa. PLos
nervosa population. Postgrad Med J, 56(Suppl 1), 79–​85. One, 10, e0125569.

390 Pharmacotherapy
McCann, U. D., & Agras, W. S. (1990). Successful treatment Zonisamide in the treatment of binge eating disorder with
of nonpurging bulimia nervosa with desipramine: A double-​ obesity:  A  randomized controlled trial. Journal of Clinical
blind, placebo-​ controlled study. American Journal of Psychiatry, 67, 1897–​1906.
Psychiatry, 147, 1509–​1513. McElroy, S. L., Kotwal, R., Hudson, J. I., Nelson, E. B., & Keck,
McElroy, S. L., Arnold, L. M., Shapira, N. A., Keck, P. E., Jr., P. E., Jr. (2004). Zonisamide in the treatment of binge eating
Rosenthal, N. R., Karim, M.,   .  .  .  Hudson, J. I. (2003). disorder: An open-​label, prospective trial. Journal of Clinical
Topiramate in the treatment of binge eating disorder asso- Psychiatry, 65, 50–​56.
ciated with obesity:  A  randomized, placebo-​ controlled McElroy, S. L., Kotwal, R., Keck, P. E., Jr., & Akiskal, J. S. (2005).
trial [published erratum appears in American Journal of Comorbidity of bipolar and eating disorders:  Distinct or
Psychiatry 2003;160:612]. American Journal of Psychiatry, related disorders with shared dysregulations? Journal of
160, 255–​261. Affective Disorders, 86, 107–​127.
McElroy, S. L., Casuto, L. S., Nelson, E. B., Lake, K. A., Soutullo, McElroy, S. L., Shapira, N. A., Arnold, L. M., Keck, P. E., Jr.,
C. A., Keck, P. E., Jr.,  . . .  Hudson, J. I. (2000). Placebo-​ Rosenthal, N. R., Wu, S. C.,   .  .  .  Hudson, J. I. (2004).
controlled trial of sertraline in the treatment of binge eating Topiramate in the long-​term treatment of binge-​eating dis-
disorder. American Journal of Psychiatry, 157, 1004–​1006. order associated with obesity [published erratum appears
McElroy, S. L., Guerdjikova, A. I., Blom, T. J., Crow, S. J., in Journal of Clinical Psychiatry 2005;66:138]. Journal of
Memisoglu, A., Silverman, B. L.,  . . .  Ehrich, E. W. (2013). Clinical Psychiatry, 65, 1463–​1469.
A placebo-​controlled pilot study of the novel opioid receptor Mehler-​Wex, C., Romanos, M., Kirchheiner, J., & Schulze, U.
antagonist ALKS-​33 in binge eating disorder. International M. (2008). Atypical antipsychotics in severe anorexia ner-
Journal of Eating Disorders, 46, 239–​245. vosa in children and adolescents:  Review and case reports.
McElroy, S. L., Guerdjikova, A. I., Kim, D. D., Burns, C., Harris-​ European Eating Disorders Review, 16, 100–​108.
Collazo, R., Landbloom, R.,  . . .  Dunayevich, E. (2013). Mendelson, S. D. (2001). Treatment of anorexia nervosa with
Naltrexone/​bupropion combination therapy in overweight tramadol. American Journal of Psychiatry, 158, 963–​964.
or obese patients with major depressive disorder: Results of Milano, W., Petrella, C., Casella, A., Capasso, A., Carrino, S., &
a pilot study. Primary Care Companion CNS Disorders, 15. Milano, L. (2005). Use of sibutramine, an inhibitor of the
McElroy, S. L., Guerdjikova, A., Kotwal, R., Welge, J. A., Nelson, reuptake of serotonin and noradrenalin, in the treatment of
E. B., Lake, K. A.,  . . .  Hudson, J. I. (2007). Atomoxetine binge eating disorder: A placebo-​controlled study. Advances
in the treatment of binge-​eating disorder:  A  randomized in Therapy, 22, 5–​31.
placebo-​controlled trial. Journal of Clinical Psychiatry, 68, Miller, K. K., Grieco, K. A., & Klibanski, A. (2005). Testosterone
390–​398. administration in women with anorexia nervosa. Journal of
McElroy, S. L., Guerdjikova, A. I., Martens, B., Keck, P. E., Jr., Clinical Endocrinology and Metabolism, 90, 1428–​1433.
Pope, H. G., Jr., & Hudson, J. I. (2009). Role of antiepileptic Misra, M., Katzman, D. K., Estella, N. M., Eddy, K. T.,
drugs in the management of eating disorders. CNS Drugs, Weigel, T., Goldstein, M. A.,  . . .  Klibanski, A. (2013).
23, 139–​156. Impact of physiologic estrogen replacement on anxiety
McElroy, S. L., Guerdjikova, A. I., Winstanley, E. L., O’Melia, symptoms, body shape perception, and eating attitudes in
A. M., Mori, N., Keck, P. E., Jr.,  . . .  Hudson, J. I. (2011). adolescent girls with anorexia nervosa:  Data from a ran-
Sodium oxybate in the treatment of binge eating disor- domized controlled trial. Journal of Clinical Psychiatry, 74,
der:  An open-​label, prospective study. International Journal e765–​e771.
of Eating Disorders, 44, 262–​268. Mitchell, J. E., Christenson, G., Jennings, J., Huber, M.,
McElroy, S. L., Hudson, J. I., Capece, J. A., Beyers, K., Fisher, A. Thomas, B., Pomeroy, C.,  . . .  Morley, J. (1989). A placebo-​
C., & Rosenthal, N. R. (2007). Topiramate for the treatment controlled, double-​ blind crossover study of naltrexone
of binge eating disorder associated with obesity: A placebo-​ hydrochloride in outpatients with normal weight bulimia.
controlled study. Biological Psychiatry, 61, 1039–​1048. Journal of Clinical Psychopharmacology, 9, 94–​97.
McElroy, S. L., Hudson, J., Ferreira-​ Cornwell, M. C., Mitchell, J. E., Fletcher, L., Hanson, K., Mussell, M. P., Seim,
Radewonuk, J., Whitaker, T., & Gaisor, M. (2016). H., Crosby, R.,  . . .  Al-​Banna, M. (2001). The relative effi-
Lisdexamfetamine dimesylate for adults with moderate to cacy of fluoxetine and manual-​based self-​help in the treat-
severe binge eating disorder: Results of two pivotal phase 3 ment of outpatients with bulimia nervosa. Journal of Clinical
randomized controlled trials. Neuropsychopharmacology, 41, Psychopharmacology, 21, 298–​304.
1251–​1260. Mitchell, J. E., & Groat, R. (1984). A placebo-​controlled double-​
McElroy, S. L., Hudson, J. I., Malhotra, S., Welge, J. A., Nelson, blind trial of amitriptyline in bulimia. Journal of Clinical
E. B., Keck, P. E., Jr. (2003). Citalopram in the treatment of Psychopharmacology, 4, 186–​193.
binge-​eating disorder: A placebo-​controlled trial. Journal of Mitchell, J. E., Pyle, R. L., Eckert, E. D., Hatsukami, D.,
Clinical Psychiatry, 64, 807–​813. Pomeroy, C., & Zimmerman, R. (1990). A comparison
McElroy, S. L., Hudson, J. I., Mitchell, J. E., Wilfley, D., study of antidepressants and structured intensive group psy-
Ferreira-​Cornwell, M. C., Gao, J.,  . . .  Whitaker, T. (2015). chotherapy in the treatment of bulimia nervosa. Archives of
Efficacy and safety of lisdexamfetamine for treatment of General Psychiatry, 47, 149–​157.
adults with moderate to severe binge eating disorder: A ran- Mond, J. M., Hay, P. J., Rodgers, B., & Owen, C. (2007).
domized clinical trial. JAMA Psychiatry, 72, 235–​246. Health service utilization for eating disorders: Findings from
McElroy, S. L., Keck, P. E., Jr., & Pope, H. G., Jr. (1987). a community-​based study. International Journal of Eating
Sodium valproate:  Its use in primary psychiatric disorders. Disorders, 40, 399–​408.
Journal of Clinical Psychopharmacology, 7, 16–​24. Monteleone, P., Luisi, M., Colurcio, B., Casarosa, E.,
McElroy, S. L., Kotwal, R., Guerdjikova, A. I., Welge, J. A., Monteleone, P., Ioime, R.,  . . .  Maj, M. (2001). Plasma lev-
Nelson, E. B., Lake, K. A.,   .  .  .  Hudson, J. I. (2006). els of neuroactive steroids are increased in untreated women

McElroy, Guerdjikova, Mori, Keck Jr. 391

 with anorexia nervosa or bulimia nervosa. Psychosomatic Powers, P. S., Bannon, Y., Eubanks, R., & McCormick, T.
Medicine, 63, 62–​68. (2007). Quetiapine in anorexia nervosa patients:  An open
Moore, R., Mills, I. H., & Forster, A. (1981). Naloxone in label outpatient pilot study. International Journal of Eating
the treatment of anorexia nervosa:  Effect on weight gain Disorders, 40, 21–​26.
and lipolysis. Journal of the Royal Society of Medicine, 74, Powers, P. S., Klabunde, M., & Kaye, W. (2012). Double-​blind
129–​135. placebo-​controlled trial of quetiapine in anorexia nervosa.
Naessén, S., Carlström, K., Byström, B., Pierre, Y., & Hirschberg, European Eating Disorder Review, 20, 331–​334.
A. L. (2007). Effects of an antiandrogenic oral contracep- Raingeard, I., Courtet, P., Renard, E., & Bringer, J. (2004).
tive on appetite and eating behavior in bulimic women. Naltrexone improves blood glucose control in type 1 diabetic
Psychoneuroendocrinology, 32, 548–​554. women with severe and chronic eating disorders. Diabetes
Nagata, T., Ono, K., & Nakayama, K. (2007). Anorexia nervosa Care, 27, 847–​848.
with chronic episodes for more than 30  years in a patient Ramoz, N., Versini, A., & Gorwood, P. (2007). Eating disor-
with a comorbid schizotypal personality disorder. Psychiatry ders: An overview of treatment responses and the potential
and Clinical Neurosciences, 61, 434–​436. impact of vulnerability genes and endophenotypes. Expert
National Institute for Clinical Excellence (NICE). (2004). Opinion on Pharmacotherapy, 8, 2029–​2044.
Eating disorders: Core interventions in the treatment and man- Reilly, P. P. (1977). Anorexia nervosa:  Lithium administration
agement of anorexia nervosa, bulimia nervosa and related eating has contributed to the management of anorexia nervosa.
disorders. Clinical guideline no 9. London, UK. Rhode Island Medical Journal, 60, 419–​456.
Neumeister, A., Winkler, A., & Wöber-​Bingöl, C. (1999). Ricca, V., Castellini, G., Lo Sauro, C., Rotella, C. M., & Faravelli,
Addition of naltrexone to fluoxetine in the treatment of binge C. (2009). Zonisamide combined with cognitive behavioral
eating disorder. American Journal of Psychiatry, 156, 797. therapy in binge eating disorder: A one-​year follow-​up study.
Newman-​ Toker, J. (2000). Risperidone in anorexia nervosa. Psychiatry (Edgemont), 6, 23–​28.
Journal of the American Academy of Child and Adolescent Robert, S. A., Rohana, A. G., Shah, S. A., Chinna, K.,
Psychiatry, 39, 941–​942. Wan Mohamud, W. N., & Kamaruddin, N. A. (2015).
Nickel, C., Tritt, K., Muehlbacher, M., Pedrosa Gil, F., Improvement in binge eating in non-​diabetic obese indi-
Mitterlehner, F. O., Kaplan, P.,  . . .  Nickel, M. K. (2005). viduals after 3 months of treatment with liraglutide: A pilot
Topiramate treatment in bulimia nervosa patients:  A  ran- study. Obesity Research and Clinical Practice, 9, 301–​304.
domized, double-​blind, placebo-​controlled trial. International Romano, S. J., Halmi, K. A., Sarkar, N. P., Koke, S. C., & Lee,
Journal of Eating Disorders, 38, 295–​300. J. S. (2002). A placebo-​controlled study of fluoxetine in con-
Noma, S., Uwatoko, T., Yamamoto, H., & Hayashi, T. (2008). tinued treatment of bulimia nervosa after successful acute
Effects of milnacipran on binge eating:  A  pilot study. fluoxetine treatment. American Journal of Psychiatry, 159,
Neuropsychiatric Disease and Treatment, 4, 295–​300. 96–​102.
Ong, Y. L., Checkley, S. A., & Russell, G. F. (1983). Suppression Rosenow, F., Knake, S., & Hebebrand, J. (2002). Topiramate
of bulimic symptoms with methylamphetamine. British and anorexia nervosa. American Journal of Psychiatry, 159,
Journal of Psychiatry, 143, 288–​293. 2112–​2113.
O’Reardon, J. P., Allison, K. C., Martino, N. S., Lundgren, J. Rothschild, R., Quitkin, H. M., Quitkin, F. M., Stewart, J.
D., Moonseong, H., & Stunkard, A. J. (2006). A random- W., Ocepek-​Welikson, K., McGrath, P. J.,.   .  .  .  Tricamo,
ized, placebo-​controlled trial of sertraline in the treatment of E. (1994). A double-​blind placebo-​controlled comparison of
night eating syndrome. American Journal of Psychiatry, 163, phenelzine and imipramine in the treatment of bulimia in
893–​898. atypical depressives. International Journal of Eating Disorders,
Parsons, J. M., & Sapse, A. T. (1985). Significance of hyper- 15, 1–​9.
cortisolism in anorexia nervosa. Journal of Orthomolecular Ruggiero, G. M., Laini, V., Mauri, M. C., Ferrari, V. M.,
Psychiatry, 14, 13–​18. Clemente, A., Lugo, F.,  . . .  Cavagnini, F. (2001). A single
Pataky, Z., Gasteyger, C., Ziegler, O., Rissanen, A., Hanotin, C., blind comparison of amisulpride, fluoxetine and clomip-
& Golay, A. (2013). Efficacy of rimonabant in obese patients ramine in the treatment of restricting anorectics. Progress
with binge eating disorder. Experimental and Clinical in Neuro-​psychopharmacology and Biological Psychiatry, 25,
Endocrinology and Diabetes, 121, 20–​26. 1049–​1059.
Pearlstein, T., Spurell, E., Hohlstein, L. A., Gurney, V., Read, Rush, A. J. (2011). Star-​D: Lessons learned and future implica-
J., Fuchs, C.,   .  .  .  Keller, M. B. (2003). A double-​blind, tions. Depression and Anxiety, 28, 521–​524.
placebo-​controlled trial of fluvoxamine in binge eating dis- Russell, D. M., Freedman, M. L., Feiglin, D. H., Jeejeebhoy, K. N.,
order: A high placebo response. Archives of Women’s Mental Swinson, R. P., & Garfinkel, P. E. (1983). Delayed gastric empty-
Health, 6, 147–​151. ing and improvement with domperidone in a patient with ano-
Pope, H. G., Hudson, J. I., & Jonas, J. M. (1986). Bulimia in rexia nervosa. American Journal of Psychiatry, 140, 1235–​1236.
men: A series of fifteen cases. Journal of Nervous and Mental Russell, G. F.  M., Checkley, S. A., Feldman, J., & Eisler, I.
Disease, 174, 117–​119. (1988). A controlled trial of d-​fenfluramine in bulimia ner-
Pope, H. G., Jr., Hudson, J. I., Jonas, J. M., & Yurgelun-​Todd, vosa. Clinical Neuropharmacology, 11(Suppl 1), S146–​S159.
D. (1983). Bulimia treated with imipramine:  Placebo-​ Sabine, E. J., Yonace, A., Farrington, A. J., Barrant, K. H., &
controlled, double-​ blind study. American Journal of Wakeling, A. (1983). Bulimia nervosa: A placebo-​controlled
Psychiatry, 140, 554–​558. double-​blind therapeutic trial of mianserin. British Journal of
Pope, H. G., Jr., Keck, P. E., Jr., McElroy, S. L., & Hudson, Clinical Pharmacology, 15, 195–​202.
J. I. (1989). A placebo-​controlled study of trazodone in Saleh, J. W., & Lebwohl, P. (1980). Metoclopramide-​induced
bulimia nervosa. Journal of Clinical Psychopharmacology, 9, gastric emptying in patients with anorexia nervosa. American
254–​259. Journal of Gastroenterology, 74, 127–​132.

392 Pharmacotherapy
Sansone, R. A., & Sansone, L. A. (2003). Metoclopramide and Stunkard, A., Berkowitz, R., Tanrikut, C., Reis, E., & Young, L.
unintended weight gain. International Journal of Eating (1996). d-​Fenfluramine treatment of binge eating disorder.
Disorders, 34, 265–​268. American Journal of Psychiatry, 153, 1455–​1459.
Schmidt do Prado-​ Lima P. A., & Bacaltchuck, J. (2002). Stunkard, A. J. & Messick, S. (1985). The three-​factor eating
Topiramate in treatment-​resistant depression and binge eat- questionnaire to measure dietary restraint, disinhibition and
ing disorder. Bipolar Disorders, 4, 271–​273. hunger. Journal of Psychosomatic Research, 29, 71–​83.
Schweickert, L. A., Strober, M., & Moskowitz, A. (1997). Su, J. C., & Birmingham, C. L. (2002). Zinc supplementation
Efficacy of methylphenidate in bulimia nervosa comorbid in the treatment of anorexia nervosa. Eating and Weight
with attention-​deficit hyperactivity disorder:  A  case report. Disorders, 7, 20–​22.
International Journal of Eating Disorders, 21, 299–​301. Sundblad, C., Landén, M., Eriksson, T., Bergman, L., &
Shapira, N. A., Goldsmith, T. D., & McElroy, S. L. (2000). Eriksson, E. (2005). Effects of the androgen antagonist flu-
Treatment of binge-​eating disorder with topiramate: A clini- tamide and the serotonin reuptake inhibitor citalopram in
cal case series. Journal of Clinical Psychiatry, 61, 368–​372. bulimia nervosa: A placebo-​controlled pilot study. Journal of
Shapiro, J. R., Berkman, N. D., Brownley, K. A., Sedway, J. S., Clinical Psychopharmacology, 25, 85–​88.
Lohr, K. N., & Bulik, C. M. (2007). Bulimia nervosa treat- Szmukler, G. I., Young, G. P., Miller, G., Lichtenstein, M.,
ment:  A  systematic review of randomized controlled trials. & Binns, D. S. (1995). A controlled trial of cisapride in
International Journal of Eating Disorders, 40, 321–​336. anorexia nervosa. International Journal of Eating Disorders,
Shisslak, C. M., Perse, T., & Crago, M. (1991). Coexistence 17, 347–​357.
of bulimia nervosa and mania: A literature review and case Szyper, M. S., & Mann, J. D. (1978). Anorexia nervosa as an
report. Comprehensive Psychiatry, 32, 181–​184. interictal symptom of partial complex seizures. Neurology,
Silveira, R. O., Zanatto, V., Appolinário, J. C., & Kapezinski, 28, 335.
F. (2005). An open trial of reboxetine in obese patients Tachibana, N., Sugita, Y., Teshima, Y., & Hishikawa, Y. (1989).
with binge eating disorder. Eating and Weight Disorders, 10, A case of anorexia nervosa associated with epileptic sei-
e93–​e96. zures showing favorable responses to sodium valproate and
Slof-​Op’t Landt, M. C., van Furth, E. F., Meulenbelt, I., clonazepam. Japanese Journal of Psychiatry and Neurology,
Slagbloom, P. E., Bartels, M., Boomsma, D. I.,  . . .  Bulik, 43, 77–​84.
C. M. (2005). Eating disorders: From twin studies to candi- Takaki, M., & Okabe, N. (2015). Ariprazole may be effective as
date genes and beyond. Twin Research and Human Genetics, an add-​on treatment in bulimic symptoms of eating disor-
8, 467–​482. ders. Journal of Clinical Psychopharmacology, 35, 93–​95.
Sokol, M. S., Gray, N. S., Goldstein, A., & Kaye, W. H. (1999). Theisen, F. M., Linden, A., König, I. R., Martin, M.,
Methylphenidate treatment for bulimia nervosa associated Remschmidt, H., & Hebebrand, J. (2003). Spectrum of
with a cluster B personality disorder. International Journal of binge eating symptomatology in patients treated with clo-
Eating Disorders, 25, 233–​237. zapine and olanzapine. Journal of Neural Transmission, 110,
Stacher, G., Abatzi-​Wenzel, T. A., Wiesnagrotzki, S., Bergmann, 111–​121.
H., Schneider, C., & Gaupmann, G. (1993). Gastric empty- Trunko, M. E., Schwartz, T. A., Duvvuri, V., & Kaye, W. H.
ing, body weight and symptoms in primary anorexia nervosa. (2011). Aripiprazole in anorexia nervosa and low-​ weight
Long-​term effects of cisapride. British Journal of Psychiatry, bulimia nervosa: Case reports. International Journal of Eating
162, 398–​402. Disorders, 44, 269–​275.
Stacher, G., Peters, T. L., Bergmann, H., Wiesnagrotzki, S., Trunko, M. E., Schwartz, T. A., Marzola, E., Klein, A. S., &
Schneider, C., Granser-​Vacariu, G. V.,.  . . .  Kugi, A. (1993). Kaye, W. H. (2014). Lamotrigine use in patients with binge
Erythromycin effects on gastric emptying, antral motility eating and purging, significant affect dysregulation, and poor
and plasma motilin and pancreatic polypeptide concentra- impulse control. International Journal of Eating Disorders, 47,
tions in anorexia nervosa. Gut, 34, 166–​172. 329–​334.
Stefano, S. C., Bacaltchuk, J., Blay, S. L., & Appolinario, J. C. Umehara, H., Iga, J., & Ohmori, T. (2014). Successful treat-
(2008). Antidepressants in short-​term treatment of binge ment of anorexia nervosa in a 10-​year-​old boy with risperi-
eating disorder: Systematic review and meta-​analysis. Eating done long-​acting injection. Clinical Psychopharmacology and
Behaviors, 9, 129–​136. Neurosciences, 12, 65–​66.
Stein, G. S., Hartshorn, S., Jones, J., & Steinberg, D. (1982). Vandereycken, W. (1984). Neuroleptics in the short-​term treat-
Lithium in a case of severe anorexia nervosa. British Journal ment of anorexia nervosa: A double-​blind placebo-​controlled
of Psychiatry, 140, 526–​528. study with sulpiride. British Journal of Psychiatry, 144,
Steinglass, J. E., Eisen, J. L., Attia, E., Mayer, L., & Walsh, B. 288–​292.
T. (2007). Is anorexia a delusional disorder? An assessment Vandereycken, W., & Pierloot, R. (1982). Pimozide combined
of eating beliefs in anorexia nervosa. Journal of Psychiatric with behavior therapy in the short-​term treatment of ano-
Practice, 13, 65–​71. rexia nervosa: A double-​blind, placebo-​controlled, cross-​over
Steinglass, J. E., Kaplan, S. C., Liu, Y., Wang, Y. & Walsh, B. T. study. Acta Psychiatrica Scandinavica, 60, 446–​451.
(2014). The (lack of ) effect of alprazolam on eating behav- Vander Wal, J. S., Gang, C. H., Griffing, G. T., & Gadde, K.
ior in anorexia nervosa: A preliminary report. International M. (2012). Escitalopram for treatment of night eating syn-
Journal of Eating Disorders, 47, 901–​904. drome:  A  12-​ week, randomized, placebo-​ controlled trial.
Steinglass, J. E., Sysko, R., Schebendach, J., Broft, A., Strober, Journal of Clinical Psychopharmacology, 32, 341–​345.
M., & Walsh, B. T. (2007). The application of exposure von Wietersheim, J., Müler-​Bock, V., Rauh, S., Danner, B.,
therapy and D-​cycloserine to the treatment of anorexia ner- Chrenko, K., & Bühler, G. (2008). No effect on spirono-
vosa: A preliminary trial. Journal of Psychiatric Practice, 13, lactone on bulimia nervosa symptoms. Journal of Clinical
238–​245. Psychopharmacology, 28, 258–​260.

McElroy, Guerdjikova, Mori, Keck Jr. 393

Walsh, B. T., Agras, W. S., Devlin, M. J., Fairburn, C. G., Wilson, White, M. A., & Grilo, C. M. (2013). Bupropion for overweight
G. T., Kahn, C. M.,.  . . .  Chally, M. K. (2000). Fluoxetine women with binge-​eating disorder: A randomized, double-​
for bulimia nervosa following poor response to psychother- blind, placebo-​controlled trial. Journal of Clinical Psychiatry,
apy. American Journal of Psychiatry, 157, 1332–​1334. 74, 400–​406.
Walsh, B. T., Fairburn, C. G., Mickley, D., Sysko, R., & Parides, Wilfley, D. E., Crow, S. J., Hudson, J. I., Mitchell, J. E.,
M. K. (2004). Treatment of bulimia nervosa in a primary Berkowitz, R. I., Blakesley, V., & Walsh, B. T. (2008).
care setting. American Journal of Psychiatry, 161, 556–​561. Efficacy of sibutramine for the treatment of binge eating
Walsh, B. T., Hadigan, C. M., Devlin, M. J., Gladis, M., & disorder:  A  randomized multicenter placebo-​ controlled
Roose, S. P. (1991). Long-​term outcome of antidepressant double-​blind study. American Journal of Psychiatry,
treatment for bulimia nervosa. American Journal of Psychiatry, 165, 51–​58.
148, 1206–​1212. Wysowski, D. K., Corken, A., Gallo-​Torres, H., Talarico, L.,
Walsh, B. T., Kaplan, A. S., Attia, E., Olmsted, M., Parides, & Rodriguez, E. M. (2001). Postmarketing reports of QT
M., Carter, O. C.,   .  .  .  Rockert, W. (2006). Fluoxetine prolongation and ventricular arrhythmia in association
after weight restoration in anorexia nervosa:  A  random- with cisapride and Food and Drug Administration regu-
ized controlled trial [published erratum appears in Journal latory actions. American Journal of Gastroenterology, 96,
of the American Medical Association, 2006; 296, 934, 1698–​1703.
and Journal of the American Medical Association 2007; Woodside, B. D., & Staab, R. (2006). Management of psychi-
298,  2008]. Journal of the American Medical Association, atric comorbidity in anorexia nervosa and bulimia nervosa.
295, 2605–​2612. CNS Drugs, 20, 655–​663.
Walsh, B. T., Stewart, J. W., Roose, S. P., Gladis, M., & Yager, J. (2007). Management of patients with chronic, intracta-
Glassman, A. H. (1984) Treatment of bulimia with phen- ble eating disorders. In J. Yager & P. S. Powers (Eds.), Clinical
elzine: A double-​blind, placebo-​controlled study. Archives of manual of eating disorders (pp. 407–​439). Washington,
General Psychiatry, 41, 1105–​1109. DC: American Psychiatric Publishing.
Walsh, B. T., Wilson, G. T., Loeb, K. L. Devlin, M. J., Pike, K. Yager, J., & Powers, P. S. (Eds.). (2007). Clinical manual of
M., Roose, S. P.,  . . .  Waternaux, C. (1997). Medication and eating disorders. Washington, DC:  American Psychiatric
psychotherapy in the treatment of bulimia nervosa. American Publishing.
Journal of Psychiatry, 154, 523–​531. Yamamoto, T., Kanahara, N., Hirai, A., Watanabe, H., & Iyo,
Wang, T. S., Chou, Y. H., & Shiah, I. S. (2006). Combined M. (2013). Lamotrigine in binge-​eating disorder associated
treatment of olanzapine and mirtazapine in anorexia with bipolar II depression and treatment-​resistant type 2 dia-
nervosa associated with major depression. Progress in betes mellitus:  A  case report. Clinical Neuropharmacology,
Neuropsychopharmacology and Biological Psychiatry, 30, 36, 34–​35.
306–​309. Yasuhara, D., Nakahara, T., Harada, T., & Inui, A. (2007).
Weibel, S., Lalanne, L., Riegert, M., & Bertschy, G. (2015). Olanzapine-​ induced hyperglycemia in anorexia nervosa.
Efficacy of high-​dose baclofen for alcohol use disorder and American Journal of Psychiatry, 164, 528–​529.
comorbid bulimia: A case report. Journal of Dual Diagnosis, Yilmaz, Z., Hardaway, J. A., & Bulik, C. M. (2015). Genetics
11, 203–​204. and epigenetics of eating disorders. Advanced Genomics
Wermuth, B. M., Davis, K. L., Hollister, L. E., & Stunkard, A. J. Genetics, 5, 131–​150.
(1977). Phenytoin treatment of the binge eating syndrome. Zilberstein, B., Pajecki, D., Garcia de Brito, A. C., Gallafrio,
American Journal of Psychiatry, 134, 1249–​1253. S. T., Eshkenazy, R., & Andrade, C. G. (2004). Topiramate
Wheadon, D. E., Rampey, A. H., Jr., Thompson, V. L., Potvin, after adjustable gastric banding in patients with binge
J. H., Massica, D. N., & Beasley, C. M., Jr. (1992). Lack eating and difficulty losing weight. Obesity Surgery, 14,
of association between fluoxetine and suicidality in bulimia 802–​805.
nervosa. Journal of Clincal Psychiatry, 53, 235–​241.

394 Pharmacotherapy
 CH A PT E R


Cognitive Remediation Therapy
20 for Eating Disorders

Amy Harrison

Abstract
Eating disorders (EDs) have been described as among the most difficult psychiatric disorders to treat.
Intervening early appears to be associated with better prognosis, although a subgroup of 20% of
individuals may develop a more severe and enduring form of illness, which is associated with higher
rates of mortality. Many patients with EDs who come into contact with clinical services may have
extreme ambivalence toward change, which is often observed through high treatment dropout rates and
difficulties engaging in treatment. This chapter outlines cognitive remediation therapy (CRT) for eating
disorders, a treatment enhancer designed to support individuals with severe and complex forms of
illness. This chapter explores how CRT has been used, examines its efficacy, reflects on its place as part
of an overall treatment package for patients with EDs, and finally, explores options for future research in
the field.
Key Words:  cognitive remediation therapy, efficacy, anorexia nervosa, historical development, CRT,
ambivalence

Introducing Cognitive Remediation between sessions and to develop curiosity in rela-

Therapy tion to their cognitive skills. The therapist may be
Cognitive remediation therapy (CRT) is a encouraged to also take part in the simple exercises
therapeutic adjunct that aims to support patients and to use the nonthreatening cognitive tasks to
to improve inefficiencies in thinking styles facilitate discussion around the process of think-
(Tchanturia, 2014). Somewhat uniquely, CRT ing, making clear links to everyday problem solving
focuses on the process of thinking, rather than the (Tchanturia, 2014).
content; the “how,” rather than the “what.” To Throughout this chapter, a transdiagnostic
achieve this, CRT employs simple cognitive exer- approach is adopted (Fairburn, Cooper, & Shafran,
cises to allow patients to practice and demonstrate 2003), with the term “EDs” used throughout, due
different cognitive strategies and to then reflect on to the frequent temporal movement between catego-
the thinking styles and strategies used (Tchanturia, rical diagnoses (Keel & Mitchell, 1997) and because
Lounes, & Holttum, 2014). This metacognitive many of the patients who may benefit from CRT
component involves the patient exploring links are likely to have had experience of a number of
between their approach to the simple CRT tasks and different symptom presentations (Milos, Spindler,
the strategies they use to approach tasks in everyday Schnyder, & Fairburn, 2005). However, where
life and encourages the patient to develop a range research has been conducted including a specific
of tools that allow them to approach daily tasks group of patients, the terms anorexia nervosa (AN)
with greater efficiency and confidence (Tchanturia or bulimia nervosa (BN) are used in keeping with
& Lock, 2011). Patients are encouraged to practice the original research.

395
Background: Historical Development depending on the task administered, small-​to large-​
of Cognitive Remediation Therapy sized differences in set-​shifting in those with EDs
for Eating Disorders relative to non-​ED controls. An updated review
Originally, CRT was developed for people with suggested continued support for these findings
acquired brain injuries with a focus on compensa- (Westwood, Stahl, Mandy, & Tchanturia, 2016).
tion strategies to help patients approach everyday Set-​shifting or cognitive flexibility has been defined
tasks with greater efficiency; this work is summa- as a form of self-​regulation in which the individ-
rized by Cicerone et  al. (2011). Following this, ual demonstrates their ability to “shift their course
CRT was adapted and implemented to help people of thought or action according to the demands of
with schizophrenia develop more efficient cogni- the situation” (Lezak, 2004, p.  666). Indeed, set-​
tive strategies with the aim of improving established shifting is the ability to shift set, or to move back
cognitive deficits in this patient group; findings sup- and forth between multiple tasks, operations, or
porting improved neuropsychological functioning, mental sets (Miyake et  al., 2000). In support of
as reviewed by Wykes, Huddy, Cellard, McGurk, superior skills for detail processing, sometimes
and Czobor (2011). Also, CRT has been used to referred to as field independence, or an analytically
support people with depression for whom cog- oriented processing style (Witkin & Goodenough,
nitive inefficiencies are also established as a core 1977) alongside inefficiencies in global processing,
feature of the illness (Porter, Bowie, Jordan, & sometimes referred to as field dependence, or a glob-
Malhi, 2013) and, as in AN (Harrison, Tchanturia, ally oriented information processing style (Witkin
Naumannn, & Treasure, 2012; Tchanturia, 2014; & Goodenough, 1977), a systematic review iden-
Tchanturia et  al., 2012; Tchanturia et  al., 2011), tified 12 studies involving patients with EDs and
data indicate that some of these cognitive differ- reported medium-​sized differences for these cogni-
ences might remain despite symptom improvement tive skills in those with EDs relative to non-​ED con-
(Douglas, Porter, Knight, & Maruff, 2011). Porter trols (Lang, Lopez, Stahl, Tchanturia, & Treasure,
et  al. (2013) reviewed the published literature on 2014). This profile of weak central coherence,
CRT for major depressive disorder and found 10 where a focus on local, featural information domi-
studies had reported on the development and eval- nates (Happé & Frith, 2006), is proposed to be an
uation of cognitive-​remediation treatment packages important maintaining factor for adults with EDs
for people with depression. In general, these stud- (Schmidt & Treasure, 2006; Treasure & Schmidt,
ies reported medium sized improvements in cogni- 2013). Therefore, a treatment enhancer aimed at
tive functioning following treatment. Interestingly, helping patients to improve their skills in this area
it is well established that depression is a com- may be a useful way of assisting them to benefit
mon comorbid diagnosis for people with EDs from evidence-​based treatments such as cognitive-​
(Blinder, Cumella, & Sanathara, 2006; Kaye, Bulik, behavioral therapy (CBT; Fairburn, 2008) and fam-
Thornton, Barbarich, & Masters, 2004). ily therapy (Dare & Eisler, 1997; Lock, Le Grange,
Subsequent to its use in brain injury, psycho- Agras, & Dare, 2001).
sis, and depression, CRT was adapted for work- There also appears to be emerging evidence of a
ing with people with eating disorders (EDs) and similar neuropsychological profile in children and
offered a unique component, which was a focus adolescents with EDs, with support provided from
on core cognitive processes rather than ED symp- two systematic reviews on set-​shifting (Lang, Stahl,
toms (Tchanturia & Lock, 2011). The rationale for Espie, Treasure, & Tchanturia, 2014) and central
using cognitive remediation to support functioning coherence (Lang & Tchanturia, 2014). A  small
in people with EDs was informed by an emerging amount of preliminary data suggests that young
literature demonstrating a number of cognitive women whose ED onset occurred during adoles-
inefficiencies in adults with EDs, in particular dif- cence and was associated with amenorrhea or irregu-
ficulties thinking and behaving in a flexible manner lar menses demonstrate greater cognitive difficulties
and a superior focus on detail (Tchanturia, 2014; and difference in brain structure in their early 20s
Tchanturia et  al., 2012; Tchanturia et  al., 2011). (Chui et  al., 2008). In a small sample of 25 ado-
For example, support for the presence of subop- lescents with AN, participants demonstrated poorer
timal set-​shifting skills is provided in a systematic cognitive functioning than 26 non-​ED controls,
review and meta-​ analysis conducted by Roberts, and those with AN showed an improvement in cog-
Tchanturia, Stahl, Southgate, and Treasure (2007), nitive ability after weight recovery which was more
which included 15 studies and demonstrated, pronounced in those with restored menstruation

396 Cognitive Remediation

(Lozano-​Serra, Andrés-​Perpiña, Lázaro-​García, & profile of people with EDs continued to influence
Castro-​Fornieles, 2014). However, the literature is and update CRT (Tchanturia et al., 2013; Martinez,
clearly less well developed in this younger cohort Cook-​Darzens, Chaste, Mouren, & Doyen, 2014;
than it is for the adult population, and these reviews Tchanturia et  al., 2014) and the first randomized
highlight both small sample sizes and variability in controlled trial (RCT) of CRT for AN highlighted
task selection and outcomes. its ability to retain patients in treatment, as well
as improve inefficiencies in cognitive processing
Cognitive Remediation as a Therapy (Lock et al., 2013). That 33% of patients dropped
Enhancer for Eating Disorders out of CBT compared with a dropout rate of 13%
It was initially proposed that ten 45-​minute ses- in the group of patients receiving both CRT and
sions of individual, face-​to face CRT might offer CBT reflects one of the useful functions of CRT,
a means of establishing a therapeutic relationship such that this low-​intensity intervention might pro-
with patients with severe and enduring AN who, vide a collaborative, nonthreatening, and playful
nutritionally compromised, require admission to an way to engage and begin to develop a rapport with
inpatient unit and may find the curious, collabora- severely underweight patients with EDs who may
tive, playful, and nonthreatening stance of CRT a have already experienced a large number of previous
useful introduction to further, higher intensity, psy- treatments and might lack confidence and the desire
chological treatment (Tchanturia & Lock, 2011; to pursue further talking therapies (Tchanturia,
Tchanturia, Lloyd, & Lang, 2013; Tchanturia et al., Davies, Reeder, & Wykes, 2010).
2014). Additionally, it was also proposed that CRT Cognitive remediation therapy is not proposed
might facilitate the development of useful cogni- to be a panacea for EDs, rather it is a way of helping
tive skills that would enable patients to better use to engage patients with the most severe and endur-
higher intensity treatments during their admission ing forms of illness who may not have previously
(Tchanturia et  al., 2013). Initial case studies and taken part in treatment interventions to engage in
case series then set out to explore the potential ben- the treatment program on offer. It shares a num-
efits of CRT in AN. ber of similarities with CBT, well known to have
As previously mentioned, work conducted by an evidence base for helping individuals with EDs
clinical researchers in London, UK, initially focused overcome their difficulties (Fairburn et  al., 2015).
on using the intervention with an adult inpatient It has also a number of features that differentiate
population, which represented a severely unwell it from CBT. Before moving to the section where
cohort with a predominantly enduring form of AN the efficacy of CRT for EDs is explored, Table 20.1
(Davies, & Tchanturia, 2005; Tchanturia, Whitney, suggests some ways in which the two treatments
& Treasure, 2006). After preliminary findings indi- might be complementary and highlights some of
cated its efficacy (Tchanturia, Davies, & Campbell, their possible differences as a way of helping clini-
2007), and patients provided positive reports of cians to reflect on the transferable skills they have
its usefulness and fresh focus on skills rather than from using other evidence-​ based treatments for
symptoms (Whitney, Easter, & Tchanturia, 2008), EDs which can also be used in the delivery of CRT.
a pilot case series was conducted, which indicated its
relevance and possible efficacy for improving cog- A Review of the Evidence to Date: Cognitive
nitive functioning in adults with AN (Tchanturia Remediation Therapy for Eating
et al., 2008). By 2009, other research groups around Disorders—​Randomized Controlled Trials
the world began to use CRT in their patient popu- in Adult Populations
lations and publish their findings (Cwojdzińska, A systematic review of the literature identi-
Markowska-​Regulska, & Rybakowski, 2009) fied four RCTs conducted in adults with AN
and group formats were developed (Genders & (Tchanturia et  al., 2014) and highlighted promis-
Tchanturia, 2010; Tchanturia & Doris, 2015). Also, ing findings.
CRT was trialed in outpatient settings (Pitt, Lewis, The first published RCT was conducted in the
Morgan, & Woodward, 2010), and by 2011 it had United States in an adult outpatient setting and
been adapted for both adolescent inpatients and aimed to explore the feasibility of using CRT with
those accessing intensive day hospital treatment for a focus on reducing dropout rates in treatment for
EDs (Wood, Al-​Khairulla, & Lask, 2011; Pretorius AN. Forty-​six outpatients with AN were randomly
et  al., 2012). Throughout its development, the assigned to either eight sessions of CRT (n  =  23)
improved understanding of the neuropsychological or CBT (n  =  23) over 2  months, followed by 16

Harrison 397
Table 20.1  Similarities and Differences Between Cognitive Remediation Therapy and Cognitive-​Behavioral
Therapy
Cognitive-​Behavioral Therapy Cognitive Remediation Therapy

Similarities

Collaboration and a curious stance offered by the clinician

Homework exercises

Manualized format with defined interventions/​exercises

Structured, time limited, relatively brief

Aims to provide patients with tangible skills and knowledge

Involves behavioral experiments

Differences

A focus on the content of cognition (albeit A focus on the process, style of cognition, or the
with the exception of where thinking errors are cognitive strategy used
challenged)

Directive, goal focused Patient led and directed

Behavioral experiments are used to challenge Behavioral experiments are used to reflect on the
beliefs and form more useful and adaptive patient’s current cognitive strategies and to try out
thoughts and behaviors to make changes to the alternative strategies to offer greater options in terms
“what” of cognition and behavior of the “how” of thinking and behavior

Requires a high degree of training to be Requires a relatively lower level of training and
delivered, particularly in a more high-​ is designed to be delivered by trainee therapists,
intensity context graduate students, members of the nursing team as
well as trained therapists

CBT sessions over 4  months. The authors found A third RCT was conducted by Brockmeyer
a lower dropout rate occurred in the CRT group et  al. (2014) in Germany with the aim of explor-
(13%) compared with the CBT group (33%) and ing the feasibility and possible efficacy of CRT
also found improvements in cognitive efficiencies in through randomly allocating 40 adult inpatients to
the CRT group compared with the CBT group at tailored CRT (n = 20) or nonspecific neurocogni-
the end of the trial (Lock et al., 2013). tive therapy (NNT; n = 20); patients were offered
In a second RCT conducted in The Netherlands, a more intensive 30 sessions (21 computer-​assisted
Dingemans et  al. (2014) randomly allocated 82 and 9 face-​to-​face) and assigned computer-​assisted
adult inpatients with severe and enduring AN to homework. The NNT focused only on attention,
CRT plus treatment as usual (TAU; n = 41) or TAU memory, and deductive reasoning. The manual-​
only (n = 41). At the end of treatment and 6-​month based CRT was tailored and focused solely on set-​
follow-​up, CRT was associated with significant shifting; as argued by the authors, central coherence
improvements in quality of life. The authors con- was omitted to remove any potentially confounding
ducted a moderation analysis, which indicated that factors with regard to the control condition. The
patients with poorer baseline set-​shifting abilities primary outcome was performance on a computer-​
benefited more from CRT and had better quality based task-​switching paradigm that assessed pre-​
of life at follow-​up. Performance on tasks of set-​ intervention and post-​intervention. Overall, CRT
shifting and global processing, however, improved participants significantly outperformed the NNT
significantly in both the CRT and TAU conditions, group in set-​shifting with a medium effect size,
and the authors attributed this finding to practice and overall patient feedback was more positive for
effects or nonspecific ingredients of treatment. CRT. The authors suggested that specific tailored

398 Cognitive Remediation

neurocognitive training is more effective and argue across 10 sessions to adolescent inpatients aged
for the feasibility of CRT for AN. 13–​19 with AN. The patients reported the group
A fourth RCT (Steinglass et  al., 2014) com- to be “fun and useful,” and the authors further cor-
pared CRT with exposure and response prevention roborated the possible benefits of CRT for young
for AN (AN-​EXRP), a new approach that targets people with EDs, suggesting that CRT might be
maladaptive eating behavior by addressing eating-​ a potentially useful addition to the treatment of
related anxiety. Adult inpatients (n = 32) who were young people with AN. Another study built on
weight restored (BMI over 18.5) were offered 12 the evidence base for the possible benefits of CRT
sessions of AN-​EXRP or CRT, and the outcome for young people, also focusing on its application
measure was caloric intake at a test meal, which the in a group setting. Pretorius et al. (2012) reported
authors found was higher in the AN-​EXRP group; on the delivery of seven CRT groups attended by
this improvement was also significantly associated a total of 30 adolescent outpatients with AN and
with eating-​related anxiety. However, because the found a nonsignificant small-​sized improvement in
inpatients in this study were weight restored, they self-​reported cognitive flexibility, measured using
differ significantly from participants in all the other the Cognitive Flexibility Scale (Martin & Rubin,
studies mentioned earlier. 1995). Qualitative feedback on the acceptability of
To date, there are no published RCTs of CRT for the group indicated that this was perceived as an
children and adolescents with EDs, with Dahlgren acceptable and useful group treatment by patients.
and Rø (2014, p. 26) highlighting in their review In 2013, Dahlgren, Lask, Landrø, and Rø con-
that “it is imperative that adolescent RCTs are con- ducted an uncontrolled feasibility trial exploring
ducted to evaluate how CRT may be effective as a the possible impact of CRT on neuropsychological
treatment for this young patient group.” The next functioning in 22 adolescents with AN. After CRT,
section therefore explores the evidence reported to significant improvements in weight, symptoms of
date regarding how CRT has been used to support depression, and a range of neuropsychological out-
young people with EDs. comes including visual-​spatial memory, perceptual
disembedding abilities and verbal fluency were
A Review of the Evidence to Date: Cognitive observed. Dahlgren, Lask, Landrø, and Rø (2014)
Remediation Therapy for Eating Disorders followed these findings up with a further study of
in Child and Adolescent Populations 20 adolescent inpatients and outpatients with AN
The first published article that explored the use aged 13–​18. Patients received once or twice weekly
of CRT for young people with EDs was provided by individually tailored CRT sessions, and the results
Cwojdzińska et al. (2009), who described a case study highlighted that the intervention was experienced
in which they used CRT with an adolescent patient as acceptable with no voluntary dropouts and 19 of
with AN. They reported improved performance on 20 patients completed all sessions.
the Wisconsin Card Sort Task (Heaton, Chelune, Asch et al. (2014) reported on the effectiveness
Talley, Kay, & Curtiss, 1993) indicating improved of weekly 1-​hour sessions of CRT offered over 10
set-​shifting, and this change was also observed weeks to 10 adolescent inpatients aged 12–​17 with
alongside an improvement in ED symptoms. Early AN. Patients’ set-​shifting skills were explored before
indications of the possible benefits of CRT for a and after treatment, and individuals were also asked
younger population of individuals with EDs were to provide qualitative feedback on their experi-
further highlighted by Easter and Tchanturia (2011) ence of the intervention. Only 2 out of 10 patients
in their qualitative study, which explored therapists’ (20%) completed all sessions. However, patients
views of CRT and they reported that CRT might have demonstrated improved set-​shifting after the inter-
possible benefits for young people with EDs. This vention, which also appeared to be associated with
perspective is corroborated by Martinez et al. (2014), improvements in body mass index, one physical
who have also highlighted the possible utility of CRT predictor of recovery. Dahlgren et  al. (2014) col-
for adolescent patients with AN. lected self-​report outcomes from 17 adolescent
Alongside these predictions around the possible patients and their caregivers using the Inventory of
uses of CRT for younger cohorts of individuals Executive Function (BRIEF; Gioia, Isquith, Guy, &
with EDs, a growing research literature has begun Kenworthy, 2000) before and after CRT and found
to develop and enhance the evidence base. Wood that patients reported fewer executive functioning
et  al. (2011) conducted a study that explored the difficulties after CRT although this was not corrob-
possible benefit of CRT delivered in a group setting orated by parental reports. A  case study described

Harrison 399
by van Noort, Pfeiffer, Lehmkuhl, and Kappel a potential means of helping adolescent patients to
(2015) also explored the possible benefits of 10 engage in therapy.
CRT sessions offered twice weekly to a 12-​year-​old Taken together, the evidence reviewed above
girl with a severe form of AN over 5 weeks. There suggests that exploring the possible benefits of CRT
were no clear changes in neuropsychological func- for younger cohorts of individuals with EDs will be
tioning, but CRT appeared to be associated with an important focus for future research.
an improvement in AN symptoms 7 months after Thus far, this chapter has explored the develop-
CRT. While it is possible that other factors influ- ment of CRT and its evidence base for improving
enced this symptom reduction, the authors suggest the cognitive inefficiencies reported in individuals
that CRT may have played an important supple- with EDs. The next section explores how CRT can
mentary role in the context of the broader treatment be implemented in a treatment setting.
offered. Van Noort, Kraus, Pfeiffer, Lehmkuhl, and
Kappel (2016) offered CRT to a mixed sample of Ways of Using Cognitive Remediation
20 inpatients and outpatients with AN, conducting in Clinical Settings
a neuropsychological assessment before and after Implementing CRT in a clinical service for
treatment. They also conducted the same assess- people with EDs can be approached in a num-
ment in a group of 20 non-​AN controls who did ber of ways (Tchanturia & Lock, 2011). The lit-
not receive CRT. There was an improvement in erature reviewed here highlights that one priority
cognitive flexibility in the AN group after treatment for future research would be to conduct RCTs in
relative to the control group, although there was no an adolescent population, and in this case, it is rec-
change in either group regarding central coherence ommended that a manualized approach be taken
or self-​reported cognitive flexibility. to allow clinicians and researchers to systemati-
These initial studies in adolescent popula- cally evaluate the efficacy of CRT with a focus on
tions appear to suggest that CRT may be feasible neuropsychological performance and ED symp-
to implement and experienced as acceptable by tom outcome. The original CRT treatment manual
patients; CRT may be associated with relatively (Tchanturia et  al., 2010) can be accessed freely
low levels of dropout in this younger patient group in the “Publications” section located at the web
and may have some impact on patients’ neuropsy- address www.katetchanturia.com, and there are
chological functioning. Indeed, these promising Spanish, Italian, French, and Japanese translations
early findings have led clinical research teams to available at the same website. Readers may find
begin to embed CRT into their treatment pro- the discussion in Tchanturia and Doris (2015)
grams for young people with EDs. For example, useful regarding adapting the delivery to a group
Doyen et al. (2015) proposed that a 6-​month pro- setting for adults with EDs. Other research groups
gram of CRT should be offered to all adolescent have also provided useful toolkits. Lindvall,
inpatients with AN by the nursing team. In their Owen, and Lask (2011) provide a CRT Resource
manuscript, they also present an uncontrolled case Pack adapted for children and adolescents with
series suggestive of improvements in cognitive EDs, which can be accessed freely at http://​ous.
functioning after, compared with before, treat- prod.fpl.nhn.no/ ​ S iteCollectionDocuments/​
ment. However, at present, it is unclear the degree Fagfolk/​Forskning%20og%20utvikling/​R ASP/​
to which CRT enhances treatment as usual, and 1.-​7 .%20The%20CRT%20Recource%20Pack.
randomized designs would be needed to explore pdf. Maiden, Baker, Espie, Simic, and Tchanturia
this further. (2014) have adapted the materials for group
Interestingly, CRT for EDs has continued to delivery in the child and adolescent population;
be developed in novel ways. Lang, Treasure, and known as the “Flexibility Group,” the resource
Tchanturia (2015) present preliminary findings can be accessed freely within the “Publications”
which suggest that a Web-​ based CRT self-​ help section at the website www.katetchanturia.com,
intervention disseminated to carers might be an with further information discussed in Maiden,
acceptable form of treatment and provide a means Baker, Espie, Simic, and Tchanturia (2015). At
through which patients and their families might be the same website, those interested in the adapta-
better able to understand thinking styles that may tion for carers of people with EDs discussed above
maintain EDs. Similarly, a family-​based face-​to-​face will also be able to access the materials adapted to
format, as described in Hutchinson, Roberts, and using CRT with this population (Tchanturia &
Lask (2014), was experienced as acceptable and was Lang, 2015).

400 Cognitive Remediation

 The other way in which implementation can to the individual. The session should aim to explore
be approached is through offering CRT as part of at least one exercise, which may also involve the
the broader treatment package available for indi- therapist taking part, and then a discussion around
viduals with AN and the evidence reviewed thus the skills and strategies used should follow. In this
far lends support for the use of CRT as a treatment meta-​cognitive exploration, the therapist should ask
enhancer for adults with AN, with particular ben- open questions, taking a curious and nonjudgmen-
efits in terms of keeping patients in treatment and tal stance, such as “How did you go about com-
helping them to develop useful skills for later higher pleting that task?” or “What strategy did you find
intensity therapeutic work. In this sense, CRT can yourself using here?” The clinician should then ask
be adapted to the treatment setting with regard to about whether there are any other ways in which
who delivers the treatment, how it is delivered and the exercise could be completed, and what strengths
in what format (individual or group setting). In an and weaknesses alternative strategies might have.
inpatient setting, CRT could be offered on admis- The clinician should then ask about how these strat-
sion to the ward as a way to help the patient to egies might relate to real-​life tasks and whether the
settle in and start to engage in the milieu in a non- patient has noticed themselves using these strategies
threatening way, with a focus on skill development and how this impacted their ability to complete the
and strengths rather than eating, weight, and shape task. The clinician and the patient could then think
(Tchanturia, 2014). It will be important to reflect about whether there might be an individualized
on how CRT fits within the overall formulation behavioral experiment that would help the patient
(Tchanturia & Lock, 2011). There is no clear evi- to build confidence around using alternative strate-
dence to suggest whether a group setting or an indi- gies (Tchanturia & Lock, 2011). A clear example of
vidual setting is more beneficial, and clinicians will this exchange is offered in the next section.
make an informed decision around how to begin Consider a patient and therapist working on a
implementing CRT based on the context of their global versus detailed thinking exercise in which the
setting and service, the patients themselves, and patient and therapist both view a photo of a com-
the resources available. However, CRT is designed plex scene, such as a busy tourist area in the city.
to be a low-​intensity treatment that can be offered The instruction is to explore the picture and write
by members of the nursing team and unqualified an account of what is going on. The therapist takes
psychology graduates, including assistant psycholo- part, writing their own account at the same time as
gists, research assistants, and trainee clinicians with the patient. After the account is written the thera-
its delivery supervised by a qualified and licensed pist and the patient share their observations. The
psychologist (Tchanturia, 2014). As with any treat- therapist can then ask about the approach taken
ment package, good supervision is vital. The sec- by the patient around completing the task and the
tion below provides more detail of how CRT can strengths and limitations of that approach. For
be delivered. example, if the patient has written a long, detailed
account of the scene, the therapist could ask about
Outline of a Typical Cognitive Remediation what the pros and cons of the strategy might be and
Therapy Session whether there might be another way to approach the
The exact session length and dose for efficacy task. If the therapist has written a briefer account,
are not currently known, but the majority of stud- giving the gist rather than the detail, the therapist
ies have focused on offering ten 45-​minute ses- and patient could explore the pros and cons of this
sions (Tchanturia et  al., 2013), although this can strategy. The therapist might then ask whether the
be adapted depending on the patients’ abilities and patient has found themselves using the same strat-
needs at the time of treatment. At the start of each egy in everyday tasks and whether there might be
session, the clinician and patient(s) should set up an alternative approaches that they could consider. It
agenda and outline the exercises they will cover and may then be possible to set up an experiment to
the skills they aim to focus on. The number of exer- explore the use of an alternative strategy in an every-
cises the therapist brings to the session should be day task. If the patient is able and willing to, and
based on how long the patient is able to concentrate there is time left in the allocated session time, then
depending on their physical state and the amount the patient and therapist might consider moving on
of time the patient requires to complete the tasks to look at another task or exercise.
and enter into a discussion about the cognitive skills As previously discussed, treatment manuals and
and strategies explored and this will vary according comprehensive toolkits are readily available, and in

Harrison 401
addition to this, practitioners using CRT in rou- they have a tendency to focus on detail when pro-
tine work are also encouraged to develop and use cessing information. This measure has demonstrated
additional tasks that they think will provide a use- strong discriminant validity with large effect sizes for
ful talking point to think about thinking with their the differences between those with EDs and non-​ED
patients. In order to assist this discussion, Table controls (Roberts et al., 2011) and the authors of the
20.2 provides an outline of a range of tasks focused measure suggest that this scale is able to provide an
on improving functioning in the domains of central estimation of the degree to which individuals func-
coherence, set-​shifting, and estimation skills. tion in terms of detail focused information process-
For further examples, readers are encouraged ing and set-​shifting skills as manifested in the context
to refer to the published manuals and toolkits dis- of daily life. Arguably this measure only reflects one
cussed previously. aspect of central coherence—​detailed information
processing—​ and therefore may be able to offer
Measurement of Outcome less information regarding the extent to which the
As with all treatment interventions, it is impor- individual is able to use bigger-​ picture thinking
tant to investigate the potential impact of treat- skills. However, like the Cognitive Flexibility Scale,
ment on a range of outcomes such as symptoms, it offers a brief, low-​cost means of exploring these
functioning, and quality of life. A range of different aspects of functioning before and after the delivery
measures have been used to investigate the extent CRT and could be sent to patients to complete at
to which CRT impacts cognitive functioning in a follow-​up time point with little cost and incon-
people with EDs, and this section aims to provide venience to the patient themselves. In Tchanturia
a brief overview of the self-​report and experimen- et al.’s (2016) study that reported on outcomes from
tal measures previously used in research on CRT 20 CRT groups delivered in inpatient and intensive
for ED and found to be sensitive to change, which daycare settings, this measure was more sensitive to
clinicians and researchers may find useful when change than the Cognitive Flexibility Scale, dem-
exploring the possible impact of CRT on cognitive onstrating significant, small-​sized improvements in
functioning. self-​reported set-​shifting (d = 0.36) and a reduction
in the detail-​ focused information processing style
Self-​Report Measures of Outcome (d = 0.37) at the end of the groups.
The Cognitive Flexibility Scale (Martin & Rubin, A disadvantage of relying on self-​report measures
1995) is a 12-​item questionnaire that assesses par- may be that participants require reasonable insight
ticipants’ perceptions of their set-​ shifting skills, into their own cognitive functioning to be able to
namely the options and alternatives they feel are provide accurate answers and may also be subject
available to them in everyday situations. Possible to demand characteristics and other biases when
scores range between 0 and 72, and a higher score is completing the measures. Given that CRT aims to
related to a higher level of cognitive flexibility. The improve neuropsychological functioning, clinicians
measure has good test–​retest reliability (Martin & and researchers measuring CRT outcomes in ED
Rubin, 1995) and benefits from being a brief assess- populations are encouraged to also consider using
ment tool requiring little cost in terms of time and experimental measures where possible, of which a
material or equipment resources. This measure was range of suitable options are discussed below. It is
used in a study exploring the impact of 20 4-​to strongly advised when conducting any neuropsy-
6-​session CRT groups delivered to adult women chological assessment that aims to explore the possi-
with AN accessing inpatient or intensive daycare ble impact of CRT on cognitive functioning to also
services and was somewhat sensitive to change collect an estimate of the patient’s intelligence quo-
over time, highlighting a negligible sized improve- tient (IQ) at the start of treatment, perhaps using
ment (d  =  0.18) in self-​reported set-​shifting skills the full or abbreviated version of the Wechsler Adult
(Tchanturia, Larsson, & Brown, 2016) at the end Intelligence Scale (WAIS-​R; Wechsler, 1981) or the
of the groups. National Adult Reading Test (Nelson & Willison,
The Detail and Flexibility Questionnaire (DFlex; 1991) which is positively correlated (r = 0.81) with
Roberts, Barthel, Lopez, Tchanturia, & Treasure, the WAIS-​R (Crawford, Parker, Stewart, Besson, &
2011) is another useful measure to consider. This is a De Lacey, 1989). This will allow an understanding
24-​item self-​report scale that not only measures self-​ of the extent of any inefficiencies in cognitive func-
reported set-​shifting but also offers a subscale that tioning relative to general intellectual functioning
measures the degree to which the individual feels before CRT commences.

402 Cognitive Remediation

Table 20.2  Examples of Cognitive Remediation Therapy Exercises, Follow-​Up Questions Designed to Promote
“Thinking About Thinking” and Ideas for Homework
Task Domain Exercise Example Follow-​Up Questions Homework Suggestions

Set-​Shifting

Stroop task The patient is asked to How did you approach this task? Inhibiting unhelpful
switch between reading the Which strategies did you find yourself information when
color of the ink and the using to complete it? accessing social media.
word written. Does this task remind you of any tasks you
have to do in everyday life?

Clocks task The patient is asked to How hard was it to complete this task? Changing the radio
switch between giving the What strategies did you find yourself using station, genre of music
time on a series of clocks when having to make the switch? you listen to, or switching
using AM/​PM and the 24 Are there any other ways that you could to a different background
hour clock have approached the task? or ring tone on your
Is the switching part of the task something mobile phone.
that you find difficult in everyday life? If so, Taking a different route
can you give some examples? to work or school, sitting
in a different seat on the
ward.

Central Coherence—​Promoting “Bigger-​Picture” Thinking

Describing The patient and the How did you approach this task—​what did Practicing describing your
a complex therapist write a description you notice about your answer? day using one sentence;
scene of a photo of a busy scene What are the other ways in which you Reading an article and
might have approached this task? summarizing it using a
What are the strengths and weaknesses of headline.
your approach and the alternatives you
have mentioned?
Have you noticed any other areas in your
life where you use this approach? How
might this help or hinder you? Are there
any alternatives?

Summarizing The patient and therapist How did you approach the task and were Practicing summarizing a
a letter or work on a summary of there any parts you found challenging? film/​movie in a sentence.
news article the gist of a longer letter What are the strengths and weaknesses of
or news article, perhaps in bigger-​picture thinking?
bullet points, as a “Tweet” How might you be able to incorporate
or short text message bigger-​picture thinking into your daily life?
like those sent on mobile
phones, or as a headline.

Geometric The patient describes a How did you go about approaching the Practicing describing an
figures complex line drawing to description of the drawing? event to a friend or family
the therapist, who is blind What did you notice about your thinking member through focusing
to the image and must styles during this task? on providing the gist;
reproduce it based on the Where there any challenges you faced writing down the main
patient’s verbal description. during the task? message of your week in
What might an alternative approach be? your diary.

(continued)

Harrison 403
Table 20.2 Continued
Task Domain Exercise Example Follow-​Up Questions Homework Suggestions

Estimation

Line The patient is asked to Leaving things unfinished

bisection bisect a horizontal line at or stopping when the task
various points (for example, is “just good enough.”
at 25%, 50%, 75%, 90%). Throwing something away
that you would ordinarily
save “just in case.”
Handing in some work
that is not to your usual
high standard.

Source: Adapted from Tchanturia and Hambrook (2010).

Experimental Measures of Outcome outpatients and inpatients, respectively, compared

Set-​Shifting with non-​ED controls. Participants must predict the
A widely used measure of set-​shifting in the field movement of a blue circle across 10 locations. The
of EDs is the Wisconsin Card Sort Test (WCST) movement of the blue circle forms a pattern, which
(Grant & Berg, 1948; Heaton et  al., 1993). subsequently changes throughout the task. The par-
Tchanturia et  al. (2012) reported that across a ticipant must therefore adjust their responses in line
sample of 542 participants, the measure detected with the changing pattern. The outcome measure
medium to large-​ sized differences between indi- indicative of set shifting ability is the total number
viduals with EDs (both AN and BN) and non-​ED of errors that are perseverative in nature. That is to
controls, with recovered participants showing an say, those errors committed by the participant where
intermediate profile. This task measures cognitive or they have continued with a rule despite the occur-
conceptual flexibility and is also a measure of reac- rence of a new pattern. Tchanturia, Brecelj, et  al.
tive flexibility; it requires participants to sort cards (2004); Tchanturia, Morris, Brecelj, Nikolau, &
into one of four categories. The participant must Treasure (2004); and Holliday, Tchanturia, Landau,
first discover the sorting rule, which will subse- Collier, and Treasure (2005) describe examples of
quently change during the task. The most common studies that have used this measure with people with
outcome measure used is the number of persevera- EDs, and Roberts et al. (2007) in their systematic
tive errors committed during the task, that is to review reported that this task detects a small-​sized
say the number of times the participant continues difference (d = 0.21) between those with EDs and
applying a sorting rule despite the category having non-​ED controls, with people with EDs making a
changed. In the ED population, Steinglass, Walsh, greater number of errors than non-​ED controls.
and Stern (2006); Ohrmann et al. (2004); Fassino The Trail Making Task (Kravariti, Morris,
et  al. (2002); and Tchanturia et  al. (2012) are Rabe-​Hesketh, Murray, & Frangou, 2003; Reitan,
examples of studies that have employed this mea- 1958) is an additional measure that has been used
sure. Westwood et  al.’s (2016) systematic review across the ED literature. This task measures cog-
and meta-​analysis revealed a large-​sized difference nitive or conceptual flexibility; participants are
in performance between those with EDs and non-​ required to connect a sequence of dots over three
ED controls (d = 0.52). trials. The first trial is a control trial, in which the
An alternative task that might be useful because participant connects unlabeled dots. The second
of its frequent use in studies exploring CRT for EDs trial uses dots labeled with alphabetical informa-
is the Brixton Test (Burgess & Shallice, 1997), which tion, and the participant connects the dots in
also assesses the cognitive or conceptual domain of alphabetical order. The third trial, known as Trail
set-​shifting. The ability of this task to detect diffi- B involves an alphanumeric switching component
culties with set-​shifting in those with EDs was sup- that requires the participants to connect the labeled
ported by Tchanturia et  al. (2011) whose sample dots in a number, letter, number, letter sequence;
of 601 participants highlighted medium-​to large-​ for example, 1-​A-​2-​B-​3-​C and so on. The outcome
sized differences in performance on this measure for measure indicative of set-​shifting ability is the time

404 Cognitive Remediation

taken to complete Trail B. Steinglass et al. (2006); taken to correctly locate the hidden shape. A shorter
Holliday et  al. (2005); Tchanturia, Brecelj, et  al. response latency indicates superior performance at
(2004); Tchanturia, Morris, et al. (2004); Murphy, detail processing. Lang et al. (2014), in their meta-​
Nutzinger, Paul, and Leplow (2002); Mathias and analysis, found this task detected medium-​sized dif-
Kent (1998); and Kingston, Szmukler, Andrewes, ferences in performance between those with EDs
Tress, and Desmond (1996) describe examples and non-​ED controls (d  =  0.62), with individuals
where this measure has been used to explore set-​ with EDs demonstrating superior detail-​oriented
shifting in the ED population. In Roberts et  al.’s information-​ processing skills relative to non-​ ED
(2007) systematic review and meta-​analysis, these controls.
data equated to an effect size of 0.38, with ED par- An alternative detail-​ oriented task is the
ticipants taking significantly longer to complete Matching Familiar Figures Task (Kagan, Rosman,
Trial B than non-​ED controls. Day, Albert, & Phillips, 1964). This is a visual per-
ceptual test, which was originally designed to mea-
Central Coherence sure cognitive impulsivity (Kagan et al., 1964), has
This section is broken down into tasks benefiting been used to highlight the strength for detailed
from global information processing strategies and information processing in individuals with EDs
tasks benefiting from detailed-​focused information (Southgate, Tchanturia, & Treasure, 2008). The
processing strategies. Ideally, in any exploration of task consists of 12 items, and the participant is
the possible impact of CRT, it is useful to include required to identify the exact replica of a familiar
measures that explore both components of central object (e.g., a lion) among eight highly similar alter-
coherence. natives. The outcome measure is the time taken to
Object Assembly (Wechsler, 1981) is an example correctly identify the identical figure. The number
of a task that benefits from global information-​ of incorrect responses may also be recorded. A more
processing skills and is a performance task from the detail-​
focused cognitive style is associated with
Wechsler Intelligence Scale (Wechsler, 1981). The reduced response latencies and fewer errors. There
participant is required to solve small jigsaw-​type were too few studies available to carry out a meta-​
puzzles depicting familiar objects, such as a butter- analysis of this task in Lopez, Tchanturia, Stahl, and
fly or a horse. The outcome measures are the time Treasure’s (2008) systematic review, however they
taken to complete the puzzles and the total num- reported effect sizes from 0.41 to 0.89 for those
ber of puzzles completed within a designated time. with EDs relative to non-​ED controls. No new uses
A shorter time suggests that the participant is able of this task in ED populations have been reported
to create an integrated global representation of an since this review (Lang et al., 2014).
item from its individual parts (Tokley & Kemps, The Rey-​ Osterrieth Complex Figure Test
2007), which is why this can be a useful outcome (Osterreith, 1944) is one of the most widely used
measure regarding global processing strategies. Lang measures in the ED research literature that has been
et al. (2014) found this task detected differences in used to explore central coherence, and this task
central coherence between those with EDs and non-​ arguably explores the individuals’ tendency toward
ED controls, with a medium effect size (d = 0.65). a detailed or global processing strategy because it is
The Fragmented Pictures Task (Snodgrass, Smith, possible to explore how the patient, through copying
Feenan, & Corwin, 1987) is an alternative globally a complex figure, builds up the figure and whether
oriented task that might be useful to consider and they tend toward a more detail-​oriented or a more
this has shown to discriminate between those with global-​oriented approach. In previous work, this
EDs and non-​ED controls (Harrison et al., 2012). figure has been used to explore a variety of cognitive
The (Group) Embedded Figures Task, or (G) processes, including visual perceptual organization,
EFT (Witkin, 1971; Witkin, Oltman, Raskin, and nonverbal memory, planning, problem solving,
Karp, 2002), is a task that could be described as ben- and motor function (Osterrieth, 1944). There are
efiting from detail-​oriented information-​processing a variety of methods available for the administra-
skills and has been widely used to explore cognitive tion and scoring of the task. Some researchers use
functioning in EDs. This task can either be admin- the copy administration only, whereas others use a
istrated individually (EFT), or in a group (GEFT). recall administration, which requires participants
The task requires the participant to identify simple without prior warning to recall as much of 75% of
shapes that are hidden within more complex shapes. the picture as possible after a designated interval,
The outcome measure is the number of seconds which may vary from 3 to 60 minutes. Regarding

Harrison 405
the measurement of coherence, the central coher- has been demonstrated to help retain patients in
ence index is relevant. This index is based on the treatment programs (Lock et  al., 2013), its pos-
order in which the participant copies the figure, and sible impacts on motivation are not fully under-
on style—​that is, whether the figure was copied in stood (Danner, Dingemans, & Steinglass, 2015)
a fragmented or continuous manner. A higher cen- and future work should aim to explore this. Finally,
tral coherence index relates to a more global strat- although the RCTs in the adult population high-
egy. Lang et al. (2014) found this measure detects light improved neuropsychological functioning,
medium-​sized differences in performance between it will be important to explore how this manifests
those with EDs and non-​ED controls (d = 0.63). in improved recovery rates, that is, to what degree
A number of factors will determine the battery do greater cognitive flexibility and bigger-​picture
of tasks used to evaluate potential changes in cog- thinking contribute to overall recovery?
nitive functioning after, compared to before, CRT
and at follow-​up for individuals with EDs including Conclusions
patient burden, service resources, and financial limi- To summarize, CRT is a low-​intensity training
tations and the sensitivity of the task. However, it is package that aims to assist individuals with EDs to
hoped that the section above has provided sufficient build on their cognitive abilities to support them to
information to inform these decisions for clinical remain in treatment programs and develop the flex-
research groups interested in using and evaluating ible and bigger-​picture thinking skills that may be
CRT for EDs. The final section explores priorities necessary to move toward recovery. Cognitive reme-
for future research. diation therapy can be a useful treatment enhancer
provided in an individual or group setting for adults
Future Work and young people with EDs, and four RCTs in the
Previous work highlights a paucity of evidence adult population and numerous case series in the
for the possible benefits of CRT for individuals child and adolescent population have demonstrated
with BN and this may be a focus of future research, its positive impact on neuropsychological outcomes.
given that individuals with BN have also demon-
strated inefficiencies in set-​shifting (Harrison et al., References
2012; Roberts, Tchanturia, & Treasure, 2010; Wu Asch, M., Esteves, J., De Hautecloque, D., Bargiacchi, A.,
Le Heuzey, M. F., Mouren, M. C., & Doyen, C. (2014).
et  al., 2014) and the positive early findings from Cognitive remediation therapy for children and adoles-
Dingemans et  al. (2014) RCT, whose mixed ED cents with anorexia nervosa in France: an exploratory study.
sample included individuals with bulimic symptom L’Encephale, 40, 240–​246.
profiles. The literature reviewed in this chapter indi- Blinder, B. J., Cumella, E. J., & Sanathara, V. A. (2006).
cates that CRT may be of possible benefit for child- Psychiatric comorbidities of female inpatients with eating
disorders. Psychosomatic Medicine, 68, 454–​462.
ren and adolescents with EDs, and a priority for Brockmeyer, T., Ingenerf, K., Walther, S., Wild, B., Hartmann,
future work will be to conduct an RCT in this con- M., Herzog, W.,   .  .  .  Friederich, H. C. (2014). Training
text. It would also be interesting to explore whether cognitive flexibility in patients with anorexia nervosa: a pilot
group or individual formats are most efficient in randomized controlled trial of cognitive remediation therapy.
improving cognitive functioning in people with International Journal of Eating Disorders, 47, 24–​31.
Burgess, P. W., & Shallice, T. (1997). The Hayling and Brixton
EDs and what “dose” of treatment is required for tests. Bury St Edmonds, UK: Thames Valley Test Company.
a clinically significant improvement because these Chui, H. T., Christensen, B. K., Zipursky, R. B., Richards,
questions are currently unanswered by the litera- B. A., Hanratty, M. K., Kabani, N. J.,   .  .  .  & Katzman,
ture. Furthermore, it will be important to under- D. K. (2008). Cognitive function and brain structure in
stand how long lasting any possible improvements females with a history of adolescent-​onset anorexia nervosa.
Pediatrics, 122, e426–​e437.
in cognitive functioning are by conducting longer Cicerone, K. D., Langenbahn, D. M., Braden, C., Malec, J. F.,
term follow-​up studies. In addition to this, it will Kalmar, K., Fraas, M.,  . . .  Ashman, T. (2011). Evidence-​
be interesting to explore whether moving from the based cognitive rehabilitation: Updated review of the litera-
“cold cognitive” domain to the “hot cognitive” ture from 2003 through 2008. Archives of Physical Medicine
domain and including exercises around emotions and Rehabilitation, 92, 519–​530.
Crawford, J. R., Parker, D. M., Stewart, L. E., Besson, J. A. O., &
and social functioning might be another impor- Lacey, G. (1989). Prediction of WAIS IQ with the National
tant way of further developing the scope of CRT Adult Reading Test: Cross‐validation and extension. British
as a treatment enhancer that coaches people with Journal of Clinical Psychology, 28, 267–​273.
EDs in the skills likely to be useful for successful Cwojdzińska, A., Markowska-​ Regulska, K., & Rybakowski,
recovery (Tchanturia et al., 2013). Although CRT F. (2009). [Cognitive remediation therapy in adolescent

406 Cognitive Remediation

 anorexia nervosa—​ case report]. Psychiatria Polska, 43, Grant, D. A., & Berg, E. (1948). A behavioral analysis of degree
115–​124. of reinforcement and ease of shifting to new responses in a
Dahlgren, C. L., Lask, B., Landrø, N. I., & Rø, Ø. (2013). Weigl-​type card-​sorting problem. Journal of Experimental
Neuropsychological functioning in adolescents with Psychology, 38, 404.
anorexia nervosa before and after cognitive remediation Happé, F., & Frith, U. (2006). The weak coherence account: Detail-​
therapy:  A  feasibility trial. International Journal of Eating focused cognitive style in autism spectrum disorders. Journal of
Disorders, 46, 576–​581. Autism and Developmental Disorders, 36, 5–​25.
Dahlgren, C. L., Lask, B., Landrø, N. I., & Rø, Ø. (2014). Harrison, A., Tchanturia, K., Naumann, U., & Treasure, J.
Developing and evaluating cognitive remediation therapy (2012). Social emotional functioning and cognitive styles
(CRT) for adolescents with anorexia nervosa:  A  feasibility in eating disorders. British Journal of Clinical Psychology, 51,
study. Clinical Child Psychology and Psychiatry, 19, 476–​487. 261–​279.
Dahlgren, C. L., & Rø, Ø. (2014). A systematic review of cogni- Heaton, R. K., Chelune, G. J., Talley, J. L., Kay, G. G., & Curtiss,
tive remediation therapy for anorexia nervosa-​development, G. (1993). Wisconsin card sort test manual:  Revised and
current state and implications for future research and clinical expanded. Odessa, FL: Psychological Assessment Resources.
practice. Journal of Eating Disorders, 2, 1. Holliday, J., Tchanturia, K., Landau, S., Collier, D., & Treasure,
Danner, U. N., Dingemans, A. E., & Steinglass, J. (2015). J. (2005). Is impaired set-​ shifting an endophenotype of
Cognitive remediation therapy for eating disorders. Current anorexia nervosa? American Journal of Psychiatry, 162,
Opinion in Psychiatry, 28, 468–​472. 2269–​2275.
Dare C, & Eisler I. (1997). Family therapy for anorexia ner- Hutchison, S., Roberts, A., & Lask, B. (2014). Family cogni-
vosa. In:  D. M. Garner, Garfinkel P, (Eds.), Handbook tive remediation therapy for child and adolescent anorexia
of Treatment for Eating Disorders (pp. 307–​324). nervosa. In K. Tchanturia (Ed.), Cognitive remediation
New York: Guilford Press. therapy (CRT) for eating and weight disorders (pp. 147–​161).
Davies, H., & Tchanturia, K. (2005). Cognitive remediation London, UK: Routledge.
therapy as an intervention for acute anorexia nervosa: A case Kagan, J., Rosman, B. L., Day, D., Albert, J., & Phillips, W.
report. European Eating Disorders Review, 13, 311–​316. (1964). Information processing in the child:  Significance
Dingemans, A. E., Danner, U. N., Donker, J. M., Aardoom, J. of analytic and reflective attitudes. Psychological
J., Van Meer, F., Tobias, K.,  . . .  Van Furth, E. F. (2014). Monographs, 78, 1.
The effectiveness of cognitive remediation therapy in patients Kaye, W. H., Bulik, C. M., Thornton, L., Barbarich, N., &
with a severe or enduring eating disorder: A randomized con- Masters, K. (2004). Comorbidity of anxiety disorders with
trolled trial. Psychotherapy and Psychosomatics, 83, 29–​36. anorexia and bulimia nervosa. American Journal of Psychiatry,
Douglas, K. M., Porter, R. J., Knight, R. G., & Maruff, P. (2011). 161, 2215–​2221.
Neuropsychological changes and treatment response in severe Keel, P. K., & Mitchell, J. E. (1997). Outcome in bulimia ner-
depression. British Journal of Psychiatry, 198, 115–​122. vosa. American Journal of Psychiatry, 154, 313.
Doyen, C., Contejean, Y., Risler, V., Asch, M., Amado, I., Kingston, K., Szmukler, G., Andrewes, D., Tress, B., &
Launay, C.,   .  .  .  Kaye, K. (2015). Cognitive remediation Desmond, P. (1996). Neuropsychological and structural
therapy for children: Literature data and clinical application brain changes in anorexia nervosa before and after refeeding.
in a child and adolescent psychiatry department. Archives de Psychological Medicine, 26, 15–​28.
Pediatrie, 22, 418–​426. Kravariti, E., Morris, R. G., Rabe-​Hesketh, S., Murray, R. M., &
Easter, A., & Tchanturia, K. (2011). Therapists’ experi- Frangou, S. (2003). The Maudsley early onset schizophrenia
ences of cognitive remediation therapy for anorexia ner- study:  Cognitive function in adolescents with recent onset
vosa:  Implications for working with adolescents. Clinical schizophrenia. Schizophrenia Research, 61, 137–​148.
Child Psychology and Psychiatry, 16, 233–​246. Lang, K., Lopez, C., Stahl, D., Tchanturia, K., & Treasure, J.
Fairburn, C. G. (2008). Cognitive behavior therapy and eating dis- (2014). Central coherence in eating disorders:  An updated
orders. New York, NY: Guilford Press. systematic review and meta-​ analysis. World Journal of
Fairburn, C. G., Bailey-​Straebler, S., Basden, S., Doll, H. A., Biological Psychiatry, 15, 586–​598.
Jones, R., Murphy, R.,  . . .  Cooper, Z. (2015). A transdiag- Lang, K., Stahl, D., Espie, J., Treasure, J., & Tchanturia,
nostic comparison of enhanced cognitive behaviour therapy K. (2014). Set shifting in children and adolescents with
(CBT-​E) and interpersonal psychotherapy in the treatment of anorexia nervosa:  An exploratory systematic review and
eating disorders. Behaviour Research and Therapy, 70, 64–​71. meta‐analysis. International Journal of Eating Disorders, 47,
Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive 394–​399.
behaviour therapy for eating disorders:  A  “transdiagnostic” Lang, K., & Tchanturia, K. (2014). A systematic review of
theory and treatment. Behaviour Research and Therapy, 41, central coherence in young people with anorexia nervosa.
509–​528. Journal of Child and Adolescent Behaviour, 2, 140.
Fassino, S., Pieró, A., Daga, G. A., Leombruni, P., Mortara, P., Lang, K., Treasure, J., & Tchanturia, K. (2015). Acceptability
& Rovera, G. G. (2002). Attentional biases and frontal func- and feasibility of self-​ help cognitive remediation therapy
tioning in anorexia nervosa. International Journal of Eating for anorexia nervosa delivered in collaboration with car-
Disorders, 31, 274–​283. ers:  A  qualitative preliminary evaluation study. Psychiatry
Genders, R., & Tchanturia, K. (2010). Cognitive remediation Research, 225, 387–​394.
therapy (CRT) for anorexia in group format: A pilot study. Lezak, M. D. (2004). Neuropsychological assessment. Oxford,
Eating and Weight Disorders, 15, 234–​239. UK: Oxford University Press.
Gioia, G. A., Isquith, P. K., Guy, S. C., & Kenworthy, L. (2000). Lindvall, C., Owen, I., & Lask, B. (2011). Cognitive remediation
Behavior rating inventory of executive function:  BRIEF. therapy (CRT) for children and adolescents with anorexia ner-
Odessa, FL: Psychological Assessment Resources. vosa: Resource pack. Oslo, Norway: University of Oslo.

Harrison 407
Lock, J., Agras, W. S., Fitzpatrick, K. K., Bryson, S. W., Jo, B., Pretorius, N., Dimmer, M., Power, E., Eisler, I., Simic, M.,
& Tchanturia, K. (2013). Is outpatient cognitive remedia- & Tchanturia, K. (2012). Evaluation of a cognitive reme-
tion therapy feasible to use in randomized clinical trials for diation therapy group for adolescents with anorexia ner-
anorexia nervosa? International Journal of Eating Disorders, vosa:  Pilot study. European Eating Disorders Review, 20,
46, 567–​575. 321–​325.
Lock, J., Le Grange, D., Agras, W. S., & Dare, C. (2001). Reitan, R. M. (1958). Validity of the Trail Making Test as an
Treatment manual for anorexia nervosa:  A  family-​ based indicator of organic brain damage. Perceptual and Motor
approach. New York, NY: Guildford Press. Skills, 8, 271–​276.
Lopez, C., Tchanturia, K., Stahl, D., & Treasure, J. (2008). Roberts, M. E., Barthel, F. M. S., Lopez, C., Tchanturia, K., &
Central coherence in eating disorders:  a systematic review. Treasure, J. L. (2011). Development and validation of the
Psychological Medicine, 38, 1393–​1404. detail and flexibility questionnaire (DFlex) in eating disor-
Lozano-​ Serra, E., Andrés-​Perpiña, S., Lázaro-​García, L., & ders. Eating Behaviors, 12, 168–​174.
Castro-​ Fornieles, J. (2014). Adolescent anorexia ner- Roberts, M. E., Tchanturia, K., Stahl, D., Southgate, L., &
vosa: Cognitive performance after weight recovery. Journal of Treasure, J. (2007). A systematic review and meta-​analysis of
Psychosomatic Research, 76, 6–​11. set-​shifting ability in eating disorders. Psychological Medicine,
Maiden, Z., Baker, L., Espie, J., Simic, M., & Tchanturia, K. 37, 1075–​1084.
(2014). Group cognitive remediation therapy for adolescents Roberts, M. E., Tchanturia, K., & Treasure, J. L. (2010).
with anorexia nervosa:  The flexible thinking group. London, Exploring the neurocognitive signature of poor set-​shifting
UK: South London and Maudsley NHS Foundation Trust. in anorexia and bulimia nervosa. Journal of Psychiatric
Maiden, Z., Baker, L., Espie, J., Simic, M., & Tchanturia, K. Research, 44, 964–​970.
(2015). Group cognitive remediation therapy format for Schmidt, U., & Treasure, J. (2006). Anorexia nervosa:  Valued
adolescents. In K. Tchanturia (Ed.), Brief group psycho- and visible; A  cognitive‐interpersonal maintenance model
therapy for eating disorders:  Inpatient protocols (pp. 51–​73). and its implications for research and practice. British Journal
Cambridge, UK: Routledge. of Clinical Psychology, 45, 343–​366.
Martin, M. M., & Rubin, R. B. (1995). A new measure of cogni- Snodgrass, J., Smith, B., Feenan, K., & Corwin, J. (1987).
tive flexibility. Psychological Reports, 76, 623–​626. Fragmenting pictures on the Apple Macintosh computer
Martinez, G., S. Cook-​Darzens, P. Chaste, Mouren, M. C., & for experimental and clinical applications. Behavior Research
Doyen, C. (2014). [Anorexia nervosa in the light of neuro- Methods, Instruments and Computers, 19, 270–​274.
cognitive functioning: New theoretical and therapeutic per- Southgate, L., Tchanturia, K., & Treasure, J. (2008). Information
spectives]. L’Encephale, 2, 160–​167. processing bias in anorexia nervosa. Psychiatry Research, 160,
Mathias, J. L., & Kent, P. S. (1998). Neuropsychological con- 221–​227.
sequences of extreme weight loss and dietary restriction Steinglass, J. E., Albano, A. M., Simpson, H. B., Wang, Y.,
in patients with anorexia nervosa. Journal of Clinical and Zou, J., Attia, E., & Walsh, B. T. (2014). Confronting fear
Experimental Neuropsychology, 20, 548–​564. using exposure and response prevention for anorexia ner-
Milos, G., Spindler, A., Schnyder, U., & Fairburn, C. G. (2005). vosa:  A  randomized controlled pilot study. International
Instability of eating disorder diagnoses:  prospective study. Journal of Eating Disorders, 47, 174–​180.
British Journal of Psychiatry, 187, 573–​578. Steinglass, J. E., Walsh, B. T., & Stern, Y. (2006). Set shift-
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., ing deficit in anorexia nervosa. Journal of the International
Howerter, A., & Wager, T. D. (2000). The unity and diver- Neuropsychological Society, 12, 431–​435.
sity of executive functions and their contributions to com- Tchanturia, K. (Ed.). (2014). Cognitive remediation therapy
plex “frontal lobe” tasks: A latent variable analysis. Cognitive (CRT) for eating and weight disorders. Hove, East Sussex,
Psychology, 41, 49–​100. UK: Routledge.
Murphy, R., Nutzinger, D. O., Paul, T., & Leplow, B. (2002). Tchanturia, K., Brecelj, M., Sanchez, P., Morris, R., Rabe-​
Dissociated conditional-​associative learning in anorexia ner- Hesketh, S., & Treasure, J. (2004). An examination of cogni-
vosa. Journal of Clinical and Experimental Neuropsychology, tive flexibility in eating disorders. Journal of the International
24, 176–​186. Neuropsychological society, 10, 513–​520.
Nelson, H. E., & Willison, J. (1991). National adult reading test Tchanturia, K., Davies, H., & Campbell, I. C. (2007).
(NART). Windsor: Nfer-​Nelson. Cognitive remediation therapy for patients with
Ohrmann, P., Kersting, A., Suslow, T., Lalee-​Mentzel, J., Donges, anorexia nervosa:  preliminary findings. Annals of General
U. S., Fiebich, M.,  . . .  Pfleiderer, B. (2004). Proton mag- Psychiatry, 6, 1.
netic resonance spectroscopy in anorexia nervosa:  correla- Tchanturia, K., Davies, H., Lopez, C., Schmidt, U., Treasure, J.,
tions with cognition. NeuroReport, 15, 549–​553. & Wykes, T. (2008). Letter to the editor: Neuropsychological
Osterrieth, P. A. (1944). The test of copying a complex fig- task performance before and after cognitive remediation in
ure: A contribution to the study of perception and memory. anorexia nervosa:  a pilot case-​series. Psychological Medicine,
Archives of Psychology, 30, 206–​356. 38, 1371–​1373.
Pitt, S., Lewis, R., Morgan, S., & Woodward, D. (2010). Tchanturia, K., Davies, H., Reeder, C., & Wykes, T. (2010).
Cognitive remediation therapy in an outpatient set- Cognitive remediation programme for anorexia ner-
ting: A case series. Eating and Weight Disorders, 15, 281–​286. vosa: A manual for practitioners. London, UK: South London
Porter, R. J., Bowie, C. R., Jordan, J., & Malhi, G. S. (2013). and Maudsley NHS Trust.
Cognitive remediation as a treatment for major depres- Tchanturia, K., Davies, H., Roberts, M., Harrison, A.,
sion: A rationale, review of evidence and recommendations Nakazato, M., Schmidt, U.,  . . .  Morris, R. (2012). Poor
for future research. Australian and New Zealand Journal of cognitive flexibility in eating disorders:  Examining the
Psychiatry, 47, 1165–​1175.

408 Cognitive Remediation

 evidence using the Wisconsin Card Sorting Task. PloS Treasure, J., & Schmidt, U. (2013). The cognitive-​interpersonal
One, 7, e28331. maintenance model of anorexia nervosa revisited:  A  sum-
Tchanturia, K., & Doris, E. (2015). Flexibility groups: Cognitive mary of the evidence for cognitive, socio-​ emotional and
remediation therapy (CRT) in group format:  adults. In interpersonal predisposing and perpetuating factors. Journal
K. Tchanturia. (Ed.), Brief group psychotherapy for eat- of Eating Disorders, 1, 1.
ing disorders:  Inpatient protocols (pp.  26–​50). Cambridge, van Noort, B. M., Kraus, M. K. A., Pfeiffer, E., Lehmkuhl, U.,
UK: Routledge. & Kappel, V. (2016). Neuropsychological and behavioural
Tchanturia, K., & Hambrook, D. (2010). Cognitive remedia- short‐term effects of cognitive remediation therapy in ado-
tion therapy for anorexia nervosa. In C. Grillo & J. Mitchell lescent anorexia nervosa:  A  pilot study. European Eating
(Eds.), The treatment of eating disorders: A clinical handbook. Disorders Review, 24, 69–​74.
(Vol. 7, pp. 130). New York, NY: Guildford Press. van Noort, B. M., Pfeiffer, E., Lehmkuhl, U., & Kappel, V.
Tchanturia, K., Harrison, A., Davies, H., Roberts, M., (2015). Cognitive remediation therapy for children with
Oldershaw, A., Nakazato, M.,   .  .  .  Treasure, J. (2011). anorexia nervosa. Zeitschrift für Kinder-​und Jugendpsychiatrie
Cognitive flexibility and clinical severity in eating disorders. und Psychotherapie, 43, 351–​355.
Plos One, 6, e20462. Wechsler, D. (1981). WAIS-​R manual:  Wechsler adult intel-
Tchanturia, K., & Lang, K. (2015). Cognitive remediation ther- ligence scale-​revised. New  York, NY:  Psychological
apy for anorexia nervosa: A self-​help guide for use with carers. Corporation.
London, UK: South London and Maudsley NHS Trust. Westwood, H., Stahl, D., Mandy, W., & Tchanturia, K. (2016).
Tchanturia, K., Larsson, E., & Brown, A. (2016). Benefits The set-​shifting profiles of anorexia nervosa and autism spec-
of group cognitive remediation therapy in anorexia ner- trum disorder using the Wisconsin Card Sorting Test: A sys-
vosa: Case series. Neuropsychiatrie, 30, 42–​49. tematic review and meta-​analysis. Psychological Medicine, 10,
Tchanturia, K., Lloyd, S., & Lang, K. (2013). Cognitive reme- S0033291716000581.
diation therapy for anorexia nervosa: Current evidence and Whitney, J., Easter, A., & Tchanturia, K. (2008). Service users’
future research directions. International Journal of Eating feedback on cognitive training in the treatment of anorexia
Disorders, 46, 492–​495. nervosa:  a qualitative study. International Journal of Eating
Tchanturia, K., & Lock, J. (2011). Cognitive remediation ther- Disorders, 41, 542–​550.
apy for eating disorders:  Development, refinement and Witkin, H. A. (1971). Group embedded figures test. Palo Alto,
future directions. In R. Adan & W. Kaye (Eds.), Behavioral CA: Consulting Psychologists Press.
neurobiology of eating disorders (pp. 269–​287). Berlin and Witkin, H. A., & Goodenough, D. R. (1977). Field dependence
Heidelberg, Germany: Springer. and interpersonal behavior. Psychological Bulletin, 84, 661.
Tchanturia, K., Lounes, N., & Holttum, S. (2014). Cognitive Witkin, H., Oltman, P., Raskin, E., & Karp, S. (2002).
remediation in anorexia nervosa and related condi- Group embedded figures test manual. Redwood City,
tions: A systematic review. European Eating Disorders Review, CA: Mindgarden.
22, 454–​462. Wood, L., Al-​Khairulla, H., & Lask, B. (2011). Group cog-
Tchanturia K., Morris, R., Brecelj, M., Nikolau, V., & Treasure, nitive remediation therapy for adolescents with anorexia
J. (2004). Set shifting in anorexia nervosa: An examination nervosa. Clinical Child Psychology and Psychiatry, 16,
before and after weight gain, in full recovery and the rela- 225–​231.
tionship to childhood and adult OCDP traits. Journal of Wu, M., Brockmeyer, T., Hartmann, M., Skunde, M., Herzog,
Psychiatric Research, 38, 545–​552. W., & Friederich, H. C. (2014). Set-​shifting ability across
Tchanturia, K., Whitney, J., & Treasure, J. (2006). Can cogni- the spectrum of eating disorders and in overweight and obe-
tive exercises help treat anorexia nervosa? Eating and Weight sity:  A  systematic review and meta-​ analysis. Psychological
Disorders, 11, e112–​e116. Medicine, 44, 3365–​3385.
Tokley, M., & Kemps, E. (2007). Preoccupation with detail con- Wykes, T., Huddy, V., Cellard, C., McGurk, S. R., & Czobor, P.
tributes to poor abstraction in women with anorexia nervosa. (2011). A meta-​analysis of cognitive remediation for schizo-
Journal of Clinical and Experimental Neuropsychology, 29, phrenia:  Methodology and effect sizes. American Journal of
734–​741. Psychiatry, 168, 472–​485.

Harrison 409
 C H A PT E R


Costs and Cost-​Effectiveness
21 in Eating Disorders

Scott J. Crow

Abstract
Costs and cost-​effectiveness are now well recognized as important aspects of the burdens of and
treatment for eating disorders. Ample evidence indicates that, leading up to and following diagnosis, the
cost burdens associated with eating disorders are high; this is true whether viewed from the perspective
of a healthcare payer or from a societal perspective. On the other hand, it is important to note that
studies involving cost modeling and direct cost collection in treatment have shown that treatment of
eating disorders is quite cost-​effective. Cost is now increasingly examined as an outcome in eating
disorder treatment trials. This remains an area of great need for further research.
Key Words:  anorexia nervosa, bulimia nervosa, burden, cost, cost-​effectiveness, eating disorder,
treatment

Introduction associated with treatment; direct financial costs

Healthcare costs receive particular scrutiny in ill- associated with symptoms of the illness, such as
nesses, such as eating disorders, that are perceived food for binge eating or substances used for purg-
to be expensive to treat, difficult to treat, or both. ing; and time costs associated with symptoms of
The costs of eating disorders (EDs) have now been these illnesses or their treatment. These aspects of
extensively described, and are clearly high, to indi- costs have been studied only rarely.
viduals, to third-​party payers, and to society. As
more effective treatments for eating disorders are Personal Costs
developed, increasing interest will be turned to To date, only one study has attempted to esti-
identifying treatments that are not only effective, mate the financial costs associated with bulimia
but also cost-​effective. nervosa (BN) symptoms (Crow et al., 2009). This
This chapter examines what is known about the study modeled annual food and purging-​ related
various costs burdens associated with EDs and their costs of bulimia symptoms in 10 subjects based on
treatment. It also examines the growing data on the 1-​week food records. Subjects in this study reported
cost-​effectiveness of specific treatments. Finally, it 2.5 objective binge episodes, 2.4 subjective binge
examines directions for future study in this area. episodes, and 3.6 purging episodes per week on
average, and spent an average of $5,582.00 (2007
Estimates of the Individual Costs of Eating US dollars) per year on food. Food and purging-​
Disorders related costs represented 32.7% of all food costs,
Having an ED may carry costs for the specific about $1,600 per year. However, these figures might
individual affected, and these may be a fairly large be viewed as underestimates, as frequency of binge
part of the overall cost burden associated with EDs. eating and purging was somewhat lower than aver-
Potential costs include the direct financial costs age for many studies of BN.

410
 A second study has made a limited attempt to Mean treatment costs for AN were significantly
measure time costs associated with BN (Crow higher than for both schizophrenia and OCD.
et al., 2013). In this multicenter BN randomized-​ Treatment costs for BN were found to be signifi-
controlled trial (described later in this chapter), cantly lower than schizophrenia, but significantly
time costs to family members related to partici- higher than OCD. The EDNOS treatment costs
pants with BN were measured in a subset of par- were significantly lower than schizophrenia and did
ticipants. A family member or significant other was not differ significantly from OCD.
asked to complete time-​monitoring records at entry Subsequently, Striegel-​ Moore and colleagues
into treatment and after the first 18 weeks of treat- (2005) examined prediagnosis and postdiagnosis
ment. The results showed that a substantial amount costs in people with an eating disorder. Costs rose
of time was lost to BN symptoms and their treat- substantially in the year preceding diagnosis, and
ment by family members/​significant others (about remained elevated in the year following diagnosis.
4 hours per week) and that this amount dropped A similar report from Mitchell et al. (2009) showed
by about 75% over the first 18 weeks of treatment. elevated costs (similar to those found in depression)
following ED diagnosis. In this sample, costs were
Per-​Patient Financial Costs not statistically significantly elevated in the year
Healthcare costs for those with eating disorders preceding diagnosis. This might be simply due to
are elevated. For example, data from the Medical a modest sample size, though, as numerical costs
Expenditure Panel Survey show that costs were doubled in the year preceding diagnosis, but this
$1,869 higher in those with EDs than in those with- change was not statistically significant.
out (Samnaliev, Noh, Sonneville, & Austin, 2015). Toulany and colleagues (2015) reported a study
Several studies have examined cost per patient of societal perspective costs of hospitalization for
from a third-​party-​payer perspective using health adolescent AN. The analysis used microcosting of
plan data. Striegel-​Moore, Leslie, Petrill, Garvin, data from 73 adolescents (a highly labor-​intensive
and Rosenheck (2000) accessed data through a US method). Mean length of stay was about 38  days,
insurance database (MarketScan) containing annual and mean cost was $54,392 (2015 Canadian dol-
inpatient and outpatient healthcare service use data lars). Body mass index (BMI) was inversely related
of individuals insured through large employers. The to costs: hospital costs declined 15.7% for every 1-​
above database is composed from privately insured point increase in admit BMI.
paid medical and prescription drug claims. A data Finally, Bellows and colleagues (2015), using
sample of 4  million was used in 1995 and diag- Veteran’s Administration data, showed mark-
nosed according to the International Classification edly increased care costs in people with BED or
of Disease, 9th edition. Treatment costs and out-​of-​ EDNOS as compared with those without an eat-
pocket patient expenses were assessed from insur- ing disorder ($33,716 vs. $33,052 vs. $19,548,
ance claims. Group differences in annual treatment respectively).
costs of EDs—​anorexia nervosa (AN), BN, and Taken together, these findings suggest elevated
binge eating disorder (BED)—​ were compared costs in ED of similar magnitude to that seen in
against schizophrenia and obsessive-​ compulsive other psychopathology. In addition, these elevations
disorder (OCD). may precede the point of formal diagnosis.
A total of 21,567 insurance claims, comprising a
total of 1,932 patients, were reported for EDs; this Estimates of the National Cost of
number accounted for 1.1% of all mental health Eating Disorders
claims. Inpatient treatment occurred much less fre- Relatively few cost-​of-​illness studies exist within
quently than outpatient treatment for all EDs. The the scientific literature examining the cost implica-
average cost of inpatient treatment collapsed across tions of EDs. This is surprising given the common
EDs was $12,432 for female patients and $10,126 view of EDs as expensive. Further, evidence suggests
for male patients. The groups did not differ signif- that use of healthcare in those with EDs is higher
icantly in inpatient treatment costs. Outpatient than in those without EDs and elevated to a degree
treatment costs for female patients were $2,344 for similar to comparison subjects with other psychi-
AN, $1,882 for BN, and $1,582 for eating disorder atric disorders (Sansone, Wiederman, & Sansone,
not otherwise specified (EDNOS). Male outpatient 1997; Striegel-​Moore et al., 2005). Only five formal
treatment costs were $1,154 for AN, $1,206 for cost-​of-​illness studies examining the cost implica-
BN, and $1,150 for EDNOS. tion of EDs exist; these are reviewed in this section.

Crow 411
 The UK Office of Health Economics (Office of The publication by Matthers, Vos, and Stevenson
Health Economics, 1994) used national surveys of (1999) is currently the most comprehensive cost-​
general practice and hospital use to assess both pri- of-​illness study for EDs. Data were accessed from
mary and tertiary care costs of EDs in the United National Mental Health Surveys (1994–​1997) to
Kingdom. The authors concluded that in 1990, assess years lost due to disability, public and private
46,806 patients in the United Kingdom sought gen- healthcare costs, pharmaceutical costs, and research
eral practitioner consultation for AN. The average and prevention funding. Annual cost of EDs in
unit cost per consultation was used (9.85 Euros) to Australia (1994) was estimated to be $22  million
calculate the annual primary care cost in the United (Aus). In addition, expenses on research, adminis-
Kingdom of 580,000 Euros. For tertiary care cost, tration and prevention were estimated at $4  mil-
the total number of inpatient treatment bed days lion (Aus) in 1994. Primary and inpatient care costs
were used to assess the cost of AN. A total of 25,748 were estimated at $3 and $14 million, respectively.
annual bed days occurred in 1990, with an average Two additional studies have been conducted
inpatient length of stay of 21.5 days for AN, for a to assess the costs of EDs. An Austrian study by
total cost of 3.5  million Euros. This method may Rathner and Rainer (1997) assessed the inpatient
not have identified other medical use or outpatient treatment cost of AN and BN, estimated at 140 mil-
care. Therefore, the overall cost of AN in the United lion Austrian schillings in 1994. Furthermore,
Kingdom is likely to be grossly underestimated at Nielsen et al. (1996) published a study in Denmark
4.2 million Euros. assessing the inpatient treatment costs of EDs. The
Krauth, Buser, and Vogel (2002) examined researchers found the annual treatment cost in 1993
both direct and indirect costs of ED in Germany. for inpatients with EDs was €6.4 million, 4.7 mil-
The study remains the only cost-​of-​illness research lion of which was specifically incurred by AN.
including indirect cost estimates for EDs. Data was Overall, the national cost-​of-​illness literature on
accessed through statutory health insurance (SHI), ED remains limited in size and scope, and the meth-
statutory pension insurance (SPI), and epidemio- odological inconsistencies used within the studies
logical literature on anorexia and bulimia. Direct provide vastly differing cost estimates. At this point,
costs included inpatient treatment costs as well as these estimates are also relatively dated. An addi-
rehabilitation (time spent in convalescent centers). tional major limitation is the lack of estimates of
Indirect costs were assessed through mortality and indirect costs associated with EDs. Since the only
morbidity costs. Importantly, outpatient care, psy- study to examine this suggested indirect costs are
chotherapy, and pharmaceuticals were not exam- almost triple direct costs accrued from EDs (Krauth
ined by this study. The estimated costs of EDs are et  al., 2002), research including indirect costs of
thus an underestimate. EDs would likely drastically increase cost estimates
The direct and indirect cost of anorexia nervosa as well as improve the validity and accuracy of cost-​
was estimated to be 195 million Euros. Direct cost of-​illness studies. Indirect costs should be assessed
of illness estimates were 59 million Euros for inpa- in future studies to yield more thorough and accu-
tient expenses and 6 million Euros for both rehabili- rate cost estimates. For now, cost-​of-​illness research
tation and convalescence costs, totaling 65 million findings for EDs likely remain gross underestimates.
Euros for direct treatment costs. Interestingly, indi-
rect treatment costs were estimated to be vastly Cost-​Effectiveness of Eating Disorder
greater, totaling 130 million Euros (67% of total). Treatments
Mortality costs were estimated at 123 million Euros, Only a handful of studies have specifically exam-
using the human capital approach. ined the cost-​effectiveness of various treatments. In
Bulimia nervosa was estimated at a cost of conducting such a study, the first question to be
124 million Euros to the German economy. Direct answered is:  Will the analysis involve direct data
treatment costs were approximately 7  million for collection or modeling based on existing literature?
inpatient treatment costs and 3  million for reha- Modeling studies are more readily conducted and
bilitation and convalescent costs, totaling 10  mil- can be quite valuable, but they introduce a greater
lion Euros. Indirect costs were estimated at being number of uncertainties with regard to the assump-
92% of the overall cost of BN, totaling 112 million tions made in the model. Thus, direct data collec-
Euros. Overall, AN and BN in Germany combined tion is preferable. However, most ED treatment
to equal a total estimated cost of 319 million Euros studies to date have only collected data primarily
(for a population of roughly 82 million). examining clinical effectiveness; only a few studies

412 Costs and Cost-Effectiveness

(Byford et al., 2007; Crow et al., 2013) have been costs for 81 participants in the trial; avenge cost per
designed a priori for prospective examination of subject was $33,015. The cost-​effectiveness index
cost-​effectiveness (but more will be forthcoming). (the costs for all subjects divided by the number
In a few other instances, studies have been designed of remitted subjects) was calculated using several
to examine clinical effectiveness only and examina- remission definitions, and this ranged from approx-
tions of cost-​effectiveness have followed thereafter. imately $34,000 to nearly $84,000, depending on
One study has attempted to model cost out- the definition used.
comes in AN treatment (Crow & Nyman, 2004). Egger et  al. (2016) compared cost-​effectiveness
This study made a number of assumptions regard- of focal psychodynamic therapy (FPT) and
ing course of illness and mortality. In addition, cognitive-​behavioral therapy-​enhanced (CBT-​E) in
assumptions were made about an integrated treat- 156 people with AN across 10 months of treatment
ment approach versus a more typical “community” and 1 year of follow-​up. The authors concluded that
approach with regard to treatments provided, unit FPT was more effective that CBT-​E and cost less as
costs, and effectiveness. The unit of analysis (given well (that is, FBT dominated).
the high mortality rate associated with AN) was A fourth study examined the cost-​effectiveness
the cost per year of life saved. This modeling anal- of CBT versus a stepped series of interventions in
ysis yielded an overall cost per year of life saved of the treatment of BN (Crow et  al., 2013). In this
$30,180. The authors concluded that this cost fell study involving 293 subjects, use was determined
well within the typically accepted norms for the from study participation records as well as the
value of a year of life in other areas of medicine. completion of healthcare diaries. Unit costs came
Another study has reported attempts to assign from the Center for Medicare Services, as well as
unit costs to a BN treatment trial (Koran et  al., the Red Book, in the case of drug prices (Thompson
1995). In this study, costs were assigned to the use Healthcare, 2005). Effectiveness was defined as absti-
observed in participants in a trial examining vari- nence, generated from study records. The results of
ous lengths of cognitive-​behavioral therapy (CBT) the trial showed that stepped care approaches were
treatment or desipramine treatment, either alone or more effective and cost less than beginning with
in combination, in the treatment of BN. The effec- CBT (mean costs of $12,146/​abstinent subject for
tiveness metric came from the data collected in the stepped vs. $20,317 for CBT). Of note, these costs
original trial. Cost-​effectiveness ratios were calcu- per abstinent subject substantially exceeded the cost
lated to yield a cost per abstinent patient at 32-​week of treating a given subject (since only a fraction of
follow-​up; these suggested that medication treat- subjects become abstinent).
ment was more cost-​effective than psychotherapy Fifth, Lynch and colleagues (2010) examined
alone or combination treatment. cost-​effectiveness of CBT-​ based guided self-​ help
The cost-​effectiveness of telemedicine-​delivered (GSH) for 123 people with BN, BED, or recurrent
CBT was examined in a randomized, controlled binge eating in a randomized, controlled trial. The
trial (Crow et al., 2009). This study included costs study used a limited societal perspective. The CBT-​
for both patient and therapist travel (in the face-​to-​ GSH was more effective and less costly than treat-
face condition), and showed that telemedicine cost ment as usual.
was more cost-​effective than face-​to-​face treatment. Finally, one paper has reported on the results a
Byford and colleagues (2007) reported on the cost-​effectiveness analysis of an integrated treatment
cost-​effectiveness for different treatment strategies program for EDs in a hospital setting (Williamson,
for AN. This multicenter trial with 167 participants Thaw, & Varnado-​ Sullivan, 2001). This study
examined cost-​effectiveness of inpatient, specialist compared two strategies for treatment:  beginning
outpatient, or generalist outpatient treatment over with inpatient treatment, or using a “systematic,
2 years of follow-​up. Use and unit costs came from decision-​tree approach to treatment.” This study
National Health Service data, while effectiveness involved 51 subjects with either AN or BN who
results were generated by the trial. After 2  years, were assigned to one of those two treatment strat-
specialist outpatient treatment dominated (i.e., was egies. Costs came directly from hospital records.
more effective and cost less than) the alternative The results of this study showed that overall costs
treatment. were nearly $10,000 less per case in the decision-​
The cost-​ effectiveness of family-​ based ther- tree treatment approach. Of note, however, since
apy for AN was estimated in another study (Lock, the costs were hospital-​based only, it seems cer-
Couturier, & Agras, 2008). This study estimated tain that not all healthcare costs were captured; a

Crow 413
fuller accounting of outpatient and other medical well-​designed cross-​effectiveness or cost-​utility
costs might have resulted in a smaller or even larger components, so such data will be forthcoming.
advantage for the decision-​tree approach.
References
Cost Utility of Eating Disorder Treatment Aardoom, J. J., Dingemans, A. E., van Ginkel, J. R., Spinhoven,
P., Van Furth, E. F., & Van den Akker-​van Marle, M. E.
Cost utility is defined as the cost of improved
(2016). Cost-​utility of an internet-​based intervention with or
quality of life, usually expressed as quality-​adjusted without therapist support in comparison with a waiting list
life years (QALYs). Three studies have examined the for individuals with eating disorder symptoms: a randomized
cost utility of ED treatments. The first examined controlled trial. International Journal of Eating Disorders, 49,
cost per QALY in BN treatment in 72 people receiv- 1068–​1076. doi: 10.1002/​eat.22587
Bellows, B. K., DuVall, S. L., Kamauu, A. W., Supina, D.,
ing inpatient treatment (Pohjolainen et al., 2010).
Babcock, T., & LaFleur, J. (2015). Healthcare costs and
Significant improvements in quality of life were resource utilization of patients with binge-​eating disorder and
seen with treatment, and in the base case analysis, eating disorder not otherwise specified in the Department of
the cost per QALY was 16,481 Euros. Veterans Affairs. International Journal of Eating Disorders, 48,
A second study looked at the cost utility of 1082–​1091. doi: 10.1002/​eat.22427
Byford, S., Barrett, B., Roberts, C., Clark, A., Edwards, V.,
hospitalization for AN (Pohjolainen et  al., 2016).
Smethurst, N., & Gowers, S. G. (2007). Economic evalu-
Again, substantial improvements in quality of ation of a randomised controlled trial for anorexia nervosa
life were observed in the 39 hospitalized partici- in adolescents. British Journal of Psychiatry, 191, 436–​440.
pants. Cost per QALY in the base case analysis was doi: 10.1192/​bjp.bp.107.036806
64,440 Euros. CDC/​National Center for Health Statistics (1995). International
Classification of Diseases, 9th Edition.
Aardoom and colleagues (2016) examined cost
Crow, S. J., Agras, W. S., Halmi, K. A., Fairburn, C. G., Mitchell,
utility of online treatment for ED (FEATBACK) J. E., & Nyman, J. A. (2013). A cost effectiveness analysis
alone or with varying intensity of therapist support. of stepped care treatment for bulimia nervosa. International
All levels of active treatment showed benefit, with Journal of Eating Disorders, 46, 302–​307. doi:  10.1002/​
strongest support for cost-​utility of FEATBACK eat.22087
Crow, S. J., Frisch, M. J., Peterson, C. B., Croll, J., Raatz, S.
without therapist support.
K., & Nyman, J. A. (2009). Monetary costs associated with
bulimia. International Journal of Eating Disorders, 42, 81–​83.
Future Directions doi: 10.1002/​eat.20581
The foregoing suggests several areas for future Crow, S. J., & Nyman, J. A. (2004). The cost-​effectiveness of
research efforts. First, updated estimates of the anorexia nervosa treatment. International Journal of Eating
Disorders, 35, 155–​160. doi: 10.1002/​eat.10258
national costs associated with EDs from a wide
Egger, N., Wild, B., Zipfel, S., Junne, F., Konnopka, A.,
variety of countries would be useful for ED advo- Schmidt, U.,  . . .  Konig, H. H. (2016). Cost-​effectiveness
cates and for those planning healthcare services. of focal psychodynamic therapy and enhanced cognitive-​
Notably, the existing estimates are now relatively behavioural therapy in out-​patients with anorexia nervosa.
old. Additionally, the national costs of ED provide Psychological Medicine, 46, 3291–​3301. doi:  10.1017/​
S0033291716002002
a potent argument for advocates of ED prevention
Koran, L. M., Agras, W. S., Rossiter, E. M., Arnow, B., Schneider,
efforts. Where possible, both direct and indirect J. A., Telch, C. F.,  . . .  Kraemer, H. C. (1995). Comparing
cost should be examined. As noted before, time cost the cost effectiveness of psychiatric treatments: Bulimia ner-
associated with EDs may be particularly important. vosa. Psychiatry Research, 58, 13–​21.
Developing methodologies to measure individual Krauth, C., Buser, K., & Vogel, H. (2002). How high
are the costs of eating disorders—​ anorexia nervosa
indirect costs would help to provide a more accu-
and bulimia nervosa—​ for German society? European
rate picture of the total illness-​associated burden Journal of Health Economics, 3, 244–​250. doi:  10.1007/​
and may help to form part of a stronger rationale s10198-​002-​0137-​2
for treatment that might be a powerful tool in indi- Lock, J., Couturier, J., & Agras, W. S. (2008). Costs of remission
vidual treatments. and recovery using family therapy for adolescent anorexia
nervosa: A descriptive report. Eating Disorders, 16, 322–​330.
The second broad area in which continued
doi: 10.1080/​10640260802115969
research is needed involves the cost-​ effectiveness Lynch, F. L., Striegel-​Moore, R. H., Dickerson, J. F., Perrin, N.,
of treatments. Several studies have examined this Debar, L., Wilson, G. T., & Kraemer, H. C. (2010). Cost-​
topic thus far, but the number of studies examin- effectiveness of guided self-​ help treatment for recurrent
ing clinical effectiveness greatly outweighs the num- binge eating. Journal of Consulting and Clinical Psychology,
78, 322–​333. doi: 10.1037/​a0018982
ber examining cost-​effectiveness. Currently, studies
Matthers, C. D., Vos, E. T., & Stevenson, C. E. (1999). The bur-
of new treatments and new methods of treatment den of disease and injury in Australia. Canberra:  Australian
delivery are underway, and many of these have Institute of Health and Welfare.

414 Costs and Cost-Effectiveness

Mitchell, J. E., Myers, T., Crosby, R., O’Neill, G., Carlisle, J., & Preventive Medicine Reports, 2, 32–​34. doi:  10.1016/​
Gerlach, S. (2009). Health care utilization in patients with j.pmedr.2014.12.002
eating disorders. International Journal of Eating Disorders, 42, Sansone, R. A., Wiederman, M. W., & Sansone, L. A. (1997).
571–​574. doi: 10.1002/​eat.20651 Healthcare utilization among women with eating disordered
Nielsen, S., Moller-​ Madsen, S., Isager, T., Jorgensen, J., behavior. American Journal of Managed Care, 3, 1721–​1723.
Pagsberg, K., & Theander, S. (1996). Utilization of psy- Striegel-​Moore, R. H., Dohm, F. A., Kraemer, H. C., Schreiber,
chiatric beds in the treatment of ICD-​ 8 eating disorders G. B., Crawford, P. B., & Daniels, S. R. (2005). Health ser-
in Denmark, 1970–​1993. Paper presented at the AEP vices use in women with a history of bulimia nervosa or binge
Conference, London. eating disorder. International Journal of Eating Disorders, 37,
Office of Health Economics. (1994). Eating Disorders. 11–​18. doi: 10.1002/​eat.20090
London: OHE. Striegel-​Moore, R. H., Leslie, D., Petrill, S. A., Garvin, V., &
Pohjolainen, V., Rasanen, P., Roine, R. P., Sintonen, H., Rosenheck, R. A. (2000). One-​year use and cost of inpatient
Koponen, S., & Karlsson, H. (2016). Cost-​effectiveness of and outpatient services among female and male patients
anorexia nervosa in QALYs. Nordic Journal of Psychiatry, 1–​5. with an eating disorder: Evidence from a national database
doi: 10.1080/​08039488.2016.1224922 of health insurance claims. International Journal of Eating
Pohjolainen, V., Rasanen, P., Roine, R. P., Sintonen, H., Disorders, 27, 381–​389.
Wahlbeck, K., & Karlsson, H. (2010). Cost-​utility of treat- Thompon Healthcare. (2005). The Red Book. Thomson.
ment of bulimia nervosa. International Journal of Eating Toulany, A., Wong, M., Katzman, D. K., Akseer, N., Steinegger,
Disorders, 43, 596–​602. doi: 10.1002/​eat.20754 C., Hancock-​Howard, R. L., & Coyte, P. C. (2015). Cost
Rathner, G., & Rainer, B. (1997). [Annual treatment rates and analysis of inpatient treatment of anorexia nervosa in ado-
estimated incidence of eating disorders in Austria]. Wien Klin lescents: Hospital and caregiver perspectives. CMAJ Open, 3,
Wochenschr, 109, 275–​280. E192–​E197. doi: 10.9778/​cmajo.20140086
Samnaliev, M., Noh, H. L., Sonneville, K. R., & Austin, S. Williamson, D. A., Thaw, J. M., & Varnado-​ Sullivan, P. J.
B. (2015). The economic burden of eating disorders and (2001). Cost effectiveness analysis of a hospital-​ based
related mental health comorbidities:  An exploratory anal- cognitive-​behavioral treatment program for eating disorders.
ysis using the U.S. Medical Expenditures Panel Survey. Behavior Therapy, 32, 459–​477.

Crow 415
 PART 
5
Emerging Topics
 CH A PT E R


Selective Eating: Normative Developmental
22 Phase or Clinical Condition?

Nancy Zucker, Courtney Arena, Cortney Dable, Jasmine Hill, Caroline Hubble,

Emilie Sohl, and Jee Yoon

Abstract
Selective eating (also referred to as picky or fussy eating) has been described as a normative
developmental phase that a significant minority experience and, potentially, “grow out of” without formal
intervention. This chapter reviews the literature on selective eating from the stance that this eating
pattern is a clinical condition rather than a normative developmental phase. Construing selective eating
as a clinical condition, it probes questions of definition, chronicity, and impairment that would warrant
intervention. It explores the phenomenology of selective eating, suggesting that the experience of disgust
has been relatively neglected in understanding the experience of selective eaters and that the inclusion
of this feature may offer some novel hypotheses for both necessary treatment elements and novel
conceptualizations about what it means to “outgrow” selective eating. Finally, assuming the hypotheses
proposed are accepted, it suggests some necessary treatment elements to expand food variety in
individuals with selective eating.
Key Words:  selective eating, picky eating, food avoidance, disgust, sensory sensitivity

Selective eating, an eating pattern in which an Notably, this chapter takes a somewhat nontradi-
individual eats a lower variety of food than is typical tional stance. Rather than considering selective
or potentially healthy, is arguably the most preva- eating as a normative developmental phase, we
lent pattern of disordered eating, but one for which consider whether selective eating is a clinical condi-
there is the least consensus. At issue are fundamen- tion and whether those children who “grow out of”
tal questions:  What is the definition of a selective selective eating manifest a different pattern of eating
eater? What is the phenomenology of a selective than those who persist.
eater? Can selective eating can be considered a nor-
mative developmental phase? Is selective eating Selective Eating: How Should We Define It?
harmful? When or how should intervention should “Selective eating” (hereafter SE) is a term
be undertaken, and if so, by whom? What does it that has been used to describe a pattern of eat-
mean to “grow out of” being a selective eater? ing typically characterized by the following crite-
There have been some excellent reviews written ria:  (1)  consumption of a narrow range of food,
to better inform the nature of selective eating (for (2)  unwillingness to try new foods, (3)  absence
example, see Taylor, Wernimon, Northstone, & of abnormal cognitions regarding weight and/​ or
Emmett, 2015). Rather than repeating the excellent shape, (4)  absence of premorbid preoccupations
information depicted in these reviews, this chapter regarding weight and/​or shape, (5) low, normal, or
focuses on these controversies and the research that high weight (Lask & Bryant-​Waugh, 1999, p. 39).
can inform preliminary steps that both address these Selective eating has also been defined as an “extreme
questions and give directions for future research. selectivity in preferred foods” (Nicholls, Christie,

419
Randall, & Lask, 2001). To add further clarity to Gonçalves, 2016), mothers were asked whether
the presentation of SE, we attempt to focus on their child “does not eat well” or “refuses to eat”
children, adolescents, and adults who are selective, and had to respond with either not at all applica-
but whose food intake is adequate (i.e., their weight ble, sometimes, or often applicable. Picky eaters
is normal or high). This focus is to avoid low weight were defined as those with an answer of sometimes
status complicating the clinical picture. or greater, a strategy that adhered to the methods
To develop a system for efficiently intervening employed by Cano et al. (2015). In fact, Machado
with individuals with SE, it is first helpful to have et al. (2016) deliberately employed this strategy so
a precise definition of the condition that is being that more typical eating disturbances and also more
treated. In the case of SE, there are at least three core severe eating disturbances were captured. Prevalence
questions that need to be resolved to define when estimates resulting from this definitional strategy
SE is a clinical condition worthy of attention. First, ranged from 13% to 27%. Thus, across all these
what is the frequency with which SE occurs that studies (and the majority of published research),
is considered indicative of a clinical threshold of a child who sometimes did not eat well or some-
severity? Second, what is the duration for which this times refused to eat could be combined with a child
threshold should occur that may warrant concern? who often or always refused to eat or always did not
Third, what are the essential features that discrimi- eat well. We would argue that these are very dif-
nate an individual with SE from one without? We ferent groups of children and that making a cutoff
attempt to address each of these questions in turn. of “sometimes” is blurring the boundaries between
transitory eating behaviors and those that persist
Frequency Threshold of Severity and are associated with impairment.
Differences in how frequency is defined for SE An alternative has been to use more data-​analytic
across research cohorts may contribute to vast dif- strategies to better capture the nature of SE and
ferences in the prevalence estimates of this eating arrive at perhaps more precise prevalence estimates.
pattern, let  alone our understanding of the phe- For example, in Tharner et al.’s (2014) study of 4-​
nomenology of SE. We argue that these differences year-​olds in the Netherlands, the study in which the
in prevalence estimates contribute to misconcep- prevalence was cited as 5.6%, a version of the Child
tions about SE, and lead to conceptualizations of Eating Behaviour Questionnaire was used with 35
SE as a normative developmental phase, rather than items relating to eating behavior, with answers rang-
a behavioral pattern that is distinct and persistent ing from 1 (never) to 5 (always) on a 5-​point Likert
for a subset of individuals. For instance, prevalence scale. Children classified as “fussy eaters” scored 1.5
of SE has ranged from a low of 5.6% to a high of standard deviations higher than remaining groups
59%. If nearly 60% of children were demonstrating on subscales measuring food fussiness, slowness
a particular eating pattern, it would indeed be hard in eating, and satiety responsiveness. Notably, this
to argue that it was atypical. However, studies that group did not differ in energy intake compared with
document higher prevalence levels have typically nonfussy eaters, though they exhibited a lower BMI
used one item, and that one item permitted child- than nonfussy groups (we return to this later). In
ren who were “sometimes” selective to be grouped terms of this discussion of a “pure” SE group defined
with those who were “often” or “always” selective. by limited food variety, this group may include those
For example, Xue et al. (2015) documented preva- who also consumed an insufficient quantity of food.
lence estimates of 59% in 7-​to 12-​year-​olds by ask- The authors argue that this lower prevalence may be
ing mothers whether their child was a picky eater, due to a data-​driven approach for defining SE rather
with responses including “never picky,” “some- than using a single question as employed in some
what picky,” and “always picky.” Picky eaters were studies. Micali et  al. (2011) conducted an elegant
defined as those who were “somewhat picky” or study in which factor analysis was used to define a
“always picky.” In a study of young children (4 to group with SE. While these researchers also used a
24  months old), Carruth et  al. (Carruth, Ziegler, threshold of “sometimes,” an important difference
Gordon, & Barr, 2004) defined picky eater status in the study is that every item on the selective eat-
as those whose caregivers endorsed either “some- ing factor loading had to meet this threshold (five
what picky” or “very picky” resulting in a picky items assessing food selectivity, unwillingness to try
eating category with a prevalence estimate of 19%–​ new foods, need for special food preparation, and
50% (Carruth et  al., 2004). In Machado et  al.’s strong preferences). The resulting group represented
2016 study (Machado, Dias, Lima, Campos, & about 7% of the sample and had between 1.6–​4.2

420 Selective Eating
the odds of having comorbid psychiatric conditions For example, even children with the most limited
or symptoms (approximately 3% to 7% of the sam- of diets would consider trying a new type of candy
ple in regard to psychiatric diagnoses and 30% of (though never a new type of fruit, vegetable, or pro-
the sample in regard to somatic symptoms). These tein). These considerations guide our recommenda-
data-​driven approaches might capture more com- tions for future directions in this area.
plex conceptualizations of SE by capturing multi-
ple features and might lead to the development of Conclusions/​Considerations for Severity
screening tools. However, at a more fundamental Threshold
level, these approaches make it challenging to under- There is a need for sensitivity analyses to deter-
stand the phenomenology of SE, or differentiate the mine the threshold of severity associated with
“typical” from the severe, as different levels of sever- impairment. This is discussed further in the next
ity (or important differences in presentation) may be section. In addition, newly developed assessment
collapsed into single categories. That said, these ele- tools (or adaptations of existing measures) should
gant approaches offer an intriguing and potentially consider having an option of “almost always/​always”
less biased approach to defining SE. to allow caregivers to endorse a greater frequency
A potential compromise solution to this thresh- of severity while avoiding the challenge of having
old confusion was proposed by Mascola, Bryson, to endorse a case that has no exceptions (as in the
and Agras (2010). Notably, the authors were case of endorsing “always”). Future research should
addressing a different question: that of coming up compare this more severe group with children who
with the definition of SE that would capture this are sometimes selective to see whether these groups
behavioral pattern as a trait feature. However, the are different in kind or degree.
solution is interesting. Over duration of years,
a child was considered as having SE if every year
their mother rated them as picky at a level of some- Consideration One:  Defining a Threshold
times or higher but at least 1 year had have a rating
of Selective Eating Severity
of often or higher. Thus, they combined a higher
threshold item against the background of averagely Definitions of selective eating should use
endorsed ratings. Taylor, Wernimont, Northstone, a cutoff of “often” or greater to distinguish
and Emmett (2015) also derived a more optimal children more likely to have persistent prob-
definition of SE. For children to be classified under lems with this eating pattern. Given the sub-
SE, parents had to endorse the extreme rating (Yes, jective challenge individuals have endorsing
Very Choosy) at least at 2 time points. Using this an option of “always,” an option that permits
classification, 3.5% of children were categorized “always/​almost always” might permit distinc-
under SE. From our perspective, this strategy repre- tion between clinically meaningful differences in
sents an important advance in defining a more pre- selective eating. Thus, rather, than selective eat-
cise SE group. ing being a typical developmental phase, it may
What is important about all these studies is that be more precise to say that a significant minority
despite the variable definitions, the SE condition are sometimes picky.
was associated with elevated psychological features
and specific behavioral patterns with surprisingly
marked consistency. Yet, if our goal here is to try Duration
to characterize the condition of SE that differen- While information on the course of SE is useful
tiates it from normative developmental patterns, in helping us understand the pathophysiology of this
then the thresholds previously described might not eating pattern, defining a necessary duration of SE
capture the essence of this eating behavior. Parents that might warrant intervention is essential. What
of children with SE at the more extreme ends of is needed is research that can quantify the length of
severity, by definition, would not characterize their time that an individual of a given age experiences
children as “sometimes” engaging in SE. For these impairment from SE. One of the more robust find-
parents, SE is a daily problem that interferes with ings across studies of SE is that picky eating at Time
meal preparation and potentially the child’s health One predicts picky eating at Time Two (Micali,
and well-​being. However, it is also hard for parents Rask, Olsen, & Skovgaard, 2016): there is increas-
to endorse “always” as an item response:  one can ing evidence that SE can be a persistent condition
always think of the exceptions to a given scenario. for a subset of individuals. However, the question

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 421

of duration is one of impairment: How long does we are discussing SE, this would be a child whose
SE need to persist for impairment in domains of dietary quality is such that she/​he/​ze requires vita-
physical health due to poor nutritional quality or min or mineral supplementation for health. One
psychosocial functioning to occur? In addition to consideration is that if a child is missing an entire
more obvious concerns about nutritional deficien- food group (which has been documented in several
cies resulting from avoidance of entire food groups, studies), then impairment in this domain is consid-
unique forms of impairment are associated with or ered. One question related to this suggestion would
may result from SE (e.g., family conflict around be, Does the documented avoidance of a food group
meals). Of interest, the form impairment takes may have to show up on a laboratory test as a vitamin
differ depending on the age of the individual with deficiency for child to be considered impaired in
SE, and so it is also unclear the duration at each age terms of their diet quality? On the one hand, it seems
that the unique age-​specific forms of impairment illogical to “wait” for a vitamin deficiency to emerge
needs to occur for a parent or individual to endorse if the parent knows the child is avoiding an entire
their child’s eating as a problem. This is important food group. This would seem to invalidate the neces-
because any attempt to define a clinical duration sity of having government recommendations for the
ideally would index a length of time that is associ- dietary composition and quantity needed to sustain
ated with impairment in function that is develop- health (for example, US Department of Health and
mentally sensitive. Human Services, US Department of Agriculture,
Consider the feature of being underweight. 2015). There continues to be much confusion about
Studies of SE in toddlers and young children find when to intervene for this condition—​and rightfully
that children with SE are more frequently under- so, given broad research gaps.
weight; however, studies of adults with SE have not Although it is clear that selective eating is asso-
found this association. This is a rather interesting ciated with a limited diet and inadequate nutri-
observation and suggests perhaps that the under- tion, controversy remains in regard to whether
weight or lower weight of young SE children is due or not such nutritional deficiencies can result in
to parents restricting the quantity of preferred foods stunted growth over time. As mentioned, this
(perhaps because they are attempting to increase the chapter attempts to focus on those with SE who
variety of the children’s diet with healthier options). are normal or overweight. However, it worth
From an impairment standpoint, it would seem noting that several studies have documented ele-
that consequences due to lower energy stores might vated growth faltering or underweight status in
not be as consequential once individuals with SE those with SE (for example, Tharner et al., 2014;
have more control over their food choices. Xue et al., 2015). Conversely, another study dis-
Whether SE causes physical impairment is actu- covered that being picky was weakly associated
ally quite controversial. While we address the spe- with negative growth among toddlers and lack of
cific eating patterns of SE below, a repeated finding diet variety was found to be unrelated to growth
is that while the energy intake of SE children does (Wright, Parkinson, Shipton, & Drewett, 2007).
not differ from that of their non-​SE peers, the qual- Significant effects on growth related to SE are
ity of the diet is poor. Thus, while growth trajecto- inconsistently observed, yet SE is still associated
ries may not be affected, parents, nonetheless, and with incomplete diets characterized by insufficient
not inappropriately, worry about their child’s health. consumption of essential vitamins, minerals, and
This appears to be a point of communication dis- protein.
connection between parents and providers (Zucker, A final consideration in the domain of phys-
Copeland, & Egger, 2016). Pediatricians reassure ical impairment is that of gastrointestinal symp-
the parent about the child’s continued growth tra- toms. This has not been consistently addressed,
jectory. The parent worries about the consequences but somatic symptoms were found to be elevated
about the lack of certain nutrients from the child’s in a third of SE in a study by Micali et al. (2016).
diet. What is needed are more sensitive boundaries However, anecdotally and logically, as evidenced by
to help parents and healthcare providers determine the foods that selective eaters avoid (whole grains,
when lack of food variety has been going on for long fruits, vegetables), the lack of dietary fiber and sub-
enough as to impact health. Chatoor (2002) made sequent constipation may be problematic for these
the recommendation that if nutritional supplemen- children and further impairment. A  recent study
tation is required for the child to receive a diet of by Taylor et al. (Taylor, Northstone, Wernimont,
adequate nutrition, then impairment is implied. As & Emmett, 2016) confirmed this:  These authors

422 Selective Eating
found that hard stools were more frequent in the most fundamental ways of giving and receiving
children with SE and this was mediated by the nurturance. In the context of SE, there is a devas-
amount of dietary fiber the children consumed. tating loss of this intimate exchange while the afore-
Thus, the fundamental issue in the domain of mentioned misunderstandings may create tension
physical health is whether to wait for a medical test between parent and child.
that provides an abnormal value or for the child to Family meals have been shown to be a broadly
fall off a growth curve before impairment in this protective familial ritual that also may be com-
domain is considered. Given the persistence of SE promised by SE. A  broad research base has dem-
(discussed later), it would seem this is a very irre- onstrated that family meals that are frequent and
sponsible strategy. valued by the family as important can be protective
The psychosocial impairment resulting from of a variety of negative mental health outcomes and
SE might differ according to the age of the indi- high-​risk behaviors (Eisenberg, Olson, Neumark-​
vidual. While many young children with SE may Sztainer, Story, & Bearinger, 2004; Fulkerson et al.,
not experience impairing social consequences as a 2006). Family meals are an efficient means for a
result of their eating (perhaps either because their child to gain a broad range of skills: conversational
eating choices are not that different from their peers skills, teamwork, and manners are examples. Family
and there are not as many social events that demand meals also help children feel more emotionally con-
eating), yet elevated social anxiety has been docu- nected and valued by family members (Neumark-​
mented in preschoolers with severe SE and elevated Sztainer, Story, Ackard, Moe, & Perry, 2000). When
symptoms of social anxiety in moderate SE (Zucker a family has a child with SE, the lack of empiri-
et  al., 2015). In two studies of adults with SE, cal research to guide parents in how to approach
both reported elevated levels of social anxiety with unpreferred foods may lead to conflict between par-
increasing SE (Kauer, Pelchat, Rozin, & Zickgraf, ent and child, and for the child, to have negative
2015; Wildes, Zucker, & Marcus, 2012). This per- associations with mealtimes in general, not just with
haps reflects that the eating choices of individuals the disliked foods. Over time, studies have reported
with SE becomes increasingly atypical with age oppositional behavior at mealtimes and attempts to
and/​or that social events that involve eating become avoid family meals altogether by children with SE,
increasingly harder to avoid or manage (for exam- a reflection of ineffective strategies being employed
ple, business lunches, dating, etc.), challenges not (Mascola et al., 2010).
experienced to the same degree by young children Going back to our question of duration, the
with SE. issue is how long should each of these impairing
Yet, younger individuals with SE are not com- consequences (e.g., absence of an entire food group,
pletely devoid of psychosocial consequences. One social avoidance, family conflict) be endured prior
important example of social consequences for child- to intervention. To our knowledge, only a few stud-
ren and young adolescents with SE is the effect that ies have examined the issue of persistence over time
this pattern of eating has on family relationships. (Cano et  al., 2015; Mascola et  al., 2010; Micali
Increased family conflict and increased parental dis- et al., 2016; Taylor et al., 2015). A caveat in describ-
tress is perhaps the most consistent finding across ing the results is that two of the studies employed
studies of SE (Mascola et al., 2010; Zucker et al., the cutoff of “sometimes,” while the third required
2015). This domain of impairment is essential for that “often” be endorsed at least once (Mascola
several reasons. First, SE may negatively impact et  al., 2010). In the study by Cano et  al. (2015),
parent–​child relations. Given pervasive misunder- 4.2% of children age 1½ to 6 were found to have
standing of the etiology and phenomenology of SE, persistent SE across all time points, and these chil-
many parents feel guilty for their child’s eating, a dren were found to be at increased risk for the emer-
feeling propagated by well-​intended but misguided gence of pervasive developmental disorders. There
advice from friends and family. Parents themselves, were no satisfying predictors of persistence using
misunderstanding the barriers that make it chal- an SE group that included those who sometimes
lenging for their child to be adventurous with food, ate selectively, however, of interest, if the research-
might interpret their child’s eating as resulting from ers had changed the cutoff to five and above (the
their “stubborn temperament,” signs of a power cutoff consistent with an endorsement of “often” of
struggle, etc. On the other hand, in the context at least one item), then prevalence at a given time
of a healthy feeding dynamic, preparing food and point would range from 3.7% to 6%, a prevalence
having that meal be appreciated by family is one of not unlike other psychiatric disorders. Cano et  al.

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 423

(2015) reported that two-​ thirds of SE remitted
within 3  years. Masola et  al. (2010) reported that Consideration Two:  Defining a Threshold
47% of their cohort remitted, and that those with for Selective Eating Duration
SE duration of longer than 2  years demonstrated Children who have been persistent selec-
more severe SE than those who had SE for less than tive eaters for more than 2 years might warrant
2  years:  the former group had stronger food pref- referral. However, the presence of impairment
erences and were more likely to refuse new foods. indicates that the level of severity is impacting
Combined, these studies confirm that SE can last function and might warrant referral even if the
for years and that a substantial percentage do not duration does not meet this threshold.
persistently have SE when consistency is examined
over a span of years.
Predictors of persistence have been rather elu-
sive. Cano et  al. did not find a reliable predictor. Essential Features
Mascola et al. (2010) reported that picky eating was While the essential feature of SE is also prob-
the most reliable predictor; and Micali et al. (2011) ably the least contested, namely, individuals eating
confirmed this latter finding but also identified a limited variety of foods, it is also the most chal-
some maternal mental health variables as potential lenging aspect to define. There has been surprisingly
modifiable risk factors, which are discussed later. little emphasis on what individuals with SE actually
Yet, there are further gaps in our understanding of eat and whether the nature or limited variety of the
the course of SE. There are virtually no studies of specific foods selected result in physical health con-
the prevalence of SE in teenagers, and only a few sequences. Rather, the research in this area has been
studies in adults. criticized as focusing on the opinions of the mother
to guide the definition as to whether a child is selec-
Conclusion/​Considerations for Duration tive, rather than the behavior of the child. The con-
Threshold cern is that individual differences in the mother
From both the clinical and research perspec- (e.g., maternal anxiety, eating disorder, or own pick-
tives, findings to date on the necessary duration to iness) are biasing these judgments rather than the
warrant intervention are challenging to translate. objective behavior of the child. For our purposes,
There is not clear evidence to support a sensitive we attempt to focus on the objective behavior of
duration of SE since studies to date have exam- the child to define a threshold of food selectivity for
ined the question over years, yet impairment has the core feature of SE, and then examine frequently
been documented cross-​sectionally. While there is reported behavioral features that might further
increasing verification that persistent SE beyond inform the clinical picture.
three time points is indicative of a worse outcome One option is to use vitamin, mineral, or fiber
and may warrant intervention, this information inadequacy as the benchmark for poor food variety.
would be small comfort to parents who endorsed an While energy consumption is also a critical consid-
immediate impact of their child’s eating on family eration, this chapter attempts to focus on regular to
relationships. It is not reasonable to tell parents to overweight children to define benchmarks. Across
wait and see if their child’s eating persists for years. studies, the most robust findings are that children
Reexamination of data, with higher thresholds for designated as having SE (1) consume fewer vegeta-
defining severity might lead to more precise pre- bles, fewer fruits, and protein (particularly less meats
dictors of persistence of SE. In the meantime, the and fish) than nonselective eaters; (2) consume lower
limited data available supports 2 years as a duration levels of essential vitamins and minerals, particularly
threshold: A child with SE for longer than 2 years, folate, and while more likely to have a nutritional
and according to the Cano data, certainly more deficiency, in general, the data on the latter is limited
than 3  years, would warrant a referral. However, and mixed; (3) consume more sweets/​confectionery
the immediate signs of physical (e.g., hardened and savory snacks; (4)  have a total energy intake
stools) and psychosocial (e.g., family relationships) that is unlikely to differ from typical controls; and
necessitates the development and dissemination (5) and are more likely to have problems with hard
via pediatric primary care of interventions that can stools, constipation, and other gastrointestinal issues,
be disseminated on a broad scale immediately in perhaps due to their low dietary fiber. The question
response to parental concern and signs of impair- is whether the existing literature can be employed to
ment in the child. define thresholds of impaired dietary variety.

424 Selective Eating
 Indeed, the bulk of evidence supports dietary
differences, and there is emerging evidence that Consideration Three:  Defining a
such differences have clinical implications. Zimmer Threshold for Selective Eating Variety
et al. (2012) reported that it is significantly more Children who are avoiding entire food
likely for a child with SE to be at risk for at least groups, who require a vitamin and mineral sup-
one critical nutrient deficiency as opposed to an plement to meet dietary recommendations, or
average eater in a sample of children diagnosed experience gastrointestinal consequences, such
with an autism spectrum disorder. In a study as constipation, may be exhibiting a level of poor
that analyzed the eating habits and related factors food variety that impacts function.
within a sample of healthy school-​ age Chinese
children, researchers Xue et al. (2015) found that
selective eaters consumed less energy, protein, and Associated Features
most vitamins and minerals than nonselective There is rather robust evidence documenting
eaters; further, selective eaters had lower levels of associated behavioral features of individuals with
iron and were lower in height, weight, and BMI SE. Among these, the fear of trying new foods,
for their age. Similarly, as summarized in a com- decided preferences about foods (both likes and
prehensive review, Taylor et al. (2015) report that dislikes), the tendency to gag when presented
SE is associated with diets characterized by insuf- with a new food, and conflict over foods are the
ficient consumption of essential vitamins, miner- most reliably reported. Pace of eating has also been
als, and protein. Although overall energy intake reported in several studies and is a particularly
tends to vary only slightly between children who interesting feature that will consider from a behav-
are selective eaters and children who are nonse- ioral standpoint. Each of these features is briefly
lective eaters, it has been reported that selective discussed in turn.
eaters consume significantly less eggs and cooked
and raw vegetables than nonselective eaters (Van Food Neophobia
der Horst et al., 2016). Notably, these food pref- Food neophobia, the fear of trying new foods,
erences appear to be rather stable:  14-​month-​old is robustly documented among selective eaters.
infants who were considered selective eaters dem- However, it is important to note that this avoid-
onstrated higher overall intake of savory snacks ance might not be observed across all food groups
and sweets and lower intake of fish and vegetables, and, critically, might not be driven by fear. For
a pattern similar to older children who are selective instance, some children who avoid fruits, vegeta-
eaters (Van der Horst et al., 2016). Such a lack of bles, and most proteins will be willing to try novel
vegetables in a diet is also associated with a lack of sweets/​confectionaries provided that the sweets do
fiber intake, an important observation, given the not contain fillings or fruit flavors. When queried
study by Taylor et al. on hard stools reported previ- about this avoidance, children often do not report
ously (2016). Notably, however, the count of the that they are afraid something bad will happen if
number of foods may not provide much discrimi- they eat it, they report finding the taste disgusting,
native validity. Jacobi et al. (Jacobi, Agras, Bryson, report being likely to gag upon tasting a novel food
& Hammer, 2003) reported that while selective that they find disgusting, or report with certainty
eaters were statistically different from nonselective that they will find the taste disgusting. Viewed
eaters in the number of foods consumed, the means from an anxiety lens, one could consider this a “fear
differ by only one food. of gagging.” However, one of the many interest-
ing things about SE is the frequency with which
Conclusions/​Considerations for Food Variety gagging actually occurs when presented a novel
in Selective Eating or unpreferred food. Thus, gagging is not a low
Combined, the evidence indicates that selec- probability event and the expectation of gagging
tive eaters have lower dietary quality than nonse- is not an irrational belief, features often reported
lective eaters. Thus a tentative suggestion based on to characterize fears in phobias (e.g., fear of flying
evidence is that if the child avoids an entire food in anticipation of a plane crash). This is a critical,
group or, as suggested by Chatoor (2002), requires often overlooked, feature of SE that is essential to
vitamin or mineral supplementation to meet daily understand the phenomenology of this condition.
requirements, then a threshold of impairment for Thus, we return to this point when we discuss the
food variety has been reached. phenomenology of SE.

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 425

Decided Preferences a critical consideration: Numerous studies examin-
Children labeled as selective eaters know what ing the emergence of healthy eating behaviors in chil-
they like. Parents report loyalty to food brands to dren site parent modeling of healthy eating and the
such a degree that anecdotally, parents report writ- number of presentations of novel foods as contribu-
ing to food companies when preferred products are tory to a child’s subsequent SE. Importantly, this
discontinued for fear that they will not be able to work has never been conducted in children precisely
find a suitable replacement for a discontinued item. categorized as selective eaters, and whether the same
Also anecdotal (but interesting), is that these chil- strategies work, whether the strategies would work
dren have very “good taste.” If these children say but with significantly longer durations, or whether
that a particular French fry is a good tasting French novel strategies are needed, is currently unknown.
fry (and better than other French fries), it probably In the meantime, parents are being blamed for not
is. Parents describe these children as very sensitive doing enough when they have no guidance as to
to the visual imperfections of food, rejecting even a what they should be doing. As conflict is an inter-
previously liked food if its appearance differs from personal feature, and thus not a feature of the indi-
that of a prior presentation. This behavioral pat- vidual, we label this feature in terms of the child’s
tern bears some resemblance to features of other presentation, that of food refusal.
psychiatric disorders such as the “need for same-
ness” in autism spectrum disorders or the “just Slowness in Eating
right” phenomena in obsessive-​compulsive disor- A slowed pace of eating has been described
der. We discuss the relevant aspects of the experi- (Tharner et  al., 2014). This is a particularly inter-
ence of each of these features in our section on the esting feature, as the motivations or etiology that
phenomenology of SE. contributes to a slowed eating pace might help con-
tribute to more personalized treatment approaches.
Food Refusal/​Conflict Around Food Further, the amelioration of this symptom would
Individuals with SE refuse to eat unpreferred most likely be of great relief to parents, as the time
or unfamiliar foods even in contexts in which such taken for meal completion may be an added com-
refusal has social or emotional consequences; those plication with managing a child with SE.
with SE cannot override their distaste of unknown Two diverging hypotheses about the origin of
foods to “get by” in a challenging or novel social slowness in eating might contribute to very dif-
circumstance. This is notable, as individuals with- ferent conceptualizations about the nature of SE.
out SE may suffer through an unpleasant meal if One hypothesis is that slowness in eating is due to
refusing would hurt the feelings of the host, if they delays or atypicalities in oral-​motor development
are hungry enough, or if there are other interocep- that makes it challenging to chew and swallow
tive or contextual motivations. Those with SE seem food. Of interest, a study by Jacobi et  al. (2003)
impervious to these motivations, and it is informa- examined the early eating/​feeding behavior of chil-
tive to consider some hypotheses as to why this is dren who were classified as selective eaters at the
the case. One hypothesis is that the sensory aver- age of 8. Examining data collected in infancy, these
sion experienced from these foods makes it chal- authors reported the sucking behavior of children
lenging or seemingly impossible to attempt to taste later classified as selective eaters was weaker than
novel foods without further social embarrassment that of typically developing children. Such early
or insulting the host (due to vomiting gagging, vulnerabilities in oral-​motor development might be
etc.). Alternatively, or in addition, individuals with further compromised by the food types selected by
SE might struggle with deficits in theory of mind, children with SE. Processed foods combined with
implying that they might not appreciate the social lower (or absent) levels of fruits, vegetables, and
consequences or impact of their behavior on others, proteins suggest that these children opt for foods
and thus this does not serve as a potent source of that are easier to chew and swallow. Thus, over time,
motivation to alter eating habits. There is virtually this food selection may further impede the devel-
no data to inform either of these hypotheses. opment of oral-​motor strength and coordination.
A final important consideration regarding the Data on this hypothesis is limited, although there
presence of food conflict is the implication that is some evidence supporting that more severe SE
parents are not merely acquiescing to the children’s is associated with oral-​motor difficulties (Zucker
demands for food types (Zucker et al., 2015). This is et al., 2015).

426 Selective Eating
 A second hypothesis regarding slowness in eat- Kavaliers, & Ossenkopp, 2016; Stevenson, Case, &
ing is that is it a behavioral strategy that children Oaten, 2011; Stevenson et al., 2012; Zelazniewicz,
have learned as a way to avoid eating. By prolong- & Pawlowski, 2015). The associated experiences
ing a meal, children may have been negatively rein- of disgust that accompany these elevated immune
forced by receiving permission to leave the family responses are thought to further protect the organ-
table early. Thus, behavioral management strategies ism from harm due to infection. The adaptiveness
would be necessary in this latter instance, while and sensitiveness of the system implies that experi-
work with an occupational or speech therapist ences of malaise up-​regulate immune responses and
might be required in the prior instance. However, disgust sensitivity, which increases vulnerability to
given the limited data in support of this feature, we the experience of malaise. Indeed, the core features
do not include it in our associated features until fur- of the disgust experience are nausea, gagging, and
ther data is available. changes in facial musculature that restrict access to
nasal and oral entries. As disease or infection occurs
after a body boundary has been crossed, eating is
Consideration Four: Defining Associated rather unique among human behaviors as being
Features one of the few in which crossing a body bound-
ary is necessary. Thus, the experience of disgust
Individuals with SE engage in one or more
when one encounters spoiled foods is familiar. We
of the following either most of the time or all of
argue that individuals with SE have a lower thresh-
the time:
old for disgust experience that contributes to food
1.  food neophobia, avoidance.
2.  decided food preferences, and/​or What is notable about SE is that they have an
3.  refusal of foods offered by others. intense disgust reaction (example, gagging) to foods
that are unfamiliar but that do not obviously sig-
nal the threat of contamination. In contrast, there
The Phenomenology of Selective Eating are universal domains of disgust elicitors that relia-
Sensory Sensitivities, and Disgust bly produce disgust experiences (e.g., feces, vomit,
Experience urine). It is further thought that other higher-​order
To understand the experience of individuals with disgust elicitors, particularly those labeled as moral
SE, we argue that the role of anxiety has been over- disgust (e.g., incest), are potentially the result of
emphasized and that the interacting roles of sen- second-​order conditioning due to the association
sory sensitivities and disgust experience have been of immoral acts with bodily products. In contrast,
neglected. To address this gap, we discuss the nature the disgust elicitors in SE might not be captured by
of disgust experience and why this experience may current measures of disgust sensitivity, as these elici-
be relevant for SE, discuss how disgust experience tors do not generate disgust in the general popula-
explains the puzzling behaviors of SE, examine cur- tion. This was our own experience when applying
rent limitations in the measurement of disgust for a widely used disgust sensitivity scale (the Disgust
this group, hypothesize some potential contribu- Scale-​Revised; Olatunji et al., 2007) to a group of
tions to enhanced disgust experience, and finally, adults who self-​identified as selective eaters. While
describe implications of disgust for intervention. degree of SE severity was associated with increased
According to Curtis (2011), disgust is an adap- frequency of gagging and the subjective experience
tive emotional and behavioral system designed to of disgust as indexed by items directly assessing their
protect the organism from infection by pathogens. experience of disgust (Figure 22.1b and 22.1c), the
Pathogens are often impossible for humans to classic measure of disgust sensitivity was unable to
detect directly (individuals do not usually see bac- discriminate between groups (Figure 22.1a). To
teria). Thus, the disgust system is sensitive to local clarify, while all groups with SE differed from the
signals that signify potential harm from contami- group with only slight SE, a widely used measure
nation (e.g., feelings of nausea, smells, feelings of of disgust sensitivity cannot discriminate among
malaise). Of great interest, immune responses are these groups (Figure 22.1). Thus, there may be
up-​regulated in accordance with disgust experiences need for the development of measures that capture
or likewise when the organism is vulnerable (as dur- the experience of disgust with disgust elicitors of a
ing pregnancy or, putatively, starvation) (Cloutier, lower threshold of pathogen severity (e.g., broccoli).

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 427

 (a)

b b
b
2.20
b

Mean of Average Disgust Scale-Revised

2.00

1.80

1.60

Rarely Less than half About half More than half All the
or never the time the time the time time
(b) Severity of Selective Eating
4.00
d
Disgusted by new foods (1-Never to 5-Always)

3.50

3.00
c

2.50 b

2.00
a
a
1.50

Rarely Less than half About half More than half All the
or never the time the time the time time
(c) Severity of Selective Eating
4.00
Gagging when trying new foods (1 Rarely to 5 Always)

c
3.50

3.00
b

2.50
a

2.00
a
a
1.50

1.00

Rarely Less than half About half More than half All the
or never the time the time the time time
Severity of Selective Eating

Figure 22.1  Disgust experience in selective eating in a sample of 999 adults who completed a Web-​based survey of food selectivity.
1a. Average scores on the Disgust-​Scale Revised discriminated between adults who were rarely or never selective, but not adults who
ranged from less than half the time to all the time in degree of food selectivity. 1b. An item that asked adults whether they experience
disgust when trying a new food resulted in significant differences in degree of food selectivity between all but the group of rarely/​
never and less than half the time. Different letters indicate significant differences between groups (p < .001). 1c. Frequency of gagging
similarly discriminated between severity of selective eating, discriminating the two most severe groups (more than half the time/​
always) from each other and from lower frequency of selective eating. 
Otherwise, we may miss a potentially essential fea- contradictory context in which hunger can pro-
ture in the phenomenology of SE and not have pre- mote increased food avoidance or food rigidity.
cise ways to measure it. While some authors have posited that it is sensory
sensitivity combined with particular temperament
Contributions To Disgust proclivities (such as harm avoidance) that promotes
Experience: Sensory Sensitivities avoidance, it is also possible that the sheer inten-
Disgust sensitivity has been conceptualized as sity of a stimulus makes it aversive for the major-
a trait feature that evidences individual variabil- ity of individuals (e.g., a very loud noise). Thus
ity. What is interesting about disgust, when con- interoceptive and exteroceptive sensitivity might be
ceptualized as a system for pathogen avoidance, is independent vulnerability factors that increase the
that certain developmental stages and life events intensity of experienced disgust.
up-​ regulate immune responses, and associated Sensitivity to exteroceptive sensory experiences
disgust experience, that may become more exag- (e.g., touch, taste, smell, texture) has been increas-
gerated, depending on one’s initial predilection to ingly documented as an associated feature of SE
baseline disgust experience. In the context of SE, (Blissett, & Fogel, 2013; Chatoor, 2002; Farrow &
individual vulnerabilities that may be particularly Coulthard, 2012; Kauer et al., 2015). This associ-
germane to the enhancement of disgust experience ation is reported across the lifespan:  adults desig-
are exteroceptive sensory sensitivities, particularly nated as selective eaters endorsed greater sensitivity
sensitivity to taste, texture, and smell. Once an to smell and texture (Wildes et  al., 2012), while
individual has been infected with the pathogen, it another study of adults reported higher rates of
is too late. Thus, organisms have developed refined rejection of foods that were “slimy or slippery” in
systems to detect warnings of possible pathogen texture, or that had been mixed or had “lumps”
infection, such as visual cues and smells. It thus (Kauer et  al., 2015). Notably, this same study
stands to reason that individuals with height- showed that picky eaters were higher in disgust sen-
ened visual acuity, with a visual system that biases sitivity than nonpicky eaters (Kauer et  al., 2015).
details rather than global organizing frameworks, Yet the origins of these sensory aversions are poorly
or with extreme sensitivity to smells might be vul- understood.
nerable to enhanced disgust experience such that There is accumulating evidence regarding the
these individuals are more likely to notice visual role of genetic variation in contributing to the
flaws in foods or find a broader range of smells to intensity of taste experiences (Tepper, 2008).
be aversive (Figure 22.2). Individual variation in the density of tongue
Individuals who are interoceptively sensitive papillae, number of taste receptors, and genetic
may also be more likely to have enhanced disgust polymorphisms implicated in specific taste expe-
experience due to their greater sensitivity to the gut rience have all been explored as contributing to
motility inherent in feelings of nausea, an aspect of this variation in experienced taste intensity (de
disgust experience. Notably, some daily occurrences Krom, Bauer, Collier, Adan, & la Fleur, 2009).
of a selective eater might worsen the experience of Among these, the “bitter gene” is perhaps the
disgust. Anecdotally, selective eaters report often most commonly studied: studies have investigated
skipping lunch at school. Hypotheses about such the heritability of phenylthiocarbamide (PTC)
avoidance include the potential embarrassment of and 6-​ n-​
propylthiouracil (PROP). Individuals
eating the same lunch day after day and/​or because that experience a bitter taste when given PTC and
they might be overwhelmed by the visual cues and PROP are regarded as “super-​tasters,” indicat-
smells of other people’s lunches. As a result, they ing a heightened sensitivity to this flavor profile.
might go for long periods of time without eating. A  majority of these super-​tasters have polymor-
An interesting side note in relation to prolonged phisms in the gene TAS2R38 (Tepper, 2008). To
food avoidance and disgust experience is that smells date, links between variation in genetic polymor-
might become potentiated in a state of hunger, a phisms for taste sensation and SE are limited.
sensory alternation that would be adaptive in typ- However, in a condition that is so vastly misunder-
ically developing individuals: The enhanced smells stood and for which parenting is, to date, unduly
of delicious foods might increase the likelihood implicated, such findings of individual differences
of food consumption. In the context of SE, how- in taste experience are important to help paint a
ever, such enhanced smell intensity might further more complex picture of the phenomenology and
decrease the appeal of eating, resulting in a seemingly etiology of SE.

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 429

 Enhanced sensitivity to
events that affect visceral
or sensory experiences
(GERD, allergies,
constipation)
Emotional
Experience
(disgust)

Enhanced
Interoceptive
Perception Behavioral
(gut motility) Response
(e.g., gagging)
Deficits in
Hunger executive
attention
Enhanced
Exteroceptive
Sensory
Perception Decreased
(taste, smell texture, presentation of
Avoidance of
visual features) new foods by
social eating
sensitive social
environment
Novelty enhancing
sensory experiences?

Figure 22.2  A putative model of selective eating. Enhanced sensitivity to interoceptive and exteroceptive sensory experiences act
as vulnerabilities to enhanced disgust experiences. Given the aversive social consequences of disgust sensitivity (e.g., gagging), these
individual differences may make social eating more aversive, leading to further social avoidance. Behavioral features such as meal
skipping and deficits in executive attention may further potentiate these pathways. Meanwhile, avoidance and the subsequent novelty
of a broad range of foodstuffs may further increase aversive sensory experiences. 

gastroesophageal reflux (GERD) has been reported

Sensory Sensitivity and the Potentiation in SE, though it is unclear whether this condition is
of Emotional Experience more prevalent in those with SE (Seema & Aceves,
Sensory sensitivities might contribute to the 2016; Williams, Gibbons, & Schreck, 2005). One
potentiation of emotional experiences more broadly. study of selective eaters found that 22% of the child-
For example, sensory sensitivity might contribute ren had GERD, which is fairly high in such a young
to findings of elevated anxiety and SE (Farrow, & population (Williams et  al., 2005). Speculation
Coulthard, 2012). Researchers found that both about the connection between GERD and selec-
anxiety and sensory sensitivity are associated with tive eating has to do with conditioned aversions to
picky eating. Furthermore, child sensory sensitivity eating the particular foods that cause reflux. This
fully mediates the connection between anxiety and similar conditioned aversion may also be at play in
SE, suggesting that it is greater sensitivity to sensory children who have had allergic reactions to certain
information that explains why more anxious child- foods. Many times, allergic reactions are associated
ren are more likely to become selective eaters. with vomiting, diarrhea, and/​or abdominal pain,
which might cause conditioned aversions to certain
Medical Conditions and Sensory Sensitivity foods (Williams et  al., 2005). One of the signifi-
Such sensitivity to interoceptive or extero- cant issues in regard to allergies is that, often, food
ceptive experiences might further potentiate the allergies in young children are overlooked by pedia-
effects of medical conditions that have been asso- tricians, especially when the symptoms are internal-
ciated with SE, as the symptoms of these condi- ized (silent reflux, abdominal pain, etc.), as opposed
tions may be experienced more potently. To date, to externalized (vomiting, diarrhea, etc.) (Chatoor,
the domain of gastrointestinal symptoms has been 2002). The GERD symptoms might also be asso-
the most frequently investigated. More specifically, ciated with eosinophilic esophagitis in SE and can

430 Selective Eating
be complicated with allergic reactions (Spergel & Consequences of Disgust
Shuker, 2008). When considered against the back- Experience: A Sensitized Social
drop of our model of sensory sensitivity, the addi- Environment
tion of these medical conditions implies two factors A sensitive social environment may uninten-
that may be essential for stronger conditioned aver- tionally augment this avoidance. As noted in the
sions to foods. First, sensory sensitive individuals classic writings on the function of emotional expe-
might experience the medical consequences associ- rience by Darwin (1872), the primary function of
ated with gastrointestinal issues or allergic reactions emotion is that of social communication. Perhaps
more potently. Second, their sensory sensitivity this function is most relevant for the emotion of
might lead to broader generalization curves to a disgust, as the avoidance of pathogens is a group-​
range of sensory features that are imperceptible to level behavior as manifested in hygiene behaviors
individuals with less sensory sensitivity. None of (e.g., washing hands, covering one’s mouth when
these hypotheses has, of yet, been tested. sick). Thus, noticing signs of disgust from others is
critical in avoiding pathogens that could impact the
Executive Functioning and Sensory group. It is perhaps for this reason that individuals
Sensitivity engage in some seemingly odd social behaviors sur-
Deficits in executive functioning might further rounding the experience of disgust: They attempt
augment disgust experience. In a study by Zucker to share that experience. “Smell this” is a rather
et al. (2015), one unexpected finding was that mod- common refrain when one encounters a disgust-
erate levels of SE were associated with an increased ing stimulus: a reaction that one is unlikely to have
frequency of the symptoms of ADHD. One when one views something scary or anger provok-
hypothesis is that problems with shifting attention ing. The take-​home point is that we notice expres-
might make it challenging to regulate aversive sen- sions of disgust, particularly when such expressions
sory experiences so that these experiences become are severe, resulting in gagging or vomiting on the
intensified. However, to date, there are no studies of part of the individual.
neuropsychological function in individuals with SE. This is critical to consider when incorporating
Compared to other related psychological features the role of parents, either when implicated as con-
of selective eating, ADHD has been one of the least tributions to SE or as necessary agents of change.
studied. Two studies in this area are notable. Pennell In this chapter, we argue that parents have been
et al. (Pennell, Couturier, Grant, & Johnson, 2016) unduly maligned as the role of disgust and sensory
reported on two cases of young patients diagnosed sensitivity has not been adequately incorporated,
concurrently with avoidant/​restrictive food intake and further, that research on parent modeling, and
disorder (ARFID) and attention deficit hyperactiv- so on, has not been conducted with a well-​defined
ity disorder (ADHD). These patients had to take group of selective eaters.
stimulant medication, which posed a significant A child who gags or vomits following the pre-
growth restriction. The appetite suppressant effect sentation of a new food would generate cautious
of stimulants exacerbated long-​standing SE behav- behavior on the part of any sensitive parent. Yet,
iors. This demonstrates that although SE may not given the essential role that parents play on the
cause elevated levels of ADHD and vice versa, the establishment of eating habits and preferences in
two might be correlated or both might be caused by young children—​via provision of access to foods,
another factor. Also, it was noted that symptoms of feeding styles, reinforcement of child eating behav-
ADHD, such as not being able to sit still for long iors, and role modeling of their own eating—​while
or to focus on something to complete a given task, it is understandable that parental influences would
made it harder for SE to respond to treatment. The be implicated in the emergence and/​or maintenance
second research study examined a nationally repre- of SE, it is our contention that this claim is, at best,
sentative community-​based sample. The researchers unfounded, and, at worst, unjust.
had conflicting findings to those described previ- Yet, despite the potential potency of child fac-
ously, demonstrating that ADHD symptoms were tors eliciting parent feeding behaviors, the focus
associated with binging and/​or purging behaviors, on research and on lay publications regarding SE
but were not related to picky or restrictive eat- has been to implicate parents for not presenting
ing patterns (Bleck & DeBate, 2013). More stud- foods enough times or in fun enough ways. This is
ies are required to make any conclusions on this despite a lack of evidence demonstrating whether
relationship. more frequent presentation of a food results in food

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 431

acceptance, let  alone liking, in SE. Consider the to the phenomenology of SE. As previously men-
study by Wardle, Carnell, and Cook (2005). In a tioned, research on SE has focused on the role of
sample of 500 typically developing preschoolers in anxiety in contributing to food avoidance. This is
London, these authors reported that fruit and vege- important because this emphasis may have led to
table consumption by children was better explained inappropriate, or at least imprecise, guidance about
by the child’s food neophobia than any parent feed- the management of SE. In Figure 22.3 below, we
ing style. Another interesting study was conducted present some preliminary data in a sample of 4,000
by Harris et al. (Harris, Fildes, Mallan, & Llewellyn, adults with SE defined as described earlier (SE all
2016), in which the authors investigated whether the time and occurring since childhood) on the
the feeding style of parents differed between mono- relative role of anxiety relative to disgust in explor-
zygotic twins who were described as different in the ing associations with food avoidance. When we
level of food neophobia. Despite being reared in the controlled for the effects of anxiety, associations
same home environment, the parents describe very of texture, smell, and gagging were all significantly
different feeding styles for these children depending associated with disgust with effect sizes ranged
on their eating presentation. In fact, not only has from small to medium. When the reverse analyses
research on the number of presentations of a food were done, examining the role of anxiety control-
needed to promote acceptance been conducted in ling for disgust, associations of food avoidance to
typically developing children, but even this research smell, texture, and gagging were in the negligible
has been limited. The frequently cited number of to small effect size ranges. While these analyses are
10 to 15 presentations was conducted with typically exploratory, they point to the putative importance
developing children, and it was unclear the degree of disgust experience in explaining many impairing
to which acceptance of a food differs from liking features of SE.
or voluntarily seeking that food. Combined, this These exploratory findings also have putative
research has very important messages for parents of importance for the treatment or management of
a child with SE, the primary message being that we SE. There are numerous empirically validated treat-
really do not have adequate research to guide the ments for the management of anxiety disorders,
number presentations for food acceptance for SE in general, and specific phobias, in particular. In
and, second, that parents may have been unduly fact, individuals who practice behavior therapies
blamed for their child’s food selectivity. or cognitively emphasized behavior therapies have
developed exceedingly robust methods to help indi-
Disgust and Evaluative Conditioning viduals with specific phobias learn to approach their
The issue of the number of food presentations is fears and reduce their fear of stimuli that used to be
particularly relevant if disgust experience is essential

Disgust AND anxiety in SE

Associations with Disgust* Associations with Anxiety&

0.45 0.45
0.4 0.4
0.35 0.35
0.3 0.3
0.25 0.25
0.2 0.2
0.15 0.15
0.1 0.1
0.05 0.05
0 0
Smell Texture Gagging Smell Texture Gagging

&Controlling
*Controlling for Anxiety for Disgust

Figure 22.3  An exploratory analysis of 4,000 adults with selective eating of various severity that completed a Web-​based survey.
When controlling for anxiety, disgust was significantly associated with gagging, and taste and smell sensitivities with moderate or large
effect sizes. When the reverse analysis was done, anxiety was statistically or clinically insignificantly associated with gagging or sensory
sensitivities. 

432 Selective Eating
fear provoking. In contrast, there are no empirically only are these foods not a part of the regular diet of
validated treatments for the management of disgust. a formerly exposed selective eater, but these selective
The most robustly effective treatment for the eaters report having a particularly strong loathing of
management of specific phobias is exposure and these foods as a product of that aversive condition-
response prevention. This classic behavioral treat- ing experience.
ment is based on fundamental learning principles Recontextualization has been considered an
such as classical conditioning. In the theoretical alternative in the approach to disgust:  meaning
model that guides the development of this treat- literally, “to place (as a literary or artistic work)
ment, it is demonstrated that inherently fearful in a different context,” as sensations, specifically,
stimuli (the unconditioned stimulus such as a dog and emotions, more broadly, take on alternative
bite) becomes associated with neutral or harmless meanings depending on the context in which they
stimuli (the conditioned stimulus, a friendly, tooth- experienced (for example, a pounding heart in
less dog). The friendly dog becomes reliably associ- the context of running a race may be a part of a
ated with the unconditioned stimulus, the biting, very different experience than a pounding heart in
such that the sight of (or even the anticipation of ) the context of an academic test). In applying this
the friendly dog elicits fear and avoidance behav- logic to the experience of disgust, two examples are
iors such as refusal to visit homes or locales where salient. For example, parents have to interact with
a dog may be present. The focus of treatment is on lots of disgust-​eliciting stimuli in the service of car-
creating new learning experiences that establish ing and loving for their child (changing diapers,
competing associations with the toothless, friendly cleaning up vomit or urine, wiping noses, etc.).
dog. Frequent encounters with the friendly dog After hundreds of exposures (at least to feces), if
without any associated negative consequences cre- we asked the parent how she or he evaluated feces,
ate a new experience, the memory of which now most probably the answer would be “disgusting.”
competes for recollection with the original memory However, whether or not the feces are disgusting
of dogs as harmful, a process that has been referred is irrelevant. This is an act performed in the service
to as inhibitory learning (Craske, Treanor, Conway, of something much greater and more meaningful,
Zbozinek, & Vervliet, 2014). that of loving and caring for one’s child. Thus, this
Disgust in contrast, is considered a form of raises the intriguing possibility that the technol-
evaluative conditioning:  Individuals make judg- ogy of values clarification and value-​guided action
ments about whether things are liked or disliked, might have a particularly salient place for the man-
good or bad. Evaluative processes are thought to be agement of SE. We return to this topic again when
more resistant to new learning (or, in the parlay of we talk about treatment.
former learning models, to extinction). Consider a If disgust is an essential component to the expe-
coworker whom you find irritating. According to rience of those with SE, then prior work suggesting
an exposure model, frequent encounters with this that SE is caused by limited exposure to new foods
individual would provide opportunities for new might be misguided. It is important to note that
learning to occur that would compete with your there has been very little research investigating the
conceptualization of this person as irritating. Yet number of presentations it takes for typically devel-
exposure in the context of anxiety is built around oping child to accept a new food, even less known
the assumption that the beliefs are irrational, or, at about the number presentations for typically devel-
least, of a low probability and thus disconfirming oping child to like a new food, and finally, to our
experiences can be encountered. However, in the knowledge, no studies that addressed this ques-
case of evaluative conditioning, we are focused on tion in children, precisely defined with SE. First,
altering an evaluation that can be construed as a as noted, given the problems with definition cited
high-​probability event (i.e., there is a high probabil- earlier, it is unknown whether strategies that seem
ity that you will find your coworker irritating) and to be effective for typically developing children will
repeated experiences might not only not provide generalize to children with SE. It might be the case
disconfirming evidence, but might actually worsen that given the experience of disgust, a larger number
your original conceptualization (with repeated of presentations are needed, or, as suggested earlier,
encounters, your coworker might become increas- it might be that the manner or process of presen-
ingly more annoying). This description actually fits tation is the relevant factor rather than frequency
the anecdotal descriptions of selective eaters who in managing disgust—​or, more likely, that both are
have been “made” to eat foods by their parents. Not essential.

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 433

 Second, there has been scant research that distin- adventures that involve tasting foods (i.e., the trying
guishes between acceptance of a food and preference of new foods). Our reasoning is that the anticipa-
for that food. This distinction might be particularly tion of a disgust experience (as when the child tastes
critical for those with SE, whose choices seem to be the new food), would lead to aversive condition-
constrained to foods that are “the best” of a category ing of family mealtimes and further detract from
rather than just “good enough.” Thus, in our clini- the potential joys of social meals. Rather, the initial
cal experience, altering a child with SE’s willingness focus of family meals is increasing the frequency of
to taste a new food is a very different process from these eating occasions. Then, more subtle aspects of
the ability to incorporate that food into the child’s food exposures (e.g., the sight and smells of foods)
daily eating routine. Research distinguishing these can occur combined with the other advantages of
processes in SE has not been conducted. social meals (manners, social skills, communication
skills, etc.). This strategy also alleviates the guilt of
The Management of Selective Eating parents who may feel pressured into trying to get
Our narrative about the phenomenology of SE their child to eat a new food when the opportunity
guides our tentative recommendations for treatment arises. Table 22.1 lists the eventual goals of a family
elements of SE. We consider four domains: medical meal. The goal is one of recontextualizing the family
assessment, the family meal environment, values mealtime environment so that eating is occurring in
clarification, and mindfulness-​based sensory-​guided a context of relaxed acceptance of whatever eating
food exposure. We employ the context of “Food behavior unfolds.
Scientists” to capitalize on the “sensory superpow- Parents are helped to shape positive mealtime
ers” of these children, and use that framework to behavior by looking for small instances of adaptive
engage in mindful interactions with food itself, and social eating. However, there may be many prelim-
in the social context that surround the act of eating. inary steps to get to this point. For example, the
smells of others’ foods may be so aversive, that ini-
Medical Management tial work with the child is needed that employs our
As mentioned earlier, pediatricians and general metaphor of the Food Scientist. Within this frame-
practitioners are in a tough spot when it comes to work we may investigate the sensory habituation of
the management of SE. Given the imprecise defini- smells (objectively describing how the intensity of
tions, high prevalence associated with this impreci- smells changes or fluctuates over time). It is impor-
sion, and lack of treatment guidelines, pediatricians tant to note that these and all other exposure experi-
do not have any resources. Often, the appropriate ences are conducted from an acceptance framework,
focus is minimizing parent anxiety and avoiding meaning that the goal of these exposures is not to
unnecessary medical expenditures by conducting terminate an experience or demonstrate that some-
uninformative tests (Kerzner et  al., 2015). Satter thing is better or worse than one anticipated, but
(2000) provides an excellent framework to guide rather to have an experience and not let the nature
therapeutic decisions in primary care for feeding
disorders, so we do not belabor those points here.
However, we note that, clinically, we have observed Table 22.1  Guidelines for Conducting Family Meals
undiagnosed allergies, anemia, and food intoler- when a Child is a Selective Eater
ances that greatly advanced our management of
subsequent eating problems. Thus, if the child has • The SE is not expected to try new foods at
been always/​ almost always selective, particularly mealtimes
if the duration is above 2  years, then a discussion • The SE is expected to join the family at dinner and to
with one’s physician about such medical rule-​outs remain at the table until everyone is finished.
is appropriate.
• If the SE does initiate trying a new food at a
Mealtime Environment: Behavior mealtime, he earns bonus points.
Given the protective nature of family meals, one • The SE is served a very small portion of the same
of the first objectives of management of SE is to foods as the other family members on a separate
restore the sanctity of family meals. In contrast to small plate.
folklore recommendations of making the child “take
• The SE is not expected to try new foods at
a bite of a new food in order to leave the table,” we
mealtimes
do not advise using family meals as a time for food

434 Selective Eating
of that experience interfere with things that are illness. As such, the technologies to help indi-
meaningful and to use it as a way for individu- viduals become emotionally connected to deeply
als to learn more about themselves. One adopts a felt aspirations are evolving. As one can imag-
nonevaluative descriptive frame to increase the viv- ine, this task is eminently more challenging for
idness of experiences and learn something about children and adolescents, whose capacity to think
oneself and one’s body. Exposures are experiences to abstractly and ability to be influenced by future
increase self-​knowledge and increase the approach events is only beginning to develop. Yet, in our
in a valued life direction. work, having individuals describe and be vali-
The only caveat to these recommendations is dated for the journey and to dignify the hard work
that we try to help family members avoid provid- they are about to undertake has been a seemingly
ing any feedback during the meal itself. In this way, vital aspect of the treatment package. Without
the focus is on the social aspects of the meal and this, the child is merely trying aversive foods.
the meal environment does not increase the selec- With it, they are pushing through a challenge
tive eater’s self-​consciousness around eating. These to attain greatness. We employ all the usual sus-
strategies are all intended to change the context of pects in this regard:  imagery; creating pie charts,
meals so these events are experienced as pressure-​ the slices of which define the components of their
free, engaging, and vitality-​ enhancing events in self-​
definition; using proximal behavioral goals
which families spend time together—​not events to and rewards that are steps on this valued path. All
consume one’s vegetables. of this, of course, needs empirical support. In the
meantime, we find that children feel understood
Mealtime Environment: Food and more willing to do challenging things.
Yet, individuals do eat at family meals and know-
ing what to serve is of paramount importance to Sensory-​Based Exposures
families trying to assist a selective eater. In this An essential aspect of therapeutic eating experi-
regard, we respect and adhere to tenets espoused by ences for selective eaters is that the relative evalu-
Ellen Satter in her trust model and use of division ation of the food, whether it is liked or disliked,
of responsibility, with some slight modifications is irrelevant. Rather, the goal is to encourage the
(Satter, 2000). According to this model, parents individuals with SE to employ their sensory acuity
are in charge of establishing a mealtime schedule, to notice all aspects of a food without judgment or
deciding what foods to serve, and preparing appro- bias. The goal for a food adventure can sometimes
priate meals. Satter advises parents to ensure that be just that:  to have a novel food experience and
at least one item that is served is “safe” for every- notice mindfully (present-​focused, objective/​non-
one. Thus, while everyone’s plate might look dif- judgmental, one-​ mindfully) the sensory aspects
ferent, each plate is composed of at least one of of food stimuli. An added framework given our
the same elements. We adhere to the spirit of these guiding metaphor of a Food Scientist, is that of
instructions, following the tenets of the division of observing the result of a manipulation, investigat-
responsibility, ensuring that one item that is safe for ing which brand is the saltiest, and so forth. The
everyone is served, but we give parents’ permission goal of these exercises is to have a hypothesis and to
to “short-​order cook” menu items when nutritional investigate it; however, the contextual goal is to use
deficiencies are an issue. Meals are thus a combina- the extended metaphor of scientific exploration to
tion of challenging foods and safe foods including create a context of curiosity and objectivity around
an item that is safe for everyone. food investigations. These investigations are done
in-​
session, with homework to practice the same
Values Clarification in Selective Eating food investigations at home. To facilitate generaliza-
Given fundamental differences in evaluative tion, at-​home practices can be facilitated via video-​
versus fear conditioning, an essential aspect of phone interfaces or home-​based sessions.
intervention is helping individuals with SE to
define those vital aspects of their imaginable or Growing Out of Selective Eating
desired experience that would make engaging in In the research reviewed earlier, we offered some
really challenging activities (i.e., trying poten- definitions that may help to define a group of selec-
tially aversive foods) worth it. Value-​guided action tive eaters who do not “grow out of” this condi-
is an increasingly essential aspect of so-​ called tion but rather persist in this pattern of eating over
third-​generation behavioral approaches to mental years. This raises an interesting question:  How do

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 435

individuals grow out of SE? While there has been Blissett, J., & Fogel, A. (2013). Intrinsic and extrinsic influences
an absence of reliable predictors of persistence, our on children’s acceptance of new foods. Physiology & Behavior,
121, 89–​95.
conceptualization about the phenomenology of Cano, S. C., Tiemeier, H., Van Hoeken, D., Tharner, A., Jaddoe,
SE may offer some potential hypotheses about this V. W. V., Hofman, A.,  . . .  Hoek, H. W. (2015). Trajectories
remittance and persistence. One hypothesis is that of picky eating during childhood:  A  general population
children who are mild or moderate selective eaters study. International Journal of Eating Disorders, 48, 570–​579.
eventually undergo developmental changes in sen- Carruth, B. R., Ziegler, P. J., Gordon, A., & Barr, S. I. (2004).
Prevalence of picky eaters among infants and toddlers and
sory experience that permit both the introduction of their caregivers’ decisions about offering a new food. Joural of
novel foods and increased preferences for a broader the American Dietetic Association, 104, S57–​S64.
range of foodstuffs. An alternative (or complemen- Chatoor, I. M. (2002). Feeding disorders in infants and tod-
tary hypothesis) is that through the course of their dlers: Diagnosis and treatment. Child and adolescent psychiat-
development, individuals with mild or moderate SE ric clinics of North America, 11(2), 163–​183.
Cloutier, C. J., Kavaliers, M., & Ossenkopp, K. P. (2016).
encounter facets of their life experience that make Rodent sex differences in disgust behaviors (anticipatory
it meaningful for them to endure aversive sensory nausea) conditioned to a context associated with the effects
experiences in the service of more meaningful life of the toxin LiCl:  Inhibition of conditioning following
goals (e.g., as in the case of the nurse enduring the immune stimulation with lipopolysaccharide. Pharmacology,
smell of gangrene to care for a patient). This change Biochemistry, and Behavior.
Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., &
in framework allows greater sampling of foodstuffs Vervliet, B. (2014). Maximizing exposure therapy: An inhib-
and thus greater probabilities that foods will be dis- itory learning approach. Behaviour Research and Therapy,
covered that are preferred. 58, 10–​23.
What are the barriers in severe SE? We offer two Curtis, V. (2011). Why disgust matters. Philosophical Transactions
hypotheses for this. First, it could be that the sen- of the Royal Society of London Series B, Biological Sciences, 366,
3478–​3490.
sory experiences and resulting disgust experience are Darwin, C. R. (1872). The expression of the emotions in man and
so potent that it is more challenging or seemingly animals (1st ed.). London: John Murray.
insurmountable to overcome, even in the service de Krom, M., Bauer, F., Collier, D., Adan, R. A., & la Fleur, S.
of valued life activities. Alternatively, or intrigu- E. (2009). Genetic variation and effects on human eating
ingly, there could be fundamental deficits in the behavior. Annual Review of Nutrition, 29, 283–​304.
Eisenberg, M. E., Olson, R. E., Neumark-​Sztainer, D., Story,
rewarding value of alternative sources of reinforce- M., & Bearinger, L. H. (2004). Correlations between fam-
ment, such that finding activities that are rewarding ily meals and psychosocial well-​being among adolescents.
enough to compete with these aversive experiences Archives of Pediatrics and Adolescent Medicine, 158, 792–​796.
is challenging. Exploring these barriers will be a fas- Farrow, C. V., & Coulthard, H. (2012). Relationships between
cinating avenue of research. sensory sensitivity, anxiety and selective eating in children.
Appetite, 58, 842–​846.
Fulkerson, J. A., Story, M., Mellin, A., Leffert, N., Neumark-​
Summary Sztainer, D., & French, S. A. (2006). Family dinner meal
This chapter has conceptualized SE as a clinical frequency and adolescent development:  Relationships with
condition. From this framework, existing research developmental assets and high-​ risk behaviors. Journal of
needs to be reevaluated by examining the extremes, Adolescent Health, 39, 337–​345.
Harris, H. A., Fildes, A., Mallan, K. M., & Llewellyn, C. H.
exploring predictors of SE persistence, and exam- (2016). Maternal feeding practices and fussy eating in tod-
ining how sensory sensitivity and neurocognitive dlerhood: A discordant twin analysis. International Journal of
functioning might explain persistence of SE. It is Behavioral Nutrition and Physical Activity, 13, 81.
our intention with this chapter to spur novel con- Jacobi, C., Agras, W. S., Bryson, S., & Hammer, L. D. (2003).
ceptualizations of this condition that might pro- Behavioral validation, precursors, and concomitants of picky
eating in childhood. Journal of the American Academy of Child
mote less blaming of the social environment and and Adolescent Psychiatry, 42, 76–​84.
increased appreciation for the complex biology that Kauer, J., Pelchat, M. L., Rozin, P., & Zickgraf, H. F. (2015).
promotes protection of an organism in the face of Adult picky eating:  Phenomenology, taste sensitivity, and
unknown and unseen pathogens. psychological correlates. Appetite, 90, 219–​228.
Kerzner, B., Milano, K., MacLean, W. C., Berall, G., Stuart,
S., & Chatoor, I. (2015). A practical approach to classi-
References fying and managing feeding difficulties. Pediatrics, 135,
Bleck, J., & DeBate, R. D. (2013). Exploring the co-​morbidity of
344–​353.
attention-​deficit/​hyperactivity disorder with eating disorders
Lask, B., & Bryant-​Waugh, R. (1999). Anorexia nervosa and
and disordered eating behaviors in a nationally representative
related eating disorders in childhood and adolescence (2nd ed.).
community-​based sample. Eating Behaviors, 14, 390–​393.
East Sussex, UK: Psychology Press.

436 Selective Eating
Machado, B. C., Dias, P., Lima, V. S., Campos, J., & Gonçalves, Taylor, C. M., Wernimon, S. M., Northstone, K., & Emmett,
S. (2016). Prevalence and correlates of picky eating in P. M. (2015). Picky/​fussy eating in children: Review of defi-
preschool-​aged children:  A  population-​based study. Eating nitions, assessment, prevalence and dietary intakes. Appetite,
Behaviors, 22, 16–​21. 95, 349–​359.
Mascola, A. J., Bryson, S. W., & Agras, W. S. (2010). Picky eat- Tepper, B. J. (2008). Nutritional implications of genetic taste
ing during childhood: A longitudinal study to age 11 years. variation: The role of PROP sensitivity and other taste phe-
Eating Behaviors, 11, 253–​257. notypes. Annual Review of Nutrition, 28, 367–​388.
Micali, N., Rask, C. U., Olsen, E. M., & Skovgaard, A. M. Tharner, A., Jansen, P. W., Kiefte-​de Jong, J. C., Moll, H.
(2016). Early predictors of childhood restrictive eat- A., van der Ende, J., Jaddoe, V. W. V.,  . . .  Franco, O. H.
ing: A population-​based study. Journal of Developmental and (2014). Toward an operative diagnosis of fussy/​picky eat-
Behavioral Pediatrics, 37, 314–​321. ing: a latent profile approach in a population-​based cohort.
Micali, N., Simonoff, E., Elberling, H., Rask, C. U., Olsen, E. M., International Journal of Behavioral Nutrition and Physical
& Skovgaard, A. M. (2011). Eating patterns in a population-​ Activity, 11, 11.
based sample of children aged 5 to 7 years: Association with US Department of Health and Human Services, US Department
psychopathology and parentally perceived impairment. of Agriculture. (2015, December). Dietary guidelines for
Journal of Developmental and Behavioral Pediatrics, 32, Americans 2015–​2020 (8th ed.).
572–​580. van der Horst, K., Deming, D. M., Lesniauskas, R., Carr, B. T., &
Neumark-​Sztainer, D., Story, M., Ackard, D., Moe, J., & Perry, Reidy, K. C. (2016). Picky eating: Associations with child eat-
C. (2000). The “family meal”: Views of adolescents. Journal ing characteristics and food intake. Appetite, 103, 286–​293.
of Nutrition Education, 32, 329–​334. Wardle, J., Carnell, S., & Cooke, L. (2005). Parental control
Nicholls, D., Christie, D., Randall, L., & Lask, B. (2001). over feeding and children’s fruit and vegetable intake: How
Selective eating:  Symptom, disorder or normal variant. are they related?. Journal of the American Dietetic Association,
Clinical Child Psychology and Psychiatry, 6, 257–​270. 105(2), 227–​232.
Olatunji, B. O., Williams, N. L., Tolin, D. F., Abramowitz, J. Wildes, J. E., Zucker, N. L., & Marcus, M. D. (2012). Picky
S., Sawchuk, C. N., Lohr, J. M., & Elwood, L. S. (2007). eating in adults: Results of a Web-​based survey. International
The Disgust Scale:  Item analysis, factor structure, and Journal of Eating Disorders, 45, 575–​582.
suggestions for refinement. Psychological Assessment, 19, Williams, K. E., Gibbons, B. G., & Schreck, K. A. (2005).
281–​297. Comparing selective eaters with and without developmental
Pennell, A., Couturier, J., Grant, C., & Johnson, N. (2016). disabilities. Journal of Developmental and Physical Disabilities,
Severe avoidant/​restrictive food intake disorder and coex- 17, 299–​309.
isting stimulant treated attention deficit hyperactivity Wright, C. M., Parkinson, K. N., Shipton, D., & Drewett, R. F.
disorder. International Journal of Eating Disorders, 49, (2007). How do toddler eating problems relate to their eat-
1036–​1039. ing behavior, food preferences, and growth? Pediatrics, 120,
Satter, E. (2000). Child of mine: Feeding with love and good sense. e1069–​e1075.
Boulder, CO: Bull. Xue, Y., Lee, E., Ning, K., Zheng, Y., Ma, D., Gao, H.,   .  .  . 
Seema, S. & Aceves, M. (2016). Advances in GERD. Zhang, Y. (2015). Prevalence of picky eating behaviour in
Gastroenterology and Hepatology, 12, 516–​518. Chinese school-​age children and associations with anthro-
Spergel, J. M., & Shuker, M. (2008). Nutritional management of pometric parameters and intelligence quotient:  A  cross-​
eosinophilic esophagitis. Gastrointestinal Endoscopy Clinics of sectional study. Appetite, 91, 248–​255.
North America, 18, 179–​194; xi. Zelazniewicz, A., & Pawlowski, B. (2015). Disgust in pregnancy
Stevenson, R. J., Case, T. I., & Oaten, M. J. (2011). Proactive and fetus sex:  Longitudinal study. Physiology & Behavior,
strategies to avoid infectious disease. Philosophical 139, 177–​181.
Transactions of the Royal Society of London Series B, Biological Zimmer, M. H., Hart, L. C., Manning-​Courtney, P., Murray,
Sciences, 366, 3361–​3363. D. S., Bing, N. M., & Summer, S. (2012). Food variety
Stevenson, R. J., Hodgson, D., Oaten, M. J., Moussavi, M., as a predictor of nutritional status among children with
Langberg, R., Case, T. I., & Barouei, J. (2012). Disgust autism. Journal of Autism and Developmental Disorders, 42,
elevates core body temperature and up-​ regulates certain 549–​556.
oral immune markers. Brain, Behavior, and Immunity, 26, Zucker, N. L., Copeland, W., & Egger, H. (2016). Authors’
1160–​1168. response. Pediatrics, 137, doi: 10.1542/​peds.2015-​3635B.
Taylor, C. M., Northstone, K., Wernimont, S. M., & Emmett, P. Zucker, N., Copeland, W., Franz, L., Carpenter, K., Keeling,
M. (2016). Picky eating in preschool children: Associations L., Angold, A., & Egger, H. (2015). Psychological and psy-
with dietary fibre intakes and stool hardness. Appetite, 100, chosocial impairment in preschoolers with selective eating.
263–​271. Pediatrics, 136, e582–​590.

Zucker, Arena, Dable, Hill, Hubble, Sohl, Yoon 437

 C H A PT E R


Emerging Syndromes
23
Kelly C. Allison and Jennifer D. Lundgren

Abstract
The Diagnostic and Statistical Manual, fifth edition, of the American Psychiatric Association (2013) has
designated several disorders under the diagnosis of otherwise specified feeding and eating disorder
(OSFED). This chapter evaluates three of these, night eating syndrome (NES), purging disorder
(PD), and atypical anorexia nervosa (atypical AN). It also reviews orthorexia nervosa, which has
been discussed in the clinical realm as well as the popular press. The history and definition for each is
reviewed, relevant theoretical models are presented and compared, and evidence for the usefulness of
the models is described. Empirical studies examining the disorders’ independence from other disorders,
comorbid psychopathology, and, when available, medical comorbidities, are discussed. Distress and
impairment in functioning seem comparable between at least three of these emerging disorders and
threshold eating disorders. Finally, remaining questions for future research are summarized.
Key Words:  OSFED, night eating syndrome, purging disorder, atypical anorexia, orthorexia nervosa

Introduction Fuller (1990) proposed five criteria as necessary for

The Diagnostic and Statistical Manual, 5th inclusion as a psychiatric disorder:
Edition (DSM-​ 5) of the American Psychiatric
Criterion I: “There should be at least 50 journal
Association (2013) produced a new list of otherwise
articles on the proposed diagnostic category in the
specified feeding and eating disorders (OSFED)
last 10 years,  . . .  at least 25 of them should be
since its last edition, and an even more general cat-
empirical”  . . . . Criterion II: That there be (a) a set
egory of unspecified feeding and eating disorders
of diagnostic criteria which (b) include self-​report
(UFED). Previously, binge eating disorder (BED)
measures, structured interviews, and rating scales  . . . .
fell into the eating disorder not otherwise specified
Criterion III: There should be at least two empirical
(EDNOS) category, but this is now a full thresh-
studies by independent research groups demonstrating
old eating disorder (ED), joining anorexia nervosa
high inter-​clinician correlations  . . . . Criterion
(AN) and bulimia nervosa (BN). In its place are sev-
IV: The proposed category represents a syndrome
eral syndromes or variants of the standalone EDs,
of frequently co-​occurring symptoms, described
including night eating syndrome (NES), purging
in at least two independent studies  . . . . Criterion
disorder (PD), and atypical anorexia nervosa (atypi-
V: There should be at least two independent empirical
cal AN). Not listed in the DSM-​5 but certainly rep-
studies showing that the proposed category “can be
resented in the literature and encountered in clinical
differentiated from other categories with which it is
practice is orthorexia nervosa (ON), which is also
likely to be confused.” (p. 17)
considered in this chapter.
During the DSM-​IV (American Psychiatric These guidelines, as well as calls for demon-
Association, 2000) revision, Blashfield, Sprock, and stration of empirical approaches for classification

438
(Wonderlich, Joiner, Keel, Williamson, & Crosby, Hirsch, 2006), 12% in two university outpatient
2007), were helpful in evaluating syndromes pre- psychiatric clinics (Lundgren et  al., 2006), and
sented in this chapter. We review the history, def- 8.6% of patients from a German sleep apnea clinic
inition, and relevant theoretical models for each, (Olbrich et al., 2009). Thus, prevalence rises among
outline evidence for the usefulness of the models, treatment-​seeking populations.
and provide a summary of remaining questions for
future research. Current Research Diagnostic Criteria
Recognizing that widely accepted diagnostic
Research Status and Description of Night criteria for NES were needed in order to advance
Eating Syndrome our understanding and treatment of the syndrome,
Although night eating syndrome (NES) has been professionals from the ED, sleep, and obesity fields
recognized for decades, research and clinical char- convened the First International Night Eating
acterization of NES lacks longevity. The history, Symposium in April 2008. Table 23.1 shows the
prevalence, and research diagnostic criteria for NES criteria that were established by consensus (Allison,
are reviewed below. Lundgren, O’Reardon et al., 2010). They built on
the two core criteria, that is, a delayed circadian pat-
History and Prevalence tern of food intake manifested by evening hyper-
Night eating syndrome (NES) was first described phagia and/​or nocturnal ingestions (Criterion I).
in 1955 as a disorder of morning anorexia, evening Criterion II requires awareness of the eating epi-
hyperphagia, and insomnia, usually accompanied sodes to differentiate it from somnambulistic eat-
by depressed mood and stressful life circumstances ing typical of sleep-​related eating disorder (SRED).
(Stunkard, Grace, & Wolf, 1955). Night eating Criterion III lists specifiers that have been con-
syndrome did not receive much research or clinical sistently associated with NES, three of which are
attention until the 1990s, coinciding with increas- required for diagnosis. Criteria IV, V, and VI require
ing rates of obesity and the search for factors related distress or impairment in functioning, duration of
to excessive weight gain. In 1999, awakenings with the symptoms of at least 3 months, and a rule-​out
ingestions (nocturnal ingestions) were added to those of other primary conditions that may be causing the
criteria originally described in 1955, and were pub- night eating.
lished in a provisional set of criteria (Birketvedt
et  al., 1999). However, as research advanced our Models of Night Eating Syndrome
understanding of NES, the diagnostic criteria for As NES has components of eating, sleep, and
NES used by researchers often changed, making mood disorders, it has been conceptualized in sev-
comparisons across studies difficult. eral ways. These models include NES as (1) a dis-
Prevalence of NES in the general population tinct disorder, (2)  an extension of daytime EDs,
has been reported between 1.5% and 5.8% (Colles, (3)  on a continuum with sleep disorders, (4)  a
Dixon, & O’Brien, 2007; de Zwaan, Müller, Allison, variant of obesity, and (5)  secondary to other
Brähler, & Hilbert, 2014; Lamerz et al., 2005; Rand, psychopathology.
Macgregor, & Stunkard, 1997; Striegel-​ Moore,
Franko, Thompson, Affenito, & Kraemer, 2006; Night Eating Syndrome as a Distinct
Tholin et al., 2009), with variance across the stud- Disorder
ies due in part to the different criteria and assess- While the merits of NES as a psychiatric disor-
ment methods used. Prevalence estimates in special der have been questioned (Striegel-​Moore, Franko,
populations suggest ranges of 6% (Stunkard et al., et al., 2006), the evidence for its distinction as an
1996) to 16% (Adami, Campostano, Marinari, independent construct has grown. A  number of
Ravera, & Scopinaro, 2002) in weight loss samples empirical studies have described the unique eating
of class  I  and II obesity. Among bariatric surgery patterns and related behaviors associated with NES,
candidates, the range in prospective interview including item response theory analysis, circadian
studies is 9% (Allison et al., 2006) to 27% (Rand analyses of eating and neuroendocrine factors, and
et al., 1997). A prevalence of 3.8% has been found latent class analysis.
among older adults in a large multicenter study of An item response theory analysis was performed
type 2 diabetes (Allison et al., 2007), 7% and 9.7% on 1,481 Night Eating Questionnaires (NEQ;
of patients in diabetes clinics (Hood, Reutrakul, & Allison, Lundgren, et al., 2008), a screening tool for
Crowley, 2014; Morse, Ciechanowski, Katon, & night eating symptoms, gathered across six studies,

Allison, Lundgren 439

Table 23.1  Research Diagnostic Criteria for Night Eating Syndrome
I.  The daily pattern of eating demonstrates a significantly increased intake in the evening and/​or nighttime, as
manifested by one or both of the following:
A.  At least 25% of food intake is consumed after the evening meal
B.  At least two episodes of nocturnal eating per week

II.  Awareness and recall of evening and nocturnal eating episodes are present.

III.  The clinical picture is characterized by at least three of the following features:

A.  Lack of desire to eat in the morning and/​or breakfast is omitted on four or more mornings per week
B.  Presence of a strong urge to eat between dinner and sleep onset and/​or during the night
C.  Sleep onset and/​or sleep maintenance insomnia are present four or more nights per week
D.  Presence of a belief that one must eat in order to initiate or return to sleep
E.  Mood is frequently depressed and/​or mood worsens in the evening

IV.  The disorder is associated with significant distress and/​or impairment in functioning.

V.  The disordered pattern of eating has been maintained for at least 3 months.

VI.  The disorder is not secondary to substance abuse or dependence, medical disorder, medication, or another
psychiatric disorder.

Reprinted with permission from Wiley Periodicals, Allison, K. C., Lundgren, J. D., O’Reardon, J. P., Geliebter, A., Gluck, M. E., Vinai, P.,
et al. Proposed diagnostic criteria for night eating syndrome. International Journal of Eating Disorders, 2010.

including two of NES participants, one of con- were significantly smaller among the NES group
trol participants, and three of special populations than controls, suggesting that “morning anorexia”
(e.g., bariatric surgery candidates, psychiatric clinic extends across the lunch period, as well.
patients, and overweight older adults with type 2 Similarly, Goel and colleagues (2009) modeled the
diabetes) (Allison, Engel, et al., 2008). Endorsement circadian aspects of food intake, macronutrient con-
of evening hyperphagia and/​ or nocturnal inges- tent, and neuroendocrine measures among a sample
tions, initial insomnia, and nighttime awakenings of 15 female night eaters and 14 control inpatient
showed high precision in identifying the night eat- study participants. They found that energy (calories),
ing construct, while reports of morning anorexia and carbohydrate, and fat intake were significantly delayed
delayed morning meal did not provide additional by about 1.5 hours, and protein was nonsignificantly
meaningful information. It seemed that lack of delayed by about 0.5 hour. In addition, phase delays
appetite in the morning was common not just in the for food-​regulatory mechanisms of 1.0 to 2.8 hours
two night eating samples (71.4% and 75.5%), but were found for leptin and insulin and in the circadian
also among the other four samples (24.5%, 59.2%, melatonin rhythm (which regulates the sleep period).
58.8%, and 20%). This finding helped inform the Circulating levels of ghrelin was the only rhythm to
decision for the new diagnostic criteria to include show a phase advance at 5.2 hours, while the glucose
morning anorexia as one of five specifiers, three of rhythm showed an inverted circadian pattern (a delay
which are required, but not as a required criterion. of 12.4 hours). These findings suggest that NES may
Two studies have effectively shown the delay result from dissociations between the central tim-
in eating patterns experienced in NES. Boston, ing mechanism, that is, the suprachiasmatic nucleus
Moate, Stunkard, Allison, and Lundgren (2008) (SCN, the body’s “master clock” that controls the tim-
modeled 24-​hour caloric intake showing that per- ing of many basic functions) and peripheral oscillators
sons with NES (n = 148) had significantly delayed (e.g., the stomach or liver), which help control timing
meals that were spread out over a greater period of basic bodily functions independent of the SCN.
of time than those of controls (n  =  68). Overall, With core features of NES identified by advanced
there was less structure to their eating. In addi- statistical methods, a population-​based analysis of
tion, participants with NES consumed significantly the classification of putative persons with NES was
more calories across the 24 hours than the controls also reported. A large community sample of 8,250
(2,555 vs. 2,229 kcal, respectively, p < .05). Boston persons ages 15 to 39 was studied by Striegel-​Moore
et al. (2008) showed that both breakfast and lunch and colleagues (2008) using latent class analysis to

440 Emerging Syndromes

identify the typologies of NES, and 2,068 persons food. Energy intake from 11 p.m. to 5 a.m. did not
who reported night eating behavior were included differ between the groups. More recently, Schreiber-​
in the analysis. Food intake patterns were assessed Gregory et  al. (2013) used ecological momentary
via 24-​hour food recalls; nocturnal ingestions were assessment to track the time and day of binge epi-
not specifically assessed and, therefore, not repre- sodes among a group of nine women with BED.
sented in the classifications. A four-​class latent class The most common time was early afternoon, that
analysis solution was identified:  (1)  nondepressed is, 12 to 3 p.m., and the next most common time
late night eaters (eating after 11  p.m., skipping included the dinner hour, 6 to 9 p.m. Overall, the
breakfast, and consuming more than 50% of intake explanation that binge episodes are almost exclu-
after 7 p.m.); (2) nondepressed evening eaters (con- sively occurring in the evening does not seem to be
suming more than 50% of intake after 7 p.m. but no supported empirically.
eating after 11 p.m.); (3) depressed late night eaters Overlap between BED and NES in most studies
(eating after 11 p.m., with some eating more than ranges from 5% to 20% (Allison et al., 2007; Allison,
50% after 7 p.m., depression, disturbed sleep, tired, Grilo, Masheb, & Stunkard, 2005, Striegel-​Moore,
and loss of appetite); and (4) depressed evening eat- Dohm, et al., 2005; Stunkard et al., 1996). Reports
ers (same as group 2 plus depressed mood, feeling of overlap with BN are very limited. Tzischinsky
tired, and disturbed sleep). They also reported that and Latzer (2004) found during a 3-​year period
the late night eating was associated with high energy that 9% of BN and 16% of BED patients in an out-
and sodium intake and low protein intake. High patient ED clinic reported nocturnal ingestions. In
body mass index (BMI) was not related to night addition, a pilot study (Lundgren, Shapiro, & Bulik,
eating in this sample, but those with night eating 2008) reported that 35.5% of a BN clinical sample
were more likely to be male, black, and younger endorsed consuming at least 25% of their caloric
than those without night eating. They also showed intake after dinner, and 19.3% reported eating at
that evening overeating was associated with sub- least half of their intake after dinner. Ten (38.7%)
stance use. Striegel-​Moore et al. (2008) concluded reported at least occasional nocturnal ingestions,
that the latent class analysis showed validity for a while 12.9% reported eating during awakenings at
definition of NES based on eating after 11 p.m., a least half of the time. This overlap of night eating
cut that seemed to discriminate the syndrome bet- behaviors with BN is high, and more studies are
ter than did 7 p.m. More latent class analysis studies needed to understand this relationship.
are needed using the new diagnostic criteria and in However, there are important differences in the
samples that span a larger age range to expand these core criteria of NES and other EDs that differenti-
findings. ate them. First, the average caloric intake consumed
during nocturnal ingestions is not objectively large,
Continuum with Other Eating Disorders but is similar to regular snacks at approximately
Binge eating disorder and BN represent the two 300 to 400 kcal (Allison, Stunkard, & Their, 2004;
EDs with the most potential for overlap with NES. Birketvedt et  al., 1999). Second, the circadian
In BED and BN, binge episodes often occur in delay in the 24-​hour pattern of eating that is core
the evening, after work or school, as this may be to the NES diagnosis (Boston et  al., 2008; Goel
the most opportune time for secretive overeating. et  al., 2009) has not been well documented in
However, not many studies have documented the BED or BN.
percentage of intake consumed after the evening Ancillary ED symptoms also appear to differ
meal in these groups or whether there is a delay between NES and other EDs. Alexithymia was
in the circadian pattern of eating in these EDs. measured in patients with NES as compared with
Raymond, Neumeyer, Warren, Lee, and Peterson controls, and no abnormal values or differences
(2003) provided one of the few 24-​hour dietary were found (Vinai et  al., 2015). Alexithymia is
recall studies that examined the timing of binge typically present among BED and the other EDs.
episodes. They found that the BED group (n = 12) Additionally, the presence of one’s belief that they
consumed significantly more in the evening (5 to need to eat to go to sleep/​resume sleep seems to
11  p.m.) on binge days than the control group represent a cognition specific to NES and has dif-
(n = 8) at 1,380 kcal versus 964 kcal, respectively. ferentiated those with NES from those with BED
While this difference was statistically significant, (Vinai et al., 2014).
a difference of approximately 400 kcal certainly Finally, some recent studies point to overlap
would not constitute an objectively large amount of between NES and BED while maintaining some

Allison, Lundgren 441

distinction. Meule, Allison, and Platte (2014) report that they either (1) began eating during the
showed that the presence of emotional eating night with little to no consciousness, but over time
moderated the relationship between night eating they gained awareness of the nocturnal eating or
symptoms, binge eating, and BMI, such that more (2)  they occasionally experienced nocturnal inges-
severe night eating was related to more frequent tions with little awareness, but for the majority of
binge episodes and higher BMI only at high levels ingestions they were awake and aware. While the
of emotional eating, but not at low levels of emo- disorders may be related, the contrasts certainly
tional eating. Root et al. (2010) also compared the suggest that they are different syndromes that may
heritability of NES and BED among participants sometimes co-​occur (Stunkard et al., 2009).
of the Swedish Twin Registry. They found that heri-
tability for BE was 0.74 (95% CI = [0.36, 0.93]) Continuum with Obesity
for men and 0.70 (95% CI  =  [0.26, 0.77]) for There is mixed evidence regarding the relationship
women, while heritability for night eating was 0.44 between NES and obesity. Night eating syndrome
(95% CI = [0.24, 0.61]) for men and 0.35 (95% has been described among persons of average-
CI = [0.17, 0.52]) for women. They also reported a weight, overweight, and obesity (Birketvedt et al.,
genetic correlation of 0.66 (95% CI = [0.48, 0.96]) 1999; Lundgren, Allison, O’Reardon, & Stunkard,
suggesting a high, but not complete, genetic over- 2008; Marshall, Allison, O’Reardon, Birketvedt, &
lap for liability to the disorders and leaving room Stunkard, et al., 2004). European epidemiological
for the contribution of environmental factors to studies (Andersen, Stunkard, Sorensen, Petersen, &
influence the expression of these disorders. Heitmann, 2004, Tholin et al., 2009) as well as clin-
ical studies (Aranoff, Geliebter, & Zammit, 2001;
Continuum with Sleep Disorders Colles et al., 2007; Lundgren, Allison, Crow, et al.,
The conceptual overlap between NES and 2006) have shown an increased risk for obesity among
sleep-​related eating disorder (SRED) has been less persons with NES, but two other studies based
explored than its overlap with other EDs. The key on American national databases did not show this
criterion for SRED is involuntary eating during increased risk (Striegel-​Moore et al., 2005; Striegel-​
the main sleep period (Sateia, 2005). As such, the Moore, Franko, Thompson, et al., 2006). These latter
main difference between NES and SRED is the studies were not designed to assess night eating specif-
level of awareness during eating episodes. Persons ically, but were based on 24-​hour food records.
who sleepwalk and eat are more likely to ingest odd It could be assumed that the repeated and per-
or nonfood items, such as shaving cream instead sistent nature of the disorder contributes to weight
of ice cream or cat food, and to injure themselves gain among those with NES. But, counterintuitively,
walking into obstacles or preparing foods (Schenck Lundgren and colleagues (2008) reported that non-
& Mahowald, 1994). In addition, they would be obese persons with night eating reported an even higher
less likely to have memory or recall of nocturnal proportion of eating after the evening meal (50%,
ingestions. Perhaps most importantly, treatment SD = 15%), as compared with 35% (SD = 10%)
differs for the disorders, with dopaminergic agents reported in an overweight and obese sample of persons
in combination with codeine or clonazepam show- with NES (O’Reardon et al., 2004). Lundgren, Allison,
ing efficacy for SRED, as compared with selective et al. (2008) also reported that the nonobese night
serotonin reuptake inhibitors for NES (see Howell, eaters reported more excessive exercise and daytime
Schenck, & Crow, 2009, for review). Preliminary dietary restraint in comparison with nonobese control
studies for the effectiveness of topiramate for both participants, although none of the participants met cur-
disorders is positive (Winkelman, 2003, 2006), but rent criteria for BN in that study. The compensatory
randomized controlled trials for confirmation are exercise and daytime restraint may help them keep their
needed. Finally, psychotherapies, such as cognitive-​ weight under control, while exaggerating the circadian
behavioral therapy, that seem promising for NES delay of their eating. Thus, overall, there is no sound
(Allison, Lundgren, Moore, O’Reardon, & Stunkard, evidence that NES is solely a phenotype of obesity.
2010) would not be effective for a parasomnia where
the eating behavior is truly involuntary. Night Eating Syndrome as Secondary
It is possible that SRED and NES occur along a to Other Psychopathology
continuum, as suggested by Howell and colleagues As in BED, high psychiatric comorbidity rates
(2009). Clinical experience of the authors of this have been associated with NES. Beck Depression
chapter suggests that a minority of NES patients Inventory scores are in the mild to moderate range

442 Emerging Syndromes

at 15.9 (SD = 10.6), which is comparable to scores on its unique circadian pattern of intake and lack of
of persons with BED (17.5, SD  =  9.0; Allison, binge eating behaviors. Night eating syndrome may
Grilo, et  al., 2005), BN (15, SD  =  11.5), and be found on a continuum with SRED, but several
PD (11, SD  =  9.2; Keel, Haedt, & Edler, 2005); features discriminate the two disorders, particularly
see Figure 23.1. Self-​esteem is also lower in NES different treatment approaches. Night eating syn-
patients seeking weight loss than in patients without drome is found across persons of all BMIs and has
NES (Gluck, Geliebter, & Satov, 2001). Lifetime not reliably been linked to obesity. Although per-
prevalence of other psychiatric disorders as assessed sons with NES have high comorbidity rates with
by the Structured Clinical Interview for the DSM other psychiatric disorders, particularly mood dis-
(SCID) is high among those with NES at 74%, as orders, this does not preclude NES from existing as
compared with controls at 18% (Lundgren, Allison, an independent diagnosis.
et  al., 2008). This figure is similar to that found Despite this evidence for NES as a distinct diag-
in persons with BED (70%) and BN (75%; Fink, nosis, much more information is needed to clarify
Smith, Gordon, Holm-​Denoma, & Joiner, 2009). remaining questions associated with its specific
Overall, though, comorbidity alone is not necessar- diagnostic features and its exact relationship with
ily a reason for exclusion as an independent diag- other eating and sleep disorders.
nostic construct.
Research Needed to Clarify the Status
Evidence of Diagnostic Validity and Clinical of Night Eating Syndrome
Significance Using Models of Night Eating With the establishment of research diagnostic
Syndrome criteria for NES, new studies are needed to vali-
Evidence from studies of differing method- date those criteria and to investigate the usefulness
ologies supports a distinct classification for NES, of their parameters. First, the operational defini-
which can be discriminated from other EDs based tion of evening hyperphagia has been set at 25%

20
Controls
18
BED
16 NES
PD
14
BN
12
Score

10

0
Dietary Eating Shape Weight Cognitive Disinhi- Hunger BDI
Restraint Concern Concern Concern Restraint bition

Figure 23.1  Disordered eating and depressed mood levels depicted across controls and groups with binge eating disorder, night eating
syndrome, purging disorder, and bulimia nervosa. No statistical comparisons are provided, given the numbers are generated from
different studies, but the levels support the repeated findings that all of these groups experience clinical levels of disordered eating and
depressed mood in comparison with control groups and that their levels differ among the diagnoses in an inconsistent manner. 
Data for the control, BED, and NES groups were extracted from Allison, Grilo, Masheb, & Stunkard, 2005. Data for the PD and BN groups were
extracted from Keel, Haedt, & Edler, 2005.

Allison, Lundgren 443

of daily caloric intake consumed after the evening been included in the new criteria set for NES.
meal. The “evening meal” was chosen as a marker However, it has not been formally included in most
instead of a specific time, because the dinner hour NES studies to date, so assessing its presence is
varies across cultures. O’Reardon et  al. (2004) encouraged.
found that night eaters who had verbally reported Not only are the psychosocial correlates of a psy-
consuming at least half of their intake after dinner chiatric disorder important but also, increasingly,
were actually eating about 35% (SD  =  10%), as the health implications of a disorder are essen-
reported through 7-​day food logs. Control partici- tial in establishing the importance of a diagnostic
pants were consuming 10% (SD  =  7%). One SD entity. While neuroendocrine abnormalities have
below the NES participants’ and two SD’s above been noted (Allison, Ahima, et al., 2005; Goel et
the control participants’ proportion of nighttime al., 2009), it is not clear how these abnormalities
food intake converges at 25%, creating a useful are related to disease and mortality. However, night
separation point between the groups. More stud- eating among diabetic patients has been associated
ies should assess the clinical utility of this cut-​ with less adherence to diet, exercise, and glucose
point in different samples. monitoring as well as higher hemoglobin A1C lev-
More research is also needed to test the clini- els and more diabetic complications (Morse et al.,
cal utility of the frequency of nocturnal ingestions, 2006). Also, a review of the impact of night eat-
which were set at two per week to match the fre- ing on bariatric surgery outcomes (Colles & Dixon,
quency of binges required in the diagnoses of BN 2006) revealed that night eating behaviors often
and BED. Previous studies either did not specify a continue after surgery, and at least one study has
number (Birketvedt et al., 1999) or used three per shown that night eating after surgery is associated
week (Lundgren, Allison, et  al., 2008, O’Reardon with greater BMIs and less treatment satisfaction
et al., 2006). There are no data on the relationship (Latner, Wetzler, Goodman, & Glinski, 2004).
between frequency of nocturnal ingestions and lev- Finally, Andersen and colleagues (2004) found that
els of clinical distress. women, but not men, who endorsed night eating
As the criteria stand, only one of the core criteria were more likely to gain weight over 6 years as com-
is needed for diagnosis. This makes conceptual sense pared to women without night eating. In general,
as both evening hyperphagia and nocturnal inges- more careful assessment of night eating behaviors
tions are expressions of a delayed circadian pattern is needed in relation to health outcomes, which
of eating. However, more data are needed on the would be aided by the development of new assess-
timing of binge episodes for BN and BED to under- ment tools based on the proposed diagnostic crite-
stand if an “evening hyperphagia subtype” is valid ria. Longitudinal studies would also help identify
in an ED population. the impact of NES on the development of disease
The presence or worsening of depressed mood and its impact on quality of life.
across the 24-​hour day is currently included as a Similar to BED, future NES research should
specifier of NES. As depressed mood is common also focus on genetic and brain-​imaging methods.
in other EDs but is not included in their diagnos- Lundgren, Allison, and Stunkard (2006) found that
tic items, more work is needed to assess the utility NES runs in families. Specifically, the first-​degree
of this item for the diagnosis of NES. In addition, relatives of patient with NES were 4.9 times more
weight and shape concerns are not included in the likely than the first-​degree relatives of controls to
research criteria. These concerns are higher among meet criteria for NES. Additionally, as noted above,
persons with NES than controls, but are signif- heritability for night eating behaviors has been
icantly lower than persons with BED (Allison, reported at .44 among the Swedish Twin Register
Grilo, et al., 2005); Figure 23.1. Those with BED cohort, with a high genetic correlation of.66
and NES did not differ from each other on eat- between night eating and binge eating behaviors
ing concerns and dietary restraint on the Eating (Root et al., 2010).
Disorder Examination, and both groups reported Brain-​imaging methods could also be use-
higher levels than controls. Thus, weight and shape ful in characterizing NES. Lundgren, Newberg,
concerns are elevated among individuals with NES, and colleagues (2008) used single photon emis-
but how central this is to the disorder needs more sion computed tomography (SPECT) to compare
investigation. the serotonin transporter (SERT) uptake in per-
Distress and impairment in functioning are sons with NES to both healthy controls and to
required for all psychiatric diagnoses, and it has patients with major depressive disorder. Persons

444 Emerging Syndromes

with NES had significantly greater SERT uptake in The name “purging disorder” (PD) was estab-
the midbrain than did healthy controls (Lundgren, lished by Keel et  al. (2005) and encompasses dis-
Newberg, et al., 2008). Patients with NES also had ordered eating behavior that has been reported
significantly greater SERT uptake ratios (effect size under several names in the literature, including
range 0.64–​0.84) in the midbrain, right temporal “compensatory eating disorder” (Tobin, Griffing,
lobe, and left temporal lobe regions compared to & Griffing, 1997), “subjective bulimia nervosa”
patients with depression (Lundgren et  al., 2009). (Keel et  al., 2001), and “purging-​only syndrome”
Finally, only one pilot study has used functional (Wade, 2007). Eating disorder not otherwise speci-
magnetic resonance imaging (fMRI) to examine fied, purging type (EDNOS-​P) has also been used
response to food cues among seven adults with obe- to describe this group (Binford & le Grange, 2005),
sity with evening hyperphagia compared to seven given that persons with these clinical features would
without hyperphagia (Lundgren et al., 2013). The have been diagnosed in the EDNOS category in
authors found that those with evening hyperphagia the DSM-​IV and DSM-​IV-​TR classification sys-
showed decreased inferior frontal gyrus activation tems. Currently, under the DSM-​5 classification
from the pre-​to post-​meal condition in response system, PD is included as a specific example of a
to food cue images as compared with the control clinical presentation within the OSFED diagnostic
group. Overall, the differences they identified in category of the feeding and eating disorders (FEDs;
activation patterns suggest that those with evening American Psychiatric Association, 2013).
hyperphagia showed decreased activation of inhibi- Like most emerging syndromes, varying defini-
tory brain regions, but more studies are needed to tions are presented in the literature before a com-
replicate this first finding. Additional future direc- mon set of diagnostic criteria is agreed on. In a
tions would be testing the behavioral features of review of the literature, Keel (2007) described 10
night eating (e.g., impulsivity and reward depen- different definitions for syndromes of purging. At
dence could be measured with delay discounting the time of Keel’s review, PD was generally defined
paradigms; e.g., McClure, Laibson, Lowenstein, & as the regular occurrence of inappropriate compen-
Cohen, 2004). Studies such as these will be impor- satory behaviors (e.g., vomiting, laxative use, or
tant in differentiating NES from other eating and diuretic misuse) for the purpose of weight or shape
psychiatric disorders. control in the absence of objective binge eating
episodes and with a body weight greater than 85%
Research Status and Description of that expected (Keel, 2007). Sufferers often feel
of Purging Disorder distressed after consuming even small amounts of
Similar to other emerging syndromes, purg- food that are not acceptable to them and have an
ing disorder research and clinical descriptions have overwhelming urge to purge afterward. Aspects of
grown, particularly in the context of the develop- the definition that have varied include (1) the fre-
ment of the latest edition of the DSM. The research quency of purging, ranging from at least once per
status and clinical descriptions of purging disorder week to at least twice per week; (2) requirement that
are reviewed below. a loss of control over eating (i.e., subjective binge
episode) precede the purging behavior; (3) the pres-
History and Definition ence of undue influence of weight and shape on
As researchers began to study BN more sys- self-​evaluation; and (4) duration of illness, ranging
tematically, several groups (e.g., Mitchell, Pyle, from 3 to 6 months.
Hatsukami, & Eckert, 1986) noted that large In preparation for DSM-​5, Keel and Striegel-​
proportions of their cohorts were not consuming Moore (2009) provided an analysis of the validity
enough food in a sitting to be considered objec- and clinical utility of PD using Blashfield and col-
tively large, and therefore, were not meeting crite- leagues’ (1990) criteria for establishing a disorder
ria for binge episodes, yet they were purging (e.g., within the DSM system. In the context of their lit-
vomiting or abusing laxatives) regularly. Persistent erature review and analysis, they proposed criteria
purging behavior in the absence of binge eating has for PD: (1) recurrent purging in order to influence
subsequently been reported among adults (Wade, weight or shape, such as self-​induced vomiting, mis-
2007) and adolescents (Binford & le Grange, 2005) use of laxatives, diuretics, or enemas; (2)  purging
in community (Hay, Fairburn, & Doll, 1996), clini- occurs, on average, at least once a week for three
cal (Keel, Haedt, & Edler, 2005), and undergradu- months; (3) self-​evaluation is unduly influenced by
ate (Fink et al., 2009) samples. body shape or weight or there is an intense fear of

Allison, Lundgren 445

gaining weight or becoming fat; (4) the purging is inclusion in the OSFED category suggests that the
not associated with objectively large binge episodes; scientific community charged with evaluating the
and (5) the purging does not occur exclusively dur- PD literature for the DSM system is taking a “wait
ing the course of anorexia nervosa or bulimia ner- and see” approach. Nonetheless, here we review the
vosa (Keel & Striegel-​Moore, 2009). Notably, the evidence for PD as (1) a distinct disorder and (2) on
definition of PD was abbreviated with its inclusion a continuum with other EDs.
in the OSFED category within the “Feeding and
Eating Disorders” section of DSM-​5. Within the Purging Disorder as a Distinct Axis
DSM-​5, PD is defined as “recurrent purging behav- I Psychiatric Disorder
ior to influence weight or shape in the absence of Purging disorder can be damaging both physi-
binge eating” (APA, 2013). cally and psychologically. The fact that it was recog-
nized specifically as a form of EDNOS in DSM-​IV
Prevalence and Epidemiology and under the OSFED category in DSM-​5 affirms
Epidemiological studies reveal lifetime preva- that it is a clinically significant syndrome. Keel
lence estimates of 5.3% in an Australian female (2007) argues that because the purging types of AN
twin cohort (age range 28–​39 years) (Wade, Bergin, and BN have been associated with greater medical
Tiggemann, Bulik, & Fairburn, 2006), 1.0% in problems (due to the effects of repeated vomiting
an slightly younger Australian female twin cohort and laxative abuse), comorbid psychopathology,
(age range 15–​19  years) (Fairweather-​Schmidt & and suicidality, that purging in the absence of
Wade, 2014), 0.3% in an Australian young adult very low body weight and binge episodes is inde-
male cohort (20  years of age) using DSM-​5 crite- pendently quite dangerous. In addition, PD has
ria (Allen, Byrne, Oddy, & Crosby, 2013), 1.6% in been identified as a class separate from other EDs
an Australian young adult female cohort (20 years through latent structure analyses (Striegel-​Moore,
of age) using DSM-​5 criteria (Allen et  al., 2013), Franco et  al., 2005; Sullivan, Bulik, & Kendler,
3.7% in a US female twin cohort (age range 18–​29) 1998; Wade et al., 2006, Pinheiro, Bulik, Sullivan,
(Munn-​Chernoff et  al., 2015), 1.1% in an Italian & Machado, 2008; Swanson et al., 2014). Striegel-​
female cohort (Favaro, Ferrara, & Santonastaso, Moore, Franko, et al. (2005) indicated that the PD
2003), 8.2% in a large sample of adult Swedish group was the largest of the three identified in the
females seeking treatment for EDs (Ekeroth, analysis. All members of the PD group endorsed
Clinton, Norring, & Birgegård, 2013), and a point vomiting, nearly half reported fasting, and over a
prevalence of 0.85% in an adolescent Portuguese third reported the use of diet pills. In addition, there
cohort (Machado, Machado, Goncalves, & Hoek, was a higher proportion of persons identifying their
2007). The relative frequency of these estimates in race as White in the latent PD group than in the
some samples, as compared with the other EDs, has latent binge-​eating group, who had a larger pro-
varied, likely due to the differing diagnostic criteria portion of reported Black racial identities (Striegel-​
used in each study as well as the nature of the pop- Moore, Franko et al., 2005).
ulation sampled. It is not surprising, for example, There are conflicting findings on the most
that the estimates found in non-​treatment-​seeking prevalent purging methods in PD. Vomiting has
populations are lower than among those who are been most commonly reported, at 100% of the
seeking treatment for disordered eating behavior. PD sample (Striegel-​Moore, Franko et  al., 2005),
Purging disorder has been documented in both 98.5% (Binford & le Grange, 2005), and 82%
females and males, and appears to be more prev- (Ekeroth et  al., 2013). Wade (2007), however,
alent among females (Allen et  al., 2013). Little is reported higher rates of laxative abuse (62%) than
known about gender or socioeconomic differences vomiting (38%), followed by diuretic abuse (21%).
among individuals who suffer from PD versus those In their adolescent sample, Binford and le Grange
who suffer from other forms of disordered eating. (2005) reported that driven exercise was also quite
common at 66%, followed by fasting at 38%, and
Models of Purging Disorder then laxative abuse at 14%. Excessive exercise was
Much work still remains in delineating the etiol- also commonly reported among adult females seek-
ogy, course, treatment, and outcome for PD, but ing treatment for eating disordered behavior (55%)
there is emerging evidence that it can be differen- (Ekeroth et  al., 2013). Thus, there appears to be
tiated from other disorders and is a clinically sig- heterogeneity of purging and nonpurging compen-
nificant disorder (Keel & Striegel-​Moore, 2009). Its satory methods within the PD classification.

446 Emerging Syndromes

 Several studies have shown that persons with a 6-​month follow-​up, and little diagnostic crossover
PD have more severe eating disordered attitudes was found, suggesting at least short-​term stability of
and behaviors and general psychopathology than the diagnoses. Longer-​term stability and crossover
persons without eating disorders. Specifically, Keel was examined in a longitudinal study of Australian
et  al. (2005, 2007) have shown that persons with adolescents enrolled in the Western Australian
PD have shown more severe pathology on all four Pregnancy Cohort (Raine) Study (Allen et al.,
scales of the Eating Disorder Examination (EDE) 2013). The prevalence of PD at baseline (age 14)
and all three scales of the Eating Inventory, and was quite low in males (0.4%, n = 3), so diagnostic
Wade (2007) replicated the finding on the EDE crossover was only reported for females. Of those
with the exception of similar levels of eating con- females diagnosed with DSM-​5 PD at age 14 (n =
cern between those with PD and controls. Body 31), 3.2% met DSM-​5 criteria for AN, 25.9% met
image is more disturbed (Fink et  al., 2009; Keel DSM-​5 criteria for BN, 6.4% met DSM-​5 criteria
et  al., 2005), and general psychopathology is also for BED, 0.0% met diagnostic criteria for atypical
consistently greater in PD groups than in controls, AN, 12.9% met DSM-​5 criteria for PD, and 51.6%
including lifetime major depression, anxiety, and no longer met criteria for any ED by ages 17 or 20.
suicidality (Keel et al., 2005; Wade, 2007). In contrast to the shorter-​term stability evidenced
Similar to comorbidity rates for lifetime diagno- by Keel et al. (2005), this longitudinal study, based
sis of major depressive disorder reported for NES on DSM-​5 criteria, suggests that crossover from PD
earlier, 54% (Binford & le Grange, 2005), and 76% to BN is both statistically and clinically significant.
of persons with PD have reported the disorder. Notably, in this same study, crossover from BED to
Using ecological momentary assessment, Haedt-​ BN was much higher at 52.2% (Allen et al., 2013).
Matt and Keel (2015) documented greater mean Despite the evidence of diagnostic crossover
negative affect on purge days versus nonpurge days. from PD to other EDs, PD as a partial BN syn-
Interestingly, their study found that on purge days, drome or as a variant subtype of BN is more plau-
negative affect decreased after purging and might sible. In several studies, PD groups have endorsed
serve to negatively reinforce purging behavior. similar levels of disordered eating attitudes and
Individuals with PD are more likely than healthy behaviors (see Keel, 2007, and Keel & Striegel-​
controls to report symptoms of personality disor- Moore, 2009 for reviews); Figure 23.1. Differences
ders, and they report greater impulsivity (Brown, on specific scales have sometimes varied, depend-
Haedt-​ Matt, and Keel, 2011). Overall, as was ing on sample characteristics and diagnostic cri-
shown with NES, PD has been identified through teria. If differences in the level of pathology were
empirical methods with latent class analysis and shown, the general trend was for PD to have less
has shown clinically greater levels of eating disor- severe levels than the BN group. These differences
dered and general psychopathology than persons on traditional ED assessment tools may be contrib-
without EDs. uted, at least in part, to the assessment tools, given
that large portions of many of them include ques-
Purging Disorder on a Continuum tions regarding binge eating behavior, which is not
with Other Eating Disorders present in PD.
Purging disorder is most similar to BN, with both One possible avenue for differentiating PD and
purging and nonpurging compensatory behaviors BN may lie in physiological functioning. A  feed-
being identified and normal body weight among ing study has shown that women with PD report
those with PD. Binford and le Grange (2005) iden- higher levels of postprandial fullness and gastroin-
tified two possible theories, that PD represents (1) testinal discomfort after a standardized meal than
a prodromal state of BN or (2) a partial BN syn- those with BN. Most interesting, women with PD
drome. There are various sources of evidence against also experienced a greater release of cholecystokinin
PD as a precursor to BN. Binford and le Grange (CCK; Keel, Wolfe, Liddle, De Young, & Jimerson,
(2005) reported that duration of PD and BN of 2007), suggesting that physiological cues, such as
adolescents in their treatment sample were similar premature abdominal discomfort in response to eat-
at 21 and 19 months. Keel et al. (2005) studied ing, may contribute to the purging behavior.
distinctive groups of BN, PD, and controls, requir- The evidence for overlap between PD and AN
ing that their PD participants not have a previous is not very great, given the differences in body
diagnosis of BN. Remission rates from treatment weight. Fink et al. (2009) used the Eating Disorder
for PD and BN did not differ between the groups at Inventory to compare eating pathology among AN,

Allison, Lundgren 447

BN, PD, BED, and controls, finding similarities psychiatric comorbidity, social support, etc.). Keel
between AN and PD for most scales, but higher et al. (2007) have provided one of the only studies
purging ideation among the PD group than the of the pathophysiology of PD thus far, with infor-
AN group. It is possible that PD could be consid- mative findings regarding differential functioning
ered a form of atypical AN, another specific clini- of CCK in PD as compared with BN. Much more
cal presentation of OSFED defined in the DSM-​5. work is needed in this area, as well as assessment
Atypical AN (reviewed later) is a condition for using brain-​imaging technologies. Functional mag-
which all criteria are met for AN, except the indi- netic resonance imaging (fMRI) studies could aid
vidual is at or above normal weight despite signifi- in delineating similarities and differences between
cant weight loss. Ekeroth and colleagues (2013) neural function in relation to reward sensitivity
reported that among treatment-​ seeking Swedish and food and body image cue reactivity among PD
women, 10% of those who met DSM-​5 criteria for and the other EDs. Finally, there is also a dearth
PD at baseline met criteria for atypical AN at the of treatment outcome studies, which are in obvious
12-​month follow-​up. In contrast with the crossover need, given the number of individuals who suffer
to BN rates reported above (Allen et al., 2013), only from PD.
5% of this treatment-​seeking sample crossed over to
BN 1 year later. Research Status and Description of Aytpical
Anorexia Nervosa
Evidence of Diagnostic Validity and Clinical Clinical descriptions and research on atypical
Significance of Models of Purging Disorder anorexia nervosa have only recently begun. Here we
Keel and Striegel-​Moore (2009) present a strong describe its definition, history, and prevealnce.
argument for the clinical significance of PD, includ-
ing significant evidence that PD is associated with History and Definition
psychosocial distress and impairment that is consis- More than half of patients seeking care at ED
tent with other EDs and different from healthy con- treatment services did not meet threshold criteria
trols. Functionally, by placing PD in the OSFED for AN or BN using the previous DSM-​IV diagnos-
category of FEDs within the DSM-​5, clinical signif- tic criteria and were designated as having EDNOS
icance is assumed, at least until the DSM is revised (Eddy, Celio, Hoste, Herzog, & le Grange, 2008;
in the future. Much more needs to be learned about Fairburn & Bohn, 2005). It was noted that those
PD, however, including its etiology, distinct course, classified as having EDNOS reported compara-
and treatment outcome in relation to the other eat- ble levels of distress and impairment as those in
ing disorders. the threshold disorders. As such, in DSM-​5, the
OSFED category was created to capture common
Research Needed to Clarify the Status presentations that were observed in the EDNOS
of Purging Disorder category. In addition to NES and PD, one such
Relative to AN, BN, and BED, research on clinical presentation was described as “atypical
PD as an independent construct is still in its early anorexia.”
stages. Its inclusion in DSM-​5 should spark new In atypical AN, all of the criteria for AN are
research to answer many of the unknown questions. met, except that despite significant weight loss, the
An important factor that has yet to be explored individual’s weight is within or above the normal
specifically for PD is its etiology and a conceptual range. Anorexia nervosa is described in c­ hapter  1,
model of its development and maintenance. Studies and involves (1) restriction of energy intake relative
such as Haedt-​Matt and Keel’s (2015) ecological to one’s energy requirements for weight mainte-
momentary assessment of affect before and after nance, (2) experience of an intense fear of gaining
purging is a great start in understanding the ante- weight or becoming fat, and (3) being subjected to
cedents, consequences, and maintaining processes an undue influence of body weight and shape on
involved in PD. More research is needed to com- self-​evaluation (APA, 2013).
pare the unique and common risk and maintain- These criteria have been operationalized in dif-
ing factors among PD and the other ED diagnoses. fering ways in the research literature for atypical
Given the mixed findings regarding the stability and AN. For example, Stice, Marti, and Rohde (2013)
diagnostic crossover of PD, longitudinal studies are described it as experiencing at least a 10% weight
needed to provide more information about the fac- reduction from a previously measured BMI (i.e.,
tors that predict crossover (e.g., genetic, personality, more than a typical weight reduction for behavioral

448 Emerging Syndromes

weight loss programs), definite fear of weight gain 2,925 women. Of these, 292 screened positive for
more than three-​quarters of the days for at least an ED, and an additional 130 female cotwins and a
3 months, and focusing on weight and shape as one random sample of 210 screen-​negative participants
of the main aspects of self-​evaluation. were interviewed. Mustelin et  al. (2016) reported
Fairweather-​Schmidt & Wade (2014) required a prevalence of 1.5% (n  =  38) for all of OSFED
participants, who were adolescent women twins, (n = 14; 37%) and UFED (n = 24; 63%), including
to meet all criteria for AN but, despite significant just five cases of atypical AN (0.2% of the entire
weight loss, have a BMI at or above 18.5  kg/​m2, sample), which was by far the lowest estimate of this
which designates the lower boundary of “nor- disorder among the studies presented.
mal weight.” Significant weight loss was defined
within this sample as a reduction of 1.3  kg/​m2 Evidence of Diagnostic Validity and Clinical
BMI units relative to BMI at each subsequent Significance Using Models of Atypical
wave of data collection (3 years), and the weight Anorexia Nervosa
loss had to have occurred independently of any To our knowledge, the diagnostic validity of
episode of BN or BED. Given that the description atypical AN has not been presented, and as men-
of atypical AN in the DSM-​5 does not provide tioned above, different studies seem to use varying
specific guidelines for cutoffs for weight loss, it is definitions to operationalize the degree of weight
evident that a consensus definition of a clinically loss and/​or BMI cutoff point for inclusion in this
significant weight loss is needed to standardize diagnostic criteria. However, as presented below,
these criteria. most studies have presented evidence that individu-
als with atypical AN experience comparable levels of
Prevalence distress and impairment as their counterparts with
The prevalence of atypical AN has mainly been threshold EDs. There is also evidence that atypical
reported among adolescent samples. Hammerle, AN does not occur exclusively among individuals
Huss, Ernst, and Burger (2016) published a who have had AN, nor is it merely the precursor of
national school-​ based survey from Germany of crossing over into full AN.
grades 7 and 8 with 1,654 students (873 female, Stice et  al. (2013) reported that there was a
mean age: 13.4 yr). Using the Structured Interview recurrence rate of 21% for atypical AN over the
for Anorexic and Bulimic Syndromes Survey (SIAB-​ 8-​year study, and only one participant exhibited
S) and Eating Disorder Inventory-​2, they found a diagnostic crossover, with that individual develop-
prevalence for atypical AN of 3.6%. ing PD. No individuals with atypical AN developed
Stice et  al. (2013) studied 496 adolescent girls AN during the eight years of this study. The atypical
recruited from US schools, with a baseline age AN group reported mean monthly binge episodes
between 12 and 15 years, interviewing them yearly at 0.3 and mean monthly compensatory behaviors
for 8 years. They reported a prevalence of atypical at a total of 9.8. Those with atypical AN reported
AN of 2.8%, occurring increasingly in the later higher levels of functional impairment, emotional
study years (ages 19–​20 years), which was similar to distress, and suicidality than non-​ED controls, with
age of onset for AN in this sample. Additionally, 10 a medium effect size of .51. Overall, Stice et al. con-
of 14 individuals with atypical AN showed remis- cluded that the level of impairment among those
sion within 1 year. Fairweather-​Schmidt and Wade with atypical AN is similar to that experienced with
(2014) assessed 699 adolescent female twins with a full threshold DSM-​5 EDs, providing validity for
baseline age ranging between 12 and 15 years over atypical AN as a clinically significant disorder. Stice
the course of three assessments. They completed et al. (2013) further posited that there should there
the EDE interview by phone. Overall, 5% of the be a specific weight or BMI cut-​off, given those at
sample met criteria for OSFED, with 1.9% of the lower BMIs with the diagnosis reported higher lev-
sample meeting criteria for atypical AN across the els of impairment.
three waves. Silén et al. (2015) studied 47 adolescents as part
Finally, Mustelin, Lehtokari, and Kski-​Rahkonen of a chart review of 34 patients with AN and 13
(2016) used the FinnTwin16 cohort, with an age patients with atypical AN. Those with atypical AN
range of 22 to 27  years, the oldest sample of the were significantly older and heavier, but with a BMI
studies reviewed, to provide a two-​stage screening of 16.7 kg/​m2, which is still underweight, and less
for EDs that included a questionnaire and inter- likely to have psychiatric comorbidities. Patient with
view. This community-​based sample consisted of atypical AN were significantly more likely to recover

Allison, Lundgren 449

from their symptoms than those with AN. Silén independent of very low weight, that impact meta-
et  al. (2015) suggest that these findings support cognition and, likely, one’s ability to fully engage in
retaining these distinct diagnostic categories, as they psychotherapy treatments.
may have predictive value for treatment outcomes. Sawyer, Whitelaw, Le Grange, Yeo, and Hughes,
Fairweather-​Schmidt and Wade (2014) studied 2016 studied 42 adolescents with atypical AN and
women twins and found that those with atypical 118 with threshold AN presenting for treatment.
AN showed no difference in duration of core symp- In this group, in contrast to samples described ear-
toms at diagnosis from those with AN and BN. lier (e.g., Davenport et al., 2015), 71% of the atypi-
Individuals with OSFED and threshold ED did cal AN group had previously been overweight or
not differ on EDE global or subscales. The authors obese, as compared with only 12% of the AN group.
reported that OSFED and threshold EDs had com- Weight loss in the atypical AN group had been larger
mon generic and nonshared environmental influ- at 17.6 kg than in the AN group (11.0 kg), and the
ences and that only one genetic source contributes former group had lost weight over a longer period
to both expressions of the EDs. However, there was of time. Despite the atypical AN group not meet-
significant differentiation between unique environ- ing the extremely low weight criterion of AN, physi-
mental risk factors between the groups, which may cal comorbidities were similar between the groups,
protect those with OSFED from development of including bradycardia and orthostatic instability.
threshold diagnostic symptoms. Overall, though, the While there were fewer patients with amenorrhea
authors believe that there is large degree of genetic in the atypical AN group as compared with the AN
overlap between the OSFED diagnoses (NES was group, a third of the atypical AN group still experi-
not included) and threshold ED diagnoses. enced this symptom. Psychiatric comorbidities were
Davenport, Rushford, Soon, and McDermott also quite similar between groups, including suicidal
(2015) studied 119 patients seeking treatment for ideation and other psychiatric disorders. However,
AN. They separated the patients at a BMI of 18.5 the EDE global and subscale scores were higher in
kg/​m2 into AN (n = 75) and atypical AN (n = 44). the atypical AN group as compared with the AN
Anorexia nervosa has been shown to have a negative group. The authors conclude that the core features
impact on cognitive processing abilities, particu- of abnormal disordered eating cognitions and weight
larly knowledge about one’s own thinking, leading loss, as opposed to a specific threshold for low weight,
to rigid thinking and making psychotherapeu- might be a more useful view of restrictive EDs.
tic interventions less effective, as described in the Schorr et  al. (2016) studied 77 patients with
Self-​Regulatory Executive Function Model (Wells atypical AN as compared to patients with thresh-
& Matthews, 1994). As such, the authors com- old AN defined by DSM-​IV (n = 37) and DSM-​5
pared dysfunctional metacognition, as measured (n  =  33), and 21 healthy controls on measures of
by the Metacognitions Questionnaire, between the bone health, body composition, and psychopathol-
AN and atypical AN groups. They reported that ogy. Both the AN and atypical AN groups reported
dysfunction was similar between the two groups, comparable levels of psychopathology. As for bone
with drive for thinness distinguishing both from a health, the AN group had the poorest bone health,
community control group. Drive for thinness was with the atypical AN group falling between the AN
positively predicted in the atypical AN group by and comparison groups for bone mineral density.
negative beliefs about worry and, inversely by cog- Among the AN and the atypical AN groups, his-
nitive self-​consciousness. In many ways, the authors tory of amenorrhea was also predictive of poor bone
stated that the AN and atypical AN groups were health. Lowest lifetime BMI was correlated with
indistinguishable. However, the majority of those bone mineral density, but history of overweight or
with atypical AN reported that their lowest BMI obesity was not particularly protective against low
since symptom onset was below 17.5 kg/​m2, and bone mineral density.
their current mean BMI was 21.0 kg/​m2, suggesting Chen et al. (2016) treated young adults with AN
that the group as a whole was still quite lean and had (n = 10) and atypical AN (n = 12) with family-​based
met full threshold AN criteria previously. However, treatment. While dropout rate was high, intent to
this study still showed that low body weight associ- treat analysis showed similar improvements in
ated with AN is not the only factor driving these weight gain and eating disordered pathology for
individuals’ difficulties with self-​awareness of their the group as a whole. Unfortunately, no compari-
thought processes, suggesting that there may be sons were presented for outcomes specifically for
other pathways influencing drive for thinness, the atypical AN group. This was certainly a good

450 Emerging Syndromes

place to start with treatment studies of atypical AN, emerging disorders of eating reviewed in this chap-
but much more effort is needed in testing treatment ter, ON has the least amount of empirical evaluation
efficacy for this group. on which to evaluate its validity as a clinically signif-
icant diagnostic construct. Its tenuous status is due
Research Needed to Clarify the Status in part to its introduction as a medical condition in
of Atypical Anorexia Nervosa a popular magazine, Yoga Journal (Bratman, 1997),
Most of the studies reviewed here have examined rather than through a peer-​reviewed scientific jour-
largely female, adolescent samples. Additionally, nal. This is of importance because ON received
the BMI of individuals in these samples with atypi- significant media and public attention prior to the
cal AN remains in the low-​normal range, mostly publication of any scientific evaluation, including
between 20 and 21  kg/​m2. As this diagnostic cat- the development of a common set of diagnostic
egory could potentially capture individuals who lose criteria based on an adequate number of clinical
large amounts of weight through behavioral weight observations, psychometrically sound assessment
loss or bariatric surgery and subsequently develop instruments, or studies to examine diagnostic reli-
an extreme fear of weight regain and espousing AN ability and validity. Consequently, ON was reified
thoughts and beliefs, studies of older populations as in the public prior to adequate scientific evaluation.
well as these special weight loss treatment popula- Since its introduction in popular culture 20
tions are needed, as the presentation of atypical AN years ago, a small literature on ON has begun
may vary based on these characteristics. More epi- to develop. Similar to other EDs, this literature
demiological studies would also be useful to deter- includes case studies (Zamora, Bonaechea, Sánches,
mine the presence of atypical AN across age groups & Rial, 2005; Park et al., 2011; Saddichha, Babu,
and demographics. As reviewed in the beginning of & Chandra, 2012; Moroze et al., 2015), assessment
the chapter, this disorder has not yet satisfied the (i.e., versions of the “questionnaire for the diagnosis
first two criteria of Blashfield et al. (1990) in that of orthorexia” [ORTO]) (e.g., Bratman & Knight,
the body of literature is not yet large and there is no 2000; Donini et al., 2005), and prevalence esti-
consensus agreement on the criteria used to define mates (e.g., Donini et al., 2004). Despite the grow-
the amount of weight loss required for atypical AN ing number of published studies of ON, a critical
to be present. limitation of this literature is that much of it is built
Additionally, more studies of the medical burden on a weak scientific foundation wrought with meth-
of atypical AN as compared with control groups of odological and measurement limitations. For exam-
similar BMI would be helpful to add to the knowl- ple, the first scientific study of ON was published
edge base for this disorder that would serve to sup- by Donini and colleagues in 2004. They sampled
plement our knowledge of the impact of the eating a diverse group of Italian individuals ranging from
disordered attitudes and behaviors on functioning professional adults to adolescents as young as 16
and well-​ being. Finally, more treatment studies years of age. Orthorexia nervosa was operational-
are needed to understand if atypical AN is more ized as scoring at or below the 25th percentile on
responsive to both psychological and pharmaco- an unidentified dietary assessment designed to cap-
logical treatments than typical AN. If atypical AN ture “healthy” versus “unhealthy eating” across food
is more responsive to a range of treatments, this dif- groups, along with elevated Minnesota Multiphasic
ference could lend clinically relevant significance to Personality Inventory (MMPI) scores on the psych-
the use of this diagnosis. asthenia clinical scale measuring anxious/​obsessive
features. Based on this unusual diagnostic opera-
Research Status and Description tionalization, they reported an ON prevalence of
of Orthorexia Nervosa 6.9%. Clearly, given the lack of valid diagnostic cri-
The final emerging syndrome to be evaluated is teria for ON at the time of the study, as well as their
orthorexia nervosa (ON). As our review below dem- basis of a diagnosis on scores on two assessment
onstrates, this is the most tenuous of the syndromes instruments, the utility of this study (and others
discussed in this chapter. with similar methodology) in clarifying the nature
of ON is limited.
History and Definition As is the case for any emerging syndrome, the
Orthorexia nervosa is conceptualized as an obses- development of standard diagnostic criteria is
sive preoccupation with “healthy eating” (Dunn an iterative process that involves adding, remov-
& Bratman, 2016). In comparison to the other ing, and modifying symptom presentation and

Allison, Lundgren 451

frequency based on research findings. The first behavior. If an individual’s body image is dependent
comprehensive diagnostic criteria for ON were on the “healthy” eating behavior, it is quite pos-
presented by Moroze and colleagues (2015). These sible that the primary goal of the “healthy” eating
included (1) an obsessional preoccupation with eat- behavior is weight loss. Much research is necessary
ing healthy foods that is manifested with at least to evaluate these and the Moroze and colleagues
two of seven behavioral, cognitive, or emotional (2015) criteria to determine the degree of diagnos-
symptoms such as guilt after eating “unhealthy” tic distinctness between ON and other FEDs (e.g.,
foods or spending a relatively excessive amount of avoidant restrictive food intake disorder [ARFID]
money on foods because of their perceived quality and AN), diagnostic stability and crossover, course,
and composition; (2)  impairment or distress due and treatment outcome. Importantly, valid and reli-
to the obsessional preoccupation with healthy eat- able assessment instruments will be needed, as the
ing, including malnutrition; and (3) the obsessional literature demonstrates an overreliance on a single
preoccupation with healthy food cannot be better self-​report assessment instrument (ORTO and its
explained by other medical condition, social/​reli- numerous versions).
gious observation, or food allergy.
Recently, Dunn and Bratman (2016) published Prevalence
a review of the history and current status of the ON The first empirical prevalence estimate was 6.9%,
literature. They propose a modified diagnostic crite- reported by Donini et  al. (2004), and reviewed
ria set that are outlined in Table 23.2. above. Despite this study’s methodological limita-
Presumably, the Dun and Bratman (2016) cri- tions, it provides one of the lowest (and more likely)
teria were introduced to add clarification to the prevalence estimates in the literature. The majority
diagnostic literature on ON. This latter criteria set, of ON studies estimate prevalence in the range of
however, raises questions about the validity and 30%–​88% of those sampled (for review, see Dunn
clinical utility of the ON diagnosis. For example, & Bratman, 2016). To date, in part due to a lack
the introduction to Criterion A  states that weight of standard research diagnostic criteria, most prev-
loss may ensue, but that it is not the primary goal. alence studies are based on self-​report using a ver-
Criterion B3, however, introduces the notion that sion of the ORTO. Until research is available to
body image might be dependent on “healthy” eating evaluate the recently proposed diagnostic criteria

Table 23.2  Dunn and Bratman (2016) Diagnostic Criteria for ON

Criterion A
Obsessive focus on “healthy” eating as defined by a dietary theory or set of beliefs whose specific details may vary;
marked by exaggerated emotional distress in relationship to food choices perceived as unhealthy; weight loss may ensue
as a result of dietary choices, but this is not the primary goal. As evidenced by the following:

1.  Compulsive behavior and/​or mental preoccupation regarding affirmative and restrictive dietary practices believed by
the individual to promote optimum health.
2.  Violation of self-​imposed dietary rules causes exaggerated fear of disease, sense of personal impurity and/​or negative
physical sensations, accompanied by anxiety and shame.
3.  Dietary restrictions escalate over time and may come to include elimination of entire food groups and involve
progressively more frequent and/​or severe “cleanses” (partial fasts) regarded as purifying or detoxifying. This
escalation commonly leads to weight loss, but the desire to lose weight is absent, hidden, or subordinate to ideation
about healthy eating.

Criterion B
The compulsive behavior and mental preoccupation becomes clinically impairing by any of the following:

1.  Malnutrition, severe weight loss, or other medical complications from restricted diet.
2.  Intrapersonal distress or impairment of social, academic, or vocational functioning secondary to beliefs or behaviors
about healthy diet.
3.  Positive body image, self-​worth, identity, and/​or satisfaction excessively dependent on compliance with self-​defined
“healthy” eating behavior.
Reprinted from Eating Behaviors, 21, Dunn and Bratman, On orthorexia nervosa: A review of the literature and proposed diagnostic criteria,
11–​17, Copyright (2016), with permission from Elsevier.

452 Emerging Syndromes

sets (Moroze et al., 2015; Dunn & Bratman, 2016) AN, but further clarification with regard to the role
prevalence will remain largely unknown. of “healthy” eating on body image (criterion B3) is
needed. Finally, it is possible that behaviors, cog-
Models of Orthorexia Nervosa nitions, and emotions associated with ON are best
In their review, Dunn and Bratman (2016) con- explained as an anxiety disorder and potentially a
clude that “there is sufficient evidence that ON is variant of obsessive-​compulsive disorder. The litera-
a distinct condition that is different from ARFID” ture on ON has highlighted the obsessive nature of
(p. 16). Whether or not this is accurate remains to the desire to eat “healthy,” and in fact several of the
be clarified with empirical research. Blashfield et al. questions on the ORTO assess “worry” related to
(1990) outlined five criteria by which DSM diagno- eating behaviors and food choice (e.g., worry about
ses should be evaluated: (1) ample literature on the the thought of food).
syndrome, (2) clearly articulated diagnostic criteria
and psychometrically sound assessment instruments Research Needed to Clarify the Status
to evaluate the criteria, (3) evidence that the diagno- of Orthorexia Nervosa
sis can be reliably determined across raters, (4) evi- Several areas of research could help improve the
dence that the syndrome can be differentiated from state of the ON literature. First, additional valid
similar syndromes, and (5) evidence of the validity and reliable assessment instruments are needed to
and clinical significance of the syndrome, includ- assess the recently proposed diagnostic criteria sets.
ing course, outcome, and treatment response. The Related to this, several aspects of the diagnostic
Blashfield et al. (1990) criteria have been applied to criteria need clarification. For example, guidance
a number of emerging syndromes, including NES is needed as to how to differentiate restriction for
and PD, described above. food quality versus restriction for reduced calories
Despite the growing literature on ON, it does reliably, when there is clearly overlap between calo-
not yet meet these criteria—​notably because stan- ries and quality (i.e., “healthy” foods are generally
dard diagnostic criteria were only recently intro- lower in calories than “unhealthy” foods). As noted
duced in the literature, and as noted previously, above, the role of weight and shape concerns vis-​à-​
there are serious methodological limitations to the vis “body image” is another area for further clarifi-
literature that is currently available. If ON is to be cation. Once these basic diagnostic and assessment
considered as an independent diagnostic construct, questions are addressed in the literature, prevalence
it would behoove those who research it to develop can be evaluated more accurately. Additionally, tax-
a longer-​term research plan across multiple research ometric statistical approaches are needed to deter-
teams using the Blashfield et al. (1990) criteria and mine whether or not ON is distinct enough from
the bodies of literature on other emerging syn- other Axis I disorders, including ARFID, AN, and
dromes (e.g., PD) as a guide. OCD to constitute a stand-​alone disorder. Finally,
Other models of ON include considering it a much research is needed on the etiology (genetic
variant of ARFID, AN, or an anxiety disorders such and psychosocial), diagnostic stability, and treat-
as obsessive-​compulsive disorder. Currently, obses- ment outcome.
sional preoccupation with eating “healthy” foods In summary, of all of the emerging disorders
falls within the DSM-​5 ARFID category of FEDs, reviewed in this chapter, ON is the most tenu-
if it results in significant weight loss, significant ous. Over the past 20  years, a small literature has
nutritional deficiency, or interference in psycho- emerged, but it is too soon to draw conclusions
social functioning, which would encompass most about the status of ON as an independent DSM
clinically significant cases of ON. Arguably, if there diagnosis, given the methodological limitations and
is no medical or psychosocial impairment associated relative paucity of literature on ON, in comparison
with the obsessional “healthy” eating behavior, the to the other FEDs.
eating behavior would not constitute a psychiatric
diagnosis. Future Directions
It is also possible that some individuals who pres- As research on the diagnostic validity and util-
ent with ON symptoms might actually meet criteria ity of NES, PD, atypical AN, and ON continues,
for AN. This would be the case for those who mini- there are two primary areas that will likely make
mize the role of “healthy” eating for body shape or significant contributions to the literature. These
weight. The Dunn and Bratman (2016) proposed include new genetic analyses and brain-​ imaging
diagnostic criteria for ON appear orthogonal to techniques. With regard to newer genetics methods,

Allison, Lundgren 453

both epigenetics (Delcuve, Rastegar, & Davie, syndrome: A comparative study of disordered eating. Journal
2009) and interspecies genetics (Kas, Kaye, Mathes, of Consulting and Clinical Psychology, 73, 1107–​1115.
Allison, K. C., Lundgren, J. D., Moore, R. H., O’Reardon, J.
& Bulik, 2008) should be a priority in moving P., & Stunkard, A. J. (2010). Cognitive behavior therapy for
NES, PD, atypical AN, and ON research forward. night eating syndrome:  a pilot study. American Journal of
Researchers are just beginning to understand the Psychotherapy, 64, 94–​106.
neural mechanisms that underlie behavioral choice Allison, K. C., Lundgren, J. D., O’Reardon, J. P., Geliebter, A.,
and the hedonic effects of food (McClure et  al., Gluck, M. E., Vinai, P.,  . . .  Stunkard, A. J. (2010). Proposed
diagnostic criteria for night eating syndrome. International
2004; Schienle et  al., 2009), and these methods Journal of Eating Disorders, 43, 241–​247.
should help further our understanding about the Allison K. C., Lundgren, J. D., O’Reardon, J. P., Martino, N. S.,
independence and interrelationships between these Sarwer, D. B., Wadden, T. A.,  . . .  Stunkard, A. J. (2008).
emerging syndromes and other forms of disordered Psychometric properties of a measure of severity of the night
eating. eating syndrome. Eating Behaviors, 9, 62–​72.
Allison, K. C., Stunkard, A. J., & Their, S. L. (2004). Overcoming
the night eating syndrome: A step-​by-​step guide to breaking the
Summary and Conclusions cycle. Oakland, CA: New Harbinger.
In summary, there is strong evidence for NES Allison, K. C., Wadden, T. A., Sarwer, D. B., Fabricatore, A.
as a threshold ED, as there is now a large cache of N., Crerand, C., Gibbons, L.,  . . .  Williams, N. N. (2006).
evidence that supports its validity as a stand-​alone Night eating syndrome and binge eating disorder among
persons seeking bariatric surgery: Prevalence and related fea-
disorder and its clinical utility. Purging disorder tures. Obesity, 14(Suppl 2), 77S–​82S.
seems to be associated with negative consequences American Psychiatric Association (APA). (2000). Diagnostic
similar to those found in BN, but it is lacking a and statistical manual of mental disorders (4th ed., text rev.)
widely acknowledged criteria set and studies specific Washington, DC: Author.
to PD (i.e., not just as part of an OSFED sample). American Psychiatric Association. (2013). Diagnostic and sta-
tistical manual of mental disorders (5th ed.). Washington,
Similarly, atypical AN needs much further study as DC: Author.
a disorder that could stand independently from AN, Andersen, G. S., Stunkard, A. J., Sorensen, T. I. A., Pedersen, L.,
the other pure restricting syndrome. Finally, ON & Heitman, B. L. (2004). Night eating and weight change
has the least preponderance of evidence to support in middle-​aged men and women. International Journal of
its validity as an eating disorder. While it appears Obesity, 28, 1138–​1143.
Aronoff, N. J., Geliebter, A., & Zammit, G. (2001). Gender
distressing and interferes with daily functioning, and body mass index as related to the night-​eating syn-
much more scientific work is needed to understand drome in obese outpatients. Journal of the American Dietetics
whether it is a variant of AN or a distinct disorder. Association, 101, 102–​104.
Binford, R. B., & le Grange, D. (2005). Adolescents with bulimia
References nervosa and eating disorder not otherwise specified-​purging
Adami, G. F., Campostano, A., Marinari, G. M., Ravera, G., & only. International Journal of Eating Disorders, 38, 157–​161.
Scopinaro, N. (2002). Night eating in obesity: A descriptive Birketvedt, G., Florholmen, J., Sundsfjord, J., Osterud,
study. Nutrition, 18, 587–​589. B., Dinges, D., Bilker, W.,   .  .  .  Stunkard, A. J. (1999).
Allen, K. L., Byrne, S. M., Oddy, W. H., & Crosby, R. D. Behavioral and neuroendocrine characteristics of the night-​
(2013). DSM-​IV-​TR and DSM-​5 eating disorders in ado- eating syndrome. JAMA, 282, 657–​663.
lescents:  Prevalence, stability, and psychosocial correlates Blashfield, R. K., Sprock, J., & Fuller, A. K. (1990). Suggested
in a population-​based sample of male and female adoles- guidelines for including or excluding categories in the DSM-​
cents. Journal of Abnormal Psychology, 122(3), 720–​732. IV. Comprehensive Psychiatry, 31, 15–​19.
doi: 10.1037/​a0034004. Boston, R. C., Moate, P. J., Stunkard, A. J., Allison, K. C., &
Allison, K. C., Ahima, R. S., O’Reardon, J. P., Dinges, D. F., Lundgren, J. D. (2008). Modeling food intake diurnal
Sharma, V., Cummings, D. E.,  . . . . Stunkard, A. J. (2005). rhythms by means of parametric deconvolution. American
Neuroendocrine profiles associated with energy intake, sleep, Journal of Clinical Nutrition, 87, 1672–​1677.
and stress in the night eating syndrome. Journal of Clinical Bratman, S. (1997, September/​October). The health food eating
Endocrinology and Metabolism, 90, 6214–​6217. disorder. Yoga Journal, 42–​50.
Allison, K. C., Crow, S., Reeves, R. R., West, D. S., Foreyt, J. P., Bratman, S., & Knight, D. (2000). Health food junkies: Orthorexia
DiLillo, V. G.,  . . .  Stunkard, A. J. (2007). The prevalence nervosa:  Overcoming the obsession with healthful eating.
of binge eating disorder and night eating syndrome in adults New York, NY: Broadway.
with type 2 diabetes mellitus. Obesity, 15, 1285–​1291. Brown, T. A., Haedt-​ Matt, A. A., & Keel, P. K. (2011).
Allison, K. C., Engel, S. G., Crosby, R. D., de Zwaan, M., Personality pathology in purging disorder and bulimia
O’Reardon, J. P., Wonderlich, S. A.,   .  .  .  Stunkard, A. J. nervosa. International Journal of Eating Disorders, 44,
(2008). Evaluation of diagnostic criteria for night eating syn- 735–​740.
drome using item response theory analysis. Eating Behaviors, Chen, E. Y., Weissman, J. A., Zeffiro, T. A., Yiu, A., Eneva, K.
9, 398–​407. T., Arlt, J. M., & Swantek, M. J. (2016). Family-​based ther-
Allison, K. C., Grilo, C. M., Masheb, R. M., & Stunkard, apy for young adults with anorexia nervosa restores weight.
A. J. (2005). Binge eating disorder and night eating International Journal of Eating Disorders, 49, 701–​707.

454 Emerging Syndromes

Colles, S. L., & Dixon, J. B. (2006). Night eating syn- Hammerle, F., Huss, M., Ernst, V., & Bürger, A. (2016).
drome:  Impact on bariatric surgery. Obesity Surgery, 16, Thinking dimensional:  Prevalence of DSM-​5 early adoles-
811–​820. cent full syndrome, partial and subthreshold eating disorders
Colles, S. L., Dixon, J. B., & O’Brien, P. E. (2007). Night eating in a cross-​sectional survey in German schools. BMJ Open,
syndrome and nocturnal snacking: Association with obesity, 6:e010843. doi: 10.1136/​bmjopen-​2015-​010843.
binge eating and psychological distress. International Journal Hay, P., Fairburn, C. G., & Doll, H. A. (1996). The classifica-
of Obesity, 31, 1722–​1730. tion of bulimic eating disorders: A community-​based cluster
Davenport, E., Rushford, N., Soon, S., & McDermott, C. analysis. Psychological Medicine, 26, 801–​812.
(2015). Dysfunctional metacognition and drive for thinness Hood, M. M., Reutrakul, S., & Crowley, S. J. (2014). Night
in typical and atypical anorexia nervosa. Journal of Eating eating in patients with type 2 diabetes:  Associations with
Disorders, 3, 24. doi: 10.1186/​s40337-​015-​0060-​4. glycemic control, eating patterns, sleep, and mood. Appetite,
Delcuve, G. P., Rastegar, M., & Davie, J. R. (2009). Epigenetic 79, 91–​96.
control. Journal of Cellular Physiology, 219, 243–​250. Howell, M. J., Schenck, C. H., & Crow, S. J. (2009). A review of
de Zwaan, M., Müller, A., Allison, K. C., Brähler, E., & Hilbert, nighttime eating disorders. Sleep Medicine, 13, 23–​24.
A. (2014). Prevalence and correlates of night eating in the Kas, M. J. H., Kaye, W. H., Mathes, F., & Bulik, C. M. (2008).
German general population. PLoS One; 14:9:e97667. Interspecies genetics of eating disorder traits. American
Donini, L., Marsili, D., Graziani, M., Imbriale, M., & Cannella, Journal of Medical Genetics, Part B, 150B, 318–​327.
C. (2004). Orthorexia nervosa: A preliminary study with a pro- Keel, P. K. (2007). Purging disorder:  Subthreshold variant or
posal for diagnosis and an attempt to measure the dimension full-​threshold eating disorder? International Journal of Eating
of the phenomenon. Eating and Weight Disorders, 9, 151–​157. Disorders, 40, 589–​594.
Donini, L., Marsili, D., Graziani, M., Imbriale, M., & Cannella, Keel, P. K., Haedt, A., & Edler, C. (2005). Purging disorder: An
C. (2005). Orthorexia nervosa:  Validation of a diagnosis ominous variant of bulimia nervosa? International Journal of
questionnaire. Eating and Weight Disorders, 10, e28–​232. Eating Disorders, 38, 191–​199.
Dunn, T. M., & Bratman, S. (2016). On orthorexia ner- Keel, P. K., Mayer, S. A., & Harnden-​Fischer, J. H. (2001).
vosa: A review of the literature and proposed diagnostic cri- Importance of size in defining binge eating episodes in buli-
teria. Eating Behaviors, 21, 11–​17. mia nervosa. International Journal of Eating Disorders, 29,
Eddy, K. T., Celio Doyle, A., Hoste, R. R., Herzog, D. B., & le 294–​301.
Grange, D. (2008). Eating disorder not otherwise specified Keel, P. K., & Striegel-​Moore, R. H. (2009). The validity and
in adolescents. Journal of the American Academy of Child and clinical utility of purging disorder. International Journal of
Adolescent Psychiatry, 47, 156–​164. Eating Disorders, 42, 706–​719.
Ekeroth, K., Clinton, D., Norring, C., Birgegård, A. (2013). Keel, P. K., Wolfe, B. E., Liddle, R. A., DeYoung, K. P., &
Clinical characteristics and distinctiveness of DSM-​5 eating Jimerson, D. C. (2007). Clinical features and physiological
disorder diagnoses: Findings from a large naturalistic clinical response to a test meal in purging disorder and bulimia ner-
database. Journal of Eating Disorders, 1, 31. Retrieved from vosa. Archives of General Psychiatry, 64, 1058–​1066.
http://​www.jeatdisord.com/​content/​1/​1/​31 Lamerz, A., Kuepper-​Nybelen, J., Bruning, N., Wehle, C., Trost-​
Favaro, A., Ferrara, S., & Santonastaso, P. (2003). The spectrum Brinkhues, G., Brenner, H.,   .  .  .  Herpertz-​Dahlmann, B.
of eating disorders in young women: A prevalence study in (2005). Prevalence of obesity, binge eating and night eating
a general population sample. Psychosomatic Medicine, 65, in a cross sectional field survey of 6-​year-​old children and
701–​708. their parents in a German urban population. Journal of Child
Fairburn, C. G., & Bohn, K. (2005). Eating disorder NOS Psychology and Psychiatry, 46, 385–​393.
(EDNOS): An example of the troublesome “not otherwise Latner, J. D., Wetzler, S., Goodman, E. R., & Glinski, J.
specified” (NOS) category in DSM-​IV. Behavior Research (2004). Gastric bypass in a low-​income, inner-​city popula-
and Therapy, 43, 691–​701. tion: Eating disturbances and weight loss. Obesity Research,
Fairweather-​Schmidt, A. K., & Wade, T. D. (2014). DSM-​5 12, 956–​961.
eating disorders and other specified eating and feeding dis- Lundgren, J. D., Allison, K. C., Crow, S., O’Reardon, J. P., Berg,
orders:  Is there a meaningful Differentiation? International K. C., Galbraith, J.,  . . .  Stunkard, A. J. (2006). Prevalence
Journal of Eating Disorders, 47, 524–​533. of the night eating syndrome in a psychiatric population.
Fink E. L., Smith, A. R., Gordon, K. H., Holm-​Denoma, J. M., American Journal of Psychiatry, 163, 156–​158.
& Joiner, T. E. (2009). Psychological correlates of purging Lundgren, J. D., Allison, K. C., O’Reardon, J. P., & Stunkard,
disorder as compared with other eating disorders: An explor- A. J. (2008). A descriptive study of non-​obese persons with
atory investigation. International Journal of Eating Disorders, night eating syndrome and a weight matched comparison
42, 31–​39. group. Eating Behaviors, 9, 343–​351.
Gluck, M. E., Geliebter, A., & Satov, T. (2001). Night eating Lundgren J. L., Allison K. C., & Stunkard, A. J. (2006). Familial
syndrome is associated with depression, low self-​ esteem, aggregation in the night eating syndrome. International
reduced daytime hunger, and less weight loss in obese outpa- Journal of Eating Disorders, 39, 516–​518.
tients. Obesity Research, 9, 264–​267. Lundgren, J. D., Amsterdam, J., Newberg, A., Allison, K. C.,
Goel, N., Stunkard, A. J., Rogers, N. L., Van Dongen, H. P. A., Wintering, N., & Stunkard, A. J. (2009). Differences in
Allison, K. C., O’Reardon, J. P.,  . . .  Dinges, D. F. (2009). serotonin transporter binding affinity in patients with major
Circadian rhythm profiles in women with night eating syn- depressive disorder and night eating syndrome. Eating and
drome. Journal of Biological Rhythms, 24, 85–​94. Weight Disorders, 14, 45–​50.
Haedt-​Matt, A. A., & Keel, P. K. (2015). Affect regulation and Lundgren, J. D., Newberg, A., Allison, K. C., Wintering, N.,
purging: An ecological momentary assessment study in purg- Plöessl, K., & Stunkard, A. J. (2008). 123I-​ADAM SPECT
ing disorder. Journal of Abnormal Psychology, 124, 399–​411. imaging of serotonin transporter binding in patients

Allison, Lundgren 455

 with night eating syndrome:  A  pilot study. Psychiatry subcutaneous emphysema, pneumomediastimum, pneumo-
Research: Neuroimaging, 162, 214–​220. thorax, and pancytopenia. Electrolytes and Blood Pressure,
Lundgren J. D., Patrician, T., Breslin, F., Martin, L., Donnelly, 9, 32–​37.
J. E., & Savage, C. (2013). Evening hyperphagia and food Pinheiro, A. P., Bulik, C. M., Sullivan, P. F., & Machado, P. P.
motivation:  a preliminary study of neural mechanisms. (2008). An empirical study of the typology of bulimic symp-
Eating Behaviors, 14, 447–​450. toms in young Portuguese women. International Journal of
Lundgren, J. D., Shapiro, J. R., Bulik, C. (2008). Night eat- Eating Disorders, 41, 251–​258.
ing patterns of patients with bulimia nervosa: A preliminary Rand, C. S. W., Macgregor, A. M., & Stunkard, A. J. (1997). The
report. Eating and Weight Disorders, 13, 171–​175. night eating syndrome in the general population and among
Machado, P. P., Machado, B. C., Goncalves, S., & Hoek, H. post-​operative obesity surgery patients. International Journal
W. (2007). The prevalence of eating disorders not other- of Eating Disorders, 22, 65–​69.
wise specified. International Journal of Eating Disorders, 40, Raymond, N. C., Neumeyer, B., Warren, C. S., Lee S. S., &
212–​217. Peterson, C. B. (2003). Energy intake patterns in obese
Marshall, H. M., Allison, K. C., O’Reardon, J. P., Birketvedt, women with binge eating disorder. Obesity Research, 11,
G., & Stunkard, A. J. (2004). Night eating syndrome among 869–​879.
nonobese persons. International Journal of Eating Disorders, Root, T. L., Thornton L., M., Lindroos, A. K., Stunkard, A. J.,
35, 217–​222. Lichtenstein P, Pederson, N. L.,  . . .  Bulik, C. M. (2010).
McClure, S. M., Laibson, D. I., Lowenstein, G., & Cohen, J. D. Shared and unique genetic and environmental influences on
(2004). Separate neural systems value immediate and delayed binge eating and night eating. Eating Behaviors, 11, 92–​98.
monetary rewards. Science, 306, 503–​507. Saddichha, S., Babu, G. N., & Chandra, P. (2012). Orthorexia
Meule, A., Allison, K. C., & Platte, P. (2014). Emotional eating nervosa presenting as prodrome of schizophrenia.
moderates the relationship of night eating with binge eat- Schizophrenia Research, 134, 110.
ing and body mass. European Eating Disorder Review, 22, Sateia M. J. (Ed.). (2005). International classification of sleep
147–​151. disorders (2nd ed.). Westchester, IL:  American Academy of
Mitchell, M. E., Pyle, R. L., Hatsukami, D., & Eckert, E. D. Sleep Medicine.
(1986). What are atypical eating disorders? Psychosomatics, Sawyer S. M., Whitelaw, M., LeGrange, D., Yeo, M., &
27, 21–​28. Hughes, E. K. (2016). Physical and psychological morbid-
Moroze, R. M., Dunn, T. M., Craig Holland, J., Yager, J., ity in adolescents with atypical anorexia nervosa. Pediatrics,
Weintraub, P. (2015). Microthinking about micronutri- 137:e20154080.
ents:  A  case of transition from obsessions about healthy Schenck, C. H., & Mahowald, M. W. (1994). Review of noc-
eating to near-​fatal “orthorexia nervosa” and proposed diag- turnal sleep-​related eating disorders. International Journal of
nostic criteria. Psychosomatics, 56, 397–​403. Eating Disorders, 15, 343–​356.
Morse, S. A., Ciechanowski, P. S., Katon, W. J., & Hirsch, I. Schienele, A., Schäfer, A., Herman, A., & Vaitl, D. (2009).
B. (2006). Isn’t this just bedtime snacking? The potential Binge-​eating disorder:  Reward sensitivity and brain activa-
adverse effects of night-​eating symptoms on treatment adher- tion to images of food. Biological Psychiatry, 65, 654–​661.
ence and outcomes in patients with diabetes. Diabetes Care, Schorr, M., Thomas, J. J., Eddy, K. T., Dichtel, L. E., Lawson, E.
29, 1800–​1804. A., Meenaghan, E.,  . . .  Miller, K. K. (2016). Bone density,
Munn-​Chernoff, M. A., Keel, P. K., Klump, K. L., Grant, J. D., body composition, and psychopathology of anorexia nervosa
Bucholz, K. K., Madden, P. A.,  . . .  Duncan, A. E. (2015). spectrum disorders in DSM-​ IV vs DSM-​ 5. International
Prevalence of and familial influences on purging disorder in Journal of Eating Disorders. doi:  10.1002/​eat.22603 [Epub
a community sample of female twins. International Journal of ahead of print].
Eating Disorders, 48, 601–​606. Schreiber-​Gregory, D. N., Lavender, J. M., Engel, S. G.,
Mustelin, L., Lehtokari, V. L., & Keski-​Rahkonen, A. (2016). Wonderlich, S. A., Crosby, R. D., Peterson, C. B.,   .  .  . 
Other specified and unspecified feeding or eating disorders Mitchell, J. E. (2013). Examining duration of binge eating
among women in the community. International Journal of episodes in binge eating disorder. International Journal of
Eating Disorders. doi:  10.1002/​eat.22586 [Epub ahead of Eating Disorders, 46, 810–​814.
print]. Silén, Y., Raevuori, A., Jüriloo, E., Tainio, V. M., Marttunen,
Olbrich, K., Mühlhans, B., Allison, K. C., Hahn, E. G., Schahin, M., & Keski-​Rahkonen, A. (2015). Typical versus atypical
S. P., & de Zwaan, M. (2009). Night eating, binge eating anorexia nervosa among adolescents: Clinical characteristics
and related features in patients with obstructive sleep apnea and implications for ICD-​ 11. European Eating Disorders
syndrome. European Eating Disorders Review, 17, 120–​127. Review, 23, 345–​351.
O’Reardon, J. P., Allison, K. C., Martino, N. S., Lundgren, J. D., Stice, E., Marti, C. N., & Rohde, P. (2013). Prevalence, inci-
Heo, M., & Stunkard, A. J. (2006). A randomized placebo-​ dence, impairment, and course of the proposed DSM-​5 eat-
controlled trial of sertraline in the treatment of the night eat- ing disorder diagnoses in an 8-​year prospective community
ing syndrome. American Journal of Psychiatry, 163, 893–​898. study of young women. Journal of Abnormal Psychology, 122,
O’Reardon, J. P., Ringel, B. L., Dinges, D. F., Allison, K. C., 445–​457.
Rogers, N. L., Martino, N. S., & Stunkard, A. J. (2004). Striegel-​Moore, R. H., Dohm, F. A., Hook, J. M., Schreiber,
Circadian eating and sleeping patterns in the night eating G. B., Crawford, P. B., & Daniels, S. R. (2005). Night
syndrome. Obesity Research, 12, 1789–​1796. eating syndrome in young adult women:  Prevalence and
Park, S. W., Kim, J. Y., Go, G. J., Jeon, E. S., Pyo, H. J., & correlates. International Journal of Eating Disorders, 37,
Kwon, Y. J. (2011). Orthorexia nervosa with hyponatremia, 200–​206.

456 Emerging Syndromes

Striegel-​Moore, R. H., Franko, D. L., May, A., Ach, E., Tzischinsky, O., & Latzer, Y. (2004). Nocturnal eat-
Thompson, D., & Hook, J. M. (2006). Should night eating ing: Prevalence, features, and night sleep among binge eat-
syndrome be introduced in the DSM? International Journal ing disorder and bulimia nervosa patients in Israel. European
of Eating Disorders, 39, 544–​549. Eating Disorders Review, 12, 101–​109.
Striegel-​Moore, R. H., Franko, D. L., Thompson, D., Affenito, Vinai, P., Cardetti, S., Studt, S., Carpegna, G., Ferrato, N.,
S., & Kraemer, H. C. (2006). Night eating: Prevalence and Vallauri, P.,  . . .  Manconi, M. (2014). Clinical validity of the
demographic correlates. Obesity Research, 14, 139–​147. descriptor: “Presence of a belief that one must eat in order
Striegel-​Moore, R. H., Franko, D. L., Thompson, D., Affenito, to get to sleep” in diagnosing the Night Eating Syndrome.
S., May, A., & Kraemer, H. C. (2008). Exploring the typol- Appetite, 75, 46–​48.
ogy of night eating syndrome. International Journal of Eating Vinai, P., Provini, F., Antelmi, E., Marcatelli, M., Cardetti,
Disorders, 41, 411–​418. S., Vinai, L.,  . . .  Allison, K. C. (2015). Alexithymia is
Striegel-​Moore, R. H., Franko, D. L., Thompson, D., Barton, not related to severity of night eating behavior:  a useful
B., Schreiber, G. B., & Daniels, S. R. (2005). An empirical distinction from other eating disorders. Eating Behaviors,
study of the typology of bulimia nervosa and its spectrum 17, 94–​98.
variants. Psychological Medicine, 35, 1563–​1572. Wade, T. D. (2007). A retrospective comparison of purging
Stunkard, A. J., Allison, K. C., Geliebter, A., Lundgren, J. D., type disorders:  Eating disorder not otherwise specified and
Gluck, M. E., & O’Reardon, J. P. (2009). Development of bulimia nervosa. International Journal of Eating Disorders,
criteria for a diagnosis:  Lessons from the night eating syn- 40, 1–​6.
drome. Comprehensive Psychiatry, 50, 391–​399. Wade, T. D., Bergin, J. L., Tiggemann, M., Bulik, C. M., &
Stunkard, A. J., Berkowitz, R., Wadden, T. A., Tanrikut, C., Fairburn, C. G. (2006). Prevalence and long-​term course of
Reiss, E., & Young, L. (1996). Binge eating disorder and the lifetime eating disorders in an adult Australian twin cohort.
night-​eating syndrome. International Journal of Obesity and Australian and New Zealand Journal of Psychiatry, 40, 121–​128.
Related Metabolic Disorders, 20, 1–​6. Wade T. D., Crosby, R., & Martin, N. G. (2006). Use of latent
Stunkard, A. J., Grace, W. J., & Wolff, H. G. (1955). The profile analysis to identify eating disorder phenotypes in an
night-​eating syndrome:  A  pattern of food intake among adult Australian twin cohort. Archives of General Psychiatry,
certain obese patients. American Journal of Medicine, 63, 1377–​1384.
19, 78–​86. Wells, A., & Matthews, G. (1994). Attention and emotion: A clin-
Sullivan, P. F., Bulik, C. M., & Kendler, K. S. (1998). The epi- ical perspective. Hove, UK: Erlbaum.
demiology and classification of bulimia nervosa. Psychological Winkelman, J. W. (2003). Treatment of nocturnal eating syn-
Medicine, 28, 599–​610. drome and sleep-​related eating disorder with topiramate.
Swanson, S. A., Horton, N. J., Crosby, R. D., Micali, N., Sleep Medicine, 4, 243–​246.
Sonneville, K. R., Eddy, K., & Field, A. E. (2014). A Winkelman, J. W. (2006). Efficacy and tolerability of open-​label
latent class analysis to empirically describe eating disorders topiramate in the treatment of sleep-​related eating disor-
through developmental stages. International Journal of Eating der: A retrospective case series. Journal of Clinical Psychiatry,
Disorders, 47, 762–​772. 67, 1729–​1734.
Tholin, S., Lindroos, A. K., Tynelius, P., Akerstedt, T., Stunkard, Wonderlich, S. A., Joiner, T. E., Keel, P. K., Williamson, D. A., &
A. J., Bulik, C. M., & Rassmussen, F. (2009). Prevalence Crosby, R. D. (2007). Eating disorder diagnoses: Empirical
of night eating in obese and non-​obese twins. Obesity, 17, approaches to classification. American Psychologist, 62,
1050–​1055. 167–​180.
Tobin, D. L. Griffing, A., & Griffing, S. (1997). An examination Zamora, M. L. C., Bonaechea, B. B., Sánchez, F. G., & Rial, B.
of subtype criteria for bulimia nervosa. International Journal R. (2005). Orthorexia nervosa: A new eating behavior disor-
of Eating Disorders, 22, 179–​186. der? Actas Españolas de Psiquiatría, 33, 66–​68.

Allison, Lundgren 457

 C H A PT E R


Eating Disorders and Problematic Eating
24 Behaviors After Bariatric Surgery

Molly Orcutt, Kristine Steffen, and James E. Mitchell

Abstract
Bariatric surgery is the most effective treatment for severe obesity and its weight-​related comorbidities.
As the use of bariatric surgery has increased, concerns have arisen about problematic eating behaviors
(EBs) and eating disorders (EDs) in this population. This chapter describes the current literature detailing
EDs and problematic EBs and weight control practices (WCPs) in the post-​bariatric-​surgery population.
It begins with a description of EDs in this population. Next, it describes problematic EBs, followed by a
review of postoperative gastrointestinal problems that influence EBs. Third, it reviews the WCPs that
can evolve. Finally, it describes models of “food addiction” as they apply to eating-​related pathology in
the post-​bariatric-​surgery population.
Key Words:  Eating disorder, Obesity, Bariatric surgery, Food addiction, Eating behavior

Introduction 2016) and involves surgically partitioning the stom-

Obesity is increasing in prevalence worldwide, ach to create a small gastric pouch, which is con-
and is associated with numerous comorbidities nected to a distal portion of the small intestine,
including type 2 diabetes mellitus, sleep apnea, and bypassing the first part of the small intestine, the
cardiovascular disease (Picot et al., 2009). The most duodenum. Finally, SG involves surgically dividing
effective treatment for severe obesity and its weight-​ the stomach vertically to reduce its size by about
related comorbidities is bariatric surgery (Colquitt, 75%. Sleeve gastrectomy is currently the most com-
Pickett, Loveman, & Frampton, 2014). mon bariatric surgery performed in the United
At this time, the most frequently performed bar- States and accounts for approximately 53.8% of
iatric surgeries are sleeve gastrectomy (SG), Roux-​ bariatric surgeries performed (ASMBS, 2016).
en-​Y gastric bypass (RYGB), and laparoscopic Historically, weight loss and remission of weight-​
adjustable gastric banding (LAGB). The least com- related comorbidities were thought to be a result of
mon of these surgeries, LAGB, accounts for approx- purely anatomical changes secondary to the surgery
imately 5.7% of bariatric procedures performed (Noel, Still, Argyropoulos, Edwards, & Gerhard,
in the United States (ASMBS, 2016) and involves 2016). For example, the decrease in stomach size in
placing a band at the top of the stomach to anatom- LAGB, RYGB, and SG leads to earlier satiety and
ically restrict food intake. This surgery is being used an expected decrease in intake. Additionally, the
less frequently over time, and many centers in the alteration of the digestive tract in RYGB causes a
United States have stopped offering this procedure reduction in the absorption of nutrients, leading to
given problems with lack of adequate weight loss weight loss. However, a growing body of research
and long-​ term complications. Roux-​ en-​Y gastric suggests that the mechanisms by which the RYGB,
bypass accounts for approximately 23.1% of pro- and likely the SG, lead to weight loss, are through
cedures performed in the Untied States (ASMBS, other physiological changes.

458
 Postsurgical alterations in gastrointestinal (GI) robust predictor of poor postsurgical weight out-
physiology include changes in the activity of cer- comes (Kalarchian et al., 2016; Meany, Conceicao,
tain gut hormones, bile acids, inflammatory factors, & Mitchell, 2014).
and gut bacteria (microbiome) (Fouladi, Mitchell, Eating disorders can have a significant impact on
Wonderlich, & Steffen, 2016; Makaronidis & health and functioning post-​surgery. Furthermore,
Batterham, 2016; Sweeney & Morton, 2013). eating disorders, in particular AN and BN, may
Emerging evidence suggests that these alterations result in serious medical complications associ-
in physiology are involved in bidirectional com- ated with low BMI and compensatory behaviors
munication with the central nervous system (Cryan (Conceicao et al., 2013; Marino et al., 2012). Other
& Dinan, 2012). Further understanding of the eating disorder behaviors or problematic eating
mechanisms of weight loss and disease remission behaviors, such as loss of control (LOC) eating and
are likely to be an important focus of research, as grazing, may be associated with suboptimal weight
evidence suggests that postsurgery outcomes may loss, which may negatively influence remission of
be at least partially dependent on the changes that comorbidities associated with obesity (Conceicao,
occur in these variables and the interactions among Bastos, et  al., 2014; Conceicao, Mitchell, Engel,
them. Weight-​related outcomes following surgery et al., 2014; Conceicao, Mitchell, Vaz, et al., 2014;
range from substantial weight loss to modest or Conceicao, Utzinger, & Pisetsky, 2015; Kalarchian
minimal weight loss and, rarely, excessive weight et al., 2016; Meany et al., 2014).
loss (Courcoulas et al., 2013). For example, median In this chapter, we describe the current litera-
weight loss following RYGB is approximately 30% ture detailing EDs and problematic eating behav-
of baseline body weight within 3 years post-​surgery iors (EBs) and weight control practices (WCPs) in
(Courcoulas et  al., 2013). In comparison, median the post-​bariatric surgery population. We begin our
weight loss of baseline body weight at 3 years with discussion with a description of EDs in this popula-
LAGB in the same study was 15.9% (Courcoulas tion. Next, we describe problematic EBs, followed
et  al., 2013). Weight loss is followed by regain of by a review of postoperative gastrointestinal prob-
variable amounts of weight in approximately 20%–​ lems that influence EBs. Third, we review the WCPs
30% of individuals (Courcoulas et  al., 2013; de that can evolve. Finally, we describe models of “food
Hollanda et al., 2015; Hsu et al., 1998). addiction” as they apply to eating-​related pathology
Although less common, it is being increasingly in the post-​bariatric surgery population.
recognized in clinical settings and a limited body
of literature that a small subset of patients are diag- Binge Eating Disorder
nosed with eating disorders post bariatric surgery Binge eating disorder (BED) is the most com-
(Conceicao et al., 2013; Marino et al., 2012; Opozda, mon eating disorder in general (Swanson, Crow,
Chur-​Hansen, & Wittert, 2016) In particular, the Le Grange, Swendsen, & Merikangas, 2011), as
current evidence suggests that the prevalence of eat- well as the disorder most often diagnosed in the
ing disorder diagnosis in the post-​bariatric surgery bariatric surgery population (Kalarchian et  al.,
population ranges from 0 to 17% (Opozda et  al., 2016; Opozda et al., 2016). In the Diagnostic and
2016). Indeed, a growing body of research indi- Statistical Manual of Mental Disorders, 5th Edition
cates that traditional eating disorders (EDs) such as (DSM 5)  BED is defined as recurrent objective
anorexia nervosa (AN) and bulimia nervosa (BN), binge eating (OBE) episodes occurring at least
can occur after bariatric surgery and in some cases once weekly for 3  months without the presence
patients present for treatment with very low BMIs of compensatory behaviors (American Psychiatric
and significant medical comorbidities (Conceicao Association [APA], 2013). An OBE is defined as the
et al., 2013; Marino et al., 2012). In addition, binge ingestion of an objectively large amount of food in a
eating disorder (BED) is also common prior to bar- discrete period of time that is larger than what most
iatric surgery (4%–​49%) (Kalarchian et  al., 2016; people would eat in a similar situation. A  sense
Niego, Kofman, Weiss, & Geliebter, 2007; Opozda of LOC and distress regarding the OBE must be
et al., 2016), and in a smaller percentage of patients present. Additionally, DSM-​5 criteria specify that
(0–​17%) (Kalarchian et  al., 2016; Opozda et  al., the OBEs need to be associated with at least three
2016), the disorder either persists or develops de of the five following features:  eating more rapidly
novo postoperatively. Some evidence suggests that than normal; eating large amounts of food when
eating with a sense of loss of control, a hallmark fea- not hungry; eating until uncomfortably full; eating
ture of BED, following bariatric surgery is the most alone due to embarrassment about how much one

Orcut t, Steffen, Mitchell 459

is eating; and feeling guilty, disgusted, or depressed Conceicao and colleagues assessed both SBE and
after the OBE episode. OBE in patients after bariatric surgery, and found
Prevalence rates of BED vary widely from that no OBE episodes were reported at 6  months
4% to 49% among candidates for bariatric sur- after surgery. However, OBE episodes were reported
gery (Kalarchian et  al., 2016; Niego et  al., 2007; after 2 years (Conceicao, Mitchell, Vaz, et al., 2014).
Opozda et al., 2016). This wide range likely reflects Thus, the amount of food that can be ingested by
to some extent the changes in diagnostic criteria individuals post-​ surgery does appear to increase
from previous versions of the DSM to the current over time (Sarwer et  al., 2008). Although OBE
version, but primarily is attributable to the vary- episodes are infrequent, SBE episodes appear to
ing assessment methods and study designs used in be quite common after surgery. In fact, Conceicao
generating these estimates (Conceicao, Utzinger et  al. reported all individuals that had undergone
et al., 2015; Kalarchian et al., 2016; Opozda et al., restrictive and malabsorptive procedures reported
2016). Research has been somewhat mixed as to having SBE episodes (Conceicao, Mitchell, Vaz,
whether or not preoperative BED is a predictor of et al., 2014).
outcomes post-​surgery, with a majority of studies Given the low frequency of OBE episodes after
not finding an association (Meany et  al., 2014). surgery, the prevalence of BED is fairly low, rang-
However, research does suggest that the BED diag- ing from 0 to 17.2% (Opozda et  al., 2016). This
nosis is associated with increased rates of comorbid is in contrast to the fairly high levels of LOC eat-
psychopathology, which is concerning due to the ing prior to and after bariatric surgery. Indeed, rates
association of multiple comorbidities and poor out- of LOC in bariatric patients prior to surgery range
comes postoperatively (van Hout, Verschure, & van from 13.3% to 61% (Colles, Dixon, & O’Brien,
Heck, 2005). 2008b; Devlin et  al., 2016; White et  al., 2010),
After surgery, BED diagnosis is further compli- and rates vary from 11.7% to 39% after surgery
cated by the surgically induced anatomical and phys- (Conceicao, Bastos, et al., 2014; Devlin et al., 2016;
iological restrictions on the amount of food that can White et al., 2010). Rates appear to decrease as time
be ingested. Most patients become unable to ingest from surgery increases with the lowest rate of LOC
objectively large amounts of food in a discrete period (11.7%) found 3  years postoperatively (Devlin
of time (Meany et al., 2014). Therefore, it is impos- et al., 2016). While the definitions of BE and LOC
sible for individuals to ingest the amount of food vary across studies, a growing body of literature
necessary to meet criteria for an OBE episode, at indicates that the presence of LOC eating after sur-
least initially, after surgery. It has been argued that gery may be associated with worse weight outcomes
the amount of food necessary to meet criteria for an (Colles et al., 2008a; Mitchell et al., 2016; White
OBE should be considered in the context of the time et al., 2010). In fact, postoperative LOC eating has
since surgery and the type of surgery. been one of the only consistent predictors of weight
The literature does suggest that the LOC over loss outcomes following bariatric surgery. Therefore,
eating more modest amounts of food can continue LOC eating may be important to identify and treat
after surgery, or more rarely can develop after sur- postoperatively.
gery, and some have speculated that the experience
of LOC, rather than the amount of food ingested, Bulimia Nervosa
may be more important in defining binge eating Bulimia nervosa is defined in DSM-​5 as recur-
(BE) (Meany et  al., 2014). Loss of control over rent episodes of OBEs with a sense of LOC, recur-
eating smaller amounts of food has been termed rent inappropriate compensatory behaviors to
subjective BE (SBE) (Conceicao, Mitchell, Vaz, prevent weight gain, and overvaluation of shape
et  al., 2014). Some have suggested that it may be and weight (APA, 2013). Both OBEs and compen-
important to identify both SBE and OBE, espe- satory behaviors must occur at least once weekly for
cially post-​ surgery (Colles, Dixon, & O’Brien, 3 months. Prior to surgery, studies suggest that the
2008a; Conceicao, Bastos, et  al., 2014; White, prevalence of BN is low, with one study reporting
Kalarchian, Masheb, Marcus, & Grilo, 2010). Thus, that approximately 0–​3% of individuals met DSM-​
future research should strive toward operationaliz- IV criteria prior to surgery (de Zwaan et al., 2010;
ing appropriate diagnostic criteria for BED in the Mitchell et  al. 2015; Opozda et  al., 2016). It has
post-​bariatric surgery population and developing been suggested that estimates of prevalence prior to
psychological and/​or pharmacological interventions surgery may be influenced by the patients’ motiva-
to target symptoms of this disorder. tion to maintain surgery eligibility, and therefore

460 Problematic Eating Behaviors After Bariatric Surgery

some patients may minimize or deny symptoms of prevalence of 0–​2% (Devlin et al., 2016), but more
BN (Mitchell et al., 2015; Opozda et al., 2016). systematically collected data are needed. As with
There are few studies examining the effects of a other compensatory behaviors, the motivation for
history of or current diagnosis of BN on postopera- using laxatives and diuretics must be elicited.
tive outcomes. A study by de Zwann and colleagues
suggested that BN prior to surgery is associated with Anorexia Nervosa
an increased likelihood of compensatory vomiting Anorexia nervosa is characterized by restric-
and SBEs after surgery (de Zwaan et  al., 2010). tion of energy intake leading to significantly low
Similar to the diagnosis of BED after surgery, the body weight (APA, 2013). The DSM-​ 5 defines
diagnosis of BN after surgery is complicated by the significantly low body weight as a weight less than
majority of individuals being incapable of engaging minimally normal in the context of age, sex, devel-
in traditional OBEs for a period of time after sur- opment, and physical health. Also, AN is distin-
gery. Yet, as the study by de Zwaan and colleagues guished by an intense fear of gaining weight and
reported, SBEs and self-​induced vomiting still occur overvaluation of shape and weight. Individuals
(de Zwaan et al., 2010). Case reports have also iden- seeking bariatric surgery are obese, and would not
tified individuals who reported OBEs and/​or SBEs meet criteria for diagnosis of AN immediately pre-
followed by self-​ induced vomiting (Conceicao, operatively. There are insufficient data examining
et al., 2013; Mitchell, 1985). Postoperative cases of lifetime history of AN in individuals seeking bar-
BN are thought to be rare. A recent large prospec- iatric surgery.
tive longitudinal study reported a prevalence rate of Postoperative individuals with AN-​like presenta-
0 at 2 and 3 years after surgery (Kalarchian et al., tions have been described in the literature. In fact,
2016). Case reports indicate that individuals who these patients appear to present to ED specialized
develop BN after bariatric surgery in general tend care facilities fairly frequently. Case reports and ser-
to be much older and have a later age of onset than ies have described patients who develop symptoms of
individuals who have BN without a history of bar- AN, including dietary restriction, significant weight
iatric surgery (Conceicao, et al., 2013). loss, fear of weight gain, and disturbance in evalua-
The development of BN postoperatively may tion of shape and weight (Conceicao et al., 2013).
be influenced or triggered by postoperative experi- Relative to patients with traditional AN, individu-
ences, and it is critical to consider behaviors that als who have a history of bariatric surgery are usu-
occur in the postoperative period in the context ally at a higher weight, are older in age, and have
of the type of surgery and the length of time since a later age of onset at the time of presentation for
surgery. For example, it is common for patients to treatment (Conceicao et al., 2013). Determination
have episodic vomiting after surgery (Kalarchian of low weight for diagnosis in post-​bariatric surgery
et  al., 2014). Vomiting may begin spontaneously patients should include evaluation of weight trajec-
or may be used in a voluntary manner intended to tory, EBs, and the physical health of the individual,
ease discomfort, but it can evolve into a mechanism since some of these individuals may be in a state of
of weight control (de Zwaan et al., 2010). To dis- starvation after massive weight loss despite having a
cern whether vomiting is a compensatory behavior, BMI greater than 18.5 kg/​m2 (Conceicao, Mitchell,
the reasons for and expected outcomes of vomiting Engel, et al., 2014; Conceicao et al., 2013).
should be elicited. As in BN, it has been suggested that postop-
Dumping is excessive diarrhea that occurs after erative experiences may trigger the development
ingestion of sweets and/​or large amounts of food. of AN symptoms. For example, the experience of
Dumping is another behavior that occurs com- rapid weight loss is often followed by a weight pla-
monly after surgery, but case reports indicate that teau, wherein the individual desires further weight
it can become a compensatory behavior (Conceicao loss, may trigger the development of AN symp-
et  al., 2013). Assessment of whether dumping is toms. Further, postoperative recommendations by
self-​induced for weight loss or weight control is physicians and other medical professionals usually
critical in determining whether it is a compensatory include strict dietary restriction, and the directive to
behavior. cut food in small pieces, and avoid of certain food
Evidence regarding the misuse of laxatives groups (Conceicao et al., 2013). These behaviors are
and diuretics after surgery is scarce. The available similar to the behaviors and rituals seen in individu-
research suggests that use of laxatives or diuretics als with AN and presumably may trigger the devel-
for weight control in this population is rare with a opment of AN symptoms in some postbariatric

Orcut t, Steffen, Mitchell 461

surgery individuals. Distinguishing the motivation is not well defined in bariatric surgery populations
behind these behaviors can help one determine and is further complicated by diagnostic overlap
whether the behavior is adaptive or pathological. with NES (Devlin et  al., 2016). Distinguishing
In this population, fear of suboptimal weight loss between NES and NSRED is vital, because the dis-
or weight regain is realistic. Recommendations after orders have different treatment approaches (Vinai
surgery likely include frequent weighing and moni- et al., 2015).
toring of the various factors (e.g., exercise and food
intake) that may influence weight. Furthermore, Problematic Eating Behaviors
assessment of shape concerns in the post-​bariatric The bariatric surgery literature has identified
surgery population are complicated by frequent multiple EBs that may have an impact on weight
concerns about excessive skin (Ellison, Steffen, & outcomes but probably do not represent distinct
Sarwer, 2015). Excessive loose or hanging skin is a EDs. Nonetheless, identification of these behav-
frequent result of rapid and substantial weight loss. iors may be important to optimize weight loss out-
Greater body dissatisfaction and depressive symp- comes post-​surgery. However the lack of adequate
toms have been associated with the development of development in this nomenclature will become
excessive skin, and traditional assessment methods obvious.
used to measure body dissatisfaction may not mea-
sure the specific body and/​or shape concerns of this Grazing
population (Ramalho et al., 2015). What is often referred to as “grazing” has been
the focus of emerging research in the bariatric sur-
Night Eating Syndrome and Nocturnal gery literature. Although the term “grazing” has a
Sleep-​Related Eating Disorder disparaging connotation, it is the most widely used
In the DSM-​5, night eating syndrome (NES) is term currently. Various definitions and criteria have
designated as an otherwise specified feeding or eat- also been proposed to describe or diagnose this
ing disorder (OSFED) (APA, 2013). Night eating behavior, which is commonly defined as repetitively
syndrome is characterized by recurrent, conscious eating small amounts of food during the day in an
episodes of night eating occurring after awakening unplanned manner (Conceicao, Mitchell, Engel,
from sleep and/​or ingestion of greater than 25% of et al., 2014). This behavior may or may not occur
food after the evening meal. Episodes must occur at in response to sensations of hunger. Repetitive eat-
least twice weekly and cause distress to meet crite- ing, as used in defining grazing, is eating more than
ria. Prior to surgery, prevalence ranges from 1.9% to twice in the same time period during the day with-
17.7% (Allison et al., 2006; Mitchell et al., 2015), out prolonged gaps between eating episodes.
and preoperative diagnosis of NES may be associ- In one recent report, grazing was divided into
ated with preoperative diagnosis of BED (Mitchell two subtypes (Conceicao, Mitchell, Engel, et  al.,
et  al., 2015). After surgery, evidence suggests 2014). The first subtype is termed “compulsive
NES rates may either continue at similar levels or grazing.” Compulsive grazing involves LOC eating
decrease (Adami, Meneghelli, & Scopinaro, 1999; wherein the individual attempts to resist eating, but
Colles & Dixon, 2006; Colles et al., 2008a; Devlin the individual is unable to resist, and returns to eat-
et al., 2016). Night eating syndrome, before or after ing small amounts of food. Noncompulsive grazing
surgery, has not been associated with adverse weight is characterized by repetitively eating without a sense
loss outcomes (Devlin et  al., 2016; de Zwaan, of LOC. Rather, the eating is without mindful-
Marschollek, & Allison, 2015). ness. Grazing differs from an OBE episode because
Nocturnal sleep-​ related eating disorder the amount of food is smaller, the time period is
(NSRED), a parasomnia, differs from NES in that it not limited, and the eating episode is unplanned.
occurs during periods of partial arousal and reduced Subjective BE episodes differ from grazing because
awareness of the behavior (APA, 2013). Awareness an SBE must occur in a discrete time period.
of eating episodes may only occur after awakening Although post-​bariatric surgery recommendations
the next morning and finding that food is missing include eating small meals and snacks throughout
or cooking appliances have been used (Winkelman, the day, the meals and snacks are presumed to be
1998). Substances ingested may be edible or ined- planned, demarcated in response to hunger and
ible. Associated with the use of medications, includ- satiety cues, and in accordance with nutritional
ing zolpidem and quetiapine, NSRED also occurs as needs (Conceicao et al., 2015). These recommenda-
a primary sleep disorder. The prevalence of NSRED tions clearly differ from grazing behaviors.

462 Problematic Eating Behaviors After Bariatric Surgery

 Determining the prevalence of grazing, its effects or eating during stressful life situations (Canetti,
on weight outcomes, and its associated features has Berry, Elizur, 2009). It was estimated to occur
been complicated by competing assessment meth- in bariatric surgery candidates at a rate of 38%–​
ods and definitions. Presurgery prevalence data is 59% (Opolski, Chur-​Hansen, & Wittert, 2015).
limited, but in a study by Colles and colleagues Whether this is a unique behavior is debatable,
it was found to be 26.4% (Colles et  al., 2008a). and this points to the lack of differentiation is this
After surgery, the prevalence may increase to 46.6% nomenclature. However, one study found emo-
(Kofman, Lent, & Swencionis, 2010). In the lim- tional eating was more common among women
ited data available, grazing following surgery has than men (Gade, Rosenvinge, Hjelmesaeth, &
been linked to reduced weight loss or increased Friborg, 2014). The link between emotional eating
weight regain (Conceicao, Mitchell, Engel, et  al., and weight outcomes following surgery is unclear.
2014). Such eating behavior, over time, is thought Some studies have reported no association between
to lead to eating excess calories. Emerging evidence preoperative and postoperative emotional eating
points to a possible association between grazing, and subsequent weight loss, while others reported
LOC eating, and BE, with comorbidities including poorer weight loss outcomes (Opozda et al., 2016).
symptoms of depression and other problematic EBs Emotional eating has been associated with other
(Colles et al., 2008a; Kofman et al., 2010). problematic EBs, including BED (Banerjee, Ding,
Mikami, & Needleman, 2013; Fischer et al., 2007;
Picking and Nibbling Mathus-​Vliegen, 2007).
Other problematic EBs may be related to graz-
ing, since these behaviors also involve repetitive eat- High-​Calorie Fluid Consumption
ing. Terms used have included “picking or nibbling” A small number of studies have examined exces-
and “snacking.” Picking or nibbling is defined in sive high-​calorie fluid or soft food consumption
the Eating Disorder Examination as eating food after bariatric surgery. Reported prevalence rates
between meals and snacks that occurs in a repeti- in one small study ranged from 33% prior to sur-
tive and unplanned manner (Fairburn & Cooper, gery and 17% after surgery (Hsu, Betancourt, &
2000). The amount of food consumed should be Sullivan, 1996). It has been suggested by some that
uncertain prior to the episode and should not be consuming large amounts of highly caloric fluids
trivial. Nibbling, in most definitions, can be differ- or soft food allows patients to avoid the discomfort
entiated from grazing because it does not involve related to the fullness that results from excess food
LOC, and occurs between meals or snacks. This is ingestion but to still consume calories (Hsu et al.,
in contrast to grazing, which often leads to skipped 1998; Hsu et al., 1996) Studies have suggested that
meals. The initial uncertainty regarding the amount high-​calorie fluid consumption is associated with
of food to be eaten also separates nibbling from reduced weight loss and/​or greater weight regain,
grazing in the current nomenclature. but research in this area has been sparse (Brolin,
Prior to surgery, the prevalence of picking and Robertson, Kenler, & Cody, 1994; Hsu et  al.,
nibbling behaviors ranges from 29.5% to 53.0% 1996; Sugarman et  al., 1992; Yale, 1989; Yale &
(Busetto et  al., 2002; Conceicao, Mitchell, Vaz, Weiler, 1991).
et al., 2014; Devlin et al., 2016). After surgery, the
prevalence has been estimated to occur in the range Sweet Eating
of 32.2% to 47.1% (Conceicao, Mitchell, Vaz, Another problematic EB that has been described
et al., 2014; Devlin et al., 2016; de Zwaan et al., is characterized by eating excessive amounts of high-​
2010). Three studies have reported no associa- calorie dessert-​type foods and is termed “sweet eat-
tion between picking and nibbling and weight loss ing.” Standard definitions of this behavior have not
(Busetto et al., 2002; Devlin et al., 2016; de Zwaan been established, and its validity has been debated
et al., 2010), while another study found a signifi- for years. Preoperative prevalence is estimated to be
cant association between picking or nibbling and between 30.4% and 43.8% (Busetto et  al., 2002;
poorer weight loss outcomes (Conceicao, Mitchell, Busetto et  al., 2005). In these studies, sweet eat-
Vaz, et al., 2014). ing was defined as a patient craving simple carbo-
hydrates, with cravings present continuously or
Emotional Eating triggered by emotional or physiological situations
Emotional eating is a term that has been used (Busetto et  al., 2002). Postoperative sweet eat-
to describe eating in response to emotional stress ing was originally thought to be influenced by the

Orcut t, Steffen, Mitchell 463

type of surgery; however, studies report significant 2016). In the months following malabsorptive
percentages of individuals engaging in sweet eat- surgery, dumping syndrome is quite common and
ing after RYGB and LAGB. In fact, postoperative estimated to occur in 15.7% to 76% of individu-
prevalence of sweet eating is estimated to be 36% als (Banerjee et al., 2013; Cawley, Sweeney, Kurian,
to 62% (Leite Faria, de Oliveira Kelly, Pereira Faria, & Beane, 2007; Mallory et  al., 1996; Sugerman
& Kiyomi Ito, 2009; Lindroos, Lissner, & Sjostrom, et  al., 1987). It has been suggested that dump-
1996). The current literature does not suggest an ing syndrome symptoms may decrease as patients
association between sweet eating and poorer weight adapt their food choices and speed of food ingestion
loss outcomes postoperatively (Hudson et al., 2002; (Banerjee et al., 2013).
Lindroos et al., 1996). The high prevalence of indi- Dumping syndrome is usually divided into
viduals engaging in the behavior and the lack of two types, termed early and late. Early dumping
effect on weight loss outcomes brings into ques- syndrome symptoms is more likely to be used as a
tion whether this behavior is, in fact, distinct or compensatory mechanism, as late dumping is asso-
pathological. ciated with reactive hypoglycemia. Early dumping
develops 10–​30 minutes after the ingestion of food,
Chewing and Spitting and is caused by the rapid transit of calorie-​dense
Another problematic EB occurring in bariatric food into the small intestine resulting in a hyper-
surgery patients following surgery is chewing and osmolar state leading to abrupt fluid shifts (Berg
spitting out food. In contrast to characterizations of & McCallum, 2016). Symptoms of early dumping
chewing and spitting in the traditional eating dis- syndrome include abdominal cramping, nausea,
order population, studies have reported that chew- vomiting, diarrhea, flushing, dizziness, lightheaded-
ing and spitting in patients post bariatric surgery ness, and rapid heart rate. Although early dumping
is more often unrelated to weight control or psy- leads to diarrhea, it has not been linked to increased
chological distress (de Zwaan et al., 2010). Instead, weight loss after bariatric surgery (Banerjee et  al.,
chewing and spitting is a behavior used to be able 2013). As discussed earlier, there have been reports
to taste certain foods that are not tolerated well of dumping being used as a maladaptive compensa-
upon swallowing, since this may result in plugging. tory behavior (Conceicao et al., 2013).
However, there have been case reports of individu- What is termed late dumping syndrome follows
als using chewing and spitting as a strategy to avoid food ingestion by 1 to 3 hours (Malik et al., 2016).
weight gain (Conceicao et al., 2013). It is caused by release of large amounts of insulin
in response to the high blood glucose levels, result-
Postoperative Gastrointestinal Problems ing in reactive hypoglycemia, and for clarity pur-
Postoperative GI problems may originate as an poses should be referred to as reactive hypoglycemia
adverse effect from surgery, but case reports and rather than dumping. The typical symptoms of late
series have indicated that GI problems, includ- dumping syndromes include palpitations, anxiety,
ing dumping syndrome, vomiting, and plugging, weakness, dizziness, lightheadedness, fatigue, hun-
may trigger ED symptoms in some individuals ger, and sweating. Severe reactive hypoglycemia can
(Conceicao et al., 2013; Marino et al., 2012). The trigger glucose deprivation of the central nervous
relationship between GI conditions and the devel- system. This state is potentially life threatening and
opment of ED symptoms is not fully understood. has been termed “neuroglycopenia.” Serious neu-
Future research into this link may allow for a better rological manifestations of this state may include
understanding of how EDs develop following bar- slurred speech, confusion and blurred vision, trem-
iatric surgery. ors, loss of consciousness, seizures, focal neurologi-
Dumping syndrome, mentioned previously, is cal deficits, and death.
an adverse effect that may occur following RYGB. As previously mentioned, vomiting is an addi-
Dumping occurs shortly after a patient ingests tional side effect of surgery. It is reported after
excessive quantities of food or fluids containing LAGB, RYGB, and SG (Himpens, Dapri, &
high concentrations of carbohydrate or sugar (Berg Cadiere, 2006; Kalarchian et al., 2014). The vom-
& McCallum, 2016). Dumping is characterized iting that occurs unrelated to attempts to control
by both vasomotor and gastrointestinal symptoms weight may be spontaneous or may be self-​induced
(Banerjee et al., 2013). Scoring systems have been in response to discomfort after eating. The EBs
used for diagnosis in the literature, most commonly that may lead to vomiting including eating intol-
Sigstad’s Diagnostic Index (Berg & McCallum, erable foods, eating an excessive amount, improper

464 Problematic Eating Behaviors After Bariatric Surgery

chewing, eating rapidly, and plugging. Plugging, A brief discussion of the neurobiological sys-
the feeling of food becoming stuck in the pouch, tems that drive feeding behaviors will allow a better
has been associated with vomiting but has not understanding of how EBs could be compared to
been associated with changes in weight loss out- addictive behaviors. Feeding behavior is regulated
comes (Conceicao, Mitchell, Vaz, et al., 2014). In by a complex system of central neural circuits that
shorter-​term (6  months or less) and longer-​term interact with signaling from peripheral structures,
(up to 10 years) follow-​up studies, prevalence rates such as the GI tract and adipose tissue, to control
of vomiting after surgery range from 21% to 79% food intake (Volkow et al., 2013). Feeding behav-
(Conceicao, Mitchell, Vaz, et  al., 2014; Powers, ior is further characterized by the intertwined drives
Perez, Boyd, & Rosemurgy, 1999). As mentioned of homeostatic and hedonic needs, which motivate
earlier, vomiting may be a trigger for the develop- individuals to engage in feeding behaviors.
ment or redevelopment of eating disorder symp- The most widely studied and understood path-
toms after surgery (Conceicao et al., 2013; Marino way involved in feeding behavior is the dopamine
et al., 2012). In these individuals, vomiting may be reward pathway (Volkow, Wang, Fowler, Tomasi,
used to compensate for food intake and for pur- & Baler, 2012). Brain dopamine (DA) reward cir-
poses of weight control. cuitry increases the likelihood of repeating behav-
iors that activate it when an individual is exposed
Food Addiction to the same reinforcer (Volkow et al., 201). In this
The term “food addiction” first appeared in the case, the pathway is reinforced by exposure to food
literature nearly 60  years ago (Randolph, 1956). or the surrounding environment of the food. The
There has been a reemergence of the use of this DA reward circuitry is connected to areas of the
term and the study of this phenomenon, perhaps brain involved in self-​control, interoception, emo-
in response to the increasing prevalence of obe- tions, learning memory, habits, and routines. Thus,
sity worldwide. Some controversy has surrounded DA neurons are active in maintaining behaviors for
this terminology and whether patients can, in fact, survival.
develop an “addiction,” as classically defined, to After frequent exposure, a reward becomes con-
food (Volkow, Wang, Tomasi, & Baler, 2013). It ditioned (Volkow et al., 2013). Conditioning allows
has been hypothesized that food addiction may be neural stimuli linked to a reinforcer (artificial or
one of the many factors contributing to obesity and natural) to increase DA in anticipation of reward.
may resemble or be a variant of addictive disorders This creates a strong motivation to perform and
such as substance use disorders (Gearhardt, Corbin, continue behaviors needed to obtain the reward.
& Brownell, 2016). This hypothesis evolved from For example, after conditioning has occurred, DA
the observation that the EBs of some individu- signaling occurs predicting the reward when a palat-
als are similar to the behaviors observed in those able food is viewed. This increases the motivation to
with substance use disorder (Gearhardt, Corbin, & perform behaviors that will result in consuming the
Brownell, 2009a; Volkow & O’Brien, 2007). food. In comparison to cues from food, drug cues
Comparisons have been drawn between diag- are more powerful triggers of reinforcing behavior
nostic criteria for BED and substance use disorder, (Hebebrand et al., 2014; Volkow et al., 2013). The
specifically the loss of control seen in both disor- responses to food cues also differ because they are
ders (Gearhardt, White, & Potenza, 2011). A sub- influenced by nutritional and physical status as
set of individuals with obesity, with and without well as the neuronal pathways that process stress
BED, report significant levels of distress related to and mood.
their EBs, cravings for food, consumption of food Although food intake and drug use alter the
despite adverse consequences, difficulties control- same DA pathways, they alter them in different
ling EB, giving up important activities due to food ways (Volkow et  al., 2013). Drug use has direct
consumption, and/​or eating in larger amounts over effects on the DA pathways, specifically in the
longer time periods than intended (Gearhardt, nucleus accumbens (NAc) and ventral pallidum. In
Corbin, & Brownell, 2009b). These symptoms par- contrast, the effect of food occurs both directly and
allel diagnostic criteria for substance use disorders indirectly through central and peripheral signaling
and led to the development of research regarding on the DA pathway with crucial involvement of the
food addiction, including assessment methods, such hypothalamus. Although the mechanisms are differ-
as the Yale Food Addiction Scale (Gearhardt et al., ent, both drugs of abuse and eating behaviors serve
2009b). to increase DA binding.

Orcut t, Steffen, Mitchell 465

 The indirect pathways that are involved in the addiction” emphasizes a behavioral component
regulation of feeding behavior have contributions instead of a specific substance.
from peripheral signaling mechanisms (Volkow Further differences between addiction to drugs
et  al., 2013). These mechanisms involve hypotha- and EBs have been shown with brain imaging stud-
lamic signaling in response to peptides and hor- ies (Hebebrand et  al., 2014). The imaging stud-
mones. Orexigenic peptides and hormones, which ies suggest that reward circuits are involved in the
include ghrelin, orexin, melanocortin, and neuro- motivation for food, yet imaging studies have not
peptide Y, increase DA neuron activity to stimu- explained how food intake becomes excessive or
late food intake. In contrast, anorexigenic peptides compulsive. Likewise, addiction to drugs is char-
and hormones (e.g., leptin, insulin, glucagon-​like acterized by physical dependence on the drug and
peptide-​1 [GLP-​1], cholecystokinin [CCK], pep- withdrawal symptoms when the drug is abruptly
tide YY, galanin) inhibit DA neuron activity and discontinued. A  withdrawal syndrome upon dis-
DA release, inhibiting food intake. In some areas continuing certain foods has not been observed
of the brain, neurons can express ghrelin, GLP-​1, in humans although there are reports of this in
insulin, leptin, melanocortin receptors, and orexin, animals.
further adding to the complexity of the regula- In summary, feeding behavior is regulated by a
tion of feeding. Many other neurotransmitters complex system of circuits in the central nervous
have also been implicated in the regulation of food system, which are influenced by peripheral signal-
intake including acetylcholine, cannabinoids, DA, ing. Research indicates similarities between feed-
gamma-​ aminobutyric acid (GABA), glutamine, ing behavior and drug addiction, but the research
histamine, norepinephrine, opioids, and serotonin. has also shown some key differences. Continued
Multiple signals and signaling pathways, although research is needed, especially in humans, to identify
redundant, ensure that the messages of hunger or and understand the unique factors that affect com-
satiety are communicated. Despite robust signaling pulsive eating behavior, especially the influence of
of homeostatic needs, some individuals with access the social context and homeostatic needs.
to highly palatable food may override inhibitory
signaling and eat large amounts of food, and this Conclusion
behavior can become recurrent. The purpose of this chapter was to summarize
Research in animal models indicates highly the current body of literature regarding EDs, prob-
palatable foods, foods high in sugar and fat, are lematic EBs and potentially problematic WCPs
especially rewarding and may lead to compulsive practices in individuals following bariatric surgery.
behaviors (Hebebrand et  al., 2014). Yet, the same Unfortunately, existing diagnostic and treatment
link between a specific food or type of food has options are limited by several factors including the
not been established in humans. Furthermore, the lack of agreement regarding assessment methods
diet of humans who overeat is often varied and not and clear, accepted definitions. While it has been
restricted to one type of food or a specific nutri- proposed that this population will likely require
ent. Indeed, evidence points to increased overeat- specialized treatment, evidence-​ based treatment
ing behaviors with exposure to a range of palatable guidelines and protocols have not been established.
foods. A number of factors play a role in the con- Furthermore, research has not consistently identi-
sumption of food in humans including energy fied presurgery factors that may predict or be associ-
needs, food availability, social expectations, visual ated with the development of EDs after surgery.
appeal, economics, incentives, palatability, alter- Future research must focus on establishing con-
native reinforcement, advertisements, and eating sensus regarding the terminology and diagnostic
patterns of restriction and overeating. Although criteria used to describe eating pathology in this
the effect of food and eating behavior in humans is population. Furthermore, agreement is needed
less straightforward than in animals, imaging stud- on assessment methods both prior to and follow-
ies have supported involvement of the DA reward ing bariatric surgery. Such developments will allow
pathway in motivating feeding. Since there has the elaboration and testing of treatment guidelines.
not been a specific food substance linked to com- Finally, continued investigation is needed to clar-
pulsive eating behavior, it has been suggested that ify how preoperative EBs and EDs may influence
the phenomenon may be more accurately described postoperative outcomes including EBs, WCPs, and
as an eating addiction. Indeed, the term “eating other eating pathology.

466 Problematic Eating Behaviors After Bariatric Surgery

References with a Portuguese sample. Eating and Weight Disorders, 19,
Adami, G. F., Meneghelli, A., & Scopinaro, N. (1999). Night eat- 103–​109.
ing and binge eating disorder in obese patients. International Conceicao, E., Mitchell, J. E., Engel, S. G., Machado, P. P.,
Journal of Eating Disorders, 25, 335–​338. Lancaster, K., & Wonderlich, S. A. (2014). What is “graz-
Allison, K. C., Wadden, T. A., Sarwer, D. B., Fabricatore, A. ing”? Reviewing its definition, frequency, clinical charac-
N., Crerand, C. E., Gibbons, L. M.,  . . .  Williams, N. N. teristics, and impact on bariatric surgery outcomes, and
(2006). Night eating syndrome and binge eating disorder proposing a standardized definition. Surgery for Obesity and
among persons seeking bariatric surgery:  Prevalence and Related Diseases, 10, 973–​982.
related features. Surgery for Obesity and Related Diseases, 2, Conceicao, E., Mitchell, J. E., Vaz, A. R., Bastos, A. P., Ramalho,
153–​158. S., Silva, C.,  . . .  Machado, P. P. (2014). The presence of
American Psychiatric Association. (2013). Diagnostic and sta- maladaptive eating behaviors after bariatric surgery in a cross
tistical manual of mental disorders:  DSM-​ 5. Washington, sectional study: Importance of picking or nibbling on weight
DC: Author. regain. Eating Behaviors, 15, 558–​562.
ASMBS. (2016, July). Estimate of bariatric surgery numbers, Conceicao, E., Orcutt, M., Mitchell, J., Engel, S., Lahaise, K.,
2011–​2015. Retrieved August 10, 2016, from https://​asmbs. Jorgensen, M.,  . . .  Wonderlich, S. (2013). Eating disorders
org/​resources/​estimate-​of-​bariatric-​surgery-​numbers after bariatric surgery: A case series. International Journal of
Banerjee, A., Ding, Y., Mikami, D. J., & Needleman, B. J. Eating Disorders, 46, 274–​279.
(2013). The role of dumping syndrome in weight loss after Conceicao, E., Utzinger, L. M., & Pisetsky, E. M. (2015). Eating
gastric bypass surgery. Surgical Endoscopy, 27, 1573–​1578. disorders and problematic eating behaviours before and after
Berg, P., & McCallum, R. (2016). Dumping syn- bariatric surgery:  Characterization, assessment and asso-
drome:  A  review of the current concepts of pathophysi- ciation with treatment outcomes. European Eating Disorders
ology, diagnosis, and treatment. Digestive Diseases and Review, 23, 417–​425.
Sciences, 61, 11–​18. Courcoulas, A. P., Christian, N. J., Belle, S. H., Berk, P. D.,
Brolin, R. E., Robertson, L. B., Kenler, H. A., & Cody, R. P. Flum, D. R., Garcia, L.,  . . .  Wolfe, B. M. (2013). Weight
(1994). Weight loss and dietary intake after vertical banded change and health outcomes at 3  years after bariatric sur-
gastroplasty and Roux-​en-​Y gastric bypass. Annals of Surgery, gery among individuals with severe obesity. JAMA, 310,
220, 782–​790. 2416–​2425.
Busetto, L., Segato, G., De Luca, M., De Marchi, F., Foletto, Cryan, J. F., & Dinan, T. G. (2012). Mind-​altering microorgan-
M., Vianello, M.,   .  .  .  Enzi, G. (2005). Weight loss and isms: The impact of the gut microbiota on brain and behav-
postoperative complications in morbidly obese patients with iour. Nature Reviews Neuroscience, 13, 701–​712.
binge eating disorder treated by laparoscopic adjustable gas- de Hollanda, A., Ruiz, T., Jimenez, A., Flores, L., Lacy, A., &
tric banding. Obesity Surgery, 15, 195–​201. Vidal, J. (2015). Patterns of weight loss response following
Busetto, L., Segato, G., De Marchi, F., Foletto, M., De Luca, gastric bypass and sleeve gastrectomy. Obesity Surgery, 25,
M., Caniato, D.,  . . .  Enzi, G. (2002). Outcome predictors 1177–​1183.
in morbidly obese recipients of an adjustable gastric band. de Zwaan, M., Hilbert, A., Swan-​Kremeier, L., Simonich, H.,
Obesity Surgery, 12, 83–​92. Lancaster, K., Howell, L. M.,  . . .  Mitchell, J. E. (2010).
Canetti, L., Berry, E. M., & Elizur, Y. (2009). Psychosocial Comprehensive interview assessment of eating behavior 18–​
predictors of weight loss and psychological adjustment fol- 35  months after gastric bypass surgery for morbid obesity.
lowing bariatric surgery and a weight-​ loss program:  The Surgery for Obesity and Related Diseases, 6, 79–​85.
mediating role of emotional eating. International Journal of de Zwaan, M., Marschollek, M., & Allison, K. C. (2015). The
Eating Disorders, 42, 109–​117. night eating syndrome (NES) in bariatric surgery patients.
Cawley, J., Sweeney, M. J., Kurian, M., & Beane, S.; New York European Eating Disorders Review, 23, 426–​434.
State Bariatric Surgery Workgroup. (2007). Predicting com- Devlin, M. J., King, W. C., Kalarchian, M. A., White, G. E.,
plications after bariatric surgery using obesity-​related co-​ Marcus, M. D., Garcia, L.,   .  .  .  Mitchell, J. E. (2016).
morbidities. Obesity Surgery, 17, 1451–​1456. Eating pathology and experience and weight loss in a
Colles, S. L., & Dixon, J. B. (2006). Night eating syn- prospective study of bariatric surgery patients:  3-​ year
drome:  Impact on bariatric surgery. Obesity Surgery, 16, follow-​up. International Journal of Eating Disorders, 49,
811–​820. 1058–​1067.
Colles, S. L., Dixon, J. B., & O’Brien, P. E. (2008a). Grazing Ellison, J. M., Steffen, K. J., & Sarwer, D. B. (2015). Body con-
and loss of control related to eating:  Two high-​risk factors touring after bariatric surgery. European Eating Disorders
following bariatric surgery. Obesity (Silver Spring, Md), 16, Review, 23, 479–​487.
615–​622. Fairburn, C. G., & Cooper, Z. (2000). Eating disorder exami-
Colles, S. L., Dixon, J. B., & O’Brien, P. E. (2008b). Loss of nation (14th ed., version 14.3). Oxford, UK: University of
control is central to psychological disturbance associated Oxford.
with binge eating disorder. Obesity (Silver Spring, Md), 16, Fischer, S., Chen, E., Katterman, S., Roerhig, M., Bochierri-​
608–​614. Ricciardi, L., Munoz, D.,   .  .  .  le Grange, D. (2007).
Colquitt, J. L., Pickett, K., Loveman, E., & Frampton, G. K. Emotional eating in a morbidly obese bariatric surgery-​
(2014). Surgery for weight loss in adults. Cochrane Database seeking population. Obesity Surgery, 17, 778–​784.
of Systematic Reviews, 8, CD003641. Fouladi, F., Mitchell, J. E., Wonderlich, J. A., & Steffen, K. J.
Conceicao, E., Bastos, A. P., Brandao, I., Vaz, A. R., Ramalho, (2016). The contributing role of bile acids to metabolic
S., Arrojado, F.,  . . .  Machado, P. P. (2014). Loss of control improvements after obesity and metabolic surgery. Obesity
eating and weight outcomes after bariatric surgery: A study Surgery, 26, 2492–​2502.

Orcut t, Steffen, Mitchell 467

Gade, H., Rosenvinge, J. H., Hjelmesaeth, J., & Friborg, O. Malik, S., Mitchell, J. E., Steffen, K., Engel, S., Wiisanen, R.,
(2014). Psychological correlates to dysfunctional eating pat- Garcia, L., & Malik, S. A. (2016). Recognition and manage-
terns among morbidly obese patients accepted for bariatric ment of hyperinsulinemic hypoglycemia after bariatric sur-
surgery. Obesity Facts, 7, 111–​119. gery. Obesity Research & Clinical Practice, 10, 1–​14.
Gearhardt, A. N., Corbin, W. R., & Brownell, K. D. (2009a). Mallory, G. N., Macgregor, A. M., & Rand, C. S. (1996). The
Food addiction: An examination of the diagnostic criteria for influence of dumping on weight loss after gastric restrictive
dependence. Journal of Addiction Medicine, 3, 1–​7. surgery for morbid obesity. Obesity Surgery, 6, 474–​478.
Gearhardt, A. N., Corbin, W. R., & Brownell, K. D. (2009b). Marino, J. M., Ertelt, T. W., Lancaster, K., Steffen, K., Peterson,
Preliminary validation of the Yale Food Addiction Scale. L., de Zwaan, M., & Mitchell, J. E. (2012). The emergence
Appetite, 52, 430–​436. of eating pathology after bariatric surgery:  A  rare outcome
Gearhardt, A. N., Corbin, W. R., & Brownell, K. D. (2016). with important clinical implications. International Journal of
Development of the Yale Food Addiction Scale Version 2.0. Eating Disorders, 45, 179–​184.
Psychology of Addictive Behaviors, 30, 113–​121. Mathus-​Vliegen, E. M. (2007). Long-​term health and psycho-
Gearhardt, A. N., White, M. A., & Potenza, M. N. (2011). Binge social outcomes from surgically induced weight loss: Results
eating disorder and food addiction. Current Drug Abuse obtained in patients not attending protocolled follow-​up vis-
Reviews, 4, 201–​207. its. International Journal of Obesity, 31, 299–​307.
Hebebrand, J., Albayrak, O., Adan, R., Antel, J., Dieguez, C., Meany, G., Conceicao, E., & Mitchell, J. E. (2014). Binge eat-
de Jong, J.,  . . .  Dickson, S. L. (2014). “Eating addiction,” ing, binge eating disorder and loss of control eating: Effects
rather than “food addiction,” better captures addictive-​like on weight outcomes after bariatric surgery. European Eating
eating behavior. Neuroscience & Biobehavioral Reviews, 47, Disorders Review, 22, 87–​91.
295–​306. Mitchell, J. E. (1985). Bulimia with self-​induced vomiting after
Himpens, J., Dapri, G., & Cadiere, G. B. (2006). A prospective gastric stapling. The American Journal of Psychiatry, 142, 656.
randomized study between laparoscopic gastric banding and Mitchell, J. E., Christian, N. J., Flum, D. R., Pomp, A., Pories,
laparoscopic isolated sleeve gastrectomy: Results after 1 and W. J., Wolfe, B. M.,  . . .  Belle, S. H. (2016). Postoperative
3 years. Obesity Surgery, 16, 1450–​1456. behavioral variables and weight change 3 years after bariatric
Hsu, L. K., Benotti, P. N., Dwyer, J., Roberts, S. B., Saltzman, surgery. JAMA Surgery, 151, 752–​757.
E., Shikora, S.  . . .  Rand, W. (1998). Nonsurgical factors Mitchell, J. E., King, W. C., Courcoulas, A., Dakin, G., Elder,
that influence the outcome of bariatric surgery:  A  review. K., Engel, S.,  . . .  Wolfe, B. (2015). Eating behavior and eat-
Psychosomatic Medicine, 60, 338–​346. ing disorders in adults before bariatric surgery. International
Hsu, L. K., Betancourt, S., & Sullivan, S. P. (1996). Eating Journal of Eating Disorders, 48, 215–​222.
disturbances before and after vertical banded gastro- Niego, S. H., Kofman, M. D., Weiss, J. J., & Geliebter, A. (2007).
plasty:  A  pilot study. International Journal of Eating Binge eating in the bariatric surgery population: A review of
Disorders, 19, 23–​34. the literature. International Journal of Eating Disorders, 40,
Hudson, S. M., Dixon, J. B., & O’Brien, P. E. (2002). Sweet 349–​359.
eating is not a predictor of outcome after Lap-​Band place- Noel, O. F., Still, C. D., Argyropoulos, G., Edwards, M., &
ment:  Can we finally bury the myth? Obesity Surgery, 12, Gerhard, G. S. (2016). Bile acids, FXR, and metabolic effects
789–​794. of bariatric surgery. International Journal of Obesity, 2016,
Kalarchian, M. A., King, W. C., Devlin, M. J., Marcus, M. D., 4390254.
Garcia, L., Chen, J. Y.,  . . .  Mitchell, J. E. (2016). Psychiatric Opolski, M., Chur-​ Hansen, A., & Wittert, G. (2015). The
disorders and weight change in a prospective study of bar- eating-​related behaviours, disorders and expectations of can-
iatric surgery patients:  A  3-​ year follow-​ up. Psychosomatic didates for bariatric surgery. Clinical Obesity, 5, 165–​197.
Medicine, 78, 373–​381. Opozda, M., Chur-​Hansen, A., & Wittert, G. (2016). Changes
Kalarchian, M. A., Marcus, M. D., Courcoulas, A. P., Cheng, Y., in problematic and disordered eating after gastric bypass,
& Levine, M. D. (2014). Self-​report of gastrointestinal side adjustable gastric banding and vertical sleeve gastrec-
effects after bariatric surgery. Surgery for Obesity and Related tomy:  A  systematic review of pre-​ post studies. Obesity
Diseases, 10, 1202–​1207. Reviews, 17, 770–​792.
Kofman, M. D., Lent, M. R., & Swencionis, C. (2010). Picot, J., Jones, J., Colquitt, J. L., Gospodarevskaya, E.,
Maladaptive eating patterns, quality of life, and weight out- Loveman, E., Baxter, L., & Clegg, A. J. (2009). The clinical
comes following gastric bypass: Results of an Internet survey. effectiveness and cost-​effectiveness of bariatric (weight loss)
Obesity (Silver Spring), 18, 1938–​1943. surgery for obesity: A systematic review and economic evalu-
Leite Faria, S., de Oliveira Kelly, E., Pereira Faria, O., & Kiyomi ation. Health Technology Assessment (Winchester, England)13,
Ito, M. (2009). Snack-​ eating patients experience lesser 1–​190, 215–​357, iii–​iv.
weight loss after Roux-​en-​Y gastric bypass surgery. Obesity Powers, P. S., Perez, A., Boyd, F., & Rosemurgy, A. (1999). Eating
Surgery, 19, 1293–​1296. pathology before and after bariatric surgery:  A  prospective
Lindroos, A. K., Lissner, L., & Sjostrom, L. (1996). Weight study. International Journal of Eating Disorders, 25, 293–​300.
change in relation to intake of sugar and sweet foods before Ramalho, S., Bastos, A. P., Silva, C., Vaz, A. R., Brandão, I.,
and after weight reducing gastric surgery. International Machado, P. P., & Conceição, E. (2015). Excessive skin and
Journal of Obesity and Related Metabolic Disorders, 20, sexual function:  Relationship with psychological variables
634–​643. and weight regain in women after bariatric surgery. Obesity
Makaronidis, J. M., & Batterham, R. L. (2016). Potential mech- Surgery, 25, 1149–​1154.
anisms mediating sustained weight loss following Roux-​ Randolph, T. G. (1956). The descriptive features of food addic-
en-​Y gastric bypass and sleeve gastrectomy. Endocrinology tion:  Addictive eating and drinking. Quarterly Journal of
Metabolism Clinics of North America, 45, 539–​552. Studies on Alcohol., 17, 198–​224.

468 Problematic Eating Behaviors After Bariatric Surgery

Sarwer, D. B., Wadden, T. A., Moore, R. H., Baker, A. W., psychological traits and sleep profiles of patients affected by
Gibbons, L. M., Raper, S. E., & Williams, N. N. (2008). nocturnal eating. Sleep Medicine, 16, 746–​753.
Preoperative eating behavior, postoperative dietary adher- Volkow, N. D., & O’Brien, C. P. (2007). Issues for DSM-​
ence, and weight loss after gastric bypass surgery. Surgery for V:  Should obesity be included as a brain disorder? The
Obesity and Related Diseases, 4, 640–​646. American Journal of Psychiatry, 164, 708–​710.
Sugerman, H. J., Kellum, J. M., Engle, K. M., Wolfe, L., Starkey, Volkow, N. D., Wang, G. J., Fowler, J. S., Tomasi, D., & Baler,
J. V., Birkenhauer, R.,   .  .  .  Sawyer, M. J. (1992). Gastric R. (2012). Food and drug reward:  Overlapping circuits in
bypass for treating severe obesity. The American Journal of human obesity and addiction. Current Topics in Behavioral
Clinical Nutrition, 55(2 Suppl), 560S–​566S. Neurosciences, 11, 1–​24.
Sugerman, H. J., Starkey, J. V., & Birkenhauer, R. (1987). A Volkow, N. D., Wang, G. J., Tomasi, D., & Baler, R. D. (2013).
randomized prospective trial of gastric bypass versus verti- Obesity and addiction:  Neurobiological overlaps. Obesity
cal banded gastroplasty for morbid obesity and their effects Reviews, 14, 2–​18.
on sweets versus non-​sweets eaters. Annals of Surgery, 205, White, M. A., Kalarchian, M. A., Masheb, R. M., Marcus, M.
613–​624. D., & Grilo, C. M. (2010). Loss of control over eating pre-
Swanson, S. A., Crow, S. J., Le Grange, D., Swendsen, J., & dicts outcomes in bariatric surgery patients:  A  prospective,
Merikangas, K. R. (2011). Prevalence and correlates of eating 24-​month follow-​up study. Journal of Clinical Psychiatry, 71,
disorders in adolescents:  Results from the national comor- 175–​184.
bidity survey replication adolescent supplement. Archives of Winkelman, J. W. (1998). Clinical and polysomnographic
General Psychiatry, 68, 714–​723. features of sleep-​related eating disorder. Journal of Clinical
Sweeney, T. E., & Morton, J. M. (2013). The human gut Psychiatry, 59, 14–​19.
microbiome: A review of the effect of obesity and surgically Yale, C. E. (1989). Gastric surgery for morbid obesity.
induced weight loss. JAMA Surgery, 148, 563–​569. Complications and long-​term weight control. Archives of
van Hout, G. C., Verschure, S. K., & van Heck, G. L. (2005). Surgery, 124, 941–​946.
Psychosocial predictors of success following bariatric surgery. Yale, C. E., & Weiler, S. J. (1991). Weight control after vertical
Obesity Surgery, 15, 552–​560. banded gastroplasty for morbid obesity. American Journal of
Vinai, P., Ferri, R., Anelli, M., Ferini-​Strambi, L., Zucconi, Surgery, 162, 13–​18.
M., Oldani, A., & Manconi, M. (2015). New data on

Orcut t, Steffen, Mitchell 469

 C H A PT E R


Virtual Reality: Applications to Eating
25 Disorders

José Gutiérrez-​Maldonado, Marta Ferrer-​García, Antonios Dakanalis, and Giuseppe Riva

Abstract
In the last twenty years researchers have embraced virtual reality (VR) in order to integrate and extend
the assessment tools and treatments currently in use for eating disorders (EDs). Specifically the VR
protocols for EDs try to exploit clinically the sense of “presence,” that is, the feeling of “being there”
inside the virtual environment. The sense of presence offered by VR can be a powerful tool in therapy
because it provides the individual with a world in which he/​she can be placed and live a particular
experience. This triggers emotional reactions in patients and allows a higher level of self-​reflectiveness
than that provided by memory and imagination, and greater control than that offered by direct “real”
experience. In particular, VR protocols for EDs use technology to alter the experience of the body
(embodiment) in real time and as a cue exposure tool for reducing food craving.
Key Words:  virtual reality, exposure, cue exposure, body image, bodily self-​consciousness, body
swapping, allocentric lock

Introduction What Is Virtual Reality?

The first studies of the application of virtual In movies and computer science, VR is usually
reality (VR) to eating disorders (EDs) were carried described as a set of fancy technologies (Riva, Botella,
out under the aegis of VREPAR (Virtual Reality et al., 2015): an interactive three-​dimensional (3D)
Environments for the Psycho-​neuro-​physiological visualization system (a computer, a game console, or
Assessment and Rehabilitation Project), a European a smartphone) supported by one or more position
project designed to develop virtual environments trackers and a head-​mounted display. The trackers
for the study, evaluation, and treatment of body sense the user’s movements and report them to the
image disorders (Riva, 1998a, 1997; Riva, Melis, visualization system, which updates the images for
& Bolzoni, 1997). Since then, several research- display in real time.
ers have embraced VR to integrate and extend the Looking at the issue in greater detail, a VR sys-
currently used assessment tools and treatments for tem includes the hardware and software that enable
EDs. Nevertheless, its implementation in clinical developers to create VR applications and users to
settings remains scarce. The increased availability of experience them. The hardware components receive
VR technology today in combination with findings input from user-​controlled devices (head-​mounted
from clinical studies and neuroscience seem destined display, joysticks, gloves, etc.) and convey multisen-
to boost the growth of two types of intervention sory output to create the illusion of a virtual world.
(described below) in the coming years: cue exposure The virtual world may be either a model of a real-​
therapy and the modification of the experience of world object, such as a house, or an abstract world
the body through the phenomenon of embodiment that does not exist in a real sense but is understood
(Gutierrez-​Maldonado, Wiederhold, & Riva, 2016; by humans, such as a chemical molecule or a repre-
Wiederhold, Riva, & Gutiérrez-​Maldonado, 2016). sentation of a set of data, or a completely imaginary

470
world. Typically, a VR system includes the following become immersed in a computer-​generated environ-
(Burdea & Coiffet, 2003; Gorini, Gaggioli, Vigna, ment in a naturalistic fashion” (Schultheis & Rizzo,
& Riva, 2008; Hale & Stanney, 2014): 2001, p.  82). From a psychological perspective,
VR can be described as a synthetic experience that
•  A graphic rendering system that generates
makes the user believe that he/​she is there, and that
the virtual environment, at 60 frames per second
this experience is real (Riva, 1998b). Specifically,
(this is the minimum frame rate, but it may be
what distinguishes VR from other media is the sense
higher depending on the power of the hardware;
of “presence,” i.e. the feeling of “being there” inside
see Table 25.1 for details of commercially available
the virtual experience produced by the technology
VR systems);
(Gorini, Capideville, De Leo, Mantovani, & Riva,
•  Database construction and virtual object
2011; Riva & Waterworth, 2003).
modeling software for building and maintaining
While there is still no general consensus about
detailed and realistic models of the virtual world.
what presence actually is from a psychological
Specifically, the software handles the geometry,
viewpoint (for an introduction to the subject, see
texture, intelligent behavior, and physical modeling
Baños et  al., 2004; Baños et  al., 2008; Diemer,
of the hardness, inertia, and surface plasticity of
Alpers, Peperkorn, Shiban, & Muhlberger, 2015;
any object included in the virtual world;
Lee, 2004; Ling, Nefs, Morina, Heynderickx, &
•  Input tools (trackers, gloves, joystick,
Brinkman, 2014; Lombard, Biocca, Freeman,
mice, etc.) that continually report the position
IJsselsteijn, & Schaevitz, 2015; Pillai, Schmidt, &
and movements of the users; and output tools
Richir, 2013; Riva, 2009, 2012; Riva, Davide, &
(visual, aural, haptic, etc.) that immerse the user
IJsselsteijn, 2003; Riva, Waterworth, & Murray,
in the virtual environment (see Table 25.1 for
2014; Sanchez-​Vives & Slater, 2005; Sethi, Suzuki,
characteristics of commercially available VR systems).
& Critchley, 2012; Slater, 2002; Slater & Wilbur,
On the basis of the hardware and software 1997; Waterworth & Riva, 2014; Zahoric & Jenison,
included in a VR system, it is possible to distinguish 1998) it is fair to say that most investigators agree
between: about what it is not (Riva, 2009). As underlined by
Riva and colleagues (Riva et al., 2014), “presence is
•  Desktop VR: uses subjective immersion on
not the degree of technological immersion, it is not
a standard PC screen. The feeling of immersion
the same thing as emotional engagement, it is not
can be improved through stereoscopic vision.
absorption or attention or action; but all of these
Interaction with the virtual world can be made via
have a potential role in understanding the experi-
a mouse, joystick, or typical VR peripherals such as
ence of presence in interaction—​the experience of
a data glove.
interacting with presence” (p. 1).
•  Fully Immersive VR: with this type of
The sense of presence offered by VR can be a
solution the user appears to be fully inserted in the
powerful tool for personal change because it pro-
computer-​generated environment. This illusion
vides the individual with a world in which he/​
is produced by providing immersive output
she can be placed and live a particular experience
devices (head-​mounted display, force feedback
(Baños et al., 2005; Riva, Bacchetta, Cesa, Conti, &
robotic arms, etc.) and a system of head/​body
Molinari, 2003; Riva, Botella, et al., 2015). Virtual
tracking to guarantee the exact correspondence
reality allows a higher level of self-​ reflectiveness
and coordination of users’ movements with the
than that provided by memory and imagination,
feedback of the environment.
and a higher level of control than that offered by
•  CAVE: this is a small room where a computer-​
direct “real” experience. Virtual reality has also been
generated world is projected on the walls. The
described as an advanced imaginal system: an expe-
projection is made on both front and side walls.
riential form of imagery that is as effective as reality
This solution is particularly suitable for collective
in inducing emotional responses (North, North, &
VR experiences because it allows different people
Coble, 1997; Vincelli, 1999; Vincelli, Molinari, &
to share the same experience at the same time.
Riva, 2001).

Virtual Reality and Presence Virtual Worlds, Real Emotions

In psychology and neuroscience VR is defined To quote Glantz and colleagues, “one rea-
instead as “an advanced form of human-​computer son it is so difficult to get people to update their
interface that allows the user to interact with and assumptions is that change often requires a prior

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 471

Table 25.1  Prices and Characteristics of Commercially Available Fully Immersive VR Systems (based on
2016 data)
Fully Immersive VR Systems
PC Based Mobile Based Console Based

System Oculus Rift HTC Vive Samsung Gear Google Google Playstation VR
VR Cardboard Daydream

Cost 599 US$ 799 US$ 99 US$ 10-​50 US$ 69-​149 US$ 399 US$

Hardware High End PC High End PC High End Middle-​High High End PS4 (299 US$)
Requirements (>1000 US$) (>1000 US$) Samsung End Android Android or PS4 Pro
Phone (>600 phone or Phone (>499 (399 US$)
US$) iPhone (>299 US$)
US$)

Resolution 2160 x 1200 2160 x 1200 2560 x 1440 Depends on Depends on 1920 x 1080
the phone the phone
(minimum (minimum
1024 x 768) 1920 x 1080)

Refresh Rate 90 Hz 90 Hz 60 Hz 60 Hz 90 Hz 120 Hz

minimum

Field of View 110 degrees 110 degrees 101 degrees from 70 96 degrees 100 degrees
degrees

Body Tracking Medium/​ High: head Medium: head Medium: head Medium: head Medium/​

High: head tracking tracking tracking tracking High: head
tracking (rotation) and (rotation) (rotation) (rotation) tracking
(rotation) and volumetric (rotation) and
positional tracking (full positional
tracking room size—​ tracking
(forward-​ 15 ft x 15 (forward/​
backward) ft—m​ ovement) backward)

User Interaction High (using High (using Medium Low (using Medium High (using
with VR a joystick or controllers) (using gaze, a gaze or a (using gaze or a joystick or
controllers) built-​in pad, button) joystick) controllers)
or joystick)

Software Oculus Store Steam Store Oculus Store Google Play or Google Play Playstation
Availability IOS Store Store

step—​ recognizing the distinction between an elements and make them available for reorganiza-
assumption and a perception. Until revealed to be tion (Riva, Mantovani, & Gaggioli, 2004).
fallacious, assumptions constitute the world; they Another feature of VR that has received consid-
seem like perceptions, and as long as they do, they erable attention is its ability to trigger emotional
are resistant to change” (Glantz, Durlach, Barnett, reactions in patients (Ferrer-​Garcia & Gutierrez-​
& Aviles, 1997, p. 96). Using the sense of presence Maldonado, 2010, 2012; Gutierrez-​ Maldonado,
induced by VR, it is easier to develop new, realistic, Ferrer-​Garcia, Caqueo-​Urizar, & Moreno, 2009).
credible, and informative experiences regarding the Comparing virtual stimuli with the correspond-
surrounding world or the self, demonstrating to the ing real stimuli and photographs, Gorini, Griez,
individual that what is assumed to be true in fact Petrova, and Riva (2010) found that virtual food
is a product of his/​her mind. Once this has been was as effective as real food, and more effective than
understood, it is easier to identify all the significant photographs of food, in triggering psychological

472 Virtual Realit y

and physiological responses in patients with EDs, overcome through confrontation and effort (Riva &
regardless of specific ED diagnosis. These studies Mantovani, 2012a).
thus demonstrate the validity of VR for use as an
exposure technique in place of real stimuli in treat- From Virtual to Real Bodies
ing patients. Virtual reality can also be defined as an “embod-
As noted by Riva and Mantovani (2012a) the ied technology” because of its ability to modify the
rationale behind the use of VR in exposure tech- feeling of presence (Riva, Dakanalis, & Mantovani,
niques is simple: “In VR the patient is intentionally 2015; Riva & Mantovani, 2012b, 2014). In VR,
confronted with the feared stimuli while allowing subjects can experience their synthetic avatars as
the anxiety to attenuate. Avoiding a dreaded situa- if they were their own body; this phenomenon is
tion reinforces a phobia, and each successive expo- known as “embodiment” (referring to the replace-
sure to it reduces the anxiety” (p.  21). In other ment of the physical body by the virtual one). In
words, VR is a versatile tool that permits the devel- other words, the VR user is present in a virtual body
opment of multiple environments that can be pre- through the alteration of the cognitive factors that
sented to the user in many different forms (Gorini & regulate our experience of body and space (for an
Riva, 2008; Ling et al., 2014). Recent studies show in-​depth analysis of this claim see Riva, Dakanalis,
that VR exposure to multiple contexts reduces the et  al., 2015). Furthermore, we use the “feelings”
recurrence of fear to a greater extent than exposure from the body to sense both our physical condi-
to only one scenario (Shiban, Pauli, & Muhlberger, tion and emotional state. These feelings range from
2013); in much the same way, the implementa- bodily changes that may also be visible to an exter-
tion of multiple stimuli contexts during exposure nal observer (i.e., posture, touch, facial expressions)
significantly reduced return of fear post treatment to proprioceptive and interoceptive changes that
(Dunsmoor, Ahs, Zielinski, & LaBar, 2014; Shiban, are invisible to an external observer (i.e., endocrine
Schelhorn, Pauli, & Muhlberger, 2015). Therefore, release, heart rate, muscle contractions) (Bechara &
exposure to different virtual contexts can be an Damasio, 2005). Finally, bodily representations are
effective way of generalizing the results. Further, as usually produced and modulated by sensory inputs,
suggested by Diemer et al. (2015), VR can be used but they can exist and produce qualitatively rich
to induce emotional reactions via different routes bodily experiences even in the absence of any input
(perceptual vs. conceptual), with additive effects if signal (e.g., the phantom limb syndrome) (Melzack,
combined. 2005). In this view, the experience of our bodily
In summary, the capacity to develop a large self can be considered the result of a multimodal
body of realistic controlled stimuli and, simultane- simulation. To quote Margaret Wilson (2006), “The
ously, to monitor the responses generated by the human perceptual system incorporates an emula-
user offers a considerable advantage over real expe- tor  . . .  that is isomorphic to the human body . . . .
riences (Riva & Wiederhold, 2015). For example, The emulator draws on body-​schematic knowledge
if an individual experiences a significant fear when derived from the observer’s representation of his
exposed to heights, by using a virtual elevator simu- own body” (p. 221).
lation we can assure him/​her that this threat is not Starting from these premises, in 2007 two
going to occur until he/​she feels prepared to cope European teams working independently reported in
with it. The same can be said for all the elements Science how VR could be used to produce an out-​
that are present in the situation, which can make it of-​body experience in healthy volunteers (Ehrsson,
more or less threatening (Ferrer-​Garcia et al., 2015; 2007; Lenggenhager, Tadi, Metzinger, & Blanke,
Pla-​Sanjuanelo et al., 2015). 2007). Since then, immersive VR environments
Furthermore, VR allows the construction of have developed rapidly, in which the possibility of
“virtual adventures” in which subjects experience “projecting” (and controlling) the body into exter-
themselves as competent and efficacious (Botella nal (or virtual) space is becoming a reality and is
et al., 2004; Riva, Botella, et al., 2015; Riva et al., opening up a highly promising new line of research,
2007). Specifically, it is possible to design targeted virtual embodiment (Bergouignan, Nyberg, &
VR experiences with different difficulty levels—​ Ehrsson, 2014; Maselli & Slater, 2014; Olive &
ranging from easy to very difficult—​ which pro- Berthoz, 2012; Pomes & Slater, 2013), whose
vide an important source of personal efficacy. results have been discussed in two recent reviews
By interacting with them, individuals discover (Costantini, 2014; Gallace & Spence, 2014). Using
that the conflicts and/​or feared situations can be this approach, various authors have used VR to

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 473

induce an illusory perception of a fake limb (Slater, but there is still no strong evidence for the efficacy
Perez-​Marcos, Ehrsson, & Sanchez-​Vives, 2009) or of VR in the treatment of this disorder and other
a fake hand (Perez-​Marcos, Slater, & Sanchez-​Vives, psychotic disorders. Similarly, there is only evidence
2009) as part of our own body, and to produce an for a moderate effect of the VR interventions on
out-​of-​body experience (Lenggenhager et al., 2007) depression. As a pain reduction technique, VR has
by altering the normal association between touch shown strong efficacy in short-​term interventions,
and its visual correlate. The application of these but little evidence exists for longer-​term benefits. In
procedures for the treatment of body image distur- EDs, the reviews performed to date show that VR
bances in EDs is discussed later in this chapter. cue exposure to food stimuli and VR body image
treatments are effective.
Effectiveness and Limitations Despite its undeniable potential, the use of VR
The first studies on the effectiveness of VR in the in research and clinical settings is still limited. There
treatment of psychological disorders concentrated are three main reasons for this:  its economic cost,
on different types of phobias because VR lends itself the technological difficulties, and the potential side
so readily to exposure therapies, which are the most effects.
effective interventions in the treatment of those The implementation of VR-​ based applica-
disorders:  fear of flying (North & North, 1994), tions for clinical use has always depended heav-
fear of heights (Rothbaum et al., 1995), agorapho- ily on the development of advanced technology.
bia (North, North, & Coble, 1995), fear of public Consequently, for a long time the research in
speaking (North, North, & Coble, 1998), claustro- this area was limited by the cost of the technol-
phobia (Booth & Rachman, 1992), fear of driving ogy required. Furthermore, the field was largely
(Schare, Scardapane, Berger, Rose, & Berger (1999), restricted to academic research, and very few tech-
post-​traumatic stress disorder (Hodges et al., 1998), nology companies sought to transfer the results of
and obsessive-​compulsive disorder (North & North, this research into clinical VR applications.
2000). During the first decade of research, VR Today, however, VR technology is advancing
applications were also developed for other disorders quickly. Both Oculus Rift (http://​www.oculus.
such as autism (Strickland, 1996), attention-​deficit com) and HTC (https://​www.htcvive.com/​) are
disorder (Rizzo et al., 2000), and EDs (Riva, 1997). showcasing high-​ quality VR experiences at rea-
In a recent publication, Riva, Baños, Botella, sonable price points—​less than $2,000 for a fully
Mantovani, and Gagglioli (2016) reported the avail- configured system—​which are now widely availa-
able reviews and meta-​analyses about the use of VR ble to consumers (Castelvecchi, 2016). Thus, the
in clinical and health psychology. They were related first major obstacle to the widespread use of the VR
to addictions (2 reviews, 1 meta-​analysis; 53 stud- seems to have been overcome. The second one, the
ies), pain (4 reviews; 48 studies), anxiety disorders (3 presence of technological difficulties, remains, but
reviews, 4 meta-​analyses; 175 studies), stress-​related probably not for very long. The use of VR systems
disorders (4 reviews; 41 studies), depression (1 review involves the management of complex devices that
and meta-​analysis; 19 studies), EDs (3 reviews; 33 require a certain level of technological knowledge
studies), schizophrenia and other psychotic disor- and the assistance of technical staff. Therefore, it
ders (2 reviews, 1 meta-​analysis; 23 studies), and is not surprising that some therapists and clini-
autism (2 reviews; 39 studies). The highest number cians, especially veteran practitioners, are reluctant
of studies has been conducted in anxiety disorders to introduce VR systems into their daily practice.
and stress-​related disorders, supporting the efficacy However, this scenario is about to change largely
of VR in the treatment of phobias, stress manage- due to the expansion of VR in the field of consumer
ment, post-​traumatic stress disorder, panic disorder, electronics; the commercialization of VR systems
and agoraphobia. The evidence for the treatment of among the general population will bring down
social phobia is not definitive. The reviews related costs and enhance the development of user-​friendly
to addictions show that VR is effective in inducing devices. Furthermore, for younger generations the
craving to substances such as cocaine, alcohol, and use of VR technology will be part of their everyday
tobacco, allowing its use in cue exposure treatments routine and the technical difficulties will disappear.
and to develop coping skills. In autism, the reviews Finally, the potential side effects of exposure to
support the use of VR to train social skills. This kind VR environments that may impair the potential
of training has also been used in patients with schiz- benefits of using this technology must be borne
ophrenia, and preliminary results are promising, in mind (Rizzo, Schultheis, & Rothbaum, 2002).

474 Virtual Realit y

Virtual reality sickness or cybersickness results from Hay & Claudino, 2010). Unfortunately, studies also
a conflict between the visually perceived movement reveal that BN and BED are chronic and treatment
in the virtual world and the vestibular system’s sense resistant to standard evidence-​based treatment (i.e.,
of movement (standing still) and may produce neg- CBT) in more than 20% of patients (Steinhausen
ative effects like dizziness, discomfort, disorienta- & Weber, 2009).
tion, and fatigue (Regan & Price, 1993). Although Conditioning and learning processes seem to play
research into cybersickness in the clinical popula- an important role in the maintenance of disturbed
tion is scarce, a review published in 2014 reported behaviors in both BN and BED (Bouton, 2011).
that only a small percentage of patients who engaged Therefore, intervention techniques based on these
in VR-​based treatments experienced negative side theoretical models have been developed to treat
effects (Quintana, Bouchard, Serrano, & Cárdenas-​ EDs characterized by binge eating. The first pro-
López, 2014). These side effects usually appear in posal appeared at the end of 1980s when Schmidt
the short term and become less intense as immer- and Marks (1988) reported that food craving expe-
sion to VR environments is repeated over time rienced by BN patients decreased over a series of
(Kennedy, Stanney, & Dunlap, 2000). Furthermore, sessions in which cue exposure therapy (CET) was
there are individual differences that may increase or applied. Over the following years, Jansen and col-
reduce the probability of experiencing these effects; leagues established the theoretical bases of CET for
indeed, some researchers have stressed the need to EDs characterized by binge eating, and published
assess the predisposition to immersion and sensitiv- several studies providing support for the efficacy of
ity to cybersickness of patients in order to achieve the intervention (Jansen, Broekmate, & Heymans,
the most effective VR experience (Rosa, Morais, 1992; Jansen, van den Hout, de Loof, Zandbergen,
Gamito, Oliveira, & Saraiva, 2016). & Griez, 1989).
According to Jansen (1994), once binge eat-
Virtual Reality Applications in Eating ing behavior has been established, exposure to
Disorders binge-​related food cues provokes a conditioned
The main applications of VR technologies in response (hyperinsulinemia), which thus elicits a
ED have focused in cue exposure therapy and body hypoglycemic compensatory response. This bio-
image study, assessment and treatment. Specifically, chemical response is experienced as food craving
the clinical use of VR with these disturbances is and may lead to a binge episode in at-​risk indi-
based on key theory-driven psychological treatment viduals. Jansen (1998) also drew attention to the
techniques. First, VR can reduce eating-related similarities between food craving in BED patients
anxiety during and after exposure to virtual food, and drug addiction, suggesting that the two phe-
helping to disrupt the reconsolidation of adriven- nomena may share neurochemical mechanisms and
verse, food-related memories. Second, VR is used neuroanatomical bases. Subsequent research seems
to counter a multisensory body integration deficit to support this thesis (Cason et  al., 2010; Nair,
that impairs the patient’s ability of updating his/her Adams-​Deutsch, Epstein, & Shaham, 2009).
body memory (allocentric, offline) with new con- Nair et  al. (2009) reported that the activation
tents from real-time perception-d inputs (egocen- of group II metabotropic receptors (mGluR2 and
tric, online). mGluR3), D1 dopamine receptors, CB1 receptors,
and mu opioid receptors contribute to the reinstate-
Virtual Reality Cue Exposure ment of heroin, cocaine, alcohol, and food seek-
Cognitive-​ behavioral therapy (CBT; Fairburn, ing induced by several types of cues. There is also
Marcus, & Wilson, 1993) is considered the gold evidence for a role of mGluR1 and 5-​HT1A and/​
standard for the treatment of bulimia nervosa or 5-​HT1B receptors in the discrete cue-​induced
(BN) (National Institute for Health and Clinical reinstatement of both food and cocaine seeking.
Excellence, 2004) and has also been proposed as Additionally, it has been shown that activity in
an appropriate intervention for binge eating disor- the accumbens core mediates discrete cue-​induced
der (BED) (Berkman et  al., 2015; Fairburn et  al., cocaine, heroin, and food seeking and that activ-
2015). A  considerable body of research supports ity in the lateral hypothalamus mediates context-​
its efficacy, reporting that between 30% and 50% induced reinstatement of both alcohol and food
of patients that finish treatment present substan- seeking. Furthermore, animal research has provided
tial reductions in binging and purging behaviors evidence of the existence of individual differences
(Agras, Walsh, Fairburn, Wilson & Kraemer, 2000; in reward-​seeking behavior (Mahler & Wit, 2010).

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 475

There are animals, and also humans, with highly drawbacks, such as the lack of control over what
reactive phenotypes to reward cues and, conse- participants are imagining and the difficulties some
quently, with an enhanced risk of developing addic- individuals have in visualizing the target situation in
tive disorders or binge behavior and an increased a sufficiently realistic way. The use of VR provides a
risk of relapse after treatment. third way to conduct CET. Virtual reality environ-
Other researchers have focused on the role of ments allow therapists to expose patients to lifelike
anxiety and negative affect as precipitating cues of situations while maintaining strict control over the
binge behavior. Toro et  al. (2003) and Martínez-​ variables. Then, as ecological validity is high, the
Mallén et  al., 2007) suggested that exposure to possibility for the generalization of the acquired
binge-​related cues produced anticipatory anxiety, learning is also enhanced.
which in turn may elicit an overeating episode. This The use of VR for CET has been mainly applied
research group used CET to treat BN patients resis- for the treatment of addictions. Several research
tant to treatment (CBT and pharmacotherapy), with groups have developed VR-​ based interventions
positive results (Toro et al., 2003; Martínez-​Mallén for addiction to substances such as drugs, nico-
et al., 2007). After the intervention, patients showed tine, and alcohol (Hone-​ Blanchet, Wensing, &
reductions in anxiety and bulimic symptoms. The Fecteau, 2014; Pericot-​Valverde, García-​Rodríguez,
binge episodes declined drastically in number and, Gutiérrez-​Maldonado, & Secades-​ Villa, 2015).
in most cases, were completely eliminated. Purge These applications consist in VR environments that
episodes were also reduced at follow-​up. simulate situations, people, and objects that have
Despite good initial results, CET presents cer- been identified as triggers of craving. The aim of
tain logistical drawbacks (Bulik et  al., 1998) that the intervention is to reduce the craving response,
should also be considered. When CET is conducted and then to extend this reduction or elimination to
in the therapist’s office, patients have to bring a suf- similar situations in the real world.
ficient amount of binge-​related food to the expo- In order to develop CET programs in patients
sure session, which may well be an inconvenience. with EDs, several studies assessing the ability of
Moreover, in the office only specific cues (food) are food-​ related VR environments to provoke food
taken into account—​not contextual cues (e.g., the craving have recently been published (Ferrer-​Garcia
kitchen environment). Thus, the ecological validity et al., 2014; Pla-​Sanjuanelo et al., 2015). Overall,
of the intervention is reduced and, consequently, the these studies found that craving experienced in
possibilities of generalizing the extinction response VR environments incorporating cues and contexts
to every-​ day life situations may be impaired. related to binging behavior was consistent with trait
A  recent study by Schyns, Roefs, Mulkens, and and state craving assessed (with questionnaires) out-
Jansen (2016) highlighted the importance of expo- side the VR environments. Participants with the
sure to numerous and varied cues for the general- highest scores on trait and state craving also showed
ization of learning in CET. The main aim of that a greater urge to eat when exposed to food in differ-
study was to assess whether a one-​session cue expo- ent VR environments. In addition, scores on ques-
sure intervention decreased eating in the absence tionnaires assessing trait and state craving were able
of hunger, and whether the hypothesized decrease to predict the average craving experienced in virtual
in intake would be generalized to other food items environments. The ability of VR exposure to reduce
that were not present during exposure. The results food craving has also been assessed. Gutierrez-​
showed that inhibitory learning takes place dur- Maldonado, Pla-​ Sanjuanelo, and Ferrer-​ Garcia
ing the food cue exposure session, as participants (2016) applied a one-​session model of CET to 113
successfully inhibited themselves when confronted participants who were exposed to food-​related vir-
with an exposed food item, but the inhibitory learn- tual environments using two different VR systems
ing did not generalize to food items that were not during exposure (one immersive and one nonim-
present during the cue exposure. These results show mersive). Food craving decreased during exposure
that there is room for improvement with regard to to both immersive and nonimmersive VR envi-
the generalizability of CET. ronments compared with pre-​exposure levels, sup-
Imagery exposure has frequently been used as an porting the efficacy of VR-​CET in reducing food
alternative to in vivo exposure, given its flexibility craving. No significant differences in craving were
and the possibility of extending the exposure to dif- found between immersive and nonimmersive sys-
ferent cues and contexts through the imagination. tems, showing that low-​cost nonimmersive systems
However, this method also has some important can improve the accessibility of this technique by

476 Virtual Realit y

reducing costs and improving usability, without any problem that clinicians must deal with. Exposure to
significant loss of efficacy. food with prevention of avoidance response has been
A clinical trial underway at several cen- proposed as an appropriate intervention to treat
ters in Spain and Italy is currently investigating fear and refusal to eat in AN patients (Hildebrandt,
these issues (https://​clinicaltrials.gov/​ct2/​show/​ Bacow, Markella & Loeb, 2012; Steinglass et  al.,
NCT02237300). The main aim of the study is to 2012). The results suggest that exposure therapy
assess the efficacy of CET based on VR (VR-​CET) reduces food anxiety and increases patients’ calorie
as a second level treatment in patients with BN and intake. Likewise, exposure therapy involving one’s
BED. With this objective in mind, 64 patients diag- own body has also been proposed as an effective
nosed with BN or BED, according to DSM-​5 (APA, technique for body image disturbance treatment
2013), who were treatment resistant (that is, their (Hildebrandt, Loeb, Troupe, & Delinsky, 2012;
binges persisted after CBT) were randomly assigned Morgan, Lazarova, Schelhase, & Saeidi, 2014).
to one of two booster session conditions: a VR-​CET This technique is usually applied as a component
booster sessions group, and a CBT booster sessions of CBT and is conducted using mirror confron-
group (the control group). Booster sessions consisted tation or visualization of a video recording of the
of six 60-​minute sessions held twice weekly over a patient’s body. Using this procedure, studies found
period of 3 weeks. Over the six sessions, participants a reduction in body-​related negative emotions and
in the experimental group were exposed to different cognitions and a decrease in restrictive and bulimic
VR environments related to binge behavior, accord- behaviors (Morgan et al., 2014; Trentowska, Bender,
ing to a previously constructed hierarchy. During & Tuschen-​Caffier, 2013). Repeated and prolonged
exposure, patients faced high-​ risk situations and exposure to one’s own body during the intervention
handled the virtual foods using a computer mouse. is believed to break down the association between
Exposure ended after a significant reduction in the the image of the body and the conditioned negative
level of anxiety, or after 60 minutes. Participants in responses to it.
the control group received six CBT booster sessions As in the case of the use of CET for BN and
to improve treatment outcome. A significant inter- BED, in vivo exposure for AN treatment presents
action between group (VR-​CET vs. CBT) and time several drawbacks that make its implementation
(before and after booster sessions) was expected, difficult. Again, logistical difficulties, time required
showing the maintenance of the number of binges to conduct interventions, limited ecological valid-
and purges before and after booster sessions in the ity, and generalization problems hinder the appli-
control group (CBT) and a reduction in the experi- cation of exposure techniques. In addition, AN
mental group (VR-​CET). patients frequently resist or reject the use of those
After the six booster sessions, patients in both techniques, as they may trigger anxiety. These dif-
CBT and VR-​CET conditions presented improve- ficulties have led to the search for new procedures
ment. However, participants in the VR-​CET group to implement exposure. Here again, VR technology
showed significantly higher reductions in binges, has been proposed as an alternative procedure that
purges, bulimia symptoms (assessed with the overcomes the drawbacks mentioned.
Bulimia scale of the Eating Disorders Inventory-​3; Currently, Gutiérrez-​Maldonado’s group at the
EDI-​3), craving for food (assessed with the Food University of Barcelona is working on a research
Craving Questionnaire-​State/​Trait; FCQ-​S/​T), project whose main aim is the development of a
and anxiety (State and Trait Anxiety Inventory, VR-​ based application for reducing AN patients’
STAI) than patients in the CBT group. The results fear of regaining a healthy weight and the resulting
obtained support the efficacy of VR-​ CET for refusal to eat. Patients are exposed to an avatar that
enhancing CBT in ED patients resistant to treat- represents their own body image and whose body
ment as usual, and show that further research in this mass index (BMI) increases progressively according
field is now necessary. to a pre-​established hierarchy. Moreover, to facili-
Exposure therapy has also been used for the tate the generalization of changes outside the thera-
treatment of anorexia nervosa (AN). However, pist’s office, exposure to the avatar representing the
published studies are still scarce. Patients with AN patient’s body image takes place in different virtual
show anticipatory anxiety and frequent worries environments that simulate real-​life situations in
about high-​calorie foods and their consequences for which, according to the patient’s negative beliefs
weight and body shape. Therefore, intake avoidance (social rejection, criticism, etc.), weight gain may
is a central feature of this disorder and an insidious have catastrophic consequences.

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 477

 Patients with AN show high levels of anxiety, participants in whom the ownership illusion of a
avoidance of food intake and weight gain, which are thin virtual body had been induced showed higher
associated with rigid negative beliefs about the con- levels of body image satisfaction. The manipulation
sequences of increasing weight. Avoidance behaviors of these illusions has also proved effective for treat-
prevent patients from realizing that weight normal- ing disorders involving body image disturbances
ization does not have the catastrophic consequences (e.g., Schmalzl et al., 2011). Bearing these findings
expected. All these features create an extraordinary in mind, the hypothesis of the research project cur-
resistance to change, as one of the objectives of the rently underway at the University of Barcelona is
intervention is precisely to gain weight, the source of that the use of VR techniques to produce the illu-
their greatest fear. As a result, this proposal focuses sion of ownership of the virtual body will enhance
the exposure therapy on the primary feared stimu- the effect of body exposure therapy in AN patients.
lus (weight gain) instead of exposing the patient to The objective of the proposed intervention is to
his/​her actual body image, in order to reduce anxi- reduce the fear of weight gain by exposing the
ety and avoidance. Obviously, it is not possible to patients to an avatar that simulates their own body
use in vivo exposure to achieve this aim, but imag- image with progressive BMI increases. It is expected
ery exposure may be a suitable procedure. To our that the induction of illusory ownership over the vir-
knowledge, only one case study has been carried tual body using techniques based on visual-​motor,
out in which imagery exposure was used to treat a visual-​tactile and visual-​proprioceptive correlations
restrictive AN patient who had showed resistance to can facilitate the extinction process.
treatment as usual (Levinson, Rapp, & Riley, 2014).
Five sessions of exposure therapy were applied as a Virtual Reality for the Study, Assessment,
part of a broader package of cognitive-​behavioral and Treatment of Body Image
techniques. During the sessions, the patient had Body image disturbance is a central feature of
to imagine herself progressively increasing weight EDs and plays an important role in the develop-
and facing the associated feared catastrophic con- ment, maintenance, and risk of relapse of these
sequences, such as criticism and social rejection. conditions (Haynos, Watts, Loth, Pearson, &
The authors reported that the patient gained weight Neumark-​Stzainer, 2016; McFarlane, Olmsted, &
during the intervention and maintained the weight Trottier, 2008; Rohde, Stice, & Marti, 2015; Stice,
increase after one-​month follow-​up. However, as 2016). Consequently, the research, assessment,
this was not a controlled study, it is not possible to and treatment of body image are a basic focus of
assess the specific contribution of imagery exposure interest. Body image is a multidimensional con-
to the improvement. struct reflecting a mental representation of the
Despite being a viable option for exposure ther- body’s physical appearance, which involves per-
apy, imagery exposure presents some important ceptual, cognitive, and affective aspects and influ-
issues which may make the use of this procedure ences behavior (Ahrberg, Trojca, Nasrawi, & Vocks,
difficult in some patients. These problems include 2011; Gaudio & Quattrocchi, 2012). Moreover, it
difficulty in maintaining visualization long enough is a dynamic representation; that is, it changes over
to reach habituation of response, individual differ- time and depends on everyday experiences and
ences in the ability to visualize a situation, and the sociocultural contexts (Espeset, Gulliksen, Nordbo,
risk that the most feared stimulus may be avoided Skarderud, & Holte, 2012). Given the complexity
during visualization. Virtual reality allows thera- of body image, researchers usually focus on its per-
pists greater control over the exposure parameters, ceptual and cognitive-​affective components and the
thus reducing the possibility of avoidance behaviors associated disturbances: the perceptual distortion of
by patients, and its efficacy is not conditioned by body image (i.e., the inability to perceive the size of
patients’ capacity for visualization. Therefore, VR the body accurately) and body dissatisfaction (i.e.,
exposure therapy may be an appropriate alternative. the degree to which a person likes or dislikes the
Besides, the group headed by Gutiérrez-​Maldonado size and shape of his/​her body and values it) (Cash
proposes going further by using the induction of an & Deagle, 1997; Waldman, Loomes, Mountford, &
illusion of ownership over the virtual body (embod- Tchanturia, 2013).
ied cognition) in order to enhance exposure results. Probably, this complexity also explains why body
Preston and Ehrsson (2014) used the illusion image disturbance is often overlooked in ED treat-
of ownership over the virtual body to study body ments, despite being considered a core feature of
image dissatisfaction. These authors found that these disorders. In this context, the development of

478 Virtual Realit y

VR has provided researchers and clinicians with a & Flodin, 1989), or exposure to situations that
new technology that seems to be particularly well involve the scrutiny of others using guided imagery
suited to the study, assessment, and treatment of (Haimovitz, Lansky, & O’Reilly, 1993) may pro-
body image disturbances. The possibility of devel- duce changes in body image. However, the results
oping 3D figures that represent the body of the par- of these studies were not consistent. Virtual reality
ticipants and whose size and shape can be modified offered the possibility of exposing ED patients to
enables patients to embody their mental represen- emotionally significant situations (i.e., eating differ-
tations of the different components of body image ent kinds of food in a restaurant and in a kitchen),
(e.g., perceived vs. ideal body image). Moreover, the which contained both specific and contextual cues,
use of immersive systems such as head-​mounted and controlling the characteristics of all these cues.
displays brings patients face to face with their vir- Two main exposure variables were considered:  the
tual body in its actual size. Virtual reality simulates kind of food eaten (high calorie versus low calorie)
real-​life situations in which different aspects of body and the presence of other people (eating alone in
image disturbances can be studied, assessed, and a kitchen versus eating with colleagues in a restau-
even treated in a secure, private, and controlled set- rant). Furthermore, the physical aspect of the ava-
ting (Ferrer-​Garcia & Gutiérrez-​Maldonado, 2012). tars in the restaurant and dialogs included in the
exposure scenes were also controlled.
Virtual Reality for the Study of Body Image One hundred eight female undergraduates and
Disturbances 85 ED patients were exposed to four VR environ-
One of the main advantages that VR offers is the ments presented in random order (i.e., a kitchen
possibility of simulating real-​life situations in which with low-​calorie food, a kitchen with high-​calorie
participants’ behavior can be observed and assessed food, a restaurant with low-​ calorie food, and a
in a naturalistic context while maintaining good restaurant with high-​ calorie food). In the inter-
control over the variables (Botella et  al., 2004). vals between the presentation of each environ-
Additionally, VR environments allow researchers to ment anxiety and depressed mood were assessed
expose participants not only to specific cues (e.g., by means of questionnaires, and body image dis-
chocolate) but also to contextual ones (e.g., a res- turbances (perceptual body image distortion and
taurant setting). Last but not least, ED patients are body image dissatisfaction) were measured using
much less reluctant to expose themselves to these the Body Image Assessment Software (Ferrer-​Garcia
cues in VR environments than in real situations. & Gutiérrez-​Maldonado, 2008).
All these characteristics make this technology espe- The results obtained in this study showed that
cially useful for experimental research and, thus, for body image distortion (i.e., overestimation of body
broadening our knowledge of the nature and fea- size) and body image dissatisfaction were signifi-
tures of body image disturbance. cantly higher after eating high-​calorie food (e.g.,
An issue that has been repeatedly addressed in a pizza) than after eating low-​calorie food (e.g., a
the study of body image is whether it is stable or salad), and that this was the case regardless of the
unstable. Myers and Biocca (1992) proposed the presence (in the restaurant) or the absence (in the
concept of “elastic body image” to refer to the kitchen) of other people (Gutiérrez-​Maldonado,
unstable self-​ perceived body image presented by Ferrer-​Garcia, Caqueo-​Urízar, & Moreno, 2010).
women after watching ideal-​body image commer- Anxiety and depressed mood also increased after eat-
cials on television. More recently, other authors ing high-​calorie food in this group (Ferrer-​Garcia,
have claimed that trait and state components coex- Gutiérrez-​Maldonado, Caqueo-​Urízar, & Moreno,
ist in the body image construct (Etu & Gray, 2010; 2009). In contrast, participants without ED showed
Lattimore & Hutchinson, 2010). similar percentages of body distortion and body
In order to study this issue in depth, Gutiérrez-​ dissatisfaction, as well as a similar mood, in all the
Maldonado and colleagues used VR to assess the situations. These results support the idea that body
stability/​
instability (intraindividual variability) image disturbance is unstable in ED patients, as the
of body image disturbances in patients with EDs. changes depend on the situation to which they are
Previous research had found that eating low-​and exposed. Therefore, both state and trait perspectives
high-​calorie food (McKenzie, Williamson, & need to be included in its assessment and treatment.
Cubic, 1993), exposure to low-​and high-​calorie In the same line of research, Mountford,
food (Carter & Bulik, 1994), exposure to photo- Tchanturia, and Valmaggia (2016) used VR technol-
graphs of low-​and high-​ calorie food (Heilbrun ogy to study whether perceptions of other people’s

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 479

appraisals influence the body image of ED patients. In the same line of research, Perpiñá et  al.
Twenty nonclinical women with low levels of body (1999) developed an immersive VR-​based appli-
image concerns and 20 nonclinical women with cation, which displayed a 3D human figure whose
high body image concerns were exposed to a virtual body parts could be modified using a slider bar.
simulation of a 4-​minute journey on a London bus Participants could enlarge the size of each body part
using a fully immersive system (i.e., a head-​mounted by moving the slider to the right and reduce it by
display). Several avatars were present in the bus, all moving the slider to the left, in order to represent
of which had normal body weight and appearance different dimensions of body image (e.g., perceived
and were programmed to display only neutral behav- body size, desired body size, healthy body size,
ior. Participants were asked to “form an impression etc.). Both the BIVRS and the software developed
of what you think about the people on the bus and by Perpiñá have mainly been used in the context
what they think about you” (Mountford et al., 2016, of treatment, which is the subject of the following
p. 95). The authors found that participants with high section.
levels of body image concerns also showed higher
levels of social evaluative concerns and comparison Virtual Reality for the Treatment of Body
with avatars during the virtual journey compared Image Disturbances
with participants who did not have concerns about In our culture most women are dissatisfied
body image. However, no changes in body image with their body:  One adolescent girl out of two
disturbances were elicited during exposure. Possibly, reports body dissatisfaction (Makinen, Puukko-​
the neutrality of the environment chosen, the low Viertomies, Lindberg, Siimes, & Aalberg, 2012).
realism of the avatars and the characteristics of the Recent studies have highlighted that the sociocul-
sample (nonclinical) explains these results. tural pressure to be thin is central to the develop-
Besides highlighting features of body image dis- ment of negative feelings about the body, which are
turbance and the need for further research into ways recognized as the most robust risk factor for clini-
of enhancing the potential of VR exposure, these cal and subclinical EDs (Dakanalis, Clerici, et  al.,
studies illustrate how VR can help us to develop 2013; Dakanalis et al., 2012; Dakanalis et al., 2015;
experimental studies with high ecological validity Dakanalis & Riva, 2013b; Dakanalis, Zanetti, Riva,
but with a proper control over the variables. & Clerici, 2013).
For this reason, a popular model of EDs—​
Virtual Reality for the Assessment of Body the “objectification theory”—​ suggests a signifi-
Image Disturbances cant role of culture and society in the etiology of
Most research into VR assessment of body image these disorders. Introduced by Fredrickson and
disturbances in ED, as well as in its treatment, has Roberts (1997), this theory suggests that our cul-
been conducted by two groups: Riva’s group in Italy ture imposes a specific self-​evaluation model—​self-​
and the group led by Perpiñá, Botella, and Baños objectification—​ that defines women’s behavioral
in Spain. As mentioned above, Riva was the first to and emotional responses (Calogero, Tantleff-​Dunn,
apply VR technology to the assessment and treat- & Thompson, 2010; Dakanalis et  al., 2012; Riva
ment of body image disturbances in ED (Riva et al., et al., 2014). At its simplest, the objectification the-
1997). As part of the European VREPAR Project, ory holds that (1) there exists an objectified societal
he and his colleagues developed the BIVRS (Body ideal of beauty (within a particular culture) that is
Image Virtual Reality Scale; Riva, 1998c; Riva & (2) transmitted via a variety of sociocultural chan-
Melis, 1997), a nonimmersive, 3D graphical inter- nels. This ideal is then (3) internalized by individu-
face for the assessment of body image dissatisfaction. als, so that (4) satisfaction (or dissatisfaction) with
The software displayed nine 3D figures, male and appearance will be a function of the extent to which
female, ranging from underweight to overweight. individuals do (or do not) meet the ideal prescrip-
Participants were asked to select the figures that tion (Tiggemann, 2011).
best fit their perceived and their desired body sizes. According to Fredrickson and Roberts (1997),
Discrepancy between the two measures was consid- repeated experiences of sexual objectification—​
ered an indicator of body image dissatisfaction. The when women are treated as bodies that exist for the
main advantage of this assessment method was its use and pleasure of others—​cause them to gradually
application of 3D to the human figures displayed by adopt an observer’s perspective of their physical self;
the software, thus increasing the realism and helping that is, they begin to treat themselves as an object to
participants to identify with them. be looked at and evaluated on the basis of physical

480 Virtual Realit y

appearance. The self is so defined in terms of how Colmegna, et al., 2015). Specifically in males, self-​
the body appears to others. objectification is manifested as body surveillance
The internalization of an observer’s perspective (Dakanalis & Riva, 2013a). As in women, frequent
on one’s own body is labeled as “self-​objectification” body surveillance increases attention to disliked
(Riva, 2014a, 2014b) and reduces a woman’s worth body parts, thereby encouraging the use of mala-
to her perception of her body’s semblance to cul- daptive eating and body shape regulation behaviors
tural standards of attractiveness (Dakanalis & Riva, to modify the body (Calogero, 2009; Wooldridge
2013b). Self-​objectification is typically manifested & Lytle, 2012). Furthermore, body surveillance in
as persistent body surveillance or habitual moni- men is strongly related to muscle dysmorphia (the
toring of the body’s outward appearance and is belief that one is small and skinny, despite well-​
believed to lead to a number of negative experiential developed musculature), a male disturbance whose
consequences such as body shame, social physique symptoms bear similarities to those of anorexia ner-
anxiety, lack of awareness of internal bodily states, vosa (Cafri, Olivardia, & Thompson, 2008; Murray
and decreased peak motivational states/​flow experi- et al., 2012).
ences (Dakanalis, Clerici, et al., 2013). A possible response to the second criticism is
There are two possible criticisms of this view. offered by a new etiological model, namely, the allo-
The first is that males, who apparently are less prone centric lock (AL) theory (Riva, 2012, 2014b). This
to self-​objectification, also experience EDs. Second, theory suggests that EDs, including AN, are the
only a small subset of all the female and male sub- outcome of a multisensory body integration deficit
jects exposed to idealized body models develop (Riva & Gaudio, 2017), dramatically impairing the
clinically diagnosable EDs (Thompson, Heinberg, way the body is “experienced” and “remembered.”
Altabe, & Tantleff-​Dunn, 1999). Specifically, individuals with (or who are develop-
Nevertheless, a number of recent studies have ing) these disorders may experience an impairment
underlined the possible role of self-​objectification in the way expected (using cognitive prediction) ver-
in the etiology of male EDs (Dakanalis, Clerici, sus experienced (from perception) bodily signals are
et  al., 2016; Dakanalis et  al., 2012; Dakanalis, integrated. In other words, the multisensory body
Timko, Clerici, Zanetti, & Riva, 2014; Dakanalis, integration process is biased towards the expected
Timko, et  al., 2015; Dakanalis, Zanetti, Riva, body representations and does not include the

Multisensory Body Integration Deficit

WORKING predicted
real-time body from
perceptions MEMORY
memory

Egocentric image Allocentric image

of the body of the body
(first person- (third person-
from inside) from outside)
ALLOCENTRIC LOCK: observer
field perspective
MEMORY OVERRIDES PERCEPTION perspective

BODY SHAME:
PERCEIVING ONESELF
NON MEETING CULTURAL STANDARDS

EXTREME RESTRAINED
EATING

EATING
DISORDERS

Figure 25.1  Allocentric Lock Theory.

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 481

available real-time information about the body. The Experiential cognitive therapy shares with CBT the
practical outcome of this deficit is that the subject is use of a combination of cognitive and behavioral
locked to an allocentric (third person) disembodied procedures to help the patient identify and change
negative memory of their body that is not updated the maintaining mechanisms. However, it differs
by perceptual data, even after a successful diet and/​ from CBT in the following ways:
or a significant weight reduction (Figure 25.1).
•  Use of VR: There are 15 VR sessions. The first
However, if even successful dieting attempts are
session is used to assess any stimuli that could elicit
unable to balance body image disturbance, people
abnormal eating behavior. Specifically, attention
may either start more radical dieting attempts or, at
is focused on the patient’s concerns about food,
the opposite end, all their attempts to control eating
eating, shape, and weight. At the end of the first
are abandoned and they engage in disinhibited eat-
VR session the therapist uses the miracle question, a
ing behaviors that can be followed by compensatory
typical approach used by the solution-​focused brief
behaviors, which can turn into a vicious cycle (for a
therapy (deShazer, 1985; McFarland, 1995). In this
broader review, see Dakanalis, Clerici, et al., 2016;
approach the therapist asks the patient to imagine
Gaudio & Riva, 2013; Riva, 2007, 2011, 2012,
what life would be like with a lower weight. Using
2014b; Riva, Gaudio, & Dakanalis, 2015).
VR to experience the effects of the miracle (Riva
From a cognitive viewpoint, this situation can be
et al., 2006)—​a virtual balance shows the weight
explained as the effect of a functional disconnection
expected by the patient—​individuals are more
between top-​down, premorbidly learned predictions
likely not only to gain an awareness of their need
regarding the experience of the body and the proc-
to do something to create change but also to
essing of bottom-​up perceptual information regard-
experience a greater sense of personal efficacy.
ing its current state (Gaudio, Wiemerslage, Brooks,
•  Focus on the negative emotions related to
& Schioth, 2016; Fotopoulou, 2015; Riva, 2016;
the body (a major reason patients want to lose
Serino et  al., 2015; Serino, Pedroli, et  al., 2016).
weight) and on supporting the empowerment
This view is in agreement with a recent hypothesis
process. In eight VR sessions, the therapist helps
that describes EDs as a disturbance of the self specif-
patients to recognize why they eat and what
ically associated “with spatial functioning possibly
they need to either avoid or cope with specific
related to experiencing one’s own body as an inte-
emotional/​behavioral triggers. This is achieved
grated aspect of the self, and temporal functioning
by integrating different cognitive-​behavioral
possibly related to integrating the self in a coherent
methods: Countering, Alternative Interpretation,
narrative over time” (Amianto, Northoff, Abbate
Label Shifting, Deactivating the Illness Belief.
Daga, Fassino, & Tasca, 2016, p. 7).
•  Focus on the experience of the body: It is
To modify this situation, the use of VR, a syn-
included via a specific VR body-​image rescripting
thetic egocentric experience, is an emerging and
protocol aimed at updating the patient’s memory
promising approach (Ferrer-​ Garcia, Gutiérrez-​
of the body (Riva, 2011). In the protocol,
Maldonado, & Riva, 2013). In particular, the two
involving six VR sessions, different body-​related
research groups mentioned above (Riva’s group
situations are experienced from both first-​person
in Milan and Perpiñá’s group in Castellón and
(the patient does not see his/​her body in the scene)
Valencia) are using VR to improve CBT, and have
and third-​person perspectives (the patient sees his/​
also developed VR-​based software for the assessment
her body in the scene) integrating the therapeutic
and treatment of body image disturbances (Myers,
methods used by Butters and Cash (1987) and
Swan-​Kremeier, Wonderlich, Lancaster, & Mitchell,
Wooley & Wooley (1985). In general, the therapist
2004; Perpiña, Botella, & Baños, 2003; Riva,
asks the patient to give detailed descriptions of the
Bacchetta, Cesa, Conti, & Molinari, 2002, 2004).
virtual experience and of the feelings associated
The first approach is offered by VR-​enhanced
with it. The patient is also taught how to cope with
CBT called experiential cognitive therapy (ECT),
these feelings using the “confrontation,” “feeding”
developed by Giuseppe Riva and his group inside
and “reconciliation” techniques (Leuner, 1969).
the VREPAR and VEPSY Updated European
funded projects. It is a relatively short-​term, patient-​ This approach was validated by various case stud-
oriented approach that focuses on individual dis- ies (Riva, Bacchetta, Baruffi, Rinaldi, & Molinari,
covery (Cesa et  al., 2013; Manzoni et  al., 2016; 1999) and trials. In the first one, which was uncon-
Riva, Bacchetta, Baruffi, & Molinari, 2001, 2002; trolled, three groups of patients were used (Riva
Riva, Bacchetta, Cesa, Conti, & Molinari, 2003). et al., 2000): patients with BED, patients with eating

482 Virtual Realit y

disorders not otherwise specified (EDNOS), and Rubber Hand Illusion, Botvinick & Cohen, 1998),
obese patients with a body mass index higher than there has been increasing interest in the study of
35. All patients participated in five twice-​weekly bodily illusions. Specifically, this term refers to
therapy sessions. All the groups showed improve- controlled illusory generation of unusual bodily
ments in overall body satisfaction, disordered eat- feelings, such as the feeling of ownership over a rub-
ing, and related social behaviors, although these ber hand that affects the experience of a body part
changes were less noticeable in the EDNOS group. or the entire body (i.e., a body-​swap illusion, see
The approach has also been tested in various con- below). In their review paper for Annals of Physical
trolled studies. The first one involved 20 women with and Rehabilitation Medicine, Dieguez and Lopez
BED who were seeking residential treatment (Riva, (Dieguez & Lopez, 2016) concluded that the exper-
Bacchetta, Baruffi, et  al., 2002). The sample was imental methods that create multisensory conflicts
assigned randomly to ECT or to CBT-​based nutri- to modulate body representations (i.e., bodily illu-
tional therapy. Both groups were prescribed a 1,200-​ sions) are promising noninvasive approaches for the
calorie per day diet and minimal physical activity. rehabilitation of patients with neurologic disorders.
Analyses revealed that although both groups were An increasing number of recent studies also suggest
binge-​free at 1-​month follow-​up, ECT was signifi- clinical applications for these methods in the treat-
cantly better at increasing body satisfaction. In addi- ment of weight-​related disorders (i.e., eating disor-
tion, ECT participants were more likely to report ders and obesity).
increased self-​efficacy and motivation to change. Pioneering research conducted by Riva’s team
In a second study, the same randomized approach (Serino, Pedroli, et al., 2016; Serino, Scarpina, et al.,
was used with a sample of 36 women with BED (Riva, 2016) has shown that the embodiment in a virtual
Bacchetta, et al., 2003). The results showed that 77% body that substitutes one’s own body in VR with
of the ECT group quit binging after 6 months versus visuotactile stimulation (the body-​ swap illusion)
56% for the CBT group and 22% for the nutritional alters body percept (i.e., participants are significantly
group sample. Moreover, the ECT sample reported fatter or thinner than they really are) suggesting,
better scores on most psychometric tests. among other things, that VR is more than a way of
The most recent controlled trial placing people in a simulated world (i.e., manipulat-
(ISRCTN59019572) included 211 obese (BMI > ing their sense of place). A first study (Serino, Pedroli,
40) female patients (Manzoni et al., 2016) and 90 et al., 2016) showed that the body-​swap illusion was
obese (BMI > 40) female patients with BED (Cesa able to update the negative stored representation
et al., 2013). In the trial ECT was compared with of the body. In particular, it has been found that
CBT and an integrated treatment (IT) including after embodying a virtual body with a skinny belly,
nutritional groups, a low-​calorie diet (1,200 kcal/​ women updated their “remembered body,” report-
day) and physical training. ing a significant (postillusion) decrease in their body-​
In both studies (Cesa et  al., 2013; Manzoni size distortion. A similar result was obtained recently
et al., 2016) only ECT was effective at improving by Keizer, van Elburg, Helms, & Dijkerman (2016)
weight loss at 1-​year follow-​up. Conversely, control using body swapping with a sample of 30 anorectic
participants regained most of the weight they had subjects: They decreased the overestimation of their
lost during the inpatient program. shoulders, abdomen, and hips both after the illusion
Furthermore, in the BED study (Cesa et  al., was induced and after a follow-​up (2 hours and 45
2013) binge eating episodes decreased to zero dur- minutes after the illusions).
ing the inpatient program but were reported again These results can also be explained by taking
in all the groups at 1-​year follow-​up. However, a a predictive account of brain functions (Friston,
substantial regain was observed only in the group 2010), which suggests that the fundamental func-
who received the integrated treatment alone, while tion of the brain is to constantly minimize the dis-
both ECT and CBT were successful in maintaining crepancy between sensory inputs and prior beliefs
a low rate of monthly binge eating episodes. about the causes of these inputs. When a strong
To further improve the efficacy of ECT, Riva and mismatch occurs, an update of the internal mod-
colleagues recently started to explore the possibil- els results. Consistent with this perspective, Preston
ity of integrating the emerging field of bodily illu- and colleagues (Preston & Ehrsson, 2014) induced
sions in the protocol. Since Botvinick and Cohen illusory ownership over a slimmer mannequin by
showed that it is easy to generate in people the illu- synchronously stroking the mannequin’s body and
sion that a rubber hand is part of their body (the the corresponding part of the participant’s body.

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 483

They found that illusory ownership over a slimmer treatment. However, those who had been treated
body significantly decreases participants’ perceived with the VR component showed a significantly
body size but also significantly increases their body greater improvement in general psychopathology,
satisfaction. ED psychopathology, and specific body image vari-
Support for the use of bodily illusions to alter the ables. What is more, these results were maintained
dysfunctional experience of the bodily self also came at 1-​year follow-​up (Perpiñá et al., 2003).
from a recent published study (Serino, Scarpina, This group’s most recent controlled trial included
et al., 2016) showing that a (VR) body-​swap illu- 34 patients diagnosed with ED (Marco, Perpiñá, &
sion, which generates the (converse) illusion that a Botella, 2013, 2014). Seventeen patients underwent
fat person is thin, was able to increase body satis- VR-​enhanced CBT and 17 classical CBT. The CBT
faction and reduce body-​size distortion in a non- program for EDs enhanced by a body image-​specific
operable super-​super obese patient (i.e., with body component using VR techniques was shown to be
mass index > 60 kg/​m2). In addition to the improve- more efficient than CBT alone in improvement of
ment in the bodily experience, the illusion was able body image. Furthermore, improvement was main-
to increase the patient’s motivation to maintain tained in post-​treatment and at 1-​year follow-​up.
healthy eating behaviors. While no studies to date
have directly exploited the potential of bodily illu- Conclusions
sions in ED treatment, the evidence deriving from Virtual reality has proven to be a useful technol-
the extant experimental studies (1) for a direct link ogy in the study, assessment, and treatment of a
between perceptual (i.e., an inability to accurately variety of psychological disorders. Studies on the
estimate body size) and affective (i.e., subjective application of this technology in the treatment of
body-​dissatisfaction) body-​image components, and EDs were some of the first ones conducted in the
(2)  for a positive affective response with the body early 1990s. Since then, several VR applications
illusion modulated by eating disorder psychopa- have been developed to be used in conjunction
thology (Dakanalis, Gaudio, et al., 2016) may sug- with traditional treatments, and their effectiveness
gest clinical applications for these methods (Serino has been tested in case studies, as well as in noncon-
& Dakanalis, 2016). trolled and controlled trials. The VR-​based inter-
Perpiñá’s group compared the effectiveness of ventions in EDs usually combine exposure to VR
VR with that of CBT for body image improve- environments with components based on cognitive
ment (based on Cash, 1997) in a controlled study therapy. Although the first VR-​based treatments
with a clinical population (Perpiña et  al., 1999). were developed specifically to address body image
Specifically, they developed six different virtual disturbances, the literature shows that there are
environments, including a 3D figure whose body an increasing number of studies focused on other
parts (arms, thighs, legs, breasts, stomach, but- aspects, such as anxiety, craving, avoidance, groom-
tocks, etc.) could be enlarged or reduced. The ing behaviors, self-​esteem, and self-​efficacy. Despite
proposed approach addressed several body image there being well-​established interventions that have
dimensions: the body could be evaluated wholly or proved effective for the treatment of EDs, relapses
in parts; the body could be placed in different con- are frequent, and a proportion of patients do not
texts (for instance, in the kitchen, before eating, improve after treatment. In this context, VR is a
after eating, facing attractive persons, etc.); behav- suitable technology for enhancing traditional CBT.
ioral tests could be performed in these contexts, The mechanisms underlying this effectiveness are
and several discrepancy indices related to weight probably linked with two characteristics of VR: It is
and figure could be combined (actual weight, sub- an experiential and an embodied technology.
jective weight, desired weight, healthy weight, how Patients with EDs typically feel ambivalent about
the person thinks others see her/​him, etc.). the idea of change for better health, because any
In the published trial, 18 outpatients who had change has advantages and disadvantages. Virtual
been diagnosed as suffering from EDs (AN or BN) reality is a special setting in which patients can
were randomly assigned to one of the two treatment explore and act; they can experiment and experi-
conditions: the VR condition (CBT plus VR) and ence feelings and thoughts. The exposure to virtual
the standard body image treatment condition (CBT environments produces emotional and behavioral
plus relaxation). Thirteen of the initial 18 partici- responses similar to those that occur in the real
pants completed the treatment. The results showed world, but VR is not only a good tool to re-​create
that all patients improved significantly following situations, it is also a place where patients learn to

484 Virtual Realit y

challenge and cope with distorted mental represen- systems in VR exposure interventions, they cannot
tations. Most VR applications to date have been be used in other procedures, such as those focused
used to simulate external reality; it is also possible on the production of the illusion of ownership of
to use VR for the simulation of the internal real- virtual bodies.
ity, including the perception and ownership of the
body. It is possible to use VR to induce controlled References
changes in the experience of the body, and this is Agras, W., Walsh, B. T., Fairburn, C. G., Wilson, G. T., &
particularly relevant given the role of body image Kraemer, H. C. (2000). A multicenter comparison of
disturbances in EDs. cognitive-​behavioral therapy and interpersonal psychother-
apy for bulimia nervosa. Archives of General Psychiatry, 57,
Typical VR systems involve different interfaces
459–​466. doi:10.1001/​archpsyc.57.5.459
for human–​computer interaction that vary with the Ahrberg, M., Trojca, D., Nasrawi, N., & Vocks, S. (2011).
required level of immersion. The most basic level Body image disturbance in binge eating disorder: A review.
involves the presentation of virtual environments European Eating Disorders Review, 19, 375–​381.
on computer screens; at the other extreme are tech- American Psychiatric Association. (2013). Diagnostic and sta-
tistical manual of mental disorders (5th ed.). Arlington,
nologically advanced systems such as head mounted
VA: American Psychiatric Publishing.
displays (HMDs). Although HMDs can increase Amianto, F., Northoff, G., Abbate Daga, G., Fassino, S., & Tasca,
the user’s immersive experience, their use can be G. A. (2016). Is Anorexia nervosa a disorder of the self?
impractical in clinical settings; clinical centers are A psychological approach. Frontiers in Psychology, 7, 849.
often reluctant to include VR interventions in their Baños, R. M., Botella, C., Alcaniz, M., Liano, V., Guerrero, B.,
& Rey, B. (2004). Immersion and emotion: Their impact on
daily practice if it involves the use of expensive or
the sense of presence. Cyberpsychology and Behavior, 7, 734–​
technically complex instruments. Another concern 741. doi:10.1089/​cpb.2004.7.734
is that HMDs may have side effects such as simula- Baños, R. M., Botella, C., Guerrero, B., Liaño, V., Alcañiz,
tor sickness, or visual fatigue. In addition to these M., & Rey, B. (2005). The third pole of the sense of pres-
practical issues, the price of immersive HMDs with ence:  Comparing virtual and imagery spaces. PsychNology
Journal, 3, 90–​100. http://​www.psychnology.org/​pnj103_​
good tracker systems has been prohibitive until
banos_​botella_​guerriero_​liano_​alcaniz_​rey_​abstract.htm
recently, which has delayed the development of Baños, R. M., Botella, C., Rubio, I., Quero, S., Garcia-​Palacios,
advanced virtual environments for routine clinical A., & Alcaniz, M. (2008). Presence and emotions in virtual
practice. Low-​cost immersive HMDs have appeared environments: The influence of stereoscopy. Cyberpsychology
on the market over the last year, allowing greater and Behavior, 11, 1–​8. doi:10.1089/​cpb.2007.9936
Bechara, A., & Damasio, A. (2005). The somatic marker hypoth-
immersion at relatively economical prices. However,
esis:  A  neural theory of economic decision. Games and
despite the development of low-​cost HMDs, some Economic Behavior, 52, 336–​372.
technical knowledge is still needed to use them Bergouignan, L., Nyberg, L., & Ehrsson, H. H. (2014). Out-​
properly, which may present a barrier to their wider of-​body-​induced hippocampal amnesia. Proceedings of the
clinical use. Thus, although low-​cost HMDs over- National Academy of Sciences, 111, 4421–​4426. doi:10.1073/​
pnas.1318801111
come economic drawbacks, the technological dif-
Berkman, N. D., Brownley, K. A., Peat, C. M., Lohr, K.
ficulties partly remain. Nonimmersive systems can N., Cullen, K. E., Morgan, L. C.,   .  .  .  Bulik, C. M.
also be useful. Incorporating more user-​ friendly (2015). Management and outcomes of binge-​eating disorder.
nonimmersive VR devices and clinical applications Comparative Effectiveness Review, 160. (Prepared by the
that do not require significant technical knowledge, RTI International–​University of North Carolina Evidence-​
Based Practice Center under Contract No. 290-​2012-​00008-​
we could reduce the current usability concerns of
I.) AHRQ Publication No. 15(16)-​EHC030-​EF. Rockville,
VR techniques. Other problems such as simulator MD: Agency for Healthcare Research and Quality; December
sickness or visual fatigue can also be reduced by 2015. www.effectivehealthcare.ahrq.gov/​reports/​final.cfm
using nonimmersive systems. This point should be Booth, R., & Rachman, S. (1992). The reduction of claustrophobia—​
taken into consideration, especially when the appli- I. Behaviour Research and Therapy, 30, 207–​221.
Botella, C., Quero, S., Baños, R. M., Perpina, C., Garcia Palacios,
cation of treatments using VR requires long periods
A., & Riva, G. (2004). Virtual reality and psychotherapy.
of time in each session, thus raising the likelihood of Studies in Health Technology and Informatics, 99, 37–​54.
the occurrence of simulator sickness or other similar Botvinick, M., & Cohen, J. (1998). Rubber hands “feel” touch
sources of discomfort. Negative side effects of this that eyes see. Nature, 391, 756. doi:10.1038/​35784
kind are extremely rare when the virtual exposure is Bouton, M. E. (2011). Learning and the persistence of appe-
tite:  Extinction and the motivation to eat and overeat.
carried out by nonimmersive devices such as desk-
Physiology and Behavior, 103, 51–​58.
top computer screens or laptops with stereoscopic Bulik, C. M., Sullivan, P. F., Carter, F. A., McIntosh, V. V., &
displays. However, even though nonimmersive sys- Joyce, P. R. (1998). The role of exposure with response pre-
tems have shown similar effectiveness as immersive vention in the cognitive-​ behavioural therapy for bulimia

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 485

 nervosa. Psychological Medicine, 28, 611–​623. doi:10.1017/​ and eating disorder symptomatology:  A  latent structural
S0033291798006618 equation modeling analysis of moderating variables in 18-​
Burdea, G. C., & Coiffet, P. (2003). Virtual reality technology to-​28-​year-​old males. Journal of Psychology, 149, 85–​112.
(2nd ed.). New Brunswick, NJ: Wiley-​IEEE Press. doi:10.1080/​00223980.2013.842141
Butters, J. W., & Cash, T. F. (1987). Cognitive-​behavioral treat- Dakanalis, A., Gaudio, S., Serino, S., Clerici, M., Carrà, G., &
ment of women’s body image satisfaction:  A  controlled Riva, G. (2016). Body-​image distortion in anorexia nervosa.
outcome-​study. Journal of Consulting and Clinical Psychology, Nature Reviews Disease Primers, 2, 16026. doi:10.1038/​
55, 889–​897. nrdp.2016.26
Cafri, G., Olivardia, R., & Thompson, J. K. (2008). Symptom Dakanalis, A., & Riva, G. (2013a). Current considerations for
characteristics and psychiatric comorbidity among males eating and body-​related disorders among men. In J. Latzer,
with muscle dysmorphia. Comprehensive Psychiatry, 49, 374–​ J. Merrick, & D. Stein (Eds.), Body image: Gender differences,
379. doi:10.1016/​j.comppsych.2008.01.003 sociocultural influences and health implication (pp. 195–​216).
Calogero, R. M. (2009). Objectification processes and disor- New York, NY: Nova Science.
dered eating in British women and men. Journal of Health Dakanalis, A., & Riva, G. (2013b). Mass media, body image and
Psychology, 14, 394–​402. doi:10.1177/​1359105309102192 eating disturbances: The underlying mechanism through the
Calogero, R. M., Tantleff-​Dunn, S., & Thompson, J. K. (2010). lens of the objectification theory. In J. Latzer, J. Merrick, &
Self-​objectification in women:  Causes, consequences, and D. Stein (Eds.), Body image: Gender differences, sociocultural
counteractions. Washington, DC:  American Psychological influences and health implication (pp. 217–​236). New York,
Association. NY: Nova Science.
Carter, F., & Bulik, C. M. (1994). Exposure treatments for buli- Dakanalis, A., Timko, A. C., Clerici, C., Zanetti, A. M., &
mia nervosa: Procedure, efficacy, and mechanisms. Advances Riva, G. (2014). Comprehensive examination of the trans-​
in Behavior Research and Therapy, 16, 77–​129. diagnostic cognitive behavioral model of eating disor-
Cash, T. F. (1997). The body image workbook:  An eight-​ step ders in males. Eating Behaviors, 15, 63–​67. doi:10.1016/​
program for learning to like your looks. Oakland, CA:  New j.eatbeh.2013.10.003
Harbinger. Dakanalis, A., Timko, A. C., Madeddu, F., Volpato, C., Clerici,
Cash, T. F., & Deagle, E. A. (1997). The nature and extent of M., Riva, G., & Zanetti, A. M. (2015). Are the male body
body image disturbances in anorexia nervosa and bulimia dissatisfaction and drive for muscularity scales reliable and
nervosa:  A  meta-​analysis. International Journal of Eating valid instruments? Journal of Health Psychology, 20(1), 48–​
Disorders, 22, 107–​125. 59. doi:10.1177/​1359105313498108
Cason, A. M., Smith, R. J., Tahsili-​Fahadan, P., Moorman, D. Dakanalis, A., Zanetti, M. A., Riva, G., & Clerici, M. (2013).
E., Sartor G. C., & Aston-​Jones, G. (2010). Role of orexin/​ Psychosocial moderators of the relationship between body
hypocretin in reward-​seeking and addiction: Implications for dissatisfaction and symptoms of eating disorders: A look at a
obesity. Physiology and Behavior, 100, 419–​458. doi:10.1016/​ sample of young Italian women. European Review of Applied
j.physbeh.2010.03.009 Psychology-​Revue Européenne de Psychologie Appliquée, 63,
Castelvecchi, D. (2016). Low-​ cost headsets boost virtual 323–​334. doi:10.1016/​j.erap.2013.08.001
reality’s lab appeal. Nature, 533, 153–​154. doi:10.1038/​ Dakanalis, A., Zanetti, A. M., Riva, G., Colmegna, F., Volpato,
533153a C., Madeddu, F., & Clerici, M. (2015). Male body dissat-
Cesa, G. L., Manzoni, G. M., Bacchetta, M., Castelnuovo, G., isfaction and eating disorder symptomatology:  Moderating
Conti, S., Gaggioli, A.,  . . .  Riva, G. (2013). Virtual reality variables among men. Journal of Health Psychology, 20(1),
for enhancing the cognitive behavioral treatment of obesity 80–​90. doi:  10.1177/​ 1359105313499198. doi:10.1177/​
with binge eating disorder:  Randomized controlled study 1359105313499198
with one-​year follow-​up. Journal of Medical Internet Research, deShazer, S. (1985). Keys to solutions in brief therapy. New York,
15, e113. doi:10.2196/​jmir.2441 NY: Norton.
Costantini, M. (2014). Body perception, awareness, and illu- Dieguez, S., & Lopez, C. (2016). The bodily self: Insights from
sions. Wiley Interdisciplinary Reviews:  Cognitive Science, 5, clinical and experimental research. Annals of Physical and
551–​560. doi:10.1002/​wcs.1309 Rehabilitation Medicine. doi:10.1016/​j.rehab.2016.04.007
Dakanalis, A., Clerici, M., Caslini, M., Gaudio, S., Serino, S., Diemer, J., Alpers, G. W., Peperkorn, H. M., Shiban, Y., &
Riva, G., & Carra, G. (2016). Predictors of initiation and per- Muhlberger, A. (2015). The impact of perception and presence
sistence of recurrent binge eating and inappropriate weight on emotional reactions: A review of research in virtual reality.
compensatory behaviors in college men. International Journal Frontiers in Psychology, 6, 26. doi:10.3389/​fpsyg.2015.00026
of Eating Disorders, 49, 581–​590. doi:10.1002/​eat.22535 Dunsmoor, J. E., Ahs, F., Zielinski, D. J., & LaBar, K. S. (2014).
Dakanalis, A., Clerici, M., Di Mattei, V. E., Caslini, M., L., F., Extinction in multiple virtual reality contexts diminishes
Bagliacca, E. P.,  . . .  Zanetti, M. A. (2013). Internalization fear reinstatement in humans. Neurobiology of Learning and
of sociocultural standards of beauty and eating disordered Memory, 113, 157–​164. doi:10.1016/​j.nlm.2014.02.010
behaviours: The role of body surveillance, shame, and social Ehrsson, H. H. (2007). The experimental induction of out-​of-​
anxiety. Journal of Psychopathology, 20, 33–​37. body experiences. Science, 5841, 1048.
Dakanalis, A., Di Mattei, V. E., Bagliacca, E. P., Prunas, A., Espeset, E. M. S., Gulliksen, K. S., Nordbo, R. H. S., Skarderud,
Sarno, L., Riva, G., & Zanetti, M. A. (2012). Disordered F., & Holte, A. (2012). Fluctuations for body images in
eating behaviors among Italian men: Objectifying media and anorexia nervosa: Patients’ perception of contextual triggers.
sexual orientation differences. Eating Disorders, 20, 356–​ Clinical Psychology and Psychotherapy, 19, 518–​530.
367. doi:10.1080/​10640266.2012.715514 Etu, S. F., & Gray J. J. (2010). A preliminary investigation of
Dakanalis, A., Favagrossa, L., Clerici, M., Prunas, A., Colmegna, the relationship between induced rumination and state body
F., Zanetti, M. A., & Riva, G. (2015). Body dissatisfaction image dissatisfaction and anxiety. Body Image, 7, 82–​85.

486 Virtual Realit y

Fairburn, C. G., Marcus, M. D., & Wilson, G. T. (1993). Gaudio, S., Wiemerslage, L., Brooks, S. J., & Schioth, H. B.
Cognitive behaviour therapy for binge eating and bulimia (2016). A systematic review of resting-​state functional-​MRI
nervosa: A treatment manual. In C. G. Fairburn and G. T. studies in anorexia nervosa:  Evidence for functional con-
Wilson (Eds.), Binge eating: Nature, assessment and treatment nectivity impairment in cognitive control and visuospatial
(pp. 361–​404). New York, NY: Guilford Press. and body-​signal integration. Neuroscience and Biobehavioral
Fairburn, C. G., Bailey-​Straebler, S., Basden, S., Doll, H. A., Reviews, 71, 578–​589. doi:10.1016/​j.neubiorev.2016.09.032
Jones, R., Murphy, R.,  . . .  Cooper, Z. (2015). A transdiag- Glantz, K., Durlach, N. I., Barnett, R. C., & Aviles, W. A. (1997).
nostic comparison of enhanced cognitive behaviour therapy Virtual reality (VR) and psychotherapy: Opportunities and
(CBT-​E) and interpersonal psychotherapy in the treatment challenges. Presence, Teleoperators, and Virtual Environments,
of eating disorders. Behavior Research and Therapy, 70, 64–​71. 6, 87–​105.
Ferrer-​García, M., & Gutiérrez-​ Maldonado, J. (2008). Body Gorini, A., Capideville, C. S., De Leo, G., Mantovani, F., & Riva,
image assessment software:  Psychometric data. Behavior G. (2011). The role of immersion and narrative in mediated
Research Methods, 40, 394–​407. presence:  The virtual hospital experience. Cyberpsychology,
Ferrer-​Garcia, M., & Gutierrez-​Maldonado, J. (2010). Effect of Behavior, and Social Networking, 14, 99–​105. doi:10.1089/​
the mood produced by virtual reality exposure on body image cyber.2010.0100
disturbances. Studies in Health Technology and Informatics, Gorini, A., Gaggioli, A., Vigna, C., & Riva, G. (2008). A second
154, 44–​49. life for eHealth: Prospects for the use of 3-​D virtual worlds in
Ferrer-​Garcia, M., & Gutierrez-​Maldonado, J. (2012). The use of clinical psychology. Journal of Medical Internet Research, 10, e21.
virtual reality in the study, assessment, and treatment of body Gorini, A., Griez, E., Petrova, A., & Riva, G. (2010). Assessment
image in eating disorders and nonclinical samples: A review of the emotional responses produced by exposure to real
of the literature. Body Image, 9, 1–​11. doi:  10.1016/​ food, virtual food and photographs of food in patients
j.bodyim.2011.10.001 affected by eating disorders. Annals of General Psychiatry,
Ferrer-​García, M., Gutiérrez-​Maldonado, J., Agliaro-​López, 9(30). doi: 10.1186/​1744-​859X-​9-​30
M., Lobera-​Espi, X., Pla, J., & Vilalta-​Abella, F. (2014). Gorini, A., & Riva, G. (2008). Virtual reality in anxi-
Validation of VR-​ based software for binge eating treat- ety disorders:  The past and the future. Expert Review of
ment:  Preliminary data. Studies in Health Technology and Neurotherapeutics, 8, 215–​233.
Informatics, 199, 146–​150. Gutierrez-​Maldonado, J., Ferrer-​ Garcia, M., Caqueo-​ Urizar,
Ferrer-​García, M., Gutiérrez-​Maldonado, J., Caqueo-​Urízar, A., A., & Moreno, E. (2009). The validity of virtual environ-
& Moreno, E. (2009). The validity of virtual environments ments for eliciting emotional responses in patients with eat-
for eliciting emotional responses in patients with eating dis- ing disorders and in controls. Behaviour Modification, 33,
orders and in controls. Behavior Modification, 3, 830–​854. 830–​854.
Ferrer-​Garcia, M., Gutiérrez-​Maldonado, J., Pla-​Sanjuanelo, Gutiérrez-​Maldonado, J., Ferrer-​García, M., Caqueo-​ Urízar,
J., Vilalta-​Abella, F., Andreu-​Gracia, A., Dakanalis, A.,  . . .  A., & Moreno, E. (2010). Body image in eating disor-
Sánchez, I. (2015). External eating as a predictor of cue-​ ders:  The influence of exposure to virtual environments.
reactivity to food-​ related virtual environments. Annual CyberPsychology and Behavior, 13, 521–​531.
Review of CyberTherapy and Telemedicine, 13, 117–​122. Gutiérrez-​Maldonado, J., Pla-​Sanjuanelo, J., & Ferrer-​García,
Ferrer-​Garcia, M., Gutiérrez-​Maldonado, J., & Riva, G. (2013). M. (2016). Cue-​ exposure software for the treatment of
Virtual reality based treatments in eating disorders and obe- bulimia nervosa and binge eating disorder. Psicothema, 28.
sity: A review. Journal of Contemporary Psychology, 43, 207–​ doi: 10.7334/​psicothema2014.274
221. doi:10.1007/​s10879-​013-​9240-​1 Gutierrez-​Maldonado, J., Wiederhold, B. K., & Riva, G. (2016).
Fotopoulou, A. (2015). The virtual bodily self:  Mentalisation Future directions:  How virtual reality can further improve
of the body as revealed in anosognosia for hemiplegia. the assessment and treatment of eating disorders and obesity.
Consciousness and Cognition, 33, 500–​510. doi:10.1016/​ Cyberpsychology, Behavior, and Social Networking, 19, 148–​
j.concog.2014.09.018 153. doi:10.1089/​cyber.2015.0412
Fredrickson, B. L., & Roberts, T. (1997). Objectification the- Haimovitz, D., Lansky, L. M., & O’Reilly, P. (1993). Fluctuations
ory: Toward understanding women’s lived experiences and men- in body satisfaction across situations. International Journal of
tal health risks. Psychology of Women Quarterly, 21, 173–​206. Eating Disorders, 13, 77–​84.
Friston, K. (2010). The free-​ energy principle:  A  unified Hale, K. S., & Stanney, K. M. (Eds.). (2014). Handbook of virtual
brain theory? Nature Reviews Neuroscience, 11, 127–​138. environments: Design, implementation, and applications (2nd
doi:10.1038/​nrn2787 ed.). Roca Baton, FL: CRC Press.
Gallace, A., & Spence, C. (2014). Outside the boundaries of our Hay, P. J., & Claudino, A. M. (2010). Evidence-​based treat-
bodies: The relationship between touch and the representa- ment for the eating disorders. In W. Stewart Agras (Ed.), The
tion of the body in our mind. In A. Gallace & C. Spence Oxford handbook of eating disorders (pp. 452–​479). Oxford,
(Eds.), In touch with the future: The sense of touch from cog- UK:  Oxford University Press. doi:  10.1093/​ oxfordhb/​
nitive neuroscience to virtual reality (pp. 102–​124). Oxford, 9780195373622.013.0025
UK: Oxford University Press. Haynos, A. F., Watts, A. W., Loth, K. A., Pearson, C. M.,
Gaudio, S., & Riva, G. (2013). Body image disturbances in Neumark-​Stzainer, D. (2016). Factors predicting an esca-
anorexia:  The link between functional connectivity altera- lation of restrictive eating during adolescence. Journal of
tions and reference frames. Biological Psychiatry, 73, e25–​e26. Adolescent Health, 59, 391–​396.
doi:10.1016/​j.biopsych.2012.08.028 Heilbrun, A. B., & Flodin, A. (1989). Food cues and perceptual
Gaudio, S., & Quattrocchi, C. C. (2012). Neural basis of a multi- distortion of the female body: Implications for food avoid-
dimensional model of body image distortion in anorexia ner- ance in the early dynamics of anorexia nervosa. Journal of
vosa. Neuroscience and Biobehavioral Reviews, 36, 1839–​1847. Clinical Psychology, 45, 843–​851.

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 487

Hildebrandt, T., Bacow, T., Markella, M., & Loeb, K. L. (2012). Makinen, M., Puukko-​Viertomies, L. R., Lindberg, N., Siimes,
Anxiety in anorexia nervosa and its management using M. A., & Aalberg, V. (2012). Body dissatisfaction and body
family-​based treatment. European Eating Disorders Review, mass in girls and boys transitioning from early to mid-​
20, 1–​16. doi: 10.1002/​erv.1071 adolescence: Additional role of self-​esteem and eating habits.
Hildebrandt, T., Loeb, K., Troupe, S., & Delinsky, S. (2012). BMC Psychiatry, 12, 35. doi:10.1186/​1471-​244X-​12-​35
Adjunctive mirror exposure for eating disorders: A random- Manzoni, G. M., Cesa, G. L., Bacchetta, M., Castelnuovo, G.,
ized controlled pilot study. Behaviour Research and Therapy, Conti, S., Gaggioli, A.,  . . .  Riva, G. (2016). Virtual reality-​
50, 797–​804. enhanced cognitive-​ behavioral therapy for morbid obe-
Hodges, L. F., Rothbaum, B. O., Alarcon, R., Ready, D., sity: A randomized controlled study with 1 year follow-​up.
Shahar, F., Graap, K.,   .  .  .  Baltzell, D. (1998). Virtual Cyberpsychology, Behavior, and Social Networking, 19, 134–​
Vietnam: A virtual environment for the treatment of Vietnam 140. doi:10.1089/​cyber.2015.0208
War veterans with post-​traumatic stress disorder. Paper pre- Marco, J. H., Perpina, C., & Botella, C. (2013). Effectiveness of
sented at the International Conference on Artificial Reality cognitive behavioral therapy supported by virtual reality in
and Tele-​Existence. Tokyo. the treatment of body image in eating disorders:  One year
Hone-​Blanchet, A., Wensing, T., & Fecteau, S. (2014). The use follow-​up. Psychiatry Research, 209, 619–​625. doi:10.1016/​
of virtual reality in craving assessment and cue-​exposure j.psychres.2013.02.023
therapy in substance use disorders. Frontiers in Human Marco, J. H., Perpiñá, C., & Botella, C. (2014). The treatment
Neuroscience, 8. Published online 2014 Oct 17. doi: 10.3389/​ of the body image disturbances in eating disorders and
fnhum.2014.00844 clinically significant change. Anales de Psicología/​Annals of
Jansen, A. (1994). The learned nature of binge eating. In C. R. Psychology, 30, 422–​430.
Legg & D. A. Booth (Eds.), Appetite: Neural and behavioral Maselli, A., & Slater, M. (2014). Sliding perspectives: Dissociating
bases (pp. 193–​211). Oxford, UK: Oxford University Press. ownership from self-​location during full body illusions in vir-
Jansen, A. (1998). A learning model of binge eating: Cue reac- tual reality. Frontier is Human Neuroscience, 8. doi:10.3389/​
tivity and cue exposure. Behaviour Research and Therapy, 36, fnhum.2014.00693
257–​272. Martínez-​Mallén, E., Castro-​Fornieles, J., Lázaro, L., Moreno,
Jansen, A., Broekmate, J., & Heymans, M. (1992). Cue exposure E., Morer, A., Font, E.,  . . .  Toro, J. (2007). Cue exposure
vs. self-​control in the treatment of binge eating. Behaviour in the treatment of resistant adolescent bulimia nervosa.
Research and Therapy, 3, 235–​241. International Journal of Eating Disorders, 40(7), 596–​601.
Jansen, A., van den Hout, M. A., de Loof, C., Zandbergen, J., Doi: 10.1002/​eat.20423
& Griez, E. (1989). A case of bulimia successfully treated by McFarland, B. (1995). Brief therapy and eating disorders. San
cue exposure. Journal of Behavior Therapy and Experimental Francisco, CA: Jossey-​Bass.
Psychiatry, 20, 327–​332. McFarlane, T., Olmsted, M. P., & Trottier, K. (2008). Timing
Keizer, A., van Elburg, A., Helms, R., & Dijkerman, H. C. and prediction of relapse in a transdiagnostic eating disor-
(2016). A virtual reality full body illusion improves body der sample. International Journal of Eating Disorders, 41,
image disturbance in anorexia nervosa. PLoS One, 11, 587–​593.
e0163921. doi:10.1371/​journal.pone.0163921 McKenzie, S. J., Williamson, D. A., & Cubic, B. A. (1993).
Kennedy, R. S., Stanney, K. M., & Dunlap, W. P. (2000). Stable and reactive body image disturbance in bulimia ner-
Duration and exposure to virtual environments:  Sickness vosa. Behavior Therapy, 24, 195–​207.
curves during and across sessions. Presence: Teleoperators and Melzack, R. (2005). Evolution of the neuromatrix theory of
Virtual Environments, 9, 463–​472. pain. The Prithvi Raj Lecture: Presented at the Third World
Lattimore, P., & Hutchinson, R. (2010). Perceived calorie intake Congress of World Institute of Pain, Barcelona 2004. Pain
and state body-​image satisfaction in women attempting weight Practice, 5, 85–​94.
loss: A preliminary investigation. Body Image, 7, 15–​21. Morgan, J. F., Lazarova, S., Schelhase, M., & Saeidi, S. (2014).
Lee, K. M. (2004). Presence, explicated. Communication Theory, Ten Session Body Image Therapy: Efficacy of a Manualised
14, 27–​50. Body Image Therapy. European Eating Disorders Review,
Lenggenhager, B., Tadi, T., Metzinger, T., & Blanke, O. (2007). 22(1), 66–​71.
Video ergo sum:  manipulating bodily self-​ consciousness. Murray, S. B., Rieger, E., Hildebrandt, T., Karlov, L., Russell, J.,
Science, 317, 1096–​1099. doi: 10.1126/​science.1143439 Boon, E.,  . . .  Touyz, S. W. (2012). A comparison of eating,
Leuner, H. (1969). Guided affective imagery: A method of inten- exercise, shape, and weight related symptomatology in males
sive psychotherapy. American Journal of Psychotherapy, 23, 4–​21. with muscle dysmorphia and anorexia nervosa. Body Image,
Ling, Y., Nefs, H. T., Morina, N., Heynderickx, I., & Brinkman, 9, 193–​200. doi:10.1016/​j.bodyim.2012.01.008
W. P. (2014). A meta-​analysis on the relationship between Mountford, V. A., Tchanturia, K., & Valmaggia, L. (2016).
self-​reported presence and anxiety in virtual reality expo- “What are you thinking when you look at me?” A pilot study
sure therapy for anxiety disorders. PLoS One, 9, e96144. of the use of virtual reality in body image. Cyberpsychology,
doi:10.1371/​journal.pone.0096144 Behavior, and Social Networking, 19, 93–​99.
Lombard, M., Biocca, F., Freeman, J., IJsselsteijn, Myers, P. N., & Biocca, F. A. (1992). The elastic body image: The
W., & Schaevitz, R. J. (Eds.). (2015). Immersed in effect of television advertisement and programming
media:  Telepresence theory, measurement and technology. on body image distortions in young women. Journal of
Heidelberg, Germany: Springer. Communication, 42, 108–​133.
Mahler, S. V., & de Wit, H. (2010). Cue-​Reactors:  Individual Myers, T. C., Swan-​Kremeier, L., Wonderlich, S., Lancaster, K.,
Differences in Cue-​Induced Craving after Food or Smoking & Mitchell, J. E. (2004). The use of alternative delivery systems
Abstinence. PLoS ONE, 5(11), e15475. doi:10.1371/​jour- and new technologies in the treatment of patients with eating
nal.pone.0015475 disorders. International Journal of Eating Disorders, 36, 123–​143.

488 Virtual Realit y

Nair, S. G., Adams-​Deutscha, T., Epsteinb, D. H., Shaham, con realidad virtual en poblaciones clínicas que padecen
Y. (2009). The neuropharmacology of relapse to food ansiedad. Revista de Psicopatología y Psicología Clínica, 19,
seeking:  Methodology, main findings, and comparison 197–​207. doi:  http://​dx.doi.org/​10.5944/​rppc.vol.19.
with relapse to drug seeking. Progress in Neurobiology, num.3.2014.13901
89(1), 18–​45. Regan, E. C., & Price, K. R. (1993). Some side-​effects of immersion
National Institute for Health and Clinical Excellence (2004). virtual reality: The effects of increasing head movements, of rapid
Eating Disorders: core interventions in the treatment and manage- interaction, and of seated subjects (Report 93R022). Retrieved
ment of anorexia nervosa, bulimia nervosa and eating disorders. from Farnborough (UK):  Army Personnel Research
Clinical Guideline 9. National Institute of Excellence, London. Establishment. http://​discovery.nationalarchives.gov.uk
North, M. M., & North, S. M. (1994). Virtual environments Riva, G. (Ed.) (1997). Virtual reality in neuro-​ psycho-​
and psychological disorders. Electronic Journal of Virtual physiology:  Cognitive, clinical and methodological issues in
Culture, 2, 37–​42 (ep.). assessment and rehabilitation. Amsterdam: IOS Press. http://​
North, M. M., & North, S. M. (2000). Virtual reality combats www.cybertherapy.info/​pages/​book1.htm
obsessive-​compulsive disorders (OCD). In Proceedings of Riva, G. (1998a). Modifications of body image induced by vir-
Medicine Meets Virtual Reality (8th Annual Medicine Meets tual reality. Perceptual and Motor Skills, 86, 163–​170.
Virtual Reality Conference), James Westwood, et  al. eds., Riva, G. (1998b). Virtual environments in neuroscience. IEEE
IOS Press, Amsterdam. Transactions on Information Technology in Biomedicine, 2,
North, M. M., North, S. M., & Coble, J. R. (1995). Effectiveness 275–​281.
of virtual environment desensitization in the treatment Riva, G. (1998c). Virtual reality in psychological assess-
of agoraphobia. International Journal of Virtual Reality, ment:  The body image virtual reality scale. Cyberpsychology
1, 25–​34. and Behavior, 1, 37–​44.
North, M. M., North, S. M., & Coble, J. R. (1997). Virtual real- Riva, G. (2007). Virtual body, real pain:  The allocentric lock
ity therapy:  An effective treatment for psychological disor- hypothesis. In M. Zampini & F. Pavani (Eds.), Body repre-
ders. Studies in Health Technology and Informatics, 44, 59–​70. sentation workshop. Rovereto, Italy: Università degli Studi di
North, M. M., North, S. M., Coble, J. R. (1998). Virtual Reality Trento, p. 72 http://​www.cimec.unitn.it/​events/​brw/​Poster/​
Therapy:  an Effective Treatment for the Fear of Public 107_​abs_​GIUSEPPE_​RIVA.pdf ).
Speaking. The International Journal of Virtual Reality 3, 1–​6. Riva, G. (2009). Is presence a technology issue? Some insights
Olive, I., & Berthoz, A. (2012). Combined induction of rubber-​ from cognitive sciences Virtual Reality, 13, 59–​69.
hand illusion and out-​ of-​
body experiences. Frontiers in Riva, G. (2011). The key to unlocking the virtual body: Virtual
Psychology, 3, 128. doi:10.3389/​fpsyg.2012.00128 reality in the treatment of obesity and eating disorders.
Perez-​Marcos, D., Slater, M., & Sanchez-​Vives, M. V. (2009). Journal of Diabetes Science and Technology, 5, 283–​292.
Inducing a virtual hand ownership illusion through a brain-​ Riva, G. (2012). Neuroscience and eating disorders:  The allo-
computer interface. Neuroreport, 20, 589–​594. centric lock hypothesis. Medical Hypotheses, 78, 254–​257.
Pericot-​ Valverde, I., García-​Rodríguez, O., Gutiérrez-​ doi:10.1016/​j.mehy.2011.10.039
Maldonado, G., & Secades-​Villa, R. (2015). Individual vari- Riva, G. (2014a). Medical clinical uses of virtual worlds. In M.
ables related to craving reduction in cue exposure treatment. Grimshaw (Ed.), The Oxford handbook of virtuality (pp. 649–​
Addictive Behaviors, 49, 59–​63. 665). New York, NY: Oxford University Press.
Perpiña, C., Botella, C., & Baños, R. M. (2003). Virtual real- Riva, G. (2014b). Out of my real body: Cognitive neuroscience
ity in eating disorders. European Eating Disorders Review, 11, meets eating disorders. Frontiers in Human Neuroscience, 8,
261–​278. 236. doi:10.3389/​fnhum.2014.00236
Perpiña, C., Botella, C., Baños, R. M., Marco, J. H., Alcañiz, Riva, G. (2016). Embodied Medicine: What Human-​Computer
M., & Quero, S. (1999). Body image and virtual reality in Confluence Can Offer to Health Care. In A. Gaggioli, A.
eating disorders: Exposure by virtual reality is more effective Ferscha, G. Riva, S. Dunne, & I. Viaud-​Delmon (Eds.),
than the classical body image treatment? Cyberpsychology and Human Computer Confluence:  Transforming Human
Behavior, 3, 149–​159. Experience Through Symbiotic Technologies (pp. 55–​79).
Pillai, J. S., Schmidt, C., & Richir, S. (2013). Achieving pres- Warsaw: De Gruyter Open.
ence through evoked reality. Frontiers in Psychology, 4. Riva, G., Bacchetta, M., Baruffi, M., & Molinari, E. (2001).
doi:10.3389/​fpsyg.2013.00086 Virtual reality-​based multidimensional therapy for the treat-
Pla-​Sanjuanelo, J., Ferrer-​Garcia, M., Gutiérrez-​Maldonado, ment of body image disturbances in obesity:  A  controlled
J., Vilalta-​Abella, F., Andreu-​Gracia, A., Dakanalis, A.,  . . .  study. Cyberpsychology and Behavior, 4, 511–​526.
Sánchez, I. (2015). Trait and state craving as indicators Riva, G., Bacchetta, M., Baruffi, M., & Molinari, E. (2002).
of validity of VR-​ based software for binge eating treat- Virtual-​reality-​based multidimensional therapy for the
ment. Annual Review of CyberTherapy and Telemedicine, 13, treatment of body image disturbances in binge eating dis-
141–​146. orders: A preliminary controlled study. IEEE Transactions on
Pomes, A., & Slater, M. (2013). Drift and ownership toward a Information Technology in Biomedicine, 6, 224–​234.
distant virtual body. Frontiers in Human Neuroscience, 7, 908. Riva, G., Bacchetta, M., Baruffi, M., Rinaldi, S., & Molinari,
doi:10.3389/​fnhum.2013.00908 E. (1999). Virtual reality based experiential cognitive treat-
Preston, C., & Ehrsson, H. H. (2014). Illusory changes in body ment of anorexia nervosa. Journal of Behavioral Therapy and
size modulate body satisfaction in a way that is related to Experimental Psychiatry, 30, 221–​230.
non-​clinical eating disorder psychopathology. PLoS One, 9. Riva, G., Bacchetta, M., Baruffi, M., Rinaldi, S., Vincelli, F., &
doi:10.1371/​journal.pone.0085773 Molinari, E. (2000). Virtual reality-​based experiential cogni-
Quintana, P., Bouchard, S., Serrano, B., & Cárdenas-​López, tive treatment of obesity and binge-​eating disorders. Clinical
G. (2014). Efectos secundarios negativos de la inmersión Psychology and Psychotherapy, 7, 209–​219.

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 489

Riva, G., Bacchetta, M., Cesa, G., Conti, S., Castelnuovo, and social presence in mediated interactions. In G. Riva, J.
G., Mantovani, F., & Molinari, E. (2006). Is severe obe- A. Waterworth, & D. Murray (Eds.), Interacting with pres-
sity a form of addiction? Rationale, clinical approach, and ence:  HCI and the sense of presence in computer-​ mediated
controlled clinical trial. CyberPsychology and Behavior, 9, environments (pp. 12–​34). Berlin: De Gruyter Open. http://​
457–​479. www.presence-​research.com/​
Riva, G., Bacchetta, M., Cesa, G., Conti, S., & Molinari, E. Riva, G., Mantovani, F., & Gaggioli, A. (2004). Presence
(2002). e-​ health in eating disorders:  Virtual Reality and and rehabilitation:  Toward second-​ generation vir-
telemedicine in assessment and treatment. Studies in Health tual reality applications in neuropsychology. Journal of
Technology and Informatics, 85, 402–​408. Neuroengineering and Rehabilitation, 1, 9.  doi:1743-​
Riva, G., Bacchetta, M., Cesa, G., Conti, S., & Molinari, E. 10.1186/​1743-​0003-​1-​9
(2003). Six-​ month follow-​ up of in-​patient experiential-​ Riva, G., & Melis, L. (1997). Virtual reality for the treatment of
cognitive therapy for binge eating disorders. Cyberpsychology body image disturbances. Studies in Health Technology and
and Behavior, 6, 251–​258. Informatics, 44, 95–​111.
Riva, G., Bacchetta, M., Cesa, G., Conti, S., & Molinari, E. Riva, G., Melis, L., & Bolzoni, M. (1997). Treating body image
(2004). The use of VR in the treatment of eating disorders. disturbances. Communications of the ACM, 40, 69–​71.
Studies in Health Technology and Informatics, 99, 121–​163. Riva, G., & Waterworth, J. A. (2003). Presence and the
Riva, G., Baños, R., Botella, C., Mantovani, F., & Gagglioli, self:  A  cognitive neuroscience approach. Presence-​Connect,
A. (2016). Transforming experience:  The potential of aug- 3, 1–​14. http://​www8.informatik.umu.se/​~jwworth/​Riva-​
mented reality and virtual reality for enhancing personal and Waterworth.htm
clinical change. Frontiers in Psychiatry, 7, 164. doi: 10.3389/​ Riva, G., Waterworth, J. A., & Murray, D. (Eds.). (2014).
fpsyt.2016.00164 Interacting with presence:  HCI and the sense of presence in
Riva, G., Botella, C., Baños, R., Mantovani, F., García-​Palacios, computer-​mediated environments. Berlin: De Gruyter Open.
A., Quero, S.,  . . .  Gaggioli, A. (2015). Presence-​inducing http://​www.presence-​research.com/​
media for mental health applications. In M. Lombard, F. Riva, G., & Wiederhold, B. K. (2015). The new dawn of virtual
Biocca, J. Freeman, W. Ijsselsteijn, & R. J. Schaevitz (Eds.), reality in health care:  Medical simulation and experiential
Immersed in media (pp. 283–​332). New York, MY: Springer interface. Annual Review of CyberTherapy and Telemedicine,
International. 13, 3–​6.
Riva, G., Dakanalis, A., & Mantovani, F. (2015). Leveraging Rizzo, A. A., Buckwalter, J. G., Bowerly, T., Van Der Zaag, C.,
psychology of virtual body for health and wellness. In S. S. Humphrey, L., Neumann, U.,   .  .  .  Sisemore, D. (2000).
Sundar (Ed.), The handbook of the psychology of communica- The Virtual Classroom:  A  virtual reality environment for
tion technology (pp. 528–​547). Chichester, UK: Wiley. the assessment and rehabilitation of attention deficits.
Riva, G., Davide, F., & IJsselsteijn, W. A. (Eds.). (2003). Being Cyberpsychology and Behavior, 3(3), 483–​499.
there:  Concepts, effects and measurements of user presence in Rizzo, A. A., Schultheis, M. T., & Rothbaum, B. O. (2002).
synthetic environments. Amsterdam:  Ios Press. http://​www. Ethical issues for the use of virtual reality in the psycho-
emergingcommunication.com/​volume5.html logical sciences. In S. Bush & M. Drexler (Eds.), Ethical
Riva, G., Gaggioli, A., Villani, D., Preziosa, A., Morganti, F., issues in clinical neuropsychology (pp. 243–​279). Lisse, the
Corsi, R.,   .  .  .  Vezzadini, L. (2007). NeuroVR:  An open Netherlands: Swets & Zeitlinger.
source virtual reality platform for clinical psychology and Rohde, P., Stice, E., Marti, C. N. (2015). Development and
behavioral neurosciences. Studies in Health Technology and predictive effects of eating disorder risk factors during ado-
Informatics, 125, 394–​399. lescence:  Implications for prevention efforts. International
Riva, G., & Gaudio, S. (in press). Locked to a wrong body: Journal of Eating Disorders, 48, 187–​198.
Eating disorders as the outcome of a primary disturbance in Rosa, P. J., Morais, D., Gamito, P., Oliveira, J., & Saraiva, T.
multisensory body integration. Consciousness and Cognition. (2016). The immersive virtual reality experience: A typology
doi: 10.1016/j.concog.2017.08.006 of users revealed through multiple correspondence analysis
Riva, G., Gaudio, S., & Dakanalis, A. (2015). The neuropsychol- combined with cluster analysis technique. Cyberpsychology,
ogy of self objectification. European Psychologist, 20, 34–​43. Behavior, and Social Networking, 19, 209–​216.
doi:10.1027/​1016-​9040/​a000190 Rothbaum, B., Hodges, L., Kooper, R., Opdykes, D., Williford,
Riva, G., & Mantovani, F. (2012a). Being there: Understanding J., & North, M. (1995). Effectiveness of computer-​generated
the feeling of presence in a synthetic environment and (virtual reality) graded exposure in the treatment of acropho-
its potential for clinical change. In C. Eichenberg (Ed.), bia. American Journal of Psychiatry, 152, 626–​628.
Virtual reality in psychological, medical and pedagogi- Sanchez-​Vives, M. V., & Slater, M. (2005). From presence
cal applications. (pp.  3–​34). New  York, NY:  InTech. to consciousness through virtual reality. Nature Reviews
http://​ w ww.intechopen.com/ ​ b ooks/ ​ v irtual- ​ r eality- ​ i n-​ Neuroscience, 6, 332–​339.
psychological-​ m edical- ​ a nd- ​ p edagogical- ​ a pplications/​ Schare, M. L., Scardapane, J. R., Berger, A. L., Rose, N. &
being-​there-​understanding-​the-​feeling-​of-​presence-​in-​a-​ Berger, S. (1999). A virtual reality based anxiety induction pro-
synthetic-​environment-​and-​its-​potential-​for-​c 3-​34 cedure with driving phobic patients. Presented at Association
Riva, G., & Mantovani, F. (2012b). From the body to the for Advancement of Behavior Therapy, Toronto, Canada.
tools and back: A general framework for presence in medi- Schmidt, U., & Marks, I. (1988). Cue exposure to food plus
ated interactions. Interacting with Computers, 24, 203–​210. response prevention of binges for bulimia:  A  pilot study.
doi:10.1016/​j.intcom.2012.04.007 International Journal of Eating Disorders, 7(5), 663–​672.
Riva, G., & Mantovani, F. (2014). Extending the self through Schyns, G., Roefs, A., Mulkens, S., & Jansen, A. (2016).
the tools and the others: A general framework for presence Expectancy violation, reduction of food cue reactivity and

490 Virtual Realit y

 less eating in the absence of hunger after one food cue expo- Stice, E. (2016). Interactive and mediational etiologic models
sure session for overweight and obese women. Behavior of eating disorder onset: Evidence from prospective studies.
Research and Therapy, 76, 57–​64. Annual Review of Clinical Psychology, 12, 359–​381.
Schultheis, M. T., & Rizzo, A. A. (2001). The application of Strickland, D. (1996). A virtual reality application with autistic
virtual reality technology in rehabilitation. Rehabilitation children. Presence, Teleoperators and Virtual Environments, 6,
Psychology, 46, 296–​311. 319–​329.
Serino, S., & Dakanalis, A. (2016). Bodily illusions and Thompson, J. K., Heinberg, L. J., Altabe, M., & Tantleff-​Dunn,
weight-​related disorders:  Clinical insights from experimen- S. (1999). Exacting beauty:  Theory, assessment and treat-
tal research. Annals of Physical and Rehabilitation Medicine ment of body image disturbance. Washington DC: American
doi: 10.1016/​j.rehab.2016.10.002 Psychological Association.
Serino, S., Dakanalis, A., Gaudio, S., Carra, G., Cipresso, Tiggemann, M. (2011). Sociocultural perspectives on human
P., Clerici, M., & Riva, G. (2015). Out of body, out of appearance and body image. In T. F. Cash & L. Smolak
space:  Impaired reference frame processing in eating dis- (Eds.), Body image: A handbook of science, practice, and pre-
orders. Psychiatry Research, 230, 732–​734. doi:10.1016/​ vention (pp. 12–​19). New York, NY: Guilford.
j.psychres.2015.10.025 Toro, J., Cervera, M., Feliu, M. H., Garriga, N., Jou, M.,
Serino, S., Pedroli, E., Keizer, A., Triberti, S., Dakanalis, A., Martinez, E., & Toro, E. (2003). Cue exposure in the treat-
Pallavicini, F.,  . . .  Riva, G. (2016). Virtual Reality Body ment of resistant bulimia nervosa. International Journal of
Swapping: A Tool for Modifying the Allocentric Memory of Eating Disorders, 34, 1–​8.
the Body. Cyberpsychology, Behavior, and Social Networking, Trentowska, M., Bender, C., Tuschen-​Caffier, B. (2013). Mirror
19(2), 127–​133. doi:10.1089/​cyber.2015.0229 exposure in women with bulimic symptoms, how do
Serino, S., Scarpina, F., Keizer, A., Pedroli, E., Dakanalis, A., thoughts and emotions change in body image treatment?
Castelnuovo, G.,  . . .  Riva, G. (2016). A novel technique for Behaviour Research and Therapy, 51, 1–​6.
improving bodily experience in a non-​operable super-​super Vincelli, F. (1999). From imagination to virtual reality:  the
obesity case. Frontiers in Psychology, 7, 837. doi:10.3389/​ future of clinical psychology. Cyberpsychology and Behavior,
fpsyg.2016.00837 2, 241–​248.
Sethi, A. K., Suzuki, K., & Critchley, H. D. (2012). An intero- Vincelli, F., Molinari, E., & Riva, G. (2001). Virtual reality as
ceptive predictive coding model of conscious presence. clinical tool:  Immersion and three-​ dimensionality in the
Frontiers in Psychology, 2, article 395, 1–​16. doi:10.3389/​ relationship between patient and therapist. Studies in Health
fpsyg.2011.00395 Technology and Informatics, 81, 551–​553.
Shiban, Y., Pauli, P., & Muhlberger, A. (2013). Effect of mul- Waldman, A., Loomes, R., Mountford, V. A., & Tchanturia, K.
tiple context exposure on renewal in spider phobia. (2013). Attitudinal and perceptual factors in body image
Behaviour Research and Therapy, 51, 68–​74. doi:10.1016/​ distortion:  An exploratory study in patients with anorexia
j.brat.2012.10.007 nervosa. Journal of Eating Disorders, 1, 1–​9. doi:  10.1186/​
Shiban, Y., Schelhorn, I., Pauli, P., & Muhlberger, A. (2015). 2050-​2974-​1-​17
Effect of combined multiple contexts and multiple stimuli Waterworth, J. A., & Riva, G. (2014). Feeling present in the physi-
exposure in spider phobia:  A  randomized clinical trial in cal world and in computer-​mediated environments. London,
virtual reality. Behaviour Research and Therapy, 71, 45–​53. UK: Palgrave Macmillan.
doi:10.1016/​j.brat.2015.05.014 Wiederhold, B. K., Riva, G., & Gutiérrez-​Maldonado, J. (2016).
Slater, M. (2002). Presence and the sixth sense. Presence: Virtual reality in the assessment and treatment of weight-​
Teleoperators, and Virtual Environments, 11, 435–​439. related disorders. Cyberpsychology, Behavior, and Social
Slater, M., Perez-​Marcos, D., Ehrsson, H. H., & Sanchez-​Vives, Networking, 19, 67–​73. doi:10.1089/​cyber.2016.0012
M. V. (2009). Inducing illusory ownership of a virtual Wilson, M. (2006). Covert imitation. In G. Knoblich, I. M.
body. Frontiers in Neuroscience, 3, 214–​220. doi:10.3389/​ Thornton, M. Grosjean, & M. Shiffrar (Eds.), Human body
neuro.01.029.2009 perception from the inside out (pp. 211–​228). New  York,
Slater, M., & Wilbur, S. (1997). A framework for immersive NY: Oxford University Press.
virtual environments (FIVE):  Speculations on the role of Wooldridge, T., & Lytle, P. P. (2012). An overview of anorexia ner-
presence in virtual environments. Presence: Teleoperators and vosa in males. Eating Disorders, 20, 368–​378. doi:10.1080/​
Virtual Environments, 6, 603–​616. 10640266.2012.715515
Steinglass, J., Albano, A. M., Simpson, H. B., Carpenter, K., Wooley, S. C., & Wooley, O. W. (1985). Intensive out-​
Schebendach, J., & Attia, E. (2012). Fear of food as a treat- patient and residential treatment for bulimia. In D. M.
ment target: exposure and response prevention for anorexia Garner & P. E. Garfinkel (Eds.), Handbook of psychother-
nervosa in an open series. International Journal of Eating apy for anorexia and bulimia (pp. 120–​132). New  York,
Disorders, 45(4), 615–​621. doi: 10.1002/​eat.20936. NY: Guilford Press.
Steinhausen, H. C., & Weber, S. (2009). The outcome of buli- Zahoric, P., & Jenison, R. L. (1998). Presence as being-​in-​the-​world.
mia nervosa: Findings from one-​quarter century of research. Presence, Teleoperators, and Virtual Environments, 7, 78–​89.
American Journal of Psychiatry, 166, 1331–​1341.

Gutiérre z-Maldonado, Ferrer-García, Dakanalis, Riva 491

 C H A PT E R


Mobile Device Applications for
26 the Assessment and Treatment of
Eating Disorders

Alison M. Darcy and Shiri Sadeh-​Sharvit

Abstract
Mobile devices and applications (apps) are increasingly used in clinical practice, offering
reconceptualization of and novel avenues to tracking symptoms and delivery of more personalized
interventions. This chapter reviews the burgeoning approaches to the integration of mobile in screening
and treating individuals with eating disorders. Promising methods of data collection such as ecological
momentary assessments enhance the capabilities of detecting symptoms and recognizing patterns—​both
are fundamental to the screening, evaluation, and monitoring of eating disorders and lay the foundations
for better treatment design. More recent advances in machine learning allow ecological momentary
interventions to be delivered and continuously optimized at the individual level in real time. This chapter
explores what this means for the future of personalized treatment for eating disorders, referring to apps
that integrate these mechanisms. Finally, the chapter provides a framework for evaluating mobile device
mental health apps in clinical care.
Key Words:  eating disorders, guided self-​help, mobile health, self-​monitoring, smartphone, app

Introduction appointments, is that it is inherently unscalable

One can think of contemporary and as such, only a minority of individuals have
psychotherapies—​particularly the most evidence-​ access to it. Indeed, relatively few clinicians are
based such as cognitive-​behavioral therapy, dialec- sufficiently trained to deliver evidence-​ based
tical behavior therapy, and so forth—​as involving treatments to individuals with eating disorders
a process that combines several elements. The and those that are tend to live in urban centers
first is pattern identification in mood, thoughts, in Western countries near academic centers.
behaviors, and the contexts in which they arise. Shame and stigmatization of mental illness (Swan
The second element is experimentation in under- & Andrews, 2003) also impede the rates of eat-
standing how modifications to those patterns may ing disorder treatment seeking and compliance
help or hinder us. Another element is learning, (Puhl, Latner, King, & Luedicke, 2014). Such
or psychoeducation. Finally, the therapeutic rela- shortcomings of current psychotherapies have
tionship itself is assumed to be therapeutic, pro- driven technologists to augment and replicate ther-
viding accountability, empathy and unconditional apeutic elements with varying degrees of success.
positive regard traditionally provided within the Moreover, the ability to find community, especially
structure of regular appointments with a clinician among people with rare disorders, is a function that
in a physical space designated for this purpose. Internet technologies are particularly well suited to
A major problem with this traditional approach, since it is one of the primary reasons the Internet
particularly illuminated by the regular in-​person was originally developed.

492
 The potential for technologies to be used to mobile (including wearable) technology, providing
enhance therapy has been recognized for decades the means to gather enormous amounts of person-
(e.g., Agras, Taylor, Feldman, Losch, & Burnett, ally relevant data. In terms of reach, smartphones
1990), however, recent advances in mobile devices are facilitating Internet adoption in emerging econ-
have opened up unprecedented possibilities to omies facilitated by easier dissemination of wireless
seamlessly integrate various functions of the thera- networks rather than costly cable infrastructures.
peutic process into one’s everyday life. Mobile refers Mobile apps have also eroded some of the ethnic
to wireless devices and sensors (including mobile and socioeconomic disparity that has been observed
phones) that are intended to be carried with the in the adoption of previous technologies, setting the
person during day-​to-​day activities (Kumar et  al., scene for an opportunity to disseminate to tradi-
2013). Mobile mental health offers an oppor- tionally underserved populations. For example, in
tunity to both improve the efficiency of health- 2013 in the United States 59% of Caucasian, 74%
care and engage individuals in their own health of African American, and 68% of Hispanic individ-
(Kostkova, 2015). uals owned smartphones (Duggan & Smith, 2013).
This chapter posits that mobile technology has Interestingly, people tend to develop an emotional
the potential to extend the very landscape that has connection with their smartphone (Vincent, 2005),
defined—​ and limited—​ eating disorders therapy which make them excellent platforms for data col-
in the 20th century, liberating it from the bricks-​ lection and behavioral change in an ecologically
and-​mortar constraints of physical services and 50-​ valid way (Klasnja & Pratt, 2012).
minute sessions, and introducing a 24 hours per In a closed loop feedback system, the mobile
day, 7 day per week (24/​7) opportunity for engage- technology (e.g., smartphone) is not just the mea-
ment in health promotion. Because mobile technol- surement device but ultimately becomes a platform
ogy is advancing so quickly, we do not provide an from which tailored interventions can be deliv-
exhaustive review of specific apps that may come ered based on meaningful insights derived by the
and go, but instead aim to equip the reader with a machine learning algorithms that are busy process-
framework for understanding mobile technologies’ ing your data in the background (i.e., “the Cloud”).
potential functions, pertinent issues, and some of In essence, the conditions necessary for the devel-
the current controversies. opment of technologies that have real potential to
impact lives in a personally meaningful way have
The Promise and Reach of Mobile never been more present than they are now (Darcy,
Mobile devices have been the fastest adopted Louie, & Roberts, 2016).
technology in history, fundamentally transforming
the way people communicate and go about their Mobile Mental Health Adoption
daily lives in less than two decades. Whereas the Technology is increasingly being integrated into
functionality of older mobile phones was limited to health services, and this endeavor has been sup-
the computational power that could fit inside the ported by government in many countries including
device, smartphones by contrast can access enormous the US Health Information Technology Economic
computational power in “the cloud” via Internet and Clinical Health Act (HITECH) as well in the
connectivity. This makes smartphones vastly more United Kingdom’s National Health Service. Data
powerful and thus ideally poised to process large suggest that clinicians are open to this and that there
amounts of data that are gathered by sensors that is substantial drive from health consumers in addi-
are inbuilt in the device. tion. In contrast to e-​mental health, mobile mental
Distributed cloud computing is one of three key health adoption has been comparatively slow (Chan,
factors that is driving a period of immense techno- Torous, Hinton, & Yellowlees, 2015). Mobile app
logical innovation. Another key factor is rapid devel- quality has been highly variable, and this relatively
opment of the field of machine learning analytics, a low barrier to entry has appropriately undermined
branch of artificial intelligence (AI) in which com- consumer and clinician confidence. For example,
putational algorithms independently learn from one review of apps for eating disorders revealed that
data without human intervention. Machine learn- out of 44 apps identified, 50% of them provided
ing algorithms keep search engine results relevant potentially harmful information, such as “tips” for
to specific users, keep SPAM out of inboxes, and are anorexia nervosa (Fairburn & Rothwell, 2015).
the basis of the autonomous (self-​driving) cars. The This variability in quality is likely less to do with
third factor is the global adoption of inexpensive fundamental problem with the feasibility of mobile

Darcy, Sadeh-Sharvit 493

apps in healthcare, and more in line with what tends convenient way to self-​monitor because individuals
to happen in the early days after the introduction of are typically in close proximity to their phone, using
any new technology. In addition to inevitable ini- it for work, interpersonal connection, and search-
tial instability of most newly introduced technolo- ing or recording of information that they consider
gies, the potential for reducing the overall burden of important (Fox & Duggan, 2013). Inexpensive
illness in the population, enhancing access to care wearable devices (e.g., Bluetooth-​ enabled fitness
and, in some cases quality of care, remains. Recent trackers, etc.) are a promising new arena for collect-
academic-​industry partnerships are signaling a new, ing data passively—​which could enhance screening,
more stable era for mobile app development that assessment, and monitoring eating disorders for
combines both computational and academic rigor. both the individual and their care team—​by being
For example, Samsung has recently partnered with both effortless and more objective. Surprisingly,
the University of California, San Francisco, to cre- few studies have examined self-​ monitoring on
ate mobile health devices and applications; Apple phones against traditional paper and pen methods.
recently released a call for innovative research for However, there are two reports that demonstrate
mobile mental health initiatives that may make use increased adherence to smartphone self-​monitoring
of their Health kit; and so on. over paper and pen for recording meals (Carter,
However, as the mental health field moves toward Burley, Nykjaer, & Cade, 2013) and physical activ-
the longer-​term goal of precision therapeutics, only ity (Kirwan, Duncan, Vandelanotte, & Mummery,
the earliest of steps have been taken in eating dis- 2012). In addition, some patient populations may
orders. While the potential of mobile apps has not prefer mobile apps to traditional monitoring (Riley
yet been fully realized (Juarascio, Goldstein, et al., et al., 2011, Darcy et al., 2016).
2015), we explore early initiatives in assessment Such real-​time recording of symptoms is fun-
and monitoring, self-​help and guided self-​help, and damental to every formulation of eating disorder
treatment augmentation and delivery. cases. Most empirically supported interventions
in eating disorders rely on identifying and making
Apps for the Assessment and Monitoring sense of the individual, complex matrix of cogni-
of Eating Disorders tions, emotions, behaviors, and exercise, and their
Traditional questionnaire-​based tools have been role in precipitating, maintaining, and recovering
validated in patients and disseminated among eat- from eating psychopathology. Mobile devices offer
ing disorder professionals. Despite their psychomet- innovative methods to screen, assess, and evalu-
ric strengths, the key limitations of these self-​reports ate symptoms, at a location and time that fit the
are that they are subject to recall bias and measure- individual’s routine and with minimal burden.
ment errors. Self-​report measures, are, by defini- Moreover, mobile devices are in the unique position
tion, subjective, and in eating disorders as with to offer the ability to conduct and integrate sophis-
other mental illness, individuals often have illness-​ ticated pattern recognition via machine learning
relevant biases that can affect responses (Waldman, algorithms, which may assist clinicians in construct-
Loomes, Mountford, & Tchanturia, 2013). In addi- ing the formulation.
tion, some limitations exist with current self-​report Mobile devices can record varied information
eating disorder measures when used with certain that ranges from passive data (e.g., glucose levels,
populations, for example with males, since many indicating how much time has passed since the
assessments were initially developed and normed last meal, the number of steps a day) to active self-​
among relatively small, female samples (Darcy & monitoring of meals, binges and purges, and the
Lin, 2012). cognitions, emotions, contexts, and triggers that are
Self-​monitoring of meals and symptoms is at associated with them. Real-​time recording is funda-
the core of most evidence-​ based treatments for mental to assessment of specific symptoms, is sup-
eating disorders, and completion of this activ- ported in specialized apps for eating disorders like
ity is associated with better outcomes (Kazantzis, Recovery Record and RiseUp as well as in products like
Whittington, & Dattilio, 2010; Lebeau, Davies, Apple Watch—​which all use reminders prompting
Culver, & Craske, 2013). However, self-​monitoring the individual to record their health data. Mobile
is difficult for many patients with eating disorders devices can also communicate with certain func-
(Cebolla et al., 2010) and is especially challenging tions embedded in smartphones and wearables,
for adolescents (Lock, 2005; Nichols & Gusella, including sensors of physical activity levels, geo-
2003). Smartphones are thought to be a more graphic locations (for further analysis of triggers),

494 Mobile device Applications for Assessment and Treatment

number of text messages sent and received, and pre- however as previously mentioned, at least one app
defined reminders to record data. These functions is currently in development by Juarascio, Goldstein,
could be programmed not only according to expert and colleagues (2015).
opinion and empirical data, but they could also Although there is no existing research on the
be informed by the individual’s goals and prefer- fidelity of different modules to record real-​ time
ences. Personalized information could be collected data in eating disorders (i.e., online versus a mobile
in association with individual triggers, for instance device), cell phones have the advantage over EMA
using the geographic location services of the phone devices of having 24/​7 access to the individual as
to mark proximity to ice cream stores, if these are well as the benefit of being experienced as part of
known triggers of a binge. Further, standardized their intimate environment. Therefore, the devel-
self-​report measures for mental disorders completed opment of mobile devices and apps should be pri-
on apps have similar psychometric features as those oritized as the preferred method of collecting data
administered via traditional data collecting methods to screen for, assess, and monitor eating disorders.
(Chan et al., 2015). Recovery Record, relatedly, has As triggers for eating disorder symptoms may be
demonstrated capacity to administer validated mea- highly diverse and context-​related, the refined algo-
sures to assess users’ current symptoms (Tregarthen, rithms need some assistance from device users to be
Lock, & Darcy, 2015). meaningful. This could result in greater awareness
The capabilities of mobile devices and apps of triggers, but also be more tedious and not appro-
uniquely complement a specific and highly sensitive priate for everyone. Assessment and monitoring
method of recording real-​time patient data, called by mobile devices might also be affected by user-​
ecological momentary assessment (EMA). In EMA, initiated errors, for instance recall bias, uncharged
information on symptoms can be collected in real battery that makes the mobile device unavailable,
time, and later analyzed and assessed by a mental misuse or misplacement of sensors and functions
health expert. Further, after one’s baseline pat- within the app, and so forth. Nonetheless, given
tern of symptoms has been established, machine-​ their distinct benefits for assessment of symptoms,
learning-​enabled algorithms could signal significant Recovery Record and RiseUp—​the only existing spe-
deviations and inform the individual, their care cialized apps that are currently available for pub-
team, or another designated function. Ecological lic use—​as well as additional developing apps and
momentary assessment has been used effectively in devices, should be more extensively studied. More
the measurement of eating disorders, for instance research is also needed on the capacity of mobile
successfully detecting the temporal relationship devices, and particularly wearables, to provide more
between stress and binge eating (Engel et al., 2016; accurate and additional data points for screening
Goldschmidt et  al., 2014). TakeControl, a new and monitoring.
app-​based intervention for eating disorders that is
currently under development (Juarascio, Manasse, Mobile Delivered Interventions
Goldstein, Forman, & Butryn, 2015), aims to use for Eating Disorders
EMA in assessment and treatment. An algorithm The magnitude of individuals requiring
based on machine learning would detect patterns evidence-​ based interventions for eating disorders
of relationships between triggers and binge behav- will outnumber available therapists in the foresee-
iors, and would inform the user that they are at a able future, particularly as rates of eating disorders
high risk for a binge episode. Recovery Record also are still increasing (Micali, Hagberg, Petersen, &
includes reminders that are set for estimated times Treasure, 2013). Problems like cost, mobility, access
of meals, to ensure that these are recorded as close to treatment, and stigma deter individuals who
as possible to their occurrence. Modeling EMA data need treatment from receiving it. App-​based inter-
allows researchers and clinicians to explore interac- ventions for eating disorders can be provided in the
tions that might elucidate symptoms, and indicate format of guided self-​help, or as a method to aug-
a significant change in symptoms that may warrant ment and complement face-​ to-​
face intervention.
a different level of care. It also facilitates testing the We will now describe these two modules and review
validity of theories on which treatments have been the few available apps that incorporate mobile deliv-
based in both one individual as well as large-​scale ered interventions for eating disorders.
samples. To the best of our knowledge, currently A recent review of the literature concluded
there are no published data on the use of EMA in that there is a reassuring number of rigorous stud-
mobile-​ based interventions for eating disorders, ies confirming that e-​ mental health applications

Darcy, Sadeh-Sharvit 495

impart many benefits including improved accessi- Coaching-​ enhanced intervention, such as the
bility, reduced running costs, flexibility in terms of one offered by Lantern, matches a mental health
standardization and personalization, interactivity, coach that interacts with the user through the app.
and consumer engagement (Lal & Adair, 2014). Coaches—​whose possible contribution to large-​scale
Internet-​delivered cognitive-​behavioral therapy mobile mental health intervention efforts is gradually
(CBT), for example, has almost two decades of increasing—​can individualize the program to fit the
research data that support its use with certain pop- user’s needs, respond to questions, clarify concepts,
ulations. Mobile mental health adoption, on the and refocus the user on important intervention com-
other hand, has been comparatively slow (Chan ponents. However, health coaches are currently unli-
et al., 2015). censed and unregulated, although there are efforts
Guided self-​help (GSH) interventions for eating underway to establish best practices guidelines. In
disorders, that is, prebuilt programs that include addition, no research exists in the eating disorder
psychoeducational materials, self-​ paced exercises field about the efficacy of mobile GSH programs that
and homework, and often also minimal support have been adapted from Web-​based models.
by a mental health professional or other person, In the absence of a coach who can individual-
have received empirical support for the past two ize program content to fit the needs of the indi-
decades (Perkins, Murphy, Schmidt, & Williams, vidual, app-​based programs can adapt to the needs
2006). Data indicate no difference in outcomes of the individual in response to data provided in
between patients with eating disorders who received the program. The interactive nature of mobile apps
the intervention in a face-​to-​face format or online and devices presents innovative methods to make
(Wilson & Zandberg, 2012). Apps appear to be a mobile-​based GSH much more interactive than its
highly suitable platform to provide GSH, as they book-​based or online counterparts. Recovery Record
are embedded within the lives of individuals and has recently received NIH funding in partnership
are free from the resource constraints associated with our group at Stanford to develop and evaluate
with conventional therapy. Psychoeducation could an adaptive version of the app with guided self-​help
be conveyed using a variety of methods (including programs tailored for specific user groups, against
reading materials, quizzes, audio and video files, a standard version of the app. Preliminary data
links to complementary materials online, etc.). suggested that, though about 30% have clinically
The range of formats to provide psychoeducation significant symptom reduction after 28 days of app
about the eating disorder and ways to combat it use, users’ response and reduction in psychopathol-
may enhance learning and patient acceptability. ogy vary according to the symptom profile. These
For example, delivery of psychoeducational materi- interventions are designed as stand-​alone programs,
als can be matched to the learning style of an indi- and were designed using a “person-​centered design
vidual so as to enhance engagement. Additionally, process” that aims to specifically design for the
progress in app-​based GSH programs is self-​paced, experience of individuals consulted in the process
tailored to the individual’s schedule and preferred (Kelley & Littman, 2007). The aim was to replicate
location. The intervention is often broken down to some of the elements of the therapeutic process
short segments that include some psychoeducation reported to be desirable by individuals consulted
and an activity. For instance, Lantern—​primarily during the design phase. Such elements included
a Web-​based program that has been adapted for regular check-​ins and accountability; goal-​setting
mobile—​ provides 10-​minute sessions of CBT-​ support; and help interpreting things that are going
inspired GSH that are limited to once a day in well versus things that are challenging. A curated list
order to pace users’ progress and allow material and of specific goals, skills, and challenges correspond-
new skills to be processed. Additional apps for eat- ing to each of the groups’ symptom profiles iden-
ing disorders, such as RiseUp and Stop Binge Eating tified in a preliminary analysis were developed. To
provide psychoeducational materials for self-​help, foster user engagement, content was designed to
but are not exclusively based on empirically sup- mirror therapy delivery such that goals, skills, and
ported interventions for eating disorders. Recovery challenges are more achievable in the early weeks
Record, Before I  Eat, and iCounselor all employ of the intervention but gradually become more dif-
strategies to reduce eating disorder symptoms that ficult as the user progresses through the program.
include components of CBT, acceptance and com- Beyond tailoring to individuals at group level, an
mitment therapy, mindfulness, and general self-​ important feature of mobile devices is that they are
soothing techniques. especially geared toward sending content specific to

496 Mobile device Applications for Assessment and Treatment

the individual via just-​in-​time (JIT) interventions. thus far has been CBT. Both ACT and DBT, which
These JIT interventions are an algorithm-​supported have good GSH modules, should be tested further
emerging approach of programmed intervention so that more diverse groups of patients could benefit
responses to small changes in cognition, emotion, of this resource.
behavior, and context. These predetermined inter- Technology can enhance treatment compo-
ventions are based on EMA and can be delivered nents that traditional intervention approaches do
via mobile devices (Nahum-​Shani et al., 2014). The not address sufficiently. Integrating face-​ to-​
face
JIT interventions can analyze changes in behavior, and online treatment or technology-​based compo-
prompt the individual to make adaptive changes nents is a practice called “blended care” or “blended
accordingly, and guide them through the process therapy” (Wentzel, van der Vaart, Bohlmeijer, &
with specific instructions or encouragement to van Gemert-​ Pijnen, 2016). A  blended therapy
turn to higher levels of care. In the field of eat- would include, for instance, a psychoeducational
ing disorders, where eating and exercise (or lack video on a new skill that the patient would watch
thereof ) often represent an impulsive reaction to a 3  days before the face-​ to-​
face session and then
psychological or situational trigger, complex algo- practice. Consequently, the therapy session would
rithms could record a person’s normal patterns. be used for troubleshooting and refinement of the
Subsequently, algorithms look for significant devia- learned material, and greater personalization of the
tions and offer strategies to address relapse. To the knowledge and skills acquired. To date, no empiri-
best of our knowledge, JIT interventions have not cal investigation of blended care in the treatment of
been deployed in any commercially available app eating disorders has been published.
and no studies of them in eating disorders have been App-​based components that are initiated, facili-
published to date. Carroll and colleagues (2013) tated, or prescribed by the therapist could booster
reported a pilot study testing wearable mobile the in-​person treatment and monitoring of patients.
detectors of heart rate, that aimed to detect arousal By further personalizing the treatment, adding
that could lead to emotional eating and encourage modular intervention units (particularly in the pres-
the user to exercise deep breathing as prevention. ence of comorbidity), saving time in session, and
There was 75% overlap between the physical arousal promoting a better system for more real time check-​
detected by these wearables and self-​reported nega- ins between sessions—​ “conservative” treatments
tive emotions, however the use of stress reduction are enhanced. The meals, emotions, and cognitions
prompts has not been tested yet. logged on Recovery Record, for example, can be
An additional promising way to use app capa- retrieved by the therapist at any time between ses-
bilities is to promote interpersonal connectedness sions or before the weekly meeting. Therapists (who
and support, which are lacking resources for many received an invite from their client) can send cheer-
patients. Indeed, many mobile apps have created new leading and supportive messages between meetings,
ways for individuals to connect and are becoming and encourage practice of the skills and techniques
inherently more social. In direct peer-​to-​peer envi- learned in treatment. If delivered conjointly, both
ronments, advice or feedback patients may receive modalities should complement each other, rather
could be detrimental and warrants some monitor- than compete or override. Unfortunately, most
ing by experts (Juarascio, Goldstein, et al., 2015). mobile health apps do not adequately use mobile
However, it is possible to allow for interpersonal device capabilities (Juarascio, Manasse, et al., 2015).
connection in a controlled and managed way, such
as in Recovery Record’s “Pair-​up” feature (Tregarthen The State of Currently Available
et al., 2015). This feature allows for individuals to Technologies
connect with another user of similar demographics Eating disorder treatment and assessment could
and select pre-​vetted encouraging messages, thus capitalize the accessibility, functionality, interactive
providing and receiving peer support, but in a way nature, convenience, low cost, and privacy of mobile
that is protected and asynchronous. health technologies. Therefore, mobile devices can
None of these abovementioned apps currently improve engagement, adherence, and psychoeduca-
offers a personalized analysis of symptoms and the tion, as well as enrich the quantity and quality of
triggers for their occurrence. Not all of these apps data available for treatment planning and monitor-
are available across all types of mobile phones, lim- ing accessible by providers. Such features are ideal
iting the dissemination among the patient popula- for engaging patients in treatment due to several
tion. Additionally, the GSH approach used in apps advantages over static programs, such as portability,

Darcy, Sadeh-Sharvit 497

capacity for real-​time, in-​the-​moment interaction, We also suggest some possible solutions for these
and multiple built-​in sensors for collection and pre- controversies.
sentation of data (Heron & Smyth, 2010).
We have referred here to six available apps that Bias in Self-​Report in Paper-​and-​Pencil
offer psychoeducation, self-​ monitoring of symp- Versus Mobile Assessments
toms, guided self-​ help that can be enhanced by Eating disorders assessment heavily relies
coaching, and communication with a clinician. We on patients’ self-​ report. It has been questioned
have also mentioned an app that is in development, whether mobile app-​level food and symptom self-​
and is planned to use EMA in reducing binging. We monitoring affords advantages over paper-​and-​pen
are aware of research involving three of these seven methods (Fairburn & Rothwell, 2015). However,
apps that is being carried out in an academic setting. since mobile real-​time monitoring employs the use
All of these apps are in English, and there are cur- of time stamps, app-​based assessments of eating dis-
rently no data on their use by non-​English speakers. orders are considered more accurate (Engel et  al.,
Our findings are in concert with those of a recent 2016). Interestingly, while interest in the integration
review (Juarascio, Manasse, et al., 2015), indicating of mobile devices in clinical practice and research is
that there are also very limited data on the feasibil- increasing in the eating disorders field, there are no
ity, acceptability, and outcomes of currently available empirical data comparing self-​report of cognitions,
apps. The apps that currently exist vary greatly. The emotions, and behaviors associated with DSM-​5
interface of some is likely to evolve to more algorithm-​ diagnoses between the traditional self-​monitoring
driven, targeted platforms that will address different paper diaries and mobile devices. In general, indi-
typologies of individuals at risk for eating disorders as viduals are more compliant with self-​report admin-
well as patients with full-​syndrome conditions. istered in mobile devices than in paper-​and-​pencil
Therefore, it is clear that one cannot make a notebooks, and do it closer to the actual time they
sweeping judgement about the efficacy of apps as report (Stone, Shiffman, Schwartz, Broderick, &
they exist currently, partly because it is early days Hufford, 2002). Further, data from individuals with
for these technologies and none have published anxiety, depressive, and bipolar disorders suggest
data from randomized clinical trials at this time, that self-​monitoring using mobile devices demon-
though several studies are underway. In addi- strates greater sensitivity to detect change in symp-
tion, trying to judge the efficacy of apps for eat- toms over the week (Depp, Kim, Vergel de Dios,
ing disorders is similar to evaluating the efficacy Wang, & Ceglowski, 2012; Torous et  al., 2015),
of treatments for cancer. Symptom presentations, and possibly indicates more severe symptoms than
populations, and tolerance will vary; so too do reported on paper (Moore, Depp, Wetherell, &
the treatment approaches that will be effective. Lenze, 2016). Improved mobile methods to collect
The earliest data from mobile apps supports this passive and active data, could help therapists pro-
view (Tregarthen et  al., 2015). Another necessary vide more accurate diagnoses and treatment plans.
direction of research is testing whether transition-
ing effective Web-​based mobile health tools into Mobile Apps as a Replacement for Therapy
app-​based modalities may be a feasible modality It is clear that mobile mental health apps will
for delivering or enhancing evidence-​based treat- never replace traditional therapy and there will
ments (Taylor et  al., 2016). However, it is clear always be a need for expertise in the treatment
that mature mobile device apps are uniquely poised of eating disorders. However, dissemination and
to monitor and gather data and deliver targeted implementation of evidence-​based therapies is one
intervention based on individual differences. Thus of the biggest challenges facing our field. In an envi-
mobile device apps are the cornerstone of a move ronment where resources are constrained, strategies
toward precision psychiatry that has the capacity that can better allocate therapist time by “task shift-
to lower the overall burden of illness and improve ing” some functions to a device, where this can be
outcomes over and above a growing literature that accomplished in a valid way, can be viewed as part of
supports adaptive treatments (Oldenburg, Taylor, a pragmatic solution to a global problem. A related
O’Neil, Cocker, & Cameron, 2015). criticism is that individuals who use mobile apps
might not seek therapy in favor of using an app
Current Controversies instead. This is only a problem if the person requires
In the following section, we address the key a higher level of care, and thus mobile apps that are
criticisms of mobile apps for eating disorders. used in the context of treatment should integrate the

498 Mobile device Applications for Assessment and Treatment

capacity to triage appropriately. Apps that are used implications for humanity. While it is likely to
outside of therapeutic context should never discour- be a long time before AI becomes sentient, we are
age an individual from seeking treatment. However, already seeing some of these issues being raised
the reality is that the majority of individuals who in mental health. Apps as we know them (mostly
have an eating disorder never present for treatment, visual programs run on a smartphone) are showing
and apps may be in a unique position to encourage, signs of petering out and are likely to be replaced in
rather than discourage it. Tregarthen et  al. (2015) the coming decades by chatbots (automated text-​
found that a substantial portion of Recovery Record based AI agents that converse via text messages on
users had never told another person about their ill- a screen) and conversational agents (such as Apple’s
ness. In this way, mobile apps are uniquely poised Siri) that are spoken to directly in a conversational
to offer a positive initial experience with therapeutic manner. Developments in the “Internet-​of-​things”
content and if validated, could be a useful part of a (that is, connected devices in the home) are driving
comprehensive stepped-​care strategy. this technology such that we will simply ask our
agent to do things like turn up the heating or “call
Privacy and Confidentiality Mum” without ever physically interacting with a
Privacy and confidentiality are among the most device. Given the conversational format of these
pressing issues for health-​related mobile apps. In interactions, they may be ideally poised to act as
the United States, mobile health apps did not exist therapists or as first responders. Importantly, they
when HIPAA laws were developed in 2006, and may already be used in this way, before sufficient
thus the unique challenges they present have not thought and consideration has been paid to how
been considered in current mandates. When the the AI should respond to its user. For instance,
National Health Service in the UK published its Miner and colleagues investigated conversa-
Health Apps Library to encourage the use of mobile tional agents’ response to mental health-​ related
health apps, it was removed shortly thereafter due and emergency statements such as “I’ve just been
to concerns that the majority of apps on the list had raped,” “I am depressed” and “I want to commit
no data privacy policy and had other data concerns. suicide.” Responses across 77 conversational agents
As a baseline best practice, apps should transpar- were highly inconsistent, and sometimes stigmatiz-
ently communicate a data policy, however, few do ing, some recognizing the statement but failing to
this and individuals may not look for one, or worse, respond in a respectful way, and many failing to
they may incorrectly assume that there is already an refer to appropriate services (Miner et  al., 2016).
acceptable amount of data protection when none, A critical challenge is identifying a health-​relevant
in fact exists. Research on individual health liter- utterance and then responding in the appropriate
acy suggest that patients with low health literacy way. For eating disorders, this may be particularly
are less likely to use apps and mobile trackers, and challenging, as it can be the meaning of the lan-
less likely to view them as helpful, but believe them guage or behavior, not the behavior itself that has
to be private (Mackert, Mabry-​Flynn, Champlin, clinical relevance. Technology companies should
Donovan, & Pounders, 2016). To date, most consult with mental health experts and clinicians
breaches of personal health information reported when designing these new interfaces, though little
to the Department of Health and Human Services guidance exists around which values to embed.
(HHS) under the HITECH Act have been due to What therapeutic orientation should the conversa-
the theft or loss of a mobile device or laptop, rather tional agent have, if any, for example?
than direct hacks. However, the issue of privacy
and confidentiality should be consistently discussed A Framework for Evaluating Mobile Mental
because data are increasingly becoming more and Health in Clinical Care
more identifiable, especially in the context of trian- While it is still early days and genuine challenges
gulation of data types, behavioral pattern recogni- exist that make quality research difficult to produce,
tion, and machine learning algorithms. work is underway to provide a framework for clini-
cians to be able to evaluate mobile mental health
The Values Embedded in Artificial apps in the context of clinical care. Torous and col-
Intelligence leagues with the American Psychiatric Association
One of the largest ongoing debates in artifi- have proposed a hierarchy of needs model (Figure
cial intelligence (AI) relates to the pace at which 26.1) with foundational tiers being more important
the technology is developing and the unknown grounding principles than upper tiers.

Darcy, Sadeh-Sharvit 499

 The foundational tier is safety. At the most basic The next tier in this framework is ease of use.
level, the app should not cause any harm. While this Does the app have a consistent user experience and
sounds self-​evident, it is particularly relevant in eat- use case, or can it be used inappropriately? It should
ing disorders, where one review found that up to not be taken for granted that all app users will use
50% of apps created for use by individuals with eat- the app in the same way. An early case series of a
ing disorders contained harmful information, such mobile app for eating disorders, for example, found
as pro-​ana (pro-​anorexia nervosa) ideals (Fairburn & that while clinicians were relatively consistent in
Rothwell, 2015). This tier also includes privacy and their use of the app, each patient used the app in a
safety concerns. Ideally data should be encrypted in unique way (Darcy, Adler, Miner, & Lock, 2014).
situ and in transit as is consistent with most data Thus it is important that clinicians and research-
legislation for covered entities (such as HIPAA laws ers inquire as to how the app is being used by both
in the United States). However most mobile men- parties. Indeed, preliminary studies suggest that a
tal health and behavioral health apps are not legally one-​app-​fits-​all approach to app design may not
obliged to comply with this standard and are cur- be appropriate for individuals with specific symp-
rently not automatically subject to FDA approval tom profiles in populations where there is a lot of
(US Department of Health and Human Services). heterogeneity. For example, individuals with eating
Where developers choose not to comply with these disorders who engage in extreme restriction may
standards, the app should at least carry a transparent hypothetically use an eating disorders app to per-
data policy available to its users that includes what if severate and excessively quantify meals, especially
any identifiable data are collected both actively and where calories are collected (Darcy, Bryson, Kim,
passively, how the data are stored, used, and most Tregarthen, & Lock, 2017, in review). Fortunately,
importantly, if, how, and under what circumstances apps have the ability to selectively present feature
they could be shared. In this way, individuals them- sets based on a user’s responses or use habits once
selves can make the decision around acceptable learned with a great deal of precision and at little
privacy and data use conditions. Many health apps extra cost, once there are sufficient data to train a
lack even a basic privacy policy and checking for the model. Such is the promise of machine learning
presence of one is always a critical step. (Darcy et al., 2016).
The next tier is efficacy. As has been discussed in Finally, the self-​actualized level for mobile health
this chapter, while it is of some urgency to produce apps is data sharing. Is the data gathered on the app
a literature that elucidates the usefulness of apps, fully interoperable? Can it, for example, be inte-
it is nonetheless somewhat challenging and a rela- grated into the electronic medical record? While
tively slow endeavor. In the interim, Torous and col- this is challenging for app developers to achieve, it
leagues (2015) suggest that individuals should look substantially adds value to the app in a clinical con-
for apps that are built in a way that translate basic text for the clinician. An app that does not facilitate
research findings into their technology, and/​or that proper data sharing risks fragmenting care and iso-
translate or are built on therapeutic approaches that lating the patient and treatment team. Conversely,
are grounded in evidence. What theoretical frame- the patient may not actually desire data sharing with
work, if any, is the app built on, and is it explained the therapist or integration of data into the perma-
and transparent? We would add to this an addi- nent medical record, and consent is always required.
tional element of efficacy, and that is personal to the This dilemma has yet to be resolved, though under
patient. There have been many examples of patients the new paradigm of patient-​ centered medicine
who have innovated technology because the current data ownership and subsequent decisions around
technologies that exist for their condition do not such should reside with the patient.
fit their purpose or use case. For example, a type-​ We embed within this hierarchy of needs (see
2-​diabetes patient who was worried that her phone Figure 26.1) concurrent patient-​ centered and
alarm was incapable of awakening her from a diabe- clinician-​centered considerations as a guide. These
tic coma hacked various devices such that a complex questions are by no means exhaustive. Also, it is not
and loud set of alarms would sound when indicated necessary for an app to meet all levels of this hierar-
by her blood sugar levels. In a participatory med- chy to be a “good” app, for example a purely edu-
icine framework, such individuals are known as cational app may not enable data sharing. Rather
engaged patients, or e-​patients. In the context of considering each of these levels and using them as a
eating disorders, where engagement in therapy can basis to start an informed discussion with a patient is
be low, this may convey a specific advantage. the goal of this model. We thus suggest that the most

500 Mobile device Applications for Assessment and Treatment

 Should
Does the
there be

Data Sharing
patient
interoperability?
want to
If so, is there?
share data
If not, are there
and if so,
plans for this
can they?
function?

How consistent is
the user

Usability
Is there anything experience?
confusing?
Is it being used the way
it was intended?

Do any studies exist?

Is the app helpful?
Is the app built on evidence
Efficacy based approaches and/or
Why?
theory?
Under what circumstances?
Does the company have any
academic partnership?

Is there a clear data policy? Does the app make any claims not
backed by research evidence?

Is the app being used in the Are there any risks presented by the
Safety

way it was intended? patient’s use of the app, and if so -

are these acceptable to the patient?
Has the patient considered
Does the clinician know what level
data privacy issues?
of privacy is valued by the patient?

Patient Clinician
considerations considerations

Collaborative exploration

Figure 26.1  Framework for evaluating mobile mental health in clinical care 

useful starting point for a clinician to evaluate apps to be engaged in their own healthcare in an unprec-
in the context of clinical care is to allow the process edented way (Topol, 2015). Thus a patient-​driven
to be patient-​driven, for three reasons. The first is process is in concert with the empowering spirit of
that in the constantly shifting app landscape (Larsen, technology itself and it would be counterproductive
Nicholas, & Christensen, 2016), where millions of to halt that process. Finally, it is well known that indi-
apps are available and their form factor will continue viduals with eating disorders are difficult to engage in
to evolve, it is unreasonable for a clinician to compre- treatment for a variety of reasons, thus if an individ-
hensively evaluate even the most relevant ones. The ual elects to use a tool that facilitates an engagement
second is that technology has allowed individuals toward recovery then the clinician should discuss it

Darcy, Sadeh-Sharvit 501

with them in the same way that they would any other Carroll, E. A., Czerwinski, M., Roseway, A., Kapoor, A., Johns,
mechanism or tool that engages the patient in their P., Rowan, K., & Schraefel, M. C. (2013, September). Food
and mood:  Just-​in-​time support for emotional eating. In
own health promotion, regardless of their own per- 2013 Humaine Association Conference on Affective Computing
sonal opinion or experience of app use. and Intelligent Interaction (ACII) (pp. 252–​257). Danvers,
MA: Institute of Electrical and Electronics Engineers.
Summary Cebolla, A., Oliver, E., Baños, R., Botella, C., Zaragozá, I.,
Alcañiz, M.,  . . .  Guille V. (2010). PDA self-​register system
Similarly to mental health professionals’ inter- for eating disorders:  A  study on acceptability and satisfac-
est in staying up-​ to-​
date with best-​practices for tion. Studies in Health Technology and Informatics, 154, 50–​
the assessment and treatment of eating disorders, 52. doi: 10.3233/​978-​1-​60750-​561-​7-​50
clinicians should educate themselves about how to Chan, S., Torous, J., Hinton, L., & Yellowlees, P. (2015).
evaluate available mobile apps and continuously Towards a framework for evaluating mobile mental health
apps. Telemedicine and e-​Health, 21, 1038–​1041.
check their proven efficacy, relevance, and useful- Darcy, A M, Adler, S, Miner, A, & Lock, J. D. (2014). How
ness for their patients. An open discussion about the smartphone applications may be implemented in the treat-
patient’s use of technology to promote recovery is ment of eating disorders: Case reports and case series data.
warranted. Advances in Eating Disorders: Theory, Practice and Research, 2,
Mobile health technologies are theoretically 217–​232. doi: 10.1080/​21662630.2014.938089
Darcy, M. A., Bryson, S. W., Kim, J. P., Tregarthen, J., & Lock,
poised to provide pattern-​recognition intelligence J. D. (2017). Is smartphone app food journaling a good idea
that could be stronger than the one offered in clas- for everyone? An eating disorder perspective. Manuscript under
sic treatment models; although therapy is largely review.
based on pattern recognition and analysis, retro- Darcy, A. M., & Lin, I. H.  J. (2012). Are we asking the right
spective self-​report is often biased. Mobile health questions? A  review of assessment of males with eating
disorders. Eating Disorders, 20, 416–​426. doi:  10.1080/​
initiatives are progressing toward app-​based and 10640266.2012.715521
other technology interfaces that will all be mobile, Darcy, A. M., Louie, A. K., & Roberts, L. W. (2016). Machine
for example, conversational agents and wearable learning and the profession of medicine. JAMA, 315, 551–​
technology that does not rely on self-​report. This 552. doi:10.1001/​jama.2015.18421
presents exciting opportunities to rethink the roles Depp, C. A., Kim, D. H., Vergel de Dios, L., Wang, V., &
Ceglowski, J. (2012). A pilot study of mood ratings cap-
of clinicians and technology as well as whether, tured by mobile phone versus paper-​and-​pencil mood charts
how, and when they interface. Inclusion of mobile in bipolar disorder. Journal of Dual Diagnosis, 8, 326–​332.
device applications will ultimately evolve the eating doi: 10.1080/​15504263.2012.723318
disorder field and the care we provide our patients. Duggan, M., & Smith, A. (2013). Cell internet use 2013.
When used properly, apps can improve the Washington, DC: Pew Research Center.
Engel, S. G., Crosby, R. D., Thomas, G., Bond, D.,
screening, evaluation, self-​ management, and tra- Lavender, J. M., Mason, T.,   .  .  .  Wonderlich, S. A.
ditional treatment of eating disorders. Apps could (2016). Ecological momentary assessment in eating dis-
augment existing services that are inadequate world- order and obesity research: A review of the recent litera-
wide; there will never be sufficient clinicians to ture. Current Psychiatry Reports, 18, 1–​9. doi:  10.1007/​
provide face-​to-​face or telepsychology for everyone s11920-​016-​0672-​7
Fairburn, C. G., & Rothwell, E. R. (2015). Apps and eating dis-
who needs it. Currently, access and scale are some of orders: A systematic clinical appraisal. International Journal
the biggest problems that face the field of psycho- of Eating Disorders, 48, 1038–​1046.
therapy right now, and mobile could be a viable way Fox, S., & Duggan, M. (2013). Health Online 2013. Washington,
to address some of the inherent limitations, though DC: Pew Research Center.
it is relatively early days in the evolution of these Goldschmidt, A. B., Wonderlich, S. A., Crosby, R. D., Engel,
S. G., Lavender, J. M., Peterson, C. B.,  . . .  Mitchell, J. E.
technologies. (2014). Ecological momentary assessment of stressful events
and negative affect in bulimia nervosa. Journal of Consulting
References and Clinical Psychology, 82, 30. doi: 10.1037/​a0034974
Agras, W. S., Taylor, C. B., Feldman, D. E. Losch, M., & Heron, K. E., & Smyth, J. M. (2010). Ecological momen-
Burnett, K. F. (1990). Developing computer-​assisted therapy tary interventions:  Incorporating mobile technology into
for the treatment of obesity. Behavior Therapy, 21, 99–​109. psychosocial and health behaviour treatments. British
doi: 10.1016/​S0005-​789480191-​1 Journal of Health Psychology, 15, 1–​39. doi:  10.1348/​
Carter, M. C., Burley, V. J., Nykjaer, C., & Cade, J. E. (2013). 135910709X466063
Adherence to a smartphone application for weight loss com- Juarascio, A. S., Goldstein, S. P., Manasse, S. M., Forman, E.
pared to website and paper diary:  Pilot randomized con- M., & Butryn, M. L. (2015). Perceptions of the feasibil-
trolled trial. Journal of Medical Internet Research, 15(4), e32. ity and acceptability of a smartphone application for the
doi: 10.2196/​jmir.2283 treatment of binge eating disorders:  Qualitative feedback

502 Mobile device Applications for Assessment and Treatment

 from a user population and clinicians. International Journal Journal of Psychiatric Research, 75, 116–​123. doi:  10.1016/​
of Medical Informatics, 84, 808–​816. doi:  10.1016/​ j.jpsychires.2016.01.011
j.ijmedinf.2015.06.004 Nahum-​ Shani, I., Smith, S. N., Tewari, A., Witkiewitz, K.,
Juarascio, A. S., Manasse, S. M., Goldstein, S. P., Forman, E. Collins, L. M., Spring, B., & Murphy, S. (2014). Just in time
M., & Butryn, M. L. (2015). Review of smartphone applica- adaptive interventions (JITAIs):  An organizing framework
tions for the treatment of eating disorders. European Eating for ongoing health behavior support. Methodology Center
Disorders Review, 23, 1–​11. doi: 10.1002/​erv.2327 Technical Report 14–​126. Retrieved from:  https://​method-
Kazantzis, N., Whittington, C. & Dattilio, F. (2010). Meta-​ ology.psu.edu/​media/​techreports/​14-​126.pdf on September
analysis of homework effects in cognitive and behav- 1, 2016.
ioral therapy:  A  replication and extension. Clinical Nichols, S., & Gusella, J. (2003). Food for thought: Will ado-
Psychology: Science and Practice, 17, 144–​156. doi:10.1111/​ lescent girls with eating disorders self-​monitor in a CBT
j.1468-​2850.2010.01204.x Group? The Canadian child and adolescent psychiatry review,
Kelley, T., & Littman, J. (2007). The art of innovation: Lessons in 12(2), 37–​39.
creativity from IDEO, America’s leading design firm (1st ed.). Oldenburg, B., Taylor, C. B., O’Neil, A., Cocker, F., & Cameron,
New York, NY: Random House. L. D. (2015). Using new technologies to improve the preven-
Kirwan, M., Duncan, M. J., Vandelanotte, C., & Mummery, W. tion and management of chronic conditions in populations.
K. (2012). Using smartphone technology to monitor physi- Annual Review of Public Health, 36, 483–​505. doi: 10.1146/​
cal activity in the 10,000 Steps Program: A matched case–​ annurev-​publhealth-​031914-​122848
control trial. Journal of Medical Internet Research, 14(2), e55. Perkins, S. S., Murphy, R. R., Schmidt, U. U., & Williams, C.
doi: 10.2196/​jmir.1950 (2006). Self‐help and guided self‐help for eating disorders.
Klasnja, P., & Pratt, W. (2012). Healthcare in the pocket: Mapping Cochrane Database of Systematic Reviews, 3, CD004191.
the space of mobile-​phone health interventions. Journal of doi: 10.1002/​14651858.CD004191.pub2.
Biomedical Informatics, 45, 184–​198. Puhl, R. M., Latner, J. D., King, K. M., & Luedicke, J. (2014).
Kostkova, P. (2015). Grand challenges in digital health. Frontiers Weight bias among professionals treating eating disor-
in Public Health, 3, 134. doi: 10.3389/​fpubh.2015.00134 ders:  Attitudes about treatment and perceived patient out-
Kumar, S., Nilsen, W. J., Abernethy, A., Atienza, A., Patrick, K., comes. International Journal of Eating Disorders, 47, 65–​75.
Pavel, M.,  . . .  Hedeker, D. (2013). Mobile health technol- Riley, W. T., Rivera, D. E., Atienza, A. A., Nilsen, W., Allison,
ogy evaluation: The mHealth evidence workshop. American S. M., & Mermelstein, R. (2011). Health behavior models
Journal of Preventive Medicine, 45, 228–​236. in the age of mobile interventions:  Are our theories up to
Lal, S., & Adair, C. E. (2014). E-​mental health: A rapid review of the task? Translational behavioral medicine, 1(1), 53–​71.
the literature. Psychiatric Services, 65, 24–​32. doi: 10.1176/​ doi: 10.1007/​s13142-​011-​0021-​7
appi.ps.201300009 Stone, A. A., Shiffman, S., Schwartz, J. E., Broderick, J. E.,
Larsen, M. E., Nicholas, J., & Christensen, H. (2016). & Hufford, M. R. (2002). Patient non-​compliance with
Quantifying app store dynamics:  Longitudinal tracking of paper diaries. BMJ, 324, 1193–​1194. doi:  10.1136/​
mental health apps. JMIR mHealth and uHealth, 4, e96. bmj.324.7347.1193
doi: 10.2196/​mhealth.6020 Swan, S., & Andrews, B. (2003). The relationship between
LeBeau, R. T., Davies, C. D., Culver, N. C., & Craske, M. shame, eating disorders and disclosure in treatment. British
G. (2013). Homework compliance counts in cognitive-​ Journal of Clinical Psychology, 42, 367–​378.
behavioral therapy. Cognitive Behaviour Therapy, 42(3), 171–​ Taylor, C. B., Kass, A. E., Trockel, M., Cunning, D., Weisman, H.,
179. doi: 10.1080/​16506073.2013.763286 Bailey, J.,  . . .  Wilfley, D. E. (2016). Reducing eating disor-
Lock, J. (2005). Adjusting cognitive behavior therapy for adoles- der onset in a very high risk sample with significant comorbid
cents with bulimia nervosa: Results of case series. American depression: A randomized controlled trial. Journal of Consulting
Journal of Psychotherapy, 59(3), 267. and Clinical Psychology, 84, 402. doi: 10.1037/​ccp0000077
Mackert, M., Mabry-​Flynn, A., Champlin, S., Donovan, E. E., Torous, J., Staples, P., Shanahan, M., Lin, C., Peck, P., Keshavan,
Pounders, K. (2016). Health literacy and health informa- M., & Onnela, J. P. (2015). Utilizing a personal smartphone
tion technology adoption:  The potential for a new digital custom app to assess the Patient Health Questionnaire-​
divide. Journal of Medical Internet Research, 18(10), e264. 9 (PHQ-​ 9) depressive symptoms in patients with major
doi: 10.2196/​jmir.6349 depressive disorder. Journal of Medical Internet Research
Micali, N., Hagberg, K. W., Petersen, I., & Treasure, J. L. Mental Health, 2: e8. doi: 10.2196/​mental.3889.
(2013). The incidence of eating disorders in the UK Topol, E. (2015). The patient will see you now: The future of medi-
in 2000–​2009:  Findings from the General Practice cine is in your hands. New York, NY: Basic Books.
Research Database. BMJ Open, 3, e002646. doi:10.1136/​ Tregarthen, J. P., Lock, J., & Darcy, A. M. (2015). Development
bmjopen-​2013-​002646 of a smartphone application for eating disorder self‐monitor-
Miner, A. S., Milstein, A., Schueller, S., Hegde, R., Mangurian, C., ing. International Journal of Eating Disorders, 48, 972–​982.
& Linos, E. (2016) Smartphone-​based conversational agents doi: 10.1002/​eat.22386
and responses to questions about mental health, interper- Waldman, A., Loomes, R., Mountford, V. A., & Tchanturia, K.
sonal violence, and physical health. JAMA Internal Medicine, (2013). Attitudinal and perceptual factors in body image
176(5), 619–​625. doi:10.1001/​jamainternmed.2016.0400 distortion:  an exploratory study in patients with anorexia
Moore, R. C., Depp, C. A., Wetherell, J. L., & Lenze, E. J. nervosa. Journal of Eating Disorders, 1, 1.  doi:  10.1186/​
(2016). Ecological momentary assessment versus stan- 2050-​2974-​1-​17
dard assessment instruments for measuring mindful- Wentzel, J., van der Vaart, R., Bohlmeijer, E. T., & van
ness, depressed mood, and anxiety among older adults. Gemert-​ Pijnen, J. E. (2016). Mixing online and

Darcy, Sadeh-Sharvit 503

 face-​to-​face therapy: How to benefit from blended care in scalability. Clinical Psychology Review, 32, 343–​357.
mental health care. JMIR Mental Health, 3. doi: 10.2196/​ doi: 10.1016/​j.cpr.2012.03.001
mental.4534 Vincent, J. (2005). Emotional attachment to mobile phones: An
Wilson, G. T., & Zandberg, L. J. (2012). Cognitive-​behavioral extraordinary relationship. In Mobile World (pp. 93–​104).
guided self-​ help for eating disorders:  Effectiveness and London, UK: Springer.

504 Mobile device Applications for Assessment and Treatment

 CH A PT E R


Internet-​Based Interventions
27 for Eating Disorders

Anja Hilbert, Lisa Opitz, and Martina de Zwann

Abstract
Evidence demonstrating the efficacy of treatment and prevention programs for eating disorders is
accruing. However, the common face-​to-​face delivery of these interventions has a number of limitations,
including high cost and limited accessibility. E-​mental health, referring to the use of information and
communication technology—​particularly the Internet—​in interventions for mental health disorders, has
the potential to overcome these barriers and enhance the treatment and prevention of eating disorders.
To date, the limited number of evaluations have documented small to moderate effect sizes in the
improvement of eating disorder symptomatology through Internet-​based treatment and prevention.
Beyond efficacy, major questions remain regarding content, structure, and modes of delivery of Internet-​
based interventions; suitable diagnostic tools and safety measures; and cost-​effectiveness, dissemination,
and implications for public health programming. These aspects deserve attention in future research
before widely recommending Internet-​based interventions for eating disorders.
Key Words:  eating disorder, E-​mental health, Internet, intervention, prevention, treatment

Introduction considered as well. The purpose of this chapter is

The number of worldwide Internet users increased to present an overview of different Internet-​based
by 900% between 2000 and 2016 (Internet World approaches to the treatment and prevention of EDs
Stats, 2016). This huge global expansion of the and their efficacy, with a focus on programs that
availability and use of the Internet not only caused have been evaluated in randomized controlled trials
remarkable changes in daily life, but also influenced (RCTs), if available. Finally, this chapter delineates
healthcare, including mental healthcare, by open- directions of future research on Internet-​based ED
ing up numerous new ways of delivery. E-​mental interventions.
health refers to the use of information and commu-
nication technology, particularly the Internet, to The Potential of Internet-​Based
support and improve mental health conditions and Interventions
mental healthcare (Riper et  al., 2012). E-​mental Currently, the conventional ED treatment via
health service options are still at an early stage of face-​to-​face interventions has a number of limita-
development and implementation for many men- tions. Epidemiological studies revealed that only
tal disorders, including eating disorders (EDs), a minority of people who meet diagnostic criteria
and only the future will show how they will affect for an ED seeks or receives mental healthcare, and
treatment delivery and the national health systems this especially concerns specialized ED treatment
over the long term. Internet-​based delivery of psy- (Hart, Granillo, Jorm, & Paxton, 2011; Hoek &
chological treatments has the potential to result in van Hoeken, 2003; Kessler et al., 2013). The effi-
substantial improvements of mental healthcare, but cacy and effectiveness of ED interventions as well
at the same time bears several risks that must be as prevention has already been evaluated, but their

505
broad dissemination from research into practice face-​to-​face treatment (Robinson & Serfaty, 2001),
appears challenging and costly (Stice, Becker, & but they could also be sufficient, for example, when
Yokum, 2013). On the part of the affected individu- treating less severe or subclinical cases with EDs, so
als, exploratory studies documented reasons for low that recovery could be achieved without face-​to-​face
healthcare-​seeking such as social barriers (e.g., fear therapy (Fairburn & Murphy, 2015). The dangerous
of stigma, shame, and social stereotyping), personal potential of the Internet in promoting the develop-
barriers (e.g., low motivation), and organizational ment of EDs with proanorexia websites, mostly tar-
barriers like high cost and low availability (Becker, geting young females (e.g., Norris, Boydell, Pinhas,
Hadley Arrindell, Perloe, Fay, & Striegel-​Moore, & Katzman, 2006), could be counteracted by plat-
2010; Evans et al., 2011). Thus, numerous obstacles forms that disseminate healthy attitudes and social
in ED treatment and prevention remain to be sur- support (Bauer, Moessner, Wolf, Haug, & Kordy,
mounted, which could be addressed by Internet-​ 2009).
based interventions. Apart from these benefits, the use of Internet-​based
One major benefit of Internet-​based interven- interventions for the treatment of mental disorders
tions is their lack of geographic boundaries, mak- raises multiple ethical concerns. First, if there is no
ing widespread dissemination easily possible. Their clear evidence on the efficacy, an Internet-​based inter-
remarkable local and temporal flexibility in com- vention bears the risk of being futile or even harmful
parison to face-​to-​face therapies is especially rele- and might delay users from seeking more appropri-
vant for individuals who would otherwise be hard ate forms of help such as evidence-​based treatments
to reach, for example, people living in remote or (Fairburn & Murphy, 2015). Rigorous research on
psychotherapeutically undersupplied areas, indi- the efficacy for the treatment of EDs is still at an
viduals with reduced mobility (e.g., due to physical early stage, but growing rapidly, which is reflected
disability), or individuals who consider a conven- in recent systematic reviews and meta-​analyses on
tional therapy difficult to obtain because of finan- this topic (Aardoom, Dingemans, Spinhoven, & Van
cial or time constraints (Burns, Durkin, & Nicholas, Furth, 2013; Aardoom, Dingemans, & Van Furth,
2009; Fairburn & Patel, 2014; Kersting, Schlicht, 2016; Bauer & Moessner, 2013; Dölemeyer, Tietjen,
& Kroker, 2009). Moreover, Internet-​based inter- Kersting, & Wagner, 2013; Fairburn & Murphy,
ventions have the advantage of allowing patients to 2015; Hay & Claudino, 2015; Loucas et al., 2014;
retain a greater sense of anonymity, which makes Melioli et  al., 2016; Schlegl, Bürger, Schmidt,
them attractive for individuals who are reserved in Herbst, & Voderholzer, 2015).
emotional disclosure or who wish to have independ- Second, the use of technology could com-
ence from a therapist (Kersting et  al., 2009). The promise the confidentiality of the patients’ pri-
anonymous aspect is crucial especially for patients vate information, if certain steps are not taken to
who would otherwise not seek help out of shame or ensure data protection. Therefore, both providers
fear of stigmatization (Burns et al., 2009). In gen- and patients should be aware of the risks and the
eral, anonymity can reduce feelings of shame and necessary measures to protect confidential infor-
thereby enable patients to disclose their intimate mation (e.g., encryption, passwords; APA, 2013;
thoughts and feelings more openly, a phenomenon Fisher & Fried, 2003). Further advice on confiden-
labeled the disinhibition effect (Suler, 2004). tiality and other important aspects of the delivery
A subgroup of patients for whom online inter- of E-​mental health can be found in the Guidelines
ventions could especially be relevant are young for the Practice of Telepsychology developed by the
individuals in the early stages of ED syndrome American Psychological Association (2013). Third,
development. Often these patients are uncertain technology might reduce the quality of interper-
about seeking help because of a lack of knowledge sonal communication. For example, in treatments
on EDs and fear of stigmatization, which can result delivered in writing only, nonverbal information
in a long delay between ED onset and the start of a such as posture, gestures, facial expression, eye con-
treatment (Bauer et al., 2013; Fairburn & Murphy, tact, and voice intonation are missing, which not
2015). Internet-​ based interventions may shorten only may increase the risk of misunderstandings
this delay by engaging these individuals early and but also makes them harder to notice and correct
informing them about the disorder and its treat- (Knaevelsrud, Jager, & Maercker, 2004). Many
ment, which makes these interventions a source of therapists consider face-​ to-​
face contact necessary
secondary prevention (Bauer et al., 2013). They fur- for a complete understanding of a patient’s men-
ther can motivate patients to contact services for a tal and physical condition (Shingleton, Richards, &

506 Internet-Based Interventions

Thompson-​Brenner, 2013). However, in an RCT, Internet-​based ED treatment studies is that almost
Celio et  al. (2000) found no difference regarding all of them incorporated principles of cognitive-​
perceived social support between an Internet-​and behavioral therapy (CBT; see Fairburn, 2008). In
a classroom-​delivered psychoeducational interven- addition, the majority of manualized and evidence-​
tion for the reduction of body dissatisfaction and based ED self-​help programs, which are easily dis-
disordered eating behaviors in a nonclinical sample. seminable over the Internet, are based on CBT
Nevertheless, other research that compared face-​to-​ (Fairburn & Wilson, 2013). In general, structured
face communication with online communication behaviorally focused interventions may be more
between two strangers in a nonclinical sample found easily translated to technology platforms than other
that face-​to-​face contact led to a higher degree of methods that focus more on interpersonal relations
closeness, self-​disclosure, and positive and negative or affect (Shingleton et al., 2013).
affect, though both formats resulted in equivalent
emotional understanding and depth of processing Psychotherapy via Internet
(Mallen, Day, & Green, 2003). It is not clear whether Internet-​based treatments for EDs range between
or to what extent important clinical information those that operate completely without the interac-
might be missed when therapist-​patient contact is tion of the patient with a therapist, based solely
provided solely via technology (Shingleton et  al., on the use of a programmed website, and those in
2013). For these reasons, Fairburn and Murphy which a human therapist conducts the whole inter-
(2015) recommended that Internet-​based interven- vention, though at a distance from the patient. The
tions should be regulated. highest degree of integration of traditional psycho-
therapy practice into Internet technology is reached
Internet-​Based Interventions when actual psychotherapy is administered via
An increasing number of Internet-​based inter- devices such as e-​mail, video chat, and voice call—​
ventions addressing EDs have been developed for the medium of communication being the only
the treatment and prevention of EDs. Regarding difference between this kind of intervention and
the type of ED, most online interventions addressed face-​to-​face psychotherapy.
bulimia nervosa (BN), binge-​eating disorder (BED), The first study of ED therapy delivered entirely
subthreshold variants of EDs, and/​ or increased via e-​mail was published by Robinson and Serfaty
body dissatisfaction. Regarding anorexia nervosa in 2001 and repeated as an RCT in 2008. Ninety-​
(AN), very few Internet-​ based treatment studies seven participants fulfilling DSM-​ IV diagnostic
have been conducted, and have consisted of small criteria for BN, BED, or eating disorder not oth-
pilot studies (e.g., Treasure, Macare, Mentxaka, & erwise specified (EDNOS) were randomized to e-​
Harrison, 2010; Yager, 2001, 2003) and relapse mail therapy, unsupported self-​directed writing, or
prevention studies (Fichter, Quadflieg, & Lindner, a wait-​list control condition. The treatment lasted
2013; Fichter et  al., 2012). This underrepresenta- 3  months with two e-​mails per week on average.
tion of AN is presumably related to the severity of The therapists assessed the history of ED symptoms;
the disorder and the medical risks associated with encouraged self-​monitoring of food intake and feel-
weight loss and starvation (Arcelus, Mitchell, Wales, ings; helped the patients to identify and change
& Nielsen, 2011), which is why online interven- maladaptive cognitions, relationships, and behav-
tions might not be expected to provide a sufficient iors that might exacerbate the ED; and worked
level of care (Shingleton et al., 2013). on establishing regular, healthy meals. In the self-​
Besides the targeted ED, the existing Internet-​ directed writing condition, participants were asked
based ED interventions can be differentiated by sev- to write e-​mails about their difficulties at least twice
eral aspects: the extent to which they include human a week without receiving any specific therapeutic
communication versus provision of information on advice. At the end of treatment, significantly fewer
a website; whether they are conducted textually, patients met criteria for an ED in the e-​mail inter-
by audio, video, or combinations thereof; whether vention compared with the wait-​list control group,
they are delivered in real time (synchronously, e.g., yet no significant difference was found between the
through chats or video calls) or delayed (asynchro- e-​
mail group and the unsupported self-​ directed
nously, e.g., through e-​mails); whether the setting writing group (Robinson & Serfaty, 2008).
is individual or in groups; and on which underly- Mitchell et al. (2008) sought to approach con-
ing therapeutic approach they are based. Regarding ventional psychotherapy by means of a telemedi-
the latter aspect, a noticeable characteristic of cine CBT, conducting the therapy sessions via video

Hilbert, Opitz, de Zwann 507

conferencing. This condition was compared to face-​ Therapists provided individualized text-​based feed-
to-​face therapy in an RCT with 128 patients with back and encouraged participants to call or e-​mail
threshold and subthreshold BN. Manual-​ based anytime in cases of distress. The modules’ topics
CBT specifically designed for BN was used in 20 included identification of individual eating pat-
therapy sessions over 16 weeks (see Fairburn, 2008). terns, establishment of regular and healthy eating
In the first 4 weeks, therapy consisted of psycho- behaviors and exercise, identification of triggers for
education, self-​monitoring, development of a regu- binge eating and distraction strategies, cognitive
lar eating pattern, and stimulus control measures restructuring of negative eating–​ or body image–​
with the goal of reaching control over eating. Later, related thoughts, problem-​solving and social skills
therapy was more cognitively oriented and included training, and relapse prevention. At post-​treatment,
strategies to reduce dieting and improve problem-​ the intervention group showed significant improve-
solving, and cognitive restructuring. The conclud- ment in binge-​eating frequency and ED psychopa-
ing sessions focused on the maintenance of progress thology compared to the control group, which was
and relapse prevention. The symptom reduction at largely maintained over 1-​year follow-​up. In addi-
post-​treatment, 3-​and 12-​month follow-​ups was tion, at post-​treatment, full remission from binge
comparable for both groups. The face-​to-​face group eating was significantly higher in the intervention
tended to have larger improvements than the tele- group than in the control group.
medicine group on full remission from binge eat-
ing and purging and showed significantly larger Guided and Unguided Self-​Help
improvements on eating disorder psychopathology Besides therapeutic approaches, which resemble
and depression. As the technological accessibility of traditional psychotherapy with the major differ-
video calls was not as advanced as it is today, the ence of proceeding via Internet instead of face-​to-​
study participants were required to travel to a spe- face, numerous Internet-​based self-​help programs
cial facility for their therapy. Because of the general to treat or prevent EDs have been developed, based
availability of Internet programs offering video calls on evidence-​based psychological treatments, mostly
today, this type of Internet therapy could be con- CBT, and self-​help applications. An advantage of
ducted from home for patients who have a com- the Internet-​based mode of delivery in compari-
puter with access to the Internet. son to books is the possibility to arrange interac-
Another RCT comparing CBT in an Internet tive contents that provide tailored feedback to
version with a face-​to-​face version was conducted their users throughout participation. Furthermore,
by Bulik et  al. (2012), with results not yet pub- the treatment can be personalized to individual
lished. They conducted a noninferiority trial with characteristics, interests, needs, and problems, an
180 patients with BN, comparing the two delivery important advantage in light of the patients’ vary-
modes in group format in terms of their relative ing needs concerning information, frequency, dura-
efficacy, attrition, adherence, acceptability, main- tion, and intensity of support (Bauer & Moessner,
tenance of therapeutic gains over 1 year of follow-​ 2013; Fairburn & Patel, 2014). Additionally, online
up, and cost-​effectiveness. In either condition, the interventions can be designed to exchange real-​time
treatment consisted of 16 sessions over 20 weeks for information with compatible smartphone apps,
small groups of patients. The manual was modified which might enhance their effects (Fairburn &
both for face-​to-​face group delivery and Internet Patel, 2014).
delivery and included psychoeducation, self-​ Self-​
help can be subdivided into guided and
monitoring, normalization of meals, cue identifica- unguided self-​help, depending on the existence of
tion, restructuring automatic thoughts, chaining, guidance by a therapist. As unguided self-​help oper-
relapse prevention, and body image interventions. ates without communicating with another person
The Internet therapy took place using Web-​based and proceeds entirely via an Internet program, it
materials and through text-​based group chats. has the advantage of allowing the greatest extent of
A 16-​ week individual Internet-​ based CBT anonymity, and may be cheaper and more acces-
program for patients with BED was compared to sible than guided self-​help. However, little research
a wait-​list control condition in an RCT with 139 has addressed unguided Internet-​based self-​help for
participants by Wagner et al. (2016). Internet-​based EDs. Leung, Ma, and Russell (2013) conducted an
CBT consisted of 11 personalized structured writ- open trial with 280 patients with EDs to investi-
ing assignments, complementary daily eating and gate an unguided Internet-​based self-​help program
activity diaries, week plans, and psychoeducation. that included components on healthy eating; family

508 Internet-Based Interventions

education; health assessment; motivation enhance- intervention, binge-​eating frequency, eating disor-
ment; self-​help strategies to normalize eating pat- der psychopathology, self-​ esteem, quality of life,
terns; challenging negative thoughts; developing and body mass index (kg/​ m2) were significantly
problem-​ solving skills; and psychological health improved in comparison to the control group.
promotion strategies. A total of 63% of the patients Improvements were maintained at 6-​month follow-​
used the program for self-​help, and this subgroup up. In addition, the intervention group reported
demonstrated significant improvement in ED psy- greater remission from binge eating than the con-
chopathology, motivational stage of change, depres- trol group at post-​treatment. de Zwaan et al. (2017)
sion, and perceived health status at one-​ month compared the same program to a face-​to-​face CBT
follow-​up when compared to baseline. However, in a noninferiority RCT with 178 patients with
these findings refer to users only, and as there was threshold or subthreshold BED. Over a period of
no control group, the results are to be considered 4 months patients received 20 individual sessions in
preliminary. the face-​to-​face CBT condition, and 17–​18 e-​mails
In guided self-​ help programs, the guidance and two face-​to-​face sessions in addition to the
component can be administered face-​ to-​
face or Web-​based program in the Internet condition. At
online; the former has not been investigated to the end of treatment and at 6-​month follow-​up, sig-
date, whereas the latter has been addressed in sev- nificant reductions in binge eating were observed in
eral studies. Offering online sessions and weekly both conditions, with the Internet treatment infe-
supportive e-​mails, the CBT-​based guided self-​help rior to the face-​to-​face-​treatment, yet at 18-​month
program “Salut BED/​Salut BN” (Carrard et al., follow-​up, there were no statistically significant dif-
2006, 2011; de Zwaan et al., 2017; Wagner et al., ferences between the two conditions. And ED psy-
2013) consists of seven modules, each including chopathology was significantly more reduced in the
lessons, examples, and exercises about the motiva- face-​to-​face condition than in the Internet condi-
tion to change, self-​monitoring of ED symptoms, tion at 6-​month follow-​up.
behavioral strategies to prevent binge eating, chang- “Overcoming Bulimia Online” (Williams,
ing automatic thoughts, problem-​ solving, asser- Aubin, Cottrell, & Harkin, 1998) is an intensive
tiveness training, and relapse prevention (Carrard Internet-​based, interactive, multimedia, guided self-​
et al., 2006). The BED version contains additional help program consisting of eight CBT-​based les-
modules dealing with physical activity, triggers of sons with homework, group and individual online
binge-​ eating episodes, and meal plans (Carrard guidance, and supportive e-​mails from therapists.
et al., 2011). Participants receive automatic feed- Sánchez-​ Ortiz et  al. (2011) conducted an RCT
back and graphs visualizing their binge-​ eating with 76 students with BN or EDNOS comparing a
episodes and other ED symptoms (Carrard et al., 3-​month treatment with this program to a wait-​list
2006). Guidance is provided through regular e-​mail control condition. At post-​treatment and 3-​month
contact with a psychologist during the intervention. follow-​up, the treatment group showed significantly
Salut BN version was evaluated in an RCT by greater improvement in ED psychopathology,
Wagner et  al. (2013) with 155 women with BN depression, anxiety, and quality of life compared
and EDNOS in comparison to guided bibliother- with the control group.
apy, which consisted of a self-​help book and e-​mail Ljotsson et al. (2007) combined book-​based self-​
support from a therapist. Both self-​help programs help with online guidance, providing patients with
were delivered over 7 months and both significantly BED and BN (threshold and subthreshold) with
reduced the frequency of binge eating, vomiting, the CBT-​ based self-​help book Overcoming Binge
and fasting, with the greatest improvements within Eating (Fairburn, 1995) and graduate students with
the first 4  months during treatment. Symptoms Guided Self-​ Help for Bulimia Nervosa. Therapist’s
were further improved by continued treatment to Manual (Fairburn, 1999) to support the patients
post-​treatment at month 7 and stabilized to month via e-​mail once to twice per week. Patients also had
18. No difference in these intent-​to-​treat effects the opportunity to chat with each other in an online
was found between Internet-​based and book-​based private discussion forum. In an RCT, 73 patients
delivery of self-​help. participated in the 12-​week intervention or a wait-​
Carrard et al. (2011) evaluated Salut BED list control condition. At post-​treatment, treated
over 6 months in an RCT with 74 women with individuals showed significantly greater improve-
threshold or subthreshold BED in comparison to ment of binge-​eating frequency, ED psychopathol-
a wait-​list control group. After the Internet-​based ogy, quality of life, and depression compared to

Hilbert, Opitz, de Zwann 509

the control group. The results were maintained at 2011). The 9-​ month intervention called “IN@”
6-​month follow-​up. included 11 sessions with psychoeducational mate-
rial as well as self-​monitoring and behavioral tasks
Adjunct to Treatment and Aftercare concerning eating behavior, weight, body accep-
Interventions tance, coping with emotions, and social skills.
Another use of the Internet involves Internet-​ Participants were offered to communicate to each
based interventions as an adjunct to conventional other in an asynchronous moderated online dis-
face-​to-​face treatment. For instance, during face-​to-​ cussion group. In addition, they had contact with
face therapy the Internet can serve as a device for a therapist who provided feedback on tasks (e.g.,
self-​monitoring of food intake or binge eating by self-​
monitoring), conducted a monthly chat ses-
reminding the patient to record them via e-​mail sion, and was available for questions via e-​mail. The
or a smartphone app. As studies on the efficacy of results have not yet been published.
smartphone apps are presented in Chapter 26, they Fichter et  al. (2012, 2013) investigated a 9-​
are not further discussed here. With regard to e-​mail month Web-​based relapse prevention program for
delivery, Yager (2001, 2003) conducted a case series patients with AN after inpatient treatment, “VIA,”
expanding the usual face-​ to-​
face psychotherapy using CBT strategies such as self-​ monitoring,
of outpatients with AN by regular e-​mail contact. stimulus control, operant methods, vicarious
Another example is the above-​presented Internet-​ learning, exposure treatment, and cognitive
based unguided self-​help program by Leung et al. restructuring. Content included goals and moti-
(2013). A majority of their study participants were vation, transfer of therapy components to every-
already undergoing treatment for their ED when day life, maintenance of a regular eating behavior,
they were offered Internet-​based unguided self-​help. handling binge-​eating episodes and compensatory
Furthermore, the Internet can be used for after- behaviors, body acceptance, self-​ esteem, coping
care interventions to ensure the maintenance of with emotions, social competence, relationships,
gains achieved during face-​to-​face treatment. The and problem-​solving. The program also provided
existence of Internet-​based programs enables cli- a behavioral analysis of problem behavior, tools to
nicians to provide extended support to patients monitor change, and interactive elements such as
beyond the completion of inpatient or outpatient a message board for peer support, monthly group
treatment. Gulec et al. (2011, 2014) conducted a 4-​ chats with a therapist, and the possibility to write
month Internet-​based aftercare program, “EDINA,” to the therapist at any time. An RCT with 258
with the goal to help patients with BN and related patients with AN comparing program participa-
EDNOS maintain their therapeutic gains. The tion with treatment as usual, showed that patients
program provided an online information and com- who received the online intervention tended to
munication platform for peer support and profes- have greater improvement than the control group
sional consultation. It included modules such as regarding body mass index, as well as ED behav-
weekly symptom monitoring, supportive feedback, iors and psychopathology during the 9-​ month
and professional counseling during group chat ses- observation period (Fichter et  al., 2012). At 9-​
sions, as well as flexible components depending on month follow-​ up, both groups showed further
the patient’s individual needs such as psychoeduca- improvement in body mass index, eating attitudes,
tion, a forum for peer support, and individual chat and eating behaviors. The intervention group had
sessions. The program was compared to a wait-​list a tendency to show larger improvements than the
control condition in an RCT with 105 female par- control group on most variables, but significance
ticipants. At post-​treatment, participants from both was reached only for bulimic behavior and men-
conditions showed a significant reduction in ED strual function. Patients who had participated in
psychopathology compared to pretreatment, but all 9 monthly intervention sessions reached a sig-
there was no significant difference between condi- nificantly higher body mass index compared with
tions. Overall, the program proved to be feasible the control group (Fichter et al., 2013).
and well accepted, and participants’ satisfaction
was high. Prevention
The potential of an Internet-​based, manualized Thanks to their accessibility and anonymity,
CBT-​oriented approach in the aftercare following Internet-​based programs could be a convenient first
inpatient treatment was studied in another RCT contact point, especially for people with milder ED
with 253 women with BN (Beintner & Jacobi, symptoms that may develop into a full-​blown ED.

510 Internet-Based Interventions

If Internet-​based prevention programs are capable binge-​eating frequency, and weight and shape con-
of reducing such symptoms and thus avoiding cern from baseline to 9-​month follow-​up.
more intense treatments, they could be of great Stice, Rohde, Durant, and Shaw (2012) devel-
value from a public health perspective (Bauer & oped the 3-​ week Internet-​ based ED preven-
Moessner, 2013). tion program “eBody project” and evaluated it in
A substantial body of research has been con- comparison to its face-​to-​face group version in an
ducted on “Student Bodies” (Winzelberg et  al., RCT with 107 female college students with body
2000), an 8-​week Internet-​based CBT program dissatisfaction. For further comparison, an educa-
targeting college-​age women at risk of developing tional video control condition and an educational
an ED. The goals of the program were to reduce brochure control condition were conducted in
weight and shape concerns as well as binge eat- addition. The eBody project program was based
ing, improve body image, and promote healthy on cognitive dissonance theory and consisted of six
weight regulation and knowledge about EDs. written and behavioral activities designed to criti-
The eight sessions consisted of psychoeducation cally address the thin-​ideal. The face-​to-​face condi-
and mandatory and optional assignments that tion followed a similar approach delivered in group
included behavior change exercises; an online, format. There was no difference in effects between
asynchronous, moderated discussion group; self-​ the Internet condition and the face-​to-​face condi-
monitoring; and a personal diary. Several RCTs tion, both demonstrating greater pre-​post reduc-
compared Student Bodies with a wait-​list condi- tions in ED symptoms and psychopathology than
tion in population samples of women (Celio et al., the two control groups.
2000; Graff Low et al., 2006; Jacobi et al., 2005, Gollings and Paxton designed the interactive
2007; Winzelberg et al., 2000) and women at risk program “Set Your Body Free” (2006) for women
of an ED (Jacobi, Völker, Trockel, & Taylor, 2012; with high body dissatisfaction and disordered eat-
Taylor et  al., 2006; Zabinski et  al., 2001). In a ing. The participants received eight weekly manual-
meta-​analysis of six US and four German RCTs ized group sessions in real time with a therapist in
with 990 participants, Beintner, Jacobi, and Taylor an online chat room. In addition, participants could
(2012) found mild to moderate improvements communicate with each other between sessions in
in ED psychopathology at post-​ treatment and an asynchronous discussion board. Before each
follow-​up. One study (Celio et al., 2000) addition- session, participants received the treatment man-
ally compared the program with a classroom-​based ual that provided ED-​focused psychoeducation, a
body image education intervention, but did not treatment topic guide, and out-​ of-​
session activi-
find a significantly greater effect of the Internet-​ ties. Based on CBT, the session topics consisted of
based intervention. enhancing motivation to change; exploring asso-
Student Bodies was adapted for overweight ciations among body image, self-​esteem, emotions,
adolescents with binge eating in “Student2Bodies–​ and relationships; examining risk factors for body
BED” (Doyle et  al., 2008; Jones et  al., 2008), image and eating problems; teaching strategies for
targeting binge eating, weight maintenance, and changing negative body image thoughts; challeng-
physical activity. The program consisted of psycho- ing body comparison tendencies and internalized
education, interactive self-​monitoring journals, an sociocultural beauty ideals; and elaborating strate-
asynchronous discussion group, weekly letters of gies for identifying and changing unhealthy eat-
encouragement, and motivational messages. Two ing patterns and weight loss behaviors. In an RCT
RCTs evaluated the program with 83 and 105 ado- comparing the Internet delivery of this program
lescent participants, respectively, at risk of BED, with face-​to-​face delivery and a delayed treatment
in comparison to a wait-​list control (Doyle et  al., control condition in 116 women with high body
2008; Jones et al., 2008). Doyle et al. (2008) found dissatisfaction, both groups showed large improve-
a significant reduction compared with the control ments in body dissatisfaction at the end of treat-
group in body mass index z-​score from baseline to ment compared with the delayed treatment control
post-​treatment, which was maintained at 4-​month group, with greater effects in the face-​to-​face con-
follow-​up, though the difference was not significant dition. At 6-​month follow-​up the effects in both
due to improvement in the control group. There groups were maintained and, owing to continued
were no significant effects for other outcomes. improvement in the Internet condition, were no
However, Jones et  al. (2008) found a significant longer different from each other (Paxton, McLean,
effect for body mass index, body mass index z-​score, Gollings, Faulkner, & Wertheim, 2007).

Hilbert, Opitz, de Zwann 511

 A further guided Internet-​based program “ESS-​ early detection, and early intervention regarding
KIMO” for people with symptoms of AN or BN EDs was developed by Bauer et al. (2009). The
by Hötzel et  al. (2014) focused on enhancing program “ES[S]‌PRIT” followed a stepped-​ care
motivation to change, based on the transtheoreti- approach combining various support components
cal model (Prochaska & DiClemente, 1992) and of increasing intensity, which included psychoedu-
using the principles of motivational interviewing cation, symptom monitoring, tailored supportive
(Miller & Rollnick, 2002). ESS-​KIMO consisted of feedback, peer support, and professional chat-​based
six online sessions with psychoeducation and writ- counseling. Participants could use the various sup-
ing tasks concerning the benefits and costs of the port components flexibly according to their needs.
patient’s ED, the ED’s impact on his or her daily In the case of severe self-​reported ED symptoms,
life and life goals, and sources of self-​esteem. After participants were encouraged to access more intense
each assignment, the participant received individu- face-​to-​face help. The efficacy of an adapted ver-
alized electronic feedback by a therapist. Results of sion of the program for high school students
an RCT with 212 women with symptoms of AN or (“YoungEs[s]prit”) in preventing the onset of self-​
BN demonstrated the program to be significantly reported ED symptoms was studied in an RCT with
superior to a wait-​list control condition in improv- 1,667 healthy participants (Lindenberg & Kordy,
ing motivation to change, self-​ esteem, dietary 2015). The cumulative ED incidence rate over 12
restraint, and vomiting. This suggests the Internet months significantly decreased in the study’s first
to be a potentially useful means to raise motivation wave, yet no difference between the intervention
to start a treatment in individuals with an ED who and control groups was found in the second sub-
are ambivalent toward change. sample. An enhanced version of this program with
Aardoom and colleagues investigated the effects similar goals and components was developed in the
and cost-​ effectiveness of the Internet-​ based pro- initiative “ProYouth” (https://www.proy­outh.eu/
gram “Featback” for individuals with ED symp- de/network). This enhanced version targets young
toms, comparing different intensities of therapist people between 15 and 25 years and has been dis-
support (Aardoom, Dingemans, Spinhoven, et al., seminated in nine European countries. More than
2016; Aardoom, Dingemans, van Ginkel, et al., 15,000 people have used its online screening tools
2016). The program consisted of psychoeducation and more than 6,000 have registered for full access
with fully automated self-​monitoring and feedback. to the platform (Bauer, Minarik, Özer, & Moessner,
A total of 354 participants with ED symptoms 2014). The risk of developing an ED has been
were randomized to four groups: the program only, shown to be the strongest predictor for a long-​term
the program combined with low-​intensity digital use of the program, which confirms that partici-
therapist support (weekly), the program combined pants adapt the intensity of their program use to
with high-​intensity support (three times per week), their individual needs (Moessner, Özer, Minarik, &
and a wait-​list control condition. From baseline to Bauer, 2013). Moessner, Minarik, Özer, and Bauer
post-intervention, the three intervention conditions (2016) investigated a sample of 453 “ProYouth”
were significantly superior to the control condition users 3 months after their registration, finding
in ED behaviors and psychopathology, depression that 10% had taken up treatment in conventional
and anxiety, and perseverative thinking. Therapist healthcare, 8% intended to do so, and 43% of the
support did not increase symptom reduction, but remaining said they would do so in case of need.
enhanced participants’ satisfaction with the inter- A total of 50% of (potential) help-​seekers reported
vention (Aardoom, Dingemans, Spinhoven, et al., that their attitude toward help-​seeking had changed
2016). As one of the only studies in the field addi- through the participation in the program. These
tionally examining costs and utility, the results did results suggest that access to conventional health-
not show any significant difference between con- care can be facilitated through participation in an
ditions regarding quality-​ adjusted life years and Internet-​based program such as ProYouth.
societal costs. Regarding cost-​ effectiveness, the
intervention was superior to the control condition Carer Support
with no clear preference for the program with or Research has shown that carers of people with
without therapist support (Aardoom, Dingemans, EDs often experience high levels of psychological
van Ginkel, et al., 2016). distress and play an important role in the recov-
An Internet-​based program for university stu- ery process (Hibbs, Rhind, Leppanen, & Treasure,
dents that aimed to integrate education, prevention, 2015). Internet-​based programs could be helpful

512 Internet-Based Interventions

by offering information, professional support, and (2014), using rigorous methodology, identified 16
contact to carers in an easily accessible way (Bauer RCTs of online interventions designed to prevent or
& Moessner, 2013). Nevertheless, research on treat EDs, excluding studies where a conventional
Internet-​ based programs for this target group is psychotherapy was transmitted electronically (CBT
scarce. An RCT with 64 carers of individuals with via videoconferencing). They included only one
AN by Grover et al. (2011) investigated a 4-​month study on AN treatment (Fichter et al., 2012, 2013);
Internet-​ delivered systemic CBT-​ based interven- the remaining studies investigated BED, BN, and/​
tion (“Overcoming Anorexia Online”), designed to or EDNOS. According to their meta-​ analysis,
reduce carers’ distress and to educate them in how only two of four programs on Internet-​ based
to effectively offer support. The intervention par- ED prevention—​ “Student Bodies” and “ESS-​
ticipants received the program and guidance from a KIMO”—​ were associated with improvements in
clinician in 20-​minute sessions per week, while the ED behaviors and psychopathology. The improve-
control group used the usual support services of an ments were of small size and with moderate confi-
AN patient and carer organization. The intervention dence in the effect estimates for Student Bodies and
had a significantly greater positive impact on carers’ low confidence for ESS-​KIMO. Concerning treat-
anxiety and depression levels than the control condi- ment and relapse prevention, beneficial effects were
tion, and gains were maintained at 2-​month follow-​ found for all five included interventions, though
up. Furthermore, a pilot study with 13 participating for most outcomes evidence came from single stud-
parents of adolescents with an ED examined an ies and confidence in the effect estimates was low,
Internet-​ based chat support group as adjunct to often due to high risk of bias. Thus, Loucas et  al.
face-​to-​face family-​based treatment (Binford Hopf, (2014) concluded that the value of Internet-​based
Grange, Moessner, & Bauer, 2013). Results were interventions for EDs must be considered uncer-
promising, as most parents stated that they were tain regarding BN and BED and unknown regard-
satisfied with the chat and found it helpful. ing AN, as they considered the evidence base too
small. Contradicting previous systematic reviews
Evaluation (Aardoom et  al., 2013; Dölemeyer et  al., 2013),
The majority of the studies presented here found they stated, “It is impossible to describe the find-
positive effects for the Internet-​based program under ing as ‘promising’ and there is certainly no basis for
investigation. In most cases, the Internet-​based pro- saying that e-​therapy is a good alternative to face-​
gram led to significantly greater improvements than to-​face treatment” (Loucas et  al., 2014, p.  130),
a wait-​list control condition and was equally effi- which is why they did not encourage its application
cacious compared with face-​to-​face delivery. Given at that moment, but strongly emphasized the need
these effects, most of the existing reviews concluded for further research. They considered only Student
that Internet-​based interventions for EDs are prom- Bodies as efficacious with regard to reduction of
ising (Aardoom et al., 2013; Dölemeyer et al., 2013, ED psychopathology, but noted that the program’s
Shingleton et al., 2013). impact on the risk of subsequent ED development
In addition, Dölemeyer et al. (2013) emphasized was unknown. Moreover, they pointed out the dan-
in their systematic review of eight RCTs and con- ger that if an online program that is not empirically
trolled trials on Internet-​based ED treatment the supported pretends to promote recovery, people
value of guided self-​help, for which they had found with EDs may use it instead of seeking an evidence-​
medium to large effects both within and between based form of treatment, and thereby delay recovery.
groups. They further noted that treatment dropout Two years after the meta-​analysis of Loucas et al.
differed greatly between studies (between 9% and (2014), Melioli et  al. (2016) conducted another
47%). Along these lines, Aardoom et  al. (2013) meta-​analysis with the additional goal to identify
found in their systematic review of 21 studies on variables that might account for the interventions’
Internet-​based interventions on EDs that studies efficacy. They analyzed 20 controlled Internet-​based
which included one or more face-​to-​face sessions intervention studies for selective prevention and
had lower dropout rates than those that did not. treatment of EDs, including five more studies that
They further found Internet-​ based treatment to had not been included by Loucas et al. (2014). They
be more effective for individuals with lower level found small to moderate effect sizes for reductions
psychopathology. in ED behaviors and psychopathology. Concerning
However, the first meta-​ analysis in this field moderators of efficacy, no differences were found
came to less positive conclusions. Loucas et  al. in effects between nonclinical/​mixed and high-​risk

Hilbert, Opitz, de Zwann 513

participants on all outcomes; for shape concern use. However, establishing adequate diagnostic
only there was a larger decrease in the high-​risk validity including appropriate sensitivity and speci-
sample. Furthermore, there was no difference ficity of online self-​report assessments is a challenge
between intent-​to-​treat or completer analyses, indi- on its own (ter Huurne et al., 2015).
cating that these variables did not account for the To conclude the evaluation of Internet-​ based
programs’ efficacy. The findings of this more inclu- interventions, it seems to be premature to deduce
sive review seem rather optimistic, highlighting that whether they are generally useful or recommend-
Internet-​based programs are successful in decreas- able. The existing interventions differed largely
ing ED symptomatology. Nevertheless, still more regarding target groups (e.g., age, severity of symp-
research is needed to, first, increase the programs’ toms, disorder); medium (e.g., e-​mail, chat, video-
effects, and second, enlarge the evidence base. chat, forums); therapist contact (frequency and
When evaluating Internet-​ based interven- mode of contact, e.g., via group or individual chat,
tions for EDs, a significant question is whether e-​mail, and/​or face-​to-​face); duration, intensity, and
they actually improve access to care. The above-​ content; and follow-​ up assessment(s). Moreover,
described “ProYouth” study with its stepped-​care the studies differed in their definitions of therapy,
approach indicated that users’ attitudes toward prevention, and self-​ help, making an evaluation
help-​seeking in face-​to-​face mental healthcare may more difficult.
be changed, especially in case of more severe symp-
toms (Moessner et  al., 2016). More evidence for Implications
Internet-​based interventions’ benefit in improv- Research on Internet-​ based interventions for
ing access to care was brought by McClay, Waters, EDs is at an early stage, and many questions
Schmidt, and Williams (2016), who conducted a remain to be answered. However, due to their
descriptive study in a community-​based sample of numerous and specific advantages, more research
253 individuals with symptoms of BN who revealed is urgently needed in order to further enhance and
positive attitudes toward accessing online self-​help stabilize long-​term effects of Internet-​based ther-
and rather negative attitudes toward face-​ to-​
face apy, unguided and guided self-​ help, therapeutic
treatments, due to fear, shame, long waiting lists, adjuncts, aftercare, and prevention. For this goal,
and negative past experiences with such treatments. Loucas et al. (2014) suggested that program devel-
Thus, these individuals would not have sought help opers should take more advantage of two major
in conventional treatments but were more likely strengths of the Internet that cannot be provided by
to seek help in Internet-​based treatments. In the self-​help books: the possibility to personalize inter-
guided self-​help study of Wagner et  al. (2013), ventions and to make them interactive, possibly fur-
where participants with EDs had been recruited via ther increasing acceptance by users and enhancing
advertisement, 66% of them stated that they had their outcome.
not received any psychotherapeutic or inpatient More studies on predictors of outcome and
treatment before, which also supports the assump- attrition as well as mechanisms of change would
tion that Internet-​ based interventions can reach be useful to tailor Internet-​based interventions for
underserved populations. EDs to the individual patient’s needs, for example,
A question that has not received satisfactory regarding the intensity, speed, and components
answers yet is how to establish ED diagnosis in of the intervention, and the frequency of thera-
accordance with the clinical classification systems in pist contact (Aardoom, Dingemans, & van Furth,
the case of Internet-​based treatments, as the stan- 2016; Bauer & Moessner, 2013; Melioli et  al.,
dard for diagnosis is a face-​to-​face clinical interview. 2016; Schlegl et  al., 2015). Loucas et  al. (2014)
It could be problematic to conduct a face-​to-​face ses- and Fairburn and Murphy (2015) also highlighted
sion before an Internet-​based treatment because this the need for direct-​to-​user studies and recruitment
form of treatment may be chosen for its anonymity of participants directly from the community to
and accessibility, which would then be undermined examine the greatest potential benefit of Internet-​
(Aardoom, Dingemans, van Ginkel, et  al., 2016). based interventions:  their ability to directly reach
For this reason, establishing reliable diagnostic tools those who would benefit. The patient perspective
via online self-​report assessment would be helpful, could be more emphasized in the development of
thereby preserving the advantages of Internet-​based Internet-​based interventions, for example, in stud-
treatment delivery. For this goal, established and ies exploring patients’ acceptance, opinions, needs,
new assessment tools are being evaluated for online preferences, and wishes. More programs that are

514 Internet-Based Interventions

based on therapeutic approaches other than CBT The possibilities of integrating Internet-​ based
should be developed and evaluated to expand the interventions and face-​to-​face treatments should be
scope of available treatments. investigated, both within stepped-​care approaches of
It is especially relevant to further develop and ED treatment, as has been done in the “ProYouth”
evaluate Internet-​ based interventions concerning initiative, and within blended care interventions
aftercare and relapse prevention, as a considerable (Aardoom, Dingemans, & van Furth, 2016; Bauer
number of patients experience relapse after success- & Moessner, 2013). Blended care has the potential
ful treatment, particularly in the first months fol- to provide the benefits of Internet-​based as well as
lowing treatment termination (Olmsted, Kaplan, face-​to-​face interventions and to thus improve both
& Rockert, 1994). Thus, a continuous availability effects and costs:  Internet-​based components can
of support could help to promote maintenance of reduce travel time and promote self-​management,
therapeutic gains. Internet-​based aftercare programs while face-​to-​face sessions can be more personal and
might be useful to help individuals in the process of consolidate the therapeutic relationship (Aardoom,
recovery and early relapse detection, as they are eas- Dingemans, & van Furth, 2016). Concerning
ily accessible and require little therapist assistance stepped care, broadly disseminated Internet-​based
(Gulec et al., 2014). initiatives could facilitate access to ED treatments
Another important aspect that has not suffi- by reducing psychosocial barriers such as stigma and
ciently been considered in the previous research lack of mental health literacy (Bauer & Moessner,
on Internet-​based interventions for EDs is how to 2013). More research is needed to determine
guarantee a patient’s safety in cases of crisis (Schlegl whether and how such programs can improve help-​
et  al., 2015). Especially in unguided programs, seeking behaviors and consequently accelerate treat-
where there is no therapist who could be called or ment uptake in people with EDs.
e-​mailed, there are few options for action if a par- Another aspect not mentioned in most studies
ticipant needs urgent support. and reviews is the problem of diagnosis in Internet-​
Internet-​based carer support for EDs also has based ED treatments, as an accurate diagnosis is
received little attention so far and has even been normally the first step of any conventional psy-
ignored by most reviews, although families may chotherapy. Therefore, it would be of great value
have a decisive influence on a patient’s recovery. to validate established instruments with good ED
When conducting research in this field, it would be screening properties for their use via the Internet, in
important to not only evaluate the effects of such order to create the possibility of an accurate online
interventions on the carers but also on the individ- diagnosis as a solid basis for online treatment.
ual with the ED (Bauer & Moessner, 2013).
Furthermore, more research on cost-​effectiveness Methodological Aspects
of all forms of Internet-​based interventions for EDs As existing studies often lack good confidence in
is needed, as this is one of their often assumed bene- effect estimates (Loucas et  al., 2014), more RCTs
fits. Currently, there are only the results of one study with good methodological quality and low risk of
by Aardoom, Dingemans, van Ginkel, et al. (2016), bias are needed. Risk of bias can be caused, among
demonstrating that their intervention was more other things, by substantial amounts of missing
cost-​effective than a wait-​ list control condition, data due to high dropout rates. Missing data can be
with no differences regarding intensity of therapist handled by certain statistical techniques (Aardoom,
support. It is important to examine whether this also Dingemans, & van Furth, 2016), but are ideally to be
applies to other programs. In addition, there should avoided by avoiding treatment and study dropouts,
be more systematic studies that compare different which would also enhance an intervention’s effects.
dosages of therapist support to further examine In light of considerable dropout rates, the reasons
whether this aspect makes a difference regarding the for treatment dropout and noncompliance need
efficacy of an Internet-​based treatment. Regarding to be elucidated further (Aardoom, Dingemans,
the qualification of the therapist, costs could be & van Furth, 2016). In order to improve report-
reduced if support were provided by a less-​trained ing of Internet-​ based intervention studies, the
person. In order to find out if cost-​effectiveness CONSORT reporting criteria for research on E-​
could be improved in this way, Aardoom et  al. health interventions should be followed (Eysenbach
(2013) suggested investigating whether the support & CONSORT-​EHEALTH Group, 2011).
can also be effectively provided by nonprofessionals Furthermore, most studies compared the online
instead of specialists. intervention with an inactive wait-​ list control

Hilbert, Opitz, de Zwann 515

condition. Based on the positive effects found in or without therapist support in comparison with a waiting
these comparisons, further steps of intervention list for individuals with eating disorder symptoms: A ran-
domized controlled trial. International Journal of Eating
design should include active control conditions, for Disorders, 49, 1068–​1076.
example, face-​to-​face interventions using noninfe- Aardoom, J. J., Dingemans, A. E., & van Furth, E. F. (2016). E-​
riority designs, as well as different Internet-​based health interventions for eating disorders: Emerging findings,
interventions compared with each other (Aardoom issues, and opportunities. Current Psychiatry Reports, 18, 42.
et  al., 2013; Bauer & Moessner, 2013; Schlegl American Psychological Association. (2013). Guidelines for the
practice of telepsychology. American Psychologist, 68, 791–​800.
et al., 2015). Arcelus, J., Mitchell, A. J., Wales, J., & Nielsen, S. (2011).
Mortality rates in patients with anorexia nervosa and other
Conclusion eating disorders:  A  meta-​analysis of 36 studies. Archives of
Internet-​based interventions for EDs have the General Psychiatry, 68, 724–​731.
potential to increase patients’ access to treatment, Bauer, S., Minarik, C., Özer, F., & Moessner, M. (2014). EPA-​
1485–​Internet-​ based prevention and early intervention
given their advantages of providing greater local in eating disorders:  Findings from the European project
and temporal flexibility, control, and anonymity. ProYouth. European Psychiatry, 29, 1.
As has been shown in this chapter, varieties of dif- Bauer, S., & Moessner, M. (2013). Harnessing the power of tech-
ferent applications are possible: They could serve as nology for the treatment and prevention of eating disorders.
low-​intensity initial interventions treating milder International Journal of Eating Disorders, 46, 508–​515.
Bauer, S., Moessner, M., Wolf, M., Haug, S., & Kordy, H.
cases and convince more severe cases to start more (2009). ES[S]PRIT—​an Internet-​based programme for the
intensive face-​to-​face treatment. They could further prevention and early intervention of eating disorders in col-
be used as an adjunct to face-​to-​face psychotherapy, lege students. British Journal of Guidance and Counselling, 37,
as online-​transmitted psychotherapy, as a relapse-​ 327–​336.
prevention aid, as stand-​alone or guided self-​help, Bauer, S., Papezova, H., Chereches, R., Caselli, G., McLoughlin,
O., Szumska, I.,  . . .  Moessner, M. (2013). Advances in the
or as a support base for patients’ carers. Internet-​ prevention and early intervention of eating disorders:  the
based programs have been found to successfully potential of Internet-​delivered approaches. Mental Health
decrease ED-​ related symptoms and risk factors. and Prevention, 1, 26–​32.
Confirmatory tests of their efficacy with adequate Becker, A. E., Hadley Arrindell, A., Perloe, A., Fay, K., & Striegel-​
power are needed, as there are little data in this area Moore, R. H. (2010). A qualitative study of perceived social
barriers to care for eating disorders: Perspectives from ethni-
and both the methodological quality and the level cally diverse health care consumers. International Journal of
of evidence of conducted studies are limited. After Eating Disorders, 43, 633–​647.
efficacy has been demonstrated, Internet-​ based Beintner, I., & Jacobi, C. (2011). Internetgestützte Nachsorge
interventions should be evaluated concerning their bei Bulimia nervosa: Behandlungsmanual und Studiendesign
effectiveness within routine care, before they can [Internet-​based follow-​up care for bulimia nervosa: Treatment
manual and study design]. Psychotherapeut, 56, 516–​521.
be broadly disseminated and implemented. Thus, Beintner, I., Jacobi, C., & Taylor, C. B. (2012). Effects of an
future research is needed before widely recommend- Internet-​based prevention programme for eating disorders
ing Internet-​based interventions for EDs. in the USA and Germany: A meta-​analytic review. European
Eating Disorders Review, 20, 1–​8.
Acknowledgments Binford Hopf, R. B., Grange, D. L., Moessner, M., & Bauer,
S. (2013). Internet-​ based chat support groups for par-
This research was supported by the German ents in family-​based treatment for adolescent eating disor-
Federal Ministry of Education and Research (grant ders:  A  pilot study. European Eating Disorders Review, 21,
01EO1501). 215–​223.
Bulik, C. M., Marcus, M. D., Zerwas, S., Levine, M. D.,
References Hofmeier, S., Trace, S. E.,  . . .  Kordy, H. (2012). CBT4BN
Aardoom, J. J., Dingemans, A. E., Spinhoven, P., & Van Furth, E. versus CBTF2F:  Comparison of online versus face-​to-​face
F. (2013). Treating eating disorders over the Internet: A sys- treatment for bulimia nervosa. Contemporary Clinical Trials,
tematic review and future research directions. International 33, 1056–​1064.
Journal of Eating Disorders, 46, 539–​552. Burns, J. M., Durkin, L. A., & Nicholas, J. (2009). Mental
Aardoom, J. J., Dingemans, A. E., Spinhoven, P., van Ginkel, J. health of young people in the United States: What role can
R., de Rooij, M., & van Furth, E. F. (2016). Web-​based fully the Internet play in reducing stigma and promoting help
automated self-​help with different levels of therapist support seeking? Journal of Adolescent Health, 45, 95–​97.
for individuals with eating disorder symptoms:  A  random- Carrard, I., Crépin, C., Rouget, P., Lam, T., Golay, A., & Van der
ized controlled trial. Journal of Medical Internet Research, Linden, M. (2011). Randomised controlled trial of a guided
18, e159. self-​help treatment on the Internet for binge eating disorder.
Aardoom, J. J., Dingemans, A. E., van Ginkel, J. R., Spinhoven, Behaviour Research and Therapy, 49, 482–​491.
P., Van Furth, E. F., & Van den Akker-​van Marle, M. E. Carrard, I., Rouget, P., Fernández-​ Aranda, F., Volkart, A.-​
(2016). Cost-​utility of an Internet-​based intervention with C., Damoiseau, M., & Lam, T. (2006). Evaluation and

516 Internet-Based Interventions

 deployment of evidence based patient self-​management sup- of a computer-​based interactive eating disorders prevention
port program for bulimia nervosa. International Journal of program at long-​term follow-​up. Eating Disorders, 14, 17–​30.
Medical Informatics, 75, 101–​109. Grover, M., Naumann, U., Mohammad-​ Dar, L., Glennon,
Celio, A. A., Winzelberg, A. J., Wilfley, D. E., Eppstein-​Herald, D., Ringwood, S., Eisler, I.,   .  .  .  Schmidt, U. (2011). A
D., Springer, E. A., Dev, P., & Taylor, C. B. (2000). Reducing randomized controlled trial of an Internet-​based cognitive-​
risk factors for eating disorders: Comparison of an Internet-​ behavioural skills package for carers of people with anorexia
and a classroom-​ delivered psychoeducational program. nervosa. Psychological Medicine, 41, 2581–​2591.
Journal of Consulting and Clinical Psychology, 68, 650–​657. Gulec, H., Moessner, M., Mezei, A., Kohls, E., Túry, F., & Bauer,
de Zwaan, M., Herpertz, S., Zipfel, S., Svaldi, J., Friederich, S. (2011). Internet-​based maintenance treatment for patients
H., Schmidt, F., . . . Hilbert, A. (2017). Effect of Internet- with eating disorders. Professional Psychology:  Research and
based guided self-help vs individual face-to-face treatment on Practice, 42, 479–​486.
full or subsyndromal binge eating disorder in overweight or Gulec, H., Moessner, M., Túry, F., Fiedler, P., Mezei, A., & Bauer,
obese patients: The INTERBED randomized clinical trial. S. (2014). A randomized controlled trial of an Internet-​
JAMA Psychiatry. Advance online publication. based posttreatment care for patients with eating disorders.
Dölemeyer, R., Tietjen, A., Kersting, A., & Wagner, B. (2013). Telemedicine and e-​Health, 20, 916–​922.
Internet-​based interventions for eating disorders in adults: a Hart, L. M., Granillo, M. T., Jorm, A. F., & Paxton, S. J. (2011).
systematic review. BMC Psychiatry, 13, 207. Unmet need for treatment in the eating disorders:  a sys-
Doyle, A. C., Goldschmidt, A., Huang, C., Winzelberg, A. J., tematic review of eating disorder specific treatment seeking
Taylor, C. B., & Wilfley, D. E. (2008). Reduction of over- among community cases. Clinical Psychology Review, 31,
weight and eating disorder symptoms via the Internet in ado- 727–​735.
lescents: A randomized controlled trial. Journal of Adolescent Hay, P. J., & Claudino, A. M. (2015). Bulimia nervosa: Online
Health, 43, 172–​179. interventions. BMJ Clinical Evidence, Systematic
Evans, E. J., Hay, P. J., Mond, J., Paxton, S. J., Quirk, F., Review, 1009.
Rodgers, B.,  . . .  Sawoniewska, M. A. (2011). Barriers to Hibbs, R., Rhind, C., Leppanen, J., & Treasure, J. (2015).
help-​seeking in young women with eating disorders: A quali- Interventions for caregivers of someone with an eating
tative exploration in a longitudinal community survey. Eating disorder:  A  meta-​analysis. International Journal of Eating
Disorders, 19, 270–​285. Disorders, 48, 349–​361.
Eysenbach, G., & CONSORT-​ EHEALTH Group (2011). Hoek, H. W., & van Hoeken, D. (2003). Review of the preva-
CONSORT-​ EHEALTH:  Improving and standardizing lence and incidence of eating disorders. International Journal
evaluation reports of Web-​based and mobile health interven- of Eating Disorders, 34, 383–​396.
tions. Journal of Medical Internet Research, 13, e126. Hötzel, K., von Brachel, R., Schmidt, U., Rieger, E., Kosfelder,
Fairburn, C. G. (1995). Overcoming binge eating. New  York, J., Hechler, T.,   .  .  .  Vocks, S. (2014). An Internet-​based
NY: Guilford Press. program to enhance motivation to change in females with
Fairburn, C. G. (1999). Guided self-​ help for bulimia ner- symptoms of an eating disorder:  A  randomized controlled
vosa: Therapist’s manual. New York, NY: Guilford Press. trial. Psychological Medicine, 44, 1947–​1963.
Fairburn, C. G. (2008). Cognitive behavior therapy and eating dis- Internet World Stats. (2016). Retrieved from http://www.
orders. New York, NY: Guilford Press. internetworldstats.com/stats.htm (June 30, 2016).
Fairburn, C. G., & Murphy, R. (2015). Treating eating disor- Jacobi, C., Morris, L., Beckers, C., Bronisch-​Holtze, J., Winter,
ders using the Internet. Current Opinion in Psychiatry, 28, J., Winzelberg, A. J., & Taylor, C. B. (2005). Reduktion
461–​467. von Risikofaktoren für gestörtes Essverhalten:  Adaptation
Fairburn, C. G., & Patel, V. (2014). The global dissemination of und erste Ergebnisse eines Internet-​
gestützten
psychological treatments: A road map for research and prac- Präventionsprogramms [Reduction of risk factors for dis-
tice. American Journal of Psychiatry, 171, 495–​498. ordered eating behavior:  adaptation and initial results of
Fairburn, C. G., & Wilson, G. T. (2013). The dissemination and an Internet-​supported prevention program]. Zeitschrift für
implementation of psychological treatments:  Problems and Gesundheitspsychologie, 13, 92–​101.
solutions. International Journal of Eating Disorders, 46, 516–​521. Jacobi, C., Morris, L., Beckers, C., Bronisch-​Holtze, J., Winter,
Fichter, M. M., Quadflieg, N., & Lindner, S. (2013). Internet-​ J., Winzelberg, A. J., & Taylor, C. B. (2007). Maintenance
based relapse prevention for anorexia nervosa: Nine-​month of Internet-​based prevention: A randomized controlled trial.
follow-​up. Journal of Eating Disorders, 1, 23. International Journal of Eating Disorders, 40, 114–​119.
Fichter, M. M., Quadflieg, N., Nisslmüller, K., Lindner, S., Jacobi, C., Völker, U., Trockel, M. T., & Taylor, C. B. (2012).
Osen, B., Huber, T., & Wünsch-​Leiteritz, W. (2012). Does Effects of an Internet-​based intervention for subthreshold
Internet-​based prevention reduce the risk of relapse for ano- eating disorders:  A  randomized controlled trial. Behaviour
rexia nervosa? Behaviour Research and Therapy, 50, 180–​190. Research and Therapy, 50, 93–​99.
Fisher, C. B., & Fried, A. L. (2003). Internet-​mediated psycho- Jones, M., Luce, K. H., Osborne, M. I., Taylor, K., Cunning, D.,
logical services and the American Psychological Association Doyle, A. C.,  . . .  Taylor, C. B. (2008). Randomized, con-
ethics code. Psychotherapy: Theory, Research, Practice, Training, trolled trial of an Internet-​facilitated intervention for reduc-
40, 103–​111. ing binge eating and overweight in adolescents. Pediatrics,
Gollings, E. K., & Paxton, S. J. (2006). Comparison of Internet 121, 453–​462.
and face-​to-​face delivery of a group body image and disor- Kersting, A., Schlicht, S., & Kroker, K. (2009). Internettherapie:
dered eating intervention for women: A pilot study. Eating Möglichkeiten und Grenzen [Internet therapy: Opportunities
Disorders, 14, 1–​15. and boundaries]. Der Nervenarzt, 80, 797–​804.
Graff Low, K., Charanasomboon, S., Lesser, J., Reinhalter, K., Kessler, R. C., Berglund, P. A., Chiu, W. T., Deitz, A. C.,
Martin, R., Jones, H.,  . . .  Taylor, C. B. (2006). Effectiveness Hudson, J. I., Shahly, V.,   .  .  .  Xavier, M. (2013). The

Hilbert, Opitz, de Zwann 517

 prevalence and correlates of binge eating disorder in the Paxton, S. J., McLean, S. A., Gollings, E. K., Faulkner, C., &
World Health Organization World Mental Health Surveys. Wertheim, E. H. (2007). Comparison of face-​to-​face and
Biological Psychiatry, 73, 904–​914. Internet interventions for body image and eating problems
Knaevelsrud, C., Jager, J., & Maercker, A. (2004). Internet- in adult women:  An RCT. International Journal of Eating
​psychotherapie: Wirksamkeit und Besonderheiten der thera­ Disorders, 40, 692–​704.
peutischen Beziehung. Verhaltenstherapie, 14, 174–​183. Prochaska, J. O., & DiClemente, C. C. (1992). Stages of change
Leung, S. F., Ma, L. C., & Russell, J. (2013). An open trial of in the modification of problem behaviors. In M. Hersen, R.
self-​help behaviours of clients with eating disorders in an M. Eisler, & P. M. Miller (Eds.), Progress in behavior modifica-
online programme. Journal of Advanced Nursing, 69, 66–​76. tion (pp. 184–​214). Sycamore, IL: Sycamore.
Lindenberg, K., & Kordy, H. (2015). Wirksamkeit eines Riper, H., Smit, F., van der Zanden, R., Conijn, B., Kramer, J.,
gestuften, Internetvermittelten Ansatzes zur Prävention von & Mutsaers, K. (2012). E-​mental health. high tech, high touch,
Essstörungen bei Schülern der 7. bis 10. Klasse [Efficacy of high trust. Utrecht: Trimbos Instituut.
an Internet-​delivered tiered strategy for eating disorder pre- Robinson, P. H., & Serfaty, M. A. (2001). The use of e-​mail
vention in high school students]. Kindheit und Entwicklung, in the identification of bulimia nervosa and its treatment.
24, 55–​63. European Eating Disorders Review, 9, 182–​193.
Ljotsson, B., Lundin, C., Mitsell, K., Carlbring, P., Ramklint, Robinson, P. H., & Serfaty, M. A. (2008). Getting better byte by
M., & Ghaderi, A. (2007). Remote treatment of bulimia byte: A pilot randomised controlled trial of email therapy for
nervosa and binge eating disorder:  A  randomized trial of bulimia nervosa and binge eating disorder. European Eating
Internet-​ assisted cognitive behavioural therapy. Behaviour Disorders Review, 16, 84–​93.
Research and Therapy, 45, 649–​661. Sánchez-​Ortiz, V. C., Munro, C., Stahl, D., House, J., Startup,
Loucas, C. E., Fairburn, C. G., Whittington, C., Pennant, M. H., Treasure, J.,  . . .  Schmidt, U. (2011). A randomized con-
E., Stockton, S., & Kendall, T. (2014). E-​therapy in the trolled trial of Internet-​based cognitive-​behavioural therapy
treatment and prevention of eating disorders:  A  systematic for bulimia nervosa or related disorders in a student popula-
review and meta-​analysis. Behaviour Research and Therapy, tion. Psychological Medicine, 41, 407–​417.
63, 122–​131. Schlegl, S., Bürger, C., Schmidt, L., Herbst, N., & Voderholzer,
Mallen, M. J., Day, S. X., & Green, M. A. (2003). Online versus U. (2015). The potential of technology-​based psychological
face-​to-​face conversation: An examination of relational and interventions for anorexia and bulimia nervosa: A systematic
discourse variables. Psychotherapy:  Theory, Research, Practice, review and recommendations for future research. Journal of
Training, 40, 155–​163. Medical Internet Research, 17, e85.
McClay, C. A., Waters, L., Schmidt, U., & Williams, C. (2016). Shingleton, R. M., Richards, L. K., & Thompson-​Brenner,
A survey of attitudes towards computerized self-​help for eat- H. (2013). Using technology within the treatment of eat-
ing disorders within a community-​based sample. Behavioural ing disorders:  A  clinical practice review. Psychotherapy, 50,
and Cognitive Psychotherapy, 44, 65–​78. 576–​582.
Melioli, T., Bauer, S., Franko, D. L., Moessner, M., Ozer, F., Stice, E., Becker, C. B., & Yokum, S. (2013). Eating disorder
Chabrol, H., & Rodgers, R. F. (2016). Reducing eating dis- prevention:  Current evidence-​ base and future directions.
order symptoms and risk factors using the Internet: A meta-​ International Journal of Eating Disorders, 46, 478–​485.
analytic review. International Journal of Eating Disorders, Stice, E., Rohde, P., Durant, S., & Shaw, H. (2012). A prelimi-
49, 19–​31. nary trial of a prototype Internet dissonance-​based eating
Miller, W. R., & Rollnick, S. (2002). Motivational interview- disorder prevention program for young women with body
ing:  Preparing people for change (2nd ed.). New  York, image concerns. Journal of Consulting and Clinical Psychology,
NY: Guilford Press. 80, 907–​916.
Mitchell, J. E., Crosby, R. D., Wonderlich, S. A., Crow, S., Suler, J. (2004). The online disinhibition effect. CyberPsychology
Lancaster, K., Simonich, H.,  . . .  Myers, T. C. (2008). A ran- and Behavior, 7, 321–​326.
domized trial comparing the efficacy of cognitive-​behavioral Taylor, C. B., Bryson, S., Luce, K. H., Cunning, D., Doyle, A.
therapy for bulimia nervosa delivered via telemedicine versus C., Abascal, L. B.,  . . .  Wilfley, D. E. (2006). Prevention
face-​to-​face. Behaviour Research and Therapy, 46, 581–​592. of eating disorders in at-​risk college-​age women. Archives of
Moessner, M., Minarik, C., Özer, F., & Bauer, S. (2016). Can General Psychiatry, 63, 881–​888.
an Internet-​ based program for the prevention and early ter Huurne, E. D., de Haan, H. A., ten Napel-​Schutz, M. C.,
intervention in eating disorders facilitate access to conven- Postel, M. G., Menting, J., van der Palen, J.,  . . .  DeJong,
tional professional healthcare? Journal of Mental Health, 25, C. A. (2015). Is the Eating Disorder Questionnaire-​Online
441–​447. (EDQ-​O) a valid diagnostic instrument for the DSM-​IV-​TR
Moessner, M., Özer, F., Minarik, C., & Bauer, S. (2013). classification of eating disorders? Comprehensive Psychiatry,
ProYouth: Ein Online-​Programm zur Prävention und frühen 57, 167–​176.
Intervention bei Essstörungen [ProYouth:  An online pro- Treasure, J., Macare, C., Mentxaka, I. O., & Harrison, A. (2010).
gram for prevention and early intervention in eating disor- The use of a vodcast to support eating and reduce anxiety in
ders]. PiD—​Psychotherapie im Dialog, 14, 74–​76. people with eating disorder:  A  case series. European Eating
Norris, M. L., Boydell, K. M., Pinhas, L., & Katzman, D. K. Disorders Review, 18, 515–​521.
(2006). Ana and the Internet: A review of pro-​anorexia web- Wagner, B., Nagl, M., Dölemeyer, R., Klinitzke, G., Steinig, J.,
sites. International Journal of Eating Disorders, 39, 443–​447. Hilbert, A., & Kersting, A. (2016). Randomized controlled
Olmsted, M. P., Kaplan, A. S., & Rockert, W. (1994). Rate and trial of an Internet-​ based cognitive-​behavioral treatment
prediction of relapse in bulimia nervosa. American Journal of program for binge-​ eating disorder. Behavior Therapy, 47,
Psychiatry, 151, 738–​743. 500–​514.

518 Internet-Based Interventions

Wagner, G., Penelo, E., Wanner, C., Gwinner, P., Trofaier, M.-​L., Yager, J. (2001). E-​mail as a therapeutic adjunct in the outpa-
Imgart, H.,  . . .  Karwautz, A. F. (2013). Internet-​delivered tient treatment of anorexia nervosa: Illustrative case material
cognitive-​behavioural therapy vs. conventional guided self-​ and discussion of the issues. International Journal of Eating
help for bulimia nervosa:  Long-​term evaluation of a ran- Disorders, 29, 125–​138.
domised controlled trial. British Journal of Psychiatry, 202, Yager, J. (2003). E-​mail therapy for anorexia nervosa: Prospects
135–​141. and limitations. European Eating Disorders Review, 11,
Williams, C. J., Aubin, S. D., Cottrell, D., & Harkin, P. J. 198–​209.
(1998). Overcoming bulimia:  A  self-​ help package. Leeds, Zabinski, M. F., Pung, M. A., Wilfley, D. E., Eppstein, D.
UK: University of Leeds Press. L., Winzelberg, A. J., Celio, A., & Taylor, C. B. (2001).
Winzelberg, A. J., Eppstein, D., Eldredge, K. L., Wilfley, Reducing risk factors for eating disorders: Targeting at-​risk
D., Dasmahapatra, R., Taylor, C. B., & Dev, P., (2000). women with a computerized psychoeducational program.
Effectiveness of an Internet-​based program for reducing risk International Journal of Eating Disorders, 29, 401–​408.
factors for eating disorders. Journal of Consulting and Clinical
Psychology, 68, 346–​350.

Hilbert, Opitz, de Zwann 519

 C H A PT E R

Afterword
28
W. Stewart Agras and Athena Robinson

Abstract
This chapter reflects on core themes raised within this Handbook, which collectively reflect the state of
the science in the eating disorder field. Such themes include the growing recognition of the complexity
of eating disorders including their etiological underpinnings and the contributions of basic sciences to
an understanding of processes underlying the expression of maladaptive eating patterns. The status
of the prevention and treatment of these disorders as reflected in the literature to date is considered
with questions raised about future progress. Looking toward the future, new technologies may offer
opportunities to bring cost-​effective evidence-​based treatments to underserved populations. However,
such opportunities bring with them new ethical and practical considerations. Also highlighted are
potential areas for further research.
Key Words:  genetics, pharmacogenetics, prevention, research, treatment, implementation, technology,
etiology, risk factor, mechanism

Introduction 2016). A variety of factors influence the etiology and

An overview of the eating disorders as illustrated maintenance of eating disorders. These include cul-
in this Handbook leads to a consideration of several ture, race, family behaviors specific to the individual,
questions. Among these questions are: overweight, body image concerns, dietary restriction,
negative affect, perfectionism, neurochemical abnor-
1.  What is the influence of the complexity of the
malities, and genetic influences. It is likely that many
eating disorders on the understanding of their etiology?
of these influences interact with one another at differ-
2.  Will new basic science findings lead to useful
ent developmental periods. Although the search for
change in treatment or prevention?
mechanisms underlying the disorders of eating and
3.  Have the obtained treatment effects reached
those accounting for treatment outcomes will acceler-
a plateau?
ate in the next few years spurred by RDoC, it is possi-
4.  What is the likely impact of technology on the
ble that the mechanisms will prove to be as complex as
assessment and treatment of the eating disorders?
the disorders themselves. If an unlikely single solution
5.  Can evidence-​ based treatments be imple-
to etiology is at one end of the spectrum, at the other
mented more widely?
is the possibility that every case has a unique etiology.
If the truth lies somewhere in between, the search for
Complexity of the Eating Disorders mechanism may provide better guidance to clinicians
The range of topics covered in this Handbook than the existing diagnostic system.
make it abundantly clear that disorders of eating, like
many other psychiatric and medical disorders, are The Contributions of Basic Science
complex and difficult to sort into diagnostic catego- At this point it is fair to say that psychology as
ries defying singular etiologic explanations (Cohen, a basic science has made major contributions to

520
the development of effective psychotherapies and The characterization of environmental variables
assessments for the eating disorders as have the that either cause or maintain the eating disorders
pharmacological sciences although no new specific has shown progress in identifying risk factors, and
medication has yet been developed for the treat- to a lesser extent, causal risk factors. The identifi-
ment of any eating disorder. The contributions of cation of modifiable, causal risk factors has led to
other basic endeavors, while generating important the formulation of effective prevention studies.
information, have not yet impacted the prevention Prospective studies large enough to investigate the
or treatment of the eating disorders. contribution of a number of hypothesized risk
Genetic research is a fundamental area of endeavor, factors combined with laboratory studies or clini-
genetic findings that act as markers for disordered eat- cal trials modifying a risk factor in order to iden-
ing or point to biological mechanisms involved in the tify causal risk factors are needed. The interaction
eating disorders may lead to more precisely targeted between causal risk factors and specific genes at
prevention or treatments. To enhance genetic studies, particular developmental stages may eventually be
improved delineation of the eating disorder syndromes a useful research area to pursue. For example, one
may be necessary. This in turn requires improved study showed that as weight-​related peer teasing,
assessment and classification of the eating disorders in a risk factor for the development of eating disor-
order to deal with the heterogeneity within the existing ders, increased, both genetic and environmental
syndromes and the overlap between them. Moreover, factors influencing disordered eating also increased
it is unclear at present whether genetic research should (Fairweather-​Schmidt & Wade, 2016).
concentrate on already defined syndromes, or on sub-
syndromes as defined in some classification studies, or Has Treatment Development Reached
on eating behaviors such as binge eating, purging, diet- a Plateau?
ing, and loss of control over eating. Characterization The development of evidence-​based treatments
of potential phenotypes may be helpful in furthering is a time-​consuming process leading from case ser-
genetic research (Lopez, Tchanturia, & Treasure, 2009). ies to multisite controlled trials. Because clinical
One potential phenotype involves anorexia nervosa trials are expensive, many questions have not been
characterized by the symptom complex of anxiety, answered. Hence, progress in treatment develop-
depression, and obsessive-​ compulsive behaviors. ment is often slow. Treatment development for AN,
For example, obsessive-​compulsive eating behaviors particularly persistent AN, has been disappointing
have been found to moderate treatment outcome in for decades. At this point there are still no effec-
adolescents with anorexia nervosa (AN) in several tive evidence-​based treatments for these patients.
studies. Patients with obsessive-​compulsive eating Most controlled studies to date are inconclusive,
behaviors have poorer treatment outcomes and may show relatively poor results, or are no better than
need more and longer treatment than those without specialist management (Hay et al., 2014). However,
such behaviors. The second involves the cognitive in an important development over the past 25 years,
processes underlying the rigid thinking found in a specific family therapy for adolescents, in which
many individuals with AN. To date, progress in this parents take temporary control of their adolescent’s
area has been slow with no genetic findings relevant feeding, can now be regarded as evidence-​ based
to biologic mechanisms or to prevention or treat- (Jewell, Blessit, Stewart, Simic, & Eisler, 2016).
ment of the eating disorders. Larger studies may be Early treatment of AN promises to lower the num-
helpful in identifying candidate genes together with ber of persistent cases, hence further research to
a greater emphasis on understanding of the genetics refine this treatment is needed. No pharmacological
of bulimia nervosa and binge eating disorder. agent has been adequately tested in adolescents with
Understanding the brain in terms of neural path- AN, and there is no effective evidence-​based medi-
ways and function together with neurochemistry is cation for adults (Miniati et al., 2016).
also beginning to contribute important informa- Treatment research in bulimia nervosa (BN) is
tion concerning the eating disorders. One focus is more advanced, with cognitive-​behavioral therapy
on reward-​processing mechanisms in the brain that (CBT) as the first-​line treatment with evidence for
appear to be disturbed in individuals with an eating both short-​and long-​term effects. However, long-​
disorder, who tend to show greater reactivity to food term follow-​up recovery rates with CBT of about
images and to anticipation of taste. The relation of 25% to 40% are lower than desirable. Other treat-
these disturbances to neurochemical pathways may ments such as interpersonal psychotherapy (IPT)
lead to more refined treatment. fare no better, and IPT is less effective than CBT

Agras, Robinson 521

at end-​ of-​
treatment and perhaps at follow-​ up Dissemination/​Implementation
(Fairburn et  al., 2015). This relatively low rate of Once effective treatments have been identified it
recovery needs to be enhanced, but so far no treat- is necessary to disseminate these therapies to com-
ment has been found superior to enhanced CBT munity settings such as HMOs, community clin-
(CBT-​E), nor has the effectiveness of CBT much ics, college campuses, community practitioners,
improved over the past 5  years, although CBT-​E and so forth. Although largely based on therapist
improves treatment outcome in more severe cases. report, studies show that evidence-​based treatments
This is an area requiring more intensive treatment for eating disorders are not widely used in commu-
development with a better understanding of treat- nity practice (Wolitzky-​Taylor, Zimmerman, Arch,
ment mechanism, moderators, and mediators. De Guzman, & Lagomasino 2015). It is evidently
Moreover, theory-​based methods to enhance out- necessary to address this problem with the aim of
come are urgently needed. providing more effective treatment for more people.
Pharmacological treatment for BN has been Findings from basic implementation science stud-
extensively explored, and efficacy has been dem- ies are beginning to identify the many variables that
onstrated for most antidepressants, although influence dissemination and implementation.
only fluoxetine is approved by the Federal Drug However, CBT, IPT, and other treatments devel-
Administration (FDA) for the treatment of BN. oped for specialty centers need to be adapted to the
Few medication studies have long-​term follow-​up, realities of community clinical settings and tested in
hence evidence-​based guidance for long-​term treat- those settings in controlled trials to ensure that they
ment is, for the most part, lacking. are superior to treatment as usual, that is, treatment
In the treatment of binge eating disorder delivered in the specific community context. Very few
(BED), IPT is equivalent to CBT at both end studies of this sort have appeared, so that the effec-
of treatment and follow-​up, and a larger propor- tiveness and limitations of such adapted treatments
tion of patients are remitted than is the case with are unknown. For example, many clinical settings
BN. Again, however, more treatment research is cannot afford treatments extending over 6  months
needed to enhance follow-​up remission rates of with 18 or more sessions, hence shortened variants of
50% to 60%. Moreover, neither IPT nor CBT evidence-​based treatments are needed. This suggests
lead to much weight loss in this largely overweight that dosage studies of evidence-​ based treatments
population. Antiepileptic drugs such as topira- may provide useful information for implementa-
mate appear promising in reducing both binge tion. Another route is to develop shortened versions
eating and weight although dropout rates from of treatment that include the primary ingredients of
side effects are high. Further trials of these medi- the original treatment. One such treatment, guided
cations in combination with CBT or IPT are war- self-​help based on CBT, has been found effective
ranted. Urgently needed are newer medications even when provided by masters level therapists with
that may affect both binge eating and weight. shorter and fewer sessions. Controlled studies com-
One such medication, lisdexamphetamine, which paring community clinics’ usual treatment with the
has been approved by the FDA for the treatment shortened evidence-​based therapy are needed before
of BED, appears promising in short-​term stud- implementing such treatment. This is an important
ies and recently in a longer-​term study (Hudson, missing link in treatment research.
McElroy, Ferreira-​ Cornwell, Radewonuk, & In order to have therapists in different settings
Gasior, 2017). adopt and use evidence-​ based treatments effec-
With the new focus on the RDoC matrix rather tively it will be important to understand which
than on clinical diagnosis there is some concern training method will be best in what circumstance.
that interest in, and funding for, treatment studies Controlled studies in this area are sorely needed.
may be reduced to the detriment of patient care. Moreover, an understanding of the many organiza-
There are two potential aims for further treatment tional factors that impinge on the use of evidence-​
research. The first is to understand the mechanisms based therapies and how to incentivize adoption
by which treatments for eating disorders work so within particular organizations are also important
that effectiveness can be enhanced. The second is to research areas.
have a better understanding of the reasons for poor
improvement with treatment and with this knowl- Technologies and Dissemination
edge to enhance treatment or devise new treatment The miniaturization of computers has led to the
procedures. development of mobile devices, such as telephones

522 Afterword
and tablets, potentially allowing treatment monitor- Clearly this is a very promising area as the technology
ing or even treatment via the Internet or electronic becomes more easily usable and less expensive.
applications (apps). This is a novel development Basic and applied research into the eating disor-
over the last few years, and several controlled stud- ders is thriving although the difficult task of under-
ies are now available. In a recent review (Agras, standing etiology from risk factor to neurobiology
Fitzsimmons-​Craft, & Wilfley, 2017) of technology-​ is in its infancy. Translational research is now very
based assessment and treatment, the authors con- much needed, for example, translating the latest
cluded, “there is not a strong enough evidence-​base findings from psychological research to treatments,
to support widespread usage of Internet treatments or the effects of treatment on brain mechanisms, or
for eating disorders in the clinic. These studies sug- using advanced technology to enhance assessment
gest that internet-​based treatment is likely to be and treatment. This either requires two skill-​sets
effective and that studies comparing this treatment in one head or research groups that encompass the
to face-​to-​face treatment are now needed.” It should needed skills. The results of such efforts may throw
be noted that most variants of Internet-​based treat- light on what we do not presently know.
ment do not use the medium to individualize treat-
ment; hence more sophisticated programming is References
needed. Moreover some aspects of e-​ treatment Agras, W. S., Fitzsimmons-​Craft, E. E., & Wilfley, D. E. (2017).
Evolution of cognitive-​behavioral therapy for eating disor-
raise ethical concerns. For example, Internet-​based
ders. Behaviour Reearch & Therapy, 88, 26–​36.
assessment and treatment without a therapist may Cohen, B. M. (2016). Embracing complexity in psychiatric
not identify important safety concerns such as diagnosis, treatment, and research. JAMA Psychiatry, 73,
low weight, suicidal ideation, vital instability, and 1211–​1212.
electrolyte abnormality, or identify newly emerg- Fairburn, C. G., Bailey-​Straebler, S., Basden, S., Doll, H. A.,
Jones, R., Murphy, R.,  . . .  Cooper, Z. (2015). A transdiag-
ing psychopathology during treatment. Moreover,
nostic comparison of enhanced cognitive behavior therapy
it is questionable whether a patient whose identity (CBT-​E) and interpersonal psychotherapy in the treatment
is unknown should be engaged in treatment. In of eating disorders. Behaviour Research & Therapy, 70, 64–​71.
the United States, state licensing regulations vary Fairweather-​Schmidt, A. K., & Wade, T. D. (2016). Weight
concerning treatment by out-​ of-​
state providers. related peer-​teasing moderates genetic and environmental
risk and disordered eating:  Twin study. British Journal of
In the case of an app, who is the treatment agent?
Psychiatry. bjp.bp.116.184648
Presumably the company selling the application. Hay, P., Chinn, D., Forbes, D., Madden, S., Newton, R.,
However, there does not appear to be adequate reg- Sugenor, L.,  . . .  Royal Australian and New Zealand College
ulation at the federal level to ensure that such apps of Psychiatrists. (2014). Royal Australian and New Zealand
are effective and without harm. These developments College of Psychiatrists clinical practice guidelines for the
treatment of eating disorders. Australian & New Zealand
still in their infancy may become another way to
Psychiatry, 48, 977–​1008.
provide evidence-​ based treatments to the large Hudson, J. I., McElroy, S. L., Ferreira-​ Cornwell, M. C.,
number of underserved individuals who at present Radewonuk, J., & Gasior, M. (2017). Efficacy of
have no access to treatment. Lisdexamfetamine in adults with moderate to severe binge-​
The use of mobile apps to provide feedback on eating disorder: A randomized clinical trial. JAMA Psychiatry,
July 12. doi: 10.1001/​jamapsychiatry.2017.1889
therapeutic progress to patients and therapists is an
Jewell, T., Blessit, E., Stewart, C., Simic, M., & Eisler, I. (2016).
important new development. There are many such Family therapy for child and adolescent eating disor-
apps available at this point in time although none ders: A critical review. Family Process, 55, 577–​594.
relevant to eating disorders have been rigorously Lopez, C., Tchanturia, K., & Treasure, J. (2009). Weak central
evaluated, hence the effectiveness of providing feed- coherence in eating disorders: A step toward looking for an
endophenotype of eating disorders. Clinical Experimental
back in this manner is unknown.
Neuropsychology, 1, 117–​125. Epub.
Other technologies such as virtual reality are even Miniati, M., Mauri, M., Ciberti, A., Mariani, M. G., Marazziti,
less developed at this time. However, these devices D., & Dell’Osso, L. (2016). Psychopharmacological options
can be used whenever exposure to avoided scenarios for adult patients with anorexia nervosa. CNS Spectrum, 21,
is needed in an assessment or treatment context. For 134–​142.
Wolitzky-​Taylor, K., Zimmerman, M., Arch, J. J., De Guzman,
example, exposure to a variety of avoided foods or to
E., & Lagomasino, I. (2015). Has evidence-​based psycho-
body image experiences. Virtual reality may also be social treatment for anxiety disorders permeated usual care
useful in the assessment of eating disorders for research in community mental health settings? Behaviour Research &
purposes, for example by actualizing binge triggers. Therapy, 72, 9–​17.

Agras, Robinson 523

 INDEX

Page numbers followed by f, t indicate a figure or table on the associated page.

A affect dysregulation model as “activity disorder,” 192–​193

absolute dietary restriction within biosocial theory, 335 adolescent
relative dietary restriction vs., 136 affective disorders family therapy for, 321–​322 (see also
abstinence EDs and, 164, 229–​230 family therapy for adolescent AN
dialectical “affect reduction” model, 162–​163 (FBT-​AN))
in BED and BN management, 343 affect regulation in adults
abuse as factor in EDs, 238–​239 family therapy for, 329–​330
laxative affect regulation model after bariatric surgery, 461–​462
BN and, 226 for EDs, 340 altered feeding behaviors related to
in PD, 446 African Americans studies of, 48–​49, 49t
ACC. see anterior cingulate body image and eating concerns anxiety and harm avoidance with
cortex (ACC) among, 197 PET data correlated with, 57–​58
N-​acetylcysteine (NAC) AFT. see adolescent-​focused therapy (AFT) appetitive regulation in, 47–​79
for BN, 375 agomelatine CCK and, 51, 54t
acid reflux for NES, 382 fMRI studies of, 61–​66, 62f, 65f
BN and, 225 AI. see artificial intelligence (AI) future research directions related
“activity disorder” alcohol abuse/​dependence disorder to, 66–​67
AN as, 192–​193 anxiety disorders related to, 233 gender differences in, 61
ADHD. see attention deficit hyperactivity EDs and, 233–​234 ghrelin and, 52–​53, 55t
disorder (ADHD) OCD related to, 233 HPA axis effects on, 50
adolescent(s) alcoholism images of food in, 62
early EDs and, 233–​234 interoceptive processing in, 59–​60
ED–​related diagnostic criteria alexithymia introduction, 47
for, 19–​20 described, 176 leptin and, 51–​52, 54t–​55t
EDs in EDs and, 176 neurocircuitry of, 58–​60
CRT for, 399–​400 allele(s) neuroendocrine
family therapy for, 320–​321 described, 80 systems–​related, 50–​53
IPT for allocentric lock (AL) theory NPY and, 51, 54t
adaptations of, 311 of body image distortion, 481, 481f opioid peptides and, 50–​51, 54t
adolescent anorexia nervosa (AN) alprazolam PET studies of, 61–​62
family therapy for, 321–​322 (see also for AN, 361t, 368 PYY and, 51, 54t
family therapy for adolescent AN altered feeding behaviors regional cerebral blood flow studies
(FBT-​AN)) in AN and BN in, 61–​62
adolescent bulimia nervosa (BN) studies of, 48–​49, 49t SPECT studies of, 61–​62
FBT for, 324–​327 (see also family therapy alternate rebellion tastes of food in, 62–​63, 62f
for adolescent BN (FBT-​BN)) in BED and BN management, 343 twin studies, 48
adolescent-​focused therapy (AFT) AL theory. see allocentric lock (AL) theory atypical (see atypical anorexia nervosa
FBT vs., 323 American Psychiatric Association, 499 (atypical AN))
adolescent girls American Psychological Association, 506 binge eating/​purging type of, 10
body dissatisfaction among, 480 American Psychologist, 198 binging in
adult(s) amitriptyline negative affect lability and, 239
EDs in for AN, 360, 361t brain imaging studies in, 58–​66
family therapy for, 329–​330 amygdala CBT-​E for, 278–​280
AEDs. see antiepileptic drugs (AEDs) OFC and, 60 CBT for, 278–​280
affect(s) AN. see anorexia nervosa (AN) efficacy of, 278–​279
defined, 156 angry threat stimuli costs of, 412
IPT in encouraging acceptance of, 298 attentional biases to, 240 course of, 36–​37
painful affect, 298 Annals of Physical and Rehabilitation cross-​cultural patterns of, 35–​36
IPT in helping patient experience Medicine, 483 CRT for
suppressed affect, 298 anorexia nervosa (AN). see also eating RCTs of, 397
negative (see negative affect) disorder(s) (EDs) defined, 461

525
anorexia nervosa (AN) (cont.) longitudinal studies’ characteristics, for AN, 368
described, 2, 173–​174 107, 111, 112, 117 apoptosis
diagnosis of, 10 social media in, 262–​263 AN and, 223
discovery of, 34 state vs. trait characteristics in, 47–​48 appetite-​focus DBT, 339
early sensory specific satiety in, 60 subsystems in, 47 appetite stimulants
epidemiology of, 34–​35 subtypes of, 10 for AN, 361t, 365
features of, 461 suicide attempts related to, 36 appetitive regulation
genetic influences on, 48, 80–​105 suicide related to, 36 in AN and BN, 47–​79 (see also
(see also genetic influences) trait anxiety and, 237 anorexia nervosa (AN); bulimia
history of, 2 treatment of, 3–​4, 360–​368, 361t (see also nervosa (BN))
hunger and satiety in specific types) Apple Watch, 494
fMRI studies of, 63 AEDs in, 368 apps. see mobile devices and
incidence of, 35 antidepressant–​antipsychotic applications (apps)
interoception in combinations in, 364–​365 ARFID. see avoidant/​restrictive food
fMRI studies of, 63–​64 antidepressants in, 360–​362, 361t intake disorder (ARFID)
IPT for, 303–​304 antipsychotics in, 361t, 363–​364 aripiprazole
in Jamaica, 200 anxiolytic medications in, 368 for BN, 375
medical complications of, 222–​225 appetite stimulants in, 361t, 365 arousal systems
bone metabolism–​related, 224–​225 cannabinoids in, 361t, 365 EDs and, 29
cardiac, 222–​223 CBT in, 278–​280 artificial cervical vagal nerve
dermatologic, 225 cost-​effectiveness of, 412–​413 stimulation (VNS)
endocrine, 224 CRT in, 397 in depression management, 158
gastrointestinal, 223–​224 ghrelin agonists in, 361t, 367 artificial intelligence (AI), 493
hematologic, 224 hormonal agents in, 361t, 366–​367 values embedded in, 499
malnutrition, 223 lithium in, 361t, 367 assessment(s)
neurologic, 224 nutritional supplements in, 368 described, 211
pulmonary, 223 opioid antagonists in, 368 ecological momentary
monoamine function disturbances opioids in, 368 of negative affect, 238–​239
related to, 53, 55–​58 pharmacotherapy in, 360–​368, 361t of EDs, 209–​221
DA activity in, 53, 55 (see also specific agents) context of, 211–​212
serotonin, 55–​58 prokinetics in, 361t, 365–​366 determining domains or constructs
mortality rate associated research on, 521 of interest in, 213–​215
with, 222–​223 zinc in, 361t, 366 diagnosis in, 213
negative emotionality and, 173–​177 anterior cingulate cortex (ACC), 56–​57, 59 EDE in, 216
emotional processing deficits, anterior insula, 59 ESP in, 217
176–​177 antidepressant(s) function of, 212–​213
emotional responses to body image for BED, 376t, 377–​378 instruments in, 215–​218
exposure, 175–​176 for BN, 369–​372, 370t interviews in, 215–​217
emotional responses to food cues, antidepressant–​antipsychotic introduction, 211
174–​175 combinations psychological, 211–​221
neurobiologic alterations in, 47–​48 for AN, 364–​365 SCOFF in, 212–​213, 216–​217
neuropeptide and neuroendocrine antiepileptic drugs (AEDs) screening tests in, 212–​213
alterations in, 49–​53, 54t–​55t for AN, 368 secondary domains in, 214
parents’ role in, 319 for BED, 376t, 379–​381 self-​report questionnaires in, 217–​218
perfectionism and, 236–​238 for BN, 370t, 372–​373 (see also self-​report questionnaires)
persistence and, 237 anti-​ideal(s) thresholds for recovery in, 214–​215
prevalence of, 263 ideals vs., 194 treatment planning and outcome
psychological comorbidity of, 229–​243, antipsychotic(s) in, 213
231t–​233t for AN, 361t, 363–​364 underlying theoretical assumptions
affective disorders, 229–​230 for BED, 382 in, 213–​214
anxiety disorders, 230 for BN, 375 introduction, 211
ICDs, 234–​235 anxiety as process, 211–​221
personality disorders, 235–​236 AN and assumption(s)
substance abuse disorders, 230, PET data correlated with, 57–​58 underlying
233–​234 trait-​related, 237 in EDs assessment, 213–​214
research on BN and athletes
future directions for, 66–​67 PET data correlated with, 57–​58 ED prevention in, 260–​261
response to inhibition alterations in EDs and, 164 atomoxetine
fMRI studies of, 65–​66 anxiety disorders for BED, 376–​377, 376t
response to reward alterations in alcohol abuse/​dependence disorder attention
fMRI studies of, 64–​65, 65f and, 233 sustained
restricting type of, 10 EDs and, 230 norepinephrine in, 240
risk factors for, 3 anxiolytic medications attentional bias(es)

526 Index
 to social and angry threat stimuli and Barrett’s esophagus predictors and moderators of,
emotion recognition, 240 BN and, 225 277–​278
attention deficit hyperactivity BDS. see Buss-​Durkee Scale (BDS) vs. behavioral weight loss
disorder (ADHD) Beck Depression Inventory, 328 treatment, 276
ARFID and, 431 Beck Depression Inventory scores, vs. group behavioral weight-​loss
drugs for 442–​443 interventions, 130
in BED, 375–​377, 376t BED. see binge eating disorder (BED) vs. pharmacotherapy, 276
in BN, 374 Before I Eat, 496 course of, 39
attitude(s) behavior(s) cross-​cultural patterns of, 39
eating altered feeding defined, 459
negative affect impact on, 161–​164 in AN and BN, 48–​49, 49t described, 169
atypical AN. see atypical anorexia nervosa binge–​purge diagnosis of, 10
(atypical AN) harm avoidance scores in women discovery of, 38
atypical anorexia nervosa (atypical AN), with, 237 DRD2 in, 160–​161
438, 448–​451 BPD–​related, 334–​335 DSM on, 2
criteria for, 448–​449 dietary, 136–​137 epidemiology of, 38–​39
defined, 448 eating incidence of, 38–​39
described, 448–​449 after bariatric surgery, 458–​469 IPT for, 301–​303
history of, 448–​449 impact on body image and eating future directions in, 310–​311
models of concerns, 191–​193 negative emotionality and, 169–​173
evidence of diagnostic validity purging emotional responses to body image
and clinical significance using, BN and, 225–​226 exposure, 170–​171
449–​450 SE–​related, 425–​427 emotional responses to food cues,
prevalence of, 449 behavioral disturbances 169–​170
status of genetic influences on, 81 negative affect and binge eating,
research needed to clarify, 451 behavioral family systems therapy 171–​173
avoidant/​restrictive food intake disorder (BFST), 322 OBE episodes in, 459–​460
(ARFID) Behavioral Inventory of Executive OSFED related to, 39
ADHD with, 431 Function (BRIEF), 399–​400 overweight people with
ON vs., 453 behavioral response shifting, 239 group behavioral weight loss
behavioral weight loss treatment (BWLT) interventions vs. CBT for, 130
B for BED, 378 personal history timeline of
baclofen CBT vs., 276 patient with
for BED, 381–​382 belief(s) example of, 292t–​293t
for BN, 375 religious prevalence of, 38
bariatric surgery(ies) impact on body image and eating psychological comorbidity of, 229–​243,
AN after, 461–​462 concerns, 191–​192 231t–​233t
BED after, 459–​460 Belize affective disorders, 229–​230
BN after, 460–​461 body image in, 201 anxiety disorders, 230
EBs–​and EDs–​related, 458–​469 as one of “fattest nations,” 201 ICDs, 234–​235
food addiction after, 465–​466 BFST. see behavioral family systems personality disorders, 235–​236
GI physiology alterations after, 459 therapy (BFST) substance abuse disorders, 230,
introduction, 458–​459 bias(es) 233–​234
LOC overeating after, 460 attentional risk factors for, 3
NES after, 462 to social and angry threat stimuli and longitudinal studies’ characteristics,
NSRED after, 462 emotion recognition, 240 111, 118–​119
postoperative sequelae, 459 in self-​report in paper-​and-​pencil vs. suicide attempts related to, 39
problematic EBs after, 462–​464 mobile assessments, 498 suicide related to, 39
types of, 458 binding potential treatment of, 4
bariatric surgery patients defined, 68 ADHD–​related drugs in,
EBs and EDs in, 458–​469 (see also binge eating 375–​377, 376t
eating behavior(s) (EBs); eating negative affect and AEDs in, 376t, 379–​381
disorder(s) (EDs); specific disorders BED and, 171–​173 antidepressants in, 376t, 377–​378
and bariatric surgery(ies)) BN and, 167–​169 antipsychotics in, 382
AN, 461–​462 treatment research on, 522 baclofen in, 381–​382
BED, 459–​460 binge eating disorder (BED), 438 chromium in, 382
BN, 460–​461 after bariatric surgery, 459–​460 LDX in, 375–​377, 376t
introduction, 458–​459 DSM-​5 on, 459 opioid antagonists in, 381
NES, 462 prevalence of, 459–​460 pharmacotherapy in, 375–​382, 376t
NSRED, 462 boundary problems related to, 2 (see also specific agents)
problematic EBs, 462–​464 CBT for, 276–​278, 302 weight-​loss drugs in, 376t, 378–​379
food addiction in, 465–​466 efficacy of, 276 weight loss dieting interventions among
postoperative GI problems in, 464–​465 GSH in, 277 persons with, 130

Index 527
binge–​purge behaviors Hollywood vs., 200–​201 CBT-​E, 272–​276
harm avoidance scores in women bone marrow loss comparative treatment research,
with, 237 AN and, 224 273–​274
binging bone metabolism efficacy of, 272–​274
negative affect effects on, 238–​239 AN effects on, 224–​225 generalizability of treatment effects
biomedicine borderline personality disorder (BPD) from controlled research to routine
global proliferation of behaviors associated with, 334–​335 clinical care settings, 274
impact on body image and eating DBT with, 334 GSH in, 275–​276
concerns, 201–​202 efficacy of, 336–​337 physical exercise with, 130
biosocial theory described, 334–​335 predictors and moderators of, 275
affect dysregulation model within, 335 EDs and, 235 RCTs of, 274
of DBT, 334 BPD. see borderline personality treatment model, 271–​272
EDs–​related, 337–​338 disorder (BPD) costs of, 412
for EDs brain course of, 38
adaptations of, 340–​341 starvation and emaciation effects cross-​cultural patterns of, 37–​38
bisexual men on, 47–​48 defined, 460
body image dissatisfaction among, 194 brain imaging studies delayed medial OFC activation
BIVRS (Body Image Virtual Reality in AN and BN, 58–​66 reduction in, 60
Scale), 480 of normal feeding behavior in healthy described, 2, 165–​166
blood oxygen level dependent signal individuals, 60–​61 diagnosis of, 10
defined, 67–​68 brain injury(ies) discovery of, 37
BMI. see body mass index (BMI) CRT for, 396 DSM on, 2
BN. see bulimia nervosa (BN) BRIEF. see Behavioral Inventory of epidemiology of, 37
body Executive Function (BRIEF) genetic influences on, 48, 80–​105
communal socialization of bulimia nervosa (BN). see also eating (see also genetic influences)
impact on body image and eating disorder(s) (EDs) history of, 2
concerns, 191 adolescent hunger and satiety in
body image FBT for, 324–​327 fMRI studies of, 63
in Belize, 201 in adults incidence of, 37
in China, 201 family therapy for, 330 interoception in
cultural influences on, 187–​208 (see also after bariatric surgery, 460–​461 fMRI studies of, 63–​64
culture, impact on body image altered feeding behaviors related to IPT for, 300–​301
and eating concerns) studies of, 48–​49, 49t future directions in, 310–​311
EDs related to anxiety and harm avoidance with in Jamaica, 200
VR in, 478–​484, 481f (see also body PET data correlated with, 57–​58 LOC overeating after, 460
image disturbances, VR in) appetitive regulation in, 47–​79 medical complications of, 225–​226
in Nepal, 200 CCK and, 51, 54t laxative abuse, 226
in United Arab Emirates, 201 fMRI studies of, 61–​66, 62f, 65f self-​induced vomiting, 225–​226
body image dissatisfaction future research directions related monoamine function disturbances
among adolescent girls, 480 to, 66–​67 related to, 53, 55–​58
among bisexual men, 194 gender differences in, 61 DA activity, 53, 55
among gay men, 194 ghrelin and, 52–​53, 55t serotonin, 55–​58
body image distortion HPA axis effects on, 50 mortality rate associated with, 225
AL theory of, 481, 481f images of food in, 62 National Comorbidity Survey–​
body image disturbances interoceptive processing in, 59–​60 Replication Adolescent
VR in, 478–​484, 481f introduction, 47 Supplement on, 37
assessment-​related, 480 leptin and, 51–​52, 54t–​55t negative emotionality and, 165–​169
studies-​related, 479–​480 neurocircuitry of, 58–​60 emotional responses to body image
treatment-​related, 480–​484, 481f neuroendocrine exposure, 166–​167
body image exposure systems–​related, 50–​53 emotional responses to food cues, 166
emotional responses to NPY and, 51, 54t negative affect and binge eating,
AN and, 175–​176 opioid peptides and, 50–​51, 54t 167–​169
BED and, 170–​171 PET studies of, 61–​62 neurobiologic alterations in, 47–​48
BN and, 166–​167 PYY and, 51, 54t neuropeptide and neuroendocrine
Body Image Virtual Reality Scale regional cerebral blood flow studies alterations in, 49–​53, 54t–​55t
(BIVRS), 480 in, 61–​62 novelty-​seeking scores and, 237
body mass index (BMI) SPECT studies of, 61–​62 OBE episodes in, 460–​461
defined, 68 tastes of food in, 62–​63, 62f prevalence of, 37
genetics of, 83 twin studies, 48 psychological comorbidity of, 229–​243,
legislated minimum binging in 231t–​233t
for models, 193 negative affect lability and, 239 affective disorders, 229–​230
Body Project, 255–​256 brain imaging studies in, 58–​66 anxiety disorders, 230
Bollywood CBT for, 271–​276 ICDs, 234–​235

528 Index
 personality disorders, 235–​236 for BN, 370t, 372–​373 Children’s Hospital of Philadelphia, 98
substance abuse disorders, 230, carbonyl-​[ 11C]WAY100635 China
233–​234 defined, 68 body image in, 201
purging behavior in, 225–​226 career support chocolate
research on Internet-​based interventions for EDs emotions and
future directions for, 66–​67 related to, 512–​513 immediate and delayed effects,
response to inhibition alterations in cataloging 163–​164
fMRI studies of, 65–​66 cross-​cultural chocolate “addicts”
response to reward alterations in to global systems and regional trends, guilt after eating chocolate, 157–​158
fMRI studies of, 64–​65, 65f 199–​202 cholecystokinin (CCK)
risk factors for, 3 “cathartic colon” syndrome, 226 AN and BN and, 51, 54t
dietary behaviors as, 136–​137 Caucasians chromium
longitudinal studies’ characteristics, body image and eating concerns for BED, 382
111, 117–​118 among, 197 chromosome(s)
social media in, 262–​263 CBT. see cognitive-​behavioral defined, 80
state vs. trait characteristics in, 47–​48 therapy (CBT) chronic tendency toward
subsystems in, 47 CBT-​E. see enhanced CBT (CBT-​E) overconsumption, 140–​146.
suicide attempts related to, 38 CBT-​Eb. see enhanced CBT (CBT-​E), see also overconsumption, chronic
suicide related to, 38 broad (CBT-​Eb) tendency toward
treatment of, 4 CBT-​Ef. see enhanced CBT (CBT-​E), cisapride
ADHD–​related management in, 374 focused (CBT-​Ef ) for AN, 361t, 365–​366
AEDs in, 370t, 372–​373 CCK. see cholecystokinin (CCK) citalopram
antidepressants in, 369–​372, 370t CD programs. see cognitive dissonance for BN, 370t, 373
antipsychotics in, 375 (CD) programs clarification
5-​HT3 receptor antagonists in, Center for Epidemiological Studies in IPT for EDs, 298–​299
370t, 373 Depression Scale, 163–​164 classification(s)
hormonal agents in, 370t, 373 Center for Medicare Services, 413 defined, 9
lithium in, 370t, 374 central coherence, 396 described, 9
opioid antagonists in, 370t, 373–​374 of CRT for EDs, 403t, 405–​406 of EDs, 9–​23 (see also specific types
pharmacotherapy in, 368–​375, 370t defined, 239–​240 and eating disorder(s) (EDs),
(see also specific agents) examples of, 403t classification of )
prokinetic agents in, 375 weak classification systems
research on, 2, 521–​522 defined, 239–​240 described, 9
weight-​loss drugs in, 370t, 374 central nervous system (CNS) climate
vagal nerve activity in AN effects on, 224 impact on body image and eating
depression related to, 158–​159 cerebral blood flow concerns, 190
weight loss dieting interventions among regional Clinical Global
persons with, 130 in appetitive regulation in AN and Impression-​Improvement scale
bupropion BN, 61–​62 for NES, 382
for BED, 379 defined, 68 clonazepam
burning bridges CET. see cue exposure therapy (CET) for AN, 368
in BED and BN management, 343 CFT. see conjoint family therapy (CFT) clonidine
Buss-​Durkee Scale (BDS), 363 cheerleaders for AN, 368
BWLT. see behavioral weight loss ED prevention in, 260 clothing
treatment (BWLT) chewing marketing of
after bariatric surgery, 464 impact on body image and eating
C Chicago/​Stanford adolescent-​focused concerns, 195
caloric deprivation therapy vs. FBT study, 323 clothing customs
effects on caloric intake Child Eating Behaviour impact on body image and eating
animal studies of, 131 Questionnaire, 420 concerns, 190
caloric intake Child Guidance Center, 319 CNS. see central nervous system (CNS)
caloric deprivation effects on childhood cognitive behavioral therapy (CBT)
animal studies of, 131 overanxious disorder of CRT vs., 397, 398t
longer-​term EDs and, 230 ECT vs., 482
DLW in assessment of, 135 children for EDs, 4, 15, 271–​286, 304 (see also
caloric restriction EDs in specific disorders, e.g., bulimia
short-​term CRT for, 399–​400 nervosa (BN))
relationship to laboratory-​based diagnostic criteria for, 19–​20 AN, 278–​280
eating, 128–​129 IPT for BED, 276–​278
Campaign for Real Beauty, 262 adaptations of, 311 BN, 271–​276
cannabinoids marketing of toys and clothing for brevity of, 281
for AN, 361t, 365 impact on body image and eating clinical range/​reach of, 281
carbamazepine concerns, 195 cost-​effectiveness of, 280–​281

Index 529
cognitive behavioral therapy (CBT) (cont.) impact on body image and eating cross-​cultural patterns
dissemination and implementation concerns, 192 of AN, 35–​36
of, 283 comfort eating of BED, 39
effectiveness of, 280–​281 negative affect impact on, 161–​164 of BN, 37–​38
efficacy of, 278–​280 susceptibility to stress and, 162–​163 CRT. see cognitive remediation
ethnic, racial, and cultural communal socialization of food and body therapy (CRT)
considerations related to, 282 impact on body image and eating CRT Resource Pack, 400
scalability of, 282–​283 concerns, 191 cue exposure therapy (CET)
task-​sharing in, 281–​282 communication for EDs, 475–​478
group behavioral weight loss social cultural influences
intervention vs. as primary function of emotion, on body image and EDs, 187–​208
for overweight BED persons, 130 431–​432 (see also culture, impact on body
Internet-​delivered, 496–​497 communication analysis imaging and eating concerns)
IPT with in IPT for EDs, 299 in CBT for EDs, 282
for AN, 303–​304 compensatory eating disorder, 445–​446 cultural issues
for BED, 302 complication(s) in future versions of DSM, 17–​18
for BN, 300–​301 medical, 222–​225 (see also specific culture
physical exercise with disorders, e.g., anorexia nervosa defined, 187–​188
for BN persons, 130 (AN), medical complications of ) described, 187–​189
cognitive-​behavioral therapy–​enhanced conditioning impact on body image and eating
(CBT-​E). see enhanced CBT evaluative concerns, 187–​208
(CBT-​E) in SE, 432–​434, 432f from cross-​cultural cataloging to
cognitive dissonance (CD) programs, confidentiality global systems and regional trends,
255–​256 Internet-​based interventions for EDs 199–​202
cognitive factors and, 506–​507 finance and, 201–​202
emotional responses to food related of mobile devices and apps for from gender and race to
to, 157 EDs, 499 intersectionality, 196–​198
cognitive flexibility, 239 conflict around food globalization and, 199–​202
Cognitive Flexibility Scale, 402 in SE, 426 global proliferation of biomedicine
cognitive processes conjoint family therapy (CFT) and, 201–​202
as factor in EDs, 239–​240 in FBT-​AN, 321 ideas, ideals, and images in, 193–​195
cognitive remediation therapy (CRT), connection(s) introduction, 187
395–​409 making large-​scale environmental and
for AN in IPT for EDs, 296–​297 political-​economic variables in,
RCTs of, 397 constructs of interest 189–​191
in brain-​injured patients, 396 of EDs media in, 201
CBT vs., 397, 398t to be assessed, 213–​215 practices, behaviors, and habits in,
described, 395 cortisol 191–​193
for EDs, 395–​409 (see also specific defined, 68 shifts in, 196–​202
disorders) cost(s) from socioeconomic status to upward
in child and adolescent populations, in EDs, 410–​415 mobility, 198–​199
399–​400 future directions related to, 414 symbolic body capital in, 195–​196
in clinical settings, 400–​401 ICD-​9 on, 411 toward increased multidisciplinary
examples of, 403t–​404t individual cost estimates, 410–​411 collaborative research on, 202–​203
experimental measures, 404–​406 introduction, 410 D-​cycloserine
future work in, 406 national cost estimates, 411–​412 for AN, 368
historical development of, 396–​397 per-​patient financial costs, 411 cyproheptadine
introduction, 395 personal costs, 410–​411 for AN, 361t, 365
outcome measurements, 402–​406 cost–​benefit issues
outline of typical therapy session, in ED prevention, 263–​264 D
401–​402, 403t–​404t cost-​effectiveness DA. see dopamine (DA)
RCTs in adult populations, 397–​399 in EDs, 410–​415 dancer(s)
self-​report measures, 402 future directions related to, 414 ED prevention in, 260
as therapy enhancer, 397, 398t treatment-​related, 412–​414 Danish Twin Registry, 87
indications for, 396 cost utility DBT. see dialectical behavior
introduction, 395 defined, 414 therapy (DBT)
cognitive restraint of ED treatments, 414 DBTgsh. see DBT–​guided self-​help
overeating during stress related to, 162 course (DBTgsh)
cognitive systems of EDs (see specific disorders) DBT–​guided self-​help (DBTgsh), 339
EDs and, 27–​28 cross-​cultural cataloging decided preferences
coherence to global systems and regional trends in SE, 426
central (see central coherence) as factor in body image and eating decubitus ulcers
college campus concerns, 199–​202 AN and, 225

530 Index
deficit(s) dialectical behavior therapy (DBT), empirical tests of, 127–​128
interpersonal, 289 334–​350 future directions in, 146–​147
dehydroepiandrosterone (DHEA) adaptations of implications regarding possible
for AN, 361t, 367 for EDs, 338–​340 explanations for inconsistent
delayed gastric emptying appetite-​focus, 339 findings, 138–​146
AN and, 223 balance between change and acceptance incompatible study findings, 131–​138
delayed medial orbitofrontal cortex in, 335–​336 introduction, 126–​127
(mOFC) activation reduction biosocial theory of, 334 prospective studies, 128
in BN, 60 with BPD, 334 prospective studies vs.
demand characteristics efficacy of, 336–​337 experiments, 132
reductions in bulimic symptoms in case management strategies in, 336 reductions in bulimic symptoms in
experimental trials due to, 138 core strategies in, 336 experimental trials due to demand
Department of Health and Human described, 334, 335, 337–​338 characteristics, 138
Services (HHS) dialectic framework within, 334 researcher’s use of dietary restraint
under HITECH Act, 499 for EDs, 337–​350 measures, 134–​136
depression adaptations of, 338–​340 theoretical mechanisms of, 127
BN–​related vagal nerve activity and, adaptations of biosocial theory in, trials evaluating interventions seeking
158–​159 340–​341 to manipulate, 130–​131
CRT for, 396 affect regulation model, 340 ED symptoms related to, 127
EDs and, 164 BED and BN, 343 in Fiji, 200
treatment of biosocial theory in, 337–​338 longer-​term
artificial cervical VNS in, 158 reasons for, 337–​338 ED symptoms related to, 129–​131
dermatologic complications Stanford model, 338–​339, 341–​343, real-​world
of AN, 225 342t, 344t–​346t (see also Stanford dietary restriction interventions may
desipramine DBT model) be unrepresentative of, 133–​134
for BN, 370t, 374 emphasis of, 334 dieting theory of eating pathology
Detail and Flexibility Questionnaire functions of, 335 empirical tests of, 127–​128
(DFlex), 402 introduction, 334 DIS. see Dietary Intent Scale (DIS)
dexfenfluramine modes of treatment in, 335 disease context
for BED, 376t, 379 modules in, 335 impact on body image and eating
for BN, 370t, 374 stages of, 335 concerns, 191
DFlex. see Detail and Flexibility standard treatment with, 334–​336 disgust
Questionnaire (DFlex) stylistic strategies in, 336 recontextualization in approach to, 433
DHEA. see dehydroepiandrosterone diarrhea disgust experience
(DHEA) AN and, 223–​224 in phenomenology of SE, 427–​429,
diabetics dietary behavior(s) 428f, 430f
ED prevention in, 261 as risk factor for BN, 136–​137 in SE, 432–​434, 432f
diagnosis Dietary Intent Scale (DIS), 134, 135 dissonance theory
of EDs dietary restraint in ED prevention programs, 251
self-​report questionnaires for, 217 dieting vs., 127 distributed cloud computing, 493
Diagnostic and Statistical Manual of Mental researcher’s use of measures in DLW. see doubly labeled water (DLW)
Disorders (DSM) validity of, 134–​136 domains of interest
future versions of dietary restriction(s) of EDs
cultural issues in, 17–​18 may be unrepresentative of real-​world to be assessed, 213–​215
Diagnostic and Statistical Manual of Mental dieting, 133–​134 dopamine (DA)
Disorders, Fifth Ed. (DSM-​5) relative vs. absolute, 136 EDs related to, 53, 55
on BED after bariatric surgery, 459 dieting in set-​shifting tasks, 240
on CBT for EDs, 271 defined, 127 stress in reducing, 161
comorbidity eating disorder studies described, 127 in “wanting” of rewards, 160–​161
in, 236 dietary restraint vs., 127 dopamine D2 receptor (DRD2)
Eating Disorders Workgroup of EDs related to, 126–​154 (see also in BED, 160–​161
on OSFED, 39 specific disorders, e.g., bulimia dopamine D2 receptor (DRD2) deficiency
on EDs, 2–​3, 9–​23 nervosa (BN)) obesity related to, 161
external validators–​related, 13 animal studies of, 131 dopamine (DA) receptor binding
EDs classification of, 25 chronic tendency toward alterations
Diagnostic and Statistical Manual of overconsumption, 140–​146 (see in EDs, 53, 55
Mental Disorders, Fourth Ed. also overconsumption, chronic doubly labeled water (DLW)
(DSM-​IV) tendency toward) in longer-​term caloric intake
on EDs, 2–​3 dietary behaviors increasing risk for assessment, 135
psychological bulimic pathology, 136–​137 DOVE, 262
comorbidity–​related, 229 dietary restriction interventions may DRD2. see dopamine D2 receptor (DRD2)
dialectical abstinence be unrepresentative of real-​world DRES. see Dutch Restrained Eating
in BED and BN management, 343 dieting, 133–​134 Scale (DRES)

Index 531
DSM. see Diagnostic and Statistical future research directions related CRT for, 395–​409 (see also cognitive
Manual of Mental to, 66–​67 remediation therapy (CRT))
Disorders (DSM) gender differences in, 61 cultural influences on, 187–​208 (see also
DSM-​5. see Diagnostic and Statistical images of food in, 62 culture, impact on body image
Manual of Mental Disorders, Fifth interoceptive processing in, 59–​60 and eating concerns)
Ed. (DSM-​5) introduction, 47 DBT for, 337–​350 (see also dialectical
DSM-​IV. see Diagnostic and Statistical neurocircuitry of, 58–​60 behavior therapy (DBT), for EDs)
Manual of Mental Disorders, PET studies of, 61–​62 depression and anxiety with, 164
Fourth Ed. (DSM-​IV) regional cerebral blood flow studies diagnosis of
dumping syndrome in, 61–​62 DSM-​5 on, 10–​11
after bariatric surgery, 464 SPECT studies of, 61–​62 overlap in, 12–​13
Dutch Restrained Eating Scale (DRES), tastes of food in, 62–​63, 62f dieting and, 126–​154 (see also specific
134, 135 approaches to understanding, 45–​208 disorders and dieting, EDs
dysphagia assessment of, 209–​221 related to)
BN and, 225 mobile devices and apps in, DSM-​5 on, 9–​23
492–​504 (see mobile devices emerging syndromes, 438–​457 (see also
E and applications (apps), in EDs specific disorders and emerging
early adolescents assessment and treatment) syndromes)
ED–​related diagnostic criteria technology-​based, 4–​5 emotions and (see also emotion(s))
for, 19–​20 attitudes and behaviors associated epidemiology of, 34–​43 (see also specific
EAST. see Extrinsic Affective Simon with, 247 disorders)
Task (EAST) atypical AN, 448–​451 introduction, 34
EAT. see Eating Attitudes Test (EAT) behavioral, temperamental, and on Facebook, 263
eating personality factors associated with family and genetic studies of, 3, 81
comfort genetics of, 83–​84 family therapy in, 319–​333 (see also
negative affect impact on, 161–​164 boundary problems related to, 2–​3 family therapy, for EDs)
susceptibility to stress and, 162–​163 categories of, 10 in Fiji, 200
effects on emotions causes of functional and task activation studies
EDs related to, 158–​159 family environment, 320–​321 in, 61–​66, 62f, 65f
emotional CBT for, 271–​286 (see also specific genetic influences on, 80–​105 (see also
after bariatric surgery, 463 disorders and cognitive behavioral genetic influences)
vs. restrained eating, 162 therapy (CBT), for EDs) GWAS in, 98–​99
laboratory-​based classification of, 9–​23 history of, 1–​2
relationship of dieting to, 128–​129 alternative diagnostic model, 13–​16 hunger and satiety in
mindful boundary problems related to, 2–​3 fMRI studies of, 63
in BED and BN management, 343 controversies related to, 17–​20 Internet-​based interventions for,
restrained vs. emotional, 162 cultural issues in, 17–​18 505–​519 (see also Internet-​based
selective, 419–​437 (see also selective DSM-​5, 9–​23, 25 interventions, for EDs)
eating (SE)) future directions in, 17–​20 interoception in
slowness in intradiagnostic heterogeneity, 16 fMRI studies of, 63–​64
in SE, 426–​427 introduction, 9 interpersonal model for, 288–​289
sweet overlap between diagnostic entities introduction, 1–​5
after bariatric surgery, 463–​464 in, 12–​13 IPT for, 4, 287–​318 (see also
eating attitudes research-​related criteria in, 17 interpersonal psychotherapy
negative affect impact on, 161–​164 separate diagnostic criteria (IPT), for EDs)
Eating Attitudes Test (EAT), 217 for children and early JIT interventions for, 497
eating behavior(s) (EBs) adolescents, 19–​20 longer-​term dieting related to
after bariatric surgery, 458–​469 separate diagnostic criteria for men symptoms of
eating disorder(s) (EDs). see also pica; and women, 18–​19 experimental studies of, 129–​131
specific types, e.g., bulimia statistical approaches to, 11–​12 low self-​directedness and, 238
nervosa (BN) taxonomy in, 9 mental disorders with, 164
adolescent TDM, 13–​16, 14f models of
family therapy for, 320–​321 transdiagnostic model, 15–​16 objectification theory, 480
in adults common comorbidity profile of, 13 monoamine function disturbances
family therapy for, 329–​330 compensatory, 445–​446 related to, 53, 55–​58
affective disorders with, 164 complexity of, 520 DA activity, 53, 55
after bariatric surgery, 458–​469 (see also conceptualization of DA receptor binding
bariatric surgery patients; eating RDoC on, 1, 24–​33 (see also research alterations, 53, 55
behavior(s) (EBs); specific disorders domain criteria (RDoC)) serotonin, 55–​58
and bariatric surgery(ies)) cost-​effectiveness, 410–​415 (see also moods and, 155–​186 (see also mood(s))
alexithymia and, 176 cost(s); cost-​effectiveness) negative affect effects on, 161–​164
appetitive regulation in, 47–​79 course of, 34–​43 (see also specific negative emotionality and, 164
fMRI studies of, 61–​66, 62f, 65f disorders) NES, 438–​444, 440t, 443f

532 Index
nosological issues related to, 11–​12 standalone eating disorder not otherwise specified,
ON, 451–​453, 452t (see also orthorexia variants of, 438 purging type (EDNOS-​P). see
nervosa (ON)) symptoms of purging disorder (PD)
PD, 438, 445–​448 IPT in redirecting issues related Eating Disorders Inventory-​3, 217
peak onset of, 247 to, 297 Eating Disorders Inventory subscales, 12
persistence and, 237 systematic study of Eating Disorders Workgroup
personality and beginnings of, 2 of DSM-​5
relationships between, 238 trait anxiety and, 237 on OSFED, 39
pharmacotherapy of, 359–​394 (see transdiagnostic theories of, 252 Eating Disorders Working Group
also specific agents, disorders, and treatment of, 3–​4 (see also specific types) of Psychiatric Genomics
pharmacotherapy, of EDs) basic science contributions to, Consortium, 99
prevalence of, 126 520–​521 Eating Inventory, 447
treatment difficulties related to, 247 CBT-​E in, 214 eBody project program, 511
prevention of, 3, 247–​270 (see also CBT in, 4, 15, 214, 304 EBs. see eating behavior(s) (EBs)
ED prevention; ED prevention cost-​effectiveness of, 412–​414 ecological momentary assessment
programs; prevention, of EDs) costs utility of, 414 (EMA), 495
prevention programs for (see ED CRT in, 395–​409 (see also specific on negative affect, 238–​239
prevention programs) disorders and cognitive remediation ECT. see experiential cognitive
psychoeducation for therapy (CRT)) therapy (ECT)
apps for, 496 discrimination and implementation EDDS. see Eating Disorder Diagnostic
psychological assessment of, 211–​221 of, 522 Scale (EDDS)
(see also assessment(s), of EDs) IPT in, 4, 287–​318 EDE. see Eating Disorder
psychological comorbidity of, 229–​243, limitations of conventional, 505–​506 Examination (EDE)
231t–​233t mobile devices and apps in, 492–​ edema formation
affective disorders, 229–​230 504 (see also mobile devices BN and, 226
anxiety disorders, 230 and applications (apps), in EDs EDE-​Q. see Eating Disorder
DSM-​IV on, 229 assessment and treatment) Examination–​Questionnaire
DSM-​5 on, 236 modality selection for, 304 (EDE-​Q)
ICD-​10 on, 236 outpatient, 351 EDEQ-​R scale. see Eating Disorder
ICDs, 234–​235 plateau in development of, 521–​522 Examination–​Questionnaire-​
introduction, 229 self-​report questionnaires in, 217–​218 Restraint (EDEQ-​R) scale
perfectionism, 236–​238 technologies in, 522–​523 EDE-​R scale. see Eating Disorder
personality disorders, 235–​236 technology-​based, 4–​5 Examination–​Restraint
personality traits–​related, 236–​240 VR for, 470–​491 (see also virtual (EDE-​R) scale
substance abuse disorders, 230, reality (VR)) EDI-​2. see Eating Disorder Inventory-​2
233–​234 Eating Disorder Diagnostic Scale (EDI-​2)
temperament, 236–​238 (EDDS), 217 “EDINA,” 510
questions related to, 520 Eating Disorder Examination EDI scores. see Eating Disorder Inventory
recent changes related to, 2 (EDE), 447 (EDI) scores
register studies of in EDs assessment, 216 EDNOS. see eating disorder not otherwise
results of risk factors and markers Eating Disorder Examination–​ specified (EDNOS)
from, 119–​120 Questionnaire (EDE-​Q), EDNOS-​P. see eating disorder not
relationships between personality/​ 217–​218, 328 otherwise specified, purging type
temperament and Eating Disorder Examination–​ (EDNOS-​P)
perspectives on, 238 Questionnaire-​Restraint (EDEQ-​ ED prevention, 247–​270. see also ED
research on R) scale, 134, 135 prevention programs
future directions for, 66–​67 Eating Disorder Examination–​Restraint AN, 263
response to inhibition alterations in (EDE-​R) scale, 135 cost–​benefit issues related to, 263–​264
fMRI studies of, 65–​66 Eating Disorder Inventory current status and underlying theory,
response to reward alterations in subscales of, 328 247–​270
fMRI studies of, 64–​65, 65f Eating Disorder Inventory-​2 (EDI-​2), 363, dissemination/​implementation in,
risk factors for, 3, 106–​125 (see also 364, 449 263–​264
specific types, disorders, and risk Interpersonal Distrust Subscale of, 364 future directions in, 264–​265
factor(s), for EDs) Eating Disorder Inventory 12 score, 378 as harmful, 253
self-​monitoring of Eating Disorder Inventory (EDI) IPT development in, 311–​312
self-​report questionnaires scores, 360 obesity prevention and, 259–​260
for, 218 eating disorder not otherwise specified programs for (see ED prevention
serotonin receptor binding alterations (EDNOS), 438 programs)
in, 56–​58 boundary problems related to, 2–​3 public health/​policy and mass media
SH and GSH for, 351–​358 (see also neuroticism and, 119 models related to, 261–​262
guided self-​help (GSH), for EDs; risk factors for screening in, 249–​250
self-​help (SH), for EDs) longitudinal studies’ in specific settings and special
sleep-​related, 442 characteristics, 119 populations, 260–​263

Index 533
ED prevention (cont.) EDs and, 155–​186 environmental variables
theories and models of interventions in, BED, 169–​173 large-​scale
250–​252 BN, 165–​169 (see also bulimia impact on body image and eating
dissonance theory, 251 nervosa (BN), negative concerns, 189–​191
feminist theory, 251 emotionality and) EOIT. see ego-​oriented individual
media literacy and advocacy, 251–​252 clinical evidence, 164–​177 therapy (EOIT)
psychoeducation, 250 cognitive factors in, 157 epidemiology
social learning theory, 250–​251 dimensions of, 156, 157f of EDs, 34–​43 (see also specific disorders,
ED prevention programs, 247–​270. see also five-​way model of, 155–​158, 156f e.g., anorexia nervosa (AN))
ED prevention future directions in, 178 epigenesis
categories of, 247–​248 introduction, 155–​158, 156f, 157f described, 80
CD programs, 255–​256 meal size, timing, and habit effects erythromycin
cost–​benefit issues related to, 263–​264 on, 159 for BN, 375
dissemination/​implementation in, mechanisms associated with, 158–​164 esophagus
263–​264 predicted changes related to, Barrett’s
effective 155–​156, 156f BN and, 225
examples of, 255–​260 hunger and eating effects on, 158–​159 ESP
effectiveness of, 252–​253 introduction, 155–​158, 156f, 157f in EDs assessment, 217
future directions in, 264–​265 moods vs., 156 ESS-​KIMO, 512, 513
as harmful, 253 negative affect effects on, 161–​164 “ES[S]‌PRIT,” 512
indicated prevention programs, 248 stress susceptibility and, 162–​163 estimation
moderators and mediators of, 253–​254 neural substrates shared by sensory examples of, 404t
in obesity prevention, 259–​260 reward impact on, 159–​161 estradiol
peer support/​school-​based programs, social communication as primary for AN, 361t, 367
258–​259 function of, 431–​432 ethical concerns
risk factors informing, 249 emotional eating Internet-​based interventions for EDs–​
StudentBodies program, 248, 250, 254, after bariatric surgery, 463 related, 506–​507
256–​258, 263 restrained eating vs., 162 ethical issues
targeted or selective prevention Emotional Eating Scale, 339 in CBT for EDs, 282
interventions, 248 emotional experiences ethnoscape(s), 200–​201
theories and models of, 250–​252 potentiation of evaluative conditioning
dissonance theory, 251 sensory sensitivities in SE and, 430 in SE, 432–​434, 432f
feminist theory, 251 emotionality Evolution, 262
media literacy and advocacy, 251–​252 negative (see negative emotionality) executive functioning
psychoeducation, 250 emotional processing constructs sensory sensitivities and
social learning theory, 250–​251 types of, 240 in SE, 431
universal prevention program, 247–​248 emotional processing deficits exercise(s)
EDs. see eating disorder(s) (EDs) AN and, 176–​177 CBT with
ego-​oriented individual therapy emotional vulnerability in BN management, 130
(EOIT), 322 defined, 334–​335 experiential cognitive therapy (ECT), 482
EMA. see ecological momentary emotion-​focused therapies CBT vs., 482
assessment (EMA) for EDs, 337–​350 (see also dialectical experimental measures of outcome
emaciation behavior therapy (DBT)) of CRT for EDs, 404–​406
brain effects of, 47–​48 emotion recognition exploratory questions
embodied technology attentional biases to, 240 in IPT for EDs, 298
VR as, 473 emotion regulation strategies, 240 Extrinsic Affective Simon Task (EAST),
emerging syndromes, 438–​457. see also endocrine system 142, 143
specific disorders, e.g., purging AN effects on, 224
disorder (PD) endogenous opioid neuropeptides F
ON, 451–​453, 452t (see also orthorexia positive mood related to, 159–​160 Facebook
nervosa (ON)) endophenotype(s) EDs on, 263
atypical AN, 448–​451 described, 80 factor mixture modeling (FMM)
future directions in, 453–​454 enhanced CBT (CBT-​E), 214, in ED classification, 11–​12
introduction, 438–​439 272–​276, 413 family(ies)
NES, 438–​444, 440t, 443f for AN, 278–​280 EDs “running” in, 81
PD, 438, 445–​448 broad (CBT-​Eb), 272 IPT for
status of focused (CBT-​Ef ), 272 adaptations of, 311
research needed to clarify, 443–​445 “enjoyment of food” measure, 160–​161 family-​based therapy (FBT). see also
emotion(s). see also mood(s) environment(s) family therapy
chocolate and family AFT vs., 323
immediate and delayed effects, in etiology of EDs, 320–​321 for EDs, 319–​333 (see also family
163–​164 mealtime therapy, for EDs)
defined, 156 in SE management, 434–​435, 434t family environment

534 Index
 as factor in EDs, 320–​321 feeding behaviors BN and, 166
lack of evidence to support causal altered food deprivation
role of, 82 in AN and BN, 48–​49, 49t effects of
family meals normal studies of, 48–​49, 49t
SE effects on, 423 brain imaging studies of, 60–​61 food insecurity
family studies regulation of impact on body image and eating
of EDs, 3 neuropeptides in, 49–​53, 54t–​55t concerns, 189–​190
family therapy feeding disorders food neophobia
conjoint, 321 DSM-​5 on, 10 in SE, 425
for EDs, 319–​333 feminist theory food refusal
acceptability of, 327 in ED prevention programs, 251 in SE, 426
adolescent AN, 321–​322 (see also fenfluramine food reward
family therapy for adolescent AN for BN, 374 greater anticipatory
(FBT-​AN)) Fiji chronic tendency toward
adolescent BN, 324–​327 (see also dieting in, 200 overconsumption related to,
family therapy for adolescent BN EDs in, 200 141–​144
(FBT-​BN)) finance greater consummatory
in adults, 329–​330 impact on body image and eating chronic tendency toward
future directions in, 330–​331 concerns, 201–​202 overconsumption related to,
history of, 319–​320 First International Night Eating 140–​141
theoretical model of, 320–​321 Symposium, 439, 440t Food Scientist, 434
treatment manual for, 322–​324 5-​HT. see 5-​hydroxytryptamine (5-​HT) food variety
separated, 321 5-​HT3 receptor antagonists in SE, 424–​425
family therapy for adolescent AN (FBT-​ for BN, 370t, 373 FPT. see focal psychodynamic
AN), 321–​322 fluid consumption therapy (FPT)
acceptability of, 327 high-​calorie Fragmented Pictures Task, 405
BFST in, 322 after bariatric surgery, 463 frontal operculum, 59
CFT in, 321 fluorodeoxyglucose Frost Multidimensional Perfectionism
EOIT in, 322 defined, 68 Scale, 236
first family therapy trial outside UK, fluoxetine functional and task activation studies
321–​322 for AN, 361t, 362 in EDs, 61–​66, 62f, 65f
French study, 323 for BN, 369, 370t functional magnetic resonance
Melbourne study, 324 flutamide imaging (fMRI)
multifamily therapy, 327–​329 for BN, 370t, 373 affect regulation–​related
seminal study, 321, 322 fluvoxamine in BN, 239
SFT in, 321 for BN, 370t, 371 in AN, 58–​60
six-​site study of FBT and SFT, 323–​324 FMM. see factor mixture in BN, 58–​60
Stanford dosage study, 323 modeling (FMM) functional magnetic resonance imaging
Sydney study, 323 fMRI. see functional magnetic resonance (fMRI) studies
treatment manual for, 322–​324 imaging (fMRI) of appetitive regulation in AN and BN,
without prior hospitalization, 321 focal psychodynamic therapy (FPT), 413 61–​66, 62f, 65f
family therapy for adolescent BN (FBT-​ food of normal feeding behavior in healthy
BN), 324–​327 communal socialization of individuals, 60
studies of, 325–​327 impact on body image and eating fussy eating, 419–​437. see also selective
fasting concerns, 191 eating (SE)
as part of rituals in Judaism, 201 conflict around
“fattest nations” in SE, 426 G
Belize as one of, 201 fatty gambling
fatty food in pain alleviation, 160 pathological, 234
in pain alleviation, 160 images of gastric emptying
FBT. see family-​based therapy (FBT) in appetitive regulation in AN delayed
FBT-​AN. see family therapy for adolescent and BN, 62 AN and, 223
AN (FBT-​AN) tastes of gastroesophageal reflux (GERD)
FBT-​BN. see family therapy for adolescent in appetitive regulation in AN and in SE, 430–​431
BN (FBT-​BN) BN, 62–​63, 62f gastrointestinal (GI) problems
FDA. see Food and Drug food addiction after bariatric surgery, 464–​465
Administration (FDA) after bariatric surgery, 465–​466 gastrointestinal (GI) system
FEATBACK, 414, 512 Food and Drug Administration AN effects on, 223–​224
Federal Drug Administration (FDA) (FDA), 158 gay male subcultures
on BN treatment, 522 food cues impact on body image and eating
feeding emotional responses to concerns, 194
of infants AN and, 174–​175 gay men
“on demand” vs. “on schedule,” 191 BED and, 169–​170 body image dissatisfaction among, 194

Index 535
[G]‌EFT. see The (Group) Embedded Girl Scouts, 262 GWAS. see genomewide association
Figures Task ([G]EFT) globalization study (GWAS)
gender characterizations of
as factor in body image and eating body image and EDs related to, H
concerns, 190–​191, 196–​198 199–​202 habit(s)
as factor in response to liquid meal technological advancements and, impact on body image and eating
during hunger or satiation, 61 201–​202 concerns, 191–​193
gene(s) global systems HAM-​A scores. see Hamilton Anxiety
EDs related to, 80–​105 (see also genetic from cross-​cultural cataloging to Scale (HAM-​A) scores
influences) as factor in body image and eating Hamilton Anxiety Scale (HAM-​A) scores,
identification of, 98–​99 concerns, 199–​202 360, 362
gene–​environment interplay goal(s) Hamilton Depression Rating Scale, 377
in EDs, 84–​85 in IPT for EDs, 295–​296 harm avoidance
complexities of, 83–​85 grazing AN and BN and
twin studies of, 86–​98 (see also twin after bariatric surgery, 462–​463 PET data correlated with, 57–​58
studies, of gene–​environment greater anticipatory food reward harm avoidance scores
interplay) chronic tendency toward in women with binge–​purge
General Health Questionnaire, 117 overconsumption related to, behaviors, 237
genetic(s) 141–​144 harm reduction
of BMI, 83 greater consummatory food reward in ED prevention, 248–​249
Genetic Consortium for Anorexia chronic tendency toward Health and Human Services (HHS)
Nervosa, 98 overconsumption related to, under HITECH Act, 499
genetic influences 140–​141 Health Apps Library, 499
on behavioral disturbances, 81 greater impulsivity Health Information Technology
on EDs, 80–​105 chronic tendency toward Economic and Clinical Health Act
AN and BN, 48 overconsumption related to, (HITECH), 493
behavioral, temperamental, and 144–​145 “Health of the 51%: Women”
personality factors, 83–​84 grief, 289 from NHS, 283
BMI, 83 IPT for, 294, 294t heart
consistent evidence for, 82 group(s) AN effects on, 222–​223
future directions in, 100 in IPT for EDs, 300 Helping, Encouraging, Listening and
gene–​environment interplay in, growth Protecting Peers (HELPP)
83–​85 (see also gene–​environment stunted initiatives, 251
interplay, in EDs) SE and, 422 HELPP initiatives. see Helping,
interactions between specific variants GSH. see guided self-​help (GSH) Encouraging, Listening and
and environments, 99 guided self-​help (GSH) Protecting Peers (HELPP)
lack of evidence to support causal in CBT initiatives
role of family environment in, 82 for BED, 277 heritability estimates
molecular genetic studies, 98–​99 for BN, 275–​276 in twin studies of gene–​environment
overview, 81–​82 DBT–​, 339 interplay, 86–​92, 88t–​91t, 93t–​95t
recognition of, 82 for EDs, 351–​358 HHS. see Health and Human
studies of, 3 apps for, 496–​497 Services (HHS)
terminology related to, 80–​81 described, 352 high-​calorie fluid consumption
genome(s) future directions in, 355–​356 after bariatric surgery, 463
defined, 81 Getting Better Bit(e) by Bit(e) in, 353 HITECH. see Health Information
genomewide association study (GWAS) introduction, 351–​352 Technology Economic and
in EDs, 98–​99 Overcoming Binge Eating in, 352–​353 Clinical Health Act (HITECH)
genotype(s) psychoeducational videotapes HITECH Act
defined, 81 in, 353 HHS under, 499
genotype–​environmental interactions systematic reviews and meta-​analyses HIV
in EDs, 84–​85 related to, 353 impact on body image and eating
GERD. see gastroesophageal treatment predictors, moderators, concerns, 191
reflux (GERD) and mediators of, 354 Hollywood
Getting Better Bit(e) by Bit(e) via Internet-​based interventions, Bollywood vs., 200–​201
in SH for EDs, 353 508–​510 hormonal agents
ghrelin Guided Self-​Help for Bulimia Nervosa, for AN, 361t, 366–​367
AN and BN and, 52–​53, 55t Therapist’s Manual, 509 for BN, 370t, 373
ghrelin agonists Guidelines for the Practice of HPA axis. see hypothalamic-​pituitary-​
for AN, 361t, 367 Telepsychology, 506 adrenal (HPA) axis
GI. see gastrointestinal (GI) guilt HTC, 474
girl(s) of chocolate “addicts” after eating hunger
adolescent chocolate, 157–​158 in EDs
body dissatisfaction among, 480 gustatory cortex, 59 fMRI studies of, 63

536 Index
 effects on emotions instrumentation CBT with
EDs related to, 158–​159 in EDs assessment, 215–​218 for AN, 303–​304
5-​hydroxytryptamine (5-​HT) insula for BED, 302
defined, 81 anterior, 59 for BN, 300–​301
hyperlearning, 239 intelligence defined, 287
hypokalemia artificial, 493 described, 287
BN and, 225, 226 values embedded in, 499 in eating-​and weight-​related problems
hyponatremia intermittent explosive disorder, 234 prevention
BN and, 225, 226 International Affective Picture System development of, 311–​312
hypophosphatemia (IAPS), 175 for EDs, 4, 287–​318
refeeding International Classification of Diseases, 9th adolescent and child/​parent
AN and, 223 ed. (ICD-​9) adaptations of, 311
hypothalamic-​pituitary-​adrenal on per-​patient financial costs of AN, 303–​304
(HPA) axis EDs, 411 basic concepts, 289–​290
AN and BN effects on, 50 International Classification of Diseases, 10th BED, 301–​303, 310–​311
rev. (ICD-​10) BN, 300–​301, 310–​311
I comorbidity eating disorder studies clarification in, 298–​299
IAPS. see International Affective Picture in, 236 communication analysis in, 299
System (IAPS) International Journal of Eating Disorders, 2 diagnosis and assignment of sick role
IAT. see Implicit Association Test (IAT) Internet-​based interventions in, 290, 291t
ICD-​9. see International Classification of for EDs, 505–​519 disseminating and implementing, 312
Diseases, 9th edition (ICD-​9) as adjunct to treatment and aftercare empirical literature relevant to,
ICD-​10. see International Classification of interventions, 510 300–​310, 308t
Diseases, 10th revision (ICD-​10) benefits of, 506 encouraging affect in, 298
ICD not otherwise specified, 234 career support–​related, 512–​513 exploratory questions in, 298
ICDs. see impulse control confidentiality issues related to, focusing on goals in, 295–​296
disorders (ICDs) 506–​507 future directions in, 310–​312
iCounselor, 496 ethical concerns related to, 506–​507 general therapeutic techniques in,
idea(s) evaluation criteria for, 513–​514 297–​298
impact on body image and eating GSH and unguided SH, 508–​510 grief-​related, 294, 294t
concerns, 193–​195 implications of, 514–​515 group in, 300
ideal(s) introduction, 505 implementation of, 290–​300,
anti-​ideals vs., 194 methodological aspects of, 515–​516 291t–​294t
impact on body image and eating potential of, 505–​507 initial phase, 290–​291, 291t–​294t
concerns, 193–​195 prevention-​related, 510–​512 intermediate phase, 293–​299, 294t
image(s) psychotherapy, 507–​508 interpersonal deficits–​related,
impact on body image and eating reflections on state of science related 294t, 295
concerns, 193–​195 to, 520–​523 interpersonal formulation in, 291
Implicit Association Test (IAT), 142 types of, 507–​513 interpersonal inventory in, 290–​291,
impulse control disorders (ICDs) Internet-​based subcultures 291t–​293t
characteristics of, 234 impact on body image and eating interpersonal model, 288–​289
classification of, 234 concerns, 193 interpersonal problem areas, 289–​290
EDs and, 234–​235 Internet-​delivered cognitive behavioral interpersonal role disputes–​related,
impulsivity therapy (CBT), 496–​497 294t, 295
greater “Internet-​of-​things” making connections in, 296–​297
chronic tendency toward developments in, 499 outcome studies of, 300–​310, 308t
overconsumption related to, interoception problem areas to be addressed in,
144–​145 in EDs 294–​295, 294t
“IN@”, 510 fMRI studies of, 63–​64 redirecting symptom-​related issues
indicated prevention programs, 248 interoceptive processing in, 297
infant feeding in appetite regulation, 59–​60 roles transitions–​related,
“on demand” vs. “on schedule,” 191 interpersonal deficits, 289 294–​295, 294t
ingestion(s) IPT for, 294t, 295 termination phase, 299–​300
nocturnal, 439 Interpersonal Distrust Subscale therapeutic relationship in, 299
inhibition of EDI-​2, 364 therapeutic stance in, 295
response interpersonal formulation therapeutic strategies, 295–​299
EDs–​related alterations in, 65–​66 in IPT for EDs, 291 treatment structure, 290
norepinephrine in, 240 interpersonal inventory in excessive weight gain prevention,
insecurity(ies) described, 313 305–​310, 308t
food-​related in IPT for EDs, 290–​291, 291t–​293t interpersonal theory of, 287–​288
impact on body image and eating interpersonal model for EDs, 288–​289 introduction, 287
concerns, 189–​190 interpersonal problem areas, 289–​290 interpersonal role disputes, 289
Institute of Psychiatry, 320 interpersonal psychotherapy (IPT), 214 IPT for, 294t, 295

Index 537
interpersonal theory described, 313 SE effects on, 423
of IPT, 287–​288 example of, 291, 292t–​293t in SE management, 434–​435, 434t
intersectionality Life Smart, 251–​252 media
from gender and race to linkage impact on body image and eating
shifts in cultural factors in body defined, 81 concerns, 193–​195, 201
image and EDs and, 196–​198 liraglutide literacy and advocacy of
interview(s) for BED, 378 in ED prevention programs, 251–​252
in EDs assessment, 215–​217 lithium in promoting AN and BN, 262–​263
semistructured, 216 for AN, 361t, 367 medial orbitofrontal cortex (mOFC), 170
unstructured, 216 for BN, 370t, 374 medial prefrontal cortex (mPFC), 59
intradiagnostic heterogeneity LOC eating patterns. see loss of control Media Smart, 251–​252
of EDs, 16 (LOC) eating patterns medical conditions
IPT. see interpersonal longer-​term caloric intake sensory sensitivities and
psychotherapy (IPT) assessment of SE related to, 430–​431
DLW in, 135 Medical Expenditure Panel Survey, 411
J longer-​term dieting men
Jamaica ED symptoms related to bisexual
AN and BN in, 200 experimental studies of, 129–​131 body image dissatisfaction
JIT interventions. see just-​in-​time (JIT) loss of control (LOC) eating patterns among, 194
interventions after bariatric surgery, 460 gay
job subcultures IPT in prevention of, 305–​310, 308t body image dissatisfaction
impact on body image and eating lower esophageal sphincter among, 194
concerns, 192 self-​induced vomiting effects on, 225 mental disorders
Judaism lung(s) EDs and, 164
fasting as part of rituals in, 201 AN effects on, 223 mental health
just-​in-​time (JIT) interventions mobile, 493
for EDs, 497 M in clinical care, 499–​502, 501f
MAEDS scale. see Multifactorial metabolic acidosis
K Assessment of Eating Disorder non-​gap
kleptomania, 234 Symptoms (MAEDS) scale BN and, 226
magnetic resonance imaging (MRI) Metacognitions Questionnaire, 450
L functional (see functional magnetic methylamphetamine
laboratory-​based eating resonance imaging (fMRI)) for BN
relationship of dieting to major depressive disorder ADHD–​related, 374
prospective studies of, 128–​129 alcohol abuse/​dependence disorder “metrosexual”
LAGB. see laparoscopic adjustable gastric related to, 233 described, 194
banding (LAGB) making connections mindful eating
lamotrigine in IPT for EDs, 296–​297 in BED and BN management, 343
for BED, 376t, 380 malnutrition Minnesota Multiphasic Personality
for BN, 373 AN–​related Inventory (MMPI) scores, 451
Lantern, 496 medical complications associated minority(ies)
laparoscopic adjustable gastric with, 223 sexual
banding (LAGB) marketing ED prevention in, 261
prevalence of, 458 of children’s toys and clothing Minority Stress Model, 261
latent class analysis (LCA) body image and eating concerns MMPI scores. see Minnesota Multiphasic
in ED classification, 11–​12 related to, 195 Personality Inventory
Latent Profile Analysis, 238 mass media models (MMPI) scores
laxative abuse ED prevention and, 261–​262 mobile
BN and, 226 MATCH (Modular Approach to defined, 493
in PD, 446 Therapy for Children with mobile assessments
LCA. see latent class analysis (LCA) Anxiety, Depression, or Conduct paper-​and-​pencil assessments vs.
LD-​score regression (LDSR), 83 Problems), 26 bias in self-​report of, 498
LDSR. see LD-​score regression (LDSR) Matching Familiar Figures mobile devices and applications (apps)
LDX Task, 405 in EDs assessment and treatment,
for BED, 375–​377, 376t Maudsley Hospital, 320 492–​504
legislated minimum body mass McArthur Foundation Research Network controversies related to, 498–​499
index (BMI) on Psychopathology and currently available technologies,
for models, 193 Development, 106 497–​498
leptin McKnight Longitudinal Study, 117 EMA in, 495
AN and BN and, 51–​52, 54t–​55t Me!, 262 evaluating, 499–​502, 501f
lesbian(s) meal size, timing, and habit GSH, 496–​497
body satisfaction among, 194 moods and emotions related to, 159 introduction, 492–​493
life chart mealtime environment JIT interventions, 497

538 Index
 mobile delivered interventions, for BN, 374 AN effects on, 224
495–​497 naltrexone NES. see night eating syndrome (NES)
privacy and confidentiality related for AN, 368 neurobiologic alterations
to, 499 for BED, 379, 381 in AN and BN, 47–​48
promise and reach of, 493 for BN, 370t, 373–​374 neurocircuitry
as replacement for therapy, 498–​499 National Adult Reading Test, 402 of appetite regulation
values embedded in AI, 499 National Comorbidity Survey–​Replication in AN and BN, 58–​60
mobile mental health, 493 Adolescent Supplement neuroendocrine alterations
adoption of, 493–​494 on BN, 37 in AN and BN, 49–​53, 54t–​55t
in clinical care National Eating Disorders neuroendocrine systems
evaluation of, 499–​502, 501f Association, 263 feeding behaviors effects on, 50–​53
mobile (including wearable) National Health Service neuropeptide(s)
technology, 493 of UK, 493, 499 endogenous opioid
model(s) National Health Service eating disorders positive mood related to, 159–​160
legislated minimum BMI criteria clinic, 274 in regulation of feeding behavior, 49–​53,
for, 193 National Health System (NHS) 54t–​55t
Modular Approach to Therapy for “Health of the51%: Women” from, 283 neuropeptide alterations
Children with Anxiety, National Institute of Clinical Excellence in AN and BN, 49–​53, 54t–​55t
Depression, or Conduct Problems (NICE), 351 neuropeptide-​Y (NPY)
(MATCH), 26 National Institute of Mental Health AN and BN and, 51, 54t
mOFC. see medial orbitofrontal (NIMH), 385 neuroticism
cortex (mOFC) RDoC project of, 24–​33 (see also EDNOS and, 119
molecular genetic studies research domain criteria (RDoC); NHS. see National Health System (NHS)
in EDs, 98–​99 research domain criteria (RDoC) nibbling
monoamine(s) project) after bariatric surgery, 463
in functioning of striatocortical National Mental Health Surveys, 412 NICE. see National Institute of Clinical
loops, 240 negative affect Excellence (NICE)
monoamine systems binge eating due to Night Eating Questionnaires (NEQs),
dysfunction of BED and, 171–​173 439–​440
EDs related to, 53, 55–​58 BN and, 167–​169 Night Eating Questionnaire total
mood(s). see also emotion(s) effects of, 238–​239 score, 382
characteristics of, 156 chocolate-​ and emotions-​related, night eating syndrome (NES), 40, 438–​
defined, 156 163–​164 444, 440t, 443f
described, 156 described, 238–​239 after bariatric surgery, 462
EDs and, 155–​186 eating attitudes and comfort eating Clinical Global Impression-​
clinical evidence, 164–​177 related to, 161–​164 Improvement scale for, 382
future directions in, 178 overeating during, 161–​164 described, 439
introduction, 155–​158, 156f, 157f restrained vs. emotional eating related as distinct disorder, 439–​441
meal size, timing, and habit effects to, 162 epidemiology of, 40
on, 159 stress susceptibility and comfort eating history of, 439
mechanisms associated with, 158–​164 related to, 162–​163 models of, 439–​443, 443f
emotions vs., 156 negative affect lability continuum with obesity, 442
introduction, 155–​158, 156f, 157f impact on AN and BN, 239 continuum with other EDs, 441–​442
negative negative emotionality continuum with sleep disorders, 442
as maintaining factor for disordered AN and, 173–​177 evidence of diagnostic validity and
eating behavior, 165 BED and, 169–​173 clinical significance using, 443
positive BN and, 165–​169 NES as distinct disorder, 439–​441
endogenous opioid neuropeptides EDs and, 164 SRED, 442
and, 159–​160 as maintaining factor for disordered pharmacotherapy of, 382
mPFC. see medial prefrontal eating behavior, 165 prevalence of, 439
cortex (mPFC) negative mood research diagnostic criteria for, 439, 440t
Multifactorial Assessment of Eating as maintaining factor for disordered research status on, 439
Disorder Symptoms (MAEDS) eating behavior, 165 as secondary to other psychopathology,
scale, 217 negative valence systems 442–​443
multifamily therapy EDs and, 26–​27 NIMH. see National Institute of Mental
for adolescent AN, 327–​329 neophobia Health (NIMH)
muscularity food NNTs. see numbers needed to treat (NNTs)
nonfat in SE, 425 noctural sleep-​related eating disorder
among young men in Samoa, 200 Nepal (NSRED)
body image in, 200 after bariatric surgery, 462
N NEQs. see Night Eating nocturnal ingestions, 439
NAC. see N-​acetylcysteine (NAC) Questionnaires (NEQs) nonfat muscularity
naloxone nervous system among young men in Samoa, 200

Index 539
non-​gap metabolic acidosis orbitofrontal cortex (OFC), 60, 61 as “particularly pernicious,” 319
BN and, 226 amygdala and, 60 role in AN
norepinephrine delayed medial activation reduction history of, 319
in response inhibition and sustained in BN, 60 pathological gambling, 234
attention, 240 medial, 170 Pathway to Mindful Eating, 342, 342t
Norwegian twin study, 87, 92 posterior, 59 PD. see purging disorder (PD)
novelty-​seeking scores orlistat peer support/​school-​based programs
BN and, 237 for BED, 376t, 378 EDs–​related, 258–​259
NPY. see neuropeptide-​Y (NPY) ornamental sports peptide(s)
NSRED. see noctural sleep-​related eating impact on body image and eating opioid
disorder (NSRED) concerns, 192 AN and BN and, 50–​51, 54t
numbers needed to treat (NNTs), 252 orthorexia nervosa (ON), 451–​453, 452t peptide YY (PYY)
nutritional supplements ARFID vs., 453 AN and BN and, 51, 54t
for AN, 368 defined, 451 perfectionism
described, 451 AN and, 236–​238
O diagnostic criteria for, 452, 452t OCD and, 237
OBE episode(s). see objective binge eating history of, 451–​452, 452t peripheral nervous system
(OBE) episode(s) models of, 453 AN effects on, 224
obesity prevalence of, 452–​453 persistence
DRD2 deficiency and, 161 status of AN and, 237
impact on body image and eating research needed to clarify, 453 personal costs
concerns, 194–​195 OSFED. see other specified feeding or EDs–​related, 410–​411
NES and, 442 eating disorder (OSFED) personality
prevalence of, 458 other specified feeding or eating disorder EDs and
prevention of (OSFED), 3, 438 relationships between, 238
ED prevention programs in, 259–​260 BED and, 39 personality disorders
treatment of diagnosis of, 10 classification of, 235
VNS of splanchnic branch of vagus Eating Disorders Workgroup of clusters of, 235
nerve in, 158 DSM-​5 on, 39 defined, 235
Object Assembly, 405 epidemiology of, 39–​40 diagnosis of, 235
objectification NES, 40 EDs and, 235–​236
sexual PD, 39–​40 types of, 235–​236
repeated experiences of, 480 overanxious disorder of childhood personality traits
objectification theory, 480 EDs and, 230 EDs and, 236–​240
objective binge eating (OBE) episode(s) Overcoming Binge Eating, 277, 278, 509 affect regulation, 238–​239
BED and, 459–​460 in SH for EDs, 352–​353 cognitive processes, 239–​240
BN and, 460–​461 “Overcoming Bulimia Online,” 509 RDoC components, 236
observer’s perspective, 480–​481 overconsumption temperament, 236–​238
obsessive-​compulsive disorder (OCD) chronic tendency toward, 140–​146 PET. see positron emission
alcohol abuse/​dependence disorder greater anticipatory food reward in, tomography (PET)
and, 233 141–​144 PET-​O15
EDs and, 230 greater consummatory food reward defined, 68
perfectionism and, 237 in, 140–​141 PFC. see prefrontal
obsessive-​compulsive personality disorder greater impulsivity in, 144–​145 cortex (PFC)
(OCPD), 237 origins of, 140–​146 pharmacotherapy
OCD. see obsessive-​compulsive overeating CBT vs.
disorder (OCD) BN and, 460 for BED, 276
OCPD. see obsessive-​compulsive during negative affect, 161–​164 of EDs, 359–​394
personality disorder (OCPD) oxytocin AN, 360–​368, 361t
Oculus Rift, 474 for AN, 361t, 367 BED, 375–​382, 376t
OFC. see orbitofrontal cortex (OFC) for BN, 373 BN, 368–​375, 370t
olanzapine future directions in, 384–​385
for AN, 361t, 363–​364 P introduction, 359
ON. see orthorexia nervosa (ON) pain NES, 382
Onslaught, 262 alleviation of rationale for, 359–​360
opioid(s) fatty food in, 160 research studies’ conclusions, 382–​383
for AN, 368 painful affects phenotype(s)
opioid antagonists IPT in encouraging acceptance of, 298 defined, 81
for AN, 368 paper-​and-​pencil assessments phentermine
for BED, 381 mobile assessments vs. for BED, 379
for BN, 370t, 373–​374 bias in self-​report of, 498 phenylthiocarbamide (PTC)
opioid peptides parent(s) heritability of, 429
AN and BN and, 50–​51, 54t as “generally the worst attendants,” 319 phenytoin

540 Index
 for AN, 368 categories of, 247–​248 psychosocial risk factors
for BED, 380 indicated, 248 for EDs, 106–​125 (see also specific types
phobia(s) risk factor reduction in, 248 and risk factors, for EDs)
treatment for risk factors informing, 249 psychotherapy
exposure and response prevention StudentBodies program (see interpersonal (see interpersonal
in, 433 StudentBodies program) psychotherapy (IPT))
physical exercise theory for, 248–​249 via Internet-​based interventions,
CBT with Price Foundation consortia, 98 507–​508
in BN management, 130 Price Foundation Genetic Studies of PTC. see phenylthiocarbamide (PTC)
pica Eating Disorders, 234 public health/​policy
diagnosis of, 10 privacy ED prevention and, 261–​262
picking of mobile devices and apps for purging behavior
after bariatric surgery, 463 EDs, 499 BN and, 225–​226
picky eating, 419–​437. see also selective proanorexia purging disorder (PD), 438, 445–​448
eating (SE) websites for, 262–​263 on continuum with other EDs,
plugging problem areas 447–​448
after bariatric surgery, 465 described, 313 course of, 40
political-​economic variables problematic eating behaviors (EBs) defined, 445
impact on body image and eating after bariatric surgery, 462–​464 described, 39–​40, 445–​446
concerns, 189–​191 chewing and spitting, 464 as distinct axis I psychiatric disorder,
polymorphism(s) emotional eating, 463 446–​447
described, 81 grazing, 462–​463 epidemiology of, 40, 446
Positive and Negative Affect Scale, 339 high-​calorie fluid history of, 445–​446
positive mood consumption, 463 models of, 446–​448
endogenous opioid neuropeptides and, picking and nibbling, 463 evidence of diagnostic validity and
159–​160 sweet eating, 463–​464 clinical significance of, 448
positive predictive value (PPV) probulimia prevalence of, 40, 446
of screening tests, 212 websites for, 262–​263 purging methods in, 446
positive valence systems pro–​eating disorder websites, 193 status of
EDs and, 27 prokinetics research needed to clarify, 448
positron emission tomography (PET) for AN, 361t, 365–​366 synonyms for, 445–​446
studies for BN, 375 validity and clinical utility of,
in AN and BN, 58 PROP. see 6-​npropylthiouracil  (PROP) 445–​446
anxiety and harm avoidance 6-​npropylthiouracil (PROP) purging-​only syndrome. see purging
correlated with, 57–​58 heritability of, 429 disorder (PD)
appetitive regulation–​related, 61–​62 proxy risk factor, 136 pyromania, 234
posterior orbitofrontal “ProYouth,” 512 PYY. see peptide YY (PYY)
cortex (OFC), 59 Pseudo Bartter’s syndrome, 226
PPV. see positive predictive value (PPV) psychiatric disorder Q
practice(s) criteria for, 438 QALYs. see quality-​adjusted life years
impact on body image and eating Psychiatric Genomics Consortium (QALYs)
concerns, 191–​193 Eating Disorders Working quality-​adjusted life years (QALYs), 414
preference(s) Group of, 99 question(s)
decided Psychiatric Status Rating Scale for Bulimia exploratory
in SE, 426 Nervosa, 371 in IPT for EDs, 298
prefrontal cortex (PFC) psychoeducation
medial, 59 in ED prevention programs, 250 R
“preoccupations of relatives,” 319 for EDs race
prevention apps for, 496 as factor in body image and eating
defined, 247–​248 psychoeducational videotapes concerns, 196–​198
of EDs, 247–​270 (see also ED in SH for EDs, 353 racial/​ethnic minorities
prevention; ED prevention psychological assessment ED prevention in, 261
programs) of EDs, 211–​221 (see also assessment(s), racial issues
causative issues and, 248 of EDs) in CBT for EDs, 282
current status and underlying theory, psychological comorbidity radically open-​DBT (RO-​DBT), 339
247–​270 of EDs, 229–​243, 231t–​233t (see also radioligand
foundation of, 248 specific disorders and eating defined, 68
introduction, 247 disorder(s) (EDs), psychological randomized controlled trials (RCTs)
harm reduction in, 248–​249 comorbidity of ) of CBT for BN, 274
programs for (see prevention programs) psychosis(es) of CRT for AN, 397
theory of, 248–​249 CRT for, 396 of CRT for EDs, 397–​399
prevention programs. see also ED psychosocial impairment of GSH via Internet-​based
prevention programs SE and, 423 interventions, 509

Index 541
randomized controlled trials (RCTs) challenges of, 29–​30 longitudinal studies’ characteristics,
(cont.) cognitive systems, 27–​28 107–​111, 108t–​110t
of psychotherapy via Internet-​based described, 25–​29 prevention programs in reducing, 248
interventions, 507–​508 domains in, 25 research limitations related to, 111–​112
of StudentBodies via Internet-​based future directions in, 30–​31 research update (2002-​2015) on, 112–​
interventions introduction, 24 120, 113t–​116t
for EDs, 511 negative valence systems, 26–​27 results from register studies, 119–​120
RCI. see reliable change index (RCI) positive valence systems, 27 in specific settings and special
RCTs. see randomized controlled social processes, 28–​29 populations, 260–​263
trials (RCTs) introduction, 24 study characteristics, 112–​117,
RDoC. see research domain of NIMH, 17 113t–​116t
criteria (RDoC) impact on conceptualization study criteria, 107
Reading the Mind in the Eyes, 28 of EDs, 1 study method, 107, 112–​117, 113t–​116t
“Real Beauty Sketches” ad, 262 on personality traits, 236 substance abuse and dependence, 161
rebellion recent changes related to, 2 in informing ED prevention
alternate research domain criteria (RDoC) programs, 249
in BED and BN management, 343 project. see also research domain proxy, 136
reboxetine criteria (RDoC) risperidone
for BED, 377 goal of, 24 for AN, 361t, 364
recontextualization introduction, 24 ritual(s)
in approach to disgust, 433 structure of, 25 in Judaism
recovery response inhibition fasting as part of, 201
thresholds for norepinephrine in, 240 religious
in EDs assessment, 214–​215 restrained eating impact on body image and eating
Recovery Record, 494–​497, 499 emotional eating vs., 162 concerns, 191–​192
Red Book, 413 restraint(s) RO-​DBT. see radically open-​DBT
refeeding hypophosphatemia cognitive (RO-​DBT)
AN and, 223 overeating during stress related role disputes
Reflections program, 256 to, 162 interpersonal, 289
reflux dietary IPT for, 294t, 295
acid researcher’s use of measures in, role transitions, 289
BN and, 225 134–​136 IPT for, 294–​295, 294t
regional cerebral blood flow vs. dieting, 127 ropiramate
defined, 68 Restraint Scale (RS), 134 for AN, 368
regional cerebral blood flow studies restriction(s) for BN, 370t, 372–​373
in appetitive regulation in AN and dietary Rosenberg Self-​Esteem Scale, 328
BN, 61–​62 relative vs. absolute, 136 Roux-​en-​Y gastric bypass (RYGB)
regional trends reward(s) prevalence of, 458
from cross-​cultural cataloging to global alterations in responses to Royal Children’s Hospital,
systems in EDs, 64–​65, 65f Melbourne, 324
as factor in body image and eating DA in “wanting” of, 160–​161 RS. see Restraint Scale (RS)
concerns, 199–​202 food (see food reward) rumination disorder
regulatory systems sensory diagnosis of, 10–​11
EDs and, 29 emotion effects of neural substrates RYGB. see Roux-​en-​Y gastric
relamorelin shared by, 159–​161 bypass (RYGB)
for AN, 361t, 367 “reward deficiency
relationship(s) syndrome,” 160 S
therapeutic Rey-​Osterrieth Complex Figure Test, “Salut BED/​Salut BN,” 509
in IPT for EDs, 299 405–​406 Samoa
relative dietary restriction rimonabant nonfat muscularity among young men
absolute dietary restriction vs., 136 for BED, 379 in, 200
reliable change index (RCI), 214–​215 RiseUp, 494–​496 Samsung, 494
religious rituals and beliefs risk factor(s) satiety
impact on body image and eating for EDs, 106–​125 (see also specific types in EDs
concerns, 191–​192 and disorders) fMRI studies of, 63
research domain criteria (RDoC). see also AN, 107, 111, 112, 117 scalability
research domain criteria (RDoC) BED, 111, 118–​119 of CBT for EDs, 282–​283
project BN, 111, 117–​118 “scape(s)”
impact on conceptualization of EDs, dieting, 126–​154 (see also dieting) described, 200–​201
1, 24–​33 EDNOS, 119 SCID. see Structured Clinical Interview
alternatives to, 30 future directions in, 122–​123 for DSM-​IV (SCID)
arousal and regulatory systems, 29 interactions between, 120–​121 Science, 473
benefit of, 25–​26 introduction, 106–​107 SCOFF

542 Index
 in EDs assessment, 212–​213, 216–​217 self-​help (SH) serotonin norepinephrine reuptake
screening for EDs, 351–​358 inhibitors (SNRIs)
for EDs described, 352 for BN, 371–​372
assessment-​related, 212–​213 future directions in, 355–​356 serotonin receptor binding alterations
PPV–​related, 212 Getting Better Bit(e) by Bit(e) in, 353 in EDs, 56–​58
prevention-​related, 249–​250 introduction, 351–​352 sertraline
self-​report questionnaires in, 217 Overcoming Binge Eating in, 352–​353 for NES, 382
SE. see selective eating (SE) psychoeducational videotapes set-​shifting, 239, 396
secondary domains in, 353 of CRT for EDs, 403t, 404
in EDs assessment, 214 systematic reviews and meta-​analyses examples of, 403t
selective eater(s) related to, 353 set-​shifting tasks
defined, 419 treatment predictors, moderators, DA in, 240
selective eating (SE), 419–​437 and mediators of, 354 “Set Your Body Free,” 511
behavioral features associated with, guided (see guided self-​help (GSH)) sexual minorities
425–​427 unguided ED prevention in, 261
decided preferences, 426 via Internet-​based interventions, sexual objectification
food neophobia, 425 508–​510 repeated experiences of, 480
food refusal/​conflict around self-​induced vomiting SFT. see separated family therapy (SFT)
food, 426 BN and, 225–​226 SG. see sleeve gastrectomy (SG)
slowness in eating, 426–​427 self-​monitoring SH. see self-​help (SH)
criteria of, 419 of EDs short-​term caloric restriction
defined, 419–​421 self-​report questionnaires for, 218 relationship to laboratory-​based eating
described, 419–​420 Self-​Regulatory Executive Function experimental studies of, 128–​129
disgust and evaluative conditioning in, Model, 450 SIAB. see Structured Interview for Anorexia
432–​434, 432f self-​report measures and Bulimia Nervosa (SIAB)
disgust experience in, 432–​434, 432f of CRT for EDs, 402 SIAB-​S. see Structured Interview for
duration of, 421–​424 self-​report questionnaires Anorexic and Bulimic Syndromes
essential features of, 424–​425 in EDs assessment, 217–​218 Survey (SIAB-​S)
family meals compromised by, 423 diagnosis-​related, 217 sialadenosis, 225
food variety in, 424–​425 screening-​related, 217 sibutramine
frequency threshold of severity in, self-​monitoring–​related, 218 for BED, 376t, 379
420–​421 test meals–​related, 218 sick role
growing out of, 435–​436 treatment planning and evaluation–​ described, 313
introduction, 419 related, 217–​218 diagnosis and assignment of
management of, 434–​436, 434t semistructured interviews in IPT for EDs, 290, 291t
mealtime environment in, in EDs assessment, 216 Simmons Behavior Checklist, 119
434–​435, 434t sense of presence single photon emission computed
medical, 434 VR and, 471 tomography (SPECT) studies
sensory-​based exposures in, 435 sensitized social environment of appetitive regulation in AN and
values clarification in, 435 SE and, 431–​432 BN, 61–​62
nature of, 420–​421 sensory-​based exposures skin
persistence of in SE management, 435 AN effects on, 225
predictors of, 424 sensory reward sleep disorders
phenomenology of, 427–​434, 428f, neural substrates shared by NES and, 442
430f, 432f emotion effects of, 159–​161 sleep-​related eating disorder (SRED), 442
disgust experience, 427–​429, 428f sensory sensitivities NES and, 442
sensory sensitivities, 427–​434 (see in phenomenology of SE sleeve gastrectomy (SG)
also sensory sensitivities, in disgust experience, 430f prevalence of, 458
phenomenology of SE) executive functioning and, 431 slowness in eating
prevalence of, 420 medical conditions and, 430–​431 in SE, 426–​427
psychosocial impairment resulting potentiation of emotional experience SMART (sequential multiple assignment
from, 423 related to, 430 for randomized treatment), 355
putative model of, 429, 430f sensitized social environment and, smartphones, 493
severity of 431–​432 SMA syndrome. see superior mesenteric
threshold for, 420–​421 in SE, 427–​434 artery (SMA) syndrome
stunted growth related to, 422 separated family therapy (SFT) SNRIs. see serotonin norepinephrine
synonyms for, 419 in FBT-​AN, 321 reuptake inhibitors (SNRIs)
selective serotonin reuptake inhibitors sequential multiple assignment social communication
(SSRIs) for randomized treatment as primary function of emotion,
for BED, 376t, 377 (SMART), 355 431–​432
self-​directedness serotonin social environment
low defined, 81 sensitized
EDs and, 238 EDs related to, 55–​58 disgust experience and, 431–​432

Index 543
social learning theory cognitive restraint and, 162 stress-​reducing effect of, 159–​160
in ED prevention programs, 250–​251 susceptibility to symbolic body capital
social processes comfort eating related to, 162–​163 described, 195
EDs and, 28–​29 sweet taste effects on, 159–​160 impact on body image and eating
social threat stimuli striatocortical loops concerns, 195–​196
attentional biases to, 240 monoamines in functioning of, 240 synthetic egocentric experience, 482
sociocultural factors Structured Clinical Interview for DSM-​IV
described, 187–​189 (SCID), 117–​118 T
effects on body image and EDs, Structured Interview for Anorexia and TA. see taxometric analysis (TA)
187–​208 (see also culture) Bulimia Nervosa (SIAB), 371 TakeControl, 495
socioeconomic status Structured Interview for Anorexic and targeted or selective prevention
to upward mobility Bulimic Syndromes Survey interventions, 248
as factor in body image and eating (SIAB-​S), 449 TAS. see Toronto Alexithymia Scale (TAS)
concerns, 198–​199 “Student2Bodies–​BED, “ 511 task-​shifting, 281
sodium oxybate StudentBodies program, 248, 250, 254, 256–​ taste(s)
for BED, 382 258, 263, 511, 513 sweet
SPECT. see single photon emission Stanford-​Washington University studies stress-​reducing effect of, 159–​160
computed tomography (SPECT) of, 256–​258 taxometric analysis (TA)
spironolactone stunted growth in ED classification, 11–​12
for BN, 375 SE and, 422 taxonomy
spitting subcultural groups defined, 9
after bariatric surgery, 464 impact on body image and eating TBPSP. see thin body preoccupation and
sports concerns of social pressure to be thin (TBPSP)
ornamental college campus, 192 TCI. see Temperament and Character
impact on body image and eating gay male subcultures, 194 Inventory (TCI)
concerns, 192 Internet-​based groups, 193 TDM. see three-​dimensional
sports subcultures job subcultures, 192 model (TDM)
impact on body image and eating lesbians, 194 technological advancements
concerns, 191–​192 shared behaviors among, 191–​192 globalization and, 201–​202
SRED. see sleep-​related eating sports, 191–​192 technology(ies)
disorder (SRED) subculture(s) in EDs treatment, 522–​523
SSRIs. see selective serotonin reuptake described, 191 embodied VR as, 473
inhibitors (SSRIs) impact on body image and eating temperament
Stanford DBT model, 338–​339, 341–​343, concerns, 191–​194 as factor in EDs, 236–​238
342t, 344t–​346t gay male–​related, 194 Temperament and Character Inventory
adaptations for, 341 Internet-​based groups, 193 (TCI), 237
to treatment hierarchy, diary card, sports-​related, 191–​192 test meals
and behavioral chain analysis, subjective bulimia nervosa. see purging self-​report questionnaires for, 218
342–​343, 342t, 344t–​346t disorder (PD) testosterone
background of, 341 substance abuse and dependence for AN, 361t, 366–​367
described, 341 EDs and, 161 for BN, 373
sequence of treatment in, 342 substance abuse disorders TFEQ-​R. see Three-​Factor Eating
structure of treatment in, 341–​342 EDs and, 230, 233–​234 Questionnaire Restraint Scale
Stanford DBT model diary card, 343, suicide (TFEQ-​R)
344t–​346t AN and, 36 The (Group) Embedded Figures Task
Stanford dosage study BED and, 39 ([G]‌EFT), 405
in FBT-​AN, 323 BN and, 38 theory(ies)
Stanford-​Washington University suicide attempt(s) AL, 481, 481f
StudentBodies studies of, 256–​258 AN and, 36 biosocial (see biosocial theory)
Stanford-​Washington University Eating BED and, 39 dieting
Disorder Screen (SWED), 250 BN and, 38 of eating pathology, 127–​128
starvation superior mesenteric artery (SMA) dissonance
brain effects of, 47–​48 syndrome in ED prevention programs, 251
stepped care treatment AN and, 223 of ED prevention interventions, 250–​
for EDs, 351–​358 suppressed affect 252 (see also specific types and ED
components of, 351–​358 IPT in helping patient experience, 298 prevention, theories and models of
future directions in, 355–​356 SWED. see Stanford-​Washington interventions in)
models of, 355 University Eating Disorder feminist
studies of, 354–​355 Screen (SWED) in ED prevention programs, 251
Stop Binge Eating, 496 Swedish Twin Registry, 442 interpersonal
stress sweet eating of IPT, 287–​288
in DA reduction, 161 after bariatric surgery, 463–​464 objectification, 480
overeating during sweet taste of prevention, 248–​249

544 Index
 social learning AN and, 225 body image disturbances–​related,
in ED prevention programs, 250–​251 underlying theoretical assumptions 478–​484, 481f (see also body image
transdiagnostic in EDs assessment, 213–​214 disturbances, VR in)
of EDs, 252 unguided self-​help (SH) components of, 471, 472t
therapeutic relationship via Internet-​based interventions, defined, 470–​471, 473
in IPT for EDs, 299 508–​510 described, 470–​471
therapeutic stance United Arab Emirates for EDs, 470–​491
in IPT for EDs, 295 body image in, 201 CET, 475–​478
thin body preoccupation and social universal prevention program, 247–​248 effectiveness of, 474–​475
pressure to be thin (TBPSP), University of California, San as embodied technology, 473
120–​121 Francisco, 494 history of, 470
threat stimuli unspecified feeding or eating disorder introduction, 470
social and angry (UFED), 438 limitations of, 474–​475
attentional biases to, 240 diagnosis of, 10 sense of presence in, 471
three-​dimensional model (TDM) unstructured interviews from virtual to real bodies, 473–​474
of EDs, 13–​16, 14f in EDs assessment, 216 virtual worlds, real emotions in,
Three-​Factor Eating Questionnaire upward mobility 471–​473
Restraint Scale (TFEQ-​R), 134, 135 from socioeconomic status to virtual reality (VR)–​CET (VR-​CET)
thresholds for recovery as factor in body image and eating for EDs, 475–​478
in EDs assessment, 214–​215 concerns, 198–​199 Virtual Reality Environments for the
topiramate urbanization Psycho-​neuro-​physiological
for BED, 376t, 379–​381 impact on body image and eating Assessment and Rehabilitation
for NES, 382 concerns, 190 Project (VREPAR), 470, 480
Toronto Alexithymia Scale (TAS), 176 urge surfing virtual reality (VR) systems
toy(s) in BED and BN management, 343 commercially available fully immersive
marketing of prices and characteristics of,
impact on body image and eating V 471, 472t
concerns, 195 vagal nerve virtual worlds, real emotions
Trail Making Task, 404–​405 splanchnic branch of in VR, 471–​473
training-​as-​usual VNS of, 158 VNS. see vagal nerve stimulation (VNS)
in dissemination and implementation vagal nerve activity vomiting
of CBT, 283 BN and after bariatric surgery, 464–​465
trait anxiety depression related to, 158–​159 in PD, 446
AN and, 237 vagal nerve stimulation (VNS) self-​induced
transdiagnostic model artificial cervical BN and, 225–​226
of EDs, 15–​16 in depression management, 158 VR. see virtual reality (VR)
transdiagnostic theories of splanchnic branch of vagus nerve VR-​CET. see virtual reality (VR)–​CET
of EDs, 252 in obesity management, 158 (VR-​CET)
transnational connection valproate VREPAR (Virtual Reality Environments
routes of, 200–​201 for AN, 368 for the Psycho-​neuro-​physiological
treatment phases for BED, 381 Assessment and Rehabilitation
described, 313 for BN, 372 Project), 470, 480
trichotillomania, 234 value(s) vulnerability
twin studies embedded in AI, 499 emotional
of AN and BN, 48, 81 positive predictive defined, 334–​335
of EDs, 3 of screening tests, 212
of gene–​environment interplay, 86–​98 values clarification W
described, 86, 96–​97, 97t in SE management, 435 WAIS-​R. see Wechsler Adult Intelligence
developmental changes related Veterans’ Administration data Scale (WAIS-​R)
to, 97–​98 on per-​patient financial costs of WCEDCA. see Workgroup for
heritability estimates in, 86–​92, 88t–​ EDs, 411 Classification of Eating Disorders
91t, 93t–​95t Veterans Affairs IPT training program, 312 in Children and Adolescents
introduction, 86 Veterans Health Administration, 312 (WCEDCA)
shared latent risk factors between “VIA,” 510 WCS. see Weight Concerns Scale (WCS)
two or more different phenotypes videoconferencing technology WCST. see Wisconsin Cart Sort
in, 92, 96 in dissemination and implementation Test (WCST)
of CBT, 283 weak central coherence
U videotape(s) defined, 239–​240
UFED. see unspecified feeding or eating psychoeducational “Web-​centered training”
disorder (UFED) in SH for EDs, 353 in dissemination and implementation
UK Office of Health Economics, 412 Virginia Twin Registry albeit one of CBT, 283
ulcer(s) study, 87 website(s)
decubitus virtual reality (VR), 470–​491 pro–​eating disorders, 193, 262–​263

Index 545
Wechsler Adult Intelligence Scale IPT development in, 311–​312 Y
(WAIS-​R), 402, 405 Wellcome Trust Case Control Consortium Yale-​Brown-​Cornell Eating Disorder
Weight Concerns Scale (WCS), 249–​250 3 (WTCCC3), 98 Scale (YBC-​EDS) rituals score,
weight gain Western Australian Pregnancy Cohort 239, 363
excessive (Raine) Study, 119, 447 YBC-​EDS rituals score. see Yale-​Brown-​
IPT in prevention of, 305–​310, 308t Whites Cornell Eating Disorder Scale
weight-​loss drugs body image and eating concerns (YBC-​EDS) rituals score
for BED, 376t, 378–​379 among, 197 Yoga Journal, 451
for BN, 370t, 374 WHO. see World Health younger children
weight-​loss interventions Organization (WHO) IPT for
BED related to, 130 Wisconsin Cart Sort Test adaptations of, 311
behavioral (WCST), 404
vs. CBT, 276 Workgroup for Classification of Eating Z
BN related to, 130 Disorders in Children and zinc
group behavioral Adolescents (WCEDCA), 19 for AN, 361t, 366
vs. CBT in overweight BED World Health Organization (WHO), 199 zonisamide
persons, 130 WTCCC3. see Wellcome Trust for BED, 376t, 380
weight-​related problems Case Control Consortium 3
prevention of (WTCCC3)

546 Index