Athena Robinson - W Stewart Agras - Oxford Handbook of Eating Disorders-Oxford University Press (2018) PDF
Athena Robinson - W Stewart Agras - Oxford Handbook of Eating Disorders-Oxford University Press (2018) PDF
Athena Robinson - W Stewart Agras - Oxford Handbook of Eating Disorders-Oxford University Press (2018) PDF
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
Neuropsychology
Kenneth M. Adams
Organizational Psychology
Steve W. J. Kozlowski
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.
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
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
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
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
Index 525
xiv Contents
Introduction
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
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
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
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
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.
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.
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
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
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.
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
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
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
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
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).
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
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
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
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
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
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
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.,
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
REC BN ↓
(Lesem, 1991) =
(Kaye, 1987) ↓
(Baranowska, 2000) ↓
(Devlin, 1997; Geracioti, 1988; Keel, 2007; Phillipp, 1991; Pirke, Ill BN ↓
1994)
(Stock, 2005) =
(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
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
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).
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
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.
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
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
(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
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
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]
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.
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
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
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
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.
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
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
.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.
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
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.
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
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)
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
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.
(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)
– 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
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.
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)
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
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
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
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
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
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
Wade, Bulik 105
C H A PT E R
Psychosocial Risk Factors
6 for Eating Disorders
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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).
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
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
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.
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.
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.
Psychological Assessment of the
10 Eating Disorders
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
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
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
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,
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).
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.
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
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
Halmi 1991 60 50
(continued)
Table 12.1 Continued
Authors Comorbid Diagnosis Anorexia Nervosa Bulimia Nervosa Binge Eating Disorder
% % %
Drug Abuse or
Dependence
Matsunaga 2000 26
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
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
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
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.
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.
Halmi 243
PART
4
Prevention and
Treatment
CH A PT E R
Prevention: Current Status and
13 Underlying Theory
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
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
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).
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
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,
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
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
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
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
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
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
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
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
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.
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.
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
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
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
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.
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.
Interpersonal Psychotherapy for the
15 Treatment of Eating Disorders
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
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
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
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
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
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”)
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
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 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
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.
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.
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,
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
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
Dialectical Behavior Therapy and
17 Emotion-Focused Therapies for
Eating Disorders
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
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
Self-Help and Stepped Care Treatments for
18 Eating Disorders
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,
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
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
Clomipramine 50 − − NDA
Fluoxetine 60 − − +/−
Antipsychotics
Risperidone 4 − − −
Appetite Stimulants
Cannabinoids
Dronabinol 5 + − NDA
Prokinetics
Hormonal Agents
Other Agents
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
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
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.
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
Tricyclic Antidepressants
Amitriptyline 150 ++ ++
Desipramine 300 ++ ++
Imipramine 300 ++ ++
Brofaromine 200 − ++
Isocarboxazid 60 ++ ++
Moclobemide 600 − −
SSRIs
Fluvoxamine 300 ++ ++
Other Antidepressants
Bupropion 450 ++ ++
Mianserin 60 − −
Trazodone 400 ++ ++
Antiepileptics
Carbamazepine NDAa − −
Other Agents
Dexfenfluraminec 60 ++ ++
Flutamide 500 ++ −
Lithium 600–1200b − −
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
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,
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
ADHD Medications
Atomoxetine 120 + +
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
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.
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
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
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.
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.
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.
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.
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
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.
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
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
Harrison 397
Table 20.1 Similarities and Differences Between Cognitive Remediation Therapy and Cognitive-Behavioral
Therapy
Cognitive-Behavioral Therapy Cognitive Remediation Therapy
Similarities
Homework exercises
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)
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
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).
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.
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.
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
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
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.
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
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
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.
Crow 415
PART
5
Emerging Topics
CH A PT E R
Selective Eating: Normative Developmental
22 Phase or Clinical Condition?
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
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.
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.
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).
b b
b
2.20
b
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.
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.
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
&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.
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
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.
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
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,
II. Awareness and recall of evening and nocturnal eating episodes are present.
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
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.
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.
Eating Disorders and Problematic Eating
24 Behaviors After Bariatric Surgery
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
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
Virtual Reality: Applications to Eating
25 Disorders
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
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).
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)
Field of View 110 degrees 110 degrees 101 degrees from 70 96 degrees 100 degrees
degrees
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
WORKING predicted
real-time body from
perceptions MEMORY
memory
BODY SHAME:
PERCEIVING ONESELF
NON MEETING CULTURAL STANDARDS
EXTREME RESTRAINED
EATING
EATING
DISORDERS
Mobile Device Applications for
26 the Assessment and Treatment of
Eating Disorders
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
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
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?
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
Patient Clinician
considerations considerations
Collaborative exploration
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
Internet-Based Interventions
27 for Eating Disorders
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
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, &
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
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
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.
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