Psychological Medicine Social anxiety in the eating disorders: a

cambridge.org/psm
systematic review and meta-analysis
Jess Kerr-Gaffney1, Amy Harrison2 and Kate Tchanturia1,3,4
1
King’s College London, Institute of Psychiatry, Psychology and Neuroscience, Psychological Medicine, London,
Review Article UK; 2Department of Psychology and Human Development, University College London, London, UK; 3South London
Cite this article: Kerr-Gaffney J, Harrison A, and Maudsley NHS Trust, National Eating Disorders Service, Psychological Medicine Clinical Academic Group,
Tchanturia K (2018). Social anxiety in the London, UK and 4Ilia State University, Tbilisi, Georgia
eating disorders: a systematic review and
meta-analysis. Psychological Medicine 1–15. Abstract
https://doi.org/10.1017/S0033291718000752
Social anxiety disorder is one of the most common comorbid conditions in eating disorders
Received: 6 December 2017 (EDs). The aim of the current review and meta-analysis is to provide a qualitative summary of
Revised: 20 February 2018
what is known about social anxiety (SA) in EDs, as well as to compare levels of SA in those
Accepted: 28 February 2018
with EDs and healthy controls. Electronic databases were systematically searched for studies
Key words: using self-report measures of SA in ED populations. In total, 38 studies were identified, 12
Anorexia nervosa; bulimia nervosa; eating of which were included in the meta-analyses. For both anorexia nervosa (AN) and bulimia
disorders; social anxiety disorder nervosa, there were significant differences between ED groups and HCs, with medium to
Author for correspondence: large effect sizes. Findings from the qualitative review indicate that levels of SA are similar
Jess Kerr-Gaffney, E-mail: jess.kerr-gaffney@ across the ED diagnoses, and SA improves with treatment in AN. In addition, high levels
kcl.ac.uk of SA are associated with more severe ED psychopathology, but not body mass index.
These findings add to the wider literature on socio-emotional functioning in EDs, and may
have implications for treatment strategies.

Introduction
Eating disorders (EDs) are associated with high levels of psychiatric comorbidity (Blinder et al.
2006); a factor linked to poorer short- and long-term outcomes (Berkman et al. 2007; Vall &
Wade, 2015). Anxiety disorders are common, with a lifetime prevalence of around 60%
reported in both anorexia nervosa (AN) and bulimia nervosa (BN) (Bulik et al. 1997). In par-
ticular, social anxiety disorder (SAD; also known as social phobia) is consistently found to be
the first or second most common comorbid anxiety disorder in EDs (Godart et al. 2000; Kaye
et al. 2004; Swinbourne et al. 2012), with prevalence rates ranging from 16 to 88.2% in AN and
17–67.8% in BN (Swinbourne & Touyz, 2007). In comparison, the lifetime prevalence of SAD
in the general population is around 12% (Kessler et al. 2005). The association between SAD
and EDs is also replicated in non-clinical populations, where disordered eating is positively
associated with social anxiety (SA) levels (Gilbert & Meyer, 2003; Gadalla & Piran, 2008;
Utschig et al. 2010; Ciarma & Mathew, 2017). High levels of SA in EDs may be part of a
wider socio-emotional phenotype hypothesised to contribute to the development and main-
tenance of EDs (Treasure & Schmidt, 2013). For example, people with AN report having impo-
verished social networks and internalising problems in childhood, problems which are further
accentuated by the ill state (Adambegan et al. 2012; Harrison et al. 2014; Westwood et al.
2016). Similarly, adolescents and young adults with ED show more insecure attachment styles
(Dias et al. 2011), which are theorised to have lasting implications on emotion regulation,
social processing, and self-evaluative processes (Gander et al. 2015).
Explanations for the link between EDs and SA have been proposed. Firstly, SA may be a
risk factor for the development of an ED. For example, anxiety around how oneself appears
to others may lead to an excessive interest in body weight and shape (Godart et al. 2000).
Another possibility is that SA may be secondary to the ED, as a consequence of ED psycho-
pathology or malnutrition. Studies examining the temporal relations between the two disor-
ders lend some support to the former hypothesis, where it is consistently reported that
SAD onset preceded the ED in the majority of those with both disorders (Bulik et al. 1997;
Godart et al. 2000; Kaye et al. 2004; Swinbourne et al. 2012). However, such studies rely on
retrospective accounts of age of onset, and are therefore subject to recall biases. Two prospect-
ive studies using representative samples provide conflicting results, and suggest that the rela-
tionship with SAD may differ as a function of ED diagnosis. Buckner et al. (2010) found that
© Cambridge University Press 2018 BN in adolescence significantly increased the risk of both SAD and panic disorder in adult-
hood, however, no anxiety or depressive disorder in adolescence predicted later BN. AN in
adolescence did not increase the risk of any anxiety disorder or depression in adulthood,
but adolescent obsessive-compulsive disorder (OCD) predicted the development of AN in
adulthood. In contrast, Ranta et al. (2017) found that both SAD and depression at age 15

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 2 Jess Kerr-Gaffney et al.

predicted BN at age 17, however, the relationship between SAD eating disorder OR binge eating disorder. No search limits were
and BN was not significant after controlling for depression. applied, except for in Web of Science, where results were filtered
Contrary to the previous study, neither AN or BN predicted by the ED term for relevance.
later SAD, however, this may be due to the far shorter follow-up
period. Thus, evidence regarding the direction of causality is
Study selection
inconsistent.
A final explanation for the comorbidity is that SAD may share Screening and selection of articles is displayed in Fig. 1. Where
common vulnerability factors with ED, as has been found to be titles of papers appeared relevant, abstracts were screened to
the case with OCD. In addition to AN being more common in check eligibility. Full texts of potentially eligible studies were
probands of individuals with OCD (an effect that increases with then retrieved. Studies that met al.l eligibility criteria but did
the degree of genetic relatedness), the moderate genetic overlap not include a HC group were included in the qualitative review,
between the two disorders has been reported in a large whereas those that included a HC group were included in both
population-based twin sample (Cederlof et al. 2015). Similarly, the meta-analysis and the qualitative review. Where a study did
SAD occurs at significantly higher rates in first-degree relatives not report means and standard deviations for SA scores, study
of probands with AN than those of healthy controls (HCs) authors were contacted. When no response was received, studies
(Strober et al. 2007). It may be that heritable vulnerability factors were excluded.
such as perfectionism partly explain the genetic overlap between
anxiety disorders and EDs. For example, perfectionism, a trait
Data collection
that is elevated in both individuals with EDs and those with
SAD (Antony et al. 1998; Lloyd et al. 2014) has been found to The following information was extracted from each paper: num-
predict both SA and disordered eating in non-clinical women ber of participants in each group, diagnosis, mean SA score, SA
(Levinson & Rodebaugh, 2016). Perfectionism has also been measure used, age, BMI, illness duration, percentage of female
shown to moderate the relationship between SA and bulimic participants, group matching technique, and recruitment source.
symptoms specifically, where those with high SA and perfection-
ism showed the most bulimic symptoms (Silgado et al. 2010). Due
Risk of bias in individual studies
to a lack of research on common vulnerability factors in clinical
ED and SAD populations, no firm conclusions for the high levels Risk of bias in individual studies was assessed by considering how
of comorbidity can be drawn. certain methodological characteristics (participant recruitment
To date, only one review has examined comorbidity between source, group matching technique, and SA measure used) might
EDs and SAD, within a general review of anxiety disorder have impacted the results of the studies.
comorbidity in EDs (Swinbourne & Touyz, 2007). However,
this review (a) only provided categorical prevalence estimates of
Summary measure and data synthesis
anxiety disorders in ED populations, and (b) did not examine
whether SA differs across ED sub-types. Further, new studies The principle summary measure used in the meta-analysis was
have become available. Therefore, the aim of the current review the difference in means and standard deviations for SA scores
and meta-analysis is to compare SA in EDs compared with between ED and HCs. The meta-analyses were performed by
HCs, and provide a qualitative synthesis of the literature, e.g. dif- pooling standardised effect sizes using a random effects model.
ferences in SA between ED sub-types, the effects of treatment on Separate meta-analyses were performed for each ED sub-type,
SA, and associations between SA and factors such as body mass and studies that included more than one ED group (e.g. AN
index (BMI) and ED psychopathology. and BN) compared with HCs were included in each of the
respective meta-analyses.
Method
Statistical analysis
The review and meta-analysis was conducted using the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses All analyses were performed using R Studio (R Core Team, 2017)
(PRISMA) statement (Liberati et al. 2009). using the metafor package (Viechtbauer, 2010). Cohen’s d was
used to estimate effect sizes and is reported with 95% confidence
intervals (CIs). Effect sizes are interpreted using Cohen’s (1988)
Eligibility criteria
definitions of small (0.2), medium (0.5), and large (0.8).
Studies using a quantitative measure of SA were included in the Positive effect sizes indicate that the ED group had SA scores
review. Inclusion criteria were: (1) at least one clinical ED sample; than HCs. Two AN studies included in the meta-analysis shared
(2) means and standard deviations reported; (3) full article avail- the same HC group, therefore, a multivariate meta-analysis was
able in English; (4) published in a peer-reviewed journal. Studies conducted using the rma.mv command. Between-study hetero-
measuring only related constructs (e.g. ‘secondary social phobia’, geneity was calculated using Cochran’s Q test. Where heterogen-
‘social appearance anxiety’) were not included. eity was found ( p < 0.05), the meta-regressions were performed
using age and SA measure as moderators. BMI could not be
used as a moderator due to missing data.
Information sources and search
The electronic databases PubMed, PsycInfo, SCOPUS, and Web
Risk of bias across studies
of Science were searched independently by JKG and AH for
papers up to February 2018. Search terms included social anxiety The presence of publication bias was assessed through visual
OR social phobia AND anorexia nervosa OR bulimia nervosa OR inspection of funnel plots, where the absence of studies in the

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 Psychological Medicine 3

Fig. 1. Systematic review search process.

bottom right corner indicates publication bias. The symmetry of Studies included in the meta-analysis and qualitative review
the funnel plots was formally assessed using Begg’s rank correl- Two of the 12 studies did not report the mean age of participants.
ation test (Begg & Mazumdar, 1994). Publication bias was also Nine studies did not report the mean BMI or percentage of Ideal
assessed using Rosenthal’s fail-safe N (Rosenthal, 1979), which Body Weight (IBW) in at least one participant group, and 10
estimates the number of unpublished studies required to change studies did not report mean illness duration of the ED group.
the significant effect size into a non-significant one. Most studies only included female participants, however, two
studies included a small proportion of males in their ED group,
and six included males in their HC group. Two studies did not
Results report participants’ gender in at least one group. ED groups
Study selection were most often inpatients (n = 6). Three studies did not report
the recruitment source for their control group, and one did not
Thirty-eight studies were included in the review (Table 1). Studies report this information for the ED group. Groups were matched
that used the same sample (Hinrichsen et al. 2004a, b; 2007a; on some characteristic in eight of the studies, most often sex.
Duclos et al. 2014; Courty et al. 2015) are combined and the
data considered together. Fourteen studies included a HC Studies included in the qualitative review
group, however, two could not be included in a meta-analysis All 26 studies reported the mean age of participants. Four studies
and are discussed in the qualitative review. Of the 12 studies did not report mean BMI or percentage IBW in at least one par-
that could be included, 10 included an AN group and five ticipant group, and half did not report mean illness duration.
included a BN group. One of the BN studies reported scores for Again, most studies included exclusively female participants,
males and females separately (Gross & Rosen, 1988), however, however, six studies included males. One study did not report
males could not be included in the meta-analysis due to too participants’ gender. ED groups were most often recruited from
few cases. Only one of the studies included in the meta-analysis inpatient services (n = 8), but specialist ED or psychiatric services
provided mean SA scores for the different AN sub-types, there- where it was ambiguous as to whether patients were outpatients or
fore, meta-analyses by sub-type could not be performed. inpatients were also common (n = 7). In studies where SA was
compared between groups (n = 13), groups were most often
matched by sex (n = 8), however, five studies did not report a
Study characteristics group matching technique.
Overall, reporting of study characteristics varied considerably. All
studies provided information on the SA measure used, and in
Synthesis of results and risk of bias
total, 12 different self-report questionnaires were used to measure
SA. The most frequently used (n = 11) was the Liebowitz Social The random-effects model with a total sample size of 1859 parti-
Anxiety Scale (LSAS; Liebowitz, 1987). cipants (AN = 281, HC = 1578) revealed that those with AN had

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at

4
Table 1. Characteristics of studies

Mean (yrs)
Mean illness
Mean (SD) SA age duration % Groups
Study Sample score Measure (SD) Mean BMI (SD) (SD) female Recruitment source matched by

Abbate-Daga et al. 56 AN Baseline 37.16 BSPS 25.03 16.31 (2.66) 7.8 (5.34) 100 Day hospital NA
(2015) (12.66) (5.75)
6 months 26.42
(9.66)
12 months 22.81
(10.30)
Allen & Craighead 11 BED Pre-treatment FNE 21 122.82 (22.86)% NR 100 University advertisement Sex
(1999) (treatment) 23.09 (4.11) (1.2)a IBW
Post-treatment
20.0 (5.0)
9 BED (waitlist) Pre-treatment 116.5 (21.98)% NR 100 University advertisement
23.11 (9.68) IBW
Post-treatment
22.33 (8.66)
Buchholz et al. (2007) 149 ED 15.02 (7.17) MASC social 15.65 NR NR 100 Tertiary care childrens NA
anxiety (1.17) hospital
subscale
Bulik et al. (1991) 23 AN 86.2 (26.4) SPAI 20.3 NR NR 100 Inpatient unit Sex
difference (8.3)
scores
54 BN 84.9 (27.9) 22.4 NR NR 100 Inpatient unit
(6.0)
43 SAD 98.7 (35.3) 36.3 NR NR 100 Social phobia clinic
(9.6)
50 HC 54.6 (35.9) 18.5 NR NA 100 University
(1.2) undergraduates
Dakanalis et al. (2016) 189 young 30.29 (12.61) SIAS 12.59 AN = 15.52 NR 100 Specialist child and NR
female (0.70)a (1.24), BN = adolescent ED service
adolescent ED 37.33 (12.49) SPS 21.18 (2.07),
39.86 (11.05) BFNE EDNOS = 20.05
(3.54)
15 young male 29.88 (12.24) SIAS NR 0
adolescent ED
32.31 (13.13) SPS

Jess Kerr-Gaffney et al.

32.01 (12.88) BFNE
444 older 37.01 (13.06) SIAS 16.74 NR 100
female (0.61)a
adolescent ED
44.41 (13.88) SPS
45.55 (10.55) BFNE
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at

55 older male 36.65 (13.32) SIAS NR 0

Psychological Medicine
adolescent ED
34.35 (12.96) SPS
34.69 (11.11) BFNE
Duclos et al. (2014) and 60 AN Baseline 48.3 LSAS 16 (1.6) 16.9 (1.1) 1.38 (0.57) 100 Inpatient unit NA
Courty et al. (2015) (31.3)
6 months 33.1
(27.9)
12 months 33.2
(29.7)
18 months 27.2
(26.3)
Flament et al. (2001) 29 AN-R 50 (10.30) LSAS 17.9 NR 2 (3) 93.1 Inpatient unit NR
(4.3)
34 BN 54 (16.51) 26.6 NR 7 (6) 100 Outpatient clinic
(6.5)
Gilboa-Schechtman 20 AN 73.05 (37.28) LSAS 16.60 NR NR 100 Outpatient clinic Age, sex,
et al. (2006) (2.48) education
20 BN 58.60 (42.48) 19.65 NR NR 100 Outpatient clinic
(5.01)
20 HC 33.90 (18.67) 19.65 NR NA 100 Community
(5.01) advertisement
Goddard & Treasure 63 ED 61.6 (31.6) LSAS 21.8 16.8 (2.5) 4.0 (2.0– 100 ED clinics Age, sex,
(2013) (5.5) 7.6)b education,
IQ
50 HC 29.8 (17.8) 21.5 21.1 (2.0) NA 100 University advertisement
(5.9)
Grabhorn et al. (2006) 30 AN 33.5 (12.9) SIAS 25.5 NR NR 100 Inpatient clinic referrals Sex
(7.7)
29.3 (11.3) SPS
30 BN 39.4 (12.9) SIAS 24.9 NR NR 100
(6.8)
35.4 (16.2) SPS
30 Depression 30.6 (14.9) SIAS 41.1 NR NR 100
(10.9)
21.3 (13.0) SPS
30 Anxiety 23.6 (14.5) SIAS 36.9 NR NR 100
(12.8)
19.5 (18.1) SPS
Gross & Rosen (1988) 65 BN (female) 9.49 (7.03) SADS NR NR NR 100 Public schools Sex
612 HC 6.98 (6.14) NR NR NA 100
(female)
(Continued )

5
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at

Table 1. (Continued.)

6
Mean (yrs)
Mean illness
Mean (SD) SA age duration % Groups
Study Sample score Measure (SD) Mean BMI (SD) (SD) female Recruitment source matched by

8 BN (male) 15.29 (5.9) NR NR NR 0

645 HC (male) 7.69 (4.5) NR NR NA 0
Hinrichsen et al. (2003) 21 AN-R 23.5 (8.68) FNE 25.7 NR NR 100 Specialist ED service Sex
(9.06)
34 AN-BP 26.7 (4.32) 28.0 NR NR 100
(6.79)
59 BN 23.6 (6.72) 26.9 NR NR 100
(6.56)
50 HC 18.5 (6.97) 19.8 NR NA 100 University
(0.86) undergraduates
Hinrichsen et al. 70 ED 88.41 (35.10) SPAI social 27.9 22.9 (11.53) NR 100 Specialist ED service NA
(2004a); (2004b); phobia (8.76) referrals (outpatient)
(2007a) subscale
Hinrichsen et al. (2007b) 191 ED 27.4 (7.84) BFNE 28.4 20.4 (6.77) NR 100 Specialist ED service NA
(8.62)
Jiménez-Murcia et al. 50 BN 13.7 (8.50) SADS 28.1 NR NR 100 Psychiatric department Sex
(2015) (8.2) referrals
49 BN + 15.7 (7.64) 26.9 NR NR 100
compulsive (9.1)
buying
MacDonald et al. (2014) 171 BN 30.26 (14.66) SPIN 26.2 22.7 (5.33) 9.2 (8.2) 96.5 Day hospital NA
(8.2)
Mattar et al. (2012a) 155 AN 57.73 (15.85) LSAS 20.90 14.43 (1.46) 4.29 (4.71) 100 Inpatient unit NA
(6.16)
Mattar et al. (2012b) 24 AN-R Baseline 47.05 LSAS 16.38 13.84 (1.26) 0.98 (0.82) 100 Inpatient unit NA
(28.30) (1.93)
EOT 24.95 (26.91)
McFarlane et al. (2015) 299 ED 31.5 (15.1) SPIN 26.0 17.1 (1.0) 8.3 (7.3) 97 Day hospital NR
(7.8)
130 ED 33.9 (15.5) 30.9 17.0 (1.0) 14.4 (11.2) 96.8
(11.1)

Jess Kerr-Gaffney et al.

Melfsen et al. (2006) 48 AN 15.35 (9.25) SPAI-C NR NR NR NR Child and adolescent NR
psychiatric departments
31 SAD 29.59 (9.79) 12.19 NR NR 58.1
(2.59)
7 AS 20.77 (13.77) 15.71 NR NR 28.6
(2.5)
1197 HC 12.51 (7.87) 12.51 NR NA 51.5 NR
(2.05)
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at

Obeid et al. (2013) 182 AN-R or 14.49 (7.32)c, 14.2 MASC social 15.6 17.02 (2.27)c, NR 100 Children’s tertiary care Sex

Psychological Medicine
EDNOS-R (7.19)d anxiety (1.39) 17.05 (2.14)d facility
subscale
99 AN-BP or 15.21 (6.04)c, 20.55 (3.50)c, NR 100
EDNOS-BP 16.14 (6.94)d 19.94 (3.71)d
63 BN 19.89 (5.71)c, 22.02 (2.62)c, NR 100
14.08 (6.78)d 22.30 (3.79)d
Ostrovsky et al. (2013) 29 BED 50.3e SPIN 36.0 33.7 (6.7)a NR 86.8a Online and university NR
(12.8)a advertisements
202 HC 32.6e NA
f e
Ohmann et al. (2013) 29 AN Baseline 24.1 SIAS 14.3 15.7 (1.3) 0.6e 100 Inpatient unit NA
(9.1)
Baseline 15.4 SPS
(9.9)
9 months 26.7 SIAS 17.8 (1.7)
(14.3)
9 months 14.6 SPS
(15.0)
Russell et al. (2018) 16 AN Baseline 59.4 LSAS 22.4 16.61 (1.77) NR 100 Inpatient unit Sex
(oxytocin) (29.1) (3.6)
Follow-up 59.4 18.00 (1.86)
(28.9)
17 AN Baseline 63.1 23.5 16.75 (1.36) NR 100
(placebo) (24.4) (10.2)
Follow-up 58.4 18.10 (1.29)
(27.3)
Sawaoka et al. (2012) 113 BED 15.13 (4.51) SCS social 45.03 37.1 (7.3) NR 77.9 Newspaper NA
anxiety (8.30) advertisements
subscale
Schmelkin et al. (2017) 19 AN 61.95 (30.53) LSAS 25.1 17.7 (0.2) NR 100 Community Sex
(1.7)
23 AN-WR 36.74 (19.15) 22.9 22.5 (0.4) NR 100
(0.5)
28 HC 22.25 (15.68) 23.9 22.6 (0.3) NA 100
(0.8)
Schneier et al. (2016) 30 AN 54.1 (26.1) LSAS 26.9 NR NR 97 Media notices and NR
(7.5) referrals from health
professionals
43 SAD 76.4 (19.8) 29.9 NR NR 53
(7.5)
50 OCD 24.2 (16.7) 29.2 NR NR 50
(5.9)
74 HC 11.5 (8.0) 28.9 NR NA 51
(7.6)
(Continued )

7
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
Table 1. (Continued.)

8
Mean (yrs)
Mean illness
Mean (SD) SA age duration % Groups
Study Sample score Measure (SD) Mean BMI (SD) (SD) female Recruitment source matched by

Schulze et al. (2009) 23 AN 17.4 (9.7) SPAI-C 14.69 14.7 (1.58) NR 100 Inpatients at a child and NR
(1.54) adolescent psychiatric
department
145 PC 16.52 (10.77) 13.29 NR NR NR NR
(2.86)
1197 HC 12.51 (7.87) 12.51 NR NA 51.5 NR
(2.05)
Schwalberg et al. (1992) 20 BN 14.7 (5.9) SCS social 26.35 104.5 (12.5)% NR 100 ED clinics Sex
anxiety (6.44) IBW
subscale
20 BED 15.1 (5.6) 41.18 157.9 (31.8)% NR 100
(7.66) IBW
20 SAD 19.6 (3.7) 34.70 NR NR 100 Anxiety disorder clinics
(9.74)
20 PD 13.9 (6.1) 31.50 NR NA 100
(6.90)
Solano et al. (2005) 35 AN & BN (SI) 19.41 (6.79) SADS 22.31 18.85 (4.12) 4.83 (3.69) 100 Outpatient clinic Sex
(4.46)
74 AN & BN (no 15.09 (9.05) 23.24 19.43 (4.70) 5.82 (5.74) 100
SI) (6.08)
Steinglass et al. (2017) 27 AN 52.44 (23.07) LSAS 27.7 17.5 (1.0) NR 100 Inpatient unit Age, sex,
(7.5) ethnicity
44 SAD 75.72 (20.05) 30.0 (4) 23.9 (6.3) NR 57 Outpatient clinic
50 OCD 24.18 (16.66) 29.2 24.6 (5.3) NR 48 Outpatient clinic
(5.8)
75 HC 11.39 (7.96) 29.0 24.1 (4.4) NA 52 NR
(7.6)
Steinman et al. (2016) 26 AN 47.81 (20.14) LSAS 26.93 NR NR 100 Inpatients NR
(7.67)
37 SAD 75.00 (19.32) 28.54 NR NR 59 Media and referrals from
(6.66) health professionals
45 OCD 23.04 (17.03) 28.80 NR NR 47
(5.89)

Jess Kerr-Gaffney et al.

62 HC 12.42 (8.01) 27.60 NR NA 53
(6.50)
Striegel-Moore et al. 34 BN 19.71 (5.39) SCS social 23.36 21.64 (2.76) 6.8 (4.15) 100 ED clinics Age, sex,
(1993) anxiety (5.8)a ethnicity,
subscale BMI
33 sub-clinical 19.24 (4.49) 22.47 (2.92) NA 100 University
ED undergraduates and
newspaper
67 HC 15.01 (5.0) 22.08 (2.87) NA 100 advertisements
 Psychological Medicine 9

significantly higher levels of SA than HCs, with a large effect

AN, anorexia nervosa; AN-BP, anorexia nervosa binge-purge sub-type; AN-R, anorexia nervosa restricting sub-type; AN-WR, anorexia nervosa weight restored; AS, Asperger’s syndrome; BED, binge eating disorder; BFNE, Brief Fear of Negative Evaluation
scale; BMI, body mass index; BN, bulimia nervosa; BSPS, brief social phobia scale; ED, eating disorder; EDNOS, eating disorder not otherwise specified; FNE, Fear of Negative Evaluation scale; HC, healthy control; IBW, ideal body weight; IQ, intelligence
quotient; LSAS, Liebowitz Social Anxiety Scale; MASC, Multidimensional Anxiety Scale for Children; NA, not applicable; NR, not reported; OCD, obsessive compulsive disorder; PC, psychiatric control; PD, panic disorder; SAD, social anxiety disorder; SADS,
Social Avoidance and Distress Scale; SA, social anxiety; SCS, Self-Consciousness Scale; SD, standard deviation; SI, self-injury; SIAS, Social Interaction Anxiety Scale; SPAI, Social Phobia & Anxiety Inventory; SPAI-C, Social Phobia & Anxiety Inventory for
size [d = 1.65, (95% CI 1.03–2.27) z = 5.20, p < 0.001] (Fig. 2).
education
Age, sex,
The funnel plot for the AN studies is displayed in Fig. 3.
There was no evidence of publication bias (Begg’s test p = 0.216,
NR

Rosenthal’s fail-safe N = 1033).

The random-effects model with a total sample size of 1031
participants (BN = 232, HC = 799) showed that those with BN
at a child and adolescent
Inpatients or outpatients
Inpatient eating disorder

had significantly higher levels of SA than HCs, with a medium

eating disorder clinic

effect size [d = 0.71, (95% CI 0.47–0.95) z = 5.74, p < 0.001]

(Fig. 4). The funnel plot for the BN studies is displayed in
Fig. 5. There was no evidence of publication bias (Begg’s test
High school

p = 0.817, Rosenthal’s fail-safe N = 112).

clinic

NR

Additional analyses
There was evidence of significant heterogeneity in the AN studies
Q(9) = 131.14, p < 0.001, therefore, meta-regressions with age and
100

100
NR

NR

SA measure as moderator variables were performed. BMI and ill-

ness duration could not be included as moderators due to studies
not reporting this information. The moderators explained a sig-
nificant amount of the variance, QM(4) = 32.56, p < 0.001. Age
had a significant influence on the size of the effect [b = 0.12,
0.92
NR

NR

NA

(95% CI 0.03–0.21), z = 2.53, p = 0.01], as did using the LSAS as

a measure of SA [b = 1.64, (95% CI 0.78–2.50), z = 3.73, p <
0.001]. The test for residual heterogeneity was not significant,
QE(4) = 7.96, p = 0.09. There was no evidence of heterogeneity
in the BN studies Q(3) = 0.67, p = 0.87.
15.7 (1.4)

14.8 (3.3)

16.2 (1.2)

NR

Qualitative review
Differences between ED and HC
A few studies comparing ED groups to HCs could not be included
(3.1)

(1.1)

(1.8)

(1.1)
21.3

15.5

15.5

14.9

in the meta-analysis due to there being too few comparisons.

Ostrovsky et al. (2013) examined SA scores in individuals with
BED compared with overweight controls, finding that those
with BED had significantly higher SA scores than controls. The
SPAI social

SPAI social
subscale

subscale

second study examined SA scores in a mixed ED group compared

phobia

phobia

with controls, finding that the ED group had significantly higher

SA scores than HCs (Goddard & Treasure, 2013). This study also
compared SA scores of parents of daughters with EDs to parents
of HCs. Parents of daughters with EDs had higher SA scores
than control parents, however, effect sizes were small and not
Four additional follow-up periods included in paper excluded here for brevity.
108.6 (41.4)

103.6 (37.6)

68.09 (25.1)

significant.
80 (31.4)

Children; SPIN, Social Phobia Inventory; SPS, Social Phobia Scale

Differences between ED diagnoses

Of the seven studies that assessed differences between AN
and BN, six found no difference in SA between groups (Bulik
48 adolescent
23 adult AN

et al. 1991; Flament et al. 2001; Solano et al. 2005; Gilboa-

10 AN-R

Schechtman et al. 2006; Grabhorn et al. 2006; Obeid et al.

22 HC

2013). The single study that reported differences between ED

AN

Values reported for groups combined.

groups found that individuals with AN-BP had significantly

higher SA scores than AN-R and BN (Hinrichsen et al. 2003).
Median and inter-quartile range.

Of the five studies that assessed differences between AN-R and

AN-BP, four found no differences in SA across AN subtypes
Zonnevylle-Bender

Zonnevylle-Bender

(Mattar et al. 2012a; Obeid et al. 2013; Duclos et al. 2014;

Abbate-Daga et al. 2015; Courty et al. 2015). As before, the single
Late-adolescents.
Mid-adolescents.

No SD reported.
et al. (2004)

et al. (2005)

study that did find a difference was Hinrichsen et al. (2003).

One study examined whether patients that met eating disorder
not otherwise specified (EDNOS) criteria in the DSM-IV but BN
criteria in the DSM-5 differed from patients who met DSM-IV
b

d
a

e
c

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 10 Jess Kerr-Gaffney et al.

Fig. 2. Forest plot of standardised mean effect size for differences (SMD) between anorexia nervosa (AN) and healthy controls (HC). CI, confidence interval; FNE,
Fear of Negative Evaluation scale; LSAS, Liebowitz Social Anxiety Scale; SPAI, Social Phobia & Anxiety Inventory; SPAI-C, Social Phobia & Anxiety Inventory for
Children.

criteria for BN (MacDonald et al. 2014). SA scores did not differ inclusion and 6-month follow-up. Further, SA levels were related
between groups. Finally, a study by Schwalberg et al. (1992) exam- to alexithymia across time, even after adjusting for depression,
ined differences between BN and BED, finding that the groups did anxiety, and BMI.
not differ in SA scores. Overall, it seems that SA is similarly ele- The third study in adolescents with AN followed 29 outpatients
vated across ED diagnostic groups. receiving group cognitive behavioural therapy (G-CBT), assessing
psychiatric, social, and emotional variables before, during (3 and
6 months), at the end of (9 months), and 1 year after completing
Treatment effects and studies with recovered patients treatment (Ohmann et al. 2013). Patients were split into groups
based on outcomes. It was found that SA significantly improved
Six studies examined change in SA over treatment, two of which
by 9 months in patients with a good outcome (defined as attaining
involved adolescent patients with AN admitted to inpatient care.
25th BMI percentile and normal eating patterns), however, SA did
Mattar et al. (2012b) assessed 24 patients at admission and dis-
not improve in those with a poor or intermediate outcomes.
charge (mean time in treatment = 3.2 months), in which time
Different from the aforementioned studies, the fourth treatment
mean SA scores significantly improved. Neither intensity of
study (Abbate-Daga et al. 2015) examined 56 adult women with
weight loss or BMI at admission, discharge, or improvements in
AN attending a day hospital service. The programme took a multi-
BMI during treatment were correlated with SA scores. The second
disciplinary approach with a focus on psychodynamic psychother-
study (Courty et al. 2015) assessed 60 patients in the second half
apy, and patients were assessed at baseline, end of treatment (EOT;
of their inpatient admission (21 weeks on average). Patients were
6 months), and at follow up, 12 months after EOT. Significant
assessed at 6, 12, and 18 months. SA scores significantly decreased
reductions in SA scores were seen at EOT and follow-up. Unlike
across time, with the largest reduction occurring between
the inpatient studies, neither of these studies examined whether
decreases in SA were due to improvements in BMI.
The final treatment study in AN was a randomised placebo-
controlled trial examining the effects of intranasal oxytocin in 33
inpatients with AN (Russell et al. 2018). Contrary to predictions,
there were no significant treatment, time, or treatment by time
effects on SA scores, however EDE eating concern scores and cog-
nitive rigidity were improved in the oxytocin group compared with
placebo. The finding that SA scores did not improve over treatment
in either group is at odds with the results of the former studies, how-
ever, this might be due to the shorter follow-up period (4–6 weeks).
The final treatment study involved 29 young adult women with
BED, who were randomly assigned to a cognitive-behavioural inter-
vention (‘Appetite Awareness Training’) or a wait-list control group
for 8 weeks (Allen & Craighead, 1999). It was found that SA scores
reduced significantly in the intervention group compared with the
control group. The intervention group also saw significant improve-
ments in various measures of binge eating.
Finally, one study examined differences in SA scores between
women with acute AN, women recovered from AN, and HCs
(Schmelkin et al. 2017). Women recovered from AN scored sig-
nificantly higher than HCs, but significantly lower than acute
Fig. 3. Funnel plot of anorexia nervosa (AN) studies included in the meta-analysis to AN on the social fear, public fear, and social avoidance sub-scales
assess publication bias. of the LSAS. However, on the public avoidance sub-scale, those

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 Psychological Medicine 11

Fig. 4. Forest plot of standardized mean effect

size for differences (SMD) between bulimia ner-
vosa (BN) and healthy controls (HC). CI, con-
fidence interval; FNE, Fear of Negative
Evaluation scale; LSAS, Liebowitz Social Anxiety
Scale; SADS, Social Anxiety and Distress Scale;
SCS, Self-Consciousness Scale; SPAI, Social
Phobia & Anxiety Inventory.

with AN scored higher than HCs and recovered AN, who did not involved AN patients only, both finding that current BMI was
differ from one another. Thus, while it seems that SA significantly not associated with SA scores in inpatient women (Mattar et al.
improves with treatment in AN, those recovered from the dis- 2012a, b). In addition, Mattar et al. (2012a) found that SA scores
order still experience high levels of SA compared with HCs. It were negatively correlated with blood albumin levels (an indicator
also appears that improvements in SA in AN are not related to of nutritional status), and positively correlated with age and dur-
a specific treatment modality, although further studies with con- ation of illness. In a sample of women with AN or BN, Bulik et al.
trol groups are required to confirm this finding. (1991) found no difference in SA scores when patients were split
into underweight and normal weight groups. Similarly, in a mixed
ED group (AN, BN, and EDNOS), SA was positively correlated
Associations with BMI with duration of illness and number of previous hospital admis-
Six studies examined whether SA was associated with BMI and sions, but not current BMI (Goddard & Treasure, 2013).
other clinical indicators of ED severity. Two of these studies Further, those who were currently on medication had significantly
higher SA than those not taking medication. Finally, in indivi-
duals with BED, higher SA was reported in those who became
overweight as children but was not correlated with current BMI
(Sawaoka et al. 2012; Ostrovsky et al. 2013). Thus, it seems that
across the ED spectrum, SA is not related to BMI. Despite the
lack of studies in this area, this finding suggests that high SA in
those with EDs is not a result of malnutrition. Instead, those
with a more severe illness may have higher levels of SA, as evi-
denced by the associations with longer duration of illness, more
hospital admissions and medication use.

Associations with psychopathology

Consistent with the hypothesis that SA may be associated with a
more severe illness, several studies have found positive associa-
tions between ED psychopathology and SA across the ED spec-
trum. In the same studies that found no association between SA
and BMI in BED, significant positive relationships were found
between SA and self-consciousness, depressive symptoms,
Eating Disorder Examination (EDE) scores, weight, shape, and
eating concerns, and binge eating severity (Sawaoka et al. 2012;
Fig. 5. Funnel plot of bulimia nervosa (BN) studies included in the meta-analysis to Ostrovsky et al. 2013). Similarly, in a mixed ED sample (AN,
assess publication bias. BN, or EDNOS), SA was significantly positively correlated with

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 12 Jess Kerr-Gaffney et al.

all eight Eating Disorder Inventory (EDI) subscales (drive for SA than HCs. Only one study compared levels of SA in BED com-
thinness, bulimia, body dissatisfaction, ineffectiveness, perfection- pared with HCs, finding that SA was also significantly elevated in
ism, interpersonal distrust, interoceptive awareness, and maturity individuals with BED compared with HCs. While there was no
fears), as well as core beliefs about abandonment and emotional evidence of publication bias in either meta-analysis, there was sig-
inhibition (Hinrichsen et al. 2004a, b; 2007a). This latter finding nificant heterogeneity across AN studies. Meta-regressions with
was partially replicated in a larger sample by Hinrichsen et al. age and SA measure as moderator variables revealed that these
(2007b), who found that core beliefs regarding abandonment variables explained a significant amount of the heterogeneity,
and defectiveness/shame (the belief that one is fundamentally such that use of the LSAS and older age of participants was asso-
flawed) explained almost a quarter of the variance in SA. ciated with larger effect sizes. The association between the SA
Further, another study using the EDI on a sample of children measure used by studies (namely the LSAS) and larger effect
and adolescents with AN found SA scores were positively asso- sizes has important implications for both research in this area
ciated with ineffectiveness and interpersonal distrust subscales, and clinical practice. Self-report measures of SA have several
as well as trait anxiety (Schulze et al. 2009). advantages: they are quicker and easier to administer, and can
Two studies examined SA and psychopathological variables in provide an estimate of SA in those who do not meet full diagnos-
adolescent mixed ED samples (AN, BN, and EDNOS). Buchholz tic criteria for SAD. However, the results from this meta-analysis
et al. (2007) demonstrated that SA was a unique predictor of body suggest that SA in EDs may be over- or under-estimated, depend-
dissatisfaction, and was also significantly positively associated ing on the measure used. While cut-off scores on self-report mea-
with ‘self-silencing’ (keeping negative thoughts or feelings to one- sures have been established in groups with a diagnosis of SAD,
self), while Obeid et al. (2013) found a negative association there are outstanding questions regarding what can be considered
between SA and self-esteem. Finally, one study found different a clinically significant level of SA in individuals with EDs.
results based on ED diagnosis. Hinrichsen et al. (2003) examined Identifying those with high SA may be useful when deciding on
emotion coping strategies in women with EDs, reporting that the type of treatment offered to patients. Further, while some
while SA was associated with greater dissociation among those scales measure SA unidimensionally, others provide sub-scores
with AN-R, this was not the case for those with AN-BP, BN, or for different aspects of SA, such as fear, avoidance, and physio-
HCs. Instead, higher levels of SA in BN and HC women were logical arousal. Which of these types of measure would be most
associated with higher bulimic psychopathology. Thus, it can be useful in ED populations is another question for future research.
seen that SA in EDs is associated with not only more severe ED Physiological arousal as it relates to SA in EDs may be a particu-
symptoms, but also beliefs and behaviours regarding self-esteem larly interesting domain to explore, given the reduced sensitivity
and emotion regulation. to interoceptive signals reported in AN (Pollatos et al. 2008).
The association between age and higher SA scores was also
found in a few AN studies not included in the meta-analysis
Associations with other comorbid symptoms
(Zonnevylle-Bender et al. 2004; Mattar et al. 2012a). One explan-
Two studies assessed differences in SA between those with and ation for this finding is that those with a longer illness duration
without comorbid symptoms. Solano et al. (2005) examined dif- may experience higher levels of SA, in agreement with the results
ferences between women with AN and BN who did and did of Goddard & Treasure (2013). This may indicate a more severe
not engage in self-injurious behaviour (SIB). Interestingly, while illness, as suggested by the positive association between SA and
there was no effect of diagnosis on SIB, it was found that those ED psychopathology (Hinrichsen et al. 2003, 2004b; Schulze
who self-injured had significantly higher ED psychopathology, et al. 2009; Sawaoka et al. 2012; Ostrovsky et al. 2013). This find-
SA scores, and body image disturbance than those who did not. ing has important implications for understanding the etiological
The second study examined differences in SA in women with link between EDs and SA. From a developmental perspective, it
BN, who either did or did not have a comorbid compulsive buy- has been postulated that there is a social phenotype for those at
ing (CB) disorder (Jiménez-Murcia et al. 2015). CB is not cur- risk of developing an ED, characterised by loneliness, shyness,
rently recognised by international diagnostic classifications, but internalising problems, inferiority, and low social support in
shares similarities with other impulse-related disorders such as childhood (Fairburn et al. 1999; Krug et al. 2013; Treasure &
BN and gambling disorder. It was found that SA scores were Schmidt, 2013). Indeed, SAD mostly occurs before ED onset in
higher in women with both BN and CB, compared with those individuals diagnosed with both disorders (Bulik et al. 1997;
with BN only. However, after adjusting for age, this difference Godart et al. 2000; Kaye et al. 2004; Swinbourne et al. 2012),
was no longer significant. Thus, although the results of Solano and may be exacerbated by the ill state. The finding that levels
et al. (2005) suggest that high SA in EDs may be associated of SA in individuals recovered from AN lie between that of
with more severe illness features such as self-harm, this might HCs and acutely ill individuals lends further support for this
not be the case for addictive-type disorders. More research is hypothesis (Schmelkin et al. 2017).
required to clarify the possible transdiagnostic nature of SA in Importantly, the lack of any association between BMI and SA
EDs and other disorders. indicates that it is not the degree of malnutrition that exacerbates
SA, but some other factor associated with the illness. One possible
explanation concerns emotional avoidance. Many have theorised
Discussion
that ED psychopathology (for example, a focus on food and
The aim of this review was to examine group differences in SA in weight, restrictive behaviours, and binge eating) helps individuals
EDs compared with HCs, and provide a qualitative synthesis of avoid having to experience negative emotions and challenging
the literature. There were significant differences in SA scores interpersonal situations (Slade, 1982). It has been demonstrated
between both AN (10 studies) and BN (five studies) compared that comorbid depressive and anxiety symptoms are associated
with HCs, with large and medium effect sizes, respectively, indi- with higher ED psychopathology in AN, and this relationship is
cating that those with AN or BN have significantly higher levels of almost fully mediated by emotional avoidance (Wildes et al.

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 Psychological Medicine 13

2010). Therefore, it could be the case that those with higher social examined, but such research may be important in improving
anxiety avoid situations that may elicit high emotion through an prognosis for EDs. Furthermore, research in this area may lead
intense focus on food and weight, therefore, reinforcing and to new insights into common illness pathways and transdiagnostic
maintaining the disorder. Further studies examining the relation- factors for AN and other disorders, such as ASD or SAD.
ship between SA (rather than general anxiety symptoms) and
Acknowledgements. JKG received financial support from the Economic and
emotional avoidance are required to test this hypothesis.
Social Research Council (ESRC), and would like to thank Jenni Leppanen for
her assistance in using the statistical programme. KT would like to acknow-
Clinical implications ledge financial support from MRC and MRF child and young adult mental
health – the underpinning aetiology of self-harm and eating disorders and
Findings from the current review contribute to the broader litera- Swiss Anorexia Nervosa Foundation (grant 58-16).
ture on socio-emotional functioning in EDs, which have demon-
strated problems in areas such as theory of mind (Bora & Declaration of interest. None.
Kose, 2016), emotion expression and recognition (Caglar-Nazali
et al. 2014; Davies et al. 2016), social anhedonia (Harrison
et al. 2014), and alexithymia (Westwood et al. 2017a). Further, References
they contribute to a growing evidence base documenting high
Abbate-Daga G et al. (2015) Day hospital treatment for anorexia nervosa:
levels of autism spectrum disorder (ASD) traits in those with
a 12-month follow-up study. European Eating Disorders Review 23,
AN (Westwood et al. 2017b). Like those with EDs, individuals 390–398.
with ASD show high levels of SA, with over half of adolescents Adambegan M et al. (2012) Internalizing and externalizing behaviour pro-
meeting clinical cut-offs on self-report measures (Kuusikko blems in childhood contribute to the development of anorexia and bulimia
et al. 2008). Because of the high degree of symptom overlap nervosa – A study comparing sister pairs. European Eating Disorders Review
between SAD and ASD, it is possible that the high levels of SA 20, 116–120.
displayed in EDs are linked to ASD traits. While it may not be Allen HN and Craighead LW (1999) Appetite monitoring in the treatment of
possible to delineate the unique contributions of SAD and ASD binge eating disorder. Behavior Therapy 30, 253–272.
traits to the social difficulties seen in AN and other EDs, these Allen KL et al. (2013) DSM–IV–TR and DSM-5 eating disorders in adoles-
difficulties may be useful targets for treatment. For example, cents: prevalence, stability, and psychosocial correlates in a population-
based sample of male and female adolescents. Journal of Abnormal
there is evidence that social skills training in adolescents with
Psychology 122, 720–732.
EDs may improve self-esteem and social withdrawal (Lázaro Antony MM et al. (1998) Dimensions of perfectionism across the anxiety dis-
et al. 2011). In addition, preliminary evidence suggests that orders. Behaviour Research and Therapy 36, 1143–1154.
Cognitive Remediation and Emotion Skills Training (CREST), Begg CB and Mazumdar M (1994) Operating characteristics of a rank correl-
an intervention designed to target emotion processing, de- ation test for publication bias. Biometrics 50, 1088–1101.
creases social anhedonia and alexithymia in adults with AN Berkman ND, Lohr KN and Bulik CM (2007) Outcomes of eating disorders:
(Tchanturia et al. 2015). Future research into the influence of a systematic review of the literature. International Journal of Eating
SA on outcomes and prognosis is warranted. Disorders 40, 293–309.
Blinder BJ, Cumella EJ and Sanathara VA (2006) Psychiatric comorbidities
of female inpatients with eating disorders. Psychosomatic Medicine 68,
Limitations 454–462.
Bora E and Kose S (2016) Meta-analysis of theory of mind in anorexia ner-
Several limitations of this review should be noted. Firstly, a con- vosa and bulimia nervosa: a specific impairment of cognitive perspective
siderable number of the studies included did not report important taking in anorexia nervosa? International Journal of Eating Disorders 49,
participant characteristics, such as BMI. Therefore, BMI could not 739–740.
be entered as a moderator variable in the meta-analysis. A further Buchholz A et al. (2007) Self-silencing in a clinical sample of female adoles-
limitation is that none of the studies included in this review exam- cents with eating disorders. Journal of the Canadian Academy of Child and
ined SA in EDNOS or OSFED (‘not otherwise specified’ categor- Adolescent Psychiatry 16, 158–163.
ies in the DSM-IV and DSM-5, respectively) compared with HCs. Buckner JD, Silgado J and Lewinsohn PM (2010) Delineation of differential
Considering such diagnoses make up a significant proportion temporal relations between specific eating and anxiety disorders. Journal of
Psychiatric Research 44, 781–787.
of those with EDs (Allen et al. 2013; Fairweather-Schmidt &
Bulik CM et al. (1991) An analysis of social anxiety in anorexic, bulimic,
Wade, 2014), establishing whether these patients show similar
social phobic, and control women. Journal of Psychopathology and
social difficulties will have important implications for their treat- Behavioral Assessment 13, 199–211.
ment. Finally, the number of studies that could be included in the Bulik CM et al. (1997) Eating disorders and antecedent anxiety disorders: a
meta-analyses (especially for BN studies) was relatively few, since controlled study. Acta Psychiatrica Scandinavica 96, 101–107.
the majority of studies did not include a HC group. Caglar-Nazali HP et al. (2014) A systematic review and meta-analysis of
‘systems for social processes’ in eating disorders. Neuroscience and
Biobehavioral Reviews 42, 55–92.
Conclusions Cederlof M et al. (2015) Etiological overlap between obsessive-compulsive dis-
Both AN and BN are characterised by high levels of SA, even in order and anorexia nervosa: a longitudinal cohort, multigenerational family
those who do not have a formal diagnosis of SAD. SA in these and twin study. World Psychiatry 14, 333–338.
Ciarma JL and Mathew JM (2017) Social anxiety and disordered eating: the
patients is associated with longer illness duration, older age, and
influence of stress reactivity and self-esteem. Eating Behaviors 26, 177–181.
higher ED psychopathology, suggesting that SA may be indicative Cohen J (1988) Statistical Power Analysis for the Behavioral Sciences, 2nd edn.,
of a more severe form of ED. Despite significant reductions fol- London: Laurence Erlbaum Associates.
lowing treatment, there is some evidence that SA remains elevated Courty A et al. (2015) Alexithymia, a compounding factor for eating and
in those recovered from AN compared with HCs. Whether SA social avoidance symptoms in anorexia nervosa. Comprehensive Psychiatry
impacts on treatment adherence and outcomes has not yet been 56, 217–228.

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 14 Jess Kerr-Gaffney et al.

Dakanalis A et al. (2016) The social appearance anxiety scale in Italian ado- Kuusikko S et al. (2008) Social anxiety in high-functioning children and ado-
lescent populations: construct validation and group discrimination in com- lescents with autism and Asperger syndrome. Journal of Autism and
munity and clinical eating disorders samples. Child Psychiatry and Human Developmental Disorders 38, 1697–1709.
Development 47, 133–150. Lázaro L et al. (2011) Effectiveness of self-esteem and social skills group ther-
Davies H et al. (2016) Facial expression to emotional stimuli in non-psychotic apy in adolescent eating disorder patients attending a day hospital treat-
disorders: a systematic review and meta-analysis. Neuroscience and ment programme. European Eating Disorders Review 19, 398–406.
Biobehavioral Reviews 64, 252–271. Levinson CA and Rodebaugh TL (2016) Clarifying the prospective relation-
Dias P et al. (2011) Autonomic correlates of attachment insecurity in a sample ships between social anxiety and eating disorder symptoms and underlying
of women with eating disorders. Attachment & Human Development 13, vulnerabilities. Appetite 107, 38–46.
155–167. Liberati A et al. (2009) The PRISMA statement for reporting systematic
Duclos J et al. (2014) Expressed emotion in anorexia nervosa: what is inside reviews and meta-analyses of studies that evaluate health care interventions:
the “black box”? Comprehensive Psychiatry 55, 71–79. explanation and elaboration. PLoS Medicine 6, e1000100.
Fairburn CG et al. (1999) Risk factors for anorexia nervosa: three integrated Liebowitz MR (1987) Social phobia. Modern Problems of Pharmacopsychiatry
case-control comparisons. Archives of General Psychiatry 56, 468–476. 22, 141–173.
Fairweather-Schmidt AK and Wade TD (2014) DSM-5 eating disorders and Lloyd S et al. (2014) Perfectionism in anorexia nervosa: novel performance
other specified eating and feeding disorders: is there a meaningful differen- based evidence. PLoS ONE 9, e111697.
tiation? International Journal of Eating Disorders 47, 524–533. MacDonald DE, McFarlane TL and Olmsted MP (2014) “Diagnostic shift”
Flament MF et al. (2001) Predictive factors of social disability in patients with from eating disorder not otherwise specified to bulimia nervosa using
eating disorders. Eating and Weight Disorders 6, 99–106. DSM-5 criteria: a clinical comparison with DSM-IV bulimia. Eating
Gadalla T and Piran N (2008) Psychiatric comorbidity in women with disor- Behaviors 15, 60–62.
dered eating behavior: a national study. Women & Health 48, 467–484. Mattar L, Huas C and Godart N (2012a) Relationship between affective
Gander M, Seveck K and Buchheim A (2015) Eating disorders in adolescence: symptoms and malnutrition severity in severe anorexia nervosa. PLoS
attachment issues from a developmental perspective. Frontiers in Psychology ONE 7, e49380.
6, 1136. Mattar L et al. (2012b) Depression, anxiety and obsessive-compulsive symp-
Gilbert N and Meyer C (2003) Social anxiety and social comparison: differen- toms in relation to nutritional status and outcome in severe anorexia ner-
tial links with restrictive and bulimic attitudes among nonclinical women. vosa. Psychiatry Research 200, 513–517.
Eating Behaviors 4, 257–264. McFarlane T et al. (2015) The effectiveness of an individualized form of day
Gilboa-Schechtman E et al. (2006) Emotional processing in eating disorders: hospital treatment. Eating Disorders 23, 191–205.
specific impairment or general distress related deficiency? Depression and Melfsen S, Walitza S and Warnke A (2006) The extent of social anxiety in
Anxiety 23, 331–339. combination with mental disorders. European Child and Adolescent
Godart NT et al. (2000) Anxiety disorders in anorexia nervosa and bulimia Psychiatry 15, 111–117.
nervosa: co-morbidity and chronology of appearance. European Obeid N et al. (2013) Self-esteem and social anxiety in an adolescent female
Psychiatry 15, 38–45. eating disorder population: age and diagnostic effects. Eating Disorders 21,
Goddard E and Treasure J (2013) Anxiety and social-emotional processing in 140–153.
eating disorders: examination of family trios. Cognitive Therapy and Ohmann S et al. (2013) Emotional aspects of anorexia nervosa: results of pro-
Research 37, 890–904. spective naturalistic cognitive behavioral group therapy. Neuropsychiatrie
Grabhorn M et al. (2006) Social anxiety in anorexia and bulimia nervosa: 27, 119–128.
the mediating role of shame. Clinical Psychology and Psychotherapy 13, Ostrovsky NW et al. (2013) Social anxiety and disordered overeating: an
12–19. association among overweight and obese individuals. Eating Behaviors 14,
Gross J and Rosen JC (1988) Bulimia in adolescents: prevalence and psycho- 145–148.
social correlates. International Journal of Eating Disorders 7, 51–61. Pollatos O et al. (2008) Reduced perception of bodily signals in anorexia ner-
Harrison A, Mountford VA and Tchanturia K (2014) Social anhedonia and vosa. Eating Behaviours 9, 381–388.
work and social functioning in the acute and recovered phases of eating dis- Ranta K et al. (2017) Social phobia, depression and eating disorders during
orders. Psychiatry Research 218, 187–194. middle adolescence: longitudinal associations and treatment seeking.
Hinrichsen H, Sheffield A and Waller G (2007a) The role of parenting Nordic Journal of Psychiatry 71, 605–613.
experiences in the development of social anxiety and agoraphobia in the Rosenthal R (1979) The “file drawer problem” and tolerance for null results.
eating disorders. Eating Behaviours 8, 285–290. Psychological Bulletin 86, 638–641.
Hinrichsen H, Waller G and Dhokia R (2007b) Core beliefs and social anx- Core Team (2017) R: A language and environment for statistical computing. R
iety in the eating disorders. Eating and Weight Disorders 12, e14–e18. Foundation for Statistical Computing, Vienna, Austria. Available at https://
Hinrichsen H, Waller G and Emanuelli F (2004a) Social anxiety and agora- www.R-project.org/
phobia in the eating disorders: associations with core beliefs. The Journal of Russell J et al. (2018) Intranasal oxytocin in the treatment of anorexia nervosa:
Nervous and Mental Disease 192, 784–787. randomized controlled trial during re-feeding. Psychoneuroendocrinology
Hinrichsen H, Waller G and van Gerko K (2004b) Social anxiety and agora- 87, 83–92.
phobia in the eating disorders: associations with eating attitudes and beha- Sawaoka T et al. (2012) Social anxiety and self-consciousness in binge eating
viours. Eating Behaviours 5, 285–290. disorder: associations with eating disorder psychopathology. Comprehensive
Hinrichsen H et al. (2003) Social anxiety and coping strategies in the eating Psychiatry 53, 740–745.
disorders. Eating Behaviors 4, 117–126. Schmelkin C et al. (2017) Low oxytocin levels are related to alexithymia in
Jiménez-Murcia S et al. (2015) Differences and similarities between bulimia anorexia nervosa. International Journal of Eating Disorders 50, 1332–1338.
nervosa, compulsive buying and gambling disorder. European Eating Schneier FR et al. (2016) Attention bias in adults with anorexia nervosa,
Disorders Review 23, 126–132. obsessive-compulsive disorder, and social anxiety disorder. Journal of
Kaye WH et al. (2004) Comorbidity of anxiety disorders with anorexia and Psychiatric Research 79, 61–69.
bulimia nervosa. American Journal of Psychiatry 161, 2215–2221. Schulze UME et al. (2009) Trait anxiety in children and adolescents with
Kessler RC et al. (2005) Lifetime prevalence and age-of-onset distributions of anorexia nervosa. Eating and Weight Disorders 14, e163–e168.
DSM-IV disorders in the national comorbidity survey replication. Archives Schwalberg MD et al. (1992) Comparison of bulimics, obese binge eaters,
of General Psychiatry 62, 593–602. social phobics, and individuals with panic disorder on comorbidity across
Krug I et al. (2013) Low social interactions in eating disorder patients in child- DSM-III—R anxiety disorders. Journal of Abnormal Psychology 101, 675–681.
hood and adulthood: a multi-centre European case control study. Journal of Silgado J et al. (2010) Social anxiety and bulimic behaviors: the moderating
Health Psychology 18, 26–37. role of perfectionism. Cognitive Therapy and Research 34, 487–492.

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752
 Psychological Medicine 15

Slade P (1982) Towards a functional analysis of anorexia nervosa and bulimia cognitive, socio-emotional and interpersonal predisposing and perpetuating
nervosa. British Journal of Clinical Psychology 21, 167–179. factors. Journal of Eating Disorders 1, 13.
Solano R et al. (2005) Self-injurious behaviour in people with eating disorders. Utschig AC et al. (2010) An investigation of the relationship between fear of nega-
European Eating Disorders Review 10, 3–10. tive evaluation and bulimic psychopathology. Eating Behaviors 11, 231–238.
Steinglass JE et al. (2017) Temporal discounting across three psychiatric dis- Vall E and Wade TD (2015) Predictors of treatment outcome in individuals
orders: Anorexia nervosa, obsessive compulsive disorder, and social anxiety with eating disorders: a systematic review and meta-analysis. International
disorder. Depression and Anxiety 34, 463–470. Journal of Eating Disorders 48, 946–971.
Steinman SA et al. (2016) Prepulse inhibition deficits only in females with Viechtbauer W (2010) Conducting meta-analyses in R with the metafor pack-
obsessive-compulsive disorder. Depression and Anxiety 33, 238–246. age. Journal of Statistical Software 36, 3.
Striegel-Moore RH, Silberstein LR and Rodin J (1993) The social self in Westwood H et al. (2017a) Alexithymia in eating disorders: systematic review
bulimia nervosa: public self-consciousness, social anxiety, and perceived and meta-analyses of studies using the Toronto alexithymia scale. Journal of
fraudulence. Journal of Abnormal Psychology 102, 297–303. Psychosomatic Research 99, 66–81.
Strober M et al. (2007) The association of anxiety disorders and obsessive Westwood H et al. (2016) Exploration of friendship experiences, before and
compulsive personality disorder with anorexia nervosa: evidence from a after illness onset in females with anorexia nervosa: a qualitative study.
family study with discussion of nosological and neurodevelopmental impli- PLoS ONE 11, e0163528.
cations. International Journal of Eating Disorders 41, 174–179. Westwood H, Mandy W and Tchanturia K (2017b) Clinical evaluation of aut-
Swinbourne J et al. (2012) The comorbidity between eating disorders and istic symptoms in women with anorexia nervosa. Molecular Autism 8, 12.
anxiety disorders: prevalence in an eating disorder sample and anxiety dis- Wildes JE, Ringham RM and Marcus MD (2010) Emotion avoidance in
order sample. Australian & New Zealand Journal of Psychiatry 46, 118–131. patients with anorexia nervosa: initial test of a functional model.
Swinbourne J and Touyz S (2007) The co-morbidity of eating disorders and International Journal of Eating Disorders 43, 398–404.
anxiety disorders: a review. European Eating Disorders Review 15, 215–221. Zonnevylle-Bender MJS et al. (2005) Adolescent anorexia nervosa patients
Tchanturia K et al. (2015) Cognitive remediation and emotion skills training have a discrepancy between neurophysiological responses and self-reported
(CREST) for anorexia nervosa in individual format: self-reported outcomes. emotional arousal to psychosocial stress. Psychiatry Research 135, 45–52.
BMC Psychiatry 15, 53. Zonnevylle-Bender MJS et al. (2004) Emotional functioning in anorexia ner-
Treasure J and Schmidt U (2013) The cognitive-interpersonal maintenance vosa patients: adolescents compared to adults. Depression and Anxiety 19,
model of anorexia nervosa revisited: a summary of the evidence for 35–42.

Downloaded from https://www.cambridge.org/core. University of Groningen, on 12 Apr 2018 at 16:50:24, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0033291718000752