Eating Behaviors 15 (2014) 42–44
Contents lists available at ScienceDirect
Eating disorder beliefs and behaviours across eating disorder diagnoses Steven Allan a,⁎, Ken Goss b,1 a b
University of Leicester, Clinical Psychology Unit, 104 Regent Road, Leicester LE1 7LT, UK Coventry & Warwickshire Partnership Trust, Coventry Eating Disorder Service, 2 Dover Street, Coventry CV1 3DB, UK
a r t i c l e
i n f o
Article history: Received 15 January 2013 Received in revised form 9 August 2013 Accepted 9 October 2013 Available online 21 October 2013 Keywords: Dietary beliefs Dietary behaviours Stirling Eating Disorder Scales Transdiagnostic
a b s t r a c t Objective: To test for differences between diagnostic groups on the severity of eating disorder beliefs and behaviours, evaluate the clinical signiﬁcance of such differences, and assess the extent to which these beliefs and behaviours may be present at clinically signiﬁcant levels across eating disorder diagnoses. Method: 136 adult women outpatients (aged 18–65, with a BMI over 15) were diagnosed with an eating disorder and completed the Stirling Eating Disorder Scale. Results: The expected pattern of statistically signiﬁcant differences was found between diagnostic groups on anorexic dietary beliefs and behaviours and bulimic dietary beliefs and behaviours. A high percentage of participants in each diagnostic group scored above the clinical cut-off on the eating disorder belief and behaviour measures and a very high percentage of participants in each group reported clinically signiﬁcant levels of restricting beliefs. Conclusions: Transdiagnostic or functional analytic approaches to treatment planning may lead to more effective interventions than current, diagnostically-based, care pathways. The high prevalence of restricting beliefs reported suggested that this may need to be a key focus for intervention for the majority of individuals presenting with an eating disorder. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Current classiﬁcatory systems for eating disorders (DSM IV; APA, 1994 and ICD-10; WHO, 2010) are argued to be problematic (Franko, Wonderlich, Little, & Herzog, 2004). In addition speciﬁc eating disorder diagnosis appears to be relatively ﬂuid over time (Braun, Sunday, & Halmi, 1994; Bulik, Sullivan, Fear, & Pickering, 1997; Milos, Spindler, Schnyder, & Fairburn, 2005). However, the overarching category of “eating disorder” does appear to remain relatively stable over time, regardless of the original, more speciﬁc, diagnosis received by a patient (Herzog, Hopkins, & Burns, 1993). Despite this the National Institute for Health and Clinical Excellence (NICE, 2004) and the American Psychiatric Association (APA, 2006) have both recommended speciﬁc care pathways and interventions based on diagnosis for Anorexia Nervosa (AN) and Bulimia Nervosa (BN). It is estimated that between 20% and 60% of those seeking treatment for an eating disorder ﬁt into the diagnostic category of Eating Disorder Not Otherwise Speciﬁed (EDNOS) (e.g., Turner & Bryant-Waugh, 2004). The levels of psychosocial distress and the impact on psychosocial functioning associated with EDNOS appear to be as severe as that found in patients with AN or BN (Herzog & Delinsky, 2001). These
⁎ Corresponding author. Tel.: +44 116 2231648. E-mail addresses: [email protected]
(S. Allan), [email protected]
(K. Goss). 1 Tel.: +44 2476 232940. 1471-0153/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eatbeh.2013.10.002
patients pose a particular challenge for treatment providers if their treatment models are based on diagnosis-speciﬁc treatment. Recent developments in CBT (Fairburn, Cooper, & Shafran, 2003) have revisited earlier conceptualizations of patients with eating disorders which saw them as having speciﬁc beliefs and behaviours in common. However, relatively little is known about the extent to which speciﬁc eating disorder beliefs are common across eating disorder diagnostic groups and the extent to which diagnostic groups share similar patterns of eating disorder behaviours. The Stirling Eating Disorder Scale (SEDS; Williams & Power, 1995) was developed to measure eating disorder beliefs and behaviours and other problems commonly associated with eating disorders, to allow comparison between eating disorder groups and controls and to monitor treatment progress. Studies using the SEDS (Campbell, Lawrence, Serpell, Lask, & Neiderman, 2002; Gamble et al., 2006; Openshaw & Waller, 2005; Williams & Power, 1995) suggested some statistically signiﬁcant differences between eating disorder diagnostic groups in terms of beliefs and behaviours. The extent to which these differences were clinically meaningful was not clear. It was also unclear what proportion of patients with a particular eating disorder diagnosis also experienced cognitive and behavioural problems associated with another eating disorder diagnosis (e.g., patients with AN who also reported bulimic cognitions and behaviours). Relatively little is known about the extent of AN and BN beliefs and behaviours in patients with EDNOS. The ﬁrst aim of the current study was to test for statistically signiﬁcant differences between diagnostic groups (AN, BN and EDNOS excluding Binge Eating Disorder; BED) on eating disorder beliefs and
S. Allan, K. Goss / Eating Behaviors 15 (2014) 42–44
behaviours. The second aim was to assess the clinical signiﬁcance of any statistical differences and the extent to which speciﬁc eating disorder beliefs and behaviours were present at clinically signiﬁcant levels across different eating disorder diagnoses.
2. Method 2.1. Participants and procedures Participants were 136 women patients aged 18–65 (mean age 27.2; SD 8.9) referred for treatment at a U.K. out-patient eating disorder service. All participants had a Body Mass Index of 15 and above and were offered treatment by the service. All participants were assessed by clinicians trained and experienced in the assessment, diagnosis and treatment of patients with eating disorders. Each participant was diagnosed using DSM IV criteria for an eating disorder (AN, BN, and EDNOS). Of the 136 female participants, 33 were diagnosed with AN, 50 were diagnosed with BN and 53 were diagnosed with EDNOS.
2.2. Measures The Stirling Eating Disorders Scale (SEDS) was chosen as it allows comparison between different diagnostic groups on eating disorder beliefs and behaviours. It provides cut-off scores for clinically signiﬁcant eating disorder symptoms and for speciﬁc AN and BN diagnoses. This study used the SEDS subscales measuring core eating disorder behaviours and beliefs. These subscales are labelled Anorexic Dietary Cognitions (ADC), Anorexic Dietary Behaviours (ADB), Bulimic Dietary Cognitions (BDC) and Bulimic Dietary Behaviours (BDB). The clinical cut-off scores for these are greater than 9, 14, 17 and 14 respectively.
3. Results The means, standard deviation, and Cronbach's alpha for each of the ADC (mean = 30.1, SD = 8.6, Alpha = 0.57), ADB (16.0, 9.9, 0.69), BDC (30.9, 10.8, 0.75) and BDB (23.8, 12.0, 0.72) subscales of the SEDS were in line with those reported in previous research (Gamble et al., 2006; Openshaw & Waller, 2005; Williams & Power, 1995). The internal reliability of the ADC subscale was somewhat lower than expected but in line with that reported for smaller sized samples (Gamble et al., 2006).
3.1. Comparison of scale scores across diagnoses The distribution of scores by diagnosis did not meet the parametric assumptions of homogeneity of variance or normality (either before or after attempts at data transformation). Therefore medians and interquartile ranges (IQRs) were calculated for each of the three diagnostic groups and the difference between these groups was tested for signiﬁcance using non-parametric Kruskal–Wallis tests. The medians, IQRs, and results of the Kruskal–Wallis tests are presented in Table 1. There were signiﬁcant differences across diagnoses for each of the scales. Post hoc analyses revealed that, as expected, the AN group had signiﬁcantly higher scores on the ADC subscales compared to either the BN or EDNOS groups. The AN group also had higher scores on the ADB subscale compared to the BN group. The BN and EDNOS group scores were not signiﬁcantly different on either of these measures. Also, as expected, the BN group had signiﬁcantly higher scores on the BDC and BDB subscales compared to either the AN or EDNOS groups. There was no signiﬁcant difference between the AN and EDNOS groups on the BDC and BDB subscales.
Table 1 Medians and interquartile ranges of dietary cognitions and behaviours by eating disorder diagnosis with signiﬁcant differences between diagnostic groups.
ADC ADB BDC BDB
AN (n = 33)
BN (n = 50)
EDNOS (n = 53)
34.8a 20.3a 34.4a 21.2a
(9.4) (15.4) (20.0) (21.1)
29.0b 12.0b 40.0b 33.1b
(10.7) (12.0) (6.3) (8.0)
31.8b 15.9ab 29.0a 17.3a
(11.2) (15.6) (19.9) (14.2)
Chi2 (df = 2)
10.2⁎ 9.9⁎ 28.9⁎⁎ 42.0⁎⁎
Notes: AN = anorexia nervosa group; BN = bulimia nervosa group; EDNOS = eating disorder not otherwise speciﬁed group; ADC = anorexic dietary cognitions; ADB = anorexic dietary behaviours; BDC = bulimic dietary cognitions; BDB = bulimic dietary behaviours. Within each row, medians with different subscripts differ at the 0.01 level of signiﬁcance according to a Mann–Whitney U test. ⁎ p b 0.01. ⁎⁎ p b 0.001.
3.2. Percentage of participants above clinical cut-off scores by diagnosis The differences between the diagnostic groups are reported in Table 1. They are in line with previous research (Gamble et al., 2006; Williams et al., 1993). However, these differences obscure the fact that the mean scores for each group on each of the scales were quite high. The percentage of participants with scores above the cut-off scores for each diagnostic group, and for all participants, is presented in Table 2. Over 95% of all participants had scores above clinical cut-off on the ADC subscale and over 82% for the BDC subscale. Nearly 53% scored above clinical cut-off on the ADB subscale, and 75% scored above clinical cut-off on the BDB subscale. Anorexic dietary cognitions appeared to be a common theme across all diagnoses, with 97% diagnosed with AN, 92.5% of those diagnosed with EDNOS, and 98% of those diagnosed with BN scoring above clinical cut-off on this subscale. Anorexic dietary behaviours are a particular problem for the majority of those in the AN (72.7%) and EDNOS (52.8%) groups who scored above the cutoff. However, these behaviours were also a problem for 40% of those diagnosed with BN. Bulimic dietary cognitions also appeared to be a common problem regardless of diagnosis, with 96% of those diagnosed with BN, 69.8% of those diagnosed with EDNOS and 81.8% of those diagnosed with AN scoring above clinical cut-off. Again bulimic dietary behaviours were also very common across diagnoses, with 98% of patients diagnosed with BN, 60.4% diagnosed with EDNOS and 63.6% of those diagnosed with AN scoring above clinical cut-off. The percentages presented in Table 2 suggested that there might be some differences between the AN group and the BN group with respect to dietary behaviours but very little difference with respect to dietary cognitions. A series of 2 × 2 Chi square tests was conducted to compare the proportions of the 83 participants diagnosed as either AN or BN who scored above or below the clinical cut-offs on the four dietary cognition and behaviour subscales of the SEDS. There was no signiﬁcant difference between those with AN or BN in the proportions reporting above and below clinical levels of anorexic dietary cognitions (Chi2 = 0.01, df = 1, ns) or reporting above or below clinical levels of bulimic dietary cognitions (Chi2 = 3.1, df = 1, ns). However, there were Table 2 Percentage of participants above the cut-off scores on dietary cognitions and behaviours by eating disorder diagnosis.
Anorexic dietary cognitions Anorexic dietary behaviours Bulimic dietary cognitions Bulimic dietary behaviours
ALL (n = 136)
AN (n = 33)
BN (n = 50)
EDNOS (n = 53)
95.6 52.9 82.4 75.0
97.0 72.7 81.8 63.6
98.0 40.0 96.0 98.0
92.5 52.8 69.8 60.4
Notes: AN = anorexia nervosa; BN = bulimia nervosa; EDNOS = eating disorder not otherwise speciﬁed.
S. Allan, K. Goss / Eating Behaviors 15 (2014) 42–44
signiﬁcant differences between the two diagnostic groups in the proportions reporting above and below clinical levels of anorexic dietary behaviour (Chi2 = 7.2, df = 1, p b0.01) and in the proportions reporting above and below clinical levels of bulimic dietary behaviour (Chi2 = 15.9, df = 1, p b 0.001).
centred approach to treatment (Fairburn et al., 2003; Kaplan, Kerr, & Maddocks, 1992).
Contributors Both authors contributed to the research (literature search, design, and analyses) and wrote the manuscript.
This study aimed to test for statistically signiﬁcant differences between diagnostic groups on eating disorder beliefs and behaviours, to evaluate the clinical signiﬁcance of such differences, and to assess the extent to which these beliefs and behaviours may be present at clinically signiﬁcant levels across eating disorder diagnoses. Signiﬁcant differences were found between diagnostic groups on anorexic dietary behaviours and cognitions and bulimic dietary behaviours and cognitions. Post hoc analysis revealed expected differences between patients diagnosed with AN and those diagnosed with BN or EDNOS in the severity score for anorexic dietary cognitions. Although differences in severity of symptoms between groups were statistically signiﬁcant they appear to be less clinically signiﬁcant given the high percentages in each diagnostic group who scored above the clinical cut-off scores on each of the eating disorder subscales. The majority of participants had clinically signiﬁcant anorexic and bulimic cognitions. Restricting behaviours were prevalent for many participants with EDNOS and a signiﬁcant minority of those with BN. This suggests that using diagnosis alone to identify food restriction as a target of intervention is likely to lead to a signiﬁcant proportion of patients not receiving targeted treatment for this type of behaviour. Anorexic dietary behaviours (such as eating low calorie foods, counting calories, hiding food eat it, ritualized eating or avoiding eating with others) was not present at a clinically signiﬁcant level for a large minority (27%) of participants diagnosed with AN. This suggests that clinicians may need to explore a wider range of weight control activities used by patients who meet diagnostic criteria for AN, such as compulsive exercise/activity, or drug misuse. Bulimic dietary behaviours were present at clinically signiﬁcant levels for 75% of the sample and were reported the majority of those with AN and EDNOS. This result is similar to that found by Eddy et al. (2002) and suggests that these behaviours may need to be addressed for most patients. Several theorists have argued that current diagnostic models may be inadequate and this study supports these critiques and indicates that those with eating disorders (at least in western cultures) have more in common than current diagnostic classiﬁcatory systems suggest (Eddy et al., 2002; Fairburn et al., 2003; Franko et al., 2004; Waller, 1993). The extremely high prevalence of restricting beliefs across all diagnostic categories suggests that these beliefs may need to be a core focus for intervention for all patients with eating disorders; particularly as changing eating disorder beliefs is seen as key to reducing relapse in most intervention programmes. The presence of high levels of comorbid restricting and bulimic behaviours suggests that the provision of treatments designed to target these problems is necessary for patients presenting for community based treatment. The results bring into question current care pathways (e.g., APA, 2006; NICE, 2004) that are diagnostically driven. They support a transdiagnostic or a functional analytic approach to treatment planning which targets speciﬁc beliefs and behaviours (Fairburn et al., 2003; Garner & Garﬁnkel, 1982; Goss & Gilbert, 2002; Waller, 1993). This may provide a more useful model when developing care pathways with the differentiation of patient groups based on symptom presentation, rather than diagnostic categories, providing a more efﬁcient and client
Role of funding sources There was no external funding source for this study.
Conﬂict of interest Both authors of this manuscript declare that they have no conﬂict of interest.
References American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.)Washington, DC: American Psychiatric Association. American Psychiatric Association (2006). Practice guideline for the treatment of patients with eating disorders (3rd ed.)Washington, DC: American Psychiatric Association. Braun, D. L., Sunday, S. R., & Halmi, K. A. (1994). Psychiatric comorbidity in patients with eating disorders. Psychological Medicine, 24(4), 859–867. Bulik, C. M., Sullivan, P. F., Fear, J., & Pickering, A. (1997). Predictors of the development of bulimia nervosa in women with anorexia nervosa. Journal of Nervous and Mental Disease, 185(11), 704–707. Campbell, M., Lawrence, B., Serpell, L., Lask, B., & Neiderman, M. (2002). Validating the Stirling Eating Disorders Scales (SEDS) in an adolescent population. Eating Behaviors, 3(3), 285–293. Eddy, K. T., Keel, P. K., Dorer, D. J., Delinsky, S. S., Franko, D. L., & Herzog, D. B. (2002). Longitudinal comparison of anorexia nervosa subtypes. International Journal of Eating Disorders, 31(2), 191–201. Fairburn, C. G., Cooper, Z., & Shafran, R. (2003). Cognitive behaviour therapy for eating disorders: A “transdiagnostic” theory and treatment. Behaviour Research and Therapy, 41(5), 509–528. Franko, D. L., Wonderlich, S. A., Little, D., & Herzog, D. B. (2004). Diagnosis and classiﬁcation of eating disorders. In J. K. Thompson (Ed.), Handbook of eating disorders and obesity (pp. 58–80). New Jersey: John Wiley & Sons. Gamble, C., Bryant-Waugh, R., Turner, H., Jones, C., Mehta, R., & Graves, A. (2006). An investigation into the psychometric properties of the Stirling Eating Disorder Scales. Eating Behaviors, 7(4), 395–403. Garner, D.M., & Garﬁnkel, P. E. (1982). Anorexia nervosa: A multidimensional perspective. New York: Brunner/Mazel. Goss, K., & Gilbert, P. (2002). Eating disorders, shame and pride: a Cognitive-behavioural functional analysis. In P. Gilbert, & J. Miles (Eds.), Body shame: Conceptualisation, research and treatment (pp. 219–255). Hove, UK: Brunner-Routledge. Herzog, D. B., & Delinsky, S. S. (2001). Classiﬁcation of eating disorders. In R. H. Striegel-Moore, & L. Smolak (Eds.), Eating disorders: Innovative directions in research and practice (pp. 13–50). Washington, DC: American Psychological Association. Herzog, D. B., Hopkins, J.D., & Burns, C. D. (1993). A follow-up study of 33 subdiagnostic eating disordered women. International Journal of Eating Disorders, 14(3), 261–267. Kaplan, A. S., Kerr, A., & Maddocks, S. E. (1992). Day hospital group treatment. In H. Harper-Giuffre, & K. R. MacKenzie (Eds.), Group psychotherapy for eating disorders (pp. 161–180). Washington, DC: American Psychiatric Press Inc. Milos, G., Spindler, A., Schnyder, U., & Fairburn, C. G. (2005). Instability of eating disorder diagnoses: Prospective study. British Journal of Psychiatry, 187, 573–578. National Collaborating Centre for Mental Health (2004). Eating disorders: Core interventions in the treatment and management of anorexia nervosa, bulimia nervosa, and related eating disorders. Leicester, UK: The British Psychological Society and Gaskell. Openshaw, C., & Waller, G. (2005). Psychometric properties of the Stirling Eating Disorder Scales with bulimia nervosa patients. Eating Behaviors, 6(2), 165–168. Turner, H., & Bryant-Waugh, R. (2004). Eating disorder not otherwise speciﬁed (EDNOS): Proﬁles of clients presenting at a community eating disorder service. European Eating Disorders Review, 12(1), 18–26. Waller, G. (1993). Why do we diagnose different types of eating disorder? Arguments for a change in research and clinical practice. European Eating Disorders Review, 1(2), 74–89. Williams, G. J., & Power, K. G. (1995). Stirling Eating Disorder Scales. London: The Psychological Corporation. Williams, G. J., Power, K. G., Millar, H. R., Freeman, C. P., Yellowlees, A., Dowds, T., et al. (1993). Comparison of eating disorders and other dietary/weight groups on measures of perceived control, assertiveness, self-esteem, and self-directed hostility. International Journal of Eating Disorders, 14(1), 27–32. World Health Organization (2010). International statistical classiﬁcation of diseases and related health problems (10th ed.)Geneva: World Health Organization.