Psychopathology of EDNOS patients: To whom do they compare?

Do the levels of psychopathology displayed by patients with an eating disorder not otherwise specified (EDNOS) more closely resemble those displayed by full-criteria anorexia and bulimia nervosa patients than they do those of non-eatingdisorder controls? Three groups of eating disorder patients (anorexia nervosa, n=27; bulimia nervosa, n=23; EDNOS, n=19) and a group of non-eating-disorder controls (n=25) were compared on four subscales of the Eating Disorders Inventory – 2 (EDI-2), and on the Eating Disorders Belief Questionnaire, the Rosenberg Self-Esteem Scale, and the Depression Anxiety Stress Scales. EDNOS patients were indistinguishable from the other two eating disorder groups on all of the measured variables, except for the Perfectionism subscale of the EDI-2. The findings of the present study support the view that EDNOS should not be considered a subclinical eating disorder. However, much remains to be understood about who EDNOS patients are. Until recently, research and treatment in eating disorders has focused on the two most widely recognised and discretely categorised disorders, namely, anorexia nervosa (AN) and bulimia nervosa (BN). However, clinical research indicates that a substantial number of those individuals presenting for treatment of eating disorders do not fulfil criteria for either AN or BN; it has been found that between 33% and 61% of patients presenting for treatment of an eating disorder do not meet full criteria for either disorder according to the Diagnostic and Statistical Manual of Mental Disorders (3rd ed., revised; DSMIII-R; American Psychiatric Association [APA], 1987) (Beumont, Kopec-Schrader, Talbot, & Touyz, 1993; Bunnell, Shenker, Nussbaum, Jacobsen, & Cooper, 1990; Williamson, Gleaves, & Savin, 1992). Despite revised criteria for AN and BN, similar findings have been reported in studies using the DSM-IV (APA, 1994) diagnostic criteria (e.g., Hay & Fairburn, 1998; Mizes & Sloan, 1998; van der Ham, Meulman, van Strien, & van Engeland, 1997). In fact, the prevalence of patients with an eating disorder who do not meet either set of diagnostic criteria but become distressed by their symptoms and ultimately seek treatment approximates twice that of the recognised AN and BN categories. The eating disorders not otherwise specified (EDNOS) category was introduced to account ‘‘for disorders of eating that do not meet the criteria for any specified Eating Disorder’’ (APA, 1994, p. 550). Like all other not otherwise specified categories in the DSM-IV, EDNOS was originally intended to classify patients with an atypical presentation of the disorders that was clinically severe enough to warrant formal diagnosis. This atypical label, however, is somewhat misleading because it denotes the EDNOS group as comprising a small percentage of eating disorder patients; in reality, as noted above, a diagnosis of EDNOS is even more common than the other diagnostic categories. Comparing EDNOS to full-syndrome eating disorders Because EDNOS patients are symptomatically quite similar to full-syndrome eating disorder patients, and because of the fact that many of the particular diagnostic criteria for eating disorders are somewhat subjective (and contentious) (Thaw, Williamson, & Martin, 2001; Walsh & Garner, 1997), researchers have been interested in whether EDNOS patients differ from their full-criteria counterparts in terms of various psychopathological features. For example, *Deceased. Correspondence: S. W. Touyz, School of Psychology, University of Sydney, Sydney, NSW 2006, Australia. Tel.: + 61 2 9351 2646. Fax: + 61 2 9351 2984. E-mail: [email protected] or [email protected] Clinical Psychologist, Vol. 8, No. 2, November 2004, pp. 70 – 75. ISSN 1328-4207 print/ISSN 1749-9552 online # The Australian Psychological Society Ltd Published by Taylor & Francis Ltd DOI: 10.1080/1328420041233130436 Bunnell et al. (1990) compared patients who met full criteria for AN with others who were ‘‘subclinical’’, that is, who did not fulfil all criteria (usually the severity of weight loss criteria), and patients who met full criteria for BN with ‘‘subclinical’’ BN cases (based on the frequency of bingeing and compensatory behaviours). There were no differences found between the two AN groups on a number of demographic, clinical, and psychological variables that were examined. The subclinical BN group was also quite similar to full-criteria BN patients, but did differ with respect to affect regulation and impulse control. Following the Bunnell et al. (1990) lead, Ricca et al. (2001) compared AN and BN patients to EDNOS patients who met all but one criteria for AN (amenorrhoea or severity of weight loss; EDNOS-A) and for BN (frequency of bingeing or compensatory behaviour; EDNOS-B) on a number of psychopathological features. There were no significant differences between the AN and EDNOS-A groups, or between the BN and EDNOS-B groups. These findings suggest that EDNOS patients display levels of psychopathology that are comparable to their full-criteria counterparts. There were, however, some reliable differences between anorexictype patients (AN and EDNOS-A) and bulimic-type patients (BN and EDNOS-B), specifically in terms of levels of depression (bulimic types were more depressed than anorexic types), and dietary restraint and eating concerns (anorexic types scored higher on both factors). The authors suggest that it would be useful to consider subdividing the EDNOS category in terms of anorexic type and bulimic type patients. Rather than dividing EDNOS patients into groups of anorexic-type and bulimic-type patients, Williamson et al. (1992) used cluster-analytic techniques to determine whether any homogeneous subgroups of EDNOS patients would emerge. They identified three distinct subgroups that were similar in symptom profiles to: (a) ‘‘subthreshold’’ AN (i.e., with relatively mild degree of emaciation), (b) nonpurging BN (now recognised in the DSM-IV), and (c) binge eating disorder (a diagnosis proposed for further study in DSM-IV). The three EDNOS subgroups were then compared with groups of AN and BN patients on psychopathological variables specific to eating disorders. The cognitive disturbances (e.g., body image and drive for thinness) of the subthreshold AN group was comparable to the disturbance found among full-syndrome AN and BN patients. The atypical, nonpurging BN group differed only from the true BN group in terms of the extent of purgative methods used to control their weight. The third atypical group, individuals with ‘‘binge eating disorder’’, binge ate but did not engage in selfinduced vomiting or extreme dieting to control their weight, and these individuals were in fact quite overweight. This group also expressed equivalent levels of body dissatisfaction compared to the AN and BN groups, but had lower drive for thinness. Overall, most cases diagnosed as EDNOS were more similar than dissimilar to full-criteria cases. Continuity versus discontinuity in eating disorders It is important to note that discussions of subthreshold eating disorders are part of a debate that has been lively for years regarding the nature of eating disorders; that is, whether eating disorders form a discrete entity, separate from normal eating patterns, or whether eating pathology exists on a continuum from nonpathological eating to eating disorders, with dieting and subthreshold eating disorders falling somewhere in between. There is some evidence in support of each view (Garfinkel et al., 1995; Lowe et al., 1996; Mintz & Betz, 1988; Ruderman & Besbeas, 1992; Stice, Killen, Hayward, & Taylor, 1998). A recent taxometric analysis by Williamson et al. (2002; see also Gleaves, Lowe, Green, Cororve, & Williams, 2000), however, suggests that it may be more appropriate to consider eating disorders as a distinct entity, qualitatively different from nonpathological forms of eating behaviour.