and Eating Disorders
CYNTHIA M. BULIK
The earliest clinical descriptions of anorexia nervosa and bulimia nervosa noted the frequent presence of both depression and anxiety. Subsequent investigations using structured diagnostic methodology in clinical and epidemiological samples and in family studies have verified those early observations. Of the many possible models of comorbidity, five are particularly plausible to explain the relation among eating, affective, and anxiety disorders. Each of these models yields a unique set of predictions.
Model One posits that depression and anxiety are sequelae of eating disorders. It predicts that depression and anxiety would not be apparent in individuals with anorexia nervosa and bulimia nervosa prior to the development of an eating disorder. Moreover, unless one posits a permanent scarring effect, this model predicts the alleviation of anxiety or depressive symptoms with recovery from the eating disorder.
Model Two, the converse of Model One, posits that eating disorders are sequelae of affective or anxiety disorders. It predicts a pattern of onset in which depression or anxiety manifest prior to the onset of eating disorders.
Model Three, the forme fruste model, posits that eating disorders are expressions of an underlying depressive or anxiety disorder. Eating disorders may be age- and genderspecific manifestations of depression or anxiety, with etiological factors completely shared among the classes of disorders.
Model Four posits that eating, anxiety, and affective disorders are different expressions of the same underlying causal factor (e.g., neuroticism or neuroendocrine disturbance).
Model Five posits that whereas eating, affective, and anxiety disorders are unique sets of conditions, they may share some etiological factors. It makes no predictions about whether the eating, depressive, or anxiety disorders manifest first but predicts the existence of both shared and independent etiological factors across the disorders.
Given these five models of comorbidity, how can extant data assist with illuminating the nature of the relationship among these three disorders?