Diagnostic efficiency of DSM-IV indicators for binge eating episodes

Research has examined various aspects of the validity of the research criteria for binge eating disorder (BED) but has yet to evaluate the utility of the 5 Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994) "indicators for impaired control" specified to help determine loss of control while overeating (i.e., binge eating). We examined the diagnostic efficiency of these indicators proposed as part of the research criteria for BED (eating until uncomfortably full; eating when not hungry; eating more rapidly than usual; eating in secret; and feeling disgust, shame, or depression after the episode).

A total of 916 community volunteers completed a battery of measures including questions about each of the indicators. Participants were categorized into 3 groups: BED (N = 164), bulimia nervosa (BN; N = 83), and non-binge-eating controls (N = 669). Four conditional probabilities (sensitivity, specificity, positive predictive power [PPP], and negative predictive power [NPP]) as well as total predictive value (TPV) and kappa coefficients were calculated for each indicator criterion in separate analyses comparing BED, BN, and combined BED + BN groups relative to controls.

PPPs and NPPs suggest all of the indicators have predictive value, with eating alone because embarrassed (PPP = .80) and feeling disgusted (NPP = .93) performing as the best inclusion and exclusion criteria, respectively. The best overall indicators for correctly identifying binge eating (based on TPV and kappa) were eating when not hungry and eating alone because embarrassed.

All 5 proposed indicators for impaired control for determining binge eating have utility, and the diagnostic efficiency statistics provide guidance for clinicians and the DSM-5 regarding their usefulness for inclusion or exclusion.