The classification of bulimic eating disorders: a community-based cluster analysis study.

There is controversy over how best to classify eating disorders in which there is recurrent binge eating. Many patients with recurrent binge eating do not meet diagnostic criteria for other of the two established eating disorders, anorexia nervosa or bulimia nervosa. The present study was designed to derive an empirically based, and clinically meaningful, diagnostic scheme by identifying subgroups from among those with recurrent binge eating, testing the validity of these subgroups and comparing their predictive validity with that of the DSM-IV scheme. A general population sample of 250 young women with recurrent binge eating was recruited using a two-stage design. Four subgroups among the sample were identified using a Ward's cluster analysis. The first subgroup had either objective or subjective bulimic episodes and vomiting or laxative misuse; the second had objective bulimic episodes and low levels of vomiting or laxative misuse; the third had subjective bulimic episodes and low levels of vomiting or laxative misuse; and the fourth was heterogeneous in character. This cluster solution was robust to replication. It had good descriptive and predictive validity and partial construct validity. The results support the concept of bulimia nervosa and its division into purging and non-purging subtypes. They also suggest a possible new binge eating syndrome. Binge eating disorder, listed as an example of Eating Disorder Not Otherwise Specified within DSM-IV, did not emerge from the cluster analysis.