The significance of self-induced vomiting as a weight control method among nonclinical samples

Abstract

The aim of the current study is to determine the viability of employing self-induced vomiting as an index of psychopathology in nonclinical samples. Eating Disorder Inventory (EDI) profiles were obtained for 77 college and ballet students who reported engaging in self-induced vomiting as a method of weight control; these profiles were compared with those of 366 women who reported never having engaged in self-induced vomiting. Cluster analyses of individuals were performed to examine the structure of EDI subscale scores for both groups of women. Three clusters of vomiting women were derived. The first consisted of approximately 10% of the vomiting sample, and they had EDI subscale scores that were at or above the mean for bulimia nervosa patients. The second cluster contained women who had elevated EDI subscales scores related to attitudes about eating and body shape but were in the normal range on other scales of psychological functioning. The third cluster was composed of 47% of the vomiting women, and they had normal to high-normal mean EDI scores. The mean frequency of vomiting did not distinguish the three clusters. Similar clusters were obtained for the nonvomiting sample of women. It is concluded that relying upon self-induced vomiting or simple measures of weight preoccupation without considering other dimensions of psychological functioning is of limited value in identifying the presence of a clinically significant eating disorder.