Objective: To determine if clinically defined multi-impulsive bulimia (MIB) can be validated using an empirical classification approach.
Method: One hundred and twenty-five women who met DSM-IV criteria for bulimia nervosa were interviewed and completed a two-week EMA protocol. Participants who reported at least three highly impulsive behaviors were compared to participants who did not report such impulsive behavior. Also, Latent Class Analysis (LCA) was conducted to determine if the MIB classification could be replicated empirically.
Results: LCA produced a two-class solution consistent with the traditional clinical approach to MIB classification. In both approaches, MIB was associated with higher levels of anxiety disorders, child abuse, and daily self-damaging behaviors than the non-MIB class.
Conclusion: Clinical classification of MIB was empirically supported through LCA. Although the classes failed to differ in bulimic behavior, MIB appears to represent a group of bulimic individuals with significant trauma histories and associated psychopathology.
Copyright 2006 by Wiley Periodicals, Inc.