Should bulimia nervosa be subtyped by history of anorexia nervosa? A longitudinal validation.

OBJECTIVE: To determine whether a past diagnosis of anorexia nervosa (AN) predicts longitudinal course and outcome among women with bulimia nervosa (BN).

METHOD: A subset (n = 176) of participants in the Longitudinal Study of Anorexia and Bulimia Nervosa who met DSM-IV criteria for BN either at study intake (n = 144) or during follow-up (n = 32; 4 had restricting AN at intake, 28 had binge/purge AN at intake) were included in this report. Over a median of 9 years, weekly eating disorder symptom data were collected from participants using the Longitudinal Interview Follow-up Examination, Eating Disorders Version.

RESULTS: While there were no between-group differences in likelihood of partial recovery, women with BN who had a history of AN were more likely to have a protracted illness, relapsing into AN during follow-up, compared to those with no AN history who were more likely to move from partial to full recovery.

CONCLUSION: Lifetime AN is an important prognostic indicator among women with BN and these longitudinal data would support the subtyping of BN on the basis of AN history.