Objective: The goal of this study was to investigate the relationship of body composition and neuroendocrine levels with clinical outcome in women with anorexia nervosa in a relapse-prevention trial.
Method: Body composition and fasting cortisol and leptin levels were assessed before random assignment in 32 weight-recovered subjects with anorexia nervosa from the New York site of the Fluoxetine to Prevent Relapse in Women With Anorexia Nervosa trial. Clinical outcome at the end of study participation was defined using modified Morgan-Russell criteria (full, good, fair, poor), then dichotomized into treatment "success" or "failure."
Results: In a binary logistic regression model examining the effect of percent body fat, body mass index, anorexia nervosa subtype, waist-to-hip ratio, and serum cortisol and leptin levels on treatment outcome, only percent body fat was significantly associated with outcome.
Conclusions: In recently weight-restored women with anorexia nervosa, lower percent body fat was associated with poor long-term outcome.