Personality as a predictor of symptomatic change in a residential treatment setting for anorexia nervosa and bulimia nervosa

PURPOSE: Although personality has been widely researched in patients with anorexia nervosa (AN) and bulimia nervosa (BN), the nature of this relationship has not yet been clearly articulated. The pathoplasty model theorizes that personality might shape symptomatic presentation and thus affect therapeutic outcomes, but more research is needed. The present study aimed at investigating the predictive value of a broad spectrum of personality traits in determining AN and BN treatment outcomes, considering both the statistical and clinical significance of the therapeutic change. METHODS: Eighty-four female patients with AN and BN treated in a residential program were evaluated at treatment onset using the Shedler-Westen Assessment Procedure-200—a clinician-rated measure of personality disorders and healthy personality functioning. At both intake and discharge, patients completed the Eating Disorder Inventory-3 to assess eating symptoms and the Outcome Questionnaire-45.2 to evaluate overall impairment. RESULTS: Considering overall ED symptomatic change, multiple regression analyses showed that, even when controlling for baseline symptoms and DSM-5 categories, schizoid (B = 0.41, p ≤ 0.01), avoidant (B = 0.31, p ≤ 0.05), and paranoid (B = 0.25, p ≤ 0.05) personality features predicted worse therapeutic outcomes. Similar results were found when applying the clinical significance approach, with the emotionally dysregulated factor as an additional negative predictor of significant/reliable change (B = − 0.09; p < 0.01). Healthy personality functioning predicted better therapeutic outcomes (B = − 0.34, p ≤ 0.001). CONCLUSIONS: Pathoplastic models and personality-based research in this clinical population have the potential to inform effective treatment strategies by targeting relevant individual factors. LEVEL OF EVIDENCE: Level III, longitudinal cohort study.