Treatment outcomes for anorexia nervosa: a systematic review and meta-analysis of randomized controlled trials

To determine the impact of specialized treatments, relative to comparator treatments, upon the weight and psychological symptoms of anorexia nervosa (AN) at end-of-treatment (EOT) and follow-up.

Randomized controlled trials (RCTs) between January 1980 and December 2017 that reported the effects of at least two treatments on AN were screened. Weight and psychological symptoms were analyzed separately for each study. PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were followed, and studies were assessed using the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) criteria and Cochrane risk of bias tool.

We identified 35 eligible RCTs, comprising data from 2524 patients. Meta-analyses revealed a significant treatment effect on weight outcomes at EOT [g = 0.16, 95% CI (0.05-0.28), p = 0.006], but not at follow-up [g = 0.11, 95% CI (-0.04 to 0.27), p = 0.15]. There was no significant treatment effect on psychological outcomes at either EOT [g = -0.03, 95% CI (-0.14 to 0.08), p = 0.63], or follow-up [g = -0.001, 95% CI (-0.11 to 0.11), p = 0.98]. There was no strong evidence of publication bias or significant moderator effects for illness duration, mean age, year of publication, comparator group category, or risk of bias (all p values > 0.05).

Current specialized treatments are more adept than comparator interventions at imparting change in weight-based AN symptoms at EOT, but not at follow-up. Specialized treatments confer no advantage over comparator interventions in terms of psychological symptoms. Future precision treatment efforts require a specific focus on the psychological symptoms of AN.