Abstract
Background: Second-generation antipsychotic drugs (SGAs) are increasingly administered to achieve weight gain in anorexia nervosa. In this meta-analysis, we aimed to determine if any evidence for this treatment option can be derived from randomized controlled trials (RCTs). Methods: Based on the ‘World Federation of Societies of Biological Psychiatry (WFSBP) Guidelines for the Pharmacological Treatment of Eating Disorders', a systematic update literature search was applied to identify all RCTs investigating the efficacy, acceptability, and tolerability of SGAs in anorexia nervosa in comparison to placebo/no treatment. The primary outcome was weight gain measured by mean change in body mass index (BMI). Secondary outcomes were mean changes in Yale-Brown-Cornell Eating Disorders Scale (YBC-EDS) total score and Eating Disorders Inventory (EDI) total score and premature discontinuation of treatment. Employing a random-effects model standardized mean differences based on Hedges's g and Mantel-Haenszel risk ratios were calculated. Results: Seven RCTs (n = 201) investigating olanzapine (N = 4), quetiapine (N = 2), and risperidone (N = 1) were included. We found no statistically significant between-group differences for mean BMI change when pooling the SGAs (N = 7, n = 161; Hedges's g = 0.13, 95% CI: -0.17 to 0.43; p = 0.4) and when examining the individual drugs. Furthermore, the SGAs failed to differentiate statistically significantly from placebo/no treatment for all secondary outcomes. Conclusions: Based on the current evidence, pharmacological treatment of anorexia nervosa with SGAs cannot be generally recommended although some individuals or subgroups of patients might benefit from an antipsychotic medication. Further research is required to identify which patients will likely benefit from such a treatment option.