Andreas Goreis

Introduction Mentalization‐based therapy (MBT) and its adapted version for adolescents (MBT‐A) are repeatedly highlighted as promising treatments for reducing self‐harm, particularly in borderline personality disorder (BPD). Despite the availability of publications providing evidence of their efficacy in reducing self‐harm, recent meta‐analyses have yielded mixed results. To inform best‐practice clinical decision‐making, we conducted a systematic review and meta‐analysis. We aimed to disentangle findings for both adolescents and adults on the efficacy of MBT(‐A) in reducing self‐harm (primary outcome) and symptoms of BPD and depression (secondary outcomes). Methods Web of Science, Scopus, Embase, PubMed/Medline, and Cochrane Review Database were searched for eligible studies published until September 2022. In total, 14 studies were identified, comprising 612 participants from nine MBT studies (six pre–post, three RCTs) and five MBT‐A studies (two pre–post, three RCTs). Aggregated effect sizes were estimated using random‐effects models. Meta‐regressions were conducted to assess the effect of moderator variables (treatment duration, drop‐out rates, and age) on effect sizes. Results Overall, both MBT and MBT‐A demonstrated promising effects in reducing self‐harm ( g = −0.82, 95% CI −1.15 to −0.50), borderline personality disorder ( g = −1.08, 95% CI −1.38 to −0.77), and depression ( g = −1.1, 95% CI −1.52 to −0.68) symptoms. However, when compared to control interventions (TAU, SCM), MBT(‐A) did not prove to be more efficacious, with the exception of MBT showing superior effects on BPD symptoms in adults ( g = −0.56, 95% CI −0.88 to −0.24). Conclusion Although the pre–post evaluations seem promising, this analysis, including RCTs, showed no superiority of MBT(‐A) to control conditions, so that prioritizing the application of MBT (‐A) for the treatment of self‐harm is not supported. Possible explanations and further implications are discussed.