The degree of efficacy, safety, quality, and certainty of meta-analytic evidence of biological non-pharmacological treatments in mental disorders is unclear.

We conducted an umbrella review (PubMed/Cochrane Library/PsycINFO-04-Jul-2021, PROSPERO/CRD42020158827) for meta-analyses of randomized controlled trials (RCTs) on deep brain stimulation (DBS), transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS), electro-convulsive therapy (ECT), and others. Co-primary outcomes were standardized mean differences (SMD) of disease-specific symptoms, and acceptability (for all-cause discontinuation). Evidence was assessed with AMSTAR/AMSTAR-Content/GRADE.

We selected 102 meta-analyses. Effective interventions compared to sham were in depressive disorders: ECT (SMD=0.91/GRADE=moderate), TMS (SMD=0.51/GRADE=moderate), tDCS (SMD=0.46/GRADE=low), DBS (SMD=0.42/GRADE=very low), light therapy (SMD=0.41/GRADE=low); schizophrenia: ECT (SMD=0.88/GRADE=moderate), tDCS (SMD=0.45/GRADE=very low), TMS (prefrontal theta-burst, SMD=0.58/GRADE=low; left-temporoparietal, SMD=0.42/GRADE=low); substance use disorder: TMS (high frequency-dorsolateral-prefrontal-deep (SMD=1.16/GRADE=moderate), high frequency-left dorsolateral-prefrontal (SMD=0.77/GRADE=very low); OCD: DBS (SMD=0.89/GRADE=moderate), TMS (SMD=0.64/GRADE=very low); PTSD: TMS (SMD=0.46/GRADE=moderate); generalized anxiety disorder: TMS (SMD=0.68/GRADE=low); ADHD: tDCS (SMD=0.23/GRADE=moderate); autism: tDCS (SMD=0.97/GRADE=very low). No significant differences for acceptability emerged. Median AMSTAR/AMSTAR-Content was 8/2 (suggesting high-quality meta-analyses/low-quality RCTs), GRADE low.

Despite limited certainty, biological non-pharmacological interventions are effective and safe for numerous mental conditions. Results inform future research, and guidelines.

None.