Mental disorders affect nearly one billion people worldwide, posing major challenges to public health. While conventional treatments like psychotherapy and pharmacotherapy are effective for many patients, they are often associated with adverse effects and high non-response rates, underscoring the need for alternative approaches. Non-invasive brain stimulation techniques such as transcranial magnetic stimulation, transcranial electric stimulation and transcranial focused ultrasound stimulation are increasingly used to treat psychiatric conditions. Although these methods show promising efficacy, data on their adverse effects remain fragmented and inconsistently reported. This meta-analysis aims to systematically compare the type and frequency of adverse effects, tolerability, and acceptability across different brain stimulation techniques and mental disorders. The findings will help improve safety monitoring and support more personalised, well-tolerated treatment strategies.

A systematic literature search of the Embase, MEDLINE(R), AMED (Allied and Complementary Medicine) and APA PsycINFO via OVID will be performed. Eligible studies include randomised controlled trials (RCTs) that compare active treatments or an active treatment with sham control, including both parallel group and cross-over studies, as well as prospective non-randomised studies such as case–control studies and pre–post studies investigating adverse effects of non-invasive brain stimulation in psychiatric populations. Included studies report on the frequency of adverse effects in a standardised manner. Primary outcomes comprise the incidence of specific adverse effects, dropout rates due to adverse effects (tolerability) and overall dropout rates (acceptability). Risk of bias will be assessed using the Cochrane RoB 2.0 tool for RCTs and the NHLBI quality assessment tool for pre–post studies. The quality of case–control studies will be assessed using the Newcastle-Ottawa scale. Provided that sufficient data are available and the network of comparisons is adequately connected, a network meta-analysis will be conducted to compare adverse effects and tolerability across interventions.

No ethical approval is needed to conduct this work. The findings will be submitted for publication in peer-reviewed journals and presented at scientific meetings.

CRD420251164554