No detectable rise in depression risk on semaglutide, pooled data suggest

Semaglutide and other GLP-1 receptor agonists have become some of the most widely used drugs in the world for type 2 diabetes, obesity and related metabolic conditions. Alongside their rise have come scattered reports and questions about possible neuropsychiatric effects, including depression and suicidal thoughts, and the published evidence has been inconsistent. Whether a drug taken by millions changes mood is a question worth answering rigorously even if any individual risk is small, so a systematic pooling of all available evidence was overdue.

Following PRISMA 2020 guidelines, the reviewers searched major biomedical databases, trial registries, pharmacovigilance databases, conference abstracts and grey literature from inception to January 2026. They included randomised controlled trials, observational studies, large database analyses and pharmacovigilance disproportionality studies involving patients receiving semaglutide, and assessed study quality with the ROBINS-I and Newcastle-Ottawa tools. Pooled risk ratios were calculated for depression, anxiety and suicidal ideation or attempt against comparator or placebo groups.

The pooled risk ratio for depression was 1.25, for anxiety 1.22 and for suicidal ideation or attempt 1.20, and none of these reached statistical significance. In plain terms, people on semaglutide were not detectably more likely to develop any of the three outcomes than people on comparators or placebo. The underlying studies disagreed with each other substantially, however, and drew on very mixed sources, including spontaneous side-effect reports, which weakens confidence in the exact figures.

The result lands on the reassuring side of an unsettled literature: pooled data do not show the psychiatric harm signal that case reports and some database analyses had hinted at. The authors are explicit about the framing, describing this as the absence of a detected increased risk rather than definitive proof of no psychiatric risk. Heterogeneity, risk of bias and reliance on voluntary reporting systems all limit certainty, and more rigorous prospective monitoring of mood outcomes in trials would settle the question properly.

For anyone taking or considering semaglutide, the balance of pooled evidence currently shows no detected increase in depression, anxiety or suicidal thoughts compared with alternatives or placebo. That is genuinely reassuring, but it is not a clean bill of mental health for the drug, because the evidence base is noisy and partly built on voluntary reports. Mood changes on any medication remain something to raise with a prescriber; this review simply suggests they are not a demonstrated group-level effect of semaglutide.