Semaglutide and depression: 2026 review

A systematic review and meta-analysis published in October 2026 in Clinical Obesity pooled evidence from randomised controlled trials, observational studies, large database analyses, and pharmacovigilance data to ask a narrow question: do patients receiving semaglutide carry a higher burden of depression, anxiety, or suicidal ideation than comparators?^[1] The answer the authors reached was nuanced: pooled risk ratios pointed upward for all three outcomes, but none crossed into statistical significance, and the heterogeneity in the data was so high that any firm conclusion in either direction is premature.

What the numbers say

Bhaduri and colleagues found a depression risk ratio of 1.25 (95% CI: 0.95-1.65) across included studies^[1], a result that spans the null. The anxiety risk ratio was 1.22 (95% CI: 0.93-1.60) and the suicidal ideation or attempt ratio was 1.20 (95% CI: 0.90-1.62). Every confidence interval crosses 1.0, meaning none of these pooled estimates would conventionally be described as a statistically significant increased risk. The I-squared values, however, tell a different story about the data quality: 98% for depression, 99% for anxiety, and 92% for suicidal ideation. I-squared statistics in that range indicate that the studies in the pool were measuring very different things in very different populations, which limits what any pooled estimate can tell us.

The authors were transparent about the drivers of heterogeneity: the pool included spontaneous adverse-event reports alongside randomised trial arms, real-world database studies, and pharmacovigilance signals. These data sources differ not just in design but in the populations they capture and the way outcomes are identified.

Where regulators stand

The regulatory picture predates the Bhaduri review. In April 2024 the European Medicines Agency's Pharmacovigilance Risk Assessment Committee concluded that 'the available evidence does not support a causal association' between GLP-1 receptor agonists, including semaglutide (Ozempic, Rybelsus, Wegovy), and suicidal or self-injurious thoughts and actions^[2]. The PRAC analysis drew on non-clinical studies, clinical trial data, post-marketing surveillance, and two independent electronic health record studies. No update to product information was required.

A separate systematic review published in Clinical Therapeutics in April 2026 that examined 82 studies on GLP-1 receptor agonists and neuropsychiatric outcomes reached a similar position: 'no associations were observed for depression or anxiety'^[3]. The same analysis found signals for potential benefit in Parkinson's disease prevention and binge eating disorder, but emphasised that evidence certainty for most outcomes remained low to very low.

A contrasting real-world signal emerged from a South Korean multicentre cohort study published in September 2026^[4]. Among 2,357 semaglutide initiators matched against 22,602 non-initiators, the study reported a hazard ratio of 3.42 for depressive disorder and 2.39 for anxiety. The authors note that residual confounding in a propensity-matched observational design cannot be excluded, and the finding sits in tension with both the PRAC review and the controlled trial data. That tension is itself informative: it suggests that people who initiate semaglutide in real-world settings may differ in mental health baseline from those enrolled in clinical trials.

What this means for patients considering semaglutide

People who are considering semaglutide for weight management or type 2 diabetes and who have a personal or family history of depression should discuss that history with their prescribing clinician before starting. Nothing in the Bhaduri review or the regulatory record establishes that semaglutide causes depression, but the real-world data are not uniformly reassuring, and the trial populations studied to date may not fully represent people with pre-existing psychiatric conditions.

The approved labels for semaglutide products (Ozempic, Wegovy, Rybelsus) do not currently carry a warning for depression or suicidal ideation in the EU or the UK. Prescribers are already expected to conduct routine mental health monitoring for patients with obesity, a condition that carries its own elevated depression burden independent of any pharmacotherapy.

What we don't yet know

Three gaps stand out. First, the Bhaduri review could not separate patients with pre-existing depression from those who developed the condition after starting semaglutide, which is the clinically relevant question. Second, all pooled estimates carry extremely high I-squared values, meaning a more refined analysis stratified by study design type, depression history, duration of use, and dose would tell a materially different story than the pooled point estimate. Third, the long-term data are thin: most included trials ran fewer than two years, and the durability question for both weight loss and any psychiatric signal over three to five years remains open.

The mechanistic picture is also incomplete. GLP-1 receptors are expressed in brain regions involved in reward, appetite, and mood regulation. Whether semaglutide's central effects on appetite could interact with mood circuits in vulnerable individuals is a hypothesis that the current body of clinical evidence cannot confirm or rule out.

Who to talk to before starting or stopping

This article is for educational purposes. If you are prescribed semaglutide and notice changes in mood, persistent low mood, or thoughts of self-harm, contact your prescribing clinician or a mental health service promptly. Do not stop a prescribed medication without first talking to your doctor: the risks of stopping abruptly may outweigh the risks of continuing, and the baseline depression burden in people managing obesity is itself a clinical concern. For country-specific prescribing context, see the relevant regulation pages on this site.