Semaglutide and mental health: what the STEP trials found on depression and suicidal thoughts

Regulators have been watching reports of depression and suicidal thoughts in people prescribed GLP-1 drugs. A pooled analysis of four semaglutide trials looked at what the trial data show. Across 3,377 adults in STEP 1, 2 and 3, depression scores and suicidal ideation were not worse on semaglutide 2.4 mg than on placebo [1].

Why this study matters

Earlier weight-loss drugs have been withdrawn over mood effects, and rimonabant is the example the authors cite. That history is why psychiatric safety is checked closely for new obesity medicines. Our other semaglutide coverage is about weight, such as STEP 1 and STEP 5, or about outcomes in disease, such as SELECT. This analysis asks a different question: whether the same trials showed any mental health signal. The wider record is in the semaglutide evidence guide.

What they did

Depressive symptoms were measured with the PHQ-9, a self-reported nine-question screen scored from 0 to 27. Suicidal ideation and behaviour were measured with the Columbia Suicide Severity Rating Scale (C-SSRS), which a clinician administers. Both were checked at screening, at baseline and at regular visits through the trials. Anyone who scored 15 or higher on the PHQ-9, or reported serious suicidal ideation or behaviour, was referred to a mental health professional [1].

The weight-loss results of the underlying trials are published separately. STEP 1, the largest, enrolled 1,961 adults without diabetes for 68 weeks and reported a mean weight change of -14.9% on semaglutide against -2.4% on placebo [2]. The psychiatric analysis also counted psychiatric and nervous system adverse events reported during the trials. Participants at the start were mostly women and the average PHQ-9 score was about 2, meaning almost no depressive symptoms [1].

What they found

PHQ-9 scores stayed in the none-to-minimal range in both groups. The authors describe the small advantage for semaglutide as statistically significant but not clinically meaningful. In STEP 5, which was much smaller, the PHQ-9 difference was -0.20 points (95% CI -0.98 to 0.59) and was not statistically significant [1].

Suicidal thoughts were uncommon in every group. Across all four trials there were two suicide attempts, one in each arm. The participant on semaglutide reported depressive symptoms linked to life stressors. The participant on placebo had a history of depression and mood swings [1].

Reading the numbers with care

The adverse event row is the one that does not lean toward semaglutide: 6 reports of suicidal ideation against 2 in STEP 1 to 3. The numbers are very small, the clinician-rated scale showed no such imbalance, and the authors did not run statistical tests on suicidal ideation because events were so rare [1]. A third measure, the PHQ-9 item on thoughts of being better off dead or self-harm, pointed the other way in STEP 1 to 3 (3.4% vs 5.0%) and slightly toward semaglutide's disadvantage in STEP 5 (2.6% vs 2.0%) [1]. Mixed, tiny counts like these mean no firm conclusion can be drawn in either direction.

Safety reported

Psychiatric adverse events were balanced. In STEP 5 they were 17.1% in both groups [1]. Nervous system events were not balanced: 29.7% on semaglutide against 21.1% on placebo in STEP 1 to 3, with headaches the largest contributor (16.8% vs 11.7%). The authors say dizziness and headaches were generally more frequent on semaglutide and list both as known effects of the treatment [1].

Caveats

• Sponsor. Novo Nordisk funded the analysis, and sponsor representatives were involved in its design, conduct, analysis and manuscript approval. Several authors are Novo Nordisk employees or shareholders [1].
• Post hoc. The analysis was not planned before the trials began, and no correction was made for multiple comparisons [1].
• Excluded groups. People with recent major depression, other severe psychiatric disorders, a lifetime suicide attempt, a PHQ-9 of 15 or more, or recent serious suicidal ideation were not enrolled. The authors note these people may be the most vulnerable, so the results do not describe them [1].
• Rare events. Suicidal ideation and behaviour were uncommon, which limits what any comparison can show [1].
• One drug and dose. The data cover semaglutide 2.4 mg for weight management in trial conditions, not other doses, other drugs or unregulated products.
• Human trials only. Nothing here comes from animal work.
• Regulators. The authors quote the European Medicines Agency as finding that the available evidence does not support a causal association, and the US FDA as not having found that these medicines cause suicidal thoughts or actions, while being unable to definitively rule out a small risk. Both continue monitoring [1].

What it means in practice

This is a description of what happened to selected trial participants under medical supervision. It is not a recommendation to use semaglutide or any medicine. Anyone experiencing low mood or thoughts of self-harm should speak to a doctor or a crisis service promptly, whatever medicines they take. The authors themselves encourage clinicians to monitor mental health in people with obesity, including those treated with these drugs [1].

What is still open

• Whether the findings hold in people with existing major depression or past suicidal behaviour.
• Whether independent, non-sponsor analyses reach the same result.
• How longer treatment affects mood outside trial conditions.

What this means

We read the analysis as reassuring for the people it studied: across 3,681 adults, semaglutide 2.4 mg was not linked to more depressive symptoms or suicidal ideation than placebo [1]. The sponsor is the manufacturer, the analysis was post hoc, and the most vulnerable people were excluded. The data do not prove there is no risk, and the regulators say as much.

Educational content only — not medical advice. Many peptides discussed on HPR are not approved for human use. Talk to a qualified clinician before making any decision about your health.