Tirzepatide and Semaglutide show comparable neuropsychiatric risks; Semaglutide lowers risk vs. other GLP-1 RAs.

Concerns regarding neuropsychiatric safety, including depression and suicidal ideation, have emerged with the increasing use of incretin-based therapies for type 2 diabetes mellitus (T2DM) and obesity. While these therapies are highly effective for metabolic management, the potential for differential neuropsychiatric risks across newer dual-agonists like tirzepatide versus established GLP-1 receptor agonists (GLP-1 RAs) like semaglutide, or even older GLP-1 RAs, remains an important clinical question. Understanding these differences is crucial for informed prescribing, especially for patients with pre-existing psychiatric conditions.

This retrospective cohort study utilized the TriNetX Global Federated Network, encompassing over 192 million patients. Researchers identified adults with T2DM, obesity, or both who initiated tirzepatide, semaglutide, or other GLP-1 RAs between July 2022 and June 2025. A new-user design with a 12-month washout period was employed. Propensity score matching was used to balance demographic, clinical, and metabolic variables. Two primary comparisons were conducted: tirzepatide versus semaglutide, and semaglutide versus other GLP-1 RAs. Follow-up included Year 1 (day 31-365) and Year 2 (day 366-730) landmark analyses. Hazard ratios (HRs) for incident depression, anxiety, and suicidal ideation were estimated using Cox models.

After rigorous propensity score matching, the study included 85,546 pairs for the tirzepatide versus semaglutide comparison and 80,115 pairs for the semaglutide versus other GLP-1 RAs comparison. Over 2 years, tirzepatide and semaglutide demonstrated similar risks for the composite psychiatric outcome, with a Year 1 HR of 0.984 [95% CI 0.950-1.019] and a Year 2 HR of 1.002 [0.960-1.046]. A nominally higher anxiety hazard was observed with tirzepatide during Year 2 (HR 1.052 [1.001-1.106]), though this finding warrants cautious interpretation due to multiple comparisons. In contrast, when compared to other GLP-1 RAs, semaglutide was associated with significantly lower risks across multiple neuropsychiatric outcomes during Year 1.

Semaglutide reduced the risk of suicidal ideation by over 50% (HR 0.488 [0.339-0.702]) compared to other GLP-1 RAs. It also showed lower risks for depression (HR 0.811 [0.770-0.855]), anxiety (HR 0.915 [0.871-0.961]), and the composite psychiatric outcome (HR 0.866 [0.832-0.901]) during Year 1.