JAMA Psychiatry study links semaglutide and tirzepatide to lower mortality in serious mental illness

A study in JAMA Psychiatry reports that adults with serious mental illness who started GLP-1 receptor agonists had lower death rates and fewer major cardiovascular events than similar patients who started another class of diabetes drugs, SGLT2 inhibitors ^[1].

The analysis is a target trial emulation built from electronic health records in the TriNetX network, a multinational federated database. It matched 764,115 pairs of adults, among them 195,184 pairs with serious mental illness — major depressive disorder, bipolar disorder or schizophrenia — and 568,931 pairs without ^[1]. Mean ages were about 60 in both treatment groups and in both cohorts, and about 58% of the serious mental illness cohort was female ^[1].

Over four years, 9,585 of 195,184 patients with serious mental illness who started a GLP-1 (4.91%) died, compared with 12,584 of 195,184 (6.45%) who started an SGLT2 inhibitor. That works out to a hazard ratio of 0.76 (95% CI, 0.74-0.78) and an absolute difference of 1.54 percentage points ^[1]. In a separately matched one-year cohort of 198,065 pairs, 1.46% of GLP-1 initiators died against 2.84% of SGLT2 initiators, a relative risk of 0.52 (95% CI, 0.49-0.54) and an absolute difference of 1.38 percentage points ^[1].

The paper also reports lower risks among GLP-1 users for three-point and five-point major adverse cardiovascular events, myocardial infarction, stroke, heart failure and coronary artery bypass grafting ^[1]. Ten-year exploratory analyses limited to people with type 2 diabetes found lower mortality with semaglutide — sold as Ozempic, Wegovy and Rybelsus — than with SGLT2 inhibitors in each psychiatric group: relative risk 0.55 for major depressive disorder, 0.57 for bipolar disorder and 0.67 for schizophrenia ^[1]. The authors write that the benefits were driven by semaglutide and tirzepatide, sold as Mounjaro and Zepbound, and that they were evident within one year and consistent across the psychiatric subgroups ^[1].

Patients were not randomly assigned to a drug, so the results cannot show that the GLP-1s caused the difference. The paper says the primary mortality finding held across prespecified sensitivity analyses, including stratification by diabetes and exclusion of people with baseline heart failure or chronic kidney disease ^[1]. Data were queried and analyzed in March 2026, and additional subgroup analyses were conducted after peer review in May 2026 ^[1].

Why it matters for patients

Bipolar disorder, major depressive disorder and schizophrenia are associated with substantial excess mortality, chiefly from cardiovascular disease, according to the study's background section ^[1]. The absolute reductions associated with GLP-1 initiation were larger for people with serious mental illness than for those without it ^[1].

The comparison group matters: patients who started an SGLT2 inhibitor were taking an active diabetes and heart treatment, not a placebo ^[1]. The reported difference is therefore between two drug classes, not between treatment and no treatment.

The study relies on prescription records and diagnosis codes. It does not report whether patients took the drugs as prescribed, how long they stayed on them, or why a clinician selected one drug over another, and it does not report insurance coverage, out-of-pocket costs or how participants obtained access. Whether the mortality gap in this population would narrow if more people with serious mental illness were prescribed GLP-1s has not been established by this kind of data.

What happens next

The authors conclude that prospective randomized trials are warranted ^[1]. No specific trial, sponsor or start date is named in the paper.