In a propensity-matched TriNetX cohort of 218,043 adults with serious mental illness, type 2 diabetes, obesity, and active antipsychotic or antidepressant treatment, GLP-1 receptor agonist users had lower incident myocardial infarction, heart failure, and all-cause death than nonusers. Three-year absolute risk differences were -1.0 points for MI, -1.8 for heart failure, and -6.5 for death. Semaglutide and tirzepatide showed similar benefit, and results held after accounting for SGLT2 and DPP-4 inhibitor use.
This population is often left out of cardiometabolic trials, even though antipsychotics drive weight gain and insulin resistance and people with SMI die roughly 10 to 20 years earlier than the general population, largely from cardiovascular causes. The direction of effect matches the large GLP-1 outcome trials, so the finding is confirmatory rather than surprising. The size of the mortality gap, however, warrants caution. A 6.5-point absolute drop in death over three years far exceeds what randomized trials show in general populations, which points to residual confounding: healthier, better-connected, more adherent patients are likelier to receive these drugs, and propensity matching on coded EHR data cannot fully capture that. Immortal-time bias and administrative case ascertainment add further uncertainty, as the authors concede. Treat this as hypothesis-strengthening evidence that people with SMI should not be denied GLP-1 therapy by default, and as a case for dedicated randomized trials in this group.