GLP-1 receptor agonists were associated with lower mortality than SGLT2 inhibitors in serious mental illness

Design

Retrospective target trial emulation, new-user active-comparator, propensity matched

Population

195,184 matched pairs of adults with serious mental illness starting a GLP-1 RA or SGLT2 inhibitor

Primary outcome

All-cause mortality at 4 years

Effect

4.91% vs 6.45%, HR 0.76 (95% CI 0.74-0.78)

This target trial emulation used the TriNetX electronic health record network. Adults starting a GLP-1 receptor agonist or an SGLT2 inhibitor were propensity matched, in separate cohorts with and without serious mental illness (bipolar disorder, major depressive disorder, schizophrenia).

Among 195,184 matched pairs with serious mental illness, four-year mortality was 4.91% with a GLP-1 receptor agonist and 6.45% with an SGLT2 inhibitor (HR 0.76, 95% CI 0.74-0.78; absolute difference 1.54 percentage points). At one year the gap was 1.46% vs 2.84%. In those with type 2 diabetes, semaglutide was associated with lower 3-point MACE (HR 0.77). Absolute reductions were larger than in people without mental illness.

Excess cardiovascular death drives much of the mortality gap in serious mental illness, and psychiatrists often defer metabolic prescribing to others. This is a signal that the choice of agent may matter.

The mortality difference emerged within a year, which is faster than a cardiovascular mechanism alone would explain and suggests that the two groups differed in ways matching could not capture.

• When metabolic treatment is being chosen for someone with serious mental illness, raise GLP-1 receptor agonists with the prescribing physician.
• Record weight, HbA1c and lipids at antipsychotic review so the case for treatment is visible.
• Check for nausea and reduced intake, which can affect lithium levels and oral medication absorption.
• Treat the finding as hypothesis-generating until randomised data exist.

Why it matters

It suggests metabolic drug choice may be one lever on the cardiovascular mortality gap that psychiatric services rarely pull.

Don't overread it

This was observational; a mortality difference visible within one year points to confounding that propensity matching did not remove.

The statistics, in plain English

A hazard ratio of 0.76 means about a quarter lower rate of death over follow-up, and the absolute difference is about 15 deaths per 1,000 people over four years. Very large databases produce very narrow confidence intervals, but narrowness reflects sample size, not freedom from bias.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.