GLP-1 receptor agonists linked to fewer psychiatric admissions in schizophrenia with diabetes

Design

Observational cohort, US electronic health records, inverse probability weighting

Population

70,257 adults with schizophrenia and type 2 diabetes

Primary outcome

Psychiatric hospitalisation and emergency department presentation

Effect

Hospitalisation HR 0.76 (95% CI 0.71 to 0.82) vs non-initiation; 0.83 (0.78 to 0.88) vs SGLT2 inhibitors

A US cohort used electronic records from 70,257 adults with schizophrenia and type 2 diabetes, comparing those who started a GLP-1 receptor agonist with non-initiators and with those who started a DPP-4 or SGLT2 inhibitor. Groups were balanced by inverse probability weighting and followed for up to two years.

Starting a GLP-1 receptor agonist was associated with fewer psychiatric hospitalisations than not starting one (weighted HR 0.76, 95% CI 0.71 to 0.82), and than starting a DPP-4 inhibitor (0.82) or SGLT2 inhibitor (0.83). Psychiatric emergency presentations followed the same pattern (HR 0.71 versus non-initiation).

The comparison with other diabetes drugs is the more useful one, because it partly controls for the fact that people prescribed any new agent are more engaged with care. Even so, residual confounding remains, and the authors are clear that this does not make these drugs a psychiatric treatment.

Where a GLP-1 receptor agonist is already indicated for diabetes or weight in someone with schizophrenia, this is a reason not to hold back. It is not a reason to prescribe one for psychosis. In India, cost and supply of these agents remain the practical barrier for many patients.

• In schizophrenia with type 2 diabetes, review whether a GLP-1 receptor agonist is metabolically indicated.
• Coordinate with the diabetes team rather than prescribing outside your usual scope.
• Watch for nausea and reduced intake, which can affect lithium and clozapine levels through dehydration.
• Do not offer these drugs as an antipsychotic or to reduce relapse on this evidence.

Why it matters

It suggests a drug given for the body may not cost anything for the mind, and could help.

Don't overread it

This was observational; it cannot show that GLP-1 receptor agonists prevent psychiatric relapse.

The statistics, in plain English

A hazard ratio of 0.76 means about a quarter fewer psychiatric hospitalisations over the follow-up period in those who started the drug. Weighting balances measured differences between groups but cannot balance unmeasured ones, such as how engaged someone is with their care.

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.