- Design
- Systematic review and frequentist network meta-analysis of randomised trials, CINeMA grading
- Population
- 5,898 antipsychotic-treated people with schizophrenia-spectrum disorders; 95 trials, 39 drugs
- Primary outcome
- Change in body weight versus placebo or standard care
- Effect
- Semaglutide -10.98 kg (-13.33 to -8.62); metformin -3.86 kg (-5.02 to -2.70)
A systematic review and network meta-analysis included 95 randomised trials of 39 drugs in 5,898 antipsychotic-treated people with schizophrenia-spectrum disorders, grading certainty with CINeMA.
Against placebo, the largest weight reductions with moderate certainty were semaglutide (-10.98 kg, 95% CI -13.33 to -8.62; 3 trials), liraglutide (-5.43 kg), topiramate (-3.95 kg), metformin (-3.86 kg; 16 trials) and exenatide (-2.97 kg). Only semaglutide and metformin produced a clinically meaningful loss of 5% or more. Gastrointestinal effects and dropout did not raise major concerns. Other agents, including aripiprazole, had only very low-certainty evidence.
Antipsychotic weight gain is often accepted as the price of treatment. This review gives a ranked, graded menu, so doing nothing is harder to justify.
Metformin has the deepest evidence base and is inexpensive and widely available, including in India, so it is the practical first step for most patients. Semaglutide gives the largest loss where it is affordable and appropriate.
- Offer metformin when weight gain on an antipsychotic reaches 5% or more, alongside lifestyle support.
- Consider semaglutide where metformin is insufficient or contraindicated and cost allows.
- Topiramate is an option but check cognitive side effects and interactions.
- Consider switching to a more weight-neutral antipsychotic where the mental state allows.
- Recheck weight and metabolic markers at 12 weeks after starting any adjunct.
Why it matters
It turns antipsychotic weight gain from a tolerated side effect into a treatable problem with a graded evidence base.
Don't overread it
The semaglutide estimate rests on three trials, so its size is less certain than metformin's 16-trial estimate.
The statistics, in plain English
A network meta-analysis compares drugs that were never tested head to head by linking them through placebo. The estimates are mean differences in kilograms against placebo. 'Moderate certainty' means the true effect is probably close to the estimate but could be substantially different; k is the number of trials behind each estimate.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for psychiatry, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free