GLP-1 drugs cut weight in antipsychotic users

People with schizophrenia carry a markedly higher cardiometabolic burden than the general population, and a large part of that is attributed to the medicines that treat their illness. Antipsychotics — clozapine and olanzapine regimens especially — are associated with weight gain, insulin resistance and disturbed blood lipids, and stopping them is usually not an option. That creates a specific clinical problem: can the metabolic damage be addressed pharmacologically while antipsychotic treatment continues? GLP-1 receptor agonists are the obvious candidate given their effects on weight and glucose elsewhere, but the individual trials in this population have been small, so a pooled estimate was needed.

The authors searched PubMed, Embase and the Cochrane Central Register of Controlled Trials from inception to 16 September 2025 for randomised clinical trials comparing GLP-1 receptor agonists with placebo or standard care in antipsychotic-treated patients with schizophrenia. The review followed PRISMA guidance and was prospectively registered with PROSPERO. Random-effects models pooled mean differences for continuous outcomes and risk ratios for events, covering anthropometric measures, glycaemic markers, fasting insulin, lipid parameters, blood pressure and adverse events. Exploratory subgroup analyses compared individual agents, separating exenatide, liraglutide and semaglutide, to see whether the class effect was uniform.

Eight randomised trials including 523 participants were analysed. GLP-1 receptor agonists significantly reduced body weight (mean difference -6.47; 95% CI -9.61 to -3.32), body mass index (-2.83; 95% CI -3.81 to -1.86), waist circumference (-5.35; 95% CI -6.60 to -4.09) and glycated haemoglobin (-0.32; 95% CI -0.39 to -0.26) relative to comparators. Several outcomes did not move: there were no significant effects on fasting insulin, lipid parameters or blood pressure. The agent subgroups diverged sharply, with exenatide showing a smaller and non-significant weight reduction, liraglutide a significant moderate reduction, and semaglutide the largest. Gastrointestinal adverse events were more frequent with GLP-1 receptor agonists, which the authors nonetheless described as acceptable tolerability.

This extends the GLP-1 evidence base into a population that is usually excluded from obesity trials despite having some of the greatest metabolic need, and the direction and size of the anthropometric effects broadly echo what is seen elsewhere. The null results for insulin, lipids and blood pressure complicate any assumption that weight loss automatically drags every cardiometabolic marker with it, at least over the durations studied. With 523 participants across 8 trials the analysis is small, and the agent-level differences are exploratory, so whether semaglutide truly outperforms the older agents in this group needs dedicated trials. Longer follow-up and cardiovascular outcomes remain untested here.

For anyone following how antipsychotic-associated weight gain might be managed, this is a reason to think GLP-1 receptor agonists produce meaningful reductions in weight, BMI, waist size and average blood glucose without stopping psychiatric treatment. It is equally a reason to temper expectations about blood pressure and blood lipids, which did not shift significantly in the pooled data. The gut side effects seen across this drug class showed up here too. All of this sits firmly within psychiatric and medical care, so its value is in understanding the current state of evidence rather than in any change made independently.