GLP-1 drugs trim weight despite antipsychotics

People with schizophrenia spectrum disorders carry a heavy burden of metabolic abnormality, and antipsychotic medication frequently makes it worse — driving weight gain, glucose dysregulation and ultimately cardiovascular morbidity and premature death. GLP-1 receptor agonists have shown clear metabolic benefits in the general population, but whether they work as well, and as safely, in this group is uncertain. That matters because psychiatric medication cannot simply be stopped to fix the metabolic problem, and because trials in psychiatric populations are historically small and sparse. This review set out to pool what randomised evidence exists.

The authors conducted a systematic review and meta-analysis of randomised controlled trials of semaglutide and other GLP-1 receptor agonists in adults with schizophrenia spectrum disorders receiving antipsychotic treatment. The work followed PRISMA 2020 guidelines with a protocol prospectively registered in PROSPERO, and electronic databases were searched from inception to the final search date. Risk of bias was assessed with the Cochrane Risk of Bias 2 tool and certainty of evidence with GRADE. Random-effects meta-analyses were used to combine results. Seven randomised controlled trials involving a total of 446 participants met the inclusion criteria, with outcomes spanning weight, body composition measures, glycaemia, lipids and blood pressure.

The primary signal was a statistically significant reduction in body weight compared with control, −6.10 kg with a 95% confidence interval from −12.10 to −0.10 and P = 0.05, but heterogeneity between trials was substantial at I² = 99%. Significant reductions were also seen in body mass index, waist circumference and fasting plasma glucose. Effects on lipid parameters, insulin-related outcomes and blood pressure were inconsistent and generally not statistically significant. Serious adverse events were less frequent in the intervention group than in controls. Overall certainty of evidence ranged from low to moderate, driven mainly by heterogeneity and imprecision.

This is broadly consistent with what GLP-1 receptor agonists do in the general population — weight and glycaemic parameters move, adiposity measures follow — and extends that evidence into a population where metabolic risk is unusually high and treatment options are constrained. But the extreme heterogeneity and a confidence interval that nearly touches zero mean the size of the weight effect here is poorly pinned down. Trials were short and small, lipid and blood pressure findings were not coherent, and the authors call for larger, longer randomised trials with standardised outcome reporting to define efficacy and safety properly.

For readers interested in how far the GLP-1 evidence now reaches, this is a reason to think the metabolic effects are not confined to otherwise healthy people with obesity, and that they may persist alongside medication that itself promotes weight gain. It is emphatically not a settled result: seven small trials with near-total heterogeneity and low-to-moderate certainty is thin ground, and only weight, BMI, waist and fasting glucose moved consistently. Anything involving antipsychotic treatment is a matter for the prescribing clinician. The wider point is that waist circumference and fasting glucose remain the parameters these drugs are judged on.