A systematic review and meta-analysis published in the British Journal of Clinical Pharmacology found that GLP-1 receptor agonists were associated with meaningful weight and metabolic improvements in adults with antipsychotic-treated schizophrenia-spectrum disorders, while effects on psychiatric symptoms remained too imprecise to call.

GLP-1 receptor agonists are the drug class that includes semaglutide (sold as Ozempic and Wegovy) and related compounds, which mimic a gut hormone that slows stomach emptying and signals fullness. The population here matters: antipsychotic medication is strongly linked to weight gain, and this group has long been underserved by obesity research.

What was pooled

The authors searched eight databases plus trial registries and pooled six trials totaling 446 participants. Only three of those trials, covering 102 people, informed the primary outcome of total symptom severity.

Key results

On the cardiometabolic side the numbers were substantial: a mean difference of 7.42 kg in body weight (95% CI -11.52 to -3.33), a BMI reduction of 2.77 kg/m2, waist circumference down 5.53 cm, and HbA1c down 0.36 percent. Gastrointestinal events were more frequent in treated participants, while serious adverse events were not.

The mental-health picture was flatter. The primary symptom-severity estimate was an SMD of -0.21 with a confidence interval running from -0.60 to 0.19, and the more conservative Hartung-Knapp-Sidik-Jonkman interval stretched from -1.02 to 0.61. Functioning, alcohol use, quality of life and cognition measures all produced intervals that included the null.

Why the weight figure needs an asterisk

Heterogeneity for the weight outcome was high at I2 = 94 percent, which the authors attribute to differences in agent, dose, duration and antipsychotic background. A 7 kg pooled average across trials that disagree that much is a direction of travel, not a number anyone should quote as an expected result.

Limitations

The pooled sample is small and the primary outcome rests on 102 participants across three trials. The authors rate certainty for body weight as limited by heterogeneity and risk of bias, and describe the secondary mental-health analyses as exploratory and unadjusted for multiplicity. Cardiovascular events were not assessed at all, so no conclusions about cardiovascular outcomes in this population can be drawn.

Bottom line

Over 12 to 36 weeks, the cardiometabolic signal is the part of this review with something behind it. Whether these drugs move psychiatric symptoms in either direction remains an open question, and answering it will take trials built for that purpose.