People who take antipsychotic drugs for schizophrenia or bipolar disorder tend to gain weight fast and die years early, and a large part of that early death is heart and metabolic disease. A new pooled analysis asked whether the GLP-1 weight-loss drugs, added on top of the antipsychotic, move those numbers. Across eight randomized trials and 664 patients, they did.
The meta-analysis, published online September 16 in the Journal of Psychopharmacology ↗, pooled trials that added a GLP-1 receptor agonist to standard antipsychotic treatment and tracked the usual cardiometabolic markers. Body weight fell by a mean of 6.74 kilograms, about 15 pounds. Body mass index dropped 2.37 points. Waist circumference came down 4.27 centimeters, roughly 1.7 inches. HbA1c, a measure of average blood sugar over about three months, fell 0.61 percentage points, and fasting blood sugar dropped as well.
One drug carried the weight column. In the subgroup analysis, semaglutide, the molecule sold as Ozempic and Wegovy, produced the biggest reductions: 11.06 kilograms of body weight and 3.60 BMI points, enough that the difference between semaglutide and the other GLP-1 agonists was itself statistically significant.
Why this population
The reason to run this analysis at all is the size of the gap it is trying to close. People with serious mental illness die roughly a decade and a half to two decades earlier than the general population, and cardiovascular and metabolic disease drive much of that. The medications that treat the psychiatric illness make the metabolic side worse. Olanzapine and clozapine, two of the most effective antipsychotics, are also among the most aggressive at packing on weight and pushing blood sugar and lipids in the wrong direction. Lifestyle programs in this group tend to underperform what they achieve elsewhere, partly because the illness and its treatment work against them.
So a drug that takes off 15 pounds without the patient having to out-diet the antipsychotic is a real thing. The other half of the result is what did not happen. The pooled trials showed no signal that the GLP-1 drugs destabilized the psychiatric illness or hurt treatment adherence, which is the fear that has kept some prescribers away. Overall adverse events, serious adverse events, and dropouts for side effects were all statistically similar between the GLP-1 and control arms.
The cost, and the ceiling
The cost is the gut. Gastrointestinal side effects, nausea, vomiting, and constipation, were significantly more common on the GLP-1 drugs, each with a p-value under 0.01. In a population already managing a heavy medication load and, often, a fragile relationship with treatment, that is not a footnote. The authors call for careful dose titration, the slow ramp that keeps the nausea manageable.
The larger caveat is what the trials measured. Weight, BMI, waist, HbA1c, and fasting glucose are surrogates. They are the things that sit upstream of the heart attacks and the early deaths, and moving them is a reasonable bet on moving the hard outcomes, but the trials did not run long enough or large enough to show fewer cardiovascular events or fewer deaths. Eight trials and 664 patients is a thin base for a claim about a group this vulnerable, and the authors, a team led from Cairo University with collaborators across several Egyptian medical schools and a US center, grade the certainty accordingly and ask for larger trials with longer follow-up.
The shape here rhymes with an earlier GLP-1 add-on story. When these drugs were added to insulin in type 1 diabetes ↗, the weight and insulin numbers moved while the harder endpoint stayed flat and the most certain finding was the harm profile. Adjunct GLP-1 keeps landing in the same place: a real effect on the metabolic surrogates, a real gastrointestinal tax, and a hard outcome the trials were never built to reach.
Both semaglutide ↗ and the older daily agent liraglutide ↗ act on the GLP-1 receptor ↗, the gut-hormone switch that slows the stomach and blunts appetite. That mechanism is exactly why they move weight in people on antipsychotics, and exactly why the nausea comes along for the ride.