GLP-1 receptor agonists cut major heart events by about a fifth in people who are not obese, the same size of benefit they give people who are. That is the headline of a new meta-analysis of 12 randomized trials and 91,167 patients. It asks a question the big drug trials never answered cleanly. Does the heart protection from these drugs depend on the weight you start at?
The short answer, from the meta-analysis ↗ published June 13 in the Journal of Diabetes and Metabolic Disorders, is no. Take patients with a body mass index under 30, the clinical cutoff for obesity. In that group, GLP-1 drugs cut major adverse cardiovascular events, the combined count of heart attack, stroke, and cardiovascular death, with a hazard ratio of 0.81 (95% CI 0.72 to 0.90). In plain terms, the treated group ran about 19 percent fewer of those events than the placebo group. Cardiovascular death alone fell by a similar margin, a hazard ratio of 0.80 (0.70 to 0.91), roughly a fifth fewer deaths.
The number that matters most is not either of those. It is the test for whether body weight changed the effect at all. For major heart events, that interaction test came back at p = 0.53; for cardiovascular death, p = 0.49. Both are far from significant, which is the statistician's way of saying the benefit in lean patients and the benefit in obese patients were the same size within the noise. The drug did not work harder in heavier people.
Why this detaches the heart from the scale
The easy story about GLP-1 drugs is that they help the heart because they shrink the patient. Lose 15 percent of body weight and blood pressure, blood sugar, and inflammation all tend to follow. If that were the whole mechanism, the cardiovascular benefit should shrink as the starting weight shrinks, and disappear in people who were never heavy. It does not. A benefit that holds flat across the body-mass range points at something the drug does directly, in the blood vessel wall and the heart muscle, that is not just a downstream echo of a smaller waistline.
That is not a new idea, but it has been hard to prove. The landmark cardiovascular outcome trials, the ones this analysis pooled, were run mostly in people with diabetes or established obesity. LEADER tested liraglutide, REWIND tested dulaglutide, and the semaglutide program carried much of the recent weight. Drugs like semaglutide ↗, liraglutide ↗, and dulaglutide ↗ all act on the same receptor, GLP-1R ↗, and the question of how much of their cardiac payoff is weight-mediated has real consequences for who should get them.
What the analysis does not show
One outcome did not move. Hospitalization for heart failure showed a hazard ratio of 1.00 (0.68 to 1.47) in the non-obese group, which is to say no effect, with a confidence interval wide enough to drive a truck through. GLP-1 drugs are not a heart failure therapy in lean patients on this evidence, and the authors do not pretend otherwise.
The deeper caveat is structural. Body-mass subgroup counts were available in only 10 of the 12 trials, and these were patients enrolled because they had diabetes or high cardiovascular risk, not healthy lean adults pulled off the street. The finding says the benefit does not require obesity. It does not say a 25-year-old with a normal BMI and no risk factors should start injecting. The authors call for long-term data in lower-risk, healthy-weight populations before anyone stretches the indication that far.
Still, the direction is clear and the registration was clean (the protocol was filed in PROSPERO before the analysis ran, which guards against fishing for a result). For a class of drugs that the public still files under weight loss, the more interesting organ may be the one that was never on the label.