For anyone on or considering this kind of treatment: why a comparison of trial averages cannot tell you whether your own weight loss sets the size of your heart protection.
Sixteen placebo-controlled randomised trials of incretin-based drugs, with 97,199 participants with and without diabetes. A team from Šibenik-Knin County General Hospital and the University of Rijeka in Croatia pooled them to ask a narrow question at the centre of metabolic medicine: in the trials where the drug produced a larger weight difference against placebo, were the hazards of heart attack, stroke and death also lower? The systematic review and meta-regression appeared online in Diabetes, Obesity and Metabolism on 3 November 2026.
Trial averages of weight and of events did not line up
The researchers compared each trial's placebo-adjusted percentage weight difference at its latest reported follow-up with its hazard ratios. A larger weight difference was not significantly associated with lower hazard ratios for three-point major adverse cardiovascular events (3P-MACE), nor for cardiovascular death, death from any cause, heart attack, stroke or hospitalisation for heart failure taken one by one.
In the standard meta-analysis of the same 16 placebo-controlled trials, the drug groups had pooled relative reductions of 15% in 3P-MACE, 13% in cardiovascular death, 12% in death from any cause, 14% in heart attack, 13% in stroke and 10% in hospitalisation for heart failure compared with placebo.
One dot per trial cannot answer for one person
This is an aggregate-data meta-regression: each trial contributes a single point, its average weight difference and its hazard ratio. The authors spell out what follows from that design: the finding does not establish an individual-level or causal relationship, and it does not exclude that weight loss contributes to the cardiovascular benefit in a given person. They call for analyses of individual participant data to settle it.
The abstract does not state who funded the work; this account rests on the published abstract.
What stays open for the individual patient
The analysis cannot say whether a weight plateau on treatment, or an unusually large loss, changes the cardiovascular effect for a particular person. Questions about starting, continuing or changing such a drug belong in a conversation with a doctor, together with blood pressure, cholesterol and blood sugar.
This article is for information only and does not replace a consultation with a doctor.
Sources
Skelin M, Perkov-Stipičin B, Bračić V, Škugor ZM, Krečak I. Association between weight loss on incretin-based therapy and cardiovascular outcomes and mortality: a systematic review and meta-regression. Diabetes Obes Metab, online 3 November 2026. doi:10.1111/dom.71511 · PubMed 42978220
Articles in this section are educational and are not medical advice, a diagnosis, or a prescription. Consult a qualified professional before acting on anything you read here.
Read next
ResearchMedicine and treatmentHormone replacement for absent parathyroids held up over 3.5 years
For adults with hypoparathyroidism who live on calcium and vitamin D tablets: what the longest follow-up of this daily hormone replacement found, and where the evidence still stops.
ResearchBiomarkersAn eye photo predicts death risk almost as well as age
For anyone curious about AI tests of biological age: what a new retinal model adds to the oldest risk factor of all, which causes of death it reads best, and what it still cannot tell one person.
ResearchMedicine and treatmentKnees, cholesterol and an AI ECG: what four new trials showed
4 stories this week: knee osteoarthritis with obesity, a cholesterol-lowering strategy in artery disease, AI screening for heart rhythm problems and clots around arm catheters. Four randomised trials, each with what it does not yet show.
Digest · 4 stories