GLP-1 Heart Failure Meta-analysis

A systematic review published in Drugs on 26 September 2026 pooled 14 trials with 15,882 participants who had heart failure. Its main result was split by heart failure phenotype: GLP-1 receptor agonists were associated with better outcomes in heart failure with preserved ejection fraction (HFpEF), especially among people with overweight or obesity, while evidence of benefit in heart failure with reduced ejection fraction (HFrEF) remained limited^[1]. The paper included a targeted analysis of semaglutide, but it did not establish one class-wide answer for every person with heart failure.

What the review pooled

The authors searched PubMed, Embase and Web of Science from database inception through 10 August 2025. They used two analytical frames. One grouped results by study design: dedicated HFpEF trials, dedicated HFrEF trials and heart failure subgroups drawn from cardiovascular outcome trials. The other grouped participants by phenotype, using a left ventricular ejection fraction threshold of 40% to separate HFpEF from HFrEF^[1]. Random-effects models produced pooled hazard ratios with 95% confidence intervals.

That structure matters because the evidence was not one large heart failure trial. It included dedicated heart failure trials and subgroup analyses from trials that enrolled people with diabetes, chronic kidney disease, obesity or established cardiovascular disease. Pooling can improve precision, but it cannot make those populations interchangeable. Readers can compare this broader evidence with our semaglutide evidence overview.

The HFpEF estimates were favorable

In the phenotype analysis for HFpEF, the pooled hazard ratio was 0.79 for major adverse cardiovascular events (95% CI 0.66 to 0.94), 0.48 for hospitalization for heart failure (95% CI 0.27 to 0.85), and 0.57 for cardiovascular death or worsening heart failure (95% CI 0.37 to 0.89). The estimate for all-cause mortality was 0.75 (95% CI 0.62 to 0.89)^[1]. A hazard ratio below 1 favors treatment, but the size of that relative estimate is not the same as an individual's absolute chance of avoiding an event.

The dedicated STEP-HFpEF trial helps explain the symptom and function findings behind this pattern. It randomized 529 people with HFpEF and obesity to semaglutide or placebo for 52 weeks. The semaglutide group improved 7.8 points more on the Kansas City Cardiomyopathy Questionnaire clinical summary score and walked 20.3 meters farther in six minutes, after adjustment against placebo^[2]. Those endpoints describe symptoms, physical limitations and exercise function. STEP-HFpEF was not sized to settle mortality on its own.

A prespecified analysis of SELECT offers a different view. Among 4,286 participants with a history of heart failure plus overweight or obesity and established cardiovascular disease, semaglutide was associated with a lower rate of major adverse cardiovascular events than placebo. The heart failure composite estimate favored semaglutide, although the phenotype-specific confidence intervals for that composite crossed 1 in both HFpEF and HFrEF^[3]. That is one reason the new review separates dedicated heart failure trials from cardiovascular outcome trial subgroups rather than treating them as one design.

HFrEF remains the unresolved half

The review did not find significant HFrEF reductions in hospitalization for heart failure, the composite of cardiovascular death or worsening heart failure, all-cause mortality, or major adverse cardiovascular events. It did report a cardiovascular death estimate of 0.71 (95% CI 0.54 to 0.95), but the authors still concluded that evidence of benefit in HFrEF remains limited and called for dedicated phenotype-specific randomized trials^[1]. One positive pooled endpoint should not erase the null results around it.

The earlier LIVE trial shows why caution is warranted. It randomized 241 people with stable chronic heart failure and reduced ejection fraction to liraglutide or placebo for 24 weeks. Liraglutide did not improve left ventricular ejection fraction. Heart rate rose by 7 beats per minute, and serious cardiac events occurred in 12 participants receiving liraglutide versus 3 receiving placebo^[4]. That trial studied liraglutide, not semaglutide, and was much smaller than SELECT. It also shows why a class label can hide drug, population and study-design differences.

What we do not know yet

The review abstract does not report the individual weight assigned to each trial, the heterogeneity statistics for every pooled endpoint, or enough detail to judge how sensitive the estimates were to any single study. Its 40% ejection-fraction threshold is described as pragmatic, and real clinical categories do not always divide cleanly at that number^[1]. The underlying studies also mixed dedicated heart failure trials with subgroup analyses from trials built around other primary questions.

We also do not know whether the HFpEF pattern reflects weight loss, direct cardiovascular effects, fewer atherosclerotic events, or some combination. The review says semaglutide benefits were concentrated among participants with overweight or obesity and HFpEF. It does not show that every GLP-1 receptor agonist has the same effect, or that the findings extend to people outside the enrolled populations. Our semaglutide safety summary covers the established adverse-event profile, which still has to be weighed in an individual clinical context.

What this means for readers

This meta-analysis supports a phenotype-specific reading rather than a blanket claim that GLP-1 drugs treat heart failure. The clearest signal sits in people with HFpEF and overweight or obesity. HFrEF needs dedicated outcome trials, and findings from one molecule should not be copied across the class. Semaglutide remains a prescription medicine; its regulatory status and approved indications depend on jurisdiction and product label.

Medical disclaimer: This article is for educational and journalistic purposes only and does not constitute medical advice. Peptides discussed may be classified as prescription medicines or research chemicals depending on your jurisdiction. Always consult a qualified healthcare professional before using any peptide product. PeptideMethods.com does not sell, distribute, or facilitate the sale of any peptide product.