Semaglutide Helps a Specific Kind of HFpEF—Not All

Semaglutide has given cardiology something unusual in heart failure with preserved ejection fraction (HFpEF): a treatment signal that matches a recognizable patient phenotype. In adults with established HFpEF and obesity, the STEP-HFpEF trials found better symptoms, greater exercise capacity, weight loss, and lower inflammatory and cardiac-stress markers over 52 weeks. That does not mean semaglutide treats every form of HFpEF, prevents heart disease in healthy people, or turns a lower C-reactive protein result into a prescription.

The distinction matters because HFpEF is a syndrome, not a single disease. Patients share signs and symptoms of heart failure despite a left ventricular ejection fraction that remains at least moderately preserved, but the drivers can differ: obesity-related inflammation, hypertension, atrial fibrillation, amyloid disease, kidney disease, and more. A drug can help one subgroup while doing little for another.

What STEP-HFpEF actually tested

The STEP-HFpEF program enrolled adults with established HFpEF, left ventricular ejection fraction of at least 45%, obesity defined as a body-mass index of at least 30, and New York Heart Association class II to IV symptoms. It consisted of two parallel, randomized trials: 529 participants without diabetes and 616 with type 2 diabetes.

Participants received semaglutide 2.4 mg once weekly or placebo in a 1:1 ratio for 52 weeks. The co-primary endpoints were change in the Kansas City Cardiomyopathy Questionnaire Clinical Summary Score (KCCQ-CSS), a patient-reported measure of symptoms and physical limitation, and six-minute walk distance. Those are meaningful outcomes for people living with HFpEF: they address whether daily life feels less restricted and whether the patient can physically do more.

The primary outcome reports described clinically meaningful improvements with semaglutide in KCCQ-CSS and six-minute walk distance, along with substantial weight loss and reductions in C-reactive protein and NT-proBNP. The trial-design publication establishes the population and protocol, but it does not provide the outcome estimates themselves. That limits how precisely this evidence can be summarized from the published design report alone.

A 2025 meta-analysis by Jia and colleagues supports the overall direction of those findings at the drug-class level. It pooled GLP-1 receptor agonists, including semaglutide and tirzepatide, in obese patients with chronic heart failure and assessed mortality, cardiovascular mortality, worsening heart failure, body weight, KCCQ-CSS, six-minute walk distance, BNP, and high-sensitivity C-reactive protein. Because it combines several drugs and heart-failure subtypes, it cannot provide a clean semaglutide-specific estimate for HFpEF. It does, however, support improvement in functional outcomes, biomarkers, and body weight in obese chronic-heart-failure populations.

Why obesity may be the gatekeeper

The biological case is plausible, though plausibility is not proof. A 2021 mechanistic review by Tschöpe and colleagues describes visceral fat expansion as a central driver of an obesity-related HFpEF phenotype. Visceral and epicardial fat can promote inflammatory signaling, adipokine disruption, fibrosis, myocyte stiffness, and mechanical restraint on diastolic filling. Those processes can make the heart stiff even when its ejection fraction looks preserved.

Semaglutide produces weight loss, and GLP-1 receptor activation may also affect inflammatory and adipose-tissue pathways. But STEP-HFpEF did not include a weight-matched placebo group, so it cannot separate the benefit of weight loss itself from any direct effect of GLP-1 receptor signaling. The mechanistic story fits the results; it does not settle the causal division of labor.

The diabetes and non-diabetes arms reportedly produced comparable functional results, which argues against glucose lowering as the only explanation. Still, comparable results across trial arms do not prove that inflammation, epicardial-fat reduction, or another proposed pathway caused the improvement. Human outcome trials answer whether treatment helped; they rarely identify one tidy molecular culprit.

The evidence does not cover all HFpEF

The most important boundary comes from a 2025 systematic review and meta-analysis by Zhang and colleagues. Across cardiovascular and renal disease populations, GLP-1 receptor agonists and tirzepatide reduced heart-failure risk significantly in obese participants but did not show a statistically significant reduction in non-obese participants.

This was not a trial confined to semaglutide-treated HFpEF, so it cannot prove that every non-obese person with HFpEF fails to benefit. It does make broad extrapolation untenable. The STEP-HFpEF trials intentionally studied patients with BMI of at least 30, and the meta-analysis found no statistically significant heart-failure-risk benefit in non-obese populations. Leaner HFpEF driven chiefly by hypertensive remodeling, amyloidosis, or atrial-fibrillation-related dysfunction may not share the adipose-inflammatory biology targeted in obesity-related HFpEF.

Semaglutide's HFpEF evidence is phenotype-specific: established HFpEF plus obesity. BMI 30 was an enrollment threshold, not proof of a biological cliff at exactly BMI 30.

That framing also helps explain why earlier HFpEF drug programs often disappointed. Broadly enrolled HFpEF populations can mix several diseases under one label. A treatment aimed at one mechanism may show a diluted average effect when many participants have another dominant mechanism. This is a reasonable explanation, not a retroactive proof that every prior neutral trial failed for the same reason.

SELECT shows cardiovascular benefit in a different population

The SELECT trial provides semaglutide's best hard cardiovascular-endpoint evidence, but it answers a different question. In 17,604 adults without diabetes who had established atherosclerotic cardiovascular disease and BMI of at least 27, semaglutide 2.4 mg weekly reduced first major adverse cardiovascular events versus placebo over a mean 40 months. The hazard ratio was 0.80 (95% confidence interval, 0.72 to 0.90; p<0.001), a 20% relative reduction.

SELECT enrolled people with prior myocardial infarction, stroke, or peripheral arterial disease. HFpEF was not an enrollment criterion. The trial therefore supports semaglutide for cardiovascular risk reduction in that specific high-risk population; it does not establish fewer HFpEF hospitalizations, lower mortality in HFpEF, or prevention of disease in people without established cardiovascular disease. A relative risk reduction also does not reveal an individual's absolute benefit without the underlying event rate.

What remains unknown

Hard HFpEF outcomes: The available evidence supports symptomatic and functional improvement over 52 weeks. It does not yet establish whether semaglutide reduces HFpEF hospitalization or death.
Durability: Both STEP-HFpEF trials lasted one year. Available evidence does not show whether benefits persist longer, plateau, or reverse after treatment stops.
Drug-specific precision: The available meta-analyses pool different GLP-1 drugs and, in some analyses, different heart-failure phenotypes. Their class-level results should not be presented as an exact semaglutide effect in HFpEF.
Comparisons with tirzepatide: No head-to-head randomized trial establishes whether tirzepatide's greater average weight loss translates into better HFpEF outcomes.

Safety changes the calculation for prevention-minded users

Semaglutide also has risks, and the relevant balance changes with the indication. A 2025 systematic review and meta-analysis of ocular adverse events identified an association between semaglutide use and non-arteritic anterior ischemic optic neuropathy (NAION), a potentially vision-threatening optic-nerve injury. The current evidence does not define the absolute risk, dose-response relationship, or which patients face the greatest susceptibility. The signal has not been incorporated into prescribing guidelines, but it deserves attention.

For someone with symptomatic obesity-related HFpEF, a treatment with demonstrated functional benefit may justify accepting uncertain or uncommon risks under clinical supervision. For a cardiovascular-risk-free person taking semaglutide solely to improve CRP, NT-proBNP, or a wearable dashboard, the equation is different. No randomized trial cited here shows hard cardiovascular benefit in that population. Biomarker movement in people with disease does not prove prevention in healthy people, and a possible irreversible eye injury is not a decorative footnote.

The practical verdict

Semaglutide 2.4 mg weekly is a credible option for patients with established HFpEF and obesity, particularly when symptoms and physical limitation remain substantial. The randomized STEP-HFpEF program tested that exact phenotype for 52 weeks, while class-level meta-analysis supports the direction of benefit in obese chronic heart failure.

Outside that group, confidence should fall fast. Non-obese HFpEF lacks demonstrated heart-failure-risk benefit in the available meta-analytic evidence. SELECT cannot be repurposed as proof of benefit for people without established atherosclerotic cardiovascular disease. If the goal is better cardiac biomarkers rather than treatment of diagnosed disease, semaglutide is being asked to do work the evidence has not shown it can do.