Pooled analysis suggests semaglutide cuts heart failure events

Researchers pooled individual participant data from four randomized, placebo-controlled semaglutide trials and found that people with heart failure with mildly reduced or preserved ejection fraction (HFpEF) had fewer cardiovascular deaths or worsening heart failure events when assigned to semaglutide instead of placebo ^[1]. The analysis was published in 2024 and funded by Novo Nordisk, which makes semaglutide ^[1].

The four trials were SELECT, FLOW, STEP-HFpEF and STEP-HFpEF DM ^[1]. SELECT enrolled people with atherosclerotic cardiovascular disease plus overweight or obesity, FLOW enrolled people with type 2 diabetes and chronic kidney disease, and the two STEP-HFpEF trials enrolled people with obesity-related HFpEF ^[1]. All used once-weekly injected semaglutide: 2.4 mg in SELECT and the two STEP-HFpEF trials, and 1.0 mg in FLOW ^[1].

What the numbers show

Of 22,282 participants across the four trials, 3,743 (16.8%) had a history of HFpEF — 1,914 assigned to semaglutide and 1,829 to placebo ^[1]. In that group, the combined endpoint of cardiovascular death or a first worsening heart failure event (defined as a hospitalization or an urgent visit for heart failure) occurred in 103 people (5.4%) on semaglutide versus 138 (7.5%) on placebo, a hazard ratio of 0.69 (95% CI 0.53–0.89; p=0.0045) ^[1]. That hazard ratio is the source of the roughly 31% relative reduction.

The effect was driven by worsening heart failure events, which occurred in 54 semaglutide participants (2.8%) versus 86 on placebo (4.7%), hazard ratio 0.59 (0.41–0.82; p=0.0019) ^[1]. Cardiovascular death alone was not significantly reduced: 59 deaths (3.1%) with semaglutide versus 67 (3.7%) with placebo, hazard ratio 0.82 (0.57–1.16; p=0.25) ^[1].

Serious adverse events were less common in the semaglutide group than the placebo group: 572 participants (29.9%) versus 708 (38.7%) ^[1]. In trials of people with chronic illness, that pattern often reflects fewer events related to the underlying disease, though the abstract does not break down the categories ^[1].

Important limits

This was a post-hoc pooled analysis, meaning the question was asked after the trials finished rather than planned in advance as the main test ^[1]. Only the two STEP-HFpEF trials were designed around HFpEF; participants from SELECT and FLOW were included based on an investigator-reported history of HFpEF, not a uniform diagnostic protocol ^[1]. The earlier STEP-HFpEF trials had shown improvements in heart failure symptoms and physical limitations, but whether semaglutide reduced clinical heart failure events in this group had not been established ^[1].

The analysis also mixes two different doses — 2.4 mg and 1.0 mg weekly — and does not, in the abstract, report results separately by dose ^[1]. All four trials used injected semaglutide; oral semaglutide (Rybelsus) was not part of this analysis ^[1]. Whether this evidence changes any US prescribing label or treatment guideline is not addressed in the source.

Why it matters for patients

HFpEF is the most common type of heart failure and carries a high risk of hospitalization and death, especially among people with overweight, obesity or type 2 diabetes — the same groups often prescribed semaglutide ^[1]. The authors note that few treatment options currently exist for these patients ^[1].

In absolute terms, the difference in the combined endpoint was about 2 percentage points over the trial periods — 5.4% versus 7.5% ^[1]. That is a modest absolute change, but it points in the same direction as the symptom improvements seen in the STEP-HFpEF trials ^[1].

What this analysis does not show is an effect on cardiovascular death by itself, which did not reach statistical significance ^[1]. It also cannot tell an individual whether heart failure benefit would apply to them, since the HFpEF subgroups inside SELECT and FLOW were identified by investigator report rather than by a standardized workup ^[1].

What happens next

All four trials — registered as NCT03574597 (SELECT), NCT03819153 (FLOW), NCT04788511 and NCT04916470 (the STEP-HFpEF trials) — are complete ^[1]. The source does not describe any planned follow-up trial designed specifically to test heart failure events in HFpEF as a primary endpoint, so any next step is not yet known from this material.

Several authors report consulting fees, research grants or honoraria from Novo Nordisk and other drugmakers ^[1].