The STEP-HFpEF trial results, published by Kosiborod et al. in the NEJM, address one of the most challenging intersections in medicine: heart failure with preserved ejection fraction (HFpEF) and obesity.[1]

Why this matters: HFpEF accounts for ~50% of all heart failure but has had no effective pharmacotherapy until recently. SGLT2 inhibitors showed modest benefit (EMPEROR-Preserved, DELIVER), but the effect was smaller than in HFrEF. HFpEF with obesity (the "obese HFpEF phenotype") may represent a distinct pathophysiology where adiposity-driven inflammation and hemodynamic overload are the primary drivers.

Trial design: Randomized, double-blind, placebo-controlled. N = 529 adults with HFpEF (EF >= 45%), NYHA class II-IV, BMI >= 30, and KCCQ-CSS < 90 (indicating symptoms). Semaglutide 2.4 mg weekly vs placebo. 52 weeks.

Dual primary endpoints:

• Change in KCCQ-CSS (symptom score): +16.6 vs +8.7 points (difference 7.8 points, 95% CI: 4.8-10.9, p < 0.001)
• Change in body weight: -13.3% vs -2.6% (difference -10.7%, p < 0.001)

Key secondary outcomes:

• 6-minute walk distance: +21.5 m vs +1.2 m (difference 20.3 m, p < 0.001)
• hsCRP: -43.5% vs -7.3% (p < 0.001)
• NT-proBNP: -20.9% vs -5.3% (p < 0.001)
• Hierarchical composite (death, HF events, KCCQ, 6MWD): win ratio 1.72 (95% CI: 1.37-2.15)

A 7.8-point improvement in KCCQ-CSS is clinically meaningful (threshold typically 5 points). The 20-meter 6MWD improvement exceeds what SGLT2i achieved in HFpEF. And the inflammatory biomarker reduction suggests mechanistic engagement beyond weight loss.

These results are genuinely exciting for the HFpEF field. But I want to raise the methodological elephant: STEP-HFpEF used patient-reported outcomes (KCCQ) and functional measures (6MWD) as primary endpoints, not hard clinical events (death, hospitalization). This is appropriate for a 529-patient trial over 52 weeks — you can't power for events with that design — but it means we still don't know if semaglutide reduces HF hospitalizations or mortality in HFpEF.

The KCCQ improvement of 7.8 points is meaningful, but KCCQ is subjective and influenced by weight loss itself (less dyspnea on exertion, better mobility, improved self-image). Disentangling the direct cardiac benefit from the general benefits of weight loss is challenging with this study design.

The NT-proBNP reduction is reassuring from a cardiac standpoint — it suggests reduced ventricular wall stress, not just symptomatic improvement. But a dedicated HF outcome trial is needed.

Cardiologist in an HFpEF clinic here. The clinical improvement I've seen in my HFpEF patients who've started semaglutide is striking — and I say this as someone who's been skeptical of most HFpEF therapies.

The typical obese HFpEF patient has BMI 35-40, is deconditioned, has significant dyspnea on exertion, and is on a cocktail of diuretics, ARNi/ACEi, and SGLT2i with marginal improvement. Adding semaglutide and seeing them lose 10-15% body weight, increase exercise tolerance, reduce diuretic requirements, and improve quality of life — that's transformative.

I'm less concerned about the lack of hard outcome data because the mechanism is so biologically plausible. In obese HFpEF, excess epicardial adipose tissue, systemic inflammation, and plasma volume expansion are driving the syndrome. Semaglutide addresses all three — weight reduction decreases cardiac loading, anti-inflammatory effects reduce epicardial fat inflammation, and the natriuretic effect reduces volume overload.

The STEP-HFpEF DM trial (companion trial in HFpEF patients with diabetes) showed similar results[2]:

• KCCQ-CSS improvement: +13.7 vs +6.4 (difference 7.3 points, p < 0.001)
• Weight loss: -9.8% vs -3.4%
• 6MWD: +14.3 m vs -0.3 m

The consistency across diabetic and non-diabetic HFpEF populations strengthens the evidence. Interestingly, the weight loss was slightly less in the diabetes cohort (as expected — semaglutide typically produces less weight loss in T2D), but the KCCQ improvement was nearly identical. This argues that the cardiac benefit is not purely proportional to weight loss.

A pooled analysis of both STEP-HFpEF trials (n = 1,145) showed a reduction in the hierarchical composite of death and HF events with a win ratio of 1.58 (95% CI: 1.29-1.94). While not powered for individual hard endpoints, this is a strong signal.

Does this mean semaglutide should be used in all HFpEF patients who are obese? My mother has HFpEF with an EF of 55% and BMI 34. She's on sacubitril-valsartan and empagliflozin. Should I ask her cardiologist about adding semaglutide?