HFpEF: what the GLP-1 trials show (semaglutide, tirzepatide)

Complete guide: GLP-1 drugs: a complete guide to semaglutide and beyond

What is heart failure with preserved ejection fraction?

Heart failure with preserved ejection fraction (HFpEF; ICD-10 I50.3 for the diastolic form) is a syndrome in which the heart pumps with a normal or near-normal ejection fraction but fills poorly and intracardiac pressures rise. The universal definition recognizes it by symptoms or signs of heart failure, a left ventricular ejection fraction ≥50%, and objective evidence of elevated filling pressures, for example natriuretic peptides or hemodynamic testing³.

It is not rare. In a study of 2,802 patients admitted with heart failure at 103 Ontario hospitals, 31% had an ejection fraction above 50%; they were older, more often women, with hypertension and atrial fibrillation, and their one-year mortality (22%) was not significantly different from that of reduced ejection fraction (26%)². Obesity is one of its drivers: in the Framingham cohort, each BMI point raised heart failure risk 5% in men and 7% in women, and obesity doubled the risk¹.

Symptoms and how it is diagnosed

Symptoms are exertional dyspnea, fatigue, exercise intolerance, edema and, later, orthopnea or admissions for congestion. Because patients are often euvolemic at the visit and ejection fraction is normal, diagnosis is difficult. Two tools help: the H2FPEF score, which adds obesity, atrial fibrillation, age >60, ≥2 antihypertensives, E/e' >9 and pulmonary artery systolic pressure >35 mm Hg (AUC 0.841; odds double with each point)⁴, and the European Society of Cardiology HFA-PEFF algorithm, which combines echocardiography and natriuretic peptides⁵.

Standard treatment today

The 2022 AHA/ACC/HFSA guideline and the 2023 European Society of Cardiology focused update recommend diuretics for congestion, treating hypertension, atrial fibrillation and sleep apnea, and an SGLT2 inhibitor as the foundation⁶⁷. Two randomized trials support it: empagliflozin lowered cardiovascular death or heart failure hospitalization from 17.1% to 13.8% (HR 0.79)⁸, and dapagliflozin from 19.5% to 16.4% (HR 0.82)⁹.

Finerenone reduced worsening events in FINEARTS-HF (rate ratio 0.84) with more hyperkalemia¹⁰. Spironolactone missed significance in TOPCAT (HR 0.89; P=0.14), although it reduced heart failure hospitalizations¹¹, and sacubitril-valsartan was borderline in PARAGON-HF (rate ratio 0.87; P=0.06)¹². A 2026 American College of Cardiology consensus also places semaglutide and tirzepatide as an option in patients with obesity, with "tempered optimism" about clinical events because there were few²².

What peptide research has shown

Semaglutide (GLP-1 receptor agonist). In STEP-HFpEF, 529 patients with preserved-ejection-fraction heart failure and BMI ≥30 received semaglutide 2.4 mg weekly or placebo for 52 weeks. The KCCQ score improved 16.6 versus 8.7 points (difference 7.8; 95% CI 4.8–10.9), weight changed −13.3% versus −2.6%, and 6-minute walk distance rose 21.5 m versus 1.2 m¹³. The pooled analysis of STEP-HFpEF and STEP-HFpEF DM (1,145 patients; 616 of them with type 2 diabetes) gave a 7.5-point KCCQ difference and −8.4% weight¹⁴, and the symptom benefit was consistent across three HbA1c levels¹⁵. These trials measured symptoms and function, not mortality.

For clinical events, a post hoc analysis of four trials (SELECT, FLOW, STEP-HFpEF and STEP-HFpEF DM) in 3,743 participants with a history of HFpEF observed cardiovascular death or a heart failure event in 5.4% with semaglutide versus 7.5% with placebo (HR 0.69); cardiovascular death alone did not change significantly (HR 0.82)¹⁷.

Tirzepatide (dual GIP and GLP-1 receptor agonist). SUMMIT enrolled 731 patients with ejection fraction ≥50% and BMI ≥30. Cardiovascular death or worsening heart failure occurred in 9.9% with tirzepatide and 15.3% with placebo (HR 0.62; 95% CI 0.41–0.95); worsening alone was 8.0% versus 14.2% (HR 0.54), and KCCQ improved 19.5 versus 12.7 points. Cardiovascular death was numerically higher with tirzepatide (2.2% versus 1.4%; HR 1.58; CI 0.52–4.83), with very few events¹⁶.

Evidence from clinical practice. Five US insurance-claims cohorts (up to 58,333 patients on semaglutide and 11,257 on tirzepatide) observed a lower risk of heart failure hospitalization or death versus sitagliptin as a placebo proxy (HR 0.58 and 0.42), with no significant difference between the two drugs (HR 0.86; CI 0.70–1.06). The design is observational¹⁸.

Natriuretic peptides as biomarkers. BNP and NT-proBNP are not treatments but diagnostic and prognostic tools, and they are part of the algorithms and of trial eligibility criteria⁵¹⁹. In 14,750 participants from four trials, higher BMI was associated with lower NT-proBNP; for the same risk (5 events per 100 person-years), NT-proBNP in participants without atrial fibrillation was nearly 3-fold lower at BMI ≥35 (158 pg/mL) than at BMI <35 (450 pg/mL)¹⁹. And HFpEF with normal natriuretic peptides is also associated with higher morbidity and mortality²⁰. A normal value does not rule it out, especially in obesity.

Safety and contraindications

In SUMMIT, adverse events (mainly gastrointestinal) leading to drug discontinuation were 6.3% with tirzepatide and 1.4% with placebo¹⁶. In the four-trial semaglutide pool, serious adverse events were less frequent with the drug (29.9% versus 38.7%)¹⁷. More on the class in GLP-1 beyond weight and tirzepatide and the dual GIP agonist. The key warning concerns reduced ejection fraction: in a systematic review of the FIGHT and LIVE trials (541 patients), liraglutide did not improve outcomes and was linked to a higher risk of adverse heart failure outcomes²¹. The chair of the ACC consensus committee noted that this concern does not apply to HFpEF²². Label-specific contraindications are summarized in our tirzepatide side-effects review.

Regulatory status

In the European Union, the European Medicines Agency's scientific committee recommended in September 2024 updating the Wegovy label with data on symptoms and functional capacity in obesity-related heart failure with preserved ejection fraction²⁴. In the US, Lilly announced in May 2025 that it withdrew its tirzepatide application because the FDA indicated that it required another confirmatory trial²³. Always check the current label.

Summary by agent

Frequently asked questions

Is HFpEF a "mild" form of heart failure? No. In the Ontario study, one-year mortality was 22%, similar to reduced ejection fraction (26%)².

Does a normal NT-proBNP rule out HFpEF? No. HFpEF can occur with normal natriuretic peptides, with higher morbidity and mortality²⁰, and obesity lowers these values¹⁹.

Does semaglutide improve HFpEF? In patients with obesity, STEP-HFpEF showed better symptoms, function and weight than placebo¹³, and a post hoc analysis observed fewer events¹⁷. There is no trial designed for mortality.

Did tirzepatide reduce hospitalizations? In SUMMIT, worsening heart failure was 8.0% versus 14.2% (HR 0.54)¹⁶. Few events and a population with obesity.

Do these drugs help in heart failure with reduced ejection fraction? There is no evidence in favor: FIGHT and LIVE with liraglutide showed no benefit²¹.