Incretin-Based Therapies and SGLT2 Inhibitors in Obesity-Associated HFpEF: A Systematic Review of Direct and BMI-Stratified Randomized Evidence

Background/Objectives: Incretin-based therapies have been tested in trials specifically enrolling patients with obesity-related heart failure with preserved ejection fraction (HFpEF), whereas much of the obesity-relevant evidence for sodium-glucose cotransporter 2 (SGLT2) inhibitors derives from body mass index (BMI) subgroups of broader heart-failure trials. We synthesized these evidence streams separately and evaluated their directness and certainty. Methods: PubMed, Web of Science Core Collection, and CENTRAL were searched from inception through 1 July 2026, with supplementary searches of ClinicalTrials.gov, reference lists, trial names, registration numbers, protocols, and appendices. Randomized trials and randomized trial-derived analyses were eligible if they provided obesity- or BMI-relevant data for health status, 6 min walk distance (6MWD), body weight, or heart-failure/cardiovascular-death outcomes. Risk of bias was evaluated with RoB 2 and certainty with GRADE. Drug classes and evidence strata were analyzed separately. Results: The database searches identified 5258 records; 4461 unique records were screened, 48 reports underwent full-text assessment, and 41 reports were linked to 12 randomized trial programs. Twelve reports from 10 independent trials contributed unique quantitative estimates. In obesity-specific trials, incretin-based therapy improved KCCQ-CSS by 7.38 points (95% CI 5.51 to 9.26; I2 = 0%; moderate certainty), increased 6MWD by 17.60 m (95% CI 11.86 to 23.35; I2 = 0%; moderate certainty), and reduced weight by 9.56 percentage points (95% CI −12.71 to −6.42; I2 = 95.0%; moderate certainty). SUMMIT reported fewer cardiovascular-death or worsening-heart-failure events (HR 0.62, 95% CI 0.41 to 0.95; moderate certainty). For SGLT2 inhibitors, the primary blinded functional synthesis based on two subgroup analyses favored treatment for 6MWD (16.85 m, 95% CI 5.19 to 28.50; I2 = 11.2%; low certainty), but inference was highly imprecise under t-based Hartung–Knapp analysis. Weight effects were not pooled across evidence tiers: the exact BMI ≥ 30 kg/m2 EMPEROR-Preserved estimate was −1.48 kg (95% CI −1.84 to −1.13; low certainty), while obesity-enriched trials provided supportive but indirect estimates. In BMI ≥ 30 kg/m2 strata from EMPEROR-Preserved and DELIVER, the exploratory pooled event estimate was HR 0.80 (95% CI 0.69 to 0.91; I2 = 12.5%; low certainty), whereas the modified Hartung–Knapp 95% CI was 0.33 to 1.95. Conclusions: Obesity-specific randomized evidence supports clinically relevant improvements in symptoms, functional capacity, and weight with incretin-based therapies. SGLT2 estimates in patients with obesity or high BMI generally favored treatment but were less direct and less certain because they were frequently subgroup-derived, sometimes included participants with LVEF 41–49%, and were based on few independent trials. The findings do not establish class superiority or test a phenotype-guided treatment strategy; rather, they generate a hypothesis for phenotype-informed treatment selection.