SGLT2 inhibitors for heart failure outcomes across baseline glycemic status: a systematic review and stratified meta-analysis of independent parent trial populations

Abstract

Sodium-glucose cotransporter 2 (SGLT2) inhibitors improve heart failure outcomes, but whether relative effects are consistent across normoglycemia, prediabetes, and diabetes/type 2 diabetes remains uncertain. Linked publications from the same randomized populations may also be double counted in publication-based syntheses. We conducted a PROSPERO-registered systematic review and stratified meta-analysis. PubMed, Embase, Web of Science Core Collection, Cochrane CENTRAL, ClinicalTrials.gov, and supplementary sources were searched from inception to June 1, 2026. Independent parent trial populations were the quantitative analysis units. Trial-reported hazard ratios were pooled with random-effects REML models, and source-defined glycemic strata were analyzed separately. Source-reported interaction tests were interpreted without estimating a new cross-trial interaction. Modified Knapp–Hartung sensitivity analysis assessed the five Wald-type models with only three or four contributing trials. Seventeen eligible reports were mapped to four independent chronic heart failure trials—DAPA-HF, EMPEROR-Reduced, EMPEROR-Preserved, and DELIVER—with 20,725 randomized participants. Dapagliflozin or empagliflozin reduced the trial-defined primary composite outcome (HR, 0.78; 95% CI, 0.73 to 0.83); the small-k sensitivity remained conclusive (95% CI, 0.70 to 0.86). Pooled estimates were HR, 0.72 (95% CI, 0.57 to 0.90) for normoglycemia or normal HbA1c, HR, 0.81 (95% CI, 0.70 to 0.95) for prediabetes, and HR, 0.77 (95% CI, 0.71 to 0.84) for diabetes/type 2 diabetes. Source-reported interaction tests showed no clear effect modification but did not establish equivalence. Strict hospitalization for heart failure was reduced (HR, 0.72; 95% CI, 0.66 to 0.78). Small-k sensitivities for cardiovascular death, all-cause death, and volume depletion were less conclusive; the renal outcome remained exploratory. Across four parent trials of dapagliflozin or empagliflozin, the primary composite and strict hospitalization for heart failure improved, with directionally consistent relative effects across source-defined glycemic strata. These subgroup findings do not prove equivalent efficacy or exclude treatment-effect modification. Small-k sensitivity supported the primary finding while increasing uncertainty for selected secondary and safety outcomes. PROSPERO CRD420261418805.