In 8,550 propensity-matched pairs of adults with metabolic dysfunction-associated steatotic liver disease (MASLD) and heart failure, those starting semaglutide, liraglutide or dulaglutide had lower five-year mortality than those starting empagliflozin or dapagliflozin: 7.0% versus 10.1%, hazard ratio 0.607. Coded acute heart failure diagnoses were roughly halved (14.2% versus 29.5%, HR 0.398), with the strongest signal in days 1-90 (HR 0.262). Results held in type 2 diabetes-only pairs. Acute pancreatitis showed no advantage.
Treat the size of this gap with suspicion. SGLT2 inhibitors are guideline-backed, trial-proven drugs for heart failure, and large randomized trials show real reductions in heart failure hospitalization. A near-60% relative difference favoring GLP-1 drugs, concentrated in the first 90 days, is biologically implausible and typical of residual confounding: prescribers may steer sicker, frailer patients toward one class, and TriNetX records diagnosis codes rather than adjudicated events. The authors themselves flag non-proportional hazards and an unverified new-user design. Outcome capture also differs when coded acute HF diagnoses can reflect care-seeking patterns rather than true incidence.
The finding is hypothesis-generating at best and should not prompt anyone to swap an SGLT2 inhibitor for a GLP-1 drug. Randomized data in heart failure with preserved ejection fraction and obesity (STEP-HFpEF, SUMMIT) support GLP-1-based therapy for symptoms, but they do not show it outperforming SGLT2 inhibitors. These drug classes are often complementary, so patients with MASLD, diabetes and heart failure should ask their clinicians whether combination therapy fits them. Incremental and confounding-prone, this adds little causal certainty.