Superior Cardiorenal Protection With Tirzepatide Over Dulaglutide in Heart Failure With Preserved Ejection Fraction: A Large‐Scale Propensity Score–Matched Real‐World Analysis

Figures

• 

  Figure 1. Patient selection flow diagram: Tirzepatide versus Dulaglutide in Heart Failure with Preserved Ejection Fraction (HFpEF).

• 

  Figure 2. Propensity score density distributions before and after 1:1 propensity score matching in patients with heart failure with preserved ejection fraction (HFpEF). The left panel shows the distribution of propensity scores in the tirzepatide cohort (purple) and dulaglutide cohort (teal) prior to matching, demonstrating substantial overlap but with noticeable shifts reflecting baseline differences. The right panel illustrates the distributions after matching (n = 24 305 per cohort), with near‐complete overlap between the two cohorts, confirming excellent post‐matching balance and successful minimization of measured confounding.

• 

  Figure 3. Love plot demonstrating covariate balance before and after 1:1 propensity score matching in patients with heart failure with preserved ejection fraction (HFpEF) initiating tirzepatide versus dulaglutide. Absolute standardized mean differences (SMDs) are shown for selected baseline characteristics. Red circles represent imbalances before matching; green triangles represent imbalances after matching. Dashed vertical line indicates the commonly accepted threshold for good balance (SMD = 0.1). Excellent balance was achieved for most covariates after propensity score matching, with minor residual differences in body mass index, body weight, HDL cholesterol, triglycerides, and left ventricular ejection fraction.

• 

  Figure 4. Forest plot of hazard ratios for clinical outcomes at 1‐year follow‐up in propensity score–matched cohorts of patients with heart failure with preserved ejection fraction (HFpEF) initiating tirzepatide versus dulaglutide (n = 24 305 per group). Hazard ratios (HRs) with 95% confidence intervals are shown, derived from Cox proportional hazards models (dulaglutide as comparison). Square points indicate statistically significant associations (log‐rank p < 0.05); circles indicate non‐significant associations. An HR < 1 favors tirzepatide.

• 

  Figure 5. Kaplan–Meier curves for 1‐year outcome‐free survival across multiple endpoints in propensity score–matched cohorts of patients with heart failure with preserved ejection fraction (HFpEF) initiating tirzepatide (blue) versus dulaglutide (orange) (n = 24 305 per group). Curves represent time‐to‐first event for all‐cause mortality, acute heart failure events, all‐cause hospitalization, myocardial infarction, stroke, new‐onset atrial fibrillation, major adverse cardiovascular events (MACE), major adverse kidney events (MAKE), acute pancreatitis, and hypoglycemia. p‐values are derived from log‐rank tests.

• 

  Figure 6. Forest plot of hazard ratios for clinical outcomes at 3‐year follow‐up in propensity score–matched cohorts of patients with heart failure with preserved ejection fraction (HFpEF) initiating tirzepatide versus dulaglutide (n = 24 305 per group). Hazard ratios (HRs) with 95% confidence intervals are shown, derived from Cox proportional hazards models (dulaglutide as comparison). Square points indicate statistically significant associations (log‐rank p < 0.05); circles indicate non‐significant associations. An HR < 1 favors tirzepatide.

• 

  Figure 7. Kaplan‐Meier curves for 3‐year outcome‐free survival across multiple endpoints in propensity score–matched cohorts of patients with heart failure with preserved ejection fraction (HFpEF) initiating tirzepatide (blue) versus dulaglutide (orange) (n = 24 305 per group). Curves represent time‐to‐first event for all‐cause mortality, acute heart failure events, all‐cause hospitalization, myocardial infarction, stroke, new‐onset atrial fibrillation, major adverse cardiovascular events (MACE), major adverse kidney events (MAKE), acute pancreatitis, and hypoglycemia. p‐values are derived from log‐rank tests.