AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Antecedent Chronic Kidney Disease and Cardiovascular Mortality in 300,000 Incident HFpEF Cases in the United States
By Karnib M, Selvaganesan P, Helmy I et al.
Background: HFpEF has been increasingly viewed as a disorder in which kidney dysfunction may play a central role. However, the distribution of incident HFpEF, across the full spectrum of antecedent kidney function remains poorly characterized, particularly among patients with intact renal function. The prognostic significant of antecedent CKD severity for cardiovascular (CV) mortality after incident HFpEF also remains uncertain. Research Question: In a contemporary incident HFpEF cohort, what is the distribution of HFpEF across antecedent KDIGO based CKD stages including those with intact renal function, and is antecedent CKD severity associated with subsequent 5-year CV mortality? Methods: This cohort included U.S. Veterans Affairs patients with incident heart failure (HF) diagnosed between January 1, 2000, and June 30, 2025. HFpEF diagnosis was ascertained using natural language processing-derived LVEF > 50% closest to the incident HF date and a validated HFpEF algorithm incorporating loop diuretic use, elevated natriuretic peptide levels, and HFpEF-ABA score. Antecedent kidney function was classified using KDIGO criteria based on 2 eGFR measurements obtained ≥90 days apart before incident HFpEF. Patients were grouped as normal kidney function (NKF) (eGFR ≥ 90 ml/min/1.73 m ² and no urine albumin-to-creatinine ratio [UACR] > 30 mg/g), CKD stage II (eGFR 60-89); and CKD stage III or > (eGFR <60). Outcomes were 5-year all-cause and CV mortality. Fine-Gray model treated non-CV death as a competing event. Results: Among 332,497 patients with incident HFpEF, 198,637 had stable antecedent kidney function and were classified as NKF (20.5%), CKD stage II (24.0%) and CKD stage III or > (55.5%). Advanced CKD was associated with older age and higher prevalence of hypertension, diabetes, CAD, whereas obesity, chronic liver and lung diseases were more frequent in the NKF group. 5-year all-cause mortality increased from 44.4% in NKF to 46.4% in CKD stage II and 63.5% in CKD stage III or > with crude mortality rates per 100 person-years of 13.6, 14.5, and 24.3. In adjusted Cox models, hazard ratios (HR) for all-cause mortality were 0.89 for CKD stage II and 1.24 for CKD stage III or > compared to NKF. CV deaths accounted for 28.1%, 36.0% and 37.8% of all deaths in the NKF, CKD stage II and CKD stage III or > groups, respectively, with corresponding adjusted subdistribution HRs for CV mortality of 1.34 and 2.02 versus NKF. Non-CV death predominated across groups. Conclusion: In the largest nationwide contemporary study of >300,000 rigorously phenotyped patients with incident HFpEF, over one fifth had “renal-intact HFpEF” (eGFR > 90ml/min/1.73 m2 & UACR <30 mg/g), suggesting that a substantial proportion of HFpEF developed in the absence of antecedent kidney dysfunction. Furthermore, worsening antecedent kidney function was associated with a marked stepwise increase in CV mortality following HFpEF diagnosis.