Atrial Fibrillation, Heart Failure Phenotypes, and Mortality Risk in the Nationwide START Registry: A Propensity Score Matching Analysis

Background Data on atrial fibrillation (AF) and heart failure (HF) with preserved ejection fraction (HFpEF) are scarce. We investigated the association of HFpEF with all-cause mortality in AF. Methods We included 10 369 patients with AF on oral anticoagulants from the nationwide ongoing START (Survey on Anticoagulated Patients Register) registry. Patients were divided into 3 groups: (1) no HF, (2) HF with reduced EF/HF with mildly reduced EF (EF <= 50%), and HFpEF (EF >50%). Patients with HF should have had a clinical diagnosis or a history of HF hospitalization. The association between HF types and all-cause mortality was investigated by Cox proportional hazards regression analysis to estimate hazard ratio (HR) and 95% CI for each factor. The Fine-Gray model and propensity score matching were used. Results Mean age was 76.4 +/- 9.4 years and 45.8% were women. Overall, 2309 (22.2%) patients had HF, of whom 47.4% had HFpEF. During 720 +/- 576 days of follow-up (20 747 patients/year), 727 deaths occurred (3.5 per 100 patient-years). After propensity score matching, both HF with mildly reduced EF/HF with reduced EF and HFpEF were associated with all-cause mortality (HR, 1.33; P=0.037 and HR, 1.49; P=0.004). HFpEF was associated with mortality in men (HR, 1.654; P=0.001) but not in women (HR, 1.243; P=0.175). In HFpEF, age >= 75 years (HR, 2.247; P=0.003), chronic respiratory disease (HR, 2.109; P<0.001), anemia (HR, 1.482; P=0.035), paroxysmal AF (HR, 0.528; P=0.012), creatinine clearance<30 mL/min (HR, 1.791; P=0.018), direct oral anticoagulants (HR, 0.575; P=0.005), and renin-angiotensin inhibitors (HR, 0.670; P=0.033) were associated with all-cause mortality. Conclusions HFpEF is frequent in patients with AF and associated with an increased mortality, especially in men. Comorbidities and treatments associated differently with mortality according to HF phenotype.