Differential Associations of Cardiac, Pulmonary, Arterial, and Muscle Physiological Parameters and Biomarkers With the Incidence of HFpEF and HFrEF.

Background: Postulated differences in heart failure with preserved ejection fraction (HFpEF) compared with heart failure with reduced ejection fraction (HFrEF) include greater contributions of noncardiac dysfunction to HFpEF.

Objectives: This study aims to evaluate associations of cardiac and noncardiac physiological measures with prevalent and incident HFpEF and HFrEF.

Methods: Among 5,484 ARIC (Atherosclerosis Risk In Communities) study participants attending the fifth visit (2011-2013), this study assessed associations of cardiac structure and function by echocardiogram, pulmonary function by spirometry, arterial stiffness by pulse wave velocity, muscle strength by handgrip, fat mass by bioimpedance, inflammation by plasma biomarkers, and renal function with prevalent and incident adjudicated HFpEF (ejection fraction [EF] ≥50%) and HFrEF (EF <50%) by using adjusted logistic and Cox models.

Results: Mean age was 75 ± 5 years, 59% were women, left ventricular (LV) EF was 65% ± 7%, 246 patients had prevalent HFpEF, and 81 patients had prevalent HFrEF. Worse LV and right ventricular systolic function and larger LV size were associated more strongly with prevalent HFrEF; a higher body mass index (BMI) was associated more strongly with prevalent HFpEF. Among heart failure-free participants, 220 incident HFpEF and 187 HFrEF events occurred over a median 7 years (range: 6-8 years) of follow-up. Worse LV systolic function and larger LV size were more strongly associated with incident HFrEF, whereas higher pulmonary artery systolic pressure demonstrated a greater association with incident HFpEF. Most noncardiac dysfunctions, including greater BMI, fat mass, and systemic inflammation, showed similar magnitudes of association with incident HFpEF and HFrEF.

Conclusions: Among older adults, subclinical LV systolic dysfunction and remodeling differentially predicted the risk of incident HFrEF, whereas diastolic and most noncardiac dysfunctions were associated similarly with both incident HFpEF and HFrEF.