Abstract MP75: Novel Heart Failure Stages Incorporating Cardiorespiratory Fitness are Differentially Associated with Heart Failure Risk in the Community

Abstract

Background: Heart failure (HF) progresses through stages as depicted by the AHA/ACC/HFSA guidelines. These stages advance from the presence of risk factors (stage A), to asymptomatic structural heart disease, elevated filling pressures, or abnormal biomarkers (stage B), to clinical HF (stage C). However, even among individuals classified as having the highest pre-clinical risk (stage B), the majority do not progress to clinical HF. This highlights the need to identify new risk factors to refine preclinical HF phenotypes. Impaired cardiorespiratory fitness (CRF) is a powerful predictor of HF risk but is not currently considered in HF staging. Accordingly, we hypothesized that HF substages incorporating CRF would be differentially associated with HF risk. Methods: We investigated Framingham Heart Study participants with submaximal exercise testing (up to Bruce protocol stage 2) at a routine exam. CRF was estimated based on stage 2 heart rate, weight, age, sex, treadmill grade, and velocity using the equation from Loe, et al. Due to the importance of HR for determining estimated CRF, we excluded individuals on atrioventricular nodal blocking agents. We evaluated the association of HF substages (categorized by CRF above vs. below sex-specific median) with incident HF using Cox models adjusted for age and sex. Results: Our sample included 1683 individuals, 55% women, mean age 58±8 yrs, with estimated CRF 35±3mL/kg/min in women and 40±4mL/kg/min in men. A total of 522 individuals (31%) were classified as having no HF risk factors (“healthy”; stage H), 693 (41%) as stage A, and 468 (28%) as stage B. Over a median follow up of 20 years, 99 incident HF events occurred. HF incidence was highest in individuals categorized as stage B with lower than median CRF (B2, Figure 1A). Stage B participants with estimated CRF above the median did not have a significantly higher risk of HF compared to stage H participants, but those with stage B and low CRF had an over 2.5-fold higher risk of HF (hazard ratio [HR] 2.75, 95% CI 1.50-5.07, P<0.01; Figure 1B). Among stage B participants, every 1-SD higher CRF was associated with lower HF risk (HR 0.57, 95% CI 0.43-0.76, P<0.001). We did not observe an association of CRF with incident HF among individuals categorized as stage A. Conclusion: Incorporating CRF into HF staging criteria may improve risk assessment and facilitate precision targeting of preventive therapies, especially among individuals classified as stage B HF.