Cardiorespiratory Fitness and Reclassification of Risk for Incidence of Heart Failure: The Veterans Exercise Testing Study.

Abstract

It is well established that cardiorespiratory fitness (CRF) is inversely associated with cardiovascular and all-cause mortality. However, little is known regarding the association between CRF and incidence of heart failure (HF). Between 1987 and 2014, we assessed CRF in 21 080 HF-free subjects (58.3±11 years) at the Veterans Affairs Medical Centers in Washington, DC, and Palo Alto, CA. Subjects were classified by age-specific quintiles of CRF. Multivariable Cox models were used to determine the association between HF incidence and clinical and exercise test variables. Reclassification characteristics of fitness relative to standard clinical risk factors were determined using the category-free net reclassification improvement and integrated discrimination improvement indices. During the follow-up (mean 12.3±7.4 years), 1902 subjects developed HF (9.0%; average annual incidence rate, 7.4 events per 1000 person-years). When CRF was considered as a binary variable (unfit/fit), low fitness was the strongest predictor of risk for HF among clinical and exercise test variables (hazard ratio, 1.91; 95% confidence interval, 1.74-2.09; P<0.001). In a fully adjusted model with the least-fit group as the reference, there was a graded and progressive reduction in risk for HF as fitness level was higher. Risks for developing HF were 36%, 41%, 67%, and 76% lower among increasing quintiles of fitness compared with the least-fit subjects (P<0.001). Adding CRF to standard risk factors resulted in a net reclassification improvement of 0.37 (P<0.001). CRF is strongly, inversely, and independently associated with the incidence of HF in veterans referred for exercise testing.