Cardiorespiratory fitness versus physical activity as predictors of all-cause mortality in men

Background Although both cardiorespiratory fitness (CRF) and physical activity (PA) are associated with mortality, whether they are associated with all-cause mortality independent of each other is unclear. Methods CRF was assessed by a maximal exercise test and PA was measured by self-report in 8,171 male veterans. The predictive power of CRF and PA, along with clinical variables, was assessed for all-cause mortality during a mean (+/- SD) follow-up 8.7 (4.4) years during which there were 1,349 deaths. Results CRF was associated with mortality after adjusting for clinical variables and remained a strong predictor of mortality after further adjusting for PA (hazard ratio 0.85, 95% CI 0.83-0.87). PA was a significant predictor of mortality after controlling for clinical variables; however, the association was eliminated after further adjusting for CRF (hazard ratio 0.98, 95% CI 0.88-1.10). In CRF-stratified analysis, being active (>= 150 min/wk) was not associated with mortality within the unfit or fit categories (P > .4). However, in PA-stratified analysis, subjects categorized as fit (>= 7 metabolic equivalents [METS]) had a lower risk of mortality regardless of PA status (P < .001). Conclusions In adult men, PA was associated with mortality independent of established risk factors, but not CRF. Conversely, CRF remained a strong predictor of mortality independent of PA status and established risk factors.