Abstract
AIMS: To estimate the relation between physical exercise volume, type, and intensity with all-cause mortality and recurrent vascular events in patients with cardiovascular disease (CVD) and to quantify to what extent traditional cardiovascular risk factors mediate these relations. METHODS AND RESULTS: In the prospective UCC-SMART cohort (N = 8660), the associations of clinical endpoints and physical exercise volume (metabolic equivalent of task hours per week, METh/wk), type (endurance vs. endurance + resistance), and intensity (moderate vs. vigorous) were estimated using multivariable-adjusted Cox models. The proportion mediated effect (PME) through body mass index, systolic blood pressure, low-density lipoprotein cholesterol, insulin sensitivity, and systemic inflammation was assessed using structural equation models. Sixty-one percent of patients (73% male, age 61 ± 10 years, >70% receiving lipid-lowering and blood pressure-lowering medications) reported that they did not exercise. Over a median follow-up of 9.5 years [interquartile range (IQR) 5.1-14.0], 2256 deaths and 1828 recurrent vascular events occurred. The association between exercise volume had a reverse J-shape with a nadir at 29 (95% CI 24-29) METh/wk, corresponding with a HR 0.56 (95% CI 0.48-0.64) for all-cause mortality and HR 0.63 (95% CI 0.55-0.73) for recurrent vascular events compared with no exercise. Up to 38% (95% CI 24-61) of the association was mediated through the assessed risk factors of which insulin sensitivity (PME up to 12%, 95% CI 5-25) and systemic inflammation (PME up to 18%, 95% CI 9-37) were the most important. CONCLUSION: Regular physical exercise is significantly related with reduced risks of all-cause mortality and recurrent vascular events in patients with CVD. In this population with high rates of lipid-lowering and blood pressure--lowering medication use, exercise benefits were mainly mediated through systemic inflammation and insulin resistance.