Introduction: Recent studies have suggested that extreme levels of physical activity (endurance athletes) are associated with subclinical atherosclerosis as well as increased mortality. The safety of continuing high levels of physical activity is uncertain once coronary artery calcification (CAC) is discovered.
Hypothesis: We hypothesized that men performing &ge3000 MET·minutes/week of physical activity would have greater all-cause and cardiovascular (CV) mortality compared to those with <1500 or 1500-<3000 MET·minutes/week of physical activity and that mortality risk would be greater in those with CAC&ge100 compared to <100 Agatston units.
Methods: The cohort studied included 16,109 men without prevalent CV disease who reported physical activity levels and underwent EBT or MDCT scan. Physical activity was categorized into &ge3000 (n=1,266), 1500-3000 (n=3,027), and <1500 (n=11,816) MET·minutes/week. CAC scanning included EBT scans (1997-2007) or MDCT scans (2007-2013), and CAC score was categorized into &ge100 (n=3,547) and <100 (n=12,562) Agatston units. We fit separate proportional hazards regression models to follow-up times for all-cause and CV mortality. The models included all combinations of CAC and physical activity categories and were adjusted for baseline age, smoking, BMI, cholesterol, HDLc, and systolic blood pressure.
Results: The average age of participants at baseline was 51.3±8.3 years. Men with the highest activity level had a lower BMI and higher HDLc. After an average follow-up of 8.9 years, there were 329 all-cause and 60 CV deaths, including 174 all-cause and 38 CV deaths in those with CAC&ge100. The sample had 80% power to detect all-cause mortality hazard ratios &ge 1.9 and 1.8 for physical activity &ge3000 versus <1500 in those with CAC<100 and &ge100, respectively. The corresponding minimum detectable CV mortality hazard ratios were 3.5 and 2.8. Comparing physical activity &ge3000 to <1500 in those with CAC&ge100, the hazard ratios (95% CI) were 0.9 (0.5, 1.5) for all-cause mortality and 0.9 (0.3, 3.1) for CV mortality. Hazard ratios were similar when comparing physical activity &ge3000 to 1500-<3000 in those with CAC &ge100. Finally, when comparing physical activity categories, there was no evidence that hazard ratios varied by CAC category, p>0.7.
Conclusions: This sample offers no evidence that levels of activity &ge3000 MET·minutes/week are associated with increased all-cause or CV mortality compared to those with <1500 or 1500- <3000 MET·minutes/week, regardless of CAC level.