Medicine & Science in Sports & Exercise:
C-14 Thematic Poster - Incident Disease and Mortality: JUNE 2, 2011 8:00 AM - 10:00 AM: ROOM: 403
1National Cancer Institute, Rockville, MD. 2American Cancer Society, Atlanta, GA. 3Karolinska Institute, Stockholm, Sweden. 4Johns Hopkins Bloomberg School of Public Health, Baltimore, MD. 5Harvard Medical School, Boston, MA.
(No relationships reported)
Context: The dose-response relation between physical activity and mortality is not fully understood, and whether physical activity can mitigate the increased risk of death associated with an elevated body mass index (BMI) is unclear.
PURPOSE: To determine the dose-response association of physical activity with mortality and to evaluate the combined relations of physical activity and BMI with mortality.
METHODS: We pooled 6 prospective cohort studies (n=654,827, ages 20-90 years, median age=61.1 years) selected based on comparable assessments of moderate and vigorous leisure time physical activity. Physical activity was categorized by metabolic equivalent (MET) hours per week. Multivariate Cox regression models estimated the hazard ratios (HR) and 95% confidence intervals (CI). Years of life gained/lost were calculated by covariate adjusted survival curves.
RESULTS: During a median 10 years of follow-up, 82,465 deaths occurred. Physical activity had a curvilinear relation with mortality, with plateuing of risk reductions after 15.0 MET-hr/wk. Compared to 0 MET-hr/wk, the HRs were 0.80 [CI:0.78-0.82], 0.74 [0.72-0.77], 0.66 [0.65-0.68], 0.60 [0.58-0.61], and 0.58 [0.56-0.59], respectively for 0.1-3.74, 3.75-7.49, 7.5-1.49, 15.0-22.4, and 22.5 MET-hr/wk. The corresponding gains in life expectancy were 2.0, 2.8, 3.8, 4.8, and 5.0 years. In analyses restricted to never smokers with no prior heart disease or cancer (16,364 deaths), physical activity appeared to benefit persons at all BMI levels. Additionally, active (15.0+ MET-hr/wk) lean (BMI 18.5-24.9) persons had a lower risk of mortality than active obese (BMI 30+) persons. Relative to the active-lean, the HRs were 1.60 [CI: 1.48, 1.73] for active-obese persons, 1.72 [1.57, 1.88] for inactive-lean (0 MET-hr/wk) persons, and 2.20 [1.99, 2.42] for inactive-obese persons. The loss in life expectancy, as compared with the active-lean, was 3.2 years for the active-obese, 4.1 years for the inactive-lean, and 6.0 years for the inactive-obese.
CONCLUSION: Increased physical activity is associated with reduced risk of mortality, but it does not completely eliminate the elevated risk of death among the obese. To attain maximal longevity, both adopting a physically active lifestyle and attaining a normal body weight are important.