Lawrence Berkeley National Laboratory

Objective: To assess in women whether exercise affects the estimated age-related increase in adiposity, and contrariwise, whether age affects the estimated exercise-related decrease in adiposity.Subjects and Design: Cross-sectional analyses of 64,911 female runners who provided data on their body mass index (97.6 percent), waist (91.1 percent), and chest circumferences (77.9 percent).Results: Age affected the relationships between vigorous exercise and adiposity. The decline in BMI per km/wk run was linear in 18-25 year olds (-0.023+-0.002 kg/m2 per km run) and became increasingly nonlinear (convex or upwardly concave) with age. The waist, hip and chest circumferences declined significantly with running distance across all age groups, but the declines were 52-58 percent greater in older than younger women (P<10-5). The relationships between body circumferences and running distance became increasingly convexity (upward concavity) in older women. Conversely, vigorous exercise diminished the apparent increase in adiposity with age. The rise in average BMI with age was greatest in women who ran less than 8 km/week (0.065+-0.005 kg/m2 per y), intermediate of women who ran 8-16 km/wk (0.025+-0.004kg/m2 per y) or 16-32 km/wk (0.022+-0.003 kg/m2 per y), and least in those who averaged over 32 km/wk (0.017+-0.001 kg/m2 per y). Before age 45, waist circumference rose 0.055+-0.026 cm in for those who ran 0-8 km/wk, showed no significant change for those who ran 8-40 km./wk, and declined -0.057+-0.012 and -0.069+-0.014 cm per year in those who ran 40 -56 and over 56 km/wk. The rise in hip and chest circumferences with age were significantly greater in women who ran under eight km/wk than longer distance runners for hip (0.231+-0.018 vs 0.136+-0.004 cm/year) and chest circumferences (0.137+-0.013 vs 0.053+-0.003 cm/year).Conclusions: These cross-sectional associations suggest that in women, age and vigorous exercise interact with each other in affecting adiposity. The extent that these cross-sectional associations are causally related to vigorous exercise or are the consequence of self-selection remains to be determined.