RELATIONSHIP BETWEEN SELF-REPORTED PHYSICAL ACTIVITY AND DIRECTLY MEASURED PHYSICAL PERFORMANCE IN OLDER ADULTS 1373

Directly measured physical performance is predictive of older individuals' level of disability and capacity to function independently. To understand the role of leisure time physical activity (LTPA) on the variability in physical performance among the elderly, cross-sectional, baseline relationships of LTPA to walking speed, balance, and overall physical function score were examined in 1,246 women and 846 men 55 years and older (mean age=70.4) participating in a longitudinal, community-based study of aging and physical function. Self-reported LTPA (METs/wk) was assessed by multiplying frequency of participation in each of 13 vigorous and 9 moderate intensity recreational activities by MET value and summing over all activities. Performance variables were measured directly: walking speed by distance covered in 1 minute, balance by a 3-level categorical score based on 4 balance tasks and overall physical function by a composite score based on a series of tasks designed to assess upper body strength and dexterity as well as balance and speed. Covariates included age, percentage of body fat as measured by bioelectrical impedance, and self-reported smoking status and health status (presence/absence of specified chronic diseases). Among both men and women, age-adjusted partial correlation coefficients showed a positive association between walking speed(r=.16 and.18 respectively, p<.001) and function score (r=.08, p<.05 for the men and r=.12, p<.001 for the women). A higher level of LTPA was also significantly associated with a slightly reduced risk of poor balance(OR=0.99, 95% CI=.98-.99). In women, these associations persisted with adjustment for other covariates, but in men, only the association with walking speed remained. These findings suggest that LTPA may contribute to maintenance of functional independence in older adults