Long-Term Aspirin Use and Self-Reported Walking Speed in Older Men: The Physicians' Health Study

BACKGROUND: Mobility limitation is a component of frailty that shares a bidirectional relationship with cardiovascular disease (CVD). Data are limited on the role of established CVD prevention therapies, such as aspirin, for prevention of frailty and mobility limitation. OBJECTIVES: Examine the association between long-term aspirin use and walking speed. DESIGN, SETTING, PARTICIPANTS: Prospective cohort of 14,315 men who participated in the Physicians' Health Study I, a completed randomized controlled trial of aspirin (1982-1988), with extended post-trial follow-up. MEASUREMENTS: Annual questionnaires collected data on aspirin use, lifestyle and other factors. Average annual aspirin use was categorized for each participant: <= 60 days/year and >60 days/year. Mobility was defined according to self-reported walking pace, categorized as: don't walk regularly (reference), easy/casual <2mph, normal >= 2-2.9mph, or brisk/very brisk >= 3mph. Propensity scoring balanced covariates between aspirin categories. Multinomial logistic regression models estimated odds of being in each self-reported walking category. RESULTS: Mean age was 70 +/- 8 years; mean aspirin use 11 years. There were 2,056 (14.3%) participants who reported aspirin use <= 60 days/ year. Aspirin use >60 days/year was associated with drinking alcohol, smoking, hypertension, heart disease and stroke, while <= 60 days/year was associated with anticoagulation use and bleeding history. In all, 13% reported not walking regularly, 12% walked <2 mph, 44% walked >= 2-2.9 mph, and 31% walked >= 3 mph. After propensity score adjustment, regular aspirin use was associated with a faster walking speed. Odds ratios (95% confidence intervals) were 1.16 (0.97 to 1.39), 1.24 (1.08 to 1.43), and 1.40 (1.21 to 1.63) for <2 mph, >= 2-2.9 mph and >= 3 mph, respectively, compared to not walking regularly (p-trend<0.001). CONCLUSIONS: In this cohort of older men, long-term aspirin use is associated with a greater probability of faster walking speed later in life.