IntroductionMobility limitation is a component of frailty that shares a bidirectional relationship with cardiovascular disease (CVD). Whether established CVD prevention therapies such as aspirin can prevent frailty and mobility limitation is unknown.MethodsProspective cohort of 14,896 participants >60 years in the Physicians’ Health Study I, a completed randomized controlled trial of aspirin (1982-1986) with extended post-trial follow-up. Annual questionnaires (1986-2001) collected data on aspirin use, lifestyle and other clinical factors. Aspirin exposure was computed from 1986-2001 when walking pace was assessed. Average aspirin use was categorized for each participant<60 d/yr, 60-180 d/yr, or >180 d/yr over 15 years. Self-assessed walking pace was categorized asdon’t walk regularly, easy casual <2mph, normal ≥2-2.9mph, or brisk or very brisk ≥3mph. Propensity scoring created covariate balance across the categories of aspirin use. Aspirin randomization group was included as a covariate in the propensity score. Multinomial logistic regression models estimated the odds of prevalent mobility limitation according to level of aspirin use.ResultsMedian age was 70 years (range 60-101); mean duration of aspirin use was 9 years (range 0-15). 15.3%, 62.1%, and 22.6% of participants reported aspirin use <60, 60-180, and ≥180 days/year, respectively. Those with greater aspirin use were more likely to drink alcohol daily, have smoked previously, and have hypertension, prevalent CVD and stroke. Lower frequency of aspirin was associated with Coumadin use and bleeding history. For walking, 12% reported not walking regularly, 12.6% walked <2 mph, 44.9% walked ≥2-2.9 mph, and 30.9% walked ≥3 mph. After propensity score adjustment, the estimated odds ratios (95% confidence intervals) of not walking regularly were 0.78 (0.69-0.90) and 0.84 (0.72-0.99) for 60-180 and >180 days/yr of aspirin use, respectively, in comparison to <60 days. In analyses considering all walking speed categories, increasing aspirin use was consistently associated with greater walking speed (p-trend=0.017).ConclusionsThese results suggest that long-term aspirin use prevents mobility limitation, defined by walking speed, possibly due to anti-inflammatory effects.