Device-based measures of physical activity are associated with cognitive function but do not inoculate against later cognitive decline

Physical activity and sleep are thought to be key protective factors against cognitive decline. However, longitudinal studies with objective measures of physical activity and sleep are still missing. 719 older adults from the Survey of Health, Ageing and Retirement in Europe (SHARE) wore accelerometers over an 8-day period. Cognitive function was measured at baseline and 26 months later. In unadjusted analyses, steps per day and intensity minutes per day at baseline were associated both contemporaneously and longitudinally with cognitive function (betas = 0.22 and 0.23, both p  < 0.001). However, in longitudinal analyses that controlled for age, education, and cognitive function at the prior wave, physical activity was no longer significantly associated with cognitive function (beta = 0.04, p  = 0.131). Sleep metrics of sleep duration and sleep efficiency were not significantly associated with cognitive function (all betas < 0.21, all p values  > 0.090). Alongside other interpretations such as reverse causality, the results are consistent with a concurrent effect of physical activity but not with inoculation effects against cognitive decline 26 months later.