• cognitive aging;
  • leisure activity;
  • physical activity


To examine the cognitively protective effect of leisure and physical activities while accounting for prior cognitive ability, a rarely considered confounder of the previously reported associations between activity and cognitive aging.


Longitudinal cohort study.


Glostrup, Denmark.


Community-dwelling sample of adults recruited into the Glostrup 1914 Cohort (baseline N = 802). All were born in 1914 and were assessed at ages 50, 60, 70, and 80. New participants were recruited during the study to counter attrition.


On each occasion, cognitive ability was assessed using four tests, which defined a general cognitive ability score. Self-reported participation in leisure and physical activities was also collected. In general, physical activity was summarized on a 3- or 4-point scale, and leisure activity as none versus some (ages 50 and 60) or according to participation in a list of common activities (age 70). The effect of activity—leisure and physical—on the level of cognitive ability and cognitive change over time from age 60 to 80 was examined in growth curve models.


Greater activity (leisure or physical) was consistently associated with a higher level of cognitive ability. Adjusting for baseline cognitive ability (age 50) attenuated these associations, suggesting that associations between activity and cognition reported in old age are largely a consequence of preserved differentiation. A small but significant association remained between greater physical activity at age 60 or 70 and less cognitive decline.


The association between more-frequent leisure activity and less cognitive decline mainly reflects the positive cross-sectional association between activity and cognition, although the link that remains between greater physical activity and a more-successful cognitive aging trajectory is of particular relevance to those who are developing interventions.