Background: Physical activity has been associated with a reduced risk of cognitive decline but the nature of this association remains obscure.
Objective: To study associations between midlife physical activity and cognition in old age for a prospective cohort of Finnish twins.
Methods: Physical activity in the Finnish Twin Cohort was assessed using questionnaire responses collected in 1975 and 1981. After a mean follow-up of 25.1 years, the subjects' (n = 3050; mean age 74.2; range 66-97) cognition was evaluated with a validated telephone interview. Both participation in vigorous physical activity, and the volume of physical activity, divided into quintiles, were used as predictors of cognitive impairment. Metrics collected by TELE were used to categorize participants as: cognitively impaired, suffering mild cognitive impairment, or cognitively healthy.
Results: Participation in vigorous physical activity compared to non-participation for both 1975 and 1981 was associated with a lower risk of cognitive impairment in individual-based analyses (fully adjusted OR 0.50, 95% CI 0.35-0.73). Pairwise analyses yielded similar but statistically non-significant associations. In terms of the volume of physical activity, the most active quintile of individuals (OR 0.69, 95% CI 0.46-1.04) had a reduced risk of cognitive decline compared with the most sedentary quintile in the fully adjusted model although no clear dose-response was found.
Conclusion: Vigorous midlife physical activity was associated with less cognitive impairment but without a clear dose-response association between the volume of physical activity and cognition.
Keywords: Cohort studies; cognition; dementia; exercise; genetics.