Leisure Time Physical Activities' Association With Cognition and Dementia: A 19 Years' Life Course Study.

Introduction: Cognitive impairment is one of the main disabilities in dementia. Physical activity (PA) has been suggested as protective for dementia. However, the findings are disparate in studies, and the question of whether this is because of reverse causality is still open. We aimed to explore the association of PA with cognition in people who later developed dementia compared to those who did not. Method: Since 2001, 11,512 (55% women) participants over the age of 50 years had taken at least one cognitive test in the Tromsø Study. Of these, 1,123 (58% women) later developed dementia. The cases were extracted from hospital journals and entered into an endpoint registry. Leisure time PA (LTPA) was self-reported. Multilevel mixed-effects linear regression was used to address whether LTPA was associated with cognition, stratified by those later developing dementia, and dementia-free in a separate analysis. Results: Leisure time PA was associated with scores in cognitive tests that were 55% (z-score 0.14) higher in those who did not develop dementia. For those in a preclinical phase of dementia, there was no association with LTPA on global cognitive scores. However, in a multifactorial test on processing speed and memory, women had a positive association with processing speed and memory. Conclusion: Leisure time PA had a positive association with global cognition function only for those who did not develop dementia. In women who were developing dementia, LTPA had a positive association with processing speed and memory, while in men, there were no such associations.