Lifestyle, cognitive aging, and brain correlates

Abstract

Inter-individual differences in level and rate of cognitive decline typically seen in aging have been linked to inter-individual differences in lifestyle factors such as leisure activities, including physical activity. The general aim of this thesis was to further our understanding of how and why leisure activity engagement is related to aging-related changes in cognitive performance. Specifically, we sought to (a) identify lifestyle components that are associated with late-life cognitive performance, (b) identify brain correlates of these lifestyle components that are also relevant for cognitive performance, and (c) explore the relative importance of lifestyle- and health-related factors for predicting cognitive change, as well as interactive effects among these factors.

In Study I and II, we investigated associations between 3-year changes in leisure activities and concurrent changes in cognitive performance and white matter microstructure in 563 (Study I) and 442 (Study II) participants aged 81 years and older. Study I documented changes in white matter microstructure in the corticospinal (CS) tract to be associated with changes in perceptual speed. In Study II, we observed that concurrent change in frequency of engagement in social activities (e.g. going out to eat in a restaurant, going to the movies, concerts, or the theater) was related to change in both white matter microstructure in the CS tract and in perceptual speed. Change in white matter microstructure in the CS tract statistically accounted for the association between changes in frequency of social leisure activities and perceptual speed.

In Study III, we turned to D2/3 dopamine receptor (D2/3DR) availability as a potential brain correlate of lifestyle and cognition in aging. We investigated D2/3DR availability, cognitive performance, and physical activity in 178 healthy adults aged 64-67 years. Participants completed tests of working memory, episodic memory, and processing speed, and a leisure activity questionnaire. Subjective intensity, but not frequency, across the activities each individual performed was associated with D2/3DR availability in caudate nucleus as well as with episodic and working memory. Episodic memory was also related to D2/3DR availability in the caudate, forming a correlative triad with physical activity intensity and caudate D2/3DR availability.

In Study IV, we applied a new data-mining technique called structural equation modelling trees and forests to investigate the relative importance of leisure activity engagement, physical activity, and other age- and health-related factors in predicting subsequent 6-year change in perceptual speed in 1046 participants aged 60 years and older. With regard to variable importance, a measure that subsumes main effects and interactions among predictors, frequency of leisure activities was not unimportant, although less important than age, retirement status, walking speed, and multimorbidity. Conceivably, the association between leisure activity engagement and subsequent cognitive decline is conditional upon age- and health-related factors included in the current analyses.

Regarding aim (a), identifying lifestyle components related to cognitive aging, we identified change in social activities to be related to change in perceptual speed (Study II). We also found that subjective intensity, but not frequency, of physical activity was related to episodic and working memory (Study III). Regarding the relative importance of frequency of leisure activity engagement as a predictor of change in cognition, we observed some importance of all types of activities, except for physical activity, in predicting change in perceptual speed (Study IV). Concerning aim (b), identifying brain correlates of lifestyle components and cognitive performance, we observed white matter microstructural changes to be related to changes in both leisure activity and perceptual speed (Study II), and D2/3DR availability (Study III) to be related to both subjective physical activity intensity and episodic memory. Regarding aim (c), exploring the relative importance of lifestyle components as predictors of subsequent cognitive decline (Study IV), we found rather small effects of the lifestyle components investigated in Studies II and III, but still found leisure activities to be informative as predictors when using a data-mining approach that takes interactive effects with other predictors into account.

The studies in this thesis contribute with new data on associations between lifestyle and cognitive aging, and on brain measures correlated with these two factors. Specifically, we are the first to show parallel changes in leisure activity, white matter microstructure, and perceptual speed. We are also the first to observe an association between physical activity intensity and D2/3DR availability. In sum, the present results indicate that engaging in social activities in very late life and physical activity intensity around retirement age are related to cognitive performance and associated brain parameters. Although the issue of causal directionality remains unresolved, leisure activities are correlates and informative predictors of cognitive decline.