부산대학교 도서관

As populations around the world are rapidly ageing, there is an urgent need to identify modifiable determinants of brain health in late life. Cognitive and social activities represent potential candidates, particularly as they have previously been linked to a lower risk of cognitive decline and dementia. The aim of this thesis was therefore to evaluate whether cognitive and social activities associate with magnetic resonance imaging (MRI) markers of brain structure and functional connectivity and performance on neuropsychological assessments. A systematic review and meta-analysis were first conducted of previous MRI studies, to evaluate whether cognitive and social activities relate to imaging indices of brain structure. The review suggested that when combined, higher activity levels were linked to greater white matter volume, fewer white matter lesions and greater hippocampal volume. With the review largely inconclusive about the independent associations of cognitive and social activities, late life activity levels were next examined in relation to brain integrity and cognitive function. In a cross-sectional analysis of Whitehall II data, cognitive activity was the only activity to associate with cognition (global and executive function), whereas social activities were related to total grey matter volume, without any cognitive correlates. Given that a cross-sectional design is insensitive to changes in activity levels that occur over time, the Whitehall II dataset was revisited in order to evaluate whether trajectories of leisure activities relate to the ageing brain. Between mid-life and late life, a curvilinear pattern of change in activities was observed. Level and change in cognitive activities also associated with global cognition, executive function, memory and processing speed, while social activities related to executive function, only. The links with brain structure and functional connectivity, however, remained small. Considering that composite measures of leisure activities are non-specific indices, the next objective was to assess whether particular activities play a significant role in maintaining the brain in late life. Examining data from the UK Biobank study, weekly leisure-time computer use during mid-life and late life was linked to performance on multiple tests (e.g. fluid intelligence and prospective memory) with mid-life computer use further related to putamen volume. Late life educational classes demonstrated links with higher fluid intelligence, while mid-life attendance at a sports club or gym was related to sensorimotor connectivity. Overall, this thesis provides evidence to suggest that leisure activities – cognitive activities in particular – are consistently linked to late life cognition. However, associations with measurable brain structure and functional connectivity were found to be scarce.