Linking brain and cognitive plasticity in aging : a longitudinal magnetic resonance imaging study of memory training in middle and late life

Abstract

The goal of the present thesis was to study how memory training is related to structural characteristics of the adult aging brain. We studied two aspects of a novel memory training intervention: 1) is the training accompanied with measureable changes in the brain? (paper I-II), and 2) is training-benefit in adults with memory problems related to structural features of the brain––pre-intervention?<br><br>

In papers I-II, we undertook a randomized controlled magnetic resonance (MR) imaging study of an intensive two-months memory-training program in healthy adults (mean age 60.3). Assessment of cognitive performance and MR-imaging was performed before and after training. Structural changes in cerebral grey and white matter were assessed using MRprotocols optimized for reliable longitudinal analysis. Following intervention, the training group improved task-specific memory performance. A unique finding is that training was accompanied by regional increases in cerebral cortical thickness and white matter integrity compared with controls. We found significant relationships between the changes in performance and brain structure characteristics, suggesting a link between the two levels of enquiry. The current assessment-interval spanned less than three months, and follow-up studies are needed to conclude on the long-term effects of the present memory training on brain and cognition.<br><br>

Earlier research indicated positive effects of cognitive intervention for adults with memory concerns, but evidence regarding who might benefit from training was lacking. In paper III, we offered the same training program for memory clinic outpatients with subjective memory impairment, and studied predictors of training benefit. The results showed that training was feasible for this patient group, including high participation rates and low dropout rates. Regional left hippocampal volumes before training were found to predict memory-training benefit. Sub-region analysis suggested that the effects were selective to the left cornu ammonis (CA) sectors CA2/3, and CA4 and dentate gyrus, which are of known importance for episodic memory. The finding implicate that structural imaging could serve useful in future trials evaluating treatment potential and in selecting candidates for intervention.