Age-related Differences in the Neural Mechanisms of Episodic Memory: Representational and Network Analyses

Advanced age is associated with substantial changes in the brain. These changes can be attributed to many difference sources, such as detrimental effects of aging, brain’s compensatory responses to such negative effects, and cognitive or neural resources acquired over lifespan. As a result, under the same cognitive task, healthy older adults (OAs) often show recruitment of brain regions that are different from healthy young adults (YAs). These observations have been drawn from functional magnetic resonance imaging (fMRI) studies on aging and cognition, which have been largely based on univariate analysis that relates experimental conditions to activity level in individual brain region. While univariate analysis reveals the age differences in the recruitment of brain regions, much remains unknown regarding how these regions are playing their roles. Meanwhile, recent methodological advances in cognitive neuroscience have provided the opportunities to examine 1) functional communications across brain regions, and 2) information stored in the distributed neural representation in a region. In this dissertation, I described age-related differences in these two novel perspectives in a series of fMRI studies on episodic memory, a domain of cognition that is particularly affected by aging. In these studies, healthy YAs and OAs encoded and later retrieved images of scenes or objects inside the scanner. Analyses on functional brain network and neural representations were conducted on the neuroimaging data. These analyses revealed three main findings. First, neural representation and functional connectivity revealed reduced involvements of the core task regions in OAs. During encoding, early visual cortex (EVC) in OAs exhibited reduced representation of visual information. During retrieval, medial temporal lobe (MTL) in OAs exhibited reduced reconfiguration of functional connectivity associated with successful remembering. Second, enhanced recruitments of additional neural resources in OAs were also observed. During encoding, anterior temporal lobe (ATL) in OAs exhibited enhanced semantic representation. During retrieval, prefrontal cortex (PFC) in OAs showed enhanced functional connectivity and stronger reconfiguration of connectivity associated with successful remembering. Finally, we found that schematic knowledge affected functional communication in PFC and semantic representation in ATL differently in the two age groups, suggesting that schema-related strategies may be preferentially selected by OAs. Taken together, these studies depicted the detrimental effect of aging and brain’s adaptive changes in two novel perspectives: functional communication and information processing, which may contribute to a more comprehensive understanding of episodic memory function in aging populations.