Effects of Age on Neural Correlates of Episodic Encoding and Brain Structure, and Their Relation to Cognitive Performance

Episodic memory – memory for unique personal events – is essential to our daily life. Relative to other forms of memory, episodic memory declines disproportionately with advancing age. One prominent account of such decline proposes a reduction in the efficacy of episodic encoding in older individuals. Numerous studies have employed functional magnetic resonance imaging to investigate the neural correlates of episodic encoding in young and older adults with the “subsequent memory procedure”. With this procedure, encoding related neural activity is contrasted based on subsequent memory performance for the study items. These studies have consistently reported that neural activity during encoding is predictive of later memory performance. Such subsequent memory effects (SMEs) take two forms: positive SMEs, where enhanced neural activity is associated with study items later remembered relative to study items that are less well remembered or forgotten; and negative SMEs, that take the opposite pattern. Studies have generally reported age-invariant positive SMEs whereas negative effects tend to be attenuated in older adults. Of importance, neural activity preceding the onset of a study item has also been shown to predict subsequent memory. Few studies have examined the effect of age on such pre-stimulus subsequent memory effects (pre-stimulus SMEs). Experiment 1 (Chapter 2) describes findings on pre-stimulus neural activity in healthy young and older adults. The results revealed age-invariant and age-dependent pre-stimulus SMEs in different brain regions, although age differences were mostly quantitative rather than qualitative. In contrast to prior reports of pre-stimulus SMEs, the effects in the present study were negative in direction. They could reflect allocation of neural resources in preparation of the upcoming study event. The study reported in Chapter 3 combined data from 2 independent experiments to examine age differences in poststimulus SMEs. The 2 regions of a priori interest were the hippocampus and left inferior frontal gyrus (IFG). Positive and negative SMEs were evident in both age groups. Of importance, the hippocampal SMEs were equivalent across age groups; and there was no evidence of age-related right-frontal over-recruitment. There was an age-invariant relationship between hippocampal SMEs and memory performance, suggesting intact hippocampal encoding activity in healthy older adults, and consistent with the notion that hippocampal activity reflects the amount of information encoded. A positive relationship between left IFG SME and memory performance was observed in older adults only. The study in Chapter 4 took an integrated approach to examine the relationship between structural and functional measures, and memory performance in young and older adults. Consistent with the literature, robust age-related decline was evident in hippocampal volume and cortical thickness. Results from an integrated statistical model revealed that hippocampal encoding activity, but not hippocampal volume, was predictive of memory performance in both age groups. On the other hand, cortical thickness negatively correlated with performance in young adults, but positively correlated with performance in older adults. Both cortical thickness and cortical SMEs explained unique variance in memory performance. Of importance, IFG thickness-memory relationships were no longer significant after controlling for global thickness. In conclusion, both pre-stimulus and encoding-related neural activity can be resistant to the effects of age, although the left IFG acts as a ‘bottleneck’ in older adults. Age differences in pre-stimulus SMEs require a nuanced interpretation, rather than appeal to a generic construct. Moreover, age differences appear to be more robust in structural rather than in functional measures. Lastly, the age-dependent cortical thickness-memory relationship was general rather than region-specific.