Ageing and Episodic Memory: Combining Neuropsychological and Event-Related Potential Approaches to Investigate Strategic Retrieval

Abstract

This thesis investigates the effect of normal ageing on the strategies adopted during episodic memory retrieval, using a combination of neuropsychological profiling and neuroimaging data measured during performance on a source memory exclusion task. The exclusion task is a type of source memory task where participants distinguish between targets (studied items from one source e.g. female voice), non-targets (studied items from another source e.g. male voice) and new items. Unlike a source memory task where three separate buttons are pressed for each item at test, in the exclusion task one button is pressed for targets and a second for non-target and new items. As this task is more complex than a normal source memory paradigm and also allows participants to perform the task in more than one way, it places high emphasis on the use of strategies to facilitate retrieval and is therefore ideal for investigating strategic retrieval. Previous source memory studies have shown that while older adults are reasonably good at recognising whether items are old or new, they show marked impairments at remembering the source in which items were presented at study. Dual process theories propose that the age-related decline in source memory occurs because recollection becomes impaired with ageing whereas familiarity remains relatively spared. The results reported in this thesis support dual process theory. Experiment 2a showed that, behaviourally, as expected, the young outperformed the elderly. Event-related potentials (ERPs), recorded while a source memory exclusion test was performed, revealed that both young and older adults showed bilateral frontal and left parietal old/new effects, thought to index familiarity and recollection respectively. Importantly, the magnitude of the left parietal effect was significantly reduced in the older adults.

The ERP findings also suggested that dual process theories represent an oversimplification of episodic memory decline with age. In Experiment 1a, three temporally and topographically distinct late frontal old/new effects were present in the younger adults: a bilateral anterior frontal effect (450-900ms post stimulus), a right prefrontal effect (900-1300ms) and a right frontal effect (1300-2000ms). Significant positive correlations between the magnitude of these effects and performance on neuropsychological tests of executive functioning in Experiment 1b, revealed that the bilateral anterior frontal effect was related to working memory, strategy use and planning; the right prefrontal effect was related to working memory and planning while the right frontal effect was related to planning. By contrast, the older adults in Experiment 2a only produced the right frontal effect, which correlated with planning across all three time windows in Experiment 2c. Post-retrieval monitoring in older adults therefore appeared to be qualitatively different than their younger counterparts. Performance on the neuropsychological tests in Experiment 2b, revealed that the older adults’ working memory and strategy use was impaired compared to the young, whereas planning was relatively intact, suggesting that age-related differences in post retrieval processing may be due to reduced executive functioning in older adults. Identifying distinct late frontal effects and demonstrating a relationship between these effects and specific executive functions is a novel finding.

The presence of a left parietal target greater than non-target difference in the young adults from Experiment 1a and 2a was interpreted as the young reducing recollection of irrelevant non-target information. The modulation did not differ in magnitude for targets and non-targets in the elderly adults from Experiment 2a, suggesting they were less able to reduce activation of goal irrelevant non-target information. The results in the young adults from Experiment 1a also highlight the importance of considering the context of source information on the processes engaged at retrieval. The bilateral frontal effect was significant for the retrieval of the intrinsic context (source information inherent to the studied item), but not the extrinsic context (source information not inherent to the studied item). This finding was interpreted within a unitisation framework, where the intrinsic context became unitised with the item and enhanced familiarity based remembering.

The findings also highlight that in order to fully understand post retrieval processing in both young and old adults, focus should move away from examining quantitative differences in the right frontal effect over long time periods and instead identify qualitatively distinct late frontal effects that may reflect the engagement of various executive functions over time.