Neural evidence for age-related differences in representational quality and strategic retrieval processes

Abstract

Mounting behavioral evidence suggests that declines in both representational quality and controlled retrieval processes contribute to episodic memory decline with age. The present study sought neural evidence for age-related change in these factors by measuring neural differentiation during encoding of paired associates and changes in regional blood oxygenation level-dependent activity and functional connectivity during retrieval conditions that placed low (intact pairs) and high (recombined pairs) demands on controlled retrieval processes. Pattern similarity analysis revealed age-related declines in the differentiation of stimulus representations at encoding, manifesting as both reduced pattern similarity between closely related events and increased pattern similarity between distinct events. During retrieval, both groups increased recruitment of areas within the core recollection network when endorsing studied pairs, including the hippocampus and angular gyrus. In contrast, only younger adults increased recruitment of, and hippocampal connectivity with, lateral prefrontal regions during correct rejections of recombined pairs. These results provide evidence for age-related changes in representational quality and in the neural mechanisms supporting memory retrieval under conditions of high, but not low, control demand.

Keywords: Aging; Cognitive control; Dedifferentiation; Episodic memory; Recollection.

Fig. 1

Schematic depicting experimental paradigm. Participants studied trial-unique adjectives paired with one of eight images pictured. At test, participants were presented with studied (intact) and nonstudied (recombined) pairs and made an old/new judgment for each. Pictures were replaced with word labels during the test phase to reduce perceptual overlap between the study and test, minimizing the ability to rely on perceptual fluency to support performance.

Fig. 2

Cortical regions of interest (ROIs). Left panel: Age-invariant results of the contrast Objects > Scenes (yellow-red) and Scenes > Objects (green-blue) contrasts during the localizer task, thresholded at p < 0.05 FWE corrected, used as a mask for analysis of representational quality during encoding. Right panel: Coordinate-based ROIs defined by Wheeler and Buckner (2003) centered on a peak coordinate in the DLPFC (yellow; peak coordinates −47, +17, +24), VLPFC (green; peak coordinates −45, +35, −4), and angular gyrus (red; peak coordinates −45, −69, −6). Abbreviations: ROIs, regions of interest; DLPFC, dorsolateral prefrontal cortex; VLPFC, ventrolateral prefrontal cortex; FWE, family-wise error corrected. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Behavioral performance. Mean proportion of hits to intact pairs and correct rejections of recombined pairs. Error bars represent standard error of the mean. Older adults made fewer hits and correct rejections relative to younger adults, *** p < 0.001.

Fig. 4

Pattern similarity at encoding as a function of event relatedness. Top: Correlation matrix depicting the Pearson correlation between individual encoding events, sorted by category, subcategory, and then by the exemplar presented on each trial. Bottom: mean correlation between events as a function of event relatedness. Relative to younger adults, older adults exhibit reduced pattern similarity for related events, coupled with increased pattern similarity for events that are more distinct. Error bars represent standard error of the mean. ***p < 0.001; * p < 0.05. Abbreviations: SE, same exemplar; SS, same subcategory; SC, same category; DC, different category.

Fig. 5

Percent signal change during Hits and CRs in each regions of interest. Both older and younger adults exhibited increased activity in the ANG and HIPP during Hits relative to CRs. Only younger adults exhibited increased activity in the DLPFC during CRs relative to Hits. Error bars represent standard error of the mean. ***p < 0.001; ** p < 0.01; * p < 0.05. Abbreviations: CRs, correct rejections; ANG, angular gyrus; HIPP, hippocampus; DLPFC, dorsolateral prefrontal cortex.

Fig. 6

Mean hippocampal connectivity (Pearson r) with each cortical region of interest during Hits and CRs. Only younger adults exhibited increased hippocampal connectivity with the VLPFC during CRs relative to Hits, whereas hippocampal connectivity did not vary across trial types in older adults. Error bars represent standard error of the mean. ***p < 0.001. Abbreviations: CRs, correct rejections; VLPFC, ventrolateral prefrontal cortex.

Fig. 7

Age-invariant univariate effects during Hits > CRs (left two panels) and CRs > Hits (right panel), p < 0.05 FWE corrected. Abbreviation: CRs, correct rejections.