Metacognition, cortical thickness, and tauopathy in aging

Abstract

We investigated self-rating of cognitive task performance (self-appraisal) and the difference between self-rating and actual task performance (appraisal discrepancy) in cognitively healthy older adults and their relationship with cortical thickness and Alzheimer's disease (AD) biomarkers, amyloid and tau. All participants (N = 151) underwent neuropsychological testing and 1.5T structural magnetic resonance imaging. A subset (N = 66) received amyloid-PET with [11C] PiB and tau-PET with [18F] Flortaucipir. We found that worse performers had lower self-appraisal ratings, but still overestimated their performance, consistent with the Dunning-Kruger effect. Self-appraisal rating and appraisal discrepancy revealed distinct relationships with cortical thickness and AD pathology. Greater appraisal discrepancy, indicating overestimation, was related to thinning of inferior-lateral temporal, fusiform, and rostral anterior cingulate cortices. Lower self-appraisal was associated with higher entorhinal and inferior temporal tau. These results suggest that overestimation could implicate structural atrophy beyond AD pathology, while lower self-appraisal could indicate early behavioral alteration due to AD pathology, supporting the notion of subjective cognitive decline prior to objective deficits.

Keywords: Cortical thickness; Metacognition; PET; Subjective cognitive decline; Tau.

Fig. 1.

Relationship between self-appraisal/appraisal discrepancy and neuropsychological task performance. Better objective performance was related to higher self-appraisal and lower appraisal discrepancy.

Fig. 2.

Vertex-wise whole brain analyses were conducted to examine the relationship between cortical thickness and (A) average appraisal discrepancy (overestimation correlated with thinner cortices) and (B) average objective performance, after adjusting for age, sex, education, and GDS (worse objective performance correlated with thinner cortices). The maps shown here used a liberal threshold of cluster-wise Monte Carlo simulation correction at p < 0.01. Overlapping regions are presented in (C). GDS, geriatric depression scale.

Fig. 3.

Relationship between self-appraisal/appraisal discrepancy and AD biomarkers, adjusting for age, sex, education, and GDS. Higher entorhinal FTP SUVR was significantly associated with lower average self-appraisal ratings (A), especially in individuals with higher global PiB DVR (B). Higher inferior temporal FTP SUVR was related to lower average self-appraisal ratings (C) but there was no significant interaction between global PiB DVR and inferior temporal FTP SUVR. (D) shows the relationship between entorhinal FTP SUVR and average appraisal discrepancy. There was a significant interaction between global PiB DVR and entorhinal FTP SUVR: higher entorhinal tau was related to greater average appraisal discrepancy (overestimation) but in individuals with lower amyloid level; this interaction is unlikely to be biologically significant (see text) (E). There was no significant association between inferior temporal FTP SUVR and average appraisal discrepancy (F). FTP. Flortaucipir; SUVR, standardized update value ratio; GDS, geriatrics depression scale; PiB, Pittsburg compound B; PiB positivity DVR threshold, 1.065.