Alterations in metamemory capacity and neural correlates in a subtype of subjective cognitive decline.

BACKGROUNDS: Subjective cognitive decline (SCD), one of the important clinical indicators for preclinical Alzheimer's disease (AD), is primarily defined as self-perceived cognitive decline without objective evidence for cognitive impairment. However, the accuracy of their self-evaluation of cognition is unclear. This study sought to investigate the capacity for self-evaluation of own cognitive performance in SCD by applying an objective metamemory paradigm. METHODS: 147 individuals with SCD were classified into four subgroups by their subjective feeling of worse performance than peers or not (P+/-) and whether they have objectively slight cognitive impairment compared to normative data (S+/-). Metamemory scores, the amplitude of the low-frequency fluctuation (ALFF), fractional low-frequency fluctuation amplitude (fALFF), and cortical thickness were compared among four subgroups. Partial correlations between neuropsychological scores and neuroimaging measures were examined, controlling for age, sex, and education years. RESULTS: SCD S+P- showed the worst performance in short-term delayed recall and the worst metamemory performance, indicated by the highest value in the degree of confidence of short-term delayed recall (DOC-N4) and long-term cued recall (DOC-N6) and the worst value in relative accuracy of judgments of short-term delayed recall (ROJ-N4). ALFF values in the bilateral superior medial frontal and olfactory cortices and the left superior orbitofrontal gyrus cortex were significantly higher in SCD P- compared with SCD P+ groups (all P < 0.05, FWE-corrected, cluster-wise level). A significant S × P interaction effect in the left hippocampus and middle cingulate cortex was found for the fALFF signals (all P < 0.05, FWE-corrected, cluster-wise level). Significant interaction and main effects on cortical thickness were reported. The parahippocampal and posterior cingulate cortices were significantly decreased in SCD S+P- (all P < 0.05). CONCLUSION: SCD S+P- showed the worst episodic memory performance, altered metamemory capacity (overconfidence and less accuracy of judgment), and altered neuroimaging measures, though they had feelings of similar performance with peers. Our results indicate that metamemory capacity is affected in a subtype of SCD with reduced cortical thickness and intensity of regional spontaneous activity in key areas for metamemory processing.