Authors: Laura Rabin, Jacqueline Mogle, Mindy J. Katz, Martin J. Sliwinski
Published: 2020-12-07
DOI: 10.1002/alz.045918
Source: Full article
AbstractBackgroundAlthough subjective cognitive concerns are common in older adults and may portend progression to Alzheimer’s disease, the cross‐sectional relationship between subjective and objective cognition is inconsistent. Many studies report statistically significant, though weak, relations between self‐reported cognition and neuropsychological test scores among cognitively normal (CN) older adults and those with mild cognitive impairment (MCI). Others have failed to find an association or have found stronger associations between subjective cognition and psychosocial variables. We applied a novel approach to investigate the subjective‐objective cognition relation, using both conventional measures and daily assessments in ethnically/racially diverse older adults.MethodParticipants (n=229, mean age=76.97, 65% female, 50% White) completed in‐person neuropsychological tests and a paper‐and‐pencil questionnaire of change in cognitive functioning over 5 years. Additionally, for 14 days, they completed daily cognitive tasks and self‐reported their experience of cognitive difficulties via smartphones. Participants were classified as MCI (n=62) or CN (n=167) using Jak/Bondi criteria. We used correlational analyses (controlling for depression) to compare subjective/objective data stratified by cognitive status and logistic regression to examine associations with diagnosis.ResultIn MCI, there were statistically significant moderate associations (r=.34 to .38; p<.01) between daily diary self‐reported cognition (retrospective and prospective memory) and ambulatory assessments of processing speed and spatial memory, respectively (performance averaged over multiple trials, across days); there was no relationship with traditional neuropsychological tests. In CN, there were statistically significant, weak‐moderate associations between self‐report of decline in various cognitive domains and traditional tests of episodic memory and language (r=‐.22 to ‐.29; p<.01). Self‐reported cognition on daily diary but not questionnaire items was associated with group membership, controlling for age, education, sex, and general cognition (p<.05).ConclusionFor individuals with MCI, daily reports of cognitive difficulties were moderately aligned with performance on objective ambulatory cognitive assessments. By contrast, in those with intact cognition, there was modest correspondence between single‐occasion retrospective reporting of cognition and traditional neuropsychological tests. Notably, only daily subjective report items predicted MCI status at baseline. Results have implications for enhancing the accuracy of assessments, particularly for those with MCI, through the inclusion of daily measurements to capture features of cognition that may otherwise be missed.