Longitudinal changes of self‐ and informant‐reported cognitive decline in elderly with regard to clinical progression and amyloid‐status

Abstract

AbstractBackgroundSelf‐ and informant‐reports of cognitive decline have been previously associated with a higher risk of cognitive decline in elderly. Yet, few studies assessed their longitudinal changes over time with regard to both incident mild cognitive impairment (iMCI) and Alzheimer’s disease (AD) pathology. This study aims at describing these links to better understand which trajectories are indicative of higher AD‐risk.MethodWe used data from 501 participants followed‐up annually (2.51±1.35 years) in the DELCODE project. All were cognitively unimpaired, and had an amyloid measurement available, at inclusion. Among them, 145 were amyloid‐positive, and 65 progressed to iMCI during the follow‐up period (31 iMCI‐Aß+). The SCD‐Interview was used to measure (i) the number of cognitive domains in which the participants or its relative perceived a decline (i.e., self‐ and informant‐SCD‐I), and (ii) the number of fulfilled SCD‐plus criteria. Linear mixed‐effect models were used to assess SCD trajectories over time, according to four groups defined by amyloid status and clinical progression (Stable‐Aß‐, Stable‐Aß+, iMCI‐Aß‐ and iMCI‐Aß+). Analyses were adjusted for age, sex and education; and their interaction with time.ResultBoth self‐SCD‐I levels and SCD‐plus scores decreased significantly over time, without significant differences between groups (Figure 1A, 1B). Conversely, informant‐SCD‐I levels did not significantly change over time in the whole sample, although slopes significantly differed between groups. Compared to the Stable‐Aß‐ group, there was a greater increase over time of informant‐SCD‐I report only in iMCI‐Aß+ [est 0.16, SE 0.05, p<.001] (Figure 1C). The resulting discrepancy score (Self‐SCD‐I minus Informant‐SCD‐I) tended to decrease over time with significant differences between groups. Compared to the Stable‐Aß‐ group, the discrepancy score showed a faster decrease over time in iMCI‐Aß+ [est ‐0.21, SE 0.056 p<.001] (Figure 1D).ConclusionLongitudinal patterns in subjective reports of cognitive decline might help identify preclinical AD subjects at higher risk of clinical progression. Specifically, an increase of informant‐reported cognitive decline over time may indicate objective cognitive decline in amyloid‐positive subjects. Conversely, the decrease of self‐reported cognitive decline over time, regardless of the group, may reflect generally reduced concerns after inclusion, or result from current limitations in longitudinal SCD assessment.