Poor objective sleep quality is associated with worse cognition cross‐sectionally and longitudinally in cognitively unimpaired adults

Abstract

AbstractBackgroundSleep disturbances are prevalent in Alzheimer’s disease (AD), with sleep quality being impaired already in the preclinical stages. Specifically, reduced sleep efficiency and sleep fragmentation have been cross‐sectionally associated with poor cognition in healthy older adults. However, there is limited knowledge of how objectively measured sleep features relate to cognitive change. We aimed to investigate the relationships of actigraphy‐measured sleep with both cross‐sectional and longitudinal cognition in a cohort of cognitively unimpaired (CU) adults at increased risk of AD.MethodThis study included 123 CU adults from the ALFASleep project (mean age 63.9 years, 60.2% females, 50.4% APOE‐ε4 carriers) without sleep medication use (Table 1). Correlates of sleep quality were measured using actigraphy (Actiwatch2®, Philips Respironics) for two weeks. We used separate multivariate linear regressions to assess the association between actigraphy‐derived sleep parameters: total sleep time (TST), sleep efficiency, sleep latency, wake after sleep onset (WASO), and cognitive composite outcomes: Preclinical Alzheimer’s Cognitive Composite (PACC), and executive, memory and attention composites. We corrected all models for age, sex, APOE‐ε4, status, education and the time difference between cognitive and sleep assessments. Furthermore, we performed analyses on cognitive change from a previous visit (performed on average 3 years before) additionally adjusting by the time difference between two cognitive assessments.ResultsIn cross‐sectional analyses, shorter TST was associated with lower PACC (p = 0.049) and executive composite (p = 0.039) scores. Also, lower sleep efficiency and higher WASO (indicating more nocturnal awakenings) were associated with lower performance on executive tasks (p = 0.015; p = 0.026) (Table 2). In longitudinal analyses, shorter TST was associated with poorer longitudinal performance in PACC and executive composite scores (p = 0.020).ConclusionsObjectively measured indicators of poorer sleep quality are associated with poorer cognitive outcomes, both cross‐sectionally and longitudinally, in middle‐aged CU adults. Importantly, the executive function seems to be the most vulnerable to sleep impairment. Further research to unravel the underlying mechanisms using multi‐modal neuroimaging and CSF biomarkers of AD pathology is undergoing.