Instability in longitudinal sleep duration predicts cognitive impairment in aged participants of the Seattle Longitudinal Study

IMPORTANCE. Sleep disturbances and clinical sleep disorders are associated with all-cause dementia and neurodegenerative conditions. It remains unclear how longitudinal changes in sleep impact the incidence of cognitive impairment. OBJECTIVE. To evaluate how longitudinal sleep patterns contribute to age-related changes in cognitive function in healthy adults. DESIGN, SETTING, PARTICIPANTS. This study utilizes retrospective longitudinal analyses of a community-based study within Seattle, evaluating self-reported sleep (1993–2012) and cognitive performance (1997–2020) in aged adults. MAIN OUTCOMES AND MEASURES. The main outcome is cognitive impairment as defined by sub-threshold performance on 2 of 4 neuropsychological batteries: Mini-Mental State Examination (MMSE), Mattis Dementia Rating Scale, Trail Making Test, and Wechsler Adult Intelligent Scale (Revised). Sleep duration was defined through self-report of ‘average nightly sleep duration over the last week’ and assessed longitudinally. Median sleep duration, change in sleep duration (slope), variability in sleep duration (standard deviation, Sleep Variability), and sleep phenotype (“Short Sleep” median ≤7hrs.; “Medium Sleep” median = 7hrs; “Long Sleep” median ≥7hrs.). RESULTS. A total of 822 individuals (mean age of 76.2 years [11.8]; 466 women [56.7%]; 216 APOE allele positive [26.3%]) were included in the study. Analysis using a Cox Proportional Hazard Regression model (concordance 0.70) showed that increased Sleep Variability (95% CI [1.27,3.86]) was significantly associated with the incidence of cognitive impairment. Further analysis using linear regression prediction analysis (R(2)=0.201, F (10, 168)=6.010, p=2.67E-07) showed that high Sleep Variability (β=0.3491; p=0.048) was a significant predictor of cognitive impairment over a 10-year period. CONCLUSIONS AND RELEVANCE. High variability in longitudinal sleep duration was significantly associated with the incidence of cognitive impairment and predictive of decline in cognitive performance ten years later. These data highlight that instability in longitudinal sleep duration may contribute to age-related cognitive decline.