The independent and combined associations of nocturnal sleep duration, sleep midpoint, and sleep onset latency with global cognitive function in older Chinese adults

Abstract

This study aims to investigate the independent and combined associations of nocturnal sleep duration, sleep midpoint, and sleep onset latency with global cognitive function in older Chinese adults. Our cross-sectional study included 4601 community-dwelling cognitively unimpaired adults aged 60 years or older from the West China Health and Aging Cohort Study. Sleep characteristics were assessed using the Pittsburgh Sleep Quality Index, and global cognitive function was evaluated using the Mini-Mental State Examination (MMSE). Multivariable linear regression models were employed, adjusting for an extensive set of demographic, lifestyle, and comorbidity factors. Subgroup analyses were performed based on sex, age, and genetic risk profiles for cognitive performance. The mean age of participants was 69.0 ± 5.53 years, with 52.1% being female. The mean MMSE total score was 24.9 ± 3.20. Compared to the reference category for each sleep variable, sleep duration < 5 h/day or > 8 h/day, sleep midpoint earlier than 1:30 AM, and sleep latency > 60 min were each independently associated with significantly lower MMSE scores (β range - 0.36 to - 0.34; 95% confidence interval range - 0.60 to - 0.10). A combined analysis revealed that individuals with concurrent extreme sleep duration, early midpoint, and/or long latency had even lower MMSE scores, especially among those with genetically predicted poorer cognitive performance (β up to - 1.86). Multiple dimensions of sleep are independently and jointly associated with global cognitive function in older Chinese adults, highlighting the importance of a holistic approach to sleep in cognitive aging research and interventions.

Keywords: Genetic susceptibility; Global cognitive function; Older people; Sleep patterns.

Fig. 1

Results of restricted cubic spline regressions of MMSE (Mini-Mental State Examination) total score on A sleep duration, B sleep midpoint, and C sleep latency. The solid blue line represents the estimated change in MMSE score relative to the reference value, and the light blue shaded region indicates the 95% confidence interval. Models adjusted for: age, income, educational level, ethnicity, marital status, body mass index, smoking status, drinking status, physical activity level, history of major diseases (type 2 diabetes, hypertension, cardiovascular diseases, and cancers), and the other two sleep phenotypes not investigated as the main exposure

Fig. 2

Joint effects of A sleep duration and sleep midpoint, B sleep midpoint and sleep latency, C sleep duration and sleep latency, D sleep duration, sleep midpoint, and sleep latency on MMSE (Mini-Mental State Examination) total score. The dots represent point estimates (difference in mean MMSE score compared to the reference group), and the horizontal lines represent 95% confidence intervals. Reference duration, duration of nocturnal sleep ≥ 5 h/day and ≤ 8 h/day; reference midpoint, midpoint of sleep no earlier than 1:30 AM; reference latency, latency of sleep shorter than 60 min. Models adjusted for: age, income, educational level, ethnicity, marital status, body mass index, smoking status, drinking status, physical activity level, history of major diseases (type 2 diabetes, hypertension, cardiovascular diseases, and cancers), and the other sleep phenotype not involved in the joint exposure (for analyses in A, B, and C). LCI, lower bound of the 95% confidence intervals; UCI, upper bound of the 95% confidence intervals

Fig. 3

Joint effects of sleep duration, sleep midpoint, and sleep latency on MMSE (Mini-Mental State Examination) total score among lower (red) and higher (blue) cognitive performance polygenic score groups. The dots represent point estimates (difference in mean MMSE score compared to the reference group), and the horizontal lines represent 95% confidence intervals. Reference duration, duration of nocturnal sleep ≥ 5 h/day and ≤ 8 h/day; reference midpoint, midpoint of sleep no earlier than 1:30 AM; reference latency, latency of sleep shorter than 60 min. Models adjusted for: age, income, educational level, ethnicity, marital status, body mass index, smoking status, drinking status, physical activity level, and history of major diseases (type 2 diabetes, hypertension, cardiovascular diseases, and cancers). LCI, lower bound of the 95% confidence intervals; UCI, upper bound of the 95% confidence intervals