Association between sleep duration, depression and cognitive decline trajectories: findings from a prospective cohort study in China.

Objective: This study investigated the relationship between sleep duration and cognitive decline trajectories among Chinese adults with age ≥ 45. Additionally, it examined whether baseline depression symptoms mediated the association between sleep duration and cognitive decline trajectories.

Methods: Data came from the China Health and Retirement Longitudinal Study (CHARLS), a nationally survey of Chinese adults. Total sleep duration was grouped into shorter (≤ 6 h), normal (6-9 h), and longer (> 9 h). Nighttime sleep duration was categorized as shorter (≤ 6 h), normal (6-8 h), and longer (> 8 h). Daytime nap duration was classified into no nap, shorter (≤ 0.5 h), normal (0.5-1.5 h), and longer (> 1.5 h). Cognitive decline trajectories were identified using a group-based trajectory model (GBTM). Depression symptoms, measured by baseline depression scores, were included as a potential mediating variable. Multinomial logistic regression models were applied to analyze the association between sleep duration and cognitive trajectories.

Results: A total of 4,094 respondents were included. Participants were classified into three trajectory groups: high-level stable (n = 1,278, 31.2%), moderate-level slow-decline (n = 2,089, 51.0%), and low-level rapid-decline (n = 727, 17.8%). Compared to the stable group, shorter total sleep duration was associated with both rapid-decline (RRR:1.510; 95%CI:1.079,2.114) and slow-decline (RRR:1.466; 95%CI:1.145,1.877) cognitive trajectories. Similarly, shorter nighttime sleep duration was associated with both rapid-decline (RRR:1.572; 95%CI: 1.089,2.271) and slow-decline (RRR:1.480; 95%CI:1.129,1.939) cognitive trajectories compared to the stable group. Baseline depression symptoms did not mediate these relationships (P > 0.05).

Conclusions: Shorter total sleep duration (≤ 6 h) and shorter nighttime sleep duration (≤ 6 h) were significantly associated with cognitive decline trajectories. However, baseline depression symptoms did not mediate these relationships.