To explore interactions between depression and sleep duration and their impact on long-term cognitive trajectories. This longitudinal study (2011−2020) analyzed depressive symptoms and sleep duration in relation to 10-year cognitive trajectories in 7649 Chinese adults aged 45–80 years. Exposures included sleep duration and depressive symptoms. Cognitive function was the primary outcome, standardized using regression models adjusted for age, sex, and education. Standardized Z -scores for global cognition, episodic memory, and mental status were derived. Growth mixture modeling identified three cognitive trajectories. Multivariable logistic regression examined exposure-trajectory associations. Three distinct cognitive trajectories were identified: persistently high, moderately stable, and persistently low. Compared with the “non-depression-optimal sleep” group, all other sleep-depression groups showed significantly reduced associations, with “depression-long sleep” exhibiting the highest risk of low cognition. Within the persistent moderate group, all combinations except “non-depression-short sleep” showed significant negative associations, with “depression-long sleep” exhibiting the greatest risk. Sleep duration had a U-shaped relationship with cognitive scores, whereas depressive symptoms were negatively associated. Optimal cognitive maintenance occurred among participants with no depression and a 6–8 h sleep duration. Promoting regular, healthy sleep and reducing depression risk may help preserve cognitive function in old age. In a 10-year cohort study of older adults, the absence of depression combined with 6–8 h of sleep was optimal for maintaining cognitive function, while depression coupled with long sleep duration posed the highest risk for cognitive decline. • Optimal sleep (6–8 h) and no depression support stable long-term cognition. • Depression plus short or long sleep accelerates cognitive decline. • Sleep duration shows a U-shaped link with cognition, peaking at ~7 h. • Depression severity is negatively associated with cognitive function.
Jiang et al. (Sun,) studied this question.