Sleep duration and timing are nonlinearly associated with depressive symptoms among older adults

Geriatric depression is a common but preventable psychiatric disorder; however, its association with specific sleep patterns remains unclear. Therefore, we examined the association of self-reported sleep duration and sleep timing with depressive symptoms in the older population. A total of 1068 older Taiwanese adults (52.7% women; 72.2 ± 5.7 y) responded to a telephone survey during 2019–2020. Self-reported data on sociodemographic characteristics, sleep duration, bedtime, wake-up time (adapted items from Pittsburgh Sleep Quality Index), and depressive symptoms (five-item Center for Epidemiological Studies–Depression scale) were included. Generalized additive models were used to examine the nonlinear associations of sleep duration and midpoint sleep time (ie, the midpoint of bedtime and wake-up time) with depressive symptoms. The means of sleep duration and midpoint sleep time in the participants were 6 h per night and 02:13 h, respectively. The results showed marked nonlinear associations of sleep patterns with depressive symptoms. Sleep duration shorter than 4 h per night was associated with a relatively higher level of depressive symptoms, with the highest risk (coefficient = 3.41; 95% confidence interval [CI] = 2.12, 4.70) while sleeping 2.06 h per night. The midpoint sleep time was positively associated with depressive symptom scores after 01:00 h. The results showed that sleep duration and fitting sleep timing were nonlinearly associated with the risks of depressive symptoms in the general older adult population. These findings have implications for targeting nonpharmacological approaches by tackling modifiable behaviors, such as adequate sleep duration and timing, with decreased risks of depressive symptoms in the older adult population.