Association between sleep duration from midlife to late life and the risk of depressive symptoms: the Singapore Chinese Health Study

Huiqi Li,Bee Choo Tai,An Pan,Woon-Puay Koh

DOI: https://doi.org/10.1192/bjo.2024.772

2024-10-11

BJPsych Open

Abstract:Background: The prospective association between sleep duration and the development of late-life depressive symptomology is unclear. Aims: To investigate sleep duration from midlife to late life in relation to risk of depressive symptoms in late life. Method: A total of 14 361 participants from the Singapore Chinese Health Study were included in the present study. Daily sleep duration was self-reported at baseline (mean age of 52.4 years; 1993-98), follow-up 2 (mean age of 65.2 years; 2006-10) and follow-up 3 (mean age of 72.5 years; 2014-16) interviews. Depressive symptoms were evaluated using the Geriatric Depression Scale at follow-up 3 interviews. Modified Poisson regression models were performed to estimate relative risks and 95% confidence intervals of late-life depressive symptoms in relation to sleep duration at baseline and the two follow-up interviews. Results: Compared with sleeping 7 h per day, a short sleep duration of ≤5 h per day at baseline (i.e. midlife) was related to a higher risk of depressive symptoms (relative risk 1.10, 95% CI 1.06-1.15), and this risk was not affected by subsequent prolongation of sleep. Conversely, a long sleep duration of ≥9 h per day at baseline was not related to risk of depressive symptoms. At follow-up 3 (i.e. late life), both short sleep (relative risk 1.20, 95% CI 1.16-1.25) and long sleep (relative risk 1.12, 95% CI 1.07-1.18) duration were cross-sectionally associated with depressive symptoms. Conclusion: Short sleep duration in midlife, regardless of subsequent prolongation, is associated with an increased risk of depression in late life. Contrariwise, both short and long sleep duration in late life co-occur with depressive symptoms.