Nonrestorative sleep, sleep duration, and chronotype in relation to depression: a cross-sectional study in Japan

Nonrestorative sleep (NRS), sleep duration, and chronotype are associated with depression. However, the combined association between NRS and sleep duration or chronotype and depression remains unclear. Thus, this study aimed to investigate the association between depression and combinations of NRS and sleep duration or chronotype. This cross-sectional study used data from the Japan “Society and new Tobacco” Internet Survey in 2025, including participants aged ≥ 20 years. Depression was defined as self-reported currently having depression under medical care with medication. Participants were classified according to the frequency of NRS, defined as a lack of subjective sleep recovery, sleep duration, and chronotype, assessed based on the sleep-corrected midpoint of sleep on free days. Multivariate logistic regression was used to assess the association between depression and combinations of NRS and sleep duration or chronotype. Of the 20,951 participants, 658 reported current depression. In those with NRS ≥ 3 times/week, the odds of depression were higher across all sleep durations, with sleeping ≥ 9 h showing the highest (adjusted odds ratio [aOR], 8.42; 95% confidence interval [CI], 5.13–13.79), compared with those without NRS who slept 7–8 h. Among those without NRS, sleeping 6–7 h had the lowest odds of depression (aOR, 0.56; 95% CI, 0.32–0.99), whereas sleeping < 5 h and sleeping 8–9 h were associated with higher odds (aOR 2.06, 95% CI: 1.04–4.06 and aOR 1.97, 95% CI: 1.21–3.19, respectively). Frequent NRS was associated with higher odds across all chronotypes; however, chronotype was not significantly associated with depression among those without NRS. Frequent NRS is consistently associated with depression across all sleep durations and chronotypes. Assessing NRS frequency may provide additional information beyond sleep duration or chronotype to identify adults at risk of depression.