Abstract
BACKGROUND: Evidence of the association between sleep duration, depressive symptoms, and mortality in cancer survivors is limited. METHODS: 2,111 participants from the National Health and Nutrition Examination Surveys were involved and linked to a mortality database up to 31 December 2019. Sleep duration was self-reported at baseline. Depressive symptom was assessed using the Patient Health Questionnaire (PHQ-9) and a total score of ≥ 10 is considered clinically relevant depressive symptoms. Cox proportional hazards regression model was the main statistical models. RESULTS: A 2% increase in all-cause mortality was found for each 1-point increase in depressive symptoms score (hazard ratio [HR], 1.02; 95% confidence interval [CI], 1.00-1.03). Compared with those patients who slept 7 h, positive associations of < 6 h (HR, 1.79; 95% CI, 1.29-2.50), 8 h (HR, 1.46; 95% CI, 1.15-1.85), and > 8 h (HR, 1.49; 95% CI, 1.09-2.05) durations with all-cause mortality were observed. In the joint analysis, cancer survivors with depressive symptoms had a higher risk of all-cause mortality when sleeping < 6 h (HR, 1.69; 95% CI, 1.09-2.62) or ≥ 8 h (HR, 1.55; 95% CI, 1.02-2.35). LIMITATIONS: Dynamics of sleep and depression data were not available. CONCLUSIONS: Long (≥ 8 h) and short (< 6 h) sleep duration was both associated with increased risks of all-cause deaths after adjusting for depressive symptoms scores. We also observed a joint effect of sleep duration and depressive symptoms. The findings may provide evidence to guide optimal sleep duration for prolonged cancer survival.