Association between total sleep time and all cancer mortality: non-linear dose-response meta-analysis of cohort studies.

ABSTRACT

OBJECTIVE: Appropriate total sleep time is reported to be associated with several important health outcomes. However, the relationship between total sleep time and all cancer mortality is not well defined because of inconsistent results from published studies, and no dose-response meta-analysis was performed to evaluate the exact dose-response relationship. METHODS: We conducted a literature search of PubMed and Web of Science to identify all relevant epidemiological studies published before August 9, 2018. We performed categorical and non-linear dose-response meta-analyses to quantify the association between total sleep time and all cancer mortality. RESULTS: Finally, we included 14 cohort studies in the present meta-analyses enrolling 866,877 participants with 43,021 cancer deaths. We found that total sleep time less than seven hours was not significantly associated with increased risk of all cancer mortality [relative risk (RR) = 1.02; 95% confidence interval (CI) = 0.99-1.05]. However, four to five hours total sleep time was related to an 8% increased risk of all cancer mortality (RR = 1.08; 95% CI = 1.02-1.13) in dose-response meta-analysis. Furthermore, long total sleep time (≥8 hours) was weakly associated with all cancer mortality (RR = 1.05; 95% CI = 1.02-1.08). However, the increment in total sleep time longer than nine hours was notably associated with an increased risk of cancer mortality. CONCLUSION: The current meta-analysis provides evidence of a positive association between total sleep time of four to five hours and total sleep time longer than eight hours with the risk of all cancer mortality among the general population. Additional studies are needed to establish causality.