Abstract
STUDY OBJECTIVES: To prospectively investigate the association between sleep traits and lung cancer risk, accounting for the interactions with genetic predisposition of lung cancer.
METHODS: We included 469 691 individuals free of lung cancer at recruitment from UK Biobank, measuring sleep behaviors with a standardized questionnaire and identifying incident lung cancer cases through linkage to national cancer and death registries. We estimated multivariable-adjusted hazard ratios (HRs) for lung cancer (2177 incident cases) across four sleep traits (sleep duration, chronotype, insomnia, and snoring) and examined the interaction and joint effects with a lung cancer polygenic risk score.
RESULTS: A U-shaped association was observed for sleep duration and lung cancer risk, with an 18% higher risk (95% confidence interval [CI]: 1.07 to 1.30) for short sleepers and a 17% higher risk (95% CI: 1.02 to 1.34) for long sleepers compared with normal sleepers (7-8 h/day). Evening preference was associated with elevated lung cancer risk compared with morning preference (HR: 1.25; 95% CI: 1.07 to 1.46), but no association was found for insomnia or snoring. Compared with participants with favorable sleep traits and low genetic risk, those with both unfavorable sleep duration (<7 hours or >8 hours) or evening preference and high genetic risk showed the greatest lung cancer risk (HRsleep duration: 1.83; 95% CI: 1.47 to 2.27; HRchronotype: 1.85; 95% CI: 1.34 to 2.56).
CONCLUSIONS: Both unfavorable sleep duration and evening chronotype were associated with increased lung cancer incidence, especially for those with low to moderate genetic risk. These results indicate that sleep behaviors as modifiable risk factors may have potential implications for lung cancer risk.