Sleep and cancer mortality in the Cancer Prevention Study-II.

Abstract

Sleep is a multi-dimensional human function that is associated with cancer outcomes. Previous work on sleep and cancer mortality have not investigated how this relationship varies by sex and cancer site. We investigated the association of sleep duration and perceived insomnia with site-specific and overall cancer mortality among participants in the Cancer Prevention Study-II. Sleep was collected at baseline in 1982 among 1.2 million cancer-free US adults. Cancer-specific mortality was determined through 2018. We used multivariable Cox proportional hazard models to calculate hazard ratios and 95% confidence intervals for overall and site-specific cancer mortality, stratified by sex. Among 983,105 participants (56% female) followed for a median of 27.9 person-years, there were 146,911 primary cancer deaths. Results from the adjusted model showed short (6h/night) and long (8 h/night and 9-14 h/night) sleep duration, compared to 7 h/night, were associated with a modest 2%, 2%, and 5% higher risk of overall cancer mortality, respectively, and there was a significant non-linear trend (p-trend < 0.01). This non-linear trend was statistically significant among male (p-trend < 0.001) but not female (p-trend 0.71) participants. For male participants, short and long sleep were associated with higher risk of lung cancer mortality and long sleep was associated with higher risk of colorectal cancer mortality. Perceived insomnia was associated with a 3-7% lower risk of overall cancer mortality. Sleep is important to consider in relation to sex- and site-specific cancer mortality. Future research should investigate other components of sleep in relation to cancer mortality.