Sleep duration and cancer risk: time to use a “sleep-years” index?

With a focus on melatonin, a recent paper in the Journal investigated the hypothesis that endometrial cancer might be associated with the duration, and ultimately, amount of sleep. The authors found that “[s]elf-reported sleep duration may not adequately represent melatonin levels.” The authors also concluded that there was “weak evidence of an association between sleep duration and endometrial cancer risk.” Overall, these are interesting observations because primarily experimental and mechanistic research from many angles supports the study’s notion that inappropriate sleep may be a determinant of cancer risk. To find out whether this is so in man, rather than assigning study individuals to fixed or average “baseline sleep categories” i.e., ≤5, 6, 7, 8, ≥9 h of habitual sleep in the present study, the accumulated amount of sleep over decades should be reconstructed in retrospective or constructed in prospective studies. To achieve this end, future epidemiological studies may want to use a sleep-years index [SYI]. This simple exposure parameter promises to be a sensible, feasible, and affordable way to approximate cumulative time spent at sleep in critical time windows over many years which we should expect to be relevant for the development of cancer. The SYI could be tested and used in observational studies which promise to be comparable and can be merged. This commentary provides roots of the index and explains why and how it should be used and how it could be interpreted in rigorous studies of biologically plausible links between sleep, on the one hand, and the development of internal cancers, on the other. This commentary also points out limitations of interpreting the SYI. It is emphasized that, where possible, the SYI should be assessed independently of (a) other sleep facets—such as quality—and of (b) known or suspected cancer risk factors. The respective contribution of (a) and (b) to risk must then be assessed during the analyses. Overall, the suggested inclusion and standardization of assessing sleep duration could be an important step forward when evaluating possible cancer risks in relation to sleep. Finally, the proposed approach may prove useful beyond sleep epidemiology per se: to exemplify, research into suggested causal links between disrupted natural sleep-wakefulness cycles and increased cancer risks in shift-workers could also benefit.