Long-term, night shiftwork has been identified as a potential carcinogenic risk factor. It is hypothesized that increased light at night exposure during shiftwork reduces melatonin production, which is associated with increased cancer risk. Sleep duration has been hypothesized to influence both melatonin levels and cancer risk, and it has been suggested that sleep duration could be used as a proxy for melatonin production. Finally, physical activity has been shown to reduce cancer risk, and laboratory studies indicate it may influence melatonin levels. A cross-sectional study of light exposure, sleep duration, physical activity, and melatonin levels was conducted among 61 female rotating shift nurses (work schedule: two 12 h days, two 12 h nights, five days off). Light intensity was measured using a light-intensity data logger, and sleep duration and physical activity were self-reported in a study diary and questionnaire. Melatonin concentrations were measured from urine and saliva samples. The characteristics of nurses working day and night shifts were similar. Light intensity was significantly higher during sleep for those working at night (p< 0.0001), while urinary melatonin levels following sleep were significantly higher among those working days (p = 0.0003). Mean sleep duration for nurses working during the day (8.27 h) was significantly longer than for those working at night (4.78 h, p< 0.0001). An inverse association (p = 0.002) between light exposure and urinary melatonin levels was observed; however, this was not significant when stratified by shift group. There was no significant correlation between sleep duration and melatonin, and no consistent relationship between physical activity and melatonin. Analysis of salivary melatonin levels indicated that the circadian rhythms of night workers were not altered, meaning peak melatonin production occurred at night. This study indicates that two nights of rotating shift work may not change the timing of melatonin production to the day among those working at night. Additionally, in this study, sleep duration was not correlated with urinary melatonin levels, suggesting it may not be a good proxy for melatonin production.
