Decreased melatonin production, due to nighttime exposure to light, has been proposed as one of the physiological mechanisms increasing cancer risk in night workers. However, few studies measured melatonin production in night workers, and most of these studies did not measure melatonin over 24 h. One study compared total melatonin production between day and night shifts in rotating night workers and did not find significant differences. However, without baseline measures, it was not possible to exclude that melatonin production was reduced during both day and night work. Here, we used data collected in a simulation study of night work to determine the effect of night work on both nighttime and 24-h melatonin production, during three consecutive days of simulated night work. Thirty-eight healthy subjects (15 men, 23 women; 26.6 ± 4.2 years) participated in a 6-d laboratory study. Circadian phase assessments were made with salivary dim light melatonin onset (DLMO) on the first and last days. Simulated day work (09:00-17:00 h) occurred on the second day, followed by three consecutive days of simulated night work (00:00-08:00 h). Light intensity at eye level was set at 50 lux during both simulated day and night work. The subjects were divided into three matched groups exposed to specific daytime light profiles that produced various degrees of circadian phase delays and phase advances. Melatonin production was estimated with the excretion of urinary 6-sulfatoxymelatonin (aMT6s). For the entire protocol, urine was collected every 2 h, except for the sleep episodes when the interval was 8 h. The aMT6s concentration in each sample was multiplied by the urine volume and then added to obtain total aMT6s excretion during nighttime (00:00-08:00 h) and during each 24-h day (00:00-00:00 h). The results showed that melatonin production progressively decreased over consecutive days of simulated night work, both during nighttime and over the 24 h. This decrease was larger in women using oral contraceptives. There was no difference between the three groups, and the magnitude of the decrease in melatonin production for nighttime and for the 24 h was not associated with the magnitude of the absolute circadian phase shift. As light intensity was relatively low and because the decrease in melatonin production was progressive, direct suppression by nighttime light exposure was probably not a significant factor. However, according to previous experimental observations, the decrease in melatonin production most likely reflects the circadian disruption associated with the process of re-entrainment. It remains to be determined whether reduced melatonin production can be harmful by itself, but long-term and repeated circadian disruption most probably is.
