Sleep and circadian phase in a ships crew

Numerous factors influence the increased health risks of seamen. This study investigated sleep (by actigraphy) and the adaptation of the internal clock in watch-keeping crew compared to day workers, as possible contributory factors. Fourteen watch keepers, 4 h on, 8 h off (0800-1200/2000-2400 h, 1200-1600/ 2400-0400 h, 1600-2000/0400-0800 h) (fixed schedule, n = 6; rotating by delay weekly, n = 8), and 12 day workers participated during a voyage from the United Kingdom to Antarctica. They kept daily sleep diaries and wore wrist monitors for continuous recording of activity. Sleep parameters were derived from activity using the manufacturer's software and analyzed by repeated-measures ANOVA using SAS 8.2. Sequential urine samples were collected for 48 h weekly for 6-sulphatoxymelatonin measurement as an index of circadian rhythm timing. Individuals working watches of 1200-1600/2400-0400 h and 1600-2000/0400-0800 h had 2 sleeps daily, analyzed separately as main sleep (longest) and 2nd sleep. Main sleep duration was shorter in watch keepers than in day workers (p < 0.0001). Objective sleep quality was significantly compromised in rotaters compared to both day workers and fixed watch keepers, the most striking comparisons being sleep efficiency (percentage desired sleep time spent sleeping) main sleep (p < 0.0001) and sleep fragmentation (an index of restlessness) main sleep (p < 0.0001). The 2nd sleep was substantially less efficient than was the main sleep (p < 0.0001) for all watch keepers. There were few significant differences in sleep between the different watches in rotating watch keepers. Circadian timing remained constant in day workers. Timing of the 6-sulphatoxymelatonin rhythm was later for the watch of 1200-1600/2400-0400 h than for all others (1200-1600/ 2400-0400 h, 5.90 +/- 0.85 h; 1600-2000/0400-0800 h, 1.5 +/- 0.64 h; 0800-1200/ 2000-2400 h, 2.72 +/- 0.76 h; days, 2.09 +/- 0.68 h [decimal hours, mean SEM]: ANOVA, p < 0.01). This study identifies weekly changes in watch time as a cause of poor sleep in watch keepers. The most likely mechanism is the inability of the internal clock to adapt rapidly to abrupt changes in schedule.