Sleepiness and sleep in a simulated "six hours on/six hours off" sea watch system

Ships are operated around the clock using rapidly rotating shift schedules called sea watch systems. Sea watch systems may cause fatigue, in the same way as other irregular working time arrangements. The present study investigated subjective sleepiness and sleep duration in connection with a 6 h on/6 h off duty system. The study was performed in a bridge simulator, very similar to those found on ships. Twelve officers divided into two groups participated in the study that lasted 66 h. Half of the subjects started with the 06:00-12:00 h watch and the other half with the 12:00-18:00 h watch. The subjects alternated between off-duty and on-duty for the remainder of the experimental period. Approximately halfway through the experiment, the 12:00-18:00 h watch was divided into two 3 h watches/off-duty periods. The effect of this was to reverse the on-duty/off-duty pattern between the two groups. This enabled all subjects to work the four possible watches (00:00-06:00 h, 06:00-12:00 h, 12:00-18:00 h, and 18:00-24:00 h) in an order that was essentially counterbalanced between groups. Ratings of sleepiness (Karolinska Sleepiness Scale; KSS) were obtained every 30 min during on-duty periods and if subjects were awake during off-duty periods. The subjectively rated duration of sleep was recorded after each off-duty period that preceded watch periods when KSS was rated. The results showed that the average level of sleepiness was significantly higher during the 00:00-06:00 h watch compared to the 12:00-18:00 h and 18:00-24:00 h watches, but not to the 06:00-12:00 h watch. Sleepiness also progressed significantly from the start toward the end of each watch, with the exception of the 06:00-12:00 h watch, when levels remained approximately stable. There were no differences between groups (i.e., the order between watches). Sleep duration during the 06:00-12:00 h off-duty period (3 h 29 min) was significantly longer than during the 12:00-18:00 h period (1 h 47 min) and the 18:00-24:00 h period (2 h 7 min). Sleep during the 00:00-06:00 h period (4 h 23 min) was longer than all sleep periods except the 06:00-12:00 h period. There were no differences between groups. In spite of sufficient opportunities for sleep, sleep was on the average around 1-1 h 30 min shorter than the 7-7 h 30 min that is considered "normal" during a 24 h period. This is probably a consequence of the difficulty to sleep during daytime due to the alerting effects of the circadian rhythm. Also, sleepiness during the night and early mornings reached high levels, which may be explained by a combination of working close to or during the circadian trough of alertness and the relatively short sleep periods obtained. An initial suppression of sleepiness was observed during all watches, except for the 06:00-12:00 h watch. This suppression may be explained by the "masking effect" exerted by the relative high levels of activity required when taking over the responsibility of the ship. Toward the end of watches, the levels of sleepiness progressively increased to relatively high levels, at least during the 00:00-06:00 h watch. Presumably, initially high levels of activity are replaced by routine and even boredom.