Sleepiness in various shift combinations of irregular shift systems.

The present study examined the occurrence of sleepiness in various shift combinations ending with a night or morning shift. Three weeks' sleep/work shift diary data, collected from 126 randomly selected train drivers and 104 traffic controllers, were used in statistical analyses. The occurrence of sleepiness at work (i.e., Karolinska Sleepiness Scale 7 or higher) was tested with a generalised linear model with repeated measurements including explanatory factors related to shifts, sleep, and individual characteristics. The prevalence of severe sleepiness varied between 25% and 62% in the combinations ending with a night shift and between 12% and 27% in the combinations ending with a morning shift. The occurrence of sleepiness did not, however, systematically vary between the shift combinations in either case. An increased risk for sleepiness was associated with high sleep need and long shift duration in the night shift and with high sleep need, short main sleep period, long shift duration and an early shift starting time in the morning shift. Also having a child was associated with an increased risk for sleepiness in the night shift. The results suggest that the shift history of 24-36 h prior to the night and the morning shift is not strongly associated with the occurrence of sleepiness at work, but there are other factors, such as shift length and starting time and sleep need, that affect a risk for sleepiness at work.