Variation in sleepiness during early morning shifts: a mixed model approach to an experimental field study of train drivers

The present study aimed to experimentally evaluate the effect of early morning shifts on sleep and sleepiness of train drivers during normal working conditions. A total of 17 experienced train drivers were studied during a 4.5 h drive in two directions with a 2.5 h break in between on three different shifts: an early shift that started at 05:49 h (train left at 06:18 h) and ended at 17:41h, a day shift (07:49-19:41 h), and an evening shift (09:49-21:41 h). Retrospective (since the last stop) ratings of mean sleepiness and peak sleepiness (Karolinska Sleepiness Scale--KSS: 1 = very alert, 9= very sleepy, fighting sleep, difficulty staying awake) were assessed at each stop during the drive. The results showed that sleep length was reduced (p <0.001) by 1 h and 2h, respectively, by the early shift compared to the day and evening shifts. The prevalence of severe sleepiness (KSS > or = 7) was high, especially during the early shift when 14 (82%) subjects reported at least one event during the drive. Application of the Generalized Linear Mixed Models (GLMM) to the sleepiness data showed that there was an increased risk for severe sleepiness during the early shift (OR = 4.9) that increased further with the length of the drive between stops (OR = 1.9, 15 min), suggesting an interaction between early morning shift and monotony. The findings have practical implications in risk assessment. Long drives without stops and other monotonous situations should have a higher risk rating for severe sleepiness in shifts with an early start before 06:00 h, compared to shifts that begin 2 h later.