The aim of the present combined field and laboratory study was to assess circadian entrainment in two groups of police officers working seven consecutive 8/8.5-h night shifts as part of a rotating schedule. Eight full-time police officers on patrol (mean age ± SD: 29.8 ± 6.5 yrs) were provided an intervention consisting of intermittent exposure to wide-spectrum bright light at night, orange-tinted goggles at sunrise, and maintenance of a regular sleep/darkness episode in the day. Orange-tinted goggles have been shown to block the melatonin-suppressing effect of light significantly more than neutral gray density goggles. Nine control group police officers (mean age ± SD: 30.3 ± 4.1 yrs) working the same schedule were enrolled. Police officers were studied before, after (in the laboratory), and during (ambulatory) a series of seven consecutive nights. Urine samples were collected at wake time and bedtime throughout the week of night work and during laboratory visits (1 × /3 h) preceding and following the work week to measure urinary 6-sulfatoxymelatonin (UaMT6s) excretion rate. Subjective alertness was assessed at the start, middle, and end of night shifts. A 10-min psychomotor vigilance task was performed at the start and end of each shift. Both laboratory visits consisted of two 8-h sleep episodes based on the prior schedule. Saliva samples were collected 2 × /h during waking episodes to assay their melatonin content. Subjective alertness (3 × /h) and performance (1 × /2 h) were assessed during wake periods in the laboratory. A mixed linear model was used to analyze the progression of UaMt6s excreted during daytime sleep episodes at home, as well as psychomotor performance and subjective alertness during night shifts. Two-way analysis of variance (ANOVA) (factors: laboratory visit and group) were used to compare peak salivary melatonin and UaMT6s excretion rate in the laboratory. In both groups of police officers, the excretion rate of UaMT6s at home was higher during daytime sleep episodes at the end compared to the start of the work week (p < .001). This rate increased significantly more in the intervention than control group (p = .032). A significant phase delay of salivary melatonin was observed in both groups at the end of study (p = .009), although no significant between-group difference was reached. Reaction speed dropped, and subjective alertness decreased throughout the night shift in both groups (p < .001). Reaction speed decreased throughout the work week in the control group (p ≤ .021), whereas no difference was observed in the intervention group. Median reaction time was increased as of the 5th and 6th nights compared to the 2nd night in controls (p ≤ .003), whereas it remained stable in the intervention group. These observations indicate better physiological adaptation in the intervention group compared to the controls.
