285 Pilot field study of Ambulatory Sleep-Staging in Shift-Working Air Traffic Controllers

Abstract

Abstract Introduction The simplicity of wrist actigraphy for sleep-wake monitoring in the field contributes to its ubiquity in shift-work research. However, devices based solely on recording activity levels are generally not suitable to quantify sleep architecture. This is a limitation as quantifying changes in sleep stages caused by circadian misalignment is important to better assess the consequences of sleep-wake disruption in shift-working populations. This pilot study was conducted to evaluate whether sleep stages vary with respect to different shift types. Methods Six male air traffic controllers aged 48.5±8.4 years (mean±SD) completed the protocol which entailed two ~9-day periods, each with up to 6 workdays. Schedules comprised 1 or 2 early night shifts (19:30–03:30h), followed by an evening shift (15:00–23:00h), day shift (09:00–17:00h), morning shift (06:30–14:30h), and 1 or 2 full night shifts (23:00–7:00h). A portable sleep-staging device (Somno-Art, Paris, France) that monitored activity levels and heart rate was worn on the non-dominant forearm during bedtime and produced estimates of REM and NREM sleep stages with a proprietary algorithm. Total sleep time (TST) and sleep stages were assessed per shift type with mixed-effects models. Results Final analyses were based on 70 sleep periods preceding workdays, standardized to 24 h to account for the different intervals between consecutive shifts. Analyses revealed significant effects of shift type for TST (p=.016), stages N1 (p=.010) and N2 (p=.043), but none for N3 (p=.055) or REM (p=.117) sleep. TST and stage N1 sleep prior to night shifts was shorter than for day, evening, or early night shifts (all p<.05). Participants obtained less stage N2 sleep prior to night shifts than days shifts (p=.049). Conclusion This pilot study suggests variations in TST across shifts were predominantly due to differences in light sleep stages, whereas no significant differences in N3 and REM sleep were observed. Thus, while TST was reduced for night shifts, participants obtained similar durations of the most recuperative stages. These findings highlight the importance of refined monitoring of sleep in field research involving shift-work. Support (if any) Project funded by NAV Canada. Devices lent by the Somno-Art company. A.K. received a postdoctoral fellowship from the Fonds de Recherche en Santé du Québec (FRQS).