Chronotype in relation to shift work: A cohort study among 37,731 female nurses

Abstract

SummaryChronotype may affect tolerance for circadian disruption induced by shift work. This study examines the association between chronotype, self‐reported sleep timing, shift type preference, and sleep problems among nurses, and studies chronotype stability over time. The study included 37,731 Dutch female nurses who completed a baseline (2011) and follow‐up questionnaire (2017), with information on shift work (e.g., job history, shift type preference [collected in 2017 only]), and sleep characteristics (e.g., chronotype, preferred sleep–wake time in a work‐free period [collected in 2017 only], and sleep problems between working days according to Medical Outcomes Study‐Sleep Problem Index II [MOS‐SPI‐II]). The association between chronotype and sleep timing was examined using (age‐adjusted) linear regression. Associations between chronotype and shift type preference and sleep problems (MOS‐SPI‐II >30) were examined using ordered logistic and Poisson regression, respectively. With later chronotype, midsleep time increased (definite evening vs. intermediate types [reference]: β = 55 min, 95% confidence interval [95% CI]: 54–55), the odds ratio (OR) for 1‐point increase in preference for night (2.68; 95% CI: 2.48–2.90) and evening shifts increased (OR 2.20; 95% CI: 2.03–2.38), while the odds for day (OR 0.17; 95% CI: 0.16–0.18) and morning shifts (OR 0.22; 95% CI: 0.21–0.24) decreased. Intermediate chronotype was associated with fewer sleep problems (median MOS‐SPI‐II = 27.2, p < 0.01), compared with definite morning (28.9) and evening types (31.7). This study shows that chronotype is associated with sleep–wake times in a work‐free period, shift type preference, and sleep problems in nurses. Future studies on the association of shift work‐induced circadian disruption and health outcomes should therefore consider chronotype as effect‐modifier.