Impact of reduced night work on shift workers' sleep using difference-in-difference estimation.

Abstract

We used a quasi-experimental design to investigate the impact of a change in the shift work system by reducing overnight work on the sleep health of workers. A difference-in-difference (DID) analysis was used to investigate changes in sleep time and sleep quality between the experimental group (shift workers, N = 116 in 2007 and N = 118 in 2013) and the control group (regular day workers, N = 256 in 2007 and N = 185 in 2013) before and after a change in the shift system that eliminated overnight work. The sleep outcomes were measured using a questionnaire that inquired about sleep duration, mid-sleep awakenings, and subjective sleep quality. We used a generalized estimating equation model to investigate differences in the prevalence of sleep-related outcomes between baseline and post-intervention. In the DID models, the sleep time per day (+0.5 h), the prevalence of awakening during sleep (-13.9%), and self-reported poor sleep quality (-34.9%) were improved with statistical significance during evening shifts in the experimental group, but there was no significant change during day shifts in the experimental group compared to the control group after the implementation of the new shift system without overnight shift. Quitting overnight work improved sleep health of shift workers.