446 Night work and common mental disorders: analysing observational data as a non-randomised pseudo trial

Abstract

Introduction This paper investigates the relationship between working regularly in night shifts for long periods and common mental disorders, including depression, anxiety disorders, severe stress, adjustment disorders, neurotic disorders and insomnia. The investigation of this association can help position intervention policies concerned with regulations of work schedule and workers’ mental health. Methods A non-randomised pseudo trial was conducted on two populations of social and healthcare employees (the Finnish Public Sector Study) for a duration of eight years. Study one tested the effect of moving from continuous day work to work in night shifts on the development of common mental disorders for 30 019 employees and study two examined the recovery rate from mental disorders when 944 night workers with common mental disorders changed their working schedule to day shifts. Data were analysed using logistic regression with GEE. Result Study one showed a 3% higher odds of acquiring common mental disorders among night workers compared to day workers; (odds ratio 1.03, 95% CI: 0.82 to 1.30) whereas study two showed a 99% higher odds of recovery from common mental disorders among night workers who changed their work schedule to day shifts compared to night workers who continued to work night shifts (odds ratio 1.99, 95% CI: 1.20 to 3.28). Both results were adjusted for several covariates. Discussion Our paper introduces a novel approach to analyse observational repeat data on night work which is closer to randomised controlled trials than conventional epidemiologic analyses. The results from study one could not support a causal association between night work and the development of common mental disorders. However, study two shows increased recovery rates from common mental disorders when night workers change their work schedule back to day work. These findings suggest that change in work schedule may have a role in the prognosis of common mental disorders.