Objective This study aimed to investigate the effects of various aspects of night and shift work regarding incident cerebrovascular disease (CeVD).

Methods The cohort included 26 667 women and 3793 men (nurses and nursing assistants) who were employed for at least one year 2008–2016 in Region Stockholm, Sweden. Information about the cohort and working hours were obtained from a computerized employee-register and diagnoses were retrieved from national and regional registers. We used discrete time proportional hazard models to assess the risk of CeVD (2009–2017), in relation to work hour characteristics, adjusting for sex, age, country of birth, education and profession.

Results We observed an excess risk of CeVD (N=223) among employees who, during the preceding year, worked night shifts >30 times [hazard ratio (HR) 1.44, 95% confidence interval (CI) 1.04–1.99] or ≥3 consecutive night shifts >15 times (HR 1.69, 95% CI 1.18–2.42) or with >30 quick returns (<28 hours) from night shifts (HR 1.52, 95% CI 1.10–2.10) compared to those who did not work nights. We also observed an excess risk among employees with a long duration (>5 years) of exposure to night shift work (HR 1.87, 95% CI 1.27–2.77), all supported by a dose–response pattern.

Conclusions Our results show that the risk of CeVD among nurses and nursing assistants is associated with night shift work. The number of years with night shift work, the frequency of night shifts per year, the frequency of consecutive night shifts, and short recovery after night shifts influenced the risk. Work schedules aiming at minimizing these aspects of night shift work may reduce the risk.

Key terms cerebrovascular disease; healthcare; night work; occupational exposure; occupational health; prospective cohort study; shift work; shift worker; Stockholm; stroke; Sweden

• Supplementary table