Objectives Our aim was to study the relationship between spontaneous abortions among nurses working three different types of shift schedules; permanent day shift, three shift rotation, and permanent night shift.

Methods From a cohort of nurses in Norway established in 2008, we identified 1083 female workers who worked the same type of shift schedule in 2008 and one year later. Information on age, years worked as a nurse, weekly work hours, job demands, job control, smoking, consumption of alcohol, caffeine, spontaneous abortions the past year and throughout life was obtained by questionnaires. The relationship between spontaneous abortion and type of shift schedule was analysed by logistic regression analyses, adjusting for job strain and life-style factors. The analyses were performed in two strata, above and below 30 years of age.

Results An increased risk for experiencing spontaneous abortions the past year was found among nurses who worked only night shifts compared to those who worked only day shifts, restricted to nurses below 30 years of age, odds ratio (OR) 3.4 (95% confidence interval [CI] 1.0–12.4). A similar increase in risk for experiencing spontaneous abortions throughout life was found among permanent night-shift workers, OR 4.4 (95% CI 1.2–16.3), also in this case for nurses below 30 years of age. No increased risk of spontaneous abortions was found among nurses above 30 years of age.

Conclusions The findings suggest that night work may cause spontaneous abortion by disrupting the circadian rhythms, but other unknown mechanisms may also play a role. More studies of night-shift workers considering different age groups are needed to supplement the findings.