Scand J Work Environ Health 2013;39(4):325-334    pdf full text


Miscarriage and occupational activity: a systematic review and meta-analysis regarding shift work, working hours, lifting, standing, and physical workload

by Bonde JP, Jørgensen KT, Bonzini M, Palmer KT

Objective Previous studies have indicated that shift work, long working hours, and prevalent workplace exposures such as lifting, standing, and physical workload increase the risk of miscarriage, but the evidence is conflicting. We conducted a systematic review of original research reports.

Methods A search in Medline and EMBASE 1966–2012 identified 30 primary papers reporting the relative risk (RR) of miscarriage according to ≥1 of 5 occupational activities of interest. Following an assessment of completeness of reporting, confounding, and bias, each risk estimate was characterized as more or less likely to be biased. Studies with equivalent measures of exposure were pooled to obtain a weighted common risk estimate. Sensitivity analyses excluded studies most likely to be biased.

Results Working fixed nights was associated with a moderately increased risk of miscarriage (pooled RR 1.51 [95% confidence interval (95% CI) 1.27–1.78, N=5), while working in 3-shift schedules, working for 40–52 hours weekly, lifting >100 kg/day, standing >6–8 hours/day and physical workload were associated with small risk increments, with the pooled RR ranging from 1.12 (3-shift schedule, N=7) to 1.36 (working hours, N=10). RR for working hours and standing became smaller when analyses were restricted to higher quality studies.

Conclusions These largely reassuring findings do not provide a strong case for mandatory restrictions in relation to shift work, long working hours, occupational lifting, standing, and physical workload. Considering the limited evidence base, however, it may be prudent to advise women against work entailing high levels of these exposures and women with at-risk pregnancies should receive tailored individual counseling.

This article refers to the following texts of the Journal: 2011;37(3):227-236  1989;15(5):345-352  1998;24 suppl 3:28-34
Download additional material