Objectives To compare results and assess the quality of recently published meta-analyses of night shift work and breast cancer risk.

Methods A comprehensive search was conducted for English or French-language meta-analyses published from 2010–2017 that included at least one meta-risk estimate (mRE) for breast cancer associated with any night shift work exposure metric and that were accompanied by a systematic literature review. mREs from each meta-analysis were ascertained and organised by various study characteristics. Assessments of heterogeneity and publication bias were also extracted. An eight-point checklist was used to evaluate quality.

Results Seven meta-analyses, published from 2013–2016, collectively included 30 cohort and case-control studies spanning 1996–2016. Five meta-analyses scored ≥6 points on the quality assessment checklist. Of these, mREs for ever/never night shift work exposure ranged from 1.15 (95% confidence interval [CI]: 1.05–1.25, n=9 studies) to 1.40 (95% CI: 1.13–1.73, n=9 studies). In these 5 reports, mREs for duration, frequency, and cumulative night shift work exposure were inconsistent. Meta-analyses of cohort, Asian, and more fully-adjusted studies generally resulted in lower mREs than case-control, European, American, or minimally-adjusted studies. Most used random effects models due to statistically significant between-study heterogeneity. Publication bias was not evident in any of the 5 meta-analyses.

Conclusions Substantial heterogeneity is to be expected in epidemiological studies done in various settings, and among diverse populations. Future evaluations of shift work carcinogenic potential need to incorporate high quality meta-analyses that better assess and account for individual study quality.