Circadian disrupting exposures and breast cancer risk: a meta-analysis

Abstract

Purpose: Shift work, short sleep duration, employment as a flight attendant, and exposure to light at night, all potential causes of circadian disruption, have been inconsistently associated with breast cancer (BrCA) risk. The aim of this meta-analysis is to quantitatively evaluate the combined and independent effects of exposure to different sources of circadian disruption on BrCA risk in women.

Methods: Relevant studies published through January 2014 were identified by searching the PubMed database. The pooled relative risks (RRs) and corresponding 95 % confidence intervals (CIs) were estimated using fixed- or random effects models as indicated by heterogeneity tests. Generalized least squares trend test was used to assess dose-response relationships.

Results: A total of 28 studies, 15 on shift work, 7 on short sleep duration, 3 on flight attendants, and 6 on light at night were included in the analysis. The combined analysis suggested a significantly positive association between circadian disruption and BrCA risk (RR = 1.14; 95 % CI 1.08-1.21). Separate analyses showed that the RR for BrCA was 1.19 (95 % CI 1.08-1.32) for shift work, 1.120 (95 % CI 1.119-1.121) for exposure to light at night, 1.56 (95 % CI 1.10-2.21) for employment as a flight attendant, and 0.96 (95 % CI 0.86-1.06) for short sleep duration. A dose-response analysis showed that each 10-year increment of shift work was associated with 16 % higher risk of BrCA (95 % CI 1.06-1.27) based on selected case-control studies. No significant dose-response effects of exposure to light at night and sleep deficiency were found on BrCA risk.

Conclusions: Our meta-analysis demonstrates that circadian disruption is associated with an increased BrCA risk in women. This association varied by specific sources of circadian disrupting exposures, and a dose-response relationship remains uncertain. Therefore, future rigorous prospective studies are needed to confirm these relationships.