Title, Date of Publication & Journal

Increased risk of breast cancer associated with long-term shift work in Canada

Occupational and Environmental Medicine

July 1^st 2013

Claim supported by evidence?

This paper provides some evidence for an association between long-term night shift working and risk of breast cancer but presents what is likely to be an over-estimate of the association. The paper does not establish a causal link.

Summary

• This is a case control study (a type of observational study) carried out in Canada to examine whether working night shifts is linked to an increased risk of breast cancer.
• Night shift working was determined from self-reported data on occupational history; in the main analysis a night shift job was defined as one in which >50% of time was spent working evening and/or night shifts.
• Different definitions of night shift working were used to see what effect this had on the results.

Study Conclusions

The main conclusion is that working night shifts for 30 years or more is associated with an increased risk of breast cancer.

However there was no evidence that working night and/or evening shifts for up to 29 years increased the risk of breast cancer.  The increased risk was only seen among women who had worked night and/or evening shifts for 30 years or more.  If there had been an increasing trend with increased exposure to night shift working, this would have added some strength to the finding. As it is, the fact that an increased risk was only seen in the small group who’d worked night shifts for 30 years or more makes it more likely that this elevated risk could be due to other (confounding) factors.

The main results presented are only adjusted for age and centre – see below under strengths and limitations for comments on this.

When the definition of night shift included only those where the shifts started or ended between 11pm and 7am, the association with breast cancer became weaker (the odds ratio was reduced from 2.21 to 1.68).

Strengths/Limitations

This is a well-designed study which has collected detailed information about the main exposure as well as information on a large number of potential confounding factors.

Despite this, the main analysis presented (the analysis from which the odds ratio of 2.21 given in the abstract comes from) is only adjusted for age and centre. The authors justify this by saying that adjustment for these factors did not individually change the estimated odds ratios by more than 10%. However, these confounding factors could potentially have a large combined effect on the odds ratio; indeed, the results (presented in Table 2) stratified by menopausal status and adjusted for BMI suggest that this is very likely to be the case. (The overall odds ratio just adjusted for these two factors would be somewhere in the region of 1.5, as opposed to 2.21).

The authors discuss possible sources of bias inherent in case control studies. However, as they have concluded, it seems unlikely that this will have had a major impact on the results.

Glossary

Case control study: an observational study in which individuals with the disease of interest (cases) are selected and compared to individuals without the disease (controls).

Confounding factors (confounders): a factor associated with the outcome (disease) of interest as well as the exposure being studied which, if not accounted for, may lead to over or under-estimation of the association between the exposure and outcome.