Can night shifts give you breast cancer?

“Night shifts 'cause 500 breast cancer deaths a year',” The Daily Telegraph reported. It said that nurses and flight attendants are the two occupations that tend to work nights most frequently.

The Telegraph’s story is based on a large project that has estimated how cancers affect people in different professions in Britain. It covered many different types of occupation that the International Agency for Research in Cancer has classified as definitely or probably associated with increased levels of cancer risk. The Telegraph, however, focused mainly on night shift working.

The analysis estimated that about 5% of cancer deaths in Britain in 2005 and 4% of reported cases of cancer in 2004 were attributable to occupation. It estimates that the additional breast cancer risk associated with night shift working would have translated into about 2,000 extra cases of breast cancer (out of a total of about 43,200 in Britain) in 2004, and about 550 extra breast cancer deaths in 2005. About 2 million women were estimated to have been exposed to shift (night) work during the risk period being assessed (1956 to 1996). The analysis estimates that the additional breast cancer risk associated with this relies on the accuracy of underlying assumptions and data, and these may produce uncertainty or bias the results.

It’s important to remember that analyses and estimates of this type are for the population as a whole. They do not mean that individual cancer cases can necessarily be singled out as being caused by the occupations. Rather, they suggest which factors might affect risk and estimate how much lower cancer rates could be in the population as a whole if those factors were removed.

Overall, this type of estimate helps policy makers and employers to identify which occupational exposures (types of working) are potentially causing the most harm and develop strategies to reduce these as much as possible. It does not tell us why night shift working may be linked with cancer.

If your job involves shift working, you should not be unduly worried by this study because the evidence of the risk of breast cancer from shift working is limited, and the reasons unclear. This analysis looks at risk at a population level rather than your individual circumstances.

Where did the story come from?

This story is based on a special issue in the peer reviewed British Journal of Cancer, which looks at cancer relating to people’s occupations in Britain. It was funded by a grant from the UK Health and Safety Executive (HSE), and carried out by various researchers who formed the British Occupational Cancer Burden Study Group. The summary results of this analysis were published in 2010 in the journal, and the current publications provide more details on the methods and results for different types of cancer. The full reports are being published on the HSE website.

The journal issue covers a range of occupational exposures that may increase the risk of cancer, although the news has focused mainly on shift work.

The methods and results are summarised in the foreword and introduction to the issue, and these are the focus of this Behind the Headlines analysis.

The Daily Telegraph mentioned that this research follows recently published Danish research suggesting that women working three or more night shifts a week for six years or more were doubling their chances of breast cancer. The current analyses combine the findings of studies such as this with the levels of exposure to these risk factors in Britain, to calculate what proportion of cases of cancer may be avoided by removing these risk factors.

What kind of research was this?

These reports are based on modelling analyses that looked at data from several different sources to work out how much occupation-related exposures contribute to overall risk of cancer in Britain. Earlier studies had made estimates of this burden, but the researchers wanted to update these estimates using more recent figures. The aim of the project is to develop appropriate practical ways to reduce health risks to workers arising from exposure to carcinogens in the workplace.

The approach is useful for estimating how many cases of cancer could be avoided by avoiding exposure to these hazards. This could inform efforts to prevent these exposures. It is important to note that these analyses and estimates are at the level of the population as a whole rather than individuals. These figures do not mean that individual cancer cases can necessarily be identified as being solely caused by the exposures in question; instead they are estimates of how much lower cancer rates would be in the population as a whole if the exposure were removed.

These estimates are based on multiple figures from various different sources, and also on certain assumptions, and therefore may not be an accurate reflection of the effects of occupational exposures.

What did the research involve?

The researchers looked at occupation-related exposures that had been classified by the International Agency for Research on Cancer (IARC) as “carcinogenic” or “probably carcinogenic” in humans in 2008, and as having “strong” or “suggestive” evidence for having an effect on specific types of cancer. The IARC is a World Health Organization body that regularly looks at the evidence about exposures that may cause cancer, and rates these exposures in terms of how convincing the evidence is and how carcinogenic the exposure seems to be to humans.

They estimated the effects of these exposures on risk of specific cancers based on published literature and by pooling the results of studies relevant to the UK. The researchers tried to obtain figures that estimated the effects of both low level and high level exposure. The figures took into account other factors that could influence results as needed.

The researchers were looking at the effects of cancer in the years 2004 and 2005. They assumed that it would take 10-50 years before the effects of the exposures on solid tumours would be seen, so they looked at exposures in 1956 to 1995. For blood cancers, they assumed that it would take 0-20 years before the effects of the exposures would be seen, so they looked at exposures between 1986 and 2005. Because of these time lags, they only looked at people aged over 25 years old in 2004/2005 for solid cancers, and at women aged 15-79 years and men aged 15-85 years in 2004/2005 for blood cancers.

The researchers also used national data sources to assess workforce exposure to the occupational hazards in the UK (the UK Carcinogen Exposure Database, the annual Labour Force Survey and the Census of Employment). Cancer deaths were assessed for 2005 and cancer reports for 2004.

The analyses used standard methods for determining what proportions of cancers are “attributable” to individual factors.

What were the basic results?

Overall, the analyses estimated that 8,010 (5.3%) total cancer deaths in Britain in 2005 and 13,598 (4%) reported cases of cancer in 2004 were attributable to occupation. The cancers most commonly related to occupation were mesothelioma (the tissue lining of organs), sinonasal, lung and bladder cancers and non-melanoma skin cancer for men, and mesothelioma, sinonasal, lung, breast and nasopharyngeal cancers for women.

The exposures responsible for the most occupation-related cancers in 2004 were:

• asbestos: 4,216 cases, mainly of lung cancer and mesothelioma
• shift work: 1,957 cases of breast cancer (this represented 4.5% of all reported cases of breast cancer)
• mineral oils: 1,730 cases, mainly of non-melanoma skin cancer and lung cancer
• solar radiation: 1,541 cases of non-melanoma skin cancer and lung cancer
• silica: 907 cases of lung cancer
• diesel engine exhaust: 801 cases, mainly of lung cancer

Other exposures responsible for between 100 and 500 cases of cancer were polycyclic aromatic hydrocarbons from coal tar and pitches, dioxins, environmental tobacco smoke encountered at work in non-smokers, radon exposure from natural exposure in workplaces, tetrachloroethylene (dry-cleaning fluid), arsenic and strong inorganic acid mists, as well as occupation as a painter or a welder.

Night shift working, which the news has focused on, is judged by the IARC to be probably carcinogenic in humans, based on “limited” evidence in humans and evidence from animal research that it increases the risk of breast cancer. The researchers used an estimate that breast cancer is 1.5 times more common in female night workers, which came from a pooling of studies. This previous research has suggested that the greatest increase in risk was in women who had been exposed to long periods of night shift employment. For details of one piece of related research, see the Behind the Headlines analysis of night shift cancer risk from 2009.

The authors estimated that 1,953,645 women had been exposed to shift (night) work between 1956 and 1996. They calculated that about 4.5% of breast cancer cases could be attributable to night shift exposure. This equated to 552 breast cancer deaths in 2005 and 1,957 breast cancer registrations in 2004 attributable to exposure to shift work at night.

How did the researchers interpret the results?

The foreword to the issue concluded that “the results should help the [Health and Safety Executive] develop an evidence-based approach for occupational cancer control”.

Conclusion

This analysis provides estimates of the numbers of cancer cases and deaths that may be attributable to occupational exposures in Britain. The researchers themselves note that their findings need to be considered in light of their limitations. These limitations included the fact that only substances classified by the IARC as definitely or probably carcinogenic were assessed. The effects of other “possibly” carcinogenic exposures was not assessed and could increase the effects of occupational exposures. The researchers also emphasised that the accuracy of the estimates relies on the accuracy of underlying assumptions and data, and these may produce uncertainty or bias the results. For example:

• In some cases where no British studies were available that estimated the effects of the exposures, data from other countries were used.
• In cases where figures on risk for women were not available, the risks for men were used.
• Assumptions were made about the level of exposure in different professions.
• Assumptions were made about how long an exposure would take to have an effect on cancer rates.

The researchers said that additional analyses are being carried out to assess the effects of these and other assumptions.

It’s worth noting that this research was looking at occupational exposures between 1956 and 1995 for solid cancers, and 1986 and 2005 for blood cancers. Occupational and other exposures in Britain may have changed since these dates. Also, these figures are based on analysis of the population as a whole rather than individuals. They do not mean that individual cancer cases can necessarily be identified as being solely caused by the exposures in question; instead they are estimates of how much lower cancer rates could be in the population as a whole if the exposure were removed.

The number of cases of breast cancer potentially attributable to night shift working is much smaller than the number of women who have worked night shifts. The researchers estimated that almost 2 million women had been "exposed" to night shift work in the period being assessed, with about 1,957 out of the 43,202 breast cancer cases reported in 2004 attributable to night shift exposure.

Overall, this type of estimate helps policy-makers and employers to identify which occupational exposures have the potential to cause the most harm, allowing them to develop strategies to reduce these exposures as much as possible.