“Second-hand tobacco smoke as a child or adult appears to increase the risk of breast cancer,” reports the BBC News.
The news is based on a large, well-conducted study among almost 80,000 women. It assessed their lifetime smoking habits and exposure to passive smoking, following them to see who developed breast cancer in the years that followed.
Researchers found that breast cancer risk had statistically significant links to both active smoking and inhaling the highest levels of passive smoke. However, there was only an unclear suggestion of a link when women exposed to passive smoke were considered as a whole.
The lack of a clear link with passive smoking overall is unsurprising and may be due to the difficulty in remembering past exposure, particularly exposure during childhood. Despite some minor limitations and the researcher’s cautious interpretation of their passive smoking results, the public health message remains clear and uncontroversial: smoking and smoking around other people, including children, has a range of harms, and an increased breast cancer risk is likely to be among those harms.
Where did the story come from?
This US study was part of the Women’s Health Initiative, which is funded by the National Heart, Lung and Blood Institute, National Institutes of Health, and the US Department of Health and Human Services. It was carried out by researchers in locations across the US, including Morgantown, Minneapolis, Buffalo, West Virginia, New York and California. The study was published in the peer-reviewed British Medical Journal.
With the exception of framing this study as controversial, the BBC has presented the details accurately and quoted experts highlighting the public health messages that have been further strengthened by this study. They say that "If we want to have a real impact on cancer deaths then we have to prevent children from starting to smoke and give smokers all the support they need to quit".
What kind of research was this?
In this prospective cohort study the researchers set out to examine the association between smoking and risk of invasive breast cancer using previously collected data on postmenopausal women’s lifetime exposure to both passive and active smoking. They were particularly interested in the passive or second-hand smoking link. The authors explain that these type of prospective studies, featuring detailed assessment of passive smoking, are needed to further explore the link. This is because so far only one such study has included sufficiently detailed information to enable lifetime exposures to be quantified.
The study collected very detailed information on the length of time people were exposed to second-hand smoke. It also included quantitative measures of exposure in childhood and adult exposure in residential and workplace settings. Together with the size and wide geographical cover of the study, the research provides the best evidence yet for this link.
What did the research involve?
The researchers analysed data gathered from 40 clinical centres in the US between 1993 and 1998. The data was collected as part of the Women’s Health Initiative Observational Study, a long-running study set up to address major causes of ill-health and death in postmenopausal women. It is from this study, for example, that the evidence regarding hormone replacement therapy and increased breast cancer risk was detailed. In total, 93,676 women aged 50–79 were recruited.
The researchers excluded those who did not plan to live locally or those whose predicted survival was less than three years. They also excluded 12,075 women who had already suffered cancer before the study started and about another 1,500 who were either lost at follow-up or had missing data. This left 79,990 women for further analysis.
All the information on smoking and other health data for the analysis was collected by questionnaire at the start of the study. The questionnaire asked about active and passive smoking, including questions on how much passive exposure to cigarette smoke the participants remembered from childhood and adult life, both in the home and at work. Collecting data on factors such as age at which women started or stopped smoking enabled the researchers to calculate the “pack years of smoking”. They did this by multiplying the total years of smoking by the number of cigarettes smoked a day, divided by twenty (the number of cigarettes in a typical pack).
Breast cancer cases were first identified by self report in the annual follow-up questionnaire mailed to participants. They were then confirmed by review of medical records, including pathology reports. They only recorded cases of invasive breast cancer (i.e. cancer that had the ability to spread outside the breast). They did not record cases of carcinoma in situ (an early form of cancer, which by definition has not invaded surrounding tissues).
The completion rate of annual questionnaires was 93%–96%.
The researchers analysed the data appropriately and adjusted the results for other important factors such as age, race, education, hormone therapy use, number of children, age of first live birth, alcohol intake, and family history of breast cancer.
What were the basic results?
The study ran for an average of 10.3 years of individual follow-up. During this time, 3,520 new cases of breast cancer were identified among 79,990 women analysed.
The researchers then compared the risks for smokers to the risks for women who had never smoked. After adjustments, breast cancer risk was:
- 9% higher among the former smokers (hazard ratio [HR] 1.09, 95% confidence interval [CI] 1.02 to 1.17)
- 16% higher among current smokers (HR 1.16, 95% CI 1.00 to 1.34)
- 35% higher, the biggest increase, among women who had smoked for 50 years or more (HR ratio 1.35, 95% CI 1.03 to1.77) compared with whole lifetime non-smokers.
For passive smokers, after adjustment for potential confounders, the risk for the combined group with the most extensive exposure to passive smoking (10 years or more exposure to smoke in childhood, 20 years or more as an adult at home, or 10 years or more as an adult at work) was:
- 32% higher compared with those who had never been exposed to passive smoking (HR 1.32, 95% CI 1.04 to 1.67).
There was no significant association in the other groups with lower passive smoking exposure. There was also no clear dose response to overall (cumulative) passive smoking exposure (i.e. no relationship where increasing exposure led to a raised risk).
How did the researchers interpret the results?
The researchers conclude that “active smoking was associated with an increase in breast cancer risk among postmenopausal women”. They go on to say that there was a suggestion of a link between passive smoking and an increased risk of breast cancer, but that since the risk was increased only in the most extensive passive smoking category with no clear dose response, this link should be considered suggestive only, and will need confirmation in other studies.
This well-conducted analysis of cohort data has confirmed the increased risk of breast cancer among smokers, particularly those with high-intensity and long duration of smoking.
This is perhaps unsurprising. The increased risk of breast cancer among non-smokers with extensive exposure to passive smoking was also statistically significant. However, the lack of significant associations in other non-smokers with lower passive exposure, and the lack of a clear ‘dose-response’ link (where increasing passive smoking would lead to increasing risk), suggests more studies might help to identify underlying reasons for this. Such a dose-response relationship is usually required to confirm a causal link.
Other minor limitations to consider include:
- The measurement of smoking exposure once at the start of the study means that any change in smoking habits over the course of the study were not captured. This could lead to inaccuracies if some people quit but were not recorded correctly. The researchers estimate that 60% of smokers continued to smoke for six years of follow-up.
- It is possible that participants had difficulty remembering precise details of their exposure to passive smoking, particularly in childhood. This too may have led to misclassification and could have made it difficult to detect the dose-response link if one existed.
Despite these limitations and the researchers’ cautious interpretation in the passive smoking category, the public health message remains clear and uncontroversial: smoking and smoking around other people, including children, has a range of harms, and an increased breast cancer risk is likely to be among those harms.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
BBC News, 3 March 2011
Links to the science
BMJ 2011; 342:d1016