Monday, January 30, 2006

California Report Concludes Secondhand Smoke Causes Breast Cancer in Pre-Menopausal Women; Major Implications for Waitresses and Female Bartenders

The California Air Resources Board has declared secondhand smoke to be a toxic air pollutant, based on a comprehensive review of the scientific evidence linking secondhand smoke to disease prepared by the Office of Environmental Health Hazard Assessment (OEHHA) of the California Environmental Protection Agency (Cal-EPA).

The review concluded, for the first time, that secondhand smoke is a cause of breast cancer among pre-menopausal women. According to the report: "Overall, the weight of evidence (including toxicology of tobacco smoke constituents, epidemiological studies, and breast biology) is consistent with a causal association between ETS exposure and breast cancer in younger, primarily premenopausal women."

The Rest of the Story

In my opinion, OEHHA provides a thorough, comprehensive, and thoughtful review of the evidence linking secondhand smoke to breast cancer and a compelling conclusion that there is a causal relationship between secondhand smoke exposure and breast cancer among pre-menopausal women (but not post-menopausal women).

This may seem surprising to many, and in fact, I myself was skeptical of the findings when originally reported and challenged by groups such as the American Cancer Society. Although I try to be quite careful in drawing causal conclusions and despite the fact that I was skeptical myself about this particular conclusion, I have to state that I am quite convinced by the evidence at this point and I feel that a causal conclusion is warranted.

There were 2 major reasons why the findings of the Cal-EPA report were questioned. First, there was inconsistency in the link between secondhand smoke and breast cancer in epidemiologic studies. Second, people asked how it could be plausible for secondhand smoke to cause breast cancer if active smoking did not cause breast cancer. And two major bodies - the U.S. Surgeon General (2004) and the International Agency for Research on Cancer (IARC) had concluded that active smoking is not a cause of breast cancer.

It turns out that there appears to be a convincing explanation for both of these phenomena, and I think the OEHHA report makes a compelling argument for why these concerns can now be dismissed.

First, many of the epidemiologic studies that examined the link between secondhand smoke and breast cancer were plagued by the problem of non-differential misclassification of exposure. Because nonsmokers are likely to be exposed to secondhand smoke, studies which compared breast cancer risk in smokers to those in nonsmokers were really comparing breast cancer risk in smokers to risk among passive smokers and non-exposed nonsmokers.

Because the magnitude of the association between smoking and breast cancer observed in many studies is small and is not much greater than that observed between secondhand smoke and breast cancer, this bias was likely to knock out most (if not all) of the effect of smoking on breast cancer risk. And in fact, OEHHA demonstrated that in studies where nonsmokers with passive smoke exposure were specifically excluded, there was a clear finding of an elevated risk of breast cancer among smokers.

In fact, studies of the relationship between smoking and breast cancer which were more accurate in assessing potential secondhand smoke exposure in women yielded a higher estimate of breast cancer risk associated with smoking than those which were not as accurate in being able to exclude passively exposed women. This suggests that exposure misclassification was a major problem in much of the literature. Taking this into account yields a quite consistent finding of an increased breast cancer risk associated with active smoking.

A similar problem occurred with respect to the secondhand smoke studies. Many nonsmokers who stated that they were not exposed to secondhand smoke actually had considerable exposure. The control group here was not truly non-exposed. Thus, the results were biased toward finding no effect, and the risk estimates in much of the epidemiologic studies were underestimated. When exposure misclassification is taken into account, the OEHHA report demonstrates a quite consistent and clear finding of increased breast cancer risk among women exposed to secondhand smoke.

Second, while the 1994 Surgeon General's report considered only 5 studies published since 2000 and the IARC report considered only 4 studies published after 2000, the OEHHA review considered 23 studies published between 2000 and 2005. Thus, the data reviewed by both the Surgeon General and the IARC were similar to that considered in the original (1997) Cal-EPA report, which in fact concluded that evidence was not sufficient to definitively link active smoking or passive smoking with breast cancer. There is, then, no inconsistency between the findings of Cal-EPA and the findings of either the Surgeon General or IARC. And clearly, the most up to date information is what one would want to use to draw any conclusions on this matter.

Third, most of the prior reviews of the evidence on smoking and breast cancer lumped pre-menopausal and post-menopausal breast cancer together. Since these represent somewhat different entities with, possibly, different etiologies, such a procedure probably obscured the observed relationship between pre-menopausal breast cancer and tobacco smoke exposure by mixing these cases with post-menopausal breast cancer, which does not appear to be related at all to smoke exposure.

The absence of a link between active smoking and breast cancer can no longer, in my view, be used to argue that a similar link between passive smoking and breast cancer is not plausible, because there is strong evidence that active smoking is a cause of breast cancer.

And the inconsistency of OEHHA's findings with those of the U.S. Surgeon General and IARC cannot be used to argue against the report's conclusion because there is no inconsistency. The OEHHA report includes a huge amount of literature not reviewed by either the Surgeon General or IARC.

Several other factors play a role in my conclusion that the evidence presented by OEHHA is compelling enough to conclude that secondhand smoke causes pre-menopausal breast cancer.

First, the Cal-EPA did not, in its 1997 report, conclude that secondhand smoke causes breast cancer. That report stated that the evidence was not sufficient to draw such a conclusion. That adds credibility to the Agency's conclusion at this time, because obviously, the Agency has demonstrated that it is going to draw conclusions based on the science.

Second, the problem of confounding is likely to work in the opposite direction (to bias results toward the null hypothesis of no effect of secondhand smoke on breast cancer). This is because unlike many other diseases, including many cancers, breast cancer is a disease that is not particularly associated with lower socioeconomic status or lower social class. Incomplete control for potential confounders in this situation might actually be expected to result in an under-estimate of the true effect, since women exposed to secondhand smoke are more likely to be in a lower socioeconomic status group.

I feel that chance, bias, and confounding are not plausible alternative explanations for the observed increase in breast cancer risk among nonsmokers exposed to secondhand smoke.

Chance cannot explain the findings because the probability of having 13 of 14 studies all find an elevated risk of breast cancer among exposed nonsmokers is miniscule, a clear dose-response effect is present, and studies which more accurately establish exposure status have higher risk estimates.

Bias cannot explain the observed findings because the major bias - non-differential misclassification of exposure - would bias the results toward the null hypothesis of no effect.

And confounding cannot explain the findings because of the consistency of findings, the magnitude of the observed risk, the specificity of the elevated risk to pre-menopausal women, the higher risks observed in studies with better classification of exposure, and the fact that residual confounding for this particular exposure and disease would probably be expected to bias the results toward the null hypothesis of no effect.

There is, in my view, no reasonable alternative explanation for the observed finding of an increased risk of breast cancer among young women exposed to secondhand smoke.

This conclusion has important implications because it means that we now have an identified and preventable cause of breast cancer. Breast cancer organizations must take account of this. There has certainly been a lot of speculation about all kinds of environmental exposures that could possibly be linked to breast cancer, but now we have one identified cause that is easily preventable.

It also means that waitresses and female bartenders, due to their high levels of secondhand smoke exposure, are at increased risk of breast cancer, in addition to lung cancer and heart disease. This makes it even more important to protect these women from secondhand smoke exposure. I'm not arguing here that without the breast cancer evidence, there was not sufficient evidence to provide protection for bar and restaurant workers. I simply feel that this adds additional support for the need to provide this protection to all workers, but especially those who are most heavily exposed: those who work in bars and restaurants, and, I would add, casinos.

No comments: