© 2002 by Oxford University Press
Journal of the National Cancer Institute, Vol. 94, No. 4, 309-310,
February 20, 2002
© 2002 Oxford University Press
CORRESPONDENCE |
RESPONSE: Re: Sex, Smoking, and Cancer: a Reappraisal
Correspondence to: Thomas V. Perneger, M.D., Ph.D., Institute of Social and Preventive Medicine, University of Geneva, and Quality of Care Unit, Geneva University Hospitals, CH-1211 Geneva 14, Switzerland (e-mail: thomas.perneger{at}hcuge.ch).
I thank Drs. Risch and Miller, and Dr. Zang, for their comments. Both letters defend the substance of the published findings, i.e., that women smokers are in fact at higher risk of lung cancer than are male smokers. This may be true, but it is beside the point. My commentary was about rigor in interpreting data from epidemiologic studies, not about risk factors for lung cancer. The cited studies (1–4) observed that the relative risk of cancer associated with smoking is greater in women than in men but concluded that the risk of cancer is higher in women who smoke than in men who smoke. I questioned the logic of this argument. That the conclusion may be true in the case of lung cancer, as shown by other data such as incidence rates of lung cancer in nonsmokers of either sex, does not validate the flawed reasoning shown above.
Furthermore, the numerical example worked out in the commentary concerned bladder cancer, not lung cancer. At times, it appears as though the authors of these correspondences have applied statements that I made about bladder cancer also to lung cancer. This assumption is incorrect. Only the principles of adequate interpretation of risk statements have general applicability.
Dr. Zang attributes sex differences in the risk of bladder cancer to misclassification of smoking status. I do not believe that this may be true. The original analysis by Castelao et al. (4), which was based on relative risks, suggests that women are at higher risk of bladder cancer associated with smoking than are men (negative interaction between male sex and smoking), whereas examination of incidence rates indicates that the absolute risk of bladder cancer attributable to smoking is greater in men than in women (positive interaction between male sex and smoking). This seeming contradiction stems from different risk scales being used (multiplicative or additive)—hence, my recommendation that the risk scale be explicitly stated when describing statistical interactions. It has nothing to do with misclassification of exposure to smoking in men and women. In fact, as shown in Table 1 of the commentary, the incidence of bladder cancer is more than 3 times higher in men than in women, regardless of smoking status. A finer stratification by amount smoked would not change this.
I am reassured that all writers agree that risks of disease in men and women cannot be compared in a sex-matched case–control study. Nevertheless, their original interpretation of the findings from their studies (1–3), quoted in my commentary, implies otherwise, and as much can be said about the study by Castelao et al. (4). Perhaps, as a profession, we epidemiologists should pay attention not only to rigorous methodology but also to clear and careful communication of our results.
REFERENCES
1
Risch HA, Howe GR, Jain M, Burch JD, Holowaty EJ, Miller AB. Are female smokers at higher risk for lung cancer than male smokers? A case-control analysis by histologic type. Am J Epidemiol 1993;138:281–93.
2 Zang EA, Wynder EL. Cumulative tar exposure. A new index for estimating lung cancer risk among cigarette smokers. Cancer 1992;70:69–76.[CrossRef][Web of Science][Medline]
3
Zang EA, Wynder EL. Differences in lung cancer risk between men and women: examination of the evidence. J Natl Cancer Inst 1996;88:183–92.
4
Castelao JE, Yuan JM, Skipper PL, Tannenbaum SR, Gago-Dominguez M, Crowder JS, et al. Gender- and smoking-related bladder cancer risk. J Natl Cancer Inst 2001;93:538–45.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||