Cigarette smoking and the risk of epithelial ovarian cancer

Abstract

Cigarette smoking may affect each of the currently proposed mechanisms of ovarian carcinogenesis. Whether cigarette smoking has any effect on the development of ovarian cancer has not been adequately evaluated. To study this issue, the authors examined data from the Cancer and Steroid Hormone Study, a multicenter, case-control study of gynecologic cancers conducted between December 1, 1980, and December 31, 1982, in eight geographic areas of the United States. This analysis utilized data on 494 women with newly diagnosed epithelial ovarian cancer and 4,238 population-based control women 20-54 years of age. There was no association of epithelial ovarian cancer with dose of cigarette smoking, age smoking started, time since smoking started, or time since smoking last occurred. Simultaneous adjustment for age, parity, history of oral contraceptive use, and other potentially confounding factors did not alter these results.

PIP: Data from the Cancer and Steroid Hormone study was used to evaluate the effect o cigarette smoking on the risk of developing epithelial ovarian cancer. This multicenter, population-based case-control study of oral contraceptive (OC) use and ovarian, breast, and endometrial cancer enrolled women between December 1, 1980, and December 31, 1982, in 8 geographic areas: Atlanta, Detroit, San Francisco, and Seattle; the states of Connecticut, Iowa, and New Mexico; and the 4 urban counties of Utah. Eligible cases were women 20-54 years old first diagnosed as having ovarian cancer of any histologic type which was ascertained through population-based tumor registries in the above-mentioned areas during the study interval. Interviews were completed with 579 of the eligible cases (71.0%). The study controls were women 20-54 years of age selected by telephoning randomly selected phone numbers of households in the same geographic areas as the cases. 4754 of those selected were interviewed. A standard questionnaire was administered to participants in their homes by trained interviewers. Women who had never smoked a total of 100 cigarettes to be nonsmokers in this analysis. Age, parity, and ever-use of OCs (for 3 or more consecutive months) were considered to be potentially confounding factors because they are known to be associated with smoking. Women with epithelial ovarian cancer were more likely than controls to be white, nulliparous, to have used OCs, and to have had a natural menopause. Women who had ever smoked cigarettes had the same risk of epithelial ovarian cancer as women who had never smoked; this was the case when current smoking and past smoking were considered. There was a slightly reduced risk of ovarian cancer among women who had stopped smoking 10 or more years earlier, but the association was not statistically significant. Cumulative lifetime exposure to cigarette smoking categorized by increasing pack-years showed no statistically significant dose effect. Among smokers, no significant linear trend was present when pack-years was used as a continuous variable. No effect of latency was found. The age that a woman began smoking had no effect on ovarian cancer risk. Stratification of the data according to age, race, education, parity, OC use, infertility, noncontraceptive estrogen use, menopausal status, alcohol use, obesity, and family history of ovarian cancer did not reveal any appreciably different effects of smoking on ovarian cancer risk in different subgroups of women. Likelihood ratio tests revealed no statistically significant interactions.