Abstract B114: Active smoking and postmenopausal breast cancer in a large prospective cohort with long-term follow-up

Introduction : The relationship of active cigarette smoking to breast cancer remains controversial, partly because of questions about residual confounding by alcohol and partly because of the absence of the strong dose-response relationships observed with other smoking-related cancers. Methods : We examined these relationships in the American Cancer Society9s Cancer Prevention Study-II Nutrition Cohort, initiated in 1992 with 180,000 cancer-free adults and followed through June, 2007. Detailed information on smoking histories and breast cancer risk factors was collected in 1992, 1997, and every two years thereafter. Median follow-up of 13.8 years identified 4,556 invasive breast cancer cases among 73,388 women at risk. Multivariable-adjusted Cox proportional hazard regression models were used with time dependent variables for smoking, use of exogenous postmenopausal hormones, and recent mammograms. Models were adjusted also for age, race, education, age at menarche, age at first birth, age at menopause, family history of breast cancer, history of breast cysts, alcohol consumption, and use of oral contraceptives. P-values for linear trend were calculated excluding the never smoker category. Results : Relative to never smokers (56.2% of women), breast cancer incidence was statistically significantly higher among both current smokers (8.2% of women, hazard ratio (HR)=1.24, 95% confidence interval [CI] 1.07–1.42) and former smokers (35.6% of women, HR=1.13, 95% CI 1.06–1.21). The association was strongest among women who initiated smoking prior to the onset of menarche (HR=1.61, 95% CI 1.10–2.34) or 11 or more years prior to age at first birth (HR=1.45, 95% CI 1.21–1.74), but did not vary appreciably by duration of smoking (p-value for linear trend=0.98) or cigarettes per day (p-value for linear trend=0.62) among current smokers, or by time since last smoked (p-value for linear trend=0.36) or age at cessation (p-value for linear trend=0.23) among former smokers. Adjustment for alcohol intake in the models minimally affected the association with smoking status and related characteristics. Relative risks associated with current and former smoking were elevated in stratified analyses of current drinkers (HR=1.37, 95% CI 1.16–1.61 and HR=1.21, 95% CI 1.11–1.32, respectively), and former drinkers (HR=1.37, 95% CI 0.87–2.15 and HR=1.26, 95% CI 1.00–1.60, respectively), but not never drinkers (HR=1.08, 95% CI 0.81–1.45 and HR=1.04, 95% CI 0.91–1.18, respectively). However, the interaction between smoking and alcohol consumption was not statistically significant (p-value for interaction=0.11). Never drinkers who smoked were slightly more likely to start smoking at an older age and around the time of their first birth. Conclusions : Our results are consistent with the accumulating literature that active smoking is modestly associated with risk of invasive breast cancer. This risk is confined to women who currently or formerly drink alcohol, although confidence intervals were wide for never drinkers. The elevation in risk is strongest for women who initiate smoking at a young age, particularly during the biologically susceptible windows before menarche and before first birth. If active cigarette smoking does play a causal role in breast carcinogenesis, these results suggest that cigarette smoking may have a greater role in the initiation than progression of breast carcinogenesis. Citation Information: Cancer Prev Res 2011;4(10 Suppl):B114.