Cigarette Smoking Is Not Associated with Breast Cancer Risk in Young Women

Introduction

Cigarette smoking has been hypothesized to have a carcinogenic effect on breast tissue, particularly when exposure occurs before sexual maturity and breast tissue differentiation (1). Epidemiologic studies of cigarette smoking on breast cancer risk during this developmental period have produced inconsistent results (2-14). However, recent cohort studies (7-10), including two with mainly young women (7, 10), suggest an effect of smoking at an early age. In this study, we examine the effect of smoking on breast cancer risk in women <50 years of age.

Materials and Methods

Details of this population-based case-control study have been described elsewhere (15). Briefly, 1,794 U.S.-born, English-speaking, non-Hispanic white, Hispanic white, and African-American Los Angeles female residents with incident invasive breast cancer, ages 20-49 years, were recruited through the Los Angeles County Cancer Surveillance Program between February 1998 and May 2003. Four hundred forty-four U.S.-born, English-speaking neighborhood controls matched by age (±5 years) and race were recruited between July 2000 and March 2003. After excluding subjects due to missing information, 1,728 cases and 441 controls remained in our analysis.

In-person interviews were conducted using a structured questionnaire to gather information on reproductive history, smoking and alcohol consumption, family history of breast and ovarian cancer, demographics, and other factors. Information was recorded up to the reference date for each subject. The reference date was the date of diagnosis for cases or the date of initial household contact for controls. All study participants signed an informed consent before interview. The study protocol was approved by the University of Southern California's Institutional Review Board.

Data Analyses

We used unconditional multivariable logistic regression to estimate the odds ratios and 95% confidence intervals as a measure of breast cancer risk associated with cigarette use. We adjusted for age, race, education, age at menarche, number of full-term pregnancies, age at first full-term pregnancy, average amount of alcohol within the recent 5-year period, and family history of breast or ovarian cancer. Additional analyses restricted to cases diagnosed after June 2000 yielded similar results (data not shown). We also conducted a case-case analysis to determine whether family history modifies the association between smoking and breast cancer risk. Statistical analyses were done using the SAS version 9 statistical package (SAS Institute, Cary, NC).

Results

Compared with women who never smoked, women who reported having ever smoked were not at increased risk for breast cancer (odds ratio, 0.99; 95% confidence interval, 0.78-1.25; Table 1 ). Likewise, no statistically significant association or trend with breast cancer risk was found for total years of smoking; age commenced smoking; lifetime average cigarettes smoked; average number of cigarettes smoked in recent 5 years or in the 5-year period after menarche, before 18 years of age, or before age at first birth; or total years of smoking before first full-term pregnancy (all trend P values were >0.25). Results were similar when we restricted the analysis to long-term smokers (20+ years) and never smokers (data not shown). We observed no significant trends between smoking and risk of breast cancer when we restricted analyses to stage I, II, or III/IV cancer cases (data not shown).

First-degree family history of breast or ovarian cancer did not modify the association between any of the smoking variables and breast cancer risk (P > 0.15).

Discussion

In this study of young women (<50 years), we did not find significant associations of breast cancer risk with any of the smoking characteristics investigated. Our results are consistent with the bulk of published literature on this topic, which do not find significant relationships between smoking and overall incident invasive breast cancer risk among young or premenopausal women (3-5, 16-21). A review of the literature by IARC (22) and a pooled analysis on breast cancer and smoking by the Collaborative Group on Hormonal Factors in Breast Cancer (23) both conclude that there is no overall association between smoking and breast cancer risk. On the other hand, four recent studies have found an effect of smoking at an early age and breast cancer risk (7-10).

In the California Teachers cohort study, there was a statistically significant increase in risk of breast cancer associated with early initiation of smoking among postmenopausal women, but this was not quite statistically significant in premenopausal/perimenopausal women (8). The Nurses' Health Studies I (9) and II (7) found statistically significant associations and/or trends with breast cancer risk among women who initiated smoking at an early age or smoked before their first pregnancy. A Norwegian-Swedish cohort study also found an increase associated with early initiation (10).

Our study strengths include the large sample size of young incident breast cancer cases and the in-person interviews conducted with a structured questionnaire to collect detailed information on smoking and potential confounders. Our study had sufficient power to detect moderate positive associations if they existed, thus our power was similar to previously published studies (4, 12, 13).

A weakness in our study is the potential for survival bias. However, there was no indication that associations differed by stage, suggesting that survival bias is unlikely to have caused these findings.

Another disadvantage of this study is the potential for recall bias. We asked women to recall smoking patterns over decades of use. Although a clear association between smoking and breast cancer risk has not been established, women in the United States are inclined to believe that smoking is an important risk factor for breast cancer (24, 25). Women diagnosed with breast cancer would be expected to overestimate their past cigarette use compared with the control subjects. This would probably have resulted in a bias away from the null, generating an inflated risk, and is therefore unlikely to have been a problem in our case-control comparison.

In conclusion, our study does not support an effect of smoking at a young age on breast cancer risk among women under 50 years.