|
|
Background It is uncertain whether the use of an oral contraceptive increases the risk of breast cancer later in life, when the incidence of breast cancer is increased. We conducted a population-based, case–control study to determine the risk of breast cancer among former and current users of oral contraceptives.
Methods We interviewed women who were 35 to 64 years old. A total of 4575 women with breast cancer and 4682 controls were interviewed. Conditional logistic regression was used to calculate odds ratios as estimates of the relative risk (incidence-density ratios) of breast cancer.
Results The relative risk was 1.0 (95 percent confidence interval, 0.8 to 1.3) for women who were currently using oral contraceptives and 0.9 (95 percent confidence interval, 0.8 to 1.0) for those who had previously used them. The relative risk did not increase consistently with longer periods of use or with higher doses of estrogen. The results were similar among white and black women. Use of oral contraceptives by women with a family history of breast cancer was not associated with an increased risk of breast cancer, nor was the initiation of oral-contraceptive use at a young age.
Methods
Study Design
The design of the study is described in detail elsewhere.3 Briefly, we conducted a population-based, case–control study with enrollment at centers in Atlanta, Detroit, Philadelphia, Los Angeles, and Seattle. The Centers for Disease Control and Prevention was the data-coordinating center. Protocols were approved by the institutional review boards at the participating centers. All the women in the study gave written informed consent.
Case Subjects
Women who were 35 to 64 years old, resided in the study locations, and had invasive breast cancer initially diagnosed between 1994 and 1998 were identified in Philadelphia by field-center staff and at other sites through local Surveillance, Epidemiology, and End Results Program cancer registries. Women from this population were selected with the use of selection probabilities that were specific for the study site, nominal self-reported identification as white or black, age, and month of diagnosis. Younger women and black women were oversampled to approximate a uniform distribution across age groups and groups of white and black women. Of 5982 eligible women selected, 4575 (76 percent) were interviewed.
Controls
We identified controls (women without a diagnosis of invasive or in situ breast cancer) in the same geographic locations as the case subjects, using random-digit dialing to contact residential households by telephone. Approximately 82 percent of the households called were screened successfully. Throughout the study, controls were sampled randomly from the group of eligible women identified during telephone screening at rates designed to match the frequency of interviews with controls to the frequency of interviews with case subjects within strata defined according to the study site, race, and age. Of 5956 eligible women selected as controls, 4682 (79 percent) were interviewed.
Interviews
Study participants were interviewed in person with the use of a standardized questionnaire that incorporated the reference date (for case subjects, the date of the initial, histologically confirmed diagnosis of breast cancer; for controls, the date of telephone screening). We obtained detailed information about the use of oral contraceptives and other hormones. In addition, we asked questions about each woman's reproductive history, health, and family history. Cards listing response categories, photographs of hormonal medications, and a life-events calendar4 were used to enhance recall.5 The interviews were conducted between August 1994 and December 1998.
Classification of Oral Contraceptives
Oral contraceptives were classified as combination contraceptives if they included estrogen and progestin in each cycle (i.e., a monophasic, multiphasic, or sequential formulation) and as progestin-only contraceptives if they contained only progestin throughout the cycle. Oral contraceptives in the United States contain estrogen in the form of either ethinyl estradiol or mestranol.6 Mestranol has 67 percent of the estrogenic activity of ethinyl estradiol.7 Formulations containing 50 µg or more of ethinyl estradiol or 75 µg or more of mestranol were classified as providing a high dose of estrogen; other preparations were classified as providing a low estrogen dose. Multiple progestins are used in oral contraceptives, and standard dose equivalencies are unavailable. In some of our analyses, we divided progestins into three groups on the basis of their chemical structure: estranes, gonanes, and others. Gonanes tend to have the most pronounced progestational effects.8,9
Among women who knew what kind of oral contraceptive they used, the combination type accounted for 99.5 percent of the months of oral-contraceptive use; the progestin-only type accounted for the other 0.5 percent. In the case of women who did not know what kind of contraceptive they used (accounting for 24.4 percent of total months of use), we classified the contraceptive as the combination type.
Statistical Analysis
With the study site, race, and age as conditioning variables, we used conditional logistic regression to calculate odds ratios as estimates of the relative risk of breast cancer (incidence-density ratios)10; odds ratios are reported with 95 percent confidence intervals. For ease of presentation, we discuss the results as relative risks rather than odds ratios. A P value of 0.05 or less was considered to indicate statistical significance.
In addition to study site, race, and age as conditioning variables, we included eight factors (Table 1) as an a priori set of confounders in all models; individual factors were omitted when subgrouping made adjustment inappropriate. In selected models, we assessed the following additional factors individually as potential confounders: educational level, income, extent of weekly exercise, number of breast biopsies, duration of breast-feeding, smoking status, amount of alcohol consumed, and the presence or absence of a history of tubal sterilization, mammography, major medical conditions, and contraceptive shots or implants. Specifically, we fitted a conditional logistic model containing the contraceptive-use variables and confounder set, fitted an expanded model with each additional factor, and compared the results obtained from the models. Because none of the additional factors altered the point estimates substantially, we excluded them from all models. For modeling, factors were categorized as shown in Supplementary Appendix 1 (available with the complete text of this article at http://www.nejm.org). Tests for linear trend were conducted with models containing the relevant ordinal categorical variable and an indicator variable for contraceptive use (any or none).
|
Results
Sixty-five percent of the women in our study were white, and 35 percent were black. Case subjects and controls had significantly different distributions for multiple characteristics, including the number of term pregnancies and the presence or absence of a family history of breast cancer (Table 1 and Supplementary Appendix 1).
Seventy-seven percent of case subjects and 79 percent of controls had used some type of oral contraceptive. The risk of breast cancer among women who had ever used any type of oral contraceptive, as compared with those who had never used oral contraceptives, was 0.9 (95 percent confidence interval, 0.8 to 1.0). Among women who currently or had previously used only one type, the relative risk of breast cancer was 0.9, 0.6, 0.9, and 0.9 for those who had used monophasic, multiphasic, sequential, and progestin-only formulations, respectively; the relative risks for monophasic and multiphasic formulations were significant (data not shown). Only 32 case subjects and 39 controls had ever used progestin-only formulations. The results of an analysis that excluded data from women who did not know what kind of contraceptive they used were similar to the results of the analysis in which the data were included and attributed to use of combination contraceptives (data not shown). The remainder of our analyses focused on combination preparations.
Examination of multiple aspects of oral-contraceptive use (any, current, or former use; duration of use; age at first use; interval since last use; and estrogen dose) revealed little evidence that oral contraceptives increase the risk of breast cancer (Table 2). More than 2500 women had begun using oral contraceptives before the age of 20 years; the relative risk of breast cancer among these women was similar to the relative risk among women who had begun using oral contraceptives at an older age. The results for high-estrogen-dose preparations did not differ markedly from those for low-estrogen-dose preparations.
|
|
There were no consistent differences in the risk of breast cancer according to the type of progestin (Table 4). The relative risks tended to be higher among women who were currently using oral contraceptives than among women who no longer used them and were similar among women who had formerly used contraceptives and those who had ever used them (data not shown). In analyses according to the type of progestin, the risks were similar for white and black women (Supplementary Appendix 4, at http://www.nejm.org).
|
We performed an analysis to determine whether the association between the use of oral contraceptives and the risk of breast cancer varied according to the presence or absence of a family history of breast cancer, the body-mass index, or menopausal status among women who had ever used oral contraceptives and those who were currently using them (Table 5). The results among these subgroups were generally similar to the results of the overall analysis. The results were also similar when women who had formerly used oral contraceptives were compared with those who had ever used them (data not shown).
|
In a pooled analysis of 54 studies, the relative risk of breast cancer among women who were currently using oral contraceptives, as compared with those who had never used them, was 1.24 (95 percent confidence interval, 1.15 to 1.33).1 Our study yielded a relative risk of 1.0. The pooled analysis, like our study, showed that the risk of breast cancer was not significantly related to the duration of oral-contraceptive use or to the dose of estrogen. In contrast to our study, the pooled analysis showed that the risk was slightly but significantly increased among women who had stopped taking oral contraceptives up to 10 years earlier. Nine percent of the women with breast cancer in the pooled analysis were less than 35 years old at the time of the diagnosis1; our study was restricted to women who were 35 to 64 years old. In both the pooled analysis and our study, 33 percent of women with breast cancer were at least 55 years old at the time of the diagnosis.1
A recent study12 suggested that among women who had a first-degree relative with breast cancer, the risk of breast cancer among those who had used oral contraceptives before 1976 (when preparations were likely to contain high doses of estrogen and progestin) was three times as high as the risk among women who had never used oral contraceptives. In our study, oral-contraceptive use was not associated with an increased relative risk of breast cancer among women with such a family history. In this group of women, the relative risk associated with the use of high-estrogen-dose preparations for less than 1 year, 1 to less than 5 years, 5 to less than 10 years, 10 to less than 15 years, and 15 or more years was 1.0, 1.2, 1.2, 0.5, and 0.8, respectively, and for low-estrogen-dose preparations, the relative risk was 0.9, 0.8, 0.7, 0.5, and 0.5, respectively; none of the relative risks were significantly increased.
Young women with BRCA1 or BRCA2 mutations who have used oral contraceptives may have an increased risk of breast cancer13; the same may be true for young women with a family history of breast cancer.14 In our study, the relative risk of breast cancer among women who were 35 to 44 years old, had a family history of breast cancer, and had ever used oral contraceptives was higher than that among older women with such a history, but this difference was not significant.
The incidence of breast cancer in the United States is higher among white women, but the rate of death from breast cancer is higher among black women.15,16,17 Access to care may account in part for this discrepancy.18 In addition, the biology of breast cancer in white women and black women may differ; black women may be at higher risk for breast cancer that is negative for estrogen and progesterone receptors19,20 and for more histologically aggressive disease.21 Because of these possible differences and the possibility that white women and black women use oral contraceptives differently,22 we evaluated these two groups separately. We did not find consistent evidence that the effect of oral contraceptives on the risk of breast cancer differs between white women and black women.
In 1990, the labeling of U.S. oral contraceptives was revised; specifically, the reference to an increased risk of death from cardiovascular causes among healthy women 40 years of age or older who do not smoke was deleted.23 This revision may have led to increased use of oral contraceptives among older women for contraceptive or noncontraceptive reasons. In our study, the oldest current users of combination and progestin-only oral contraceptives were 54 and 62 years old, respectively, and among women who were 45 to 64 years, the risk of breast cancer was not significantly higher among the women who were currently using oral contraceptives containing a low dose of estrogen than among those who had never used oral contraceptives. However, our finding in this older group may not be definitive, and further investigation of this question is warranted.
It has been reported that the relation between the presence or absence of a history of oral-contraceptive use and the risk of breast cancer varies according to age, with older women having a slightly lower risk.24 Similarly, we found that women 45 to 64 years old who had ever used oral contraceptives had a small but significant reduction in the relative risk of breast cancer.
We cannot explain why our results varied according to the study site. Chance, biology, or bias could account for these findings. The study site did not influence estimates of relative risk for a variety of other factors, including use of hormone-replacement therapy.
The population-based design of our study minimized the potential for a biased selection of cases and controls. We interviewed 76 percent of eligible women with cancer and 79 percent of eligible controls. Because the controls were selected by random-digit dialing and 82 percent of households were screened, the actual response rate among the controls was 65 percent. To the degree that women who participated in our study differed from those who did not, our results may be biased. We have no reason to believe, however, that the participation of case subjects and controls was influenced differently by their histories of hormone use.
We did not validate information on the use of oral contraceptives. However, we used memory aids that increase recall.5,25 Other limitations include representation of only white and black women, the absence of information on diet and environmental exposures (e.g., radiation and toxic chemicals), and small subgroups. We have no information on women under the age of 35 years. In the pooled analysis,2 the relative risk of breast cancer was highest among women under the age of 35 years who were current or recent users (with recent use defined as use within the previous 5 years) and who had started using oral contraceptives before the age of 20. When we examined the data for our youngest subgroup of women, those who were 35 to 39 years old, the relative risk of breast cancer did indeed tend to be higher than that in older subgroups. However, we found little evidence that the initiation of oral-contraceptive use at a young age was associated with a substantially increased risk of breast cancer, even among current users. In the group of current users who had started using oral contraceptives before the age of 20 years, the relative risk was 1.0, 1.0, and 1.1 among women who were 35 to 39, 40 to 44, and 45 to 64 years old, respectively.
In conclusion, current or former use of oral contraceptives among women 35 to 64 years old did not significantly increase the risk of breast cancer. Our data provide strong evidence that former oral-contraceptive use does not increase this risk later in life, when the incidence of breast cancer is higher.
Supported by the National Institute of Child Health and Human Development, with additional support from the National Cancer Institute, through contracts with Emory University (N01-HD-3-3168), the Fred Hutchinson Cancer Research Center (N01-HD-2-3166), the Karmanos Cancer Institute at Wayne State University (N01-HD-3-3174), the University of Pennsylvania (N01-HD-3-3176), and the University of Southern California (N01-HD-3-3175) and through an intraagency agreement with the Centers for Disease Control and Prevention (Y01-HD-7022). The Centers for Disease Control and Prevention contributed additional staff and computer support. The Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute provided assistance for study sites in Atlanta (N01-PC-67006), Detroit (N01-CN-65064), Los Angeles (N01-PC-67010), and Seattle (N01-CN-0532).
We are indebted to the women who participated in this study for their generosity and to all past and present members of the Women's CARE Study team for their diverse contributions.
Source Information
From the Division of Reproductive Health, Centers for Disease Control and Prevention, Atlanta (P.A.M., J.A.M., H.G.W., S.G.F., M.G.M.); the Fred Hutchinson Cancer Research Center, Seattle (J.R.D., K.E.M.); the Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles (L.B., G.U.); the Center for Clinical Epidemiology and Biostatistics and the Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia (B.L.S., S.A.N.); and the Division of Epidemiology, Karmanos Cancer Institute, Wayne State University, Detroit (L.K.W.).
Other authors were Phyllis A. Wingo, Ph.D., Department of Epidemiology and Surveillance Research, American Cancer Society, Atlanta; Ronald T. Burkman, M.D., Department of Obstetrics and Gynecology, Bay State Medical Center, Springfield, Mass.; Jesse A. Berlin, Sc.D., Center for Clinical Epidemiology and Biostatistics and Department of Biostatistics and Epidemiology, University of Pennsylvania, Philadelphia; Michael S. Simon, M.D., M.P.H., Division of Hematology and Oncology, Karmanos Cancer Institute, Wayne State University, Detroit; and Robert Spirtas, Dr.P.H., Contraception and Reproductive Health Branch, Center for Population Research, National Institute of Child Health and Human Development, Bethesda, Md.
References
In addition to the authors, the Women's CARE Study included the following investigators: J.M. Liff, D.M. Deapen, E.W. Flagg, M.F. Press, and R.J. Coates. Members of the Scientific Advisory Committee included B.S. Hulka, C. Hunter, D. Lezotte, and J. Schlesselman.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Oral Contraceptives and the Risk of Breast Cancer
Althuis M. D., Brinton L. A., Grant E. C.G., Friedenson B., Marchbanks P. A., McDonald J. A., Wilson H. G.
Extract |
Full Text |
PDF
N Engl J Med 2002;
347:1448-1449, Oct 31, 2002.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
