For some time, there has been concern regarding the effects of menopausal hormones on the risk of breast cancer. Although many investigations on this issue have been conducted, the relationship remains unresolved.1 Some studies suggest that risk may be elevated among either long-term2-10 or recent11-13 users, although contradictory results prevail. The most recent results, which derive from a collaborative re-analysis of data from 51 studies,14 indicate an effect of duration of use. These results raise concern because, in response to the recognized benefits of estrogens with respect to cardiovascular diseases and osteoporosis, many women are being placed on hormone replacement therapy (HRT) at younger ages and continuing usage for extended periods of time. Many of these women spent their reproductive years in an era when oral contraceptives (OCs) were widely available, leading to large numbers of women jointly exposed to menopausal hormones and OCs. Because breast cancer is a hormonally responsive tumor, the combined effects of two synthetic hormones become a major concern, especially given recent studies indicating that certain patterns of OC use may lead to increases in breast cancer risk.15-18
Despite this concern, few studies have attempted to evaluate breast cancer risk among women exposed to both OCs and HRT. In a large case-control study of breast cancer that focused primarily on women younger than the age of 45 years, we previously observed a 30% elevation in risk associated with use of OCs, which was higher for current users, long-term users, and women whose cancers were diagnosed prior to age 35.15 In this current investigation, we evaluated relationships of breast cancer risk to HRT among women under 55 years of age, and assessed how these relationships were modified by additional usage of OCs.
As reported previously,15 a population-based, case-control study was conducted in the two metropolitan areas of Atlanta, Georgia, and Seattle/Puget Sound, Washington, and in five counties of central New Jersey. Although the focus of this investigation was on subjects 20-44 years of age, in Atlanta the age range was extended through age 54. To maximize opportunities to evaluate exposure to menopausal hormones, only subjects from Atlanta are included in this present analysis.
Cases for study were identified through rapid ascertainment systems, with periodic checks made against existing cancer registries to determine the completeness of case identification. Included for study were all women in the Atlanta metropolitan area under the age of 55 years newly diagnosed with in situ or invasive breast cancer during the period May 1, 1990 through December 31, 1992. Hospital records of these patients were abstracted to document details on the clinical and pathologic characteristics of the breast tumors.
Controls were frequency matched by geographic area and age to the expected distribution of cases and were identified through random digit dialing.19 In Atlanta, a 91.0% response rate to the telephone screener was obtained from the 6,920 telephone numbers assessed as residential.
Personal interviews, which lasted an average of 71 minutes, collected detailed information regarding demographic factors; reproductive, breast-feeding, and menstrual histories; contraceptive behavior; use of exogenous hormones; medical and screening history; anthropometry and physical activity; adolescent diet; alcohol consumption; smoking; occupations; family history of cancer; and certain lifestyle factors and opinions about cancer causation. To aid recall of use of OCs, a month-by-month calendar was used to document all contraceptive methods used since menarche. Pregnancies and other life events were marked on the calendar to serve as a frame of reference for changes in contraceptive use over time. Information on HRT was obtained by asking subjects whether they had ever taken any type of estrogen for menopausal symptoms, irregular periods, or prevention of diseases, such as bone loss. Subjects who responded positively to this screener question were then asked more detailed questions regarding forms of estrogen use, dates of use, and whether estrogen pill use had been accompanied by use of a progestin pill. Color photographs of both OCs and HRT as marketed (i.e., by year introduced and color of pill) were shown to assist participants in identifying the specific types of preparations used during each episode of usage.
In Atlanta, completed interviews were obtained from 1,048 of the 1,171 identified eligible cases (89.5%) and 934 of the 1,185 eligible controls (78.8%). Reasons for nonresponse included refusals (3.8% physician refusal and 4.5% subject refusal in cases vs. 17.0% subject refusal in controls), death (0.4 vs. 0.2%), illness (0.8 vs. 0.2%), a move outside the study area (0.7 vs. 2.4%), and other miscellaneous reasons (0.3 vs. 1.3%). Among the controls, an overall response rate of 71.7% was achieved through multiplication of the telephone screener and interview response rates.
Because controls were identified through telephone sampling, the 17 cases without residential telephones at diagnosis were eliminated, leaving 1,031 cases available for analysis. A total of 15 controls with a history of breast cancer were eliminated, leaving 919 controls for analysis.
All risk factor information was truncated at the date of diagnosis for cases or the date of completion of the random digit dialing screener for controls (known as the reference date). The relationship of hormone use to breast cancer risk was assessed through calculation of odds ratios, as approximators of relative risks (RRs). Logistic regression analyses were used to control for effects of extraneous variables and to derive maximum likelihood estimates of relative risks and their 95% confidence intervals (CI).20 These adjustment factors included age, race, a combined variable of number of births and age at first childbirth (necessary in order to simultaneously control for both variables), and a history of ever having received a mammogram.
Relative risks associated with varying patterns of use of OCs are presented in Table 1. After adjustment for reproductive parameters and mammographic screening history, ever use of OCs was associated with a RR of 1.1 (95% CI 0.9-1.4). Risk increased with duration of use. Slightly more than 15% of subjects reported usage of OCs for 10 or more years, with such usage being associated with nearly a 30% elevation in risk. Somewhat higher risks were observed for those who had initiated use within the last 15 years and who were still using OCs within the last 5 years. The risk associated with OCs was elevated further among women whose breast cancers were diagnosed at younger ages (<35 years of age) (RR = 2.2, 95% CI 0.9-5.0), particularly if use extended for more than 10 years (RR = 3.2, 95% 0.9-10.2) (data not shown). These results for Atlanta were similar to those previously presented for all three study sites originally included in this investigation.15
A total of 248 of the cases (24.0%) and 248 of the controls (27.0%) reported ever use of HRT. Examination of predictors of hormone use among the controls showed, not surprisingly, that rates of use increased with age. After adjustment for age, hormone use was significantly more common among subjects who had frequent mammograms and somewhat more common among women who had multiple births or an early age at first birth, a history of a bilateral oophorectomy or a hysterectomy, a previous breast biopsy, smoking histories, or regular breast self-examinations. Hormone use was more common among those without as compared to with affected relatives with breast cancer, although the difference was not statistically significant. Other factors, including education, family income, body weight, and alcohol consumption, were not strongly related to the prevalence of estrogen use. Of those factors related to hormone use, only control for reproductive parameters, type of menopause, and mammographic screening history affected the breast cancer risk estimate for hormone use. As shown in Table 2, the adjusted RR associated with HRT was 0.9 (95% CI 0.7-1.2). When more detailed parameters of hormone usage were examined, there was no systematic variation in risk according to years of use. Most women (56.8%) reported usage of less than 3 years. Of those with longer durations of usage, the mean duration of use was only 7.2 years. Breast cancer risk also did not vary substantially by years since first or last use. Most women using hormones were current or recent users (within the last year) and relatively few reported usage for more than 5 years. The risk was slightly lower for women who had used estrogens alone (RR = 0.7) as compared with those who had used estrogens in combination with a progestin (RR = 1.0).
A total of 25.7% of the OC users reported usage of HRT, whereas 71.9% of the HRT users reported prior use of OCs. A total of 17.4% of the cases versus 19.4% of the controls reported having used both OCs and HRT. When risk was examined by combined parameters of use of OCs and hormone replacement (Table 3), there was some indication of a modest increase in risk among exclusive users of OCs (RR = 1.2, 95% CI 0.9-1.5), but subjects using only hormones were not at elevated risk. Women who reported using both OCs and hormone replacement did not experience any alteration in risk relative to subjects unexposed to either agent (RR = 1.0, 95% CI 0.7-1.4)
Further analyses considered effects of combined parameters of use of both types of hormones according to the presence of other breast cancer risk factors, including age, race, type of menopause, body size, and family history of breast cancer (Table 4). The relationship of combined exposure to OCs and HRT did not vary substantially by age or race. However, there was some indication of a stronger relationship among women who had undergone a natural menopause (RR = 2.0,95% CI 0.7-5.6), thin women (RR = 1.5, 95% CI 0.8-3.0), and those with a family history of breast cancer in a first degree relative (RR = 2.0, 95% 0.6-7.0), although none of the individual risk estimates was statistically significant. The risks associated with hormone usage among menopausal subgroups based on type of menopause were not affected by further adjustment for ages at menopause. Further, although not shown, we did not observe substantial variation in the effects of joint exposure to OCs and HRT by other risk factors examined, including parity status, history of benign breast disease, consumption of alcoholic beverages, or cigarette smoking.
Risks were further examined according to joint effects of years of use of OCs and HRT (Table 5). Most risks were close to the null, although subjects who reported using OCs for at least 10 years and who subsequently took HRT were at elevated risk. The risk was statistically significantly elevated among subjects who reported 10 or more years of use of OCs and 3 or more years of hormone replacement (RR = 3.2, 95% CI 1.4-7.4). We attempted to discern whether this elevated risk might derive from these women being at unusual risk for other reasons; analyses specifically focused on risks according to type of menopause, body size, and family history of breast cancer. The data for such subgroups analyses became sparse, creating interpretative difficulties. However, it did not appear as though the elevated risks associated with long-term use were due to the subjects being at increased risk for other reasons. In addition, we attempted to assess whether the elevated risk could be a consequence of exposures other than merely duration of use, including whether timing of usage might be involved. However, when we examined time intervals between discontinuation of OCs and HRT, we saw no indication that subjects who had limited intervals were at any unusual risk compared with those with more extended periods (data not shown).
We attempted to determine if hormone risks varied by stages at breast cancer diagnosis (16.7% of the tumors were diagnosed as in situ) and by the means by which the cancers had been detected (breast self-examination, accidental discovery, routine mammography, or physical examination). We could find no evidence of substantial variation in hormone effects by either of these parameters.
This study evaluated the effects on breast cancer risk among OC users who were subsequently exposed to HRT. Although this is becoming an increasingly prevalent exposure pattern, few previous epidemiological studies have evaluated the joint effects of these two exogenous hormones. Those studies that have examined combined exposure have found no evidence of any substantial interactive effects,12,21-24 although the numbers of women involved were usually small. Our study did have a sizable proportion of women exposed to both agents, but we were limited in our ability to generalize results to a wide age-range of potential users because relatively few women reported usage of HRT for extended periods of time. Nonetheless, our results provide impetus for further investigations that can assemble large numbers of women exposed to both agents, particularly when usage of both agents continues over extended periods of time.
This investigation at three study sites has previously demonstrated an elevated risk of breast cancer associated with OC usage, particularly among younger women (i.e., < 35 years of age).15 In the current investigation, which was restricted to only one of the geographic areas (Atlanta) of our previous study, risk also appeared to be somewhat elevated among users of OCs, particularly if use extended for long periods of time. In addition, the present investigation presents for the first time an evaluation of breast cancer risk related to HRT. We did not find that use of HRT (either estrogens alone or estrogens combined with a progestin) was associated with breast cancer risk, a finding that agrees with many but not all investigations.1 However, in many of the investigations that have found elevated risks, these were detected among either long-term users3,8,9,14 or older3,8,11,25-27 women, subgroups in whom we could not fully evaluate effects.
Although we were limited in our ability to evaluate effects of longer-term use of HRT and to evaluate effects in older women, a strength of our study was in having adequate power to evaluate effects of joint exposure to ever use of both OCs and HRT, because nearly 20% of the study participants reported usage of both agents. Although there was no indication that combined exposure to both OCs and hormone replacement was related to breast cancer risk, our findings were not as reassuring when we examined joint effects of varying durations of use of OCs and replacement hormones. Thus, long-term users of OCs who subsequently were placed on HRT did show some elevation in risk, with the relative risk being statistically significantly elevated in subjects exposed for 10 or more years to OCs and 3 or more years to replacement hormones (RR = 3.2, 95% CI 1.4-7.4). It is possible that this result merely represented a chance finding, because only 25 women had this joint exposure. However, given observations from previous studies that long-term exposure to either OCs or menopausal hormones can lead to increases in risk, the combined exposure to long-term effects of both agents would be the exposure hypothesized a priori to be of most concern. Thus, although our isolated finding must be cautiously interpreted, it is definitely an exposure pattern that bears future assessment in other investigations that can accrue large numbers of older women exposed for long periods of time to both agents.
In future investigations, it will also be important to assess risks associated with joint effects of hormones according to other breast cancer risk factors, especially because there is some indication that effects of individual agents may be influenced by host factors. Findings on this issue have been quite conflicting, probably because of the statistical power limitations involved in examining subgroup associations. However, there is evidence from some investigations that subjects with a family history of breast cancer may be most susceptible to the effects of either OCs17,18,22,28,29 or menopausal hormones.5,7,10,25,27,30,31 Thus, our finding that family history positive subjects who reported usage of both agents were at a twofold risk relative to family history positive subjects without a history of hormone usage may be noteworthy. Although the risk was not statistically significant, it would appear to be a subgroup on which future investigations should focus. Future investigations should also probably evaluate interactions of hormone use by body weight, based both on results from this study and a recent publication showing an enhanced breast cancer risk associated with use of menopausal hormones among women who had not gained weight during adulthood.32
In summary, this investigation provides some clues for future research directions regarding the joint effects of OCs and HRT on breast cancer risk. Although our results must be cautiously interpreted due to the focus on younger women and the small percentage who reported extended usage of HRT, it is of concern that women who did report being exposed to both agents for longer periods of time were at a significantly elevated risk of breast cancer. Future studies should specifically examine breast cancer risk in this subgroup of women, as well as in other groups in which effects of hormones might be enhanced, including those with a family history of the disease.
1. Brinton LA. Hormone replacement therapy
for breast cancer
. Endocrinol Metab Clin North Am
2. Bergkvist L, Adami H-O, Persson I, Hoover R, Schairer C. The risk
of breast cancer
after estrogen and estrogen-progestin replacement. N Engl J Med
3. Brinton LA, Hoover R, Fraumeni JF Jr. Menopausal oestrogens and breast cancer risk
: an expanded case-control study. Br J Cancer
4. Ewertz M. Influence of non-contraceptive exogenous and endogenous sex hormones on breast cancer risk
in Denmark. lnt J Cancer
5. Hoover R, Glass A, Finkle WD, Azevedo D, Milne K. Conjugated estrogens and breast cancer risk
in women. J Natl Cancer Inst
6. Hoover R, Gray LA, Cole P, MacMahon B. Menopausal estrogens and breast cancer
. N Engl J Med
7. Hulka BS, Chambless LE, Deubner DC, Wilkinson WE. Breast cancer
and estrogen replacement therapy. Am J Obstet Gynecol
8. La Vecchia C, Negri E, Franceschi S, et al. Hormone replacement treatment and breast cancer risk
: a cooperative Italian study. Br J Cancer
9. Risch HA, Howe GR. Menopausal hormone usage and breast cancer
in Saskatchewan: a record-linkage cohort study. Am J Epidemiol
10. Rohan TE, McMichael AJ. Non-contraceptive exogenous oestrogen therapy and breast cancer
. Med J Austral
11. Colditz GA, Hankinson SE, Hunter DJ, et al. The use of estrogens and progestins and the risk
of breast cancer
in postmenopausal women. N Engl J Med
12. Mills PK, Beeson L, Phillips RL, Fraser GE. Prospective study of exogenous hormone use and breast cancer
in Seventh-Day Adventists. Cancer
13. Yang CP, Daling JR, Band PR, Gallagher RP, White E, Weiss NS. Noncontraceptive hormone use and risk
of breast cancer
. Cancer Causes Control
14. Collaborative Group on Hormonal Factors in Breast Cancer
. Breast cancer
and hormone replacement therapy
: collaborative reanalysis of data from 51 epidemiological studies of 52,705 women with breast cancer
and 108,411 women without breast cancer
15. Brinton LA, Daling JR, Liff JM, et al. Oral contraceptives
and breast cancer risk
among younger women. J Natl Cancer Inst
16. Collaborative Group on Hormonal Factors in Breast Cancer
. Breast cancer
and hormonal contraceptives: collaborative reanalysis of individual data on 53,297 women with breast cancer
and 100,239 women without breast cancer
from 54 epidemiological studies. Lancet
17. UK National Case-Control Study. Oral contraceptive use and breast cancer risk
in young women. Lancet
18. White E, Malone KE, Weiss NS, et al. Breast cancer
among young US women in relation to oral contraceptive use. J Natl Cancer Inst
19. Waksberg J. Sampling methods for random digit dialing. J Am Stat Assoc
20. Breslow NE, Day NE. Statistical methods in cancer research. Volume 1—the analysis of case-control studies. IARC Sci Publ
21. Colditz GA, Stampfer MJ, Willett WC, et al. Type of postmenopausal hormone use and risk
of breast cancer
: 12-year follow-up from the Nurses’ Health Study. Cancer Causes Control
22. Schuurman AG, van den Brandt PA, Goldbohm RA. Exogenous hormone use and the risk
of postmenopausal breast cancer
: Results from the Netherlands cohort study. Cancer Causes Control
23. Stanford JL, Brinton LA, Hoover RN. Oral contraceptives
and breast cancer
: Results from an expanded case-control study. Br J Cancer
24. Stanford JL, Weiss NS, Voigt LF, Daling JR, Habel LA, Rossing MA. Combined estrogen and progestin hormone replacement therapy
in relation to risk
of breast cancer
in middle aged women. JAMA
25. Kaufman DW, Palmer JR, de Mouzon J, et al. Estrogen replacement therapy and the risk
of breast cancer
: results from the case-control surveillance study. Am J Epidemiol
26. Palmer JR, Rosenberg L, Clarke EA, Miller DR, Shapiro S. Breast cancer risk
after estrogen-replacement therapy: results from the Toronto breast cancer
study. Am J Epidemiol
27. Wingo PA, Layde PM, Lee NC, Rubin G, Ory HW. The risk
of breast cancer
in postmenopausal women who have used estrogen replacement therapy. JAMA
28. Miller DR, Rosenberg L, Kaufman DW, et al. Breast cancer
before age 45 and oral contraceptive use: new findings. Am J Epidemiol
29. Rosenberg L, Palmer JR, Rao RS, et al. Case-control study of oral contraceptive use and risk
of breast cancer
. Am J Epidemiol
30. Newcomb PA, Longnecker MP, Storer BE, et al. Long-term hormone replacement therapy
of breast cancer
in postmenopausal women. Am J Epidemiol
31. Nomura AMY, Kolonel LN, Hirohata T, Lee J. The association of replacement estrogens with breast cancer
. IntJ Cancer
32. Huang Z, Hankinson SE, Colditz GA, et al. Dual effects of weight and weight gain on breast cancer risk