|
|
| |||||||||||||||||||||||||||||||||||||||
Background The evidence of an association of lactation with a reduction in the risk of breast cancer among women has been limited and inconsistent. The effect of lactation appears to be confined to premenopausal women with a history of long lactation, but most studies of this relation have been limited in statistical power. We conducted a multicenter, population-based, case-control study with a sample large enough for us to describe more precisely the association between lactation and the risk of breast cancer.
Methods Patients less than 75 years old who had breast cancer were identified from statewide tumor registries in Wisconsin, Massachusetts, Maine, and New Hampshire. Controls were randomly selected from lists of licensed drivers if the case subjects were less than 65 years old, and from lists of Medicare beneficiaries if they were 65 through 74 years old. Information on lactation, reproductive history, and family and medical history was obtained by means of telephone interviews. After the exclusion of nulliparous women, 5878 case subjects and 8216 controls remained for analysis.
Results After adjustment for parity, age at first delivery, and other risk factors for breast cancer, lactation was associated with a slight reduction in the risk of breast cancer among premenopausal women, as compared with the risk among women who were parous but had never lactated (relative risk, 0.78; 95 percent confidence interval, 0.66 to 0.91); the relative risk of breast cancer among postmenopausal women who had lactated, as compared with those who had not, was 1.04 (95 percent confidence interval, 0.95 to 1.14). With an increasing cumulative duration of lactation, there was a decreasing risk of breast cancer among premenopausal women (P for trend <0.001) but not among postmenopausal, parous women (P for trend = 0.51). A younger age at first lactation was significantly associated with a reduction in the risk of premenopausal breast cancer (P for trend = 0.003). As compared with parous women who did not lactate, the relative risk of breast cancer among women who first lactated at less than 20 years of age and breast-fed their infants for a total of six months was 0.54 (95 percent confidence interval, 0.36 to 0.82).
Conclusions There is a reduction in the risk of breast cancer among premenopausal women who have lactated. No reduction in the risk of breast cancer occurred among postmenopausal women with a history of lactation.
Methods
Identification of Case Subjects
All female residents of Wisconsin, western Massachusetts, Maine, and New Hampshire who were given a new diagnosis of breast cancer and who were less than 75 years of age were eligible for this study. Case subjects were identified by each state's cancer registry from April 1989 through December 1991, except for New Hampshire, where subjects were enrolled beginning in January 1990. Case subjects in whom breast cancer was diagnosed more than two years before the registry report were excluded from the case group. Information was available from each state registry on cancer site, histologic features, extent of disease, demographic variables, and the name of the follow-up physician. According to a protocol approved by the institutional review board of each participating center, the physician of record for each eligible case subject was contacted by mail to obtain permission to interview the patient. Eligibility was limited to case subjects with listed telephone numbers, driver's licenses verified by self-report (if less than 65 years of age), and known dates of diagnosis. Of the 8532 eligible case subjects, physicians refused participation for 710 (8.3 percent), 463 (5.4 percent) had died, 66 (0.8 percent) could not be located, and 405 (4.7 percent) declined to participate (Table 1). Thus, data for 6888 women were available for analysis, for an overall response rate of 80.7 percent. Of these women, 98 percent had histologic confirmation of invasive breast carcinoma.
|
In each state, controls were randomly selected from the community in two ways: those under 65 years of age were selected from a list of licensed drivers, and those 65 through 74 years of age were selected from a roster of Medicare beneficiaries compiled by the Health Care Financing Administration. Computer files of potential controls were obtained annually. The randomly selected controls were intended to have an age distribution similar to that of the case subjects, but controls were oversampled among younger women in the New England states in order to increase the statistical power of the study. Controls were required to have no history of breast cancer and to have a listed telephone number. Of the 11,319 potential controls, 126 (1.1 percent) had died, 153 (1.4 percent) could not be located, and 1521 (13.4 percent) declined to participate. The overall rate of participation was 84.2 percent, and it varied somewhat among the study sites (Table 1).
Data Collection
Case subjects and controls were sent letters briefly describing the study before they were contacted by telephone. The 25-minute telephone interview elicited information on each woman's reproductive history, including, for each pregnancy, the duration of lactation, the reasons for stopping breast-feeding, and any medications used to inhibit lactation. The interview also covered the use of hormones, physical activity, alcohol use, selected dietary items, height and weight, medical history, and demographic factors. Information about the woman's personal and family history of breast cancer was obtained at the end of the interview to maintain blinding. The interviewers of 78 percent of the case subjects and 90 percent of the controls were unaware of the woman's case-control status until the end of the interview.
Statistical Analysis
Only lactation before an assigned reference date was included in this analysis. For case subjects this was the date of the diagnosis of breast cancer. For comparability, control subjects were assigned a reference date that corresponded to the average length of time from diagnosis to interview for the case group in each state (range, 8 to 21 months). Age was defined as the age at diagnosis or on the reference date. Parity was the number of full-term pregnancies (defined as pregnancies longer than six months resulting in a live birth or still-birth). The women were defined as postmenopausal if they reported a natural menopause or a bilateral oophorectomy before the diagnosis or the reference date. Women who reported hysterectomy alone were classified as postmenopausal if their age at surgery was greater than or equal to the 90th percentile for age at natural menopause in the control group (54 years for smokers and 55 for nonsmokers). The women's menopausal status was considered unknown if they had undergone hysterectomy without bilateral oophorectomy at 42 to 54 years of age for smokers or 42 to 55 years for nonsmokers. The lifetime duration of lactation was defined as the cumulative total of the periods of lactation after each live birth. The age at first lactation was the subject's age at the time of the first delivery after which breast-feeding was reported. Insufficient milk -- a cause of the termination of lactation -- was defined as an insufficient milk supply within the first three months after either of the first two pregnancies.
Odds ratios and 95 percent confidence intervals from logistic-regression models were used to evaluate relative risks. Conditional models stratified according to age (to intervals of approximately 0.10 year) and state were used to accommodate the different age distribution of the case subjects and the controls in each state19. Interactions involving the duration of lactation were evaluated with use of logarithmic transformation on a continuous scale, with the other factor also continuous (except for a personal history of benign breast disease, a family history of breast cancer, and menopausal status). Subjects with unknown values for any variables in the multivariate analyses were excluded unless they were incorporated as a separate category (as was the case for a personal history of benign breast disease, a family history of breast cancer, and age at menopause).
Results
As compared with the controls, the women with breast cancer had a younger mean age at menarche, were older at the delivery of their first child, had lower parity, were more likely to have a family history of breast cancer or a history of benign breast disease, had a higher mean body-mass index (the weight in kilograms divided by the square of the height in meters), and were older at menopause (Table 2). These characteristics were also related to lactation, and we therefore controlled for them in all analyses. One thousand ten of the case subjects (15 percent) and 1313 of the controls (14 percent) either were nulliparous and therefore had never had the opportunity to lactate or could not provide a complete reproductive history. These women were excluded from all analyses, including that shown in Table 2.
|
|
|
Since the timing of exposure appears to be important in the development of breast cancer, we evaluated age at first lactation (Table 4). Younger ages at first lactation were associated with a decreased risk of breast cancer (P for trend = 0.003). Among women who were 20 years of age or younger when they first lactated, the risk of breast cancer was 0.54 for a total of six months of lactation, as compared with the risk among women who had never lactated. We considered the possibility that a residual association with the women's age at the first delivery was confounding this relation. Even after age at the first delivery was categorized in various ways, including by single years, however, the relation remained. The association with age at first lactation was not influenced by adjustment for the duration of lactation, nor did the risk vary according to the duration of lactation; therefore, we believe these two factors have independent relations to the risk of breast cancer.
|
We considered the possibility that women who breast-fed for a short time may have been unable to breast-feed because of an insufficient milk supply. For women who reported insufficient milk within the first three months after the first or second delivery, as compared with women who lactated after the birth of the first or second child, but then stopped for reasons unrelated to the adequacy of the milk supply, the relative risk of premenopausal breast cancer was 0.99 (Table 5). Among postmenopausal women, the odds ratio was slightly, but significantly, reduced (relative risk = 0.86). The associations were not altered by adjustment for age, parity, or age at the first full-term pregnancy.
|
Discussion
Overall, in this large study, lactation was not associated with the risk of breast cancer. Among premenopausal women, however, a slight inverse association was observed, and lactation at early ages and for long periods was associated with more substantial reductions in risk. Such relations were not found among postmenopausal women.
The relation between the risk of breast cancer and the duration of lactation among premenopausal women was similar to that in other reports,1,2,3,5,7,9,10,11 though in those studies the duration of lactation necessary to achieve a significant reduction in the risk of breast cancer ranged from 4 to 12 months2 to 6 to 8 years7. Previous studies demonstrating a protective effect of lactation included relatively high proportions of premenopausal women or carefully defined groups of premenopausal women1,2,10. We believe that the failure to examine the relation of lactation and the risk of breast cancer according to menopausal status may explain the null findings of MacMahon et al.13 and other large case-control studies4,9,14,16. In our study, only 25 percent of the women were premenopausal; thus, as we observed, a relatively small effect in this smaller group of women could be obscured in an unstratified analysis.
Our findings appear to conflict with the results of two recent cohort studies12,18. Although both the cohorts in both these studies were large, the number of premenopausal women with breast cancer who had a history of long-term lactation was small; thus, the statistical power of the studies to assess the association of greatest interest was limited. In addition, the study of Kvale and Heuch12 reported results without considering a modifying effect of menopausal status, and premenopausal exposures were not updated over the 30 years of follow-up. Thus, it is likely that misclassification of lactation history occurred, particularly among the youngest members of the cohort. In the Nurses' Health Study,18 data on most types of exposure were obtained prospectively, but lactation was assessed retrospectively. Also, because of their education level, few of these women were likely to have lactated at an early age, and it is in the group of women who were young at first lactation that we observed the strongest protective effect. Finally, women's employment profoundly affects factors related to breast-feeding, including the use of supplemental bottle feeding20,21. Although the proportion of women who ever lactated in the Nurses' Health Study was similar to that in our study, the use of supplemental feeding may have been more frequent among the nurses, and the return to regular menses correspondingly more rapid. Restricting our analysis to lifetime homemakers confirmed the protective association observed in the entire group of premenopausal women.
No previous studies have examined the relation between age at first lactation and the risk of breast cancer. Our finding of an association may be consistent with the observation that an early age at the first delivery reduces the risk of breast cancer13. In our data, however, age at first lactation is a stronger determinant of the risk of premenopausal breast cancer than any other reproductive factor. Since our investigation of the relation between breast cancer and age at first lactation was stimulated in part by post hoc findings, further studies will be necessary to confirm this association.
Byers et al., who reported a twofold increase in risk among women who reported that they had had insufficient milk, suggested that the inability to lactate, rather than lactation itself, might explain the difference in risk between women who lactate and those who do not1. We found no increased risk among women with insufficient milk, however, and other studies have also not found this effect9.
The use of hormones to suppress lactation was associated with a very small increase in risk among postmenopausal women only, though with no gradient of effect. Specific hormones were not identified. The use of diethylstilbestrol during pregnancy has been associated with a modest increase in the risk of breast cancer among older women22,23,24,25.
Some limitations should be considered in interpreting our results. The response rates in this study (81 percent for case subjects, and 84 percent for controls) were high, which suggests that selection bias, if any, would be limited. We restricted this study to women who had a driver's license or were Medicare recipients and who had a listed telephone number; we do not believe the use of these criteria compromised the generalizability of our findings. Controls selected from the appropriate lists were similar in education and mean income to census estimates (Applied Population Laboratory, Department of Sociology, University of Wisconsin-Madison: unpublished data). However, nonwhites were underrepresented in both the control group and the case group, perhaps because we established eligibility through telephone directories. Finally, all information included in this analysis was based on reports by the case and control subjects themselves. For most subjects these well-defined events occurred many years before the interview, but reproductive histories tend to be reported with a high degree of accuracy26.
Lactation may reduce the risk of breast cancer simply by interrupting ovulation or by modifying pituitary and ovarian hormone secretion1,2,27,28. Direct physical changes in the breast that accompany milk production may also contribute to the protective effect2,28. In particular, in early reproductive life breast tissue may be favorably influenced by these changes29. Lactating rats and mice are relatively resistant to the effects of chemical carcinogens, as compared with nonlactating controls30,31,32. Such effects may reflect low rates of DNA synthesis during lactation33,34 or increased elimination of the carcinogen by the secretory mammae30,32. Similar mechanisms could be important in breast cancer in premenopausal women, the risk factors for which appear to be somewhat different from those for postmenopausal disease8,35. We were not able to determine why lactation is associated with the risk of breast cancer only in premenopausal women.
If lactation, a potentially modifiable factor, indeed has a protective effect against premenopausal breast cancer, an important reduction in the incidence of this disease among women who have borne children could occur. If women who do not breast-feed or who breast-feed for less than 3 months were to do so for 4 to 12 months, breast cancer among parous premenopausal women could be reduced by 11 percent, judging from current rates. If all women with children lactated for 24 months or longer, however, then the incidence might be reduced by nearly 25 percent. This reduction would be even greater among women who first lactate at an early age.
Supported by grants (R01 CA 47147 and R01 CA 47305) from the Public Health Service.
We are indebted to Drs. John Baron, Daniel H. Freeman, Henry Anderson, and Greg Bogdan; Jerri Linn Phillips and the staff of the Wisconsin Cancer Reporting System; Susan T. Gershman and the staff of the Massachusetts Tumor Registry; Melanie Lanctot and the staff of the Maine Cancer Registry; the staff of the New Hampshire Cancer Registry; Barbara Weitz, Ron Bruno, and Kim Damakosh for study management; Amy Benedict, Emogene Dodsworth, Felicia Roberts, Lisa Sieczkowski, Debra Savage, Maureen D'Alessandro, Lorraine Carey, Roxanne Haecker, Kathleen O'Brien, Megan O'Brien, Tracey Westbrook, Betty Nelson, and Jean Dodge for data collection; Pamela Marcus and Jerry Phipps for computing and statistical assistance; and Mary Pankratz for technical support.
Source Information
From the University of Wisconsin Comprehensive Cancer Center, Madison, (P.A.N., B.E.S.); the Department of Epidemiology, UCLA School of Public Health, Los Angeles (M.P.L.); the Department of Obstetrics and Gynecology, Pritzker School of Medicine, University of Chicago, Chicago (R.M.); the Departments of Community and Family Medicine and Medicine, Dartmouth Medical School, Hanover, N.H. (E.R.G.); the Massachusetts Cancer Registry, Massachusetts Department of Public Health, Boston (R.W.C.); the Division of Disease Control, Maine Bureau of Health, Augusta (K.P.B.); the Departments of Epidemiology (W.C.W., B.M.) and Nutrition (W.C.W.), Harvard School of Public Health, Boston; the Channing Laboratory, Harvard Medical School, Boston (W.C.W.); and the Department of Medicine, Brigham and Women's Hospital, Boston (W.C.W.).
Address reprint requests to Dr. Newcomb at the University of Wisconsin Comprehensive Cancer Center, 1300 University Ave., No. 4780, Madison, WI 53706.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Breast-Feeding and Breast Cancer
London S. J., Mutgi A. B., Fenster D. L., Ross R. K., Yu M. C., Newcomb P. A., Storer B. E., Longnecker M. P.
Extract |
Full Text
N Engl J Med 1994;
330:1682-1684, Jun 9, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
