These three studies that reported on the risk of breast cancer in 3 highly respected medical journals recently received wide media coverage.(1)(2)(3) One study showed there was no significant risk of breast cancer among users of oral contraceptives.1 another study was halted because of the increased risk of using a combined form of estrogen plus progestin for hormonal replacement.2 The third study showed that both breast feeding and childbearing significantly lowered the risk of breast cancer.3 These three studies are briefly reviewed below.
A pooled analysis of 54 studies on the relative risk of breast cancer among women who were currently using oral contraceptives was reported in a 1996 article in the medical journal Lancet.4 The study showed a 24% increased risk of breast cancer among users and former users of oral contraceptives as compared to women who never used oral contraceptives. However, many of the studies in this article are old and the data was pooled -- i.e., came from many studies. New evidence was needed to validate this risk and to determine if it existed among the large numbers of women who took oral contraceptives at an early age and are now reaching the age in which breast cancer is the highest. Therefore, a research group that was coordinated through the Centers for Disease Control and Prevention conducted what is called the National Institute of Child Health and Human Development Women's Contraceptive and Reproductive Experiences (Women's CARE) Study to determine if the use of oral contraceptives was a risk factor for breast cancer in women who were 35 to 64 years old.
The study was a large population-based case control investigation that enrolled eligible women participants from centers in Atlanta, Detroit, Philadelphia, Los Angeles, and Seattle. Participants were 35 to 64 years old and had an initial diagnosis of invasive breast cancer between 1994 and 1998. A total of 5,982 eligible women were selected and 76% (4575) were interviewed. The control group consisted of women who did not have a diagnosis of invasive breast cancer. They were selected from the same geographic location as the breast cancer group by use of a random digit dialing process and were matched (on age and race) with the breast cancer group. Of the 5,956 eligible controls selected, 79% (4682) were included in the study. All of the participants were interviewed to determine if they used or had used oral contraceptives in the past, the type of oral contraceptive used, length of use, and multiple demographics.
Seventy-seven percent of the breast cancer group and 79% of the control group participants had used some type of oral contraceptive. The relative risk of breast cancer among women who ever had used oral contraceptives as compared to those women who had never used oral contraceptives was 0.9 (95% CI, 0.8 to 1.0) -- i.e., there essentially was no demonstrated risk. The relative risk of breast cancer was 1.0 (95% CI; 0.8 to 1.3) for women who were currently using oral contraceptives and 0.9 (95% CI; 0.8 to 1.0) for those who previously had used oral contraceptives, i.e., again no risk. The risk of breast cancer did not increase with longer periods of use or with higher dose formulation of estrogen. The researchers also did not find an increased risk with initiation of oral contraceptives at a young age, duration of use before the first term pregnancy, or among those who had a family history of breast cancer.
The authors concluded that for women between the ages of 35 to 64 years, current or former use of oral contraceptives was not associated with an increased risk of invasive breast cancer. They also emphasized that their study provided strong evidence that former oral contraceptive use does not increase the risk of breast cancer later in life when the incidence of breast cancer is the highest.
I first heard about this study listening to the evening news. The announcer excitingly stated that this was "the definitive" study on the risk of breast cancer and the use of oral contraceptives. After reading the study, I was not at all convinced that this was the definitive study, or that there was little or no risk of breast cancer among current and ever users of oral contraceptives. The following are my reasons for not being convinced that the authors produced the definitive results.
One of the most obvious flaws in the study is that the comparison group of women without the diagnosis of breast cancer is not from the same pool of subjects that the breast cancer group came from. Although the researchers tried to match the participants (on race and age) through a randomized case-control method, in my opinion they did not reach that goal. The women participants in the invasive cancer group compared to the case control group women differed significantly on 7 different variables, including age at menopause, age at first term pregnancy, number of term pregnancies, family history of breast cancer, and current or past use of hormone replacement therapy. All of these variables could influence the rate of invasive breast cancer. Furthermore, the researchers did not measure use or duration of breastfeeding among the participants. As will be seen in the third study reviewed below, breastfeeding can significantly reduce relative risk for breast cancer.
The researchers did not use a strong design to detect the comparable rates of breast cancer. Instead of investigating the differences in the rate of invasive breast cancer between groups of women who never used oral contraceptives versus women who did use oral contraceptives prospectively, they looked at the rates between those women who had invasive breast cancer and those who did not retrospectively (i.e., from past history). Both groups had a similar rate of use or non-use of oral contraceptives. Previous studies that detected an increased rate of breast cancer among oral contraceptive users did so by comparing women who never used oral contraceptive with those that had a history of use.(4) The researchers of this study admit that while they had a 65% response rate with the control group of non-invasive cancer participants, they do not know how these women differed from those who responded and thus could be a potential source of bias. Finally, much of the data collected was from recall. The researchers did not validate the use or non-use of contraception nor the type or amounts or lengths of use. Recall data could be and often is a source of error.
A study that received even more press than the oral contraceptive study was the Women's Health Initiative's (WHI) randomized control trial of estrogen plus progestin in healthy postmenopausal women. This study received cover story status in both the July 14-20th, 2002 Newsweek and Time magazines. The study deserved cover story status because of the magnitude of the findings and because the prospective randomized control group design is one of the strongest research designs that can be used to detect differences and cause and effect.
This study was sponsored by the National Heart, Lung, and Blood Institute of the National Institutes of Health and involved the recruitment of 373,002 women with intact uteruses between the ages of 50 and 79 from 40 participating clinical centers throughout the United States. Of these women, 18,845 provided consent and reported no hysterectomy, but 2,237 women were excluded due to compliance risks (e.g., alcoholism, dementia) and serious health risks. The remaining 16,608 women participants were randomized into a placebo group (N= 8,102) or a treatment group that received a daily dose of 0.625 of conjugated equine estrogen plus 2.5 mg of medroxyprogesterone acetate. The participants were enrolled in the study from 1993 through 1998. The objective of the study was to examine the effect of estrogen plus progestin on the prevention of heart disease and hip fractures, and any associated change in risk for breast and colon cancer.
The initial intent of the research group was to continue the study until 2005, but the study was halted on May 31,2002 due to the increased risk of breast cancer. The specific findings for the estrogen plus progestin group compared to the placebo group include:
According to Jacques Rossouw, MD (the acting director of the WHI), these results translate into a relatively small risk for the individual woman (http://www.nlm.nih.gov/databases/alerts/estrogen_progestin.html). For example, during a 1 year period, among 10,000 postmenopausal women with an intact uterus who are taking daily estrogen plus progestin, 8 more will have invasive breast cancer, 7 more will have a heart attack, 8 more will have a stroke, and 18 more will have blood clots than a similar group of women who do not take hormone replacements. However, he also indicated that with approximately 6 million women taking estrogen plus progestin replacements across the United States, this could add up to tens of thousands of serious health cases in a one year period.
The authors of this study concluded that the risks of taking estrogen plus progestins as replacement therapy among healthy women with uteruses for an average of 5.2 years exceeded the benefits. Furthermore, the risks of taking these hormonal replacements is not consistent with the requirements for a viable intervention for main prevention of primary disease. The results indicate that this therapy should not be initiated or continued for primary prevention of coronary heart disease.
The authors reported that the results of this study do not necessarily apply to lower dosages of these drugs, or to other formulations of estrogen and progestin, nor to other routes of administration of these hormones, e.g., through a transdermal patch. They also mention that this study could not distinguish the effects of estrogen from progestin. The results only apply to the combination of the two. There is a concomitant study that the WHI group is conducting with 10,739 women who had a hysterectomy to determine if oral estrogen (alone) will prevent coronary heart disease. The authors did mention that the relatively high rate of discontinuation in the treatment group (42%) and a cross over to active treatment in the placebo group (10.7%) are a limitation to the study. They mentioned that the result of this limitation may underestimate the magnitude of both the adverse effects on coronary heart disease and invasive breast cancer.
The Collaborative Group on Hormonal Factors in Breast Cancer (from the United Kingdom) recently reported a study that can shed further light on the development of breast cancer and the increasing rates of this cancer in developing countries. The United Kingdom Cancer Research group wanted to determine the contribution thatbreastfeeding had on the protective effect against breast cancer. They also intended to differentiate this effect (if any) from the known protective effect that childbearing has on breast cancer. The UK research group collected individual data from 47 epidemiological studies on breast cancer and breastfeeding from 30 different countries. These studies resulted in a data set of 50,302 women with breast cancer and a control group of 96,973 women without the disease. Among the women with breast cancer the average age at diagnosis was 50.1 years. The breast cancer group on average had fewer births than the control (2.2 vs 2.6), a greater portion never had children (16% vs 14%), and a lesser portion of the breast cancer group had never breastfed (71% vs 79%). One of the objectives of the study was to differentiate the relative protective effect of breastfeeding as opposed to childbearing.
There were 12,214 (29%) women in the breast cancer group and 16,900 (21%) control cases that had never breastfed. The results showed that the younger a woman began childbearing the lower was the relative risk for breast cancer. The relative risk declined by about 3% for every year younger a woman was when her first child was born. The relative risk also decreased with the number of births that a woman had at a rate of 7% per birth.
The relative risk of breast cancer declined with increasing parity in women who had ever breastfed. The relative risk of breast cancer declined by 3.4% for each child that was breastfed. Furthermore, the relative risk of breast cancer declined with the increasing duration of breastfeeding. The relative risk declined by 4.5% with 12 months of breastfeeding.
The authors calculated that the incidence of breast cancer in developing countries could be reduced by 42% solely by breastfeeding and increasing the duration of breastfeeding. In the United States (which has one of the highest rates of breast cancer) the portion of childbearing women who had ever breastfed is only around 50%. Whereas in Japan and in developing countries more than 90% of childbearing women breastfeed their infants. The authors also estimated that the incidence of breast cancer could be reduced by half if women in developed countries had the average number of children and duration of breastfeeding that was prevalent in developing countries until recently.
The authors admitted that the reporting of the duration of breastfeeding is not precise with women having a tendency to round up to the nearest 6 months. They felt, however, that this would lead to an underestimation of the true protective effect of breastfeeding. The authors also said out that there was no differentiation between exclusive breastfeeding and breastfeeding with supplements. They recommended that future studies measure the length of breastfeeding with more precision and differentiate between total breastfeeding and breastfeeding with supplements. The authors are aware that it might be unrealistic for women in today's modern society to return to childbearing and breastfeeding patterns that were characteristic a century ago.
Of the three studies reviewed, by far the best is the hormone replacement study that utilized a randomized treatment versus control group (placebo) design. That study had a prospective design, a large group of subjects and was able to detect significant differences and relative risk. The breastfeeding study, by pooling indi vidual data from many studies was able to have a very large data set that allowed the researchers to determine relative risk and show that there is a good likelihood that the increased incidence of breast cancer in modern developed countries is likely a result of decreasing parity, delayed child birth, and lack of breastfeeding. I believe that the jury is still out on the relative risk of oral contraceptive use and breast cancer. I found the weakest of the three studies to be the NEJM study on the risk of breast cancer and oral contraception. I do not see how the authors can claim that they had matched the two groups of participants when they differed significantly on 7 different variables. The authors also failed to determine breastfeeding patterns and the history of abortion. One has to wonder why the WHI researchers were able to find an increased risk of breast cancer among women who use a combination of estrogen and progestins for hormonal replacement, but the Women's CARE study researchers were not able to find an increased risk when higher levels of estrogen and progestins are used for contraception. Nobody seems to provide an explanation for this contradiction or to even mention that there is a contradiction.
The Kaiser Daily Reproductive Report (3/26/2002) recently reported a study in which 103,027 women between the ages of 30 and 49 were followed by researchers for the development of breast cancer from 1991-1992 to December, 1999. They found that women who reported using oral contraceptives at some time during that period of time had a 26% greater risk of developing breast cancer than women who never used birth control pills. The risk increased to 58% for women who used oral contraceptives for the duration of the study. This study, unlike the Women's CARE study was prospective (over time) and compared users of oral contraception with non-users. This was a much stronger designed study (than the NEJM oral contraceptive use and risk for breast cancer study) and as a result yielded increased risk with use of oral contraceptives. I am looking forward to seeing the study reported in a peer reviewed journal so that it can receive further scrutiny and coverage.
1 Marchbanks, P.A., McDonald, J.A., & Wilson, H.G. et al. Oral contraceptives and the risk of breast cancer. The New England Journal of Medicine. 346 (2002): 2025-2032. [Back]
2 Writing Group for the Women's Health Initiative. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. JAMA. 288 (2002): 321-333. [Back]
3 Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and breastfeeding: collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50,302 women with breast cancer and 96,973 women without the disease. Lancet. 360 (2002): 187-195. [Back]
4 Collaborative Group on Hormonal Factors in Breast Center. 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. 347 (1996): 1713-27. [Back]