The committee was asked to consider the potential for evidence-based actions to reduce the risk of breast cancer. Individual women, health care providers, advocacy organizations, and many other stakeholders are all eager to know what concrete steps can be taken to reduce the risk of breast cancer for an individual or the population, and when during the life course those actions might be most effective. This chapter outlines several evidence-based actions that women can take. However, the scientific community still has only limited understanding of which exposures might best be avoided and when, and which actions might have a long-term positive benefit in reducing risk for breast cancer.
Even when research strongly supports classifying an exposure as a risk factor for breast cancer, that research does not necessarily provide the information needed to determine the appropriate response to reduce risk. Should exposure be avoided completely? Will reducing or eliminating exposure in adulthood reduce a risk that has accrued from exposure at younger ages? Will the presence or absence of other risk factors for breast cancer influence the likely benefit or harm from a change in exposure to a given risk factor? Will changing one type of exposure lead to another that carries new and possibly as yet unrecognized risks for breast cancer, other diseases, or perhaps some other adverse economic or environmental outcome?
Finding ways to reduce risk and avert cases of breast cancer is a high priority for everyone concerned about this disease. Although some definite actions can be taken to reduce risk, the committee found overall that evidence-based options are limited because few studies have been done to test the effectiveness of actions that may be hypothesized to reduce risk. In this
chapter, the committee discusses some specific areas where action appears warranted, but it first summarizes the significance of the uncertainty around preventive action.
Potential for Introducing New Hazards or Risks
A key concept to remember when evaluating a particular risk associated with a particular factor is that an action that is aimed at eliminating the specific risk of concern may result in a substitution of one risk for another, or perhaps shifting risk from one group to another. Any risk of the alternative action thus needs to be considered and weighed against the risk that the change is intended to reduce or eliminate.
The complexity of trade-offs from substitutions can be illustrated with the case of contamination of potable ground water sources with pesticides or industrial chemicals shown to be carcinogenic in experimental animals or humans. Reducing exposures to potentially carcinogenic substances in drinking water from groundwater sources seems to be a logical, health-protective action, even if the actual or perceived risk from the contaminants is small. A typical action to reduce the potential cancer risk from using the contaminated ground water is to switch the consumer to an alternative source of potable water, such as a public water supply system. However, such systems require disinfection, usually by chlorination, and chlorination of surface water introduces trace levels of disinfection by-products (DBPs). Several DBPs have been found to be carcinogenic in animal bioassays (e.g., NTP, 2007a,b), and some epidemiologic studies have suggested that long-term exposure to DBPs is associated with an increase in bladder cancer (reviewed in Richardson et al., 2007), especially in a subset of the population with specific genetic polymorphisms (Cantor et al., 2010).
In this scenario, one would have to consider many factors, including (1) the relative carcinogenic potency of the groundwater contaminant(s) versus that of the DBPs, (2) the concentrations of the groundwater contaminants or the DBPs in the drinking water and indoor air following use, and (3) the duration and frequency of likely exposure to a given drinking water source over a lifetime. Depending on these values, it is possible that a comparative risk assessment would show that switching from the contaminated groundwater supply to the uncontaminated but disinfected surface water supply actually increased, rather than decreased, potential cancer risks to the exposed population.
This example, however, also illustrates the challenges in assessing trade-offs in population- and individual-level risks and benefits. There are certainly no clear benefits, at least to the individual consumer, of drinking
groundwater contaminated with low levels of pesticides. But the potential cancer risk associated with the presence of DBPs in a public water system must be assessed against the very real risks of widespread acute illness from microbial contamination that could result in the absence of disinfection (Gibbons and Laha, 1999; Schoeny, 2010).
This pattern of trading one hazardous substance for another is not uncommon. Although federal agencies evaluate the toxicity and carcinogenicity of new pesticides and prescription drugs before they are approved for sale, the United States does not have a comprehensive program to evaluate the safety of chemicals before their widespread use in consumer products. In the face of consumer concern about bisphenol A (BPA), for example, BPA-free plastics are now available, but new research appears to show that BPA-free plastics may leach other chemicals with estrogenic activity comparable to that of BPA (Yang et al., 2011). As noted in Chapter 2, the European Union has adopted a program (Registration, Evaluation, Authorisation and Restriction of Chemical Substances, or REACH) for broader safety testing by manufacturers of their products before they are approved for use. In the United States, the Government Accountability Office (GAO, 2009a,b) has recommended changes to improve the effectiveness of federal regulation of chemicals.
Risk trade-offs may also be hard to judge because a given factor can have both positive and negative health effects. For example, there is fairly compelling evidence that moderate alcohol consumption is associated with a small but consistently observed increase in the risk of breast cancer.1 However, there is also compelling evidence that consumption of the same moderate amounts of alcohol is associated with a reduction in mortality from cardiovascular disease (Maskarinec et al., 1998; Gunzerath et al., 2004; Klatsky, 2009; Ronksley et al., 2011). The risks associated with any specific environmental exposure occur against the background of a woman’s genetic susceptibility, reproductive history, and lifestyle.
Challenges in Public Health Policy Aimed at Risk Reduction
A significant challenge—relevant to the discussion of environmental risk factors for breast cancer and frequently faced by regulators of environmental pollutants and public health officials—is a lack of information about the nature of the effects of many exposures on risks for breast cancer. Chapter 4 reviewed the diverse challenges in trying to generate and interpret relevant information. Noted here are a few specific areas where substantial uncertainty faces policy makers.
Assessing the net effect of environmental exposures is one challenge.
1The trade-offs associated with alcohol use are discussed further later in this chapter.
Individuals and populations are never exposed to only one risk or protective factor at a time, but complex combinations of exposures are rarely the subject of laboratory or epidemiologic studies. The U.S. Environmental Protection Agency (EPA) and other environmental regulatory agency scientists often assume that cancer risks in a population are simply the sum of risks estimated for each individual chemical in the absence of data on risks of co-exposures. In practice, this means that one might be confident that a lifetime of exposure to a single chemical that causes a theoretical increase in cancer risk of 1 additional case per 1 million exposed people is essentially lost in the background and that the risk from that exposure may be considered minimal. But what if a population is exposed to 1,000 of these types of “1 in 1 million” lifetime risks? Are the risks simply additive (e.g., the increase in lifetime risk becomes 1,000 per 1 million, or 1 in 1,000), or is it possible that the chemicals can interact to alter risk in some nonadditive manner, either by reducing each other’s effects (e.g., competition for receptor binding) or by mutually enhancing each other’s effects? For example, smoking and asbestos exposure are each well-recognized risk factors for lung cancer, but exposure to both multiplies the risk of lung cancer, making the risk far greater than the addition of the individual effects of these two exposures. A study of asbestos workers found that the lung cancer mortality rate was 122.6 per 100,000 among men with a history of smoking and no asbestos exposure; 58.4 per 100,000 among those with asbestos exposure but no history of smoking; and 601.6 per 100,000 with exposure to both smoking and asbestos (Hammond et al., 1979). As a qualitative example, cigarette smoke is a mixture of relatively low levels of numerous carcinogens, and both direct and passive exposure to cigarette smoke are associated with a variety of cancers, including breast cancer. On the other hand, some exposures may increase risks for breast cancer, but reduce them for other cancers. In the vast majority of instances, the scientific information is typically not sufficiently developed to calculate the joint effects of multiple exposures with confidence.
Another area of uncertainty is whether risks occur at very low levels of exposures and whether those risks can be estimated from information on hazards and risks that are determined for high-dose exposures. The EPA and other regulatory agencies have made specific assumptions about the shape of the dose–response curve in relation to certain mechanisms of action. For example, a linear extrapolation of cancer risk from high doses to low doses is used for carcinogenic agents determined to be mutagenic. Although linear extrapolation may not apply in some circumstances, it is considered a protective approach in the absence of evidence to guide the selection of an alternative model (EPA, 2005). Directly detecting small differences in risk in human or animal studies may be difficult, if not impossible, because of challenges that include the need for very large study
populations, the potential for errors in measuring exposure, and the possibility of unrecognized confounding.
Finally, as highlighted in earlier chapters, the risk from a given exposure may depend on the age at which it occurs, although current knowledge about susceptible windows pertains to few exposures. Perhaps most salient is the difficulty in assessing whether reduction or elimination of an exposure will alter long-term risk of breast cancer, and if so, by how much.
Identifying evidence-based opportunities for action to reduce risk of breast cancer depends, ideally, on a convergence of several elements, including
• sufficient evidence to demonstrate that a specific factor is associated with increasing or decreasing breast cancer risk;
• a means by which to modify exposure to the risk factor;
• an understanding of whether effective changes can be made by an individual or would require instead, or in addition, changes at governmental, social, or cultural levels;
• evidence that a specific action to modify exposure will result in the desired impact on breast cancer risk, the characteristics of women who could be expected to benefit, and when the intervention needs to occur; and
• awareness of the trade-offs (potentially as yet unrecognized) that may occur in terms of other health outcomes, personal preferences, or economic consequences.
As illustrated in the reviews in Chapter 3, the evidence on many of the environmental factors that have been investigated as potential risk factors for breast cancer remains inconclusive. But for a modest set, the evidence is relatively strong and points to likely opportunities for prevention when these factors are modifiable. What, then, is the distinction between “modifiable” and “nonmodifiable” risk factors? For example, the age at which women have a first full-term pregnancy is known to influence the risk of breast cancer, with later age at first birth generally associated with higher risk. At the individual level, women can make decisions as to when they will have their first pregnancy, but changes at the population level are influenced by a range of social, economic, educational, cultural, and personal forces, and any effort to influence personal choices could have unexpected consequences. Furthermore, opportunities for modification of some factors may
be limited to certain ages—a woman’s age at a first full-term pregnancy may be modifiable before menopause, but obviously it is not after menopause.
Another question is whether particular actions to change a modifiable risk factor will actually translate into lowered risk. For example, although the evidence is relatively strong that greater body fatness is associated with increased risk of breast cancer for postmenopausal women (WCRF/AICR, 2007), it is less clear whether these women can reduce their risk if they lose weight during the postmenopausal period. It is possible that the adverse effects of being overweight are hard to reverse at older ages and can best be prevented by avoiding overweight and obesity throughout life.
Overweight and weight reduction also illustrate the complexity of framing guidance on action when the consequences of an exposure differ among groups in a population. Whereas evidence indicates that greater weight is associated with an increased risk for breast cancer for postmenopausal women, it also indicates that greater weight is associated with a lower risk of breast cancer for premenopausal women (WCRF/AICR, 2007). Therefore, avoiding overweight is not a reasonable strategy for reducing the low, but still present, risk of premenopausal breast cancer, although avoiding overweight has many other important health benefits for women of all ages.
The association between shift work and increased risk of breast cancer concerns the committee, but it does not see a sound basis, at this time, for proposing action. More research is needed to understand the mechanisms underlying the association between shift work and breast cancer and to develop a clearer, more consistent characterization of the kind of work or work schedule that is associated with increased risk. This deeper understanding is needed to guide any effort to frame and test interventions in a realm with significant socioeconomic ramifications. A specific call for research on shift work is included in the recommendations in Chapter 7.
In some cases, the available evidence from animal or mechanistic studies suggests that a chemical or other factor may be a hazard, but evidence to directly assess the breast cancer risk for women is lacking (or perhaps not possible to obtain). In such circumstances, policy makers may use formal risk assessments to gauge the magnitude of possible risk and the appropriateness of actions to mitigate it. The identification of hazards—factors that have the ability to cause adverse effects—is an essential element in risk assessment and is often based on laboratory studies of biological mechanisms and effects of exposures on laboratory animals. Estimates of risk represent the probability that a particular adverse outcome—breast cancer in this case—will occur in an individual person or a population as a result of defined exposures to a hazard. A risk assessment considers not only the hazard of the substance, but also its potency (roughly speaking, how strong its effect is for a given dose) and the magnitude, nature, and timing of expected human exposure. A highly potent carcinogen may pose
substantial risk to an exposed individual, but if exposure to the general public is very low or extremely uncommon, the population risk will tend to be low. Alternatively, a low-potency carcinogen may pose risks that are low, but if exposures are common, it may be associated with a measureable effect in the population as a whole.
The committee did not undertake formal risk assessments for the environmental chemicals it found to be biologically plausible or possible contributors to breast cancer. Critical pieces of information were lacking, particularly robust data for estimating the magnitude of human breast cancer risk for a given dose (potency) at different life stages, and the prevalence and magnitude of the exposures across the population at different life stages. These data gaps were an obstacle to proposing evidence-based action that women could take to reduce risks from exposure to any particular chemical.
With these limitations to the evidence in mind, the committee highlights here the areas where it sees the clearest indications of opportunities for actions to reduce breast cancer risk. These actions are reviewed in this section and summarized in Table 6-1. It is important to recognize that the evidence is generally more extensive and therefore stronger for postmenopausal women than for premenopausal women, and for white, non-Hispanic women than for women of other races and ethnicities. In addition, some of the prevention opportunities that the committee points to appear more likely to apply to the prevention of the more common estrogen receptor– positive (ER+) tumors than estrogen receptor–negative (ER–) tumors. Younger women, however, tend to have ER– forms of breast cancer, as do women who have strong inherited susceptibility to breast cancer, such as carrying a mutation in BRCA1. All women should know their personal risk factors for breast cancer and seek clinical guidance from their health care providers regarding their breast cancer risk and how to modify it.
Among the strongest evidence reviewed by the committee regarding environmental exposures that have been causally linked to breast cancer was the evidence on ionizing radiation. Based on standard models developed from the radiation exposures of the Japanese atomic bomb survivors, it is commonly assumed that the breasts are most sensitive to carcinogenic effects of radiation at early ages (e.g., below ages 20–30). Nevertheless, models also predict elevated risks after exposure, even in middle age (Berrington
|Opportunity for Action||Strength of Evidence That Exposure Is Associated with Breast Cancer Riska||Modification of Exposure|
|Personal Action Possible||Requires Action by Others|
|Avoid inappropriate medical radiarion exposured||+++||Yes||Yes|
|Avoid combination menopausal hormone therapy, unless medically appropriatee||+++||Yes||Confer with physician|
|Avoid or end active smoking||+||Yes||Others can facilitate|
|Avoid passive smoking||(no committee consensus)||Varies||Yes|
|Limit or eliminate alcohol consumption||++||Yes||Others can facilitate|
|Maintain or increase physical activity||--f||Yes||Others can facilitate|
|Maintain healthy weight or reduce overweight or obesity to reduce postmenopausal risk||+++||Yes||Others can facilitate|
|Target Population Defined||Effective Form and Timing Establishedb||Affects Risk for Specific Subtype||Other Prominent Known Risks or Benefits from Taking Actionc|
|All ages||Yes, especially at younger ages||?||May result in loss of clinically useful information in some instances Likely to decrease risk for other cancers|
|Postmenopausal women||Yes||ER+||May experience moderate to severe menopausal symptoms, continued menopausal associated bone loss|
|All ages, especially before first pregnancy||Yes (form)
|?||Likely to reduce risk for other cancers, heart disease, stroke|
|All ages||Yes||?||Likely to reduce risk for other cancers, heart disease|
|All women||Yes (form)
|ER?||May increase risk for cardiovascular disease No known benefit of high alcohol consumption|
|All ages||No||?||Likely to reduce risk for cardiovascular disease, diabetes May increase risk for injury|
|Unclear||No||ER+?||Likely to reduce risk for cardiovascular disease, diabetes, other cancers|
|Opportunity for Action||Strength of Evidence That Exposure Is Associated with Breast Cancer Riska||Modification of Exposure|
|Personal Action Possible||Requires Action by Others|
|Limit or eliminate workplace, consumer, and environmental exposure to chemicals that are plausible contributors to breast cancer risk while considering risks of substitutesg||Varies by chemical||Varies||Yes|
|If at high risk for breast cancer, consider use of chemoprevention||---h||Yes||Confer with physician|
aThe assessments of the evidence of an association between an exposure and risk of breast cancer are qualitative representations of the committee’s conclusions from its review of available evidence: strong conclusion of increased risk, +++; moderately strong conclusion of increased risk, ++; conclusion of increased risk, +; unclear, ?; conclusion of reduced risk, –; moderately strong conclusion of reduced risk, – –; strong conclusion of reduced risk, – – –.
bActions to address risk factors can take various forms, some of which may be more effective than others. For example, increasing physical activity might be based on amount of time spent in any one exercise opportunity, on increasing specific types of exercise, or increasing the frequency of exercise, or perhaps some combination of any of these. Studies have not been done that provide evidence that a specific form of physical activity is optimal for reducing breast cancer risk.
cThe committee’s comments on other benefits or risks highlight major considerations, but are not intended to be exhaustive.
de Gonzalez et al., 2009; Shuryak et al., 2010). Recent models suggest the possibility that exposures to ionizing radiation across the age range of 10 to 50 years may result in excess relative risks of tumors in breast tissue that are more similar than previously estimated (Preston et al., 2007).
For the U.S. population, about half of the exposure to ionizing radiation comes from medical radiation, primarily in the diagnostic setting and
|Target Population Defined||Effective Form and Timing Establishedb||Affects Risk for Specific Subtype||Other Prominent Known Risks or Benefits from Taking Actionc|
|Varies||No||?||May reduce risk for other forms of cancer May result in replacement with products that have health or other risks not yet identified|
|High-risk women||Yes||ER+||Depending on the agent, increased risk of endometrial cancer, stroke, deep-vein thrombosis among others|
dWhile recognizing the risks of ionizing radiation exposure, particularly for certain higher dose methods (e.g., CT scans), it was not the committee’s intent to dissuade women from routine mammography screening, which aids in detecting early-stage tumors.
eCombination hormone therapy with estrogen and progestin increases the risk of breast cancer and the associated risk is reduced upon stopping therapy. Oral contraceptives are also associated with an increased risk of breast cancer while they are being used. This risk is superimposed on a low background risk for younger women, who are most likely to use oral contraceptives. These contraceptives are associated with long-term risk reduction for ovarian and endometrial cancer.
fReflects reduced risk of breast cancer associated with greater physical activity.
gPlausibility may be indicated by epidemiologic evidence, animal bioassays, or mechanistic studies.
hReflects reduced risk of breast cancer associated with use of chemopreventive agents.
especially from computed tomography (CT) scans and myocardial perfusion imaging (Fazel et al., 2009). As outlined in a paper commissioned by the committee (see Appendix F), the average annual dose of radiation from medical diagnostic sources in the U.S. population approximately doubled from 1985 to 2006 (Smith-Bindman, 2011). As further elaborated in that paper, there is evidence that exposure doses for the same imaging tests vary
widely among institutions. Superimposed on this variability is the element of human error, which has resulted in very high doses of radiation inadvertently delivered to patients by inadequately trained or supervised technicians and poorly designed equipment (e.g., Bogdanich and Rebelo, 2010; Smith-Bindman, 2010; Bogdanich, 2011). Extrapolating from estimates based on CT scans (Berrington de Gonzalez et al., 2009), Smith-Bindman (2011) estimated that among women in the United States in 2007, about 2,800 future breast cancers would result over their remaining lifetimes from exposure to all sources of medical diagnostic radiation delivered in 2007.2
There have been successful efforts through the Mammography Quality Standards Act (MQSA) to standardize and minimize radiation doses received from mammography, but the United States has no federal oversight for other imaging examinations and no guidelines on optimal doses. Evidence shows that physicians are insufficiently informed about radiation doses or the cancer risks attributable to the medical imaging they order (Lee et al., 2004). Evidence from patient surveys also shows that the public has little appreciation that just two to three abdominal CT scans can deliver radiation doses in the range of exposure experienced by Hiroshima survivors, doses that have been associated with elevated risks of breast and other cancers (Baumann et al., 2011).
The committee is encouraged that in 2010, the Food and Drug Administration (FDA) launched an Initiative to Reduce Unnecessary Radiation Exposure from Medical Imaging that addresses many of these areas (FDA, 2010).3 Sufficient resources and staff are needed for the development of detailed programs, full implementation of all components, and prospective evaluation of outcomes. The patient perspectives need to be incorporated into the planning of the programs, and breast cancer and other patient advocacy groups could provide important contributions to the development and evaluation of the programs. Radiology and imaging professionals are also seeking improvements. The Image Wisely program (http://www.image wisely.org/) is an effort to inform patients and professional colleagues of the importance of minimizing unnecessary exposure to ionizing radiation in imaging procedures. The Image Gently program (Alliance for Radiation Safety in Pediatric Imaging, 2011) specifically focuses on maximizing safety for children (Don, 2011; Moreno, 2011).
Used properly, medical imaging, including mammography, is a valu-
2Ionizing radiation is also an important tool for treatment of breast and other cancers and other conditions. For individuals who have been diagnosed with cancer, the benefits and risks of exposure to ionizing radiation are different from those for an individual who does not have cancer. However, even in treatment settings, patients can be exposed to excessively high doses of radiation because of errors or equipment malfunctions.
3A broad description of this initiative is available at http://www.fda.gov/Radiation-EmittingProducts/RadiationSafety/RadiationDoseReduction/ucm199994.htm#_Toc253092879.
able tool in diagnosing illness and guiding treatment, but unnecessary or improper use may increase risks because of the exposure to ionizing radiation. The committee sees important opportunities for actions at several levels that may contribute to lowering the risk of breast cancer due to exposure to ionizing radiation by improving medical imaging procedures.
Patients and Families: Individuals can question health providers specifically about what is known regarding the health benefits and harms associated with proposed diagnostic tests that involve exposure to ionizing radiation. They can request information about the relative doses of radiation associated with each type of procedure they undergo. Informed women may be able to avoid unnecessary tests for themselves and their families.
Health Professionals: Medical education programs and training can be created to enhance health care providers’ and students’ understanding of the doses of radiation involved in diagnostic imaging tests, and the health risks associated with those doses. Continuing education opportunities can include evaluation of the medical literature on the health benefits and harms of diagnostic imaging. Technicians can be trained in the avoidance of radiation overdoses and in methods to minimize dose while maintaining image quality.
Hospitals and Medical Practices: Hospitals and medical practices can make every effort to obtain previous imaging tests that have recently been done in other settings and avoid repeating imaging studies only for convenience. Expanding use of electronic medical records and ensuring compatibility and interoperability of records and digital films may facilitate the transmission of images between facilities.
Industry: Manufacturers could be encouraged to engineer diagnostic imaging devices to maximize safety and minimize human error. Through organizations and other collaborative mechanisms that promote the development of industry standards, manufacturers could adopt design standards that promote safe operation of the equipment. They might also take steps such as increasing the similarity of the “look and feel” of imaging equipment so that technicians can better transfer operating skills across manufacturers’ machines, thereby reducing errors.
Public Health: Public education campaigns could help inform consumers about levels of exposure to ionizing radiation that result from relevant medical procedures and the cumulative risks of such exposures. Educational campaigns could also encourage consumers to work with their health care providers to minimize exposures and might provide consumer tools to help track those exposures.
Regulation: Relevant agencies can work with the appropriate professional societies and scientists trained in evidence-based decision making
to develop standardization of dosimetry and evidence-based guidelines for appropriate use of tests for screening, diagnosis, and follow-up.
Menopausal and Contraceptive Hormone Use
Menopausal Hormone Therapy
The Women’s Health Initiative (WHI), which included a randomized clinical trial to assess the health effects of combination hormone therapy (i.e., a product with estrogen and progestin), demonstrated an increased risk of breast cancer among postmenopausal women taking combination hormone therapy (Writing Group for the Women’s Health Initiative Investigators, 2002). The annualized incidence rate of invasive breast cancer over the intervention period was 0.38 percent for women taking combination estrogen–progestin hormone therapy and 0.30 percent for women on placebo (HR = 1.26, 95% CI, 1.00–1.59). The association between combination hormone therapy and increased risk of breast cancer confirmed prior results from observational studies (Collaborative Group on Hormonal Factors in Breast Cancer, 1997).
The increased risk associated with current use of combination hormone therapy has been found to decline when use stops (e.g., Chlebowski et al., 2009; Beral et al., 2011). Follow-up of the WHI clinical trial participants demonstrated that the risk of breast cancer declined rapidly after combination hormone treatment ended (Chlebowski et al., 2009). Two or more years after the end of the intervention period, the annualized incidence of breast cancer among those assigned to the combination hormone therapy group was 0.49 percent compared to 0.42 percent among women assigned to placebo (HR = 1.19, 95% CI, 0.59–2.42). In the observational Million Women Study in the United Kingdom (Beral et al., 2011), former users’ risk became comparable to that of never users within 4 years. After publication of the WHI results in 2002, rates of hormone therapy use declined rapidly, and coincident with that, U.S. breast cancer rates among women ages of 50 to 69 years were observed to decline by 11.8 percent (95% CI, 9.2–14.5) between 2001 and 2004 (Ravdin et al., 2007).
The committee is confident in urging that women avoid or minimize use of combination hormone therapy, thereby avoiding increasing their risk of breast cancer. The WHI considered other outcomes in addition to breast cancer, including heart disease, fractures, stroke, and colorectal cancer. Overall, the increased risks for breast cancer, heart disease, and stroke were considered to outweigh the reduction in risk for hip fractures and colorectal cancer (Writing Group for the Women’s Health Initiative Investigators, 2002). Managing menopausal symptoms is a common reason for women to consider taking hormone therapy, but women should confer with their
health care providers to determine the most appropriate way to manage these or other symptoms.
This position is consistent with the guidance from the U.S. Preventive Services Task Force (USPSTF, 2005). The USPSTF guidelines recommend against the routine use of combination hormone therapy for the prevention of chronic conditions because the increased risks of breast cancer, stroke, and other conditions are considered to outweigh the potential benefits of reduced risks for fractures and colorectal cancer. However, this guidance specifically excludes consideration of management of menopausal symptoms. The USPSTF advises women and their health care providers to give individualized consideration to personal risk factors and preferences in deciding whether use of hormone therapy is medically appropriate.
The Endocrine Society (Santen et al., 2010) issued a scientific statement that provides a detailed review and grading of evidence on benefits and harms associated with use of postmenopausal hormone therapy (HT). The highest-quality evidence concerning combination HT and breast health is that it increases mammographic density, which is associated with increased risk of breast cancer (e.g., Cuzick, 2008; Boyd et al., 2009, 2010). Other evidence concerning breast cancer risks was considered less strong. The statement advocates individualized assessments of women’s potential risks and benefits but suggests that some form of HT may be appropriate for menopausal women younger than age 60.
The WHI also included a trial of estrogen-only hormone therapy among women who had a hysterectomy. Women with a hysterectomy who took estrogen-only therapy were less likely to develop invasive breast cancer during the intervention period and the subsequent period after the intervention ended than women who took the placebo, but the absolute differences were small (LaCroix et al., 2011). Over the entire follow-up period (during the intervention and afterward), the incidence of breast cancer in the group that took conjugated estrogen was 0.27 percent, compared with 0.35 percent in the placebo group (HR = 0.77, 95% CI, 0.62–0.95). Although this WHI intervention trial observed no excess risk with estrogen-only therapy among women who had a hysterectomy, observational studies have found a small increased risk of breast cancer (Million Women Study Collaborators, 2003; Beral et al., 2011). One reason for this discrepancy between observational studies and the randomized clinical trials may be that observational studies are more likely to have misclassification of exposure between combination hormone therapy and estrogen-only therapy. Observational studies also suggest that these risks are higher in lean compared with obese women (Huang et al., 1997; Reeves et al., 2006; Brinton et al., 2008; Beral et al., 2011).
A criticism of the combination hormone therapy products that are prescribed most often and used in clinical trials of menopausal hormone
therapy is that they are synthetic. The concern is that they therefore may be associated with risk not present for products that are considered more “natural,” often referred to as bioidentical hormones. Bioidentical hormones are derived from plants and treated to have the same chemical structure as endogenous human hormones (Cirigliano, 2007). They may be commercially available or individually compounded in pharmacies. The Endocrine Society (2006) has issued a position statement on bioidentical hormones, emphasizing the lack of data about their safety and effectiveness, expressing concern about potentially misleading or inaccurate claims, and supporting FDA regulation of all hormone products. A review of bioidentical hormone therapy (Cirigliano, 2007) supports the concerns raised by The Endocrine Society, citing a lack of evidence to support claims of improved safety with bioidentical hormones. Because observational studies have consistently shown an increased risk of breast cancer among women with higher endogenous estrogen and androgen serum concentrations (Key et al., 2002; Hankinson, 2005–2006), there is little to suggest that use of bioidentical hormones would be a safe alternative to other forms of hormone therapy.
Oral contraceptives with combination estrogen and progestin are also associated with an increased risk of breast cancer while women are using them (Collaborative Group on Hormonal Factors in Breast Cancer, 1996; Marchbanks et al., 2002; Hunter et al., 2010), and the risk may be greater for certain product formulations (Hunter et al., 2010). The increased risks associated with oral contraceptives are short term, and they decline after use ends, as is the case with hormone therapy after menopause. This increased risk occurs against a background of low risk for the younger women who are most likely to be taking oral contraceptives. The result is that the overall impact on the incidence of breast cancer is small. In addition, oral contraceptive use is associated with a long-term reduction in the risk of both ovarian and endometrial cancers (reviewed in La Vecchia, 2001). Use of oral contraceptives in the perimenopausal period would be expected to be associated with risks similar to hormone therapy use during the same window, although data on this practice are limited (Davidson and Helzlsouer, 2002).
Chemoprevention for Women at Increased Risk of Breast Cancer
The committee noted that for women at increased risk of breast cancer, chemoprevention with tamoxifen or raloxifene has been shown in clinical trials to reduce the risk of developing breast cancer. The clinical trials evaluating the ability of these medications to reduce the risk of breast cancer
considered women with an estimated risk of being diagnosed with breast cancer within 5 years of 1.7 percent or greater to be at increased risk and eligible for the trials.4
Research has demonstrated that drugs that alter responses to estrogen (e.g., selective estrogen receptor modulators or SERMs) or production of estrogen (e.g., aromatase inhibitors) can substantially reduce risk of ER+ breast cancer (e.g., Cummings et al., 2009; Nelson et al., 2009; Goss et al., 2011). Tamoxifen and raloxifene, both SERMS, are two of the best known and best studied products of this type. The FDA has approved their use for this purpose by women who are considered at increased risk of breast cancer and are not at increased risk for cerebrovascular disease. Raloxifene is approved for use only after menopause. Other medications not currently approved for use for breast cancer risk reduction are also being evaluated. One category of drugs is aromatase inhibitors, designed to inhibit the conversion of androgen to estrogen, and early results from a study of the aromatase inhibitor exemestane show a reduced risk of breast cancer among high-risk women (Goss et al., 2011). Other medications being studied include other SERMs and aromatase inhibitors, bisphosphonates, and metformin (Cuzick et al., 2011).
A meta-analysis of randomized controlled trials of breast cancer prevention reported that with 5 years of tamoxifen use, women at high risk had a statistically significant reduction in the risk of invasive ER+ breast cancer (meta-analysis risk ratio 0.70, 95% CI, 0.59–0.82) compared with women who had not used the drug (Nelson et al., 2009). The same authors also reported a meta-analysis of randomized controlled trials for raloxifene, finding a statistically significant reduction in invasive ER+ breast cancer for women who used the drug compared to those who did not (meta-analysis risk ratio 0.44, 95% CI, 0.27–0.71). However, they noted that the mean age at entry into the tamoxifen studies ranged from 47 to 51 years, compared with a mean ranging from 67 to 68 years for the raloxifene studies. The authors estimated that use of tamoxifen or raloxifene for 5 years would be expected to result in 7 to 10 fewer breast cancer cases per 1,000 women per year (Nelson et al., 2009).
The Study of Tamoxifen and Raloxifene (STAR) trial was designed to provide a direct comparison of the two products. The study population consisted of postmenopausal women who were at increased risk of breast cancer, but did not have a history of cancer or various other condi-
4The risk of developing breast cancer can be estimated from statistical models that consider factors such as age, reproductive history, and personal and family history of breast cancer. In the United States, a commonly used model is the Breast Cancer Risk Assessment Tool (Gail et al., 1989; NCI, 2011) (available at http://www.cancer.gov/bcrisktool/). It is discussed further later in this chapter.
tions, including stroke, uncontrolled diabetes, or uncontrolled hypertension (Vogel et al., 2010). With 6.75 years of follow-up, the ratio of risk for invasive breast cancer with use of raloxifene to that with use of tamoxifen was 1.24 (95% CI, 1.05–1.47) (Vogel et al., 2010). Although the use of raloxifene reduced risk less than use of tamoxifen, raloxifene had fewer adverse effects than tamoxifen. Relatively few eligible women have chosen to use tamoxifen or raloxifene, at least in part because of their association with increased risk for serious adverse health effects, including endometrial cancer (tamoxifen) and stroke (Fisher et al., 2005; Vogel et al., 2010).
The committee endorses recommendations of the USPSTF (2002) that women have their breast cancer risk assessed and discuss with their health care providers whether use of tamoxifen or raloxifene as chemoprevention to reduce their risk of breast cancer is appropriate for them. Risk assessment to weigh the potential benefits and risks should be available to all women. Use may be appropriate for women who are at increased risk of breast cancer (a 5-year risk of at least 1.7 percent) and who have low risk for the adverse effects associated with these medications. The adverse effects can include menopausal symptoms, risk of deep vein thrombosis (blood clots), endometrial hyperplasia and cancer (for tamoxifen), and stroke. Benefits for menopausal women, in addition to breast cancer risk reduction, include lower fracture risks. Statistical models are available to help guide decision making regarding the use of chemoprevention for premenopausal women ages 35 and older (Gail et al., 1989) and for menopausal women ages 50 and older who are at increased risk for breast cancer (Freedman et al., 2011).
Active and Passive Smoking
Accumulating evidence points to active smoking being associated with an increase in risk for breast cancer (Reynolds et al., 2004; CalEPA, 2005; Ha et al., 2007; Collishaw et al., 2009; Secretan et al., 2009). Some evidence indicates the most consistent findings are for earlier initiation of smoking and smoking before a first full-term pregnancy (DeRoo et al., 2011; Luo et al., 2011; Xue et al., 2011). In addition, some expert reviews have concluded that the evidence is consistent with a causal association between passive smoking and increased risk for premenopausal breast cancer (CalEPA, 2005; Collishaw et al., 2009), while the evidence regarding passive smoking is described as inconclusive in the most recent review by the International Agency for Research on Cancer (IARC) review (Secretan et al., 2009). Some evidence also suggests a possible association between high levels of exposure to passive smoking and postmenopausal breast cancer (Reynolds et al., 2009; Luo et al., 2011).
Smoking poses substantial health risks in addition to any contribution
it may make to increased risk of breast cancer, and the committee has no hesitation in urging women not to begin smoking, to stop smoking if they are current smokers, and to protect themselves and their children from exposure to secondhand smoke. Women who have become smokers are certainly likely to gain health benefits by ceasing to smoke.
Exposure to secondhand tobacco smoke increases the risk of several diseases (HHS, 2006), and so it should be avoided. Public and private policies that call for smoke-free environments in public spaces and workplaces help reduce exposure to secondhand smoke, especially for adults. However, children and nonsmoking adults who live with smokers may still be exposed within the home and in private cars. The evidence of increased potency of smoking before pregnancy suggests the possibility that a similar window of greater vulnerability may exist at younger ages for exposure to secondhand smoke, although exposure to secondhand smoke only during childhood does not appear to increase the risk of breast cancer (HHS, 2006; Chuang et al., 2011; Luo et al., 2011).
There is opportunity for improving health at both the individual and societal level through reduction in both active and passive exposure to tobacco smoke.
Individuals: Girls and women can avoid beginning to smoke, and those who smoke can quit. Individuals can also avoid exposing themselves and their children to secondhand smoke.
Public and private sectors: Efforts can be made to expand smoke-free environments in workplaces and public spaces. Educational programs can inform smokers and nonsmokers of the dangers that secondhand smoke presents. Efforts can also be made to encourage smoke-free homes and cars. Given that initiation of smoking generally occurs in adolescence or earlier, the evidence linking smoking before a first full-term pregnancy with increased risk of breast cancer underscores the need for effective programs geared towards smoking prevention in preteen and teenage girls.
Alcohol consumption has been shown to modestly increase risk for both pre- and postmenopausal breast cancer (IARC, 2010), with the largest studies suggesting a linear relation between intake and risk. Risk was estimated to increase approximately 7 percent (Collaborative Group on Hormonal Factors in Breast Cancer, 2002) to 9 percent (Smith-Warner et al., 1998) for each additional 10 grams of alcohol consumed per day. (In the United States one drink is considered to contain approximately 14 grams of alcohol [CDC, 2011].) An analysis of data from 53 studies found that women who had substantial levels of daily alcohol consumption (. 45 g per day) had a relative risk of breast cancer of 1.46 (95% CI, 1.3–1.6), com-
pared to those who reported drinking no alcohol (Collaborative Group on Hormonal Factors in Breast Cancer, 2002). For those whose consumption was approximately one to two drinks per day (15–24 g), the relative risk was 1.13 (95% CI, 1.08–1.19) (Collaborative Group on Hormonal Factors in Breast Cancer, 2002).
Questions remain unresolved about whether the association between breast cancer and alcohol consumption is cumulative over years of exposure, or a time-limited and reversible association (IARC, 2010). Some studies also suggest that the increased risk associated with higher alcohol consumption (> 20 g/day) is primarily among women who use menopausal hormone therapy (Gapstur et al., 1992; Chen et al., 2002; Horn-Ross et al., 2004); however, an IARC (2010) review reported no significant variation. The health risks and potential benefits of moderate alcohol intake were evaluated and published as a formal position paper by the National Institutes of Health (Gunzerath et al., 2004). In addition to breast cancer, consumption of more than one to two drinks per day for women (more than two to three drinks per day for men) is associated with an increased risk for a variety of other cancers and other adverse health conditions (e.g., WCRF/AICR, 2007; Gronbaek, 2009; IARC, 2010). However, the moderate levels of consumption that are associated with an increased risk of breast cancer are also associated with positive outcomes such as lower mortality from cardiovascular disease (Maskarinec et al., 1998; Klatsky, 2009; Ronksley et al., 2011), which is a much larger contributor to morbidity and mortality among women than breast cancer. A meta-analysis found that for those who consumed an average of one drink or less (2.5–14.9 grams) per day, the relative risk of cardiovascular disease mortality was 0.77 (95% CI, 0.71–0.83) compared with those who consumed no alcohol (Ronksley et al., 2011). A similar reduction in risk was seen for the incidence of coronary heart disease, one specific type of cardiovascular disease.
With respect to balancing alcohol’s risk of breast cancer with potential benefits, Gunzenrath and colleagues advised in their position paper, “individual women, with the help of their physicians, must weight their potential increased risk for breast cancer against their potential reduced risk for CHD [coronary heart disease] in determining whether alcohol consumption should be reduced” (Gunzerath et al., 2004, p. 833).
The committee concluded that the consistent evidence that even moderate consumption of alcohol is associated with an increased risk of breast cancer warranted note in this discussion of modifiable risk factors. However, with the lack of evidence regarding the impact on breast cancer risk of changes in consumption and the evidence supporting beneficial effects of moderate alcohol consumption related to cardiovascular disease, the merits of restricting or eliminating moderate alcohol consumption as a breast cancer risk reduction strategy are hard to judge for individual women. The
committee urges women to confer with their health care providers about the potential benefits and risks of reducing their alcohol consumption.
Reviews by the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR, 2007, 2010) characterized as probable an association between greater physical activity and a reduction in risk for postmenopausal breast cancer. The evidence regarding reduction in risk for premenopausal breast cancer is described as limited. Other reviews (e.g., Monninkhof et al., 2007; Physical Activity Guidelines Advisory Committee, 2008; Friedenreich, 2010) have produced similar assessments of the available evidence. The beneficial effects of physical activity appear to be stronger for women of normal weight and without a family history of breast cancer; they are, however, observed in women of all races and ethnicities (Friedenreich, 2010).
Additional research is needed to clarify the type of activity, the amount, and the timing of physical activity over the life course that can produce a reduction of breast cancer risk. Three primary prevention studies (McTiernan et al., 2004a,b; Monninkhopf et al., 2009; Friedenreich et al., 2010, 2011; also reviewed in Winzer et al., 2011) offer some initial insight into the feasibility of exercise interventions to reduce risk among inactive postmenopausal women, most of whom were overweight. In these 1-year trials, it was not possible to measure changes in breast cancer risk directly. Outcomes were assessed on the basis of a variety of biomarkers considered relevant to breast cancer risk. Among the biomarkers were weight, body mass index (BMI), sex hormone concentrations, mammographic density, and insulin concentrations. For example, moderate to vigorous aerobic exercise of approximately 3 hours per week (3 to 4 days per week) among previously sedentary women ages 50–74 resulted in statistically significant decreases in weight, BMI, and abdominal fat (Friedenreich et al., 2011). In another study (McTiernan et al., 2004b), an average of nearly 3 hours per week of moderate intensity exercise among postmenopausal women resulted after a year in a statistically significant decline in serum estrogen levels, but only among the women whose percent body fat decreased by at least 2 percentage points. But a study that tested a program of 2.5 hours per week of combined aerobic exercise and strength training did not detect a significant change in serum estrogen levels, even among the women whose percent body fat declined (Monnihkhopf et al., 2009). The results of studies such as these suggest that changes considered likely to be indicative of reduced risk for breast cancer can be achieved, but only with greater frequency and duration of exercise.
Physical activity throughout the life course is generally recognized as
having wide-ranging health benefits, which include the likely reduction in risk for postmenopausal breast cancer among women who are more active. The committee endorses the guidance of the Department of Health and Human Services (HHS, 2008) for regular physical activity at all ages.
Excess Weight and Weight Gain
As discussed in Chapter 3, data from 2007–2008 indicate that approximately 36 percent of adult women of all ages can be considered obese and another 29 percent as overweight (Flegal et al., 2010). The systematic review by the WCRF/AICR (2007) and subsequent updates (WCRF/AICR, 2008, 2010) classified greater body fatness5 as convincingly associated with greater risk for postmenopausal breast cancer and adult weight gain as probably associated with increased risk. Some studies have found that the increased risk associated with weight gain is stronger for women who have not used HT (Eliassen et al., 2006; Ahn et al., 2007).
For younger women, however, greater body fatness is probably associated with reduced risk of premenopausal breast cancer (WCRF/AICR, 2007), although this is a time of life for which breast cancer risk is much lower than for older women. But weight gain earlier in life may be difficult to reverse later in life when the increase in risk caused by body fatness may have a greater effect because of the higher breast cancer risk at this life stage. It also appears that the association of greater weight and adult weight gain with increased postmenopausal breast cancer risk is dominated by the experience of white women and may not hold for African American women (Palmer et al., 2007).
The committee is persuaded that maintaining weight within what is considered a normal range (a BMI of 18.5–24.9) is appropriate guidance for all women. Overweight and obesity are associated with increased risk for a wide range of adverse health consequences beyond the specific relation to breast cancer. Preventing weight gain may be especially important because it is less clear whether overweight and obese women can reduce their risk of postmenopausal breast cancer by losing weight. The Nurses’ Health Study (Eliassen et al., 2006) and the Iowa Women’s Health Study (Harvie et al., 2005) found evidence of reduced risk for women who lost weight compared with those who maintained a stable weight. Other studies (Ahn et al., 2007; Teras et al., 2011), however, failed to find reduced risk among women who lost weight. For African American women, and perhaps other population
5Body fatness and overweight and obesity are commonly measured using body mass index (BMI). BMI is defined as body weight in kilograms divided by height in meters squared. The following weight categories are based on BMI values: underweight, <18.5; normal weight, 18.5–24.9; overweight, 25–29.0; and obese, .30.
groups for which data on weight-related breast cancer risk factor patterns are still limited, different or additional prevention strategies may be needed.
Chemicals and Consumer Products
The committee evaluated the potential role that some individual exogenously produced chemicals found in the diet, air, water, household products, and workplaces may play in the development of breast cancer in humans. Because there are vast numbers of such chemicals and often very limited evidence regarding breast cancer, the committee chose to examine the evidence for only a selected set (see Chapter 3). The comments that follow are typically specific to the chemicals that the committee reviewed. It is not possible for the committee to comment on the chemicals that it did not review. For some chemicals, relevant information may be available from other sources (e.g., Brody et al., 2007; California Breast Cancer Research Program, 2007; WCRF/AICR, 2007; EPA, 2011; IARC, 2011; NTP, 2011).
Ethylene Oxide, Benzene, and 1,3-Butadiene
Among the chemicals considered, the evidence for an association with increased risk of breast cancer was clearest for ethylene oxide. Benzene and 1,3-butadiene are also probably human breast carcinogens. Cigarette smoke is a source of exposure, either through active or passive smoking, to these chemicals (Fennell et al., 2000). All three substances are raw materials used in the production of numerous industrial chemicals. Ethylene oxide is also used for sterilization in industrial and medical settings. Benzene has been used as a fuel additive and is a natural constituent of crude oil. Vehicular emissions and gasoline vapors at filling stations are a source of exposure to both benzene and 1,3-butadiene. Although ambient air levels have been substantially curtailed through regulatory actions, widespread, low-level environmental exposure, especially to benzene, continues. While recognizing potential hazards, the committee did not have the capacity to estimate breast cancer risks at these low doses because the information necessary to do so is insufficient.
Because these chemicals are recognized carcinogens (NTP, 2011), steps are taken to reduce occupational and public exposures. However, there is limited awareness of the possible association between these chemicals and increased risk for breast cancer, and federal occupational health standards do not call for medical surveillance for breast cancer for exposed workers.6 Women whose work involves the potential for exposure to these chemicals
6The medical surveillance guidelines for workers exposed to these chemicals are available at 29 CFR 1910.1028 App C, 29 CFR 1910.1051 App C, and 29 CFR 1910.1047 App C.
may be able to take additional steps to minimize their exposure. Accomplishing that goal, however, will require awareness of the possibility of exposure and access to appropriate resources, procedures, and policies to make minimizing exposure possible. Although women can accomplish some of this on their own, they will also have to depend on actions by employers, equipment manufacturers, and agencies responsible for ensuring workplace and environmental safety to limit or eliminate exposure. The general public can minimize exposure through avoidance of tobacco smoke and by limiting exposure to gasoline vapors and vehicular exhaust.
Other Environmental Agents
For many of the other chemicals that the committee considered, as well as those discussed in reviews by others (e.g., Brody et al., 2007; Rudel et al., 2007; Gray, 2010), little or no epidemiologic evidence on breast cancer risk is available. However, evidence from in vivo cancer bioassays, mechanistic studies, or both may suggest the potential for exposure to contribute to breast cancer in humans. Where these indications exist, the committee recommends further research to improve understanding of the relevance of the findings for humans (see Chapter 7). Many of these chemicals have been identified as probable or likely carcinogenic hazards by authoritative organizations (e.g., IARC, EPA, National Toxicology Program), but the findings are not specific to breast cancer hazard.
For two of the agents—hair dyes and non-ionizing radiation—substantial epidemiologic evidence from large populations studied over long periods of time has consistently failed to identify a significant increase in risk of breast cancer associated with exposure. The committee concluded that avoiding exposure to either hair dyes or non-ionizing radiation has little potential to contribute to a substantial reduction in breast cancer risk for individuals or the population. For certain other compounds, such as dioxins, polychlorinated biphenyls (PCBs), some metals, and vinyl chloride, regulatory actions taken many years ago have greatly reduced exposures. However, low-level exposure continues because these chemicals persist in the environment or some sources are difficult to eliminate even if they are subject to regulatory controls. Although individuals may be able to control some sources of exposure to some of these persistent chemicals (e.g., by avoiding certain types of fish known to have high levels of PCBs or dioxins), it may be difficult for individuals to act on their own to avoid or limit many of these low-level exposures.
For many of the reviewed compounds, including those discussed above, evidence of hazard may be present, but information to assess the magnitude of risk, particularly at environmentally relevant doses is lacking or
inadequate, posing a substantial challenge for gauging the extent to which an individual’s actions may reduce risk (Table 6-1). For example, for BPA, epidemiologic data are largely lacking, and the available studies on timing of exposure do not adequately address potentially important windows such as fetal and early life that may influence adult disease. Avoidance or reduction of exposures to such substances at the individual level may be difficult or infeasible in some cases, but eminently possible in others. For example, a small study demonstrated substantial reduction in urinary levels of BPA when participants shifted to use of minimally packaged foods (Rudel et al., 2011). Levels of BPA increased when the study participants resumed eating packaged foods. Determining the sustainability of such changes, their acceptability to a broader population, and whether such reductions would actually decrease breast cancer risk would require further investigation.
The committee recognizes, however, that existing data indicate that BPA and some other substances may be hazards to human health and may well warrant consideration of actions by regulatory agencies that are aimed at reducing future population-based exposures. Other considerations for regulators may include the possibility that exposure to multiple chemicals that contribute to mechanisms involved in breast cancer (e.g., mutagens, endocrine disruptors, etc.) may present a cumulative risk that could be controlled in part through regulatory actions on individual substances. Even where evidence regarding breast cancer is limited, evidence related to other health effects (e.g., developmental effects or other types of cancer) may provide a stronger basis for regulatory action or individual efforts to avoid exposure.
Such policy action would be based on many factors, including taking into account the impact of foreseeable substitutions for a regulated substance and the likely prospect of unanticipated substitutions of substances with as yet unknown properties. Given the limits in the evidence base regarding breast cancer, and the complexity of the analysis it would entail, it is beyond the charge and capacity of this committee to make specific recommendations for regulatory action. However, it notes that GAO (2005, 2006, 2007, 2009a,b) has called several times for improvements in monitoring and regulation of toxic chemicals, citing both constraints resulting from the 1976 Toxic Substances Control Act (TSCA) and a need for better use of the authority it does provide. Under TSCA, EPA has limited authority to require that manufacturers test products for carcinogenicity (or other health hazards), and its authority to share information that may be provided is also limited. Interested organizations can help inform the public about the current provisions for testing chemicals and encourage manufacturers to improve testing and make existing information on their products more readily available.
Dietary Supplements and Cosmetics
Dietary supplements and cosmetics are widely used products, but the FDA has limited authority to test their safety before they are marketed. Rules regarding FDA regulation of dietary supplements and cosmetics differ from those covering pharmaceutical agents. Since the passage of the 1994 Dietary Supplement Health and Education Act (DSHEA), manufacturers are responsible for ensuring that their supplement products are safe and that product label information is truthful and not misleading. No FDA approval or proof of safety or efficacy is required before dietary supplements are marketed (FDA, 2009). Similarly for cosmetics, the products are not subject to premarket approval by the FDA under the laws governing the sale and use of cosmetics (FDA, 2005).
Data from the National Health Interview Survey indicate that about 114 million Americans, or more than half of the U.S. adult population, consume dietary supplements (Cohen, 2009). Vitamin and some nutrient supplements have been well studied, but many supplements marketed as alternatives to prescription hormone therapies (e.g., for control of perimenopausal symptoms or weakness in old age), or for improvement of athletic performance or weight loss, have not been tested for safety or effectiveness. Interest in such supplements could be amplified by messages that hormone therapies, physical inactivity, and overweight are risk factors for breast cancer. Similarly, cosmetics that are widely used by girls and women of all ages may also contain hormonally active ingredients that are intended to produce a more youthful appearance.
The limited role for the FDA in the marketing of dietary supplements and cosmetics is poorly understood. In a 2002 Harris poll, a majority of respondents believed that dietary supplements are approved by a federal regulatory agency (Taylor and Leftman, 2002). Moreover, in an online questionnaire completed by medical residents affiliated with 15 internal medicine programs, baseline knowledge about regulation of dietary supplements was poor and did not vary by training year of residency (Ashar et al., 2007). A third of the residents were not aware that the FDA does not require premarketing submission of safety or efficacy data.
The FDA does have the authority to withdraw dietary supplements and cosmetics from the market if product adulteration is discovered or if cosmetics are found to be misbranded. The FDA can declare such products adulterated when they present an unreasonable risk of illness or injury under the conditions of use. For example, a number of side effects—breast enlargement, loss of libido, cardiovascular side effects, thromboembolism, and bleeding—were found in men with prostate cancer who were taking the herbal dietary supplement PC-SPES. When chemical and bioassays of PC-SPES lots were performed, pharmacologic levels of diethylstilbestrol
(DES), warfarin, and indomethacin were found, leading to product withdrawal (Sovak et al., 2002; White, 2002). Currently, no prospective system is in place to routinely detect product adulteration prior to marketing.
The committee sees a need for better means for the FDA to prospectively survey or detect contaminants or ingredients in cosmetics and dietary supplements, including estrogenic substances that are known or possible causes of breast cancer, or otherwise monitor products designed to have pharmacologically active levels of such substances. It also urges consumer organizations and other interested groups to develop educational programs for the public and the health professions to enhance awareness of the rules governing marketing of dietary supplements and cosmetics, including that manufacturers are responsible for establishing the safety of these products and that the FDA has limited authority to act before they are marketed. Consumers and interested organizations can also urge manufacturers to provide consumers with more information regarding the presence of potentially hormonally active ingredients in dietary supplements and cosmetics, ideally by identifying such ingredients on product labels.
The committee has proposed several actions that women could take that may reduce their risk of breast cancer, but based on the existing literature, found it difficult to estimate the magnitude of the potential impact of these actions for either individuals or population groups. Although numerous studies have established associations between risk factors and breast cancer incidence, those associations may or may not be causal. If a risk factor is not causally linked to breast cancer, then changing exposure to that factor will not have a direct impact on breast cancer risk. In addition, there is limited research demonstrating that the effect of an exposure on breast cancer risk can be reversed by removing the exposure or, if it could be reversed, the magnitude of risk reduction that could be achieved by modifying or preventing the exposure. For example, for combination hormone therapy, for which a clinical trial has been conducted, the magnitude of breast cancer risk has been quantified, and therefore the excess risk that can be avoided by refraining from use of combination HT can be quantified.
Here, the committee offers some perspective on levels of breast cancer risks, interrelationships among risk factors, and what we know about risk reduction.
Average Risk for Breast Cancer
Estimates of risk and changes in risk should be viewed with an understanding of the incidence of the disease in the population. For breast cancer, incidence is very low until women reach their thirties, when it begins to rise steadily into older ages. But even among women in their seventies, only a small minority will be diagnosed with breast cancer. The data in Table 6-2 show the percentage of women who on average would be expected to receive a diagnosis of invasive breast cancer within 10 years of a given age. For example, for a group of 50-year-old white women at average risk, this 10-year risk is 2.43 percent Thus, out of 100 white women aged 50 years who are followed for 10 years, 2 to 3 will be diagnosed with breast cancer and 97 to 98 will not. It is also possible to calculate risk for longer periods, or even for a lifetime: the cumulative risk from birth to the end of life is 12.57 percent in white women (not shown in the table). This is how the familiar statistic of about “one in eight” white women expected to be diagnosed with breast cancer over a lifetime is derived (NCI, 2010, Table 4-18).
Because each number in Table 6-2 is an average among women in that age and race/ethnic group, there are obviously women whose risk is higher than the average and women whose risk is lower. The concept of relative risk (see Chapter 2) can be illustrated in this context. In a hypothetical 10-year study in a group of 50-year-old white women who have the average risk shown in Table 6-2, a certain risk factor might have a relative risk of 1.5—which is a 50 percent increase in risk. If half of the women have that risk factor and half do not, then the overall 2.43 percent 10-year risk of breast cancer would be just the middle ground between a risk of approximately 2.9 percent for those with the risk factor, and approximately 1.9 percent for those without it. In other words, for 50-year-old white women, the risk of being diagnosed with breast cancer in the next 10 years is about 3 out of 100 women in those who have the risk factor, compared with 2 of 100 women who do not.
Risk Estimates for Individuals
As in the hypothetical example just described, observational studies or controlled trials in groups of women produce estimates of the risk of breast cancer associated with given exposures that are based on the experience of the overall study population. A separate but related question is what this means for the individual who is exposed. Because many factors can increase or decrease an individual’s risk of cancer, the risk associated with a single factor has to be put into context.
The Breast Cancer Risk Assessment Tool (Gail et al., 1989; NCI, 2011a) and the Tyrer-Cuzick breast cancer risk assessment model (Tyrer et
|White||Black||Asian/ Pacific Islander||American Indian/Alaska Native||Hispanic (any race)|
NOTES: A percent of 0.00 represents a value that is less than 0.005. Incidence data are from the SEER 17 areas (San Francisco, Connecticut, Detroit, Hawaii, Iowa, New Mexico, Seattle, Utah, Atlanta, San Jose-Monterey, Los Angeles, Alaska Native Registry, Rural Georgia, California excluding SF/SJM/LA, Kentucky, Louisiana, and New Jersey).
SOURCE: NCI (2010).
al., 2004) are designed to generate estimates of absolute risk for individualsin conjunction with absolute risk estimates for the general population as a reference.7 The Breast Cancer Risk Assessment Tool (NCI, 2011a) uses a limited set of characteristics (e.g., age, reproductive history, family history) to assess risk for an individual who is a member of a group with these characteristics and generates an estimate of the absolute risk over the next 5 years for women in that group. For comparison, the model generates an estimate of average risk for women of the same age. An individual woman’s characteristics may put her in a group at higher or lower risk than the average. The performance of this model is better for white women than women of other races or ethnicities and is intended for women who are at least 35 years old (NCI, 2011b).
These tools are used primarily to guide decisions about medical care in clinical practice, including whether a woman’s risk of breast cancer is high enough to make her eligible for chemopreventive medications, such as
7The Breast Cancer Risk Assessment Tool is based on breast cancer rates in the U.S. population, and the Tyrer-Cuzick model is based on breast cancer rates from the United Kingdom.
tamoxifen or raloxifene. Because these risk assessment tools do not make use of information about environmental exposures, the committee did not examine them in detail.
Population Attributable Risk for Understanding the Relative Contributions of Different Factors
Table 6-2 provides not only estimates of risk for individual women, but also provides a framework for understanding the concept of population attributable risk (PAR), which was introduced in Chapter 2 and discussed in Chapter 4. The PAR represents a population-based measure of the percentage of excess cases associated with the exposure of interest (i.e., among the exposed in comparison with the unexposed) that also takes into account the distribution of the exposure within the population. While it has sometimes been defined as the proportion of all cases that would not have occurred if exposure to a causal factor was removed from the population (and Greenland, 1998), this definition represents the ideal: it assumes that all observed associations are actually causal. In reality, many associations observed in epidemiologic studies are confounded by other factors, and when those studies are used to estimate the PAR, the PAR becomes confounded. The PAR is useful for summarizing current understanding of the relative contributions of different factors to the overall “burden” or incidence of breast cancer in the population, when confounding has been adequately controlled. It is helpful for researchers and policy makers in assessing possible opportunities to reduce disease burden through public health interventions that target specific modifiable risk factors in the population as a whole, but has limited applicability for an individual.
For instance, the PAR can be used to estimate how many cases of breast cancer might be prevented if half of the women offered unnecessarily high levels of medical radiation were able to avoid those high levels, or if an additional 25 percent of women did not gain weight and become obese or overweight by the time they reached menopause. The PAR itself is an estimate of the maximum potential benefit of eliminating a risk factor; it is not an anticipated outcome, partly because most risk factors are unlikely to be completely eliminated and partly because some risk factors are proxies for others that are causal. In other words, the PAR estimates assume that the studies identified truly causal associations and that any remaining confounding or other biases would have little impact on the estimated role of the factor under study. Finally, the PAR may be different in other populations with a different combination of characteristics, even if the proportion with a specific risk factor is the same.
Unlike the risk estimates of Table 6-2, which are for women in the population before any of them develop breast cancer, the calculation of
the PAR is based on all cases of the disease observed in a specific population of women after they have been diagnosed. These cases of disease then represent 100 percent, and the PAR for a given risk factor is the proportion or percentage of these women who have breast cancer in whom that factor may play a causal role. Sometimes the PAR is calculated for a group of factors rather than a single one, but it is always calculated as a percentage of all persons in a specific population with the disease (e.g., women with breast cancer). Values calculated for individual risk factors cannot be summed to generate an estimate of their combined contribution to risk. This is because many cases of breast cancer are the result of multiple risk factors that interact with each other. Therefore, an estimate of the combined contribution of several individual risk factors to the total number of cases in the population must allow for the nature of the interaction of those factors among women who have breast cancer.
Estimates of PARs vary across studies. (A table summarizing estimates from several studies appears in Appendix D.) For example, alcohol ranges from 2 percent (Tseng et al., 1999) to 11 percent (Mezzetti et al., 1998), whereas hormone therapy has PAR estimates from approximately 4 percent in the United States in about 2001 (Clarke et al., 2006) to 27 percent in Norway in the 1990s (Bakken et al., 2004), and physical inactivity, from 6 percent in Canada in 2006 (Neutel and Morrison, 2010) to 20 percent in a combination of several European countries in 2002 (Friedenreich et al., 2010). The variation in PAR estimates arises from differences across studies in the prevalence of the risk factors in the study populations as well as in other characteristics of the study populations and in the design and quality of the studies. For example, if combination hormone therapy is widely used in a study population, its PAR would tend to higher than the PAR for hormone therapy in a study population with relatively limited use. PARs should at best be viewed as ballpark estimates of potential impact on breast cancer risk on a population level, under the assumption that the associations are causal.
Implications for Breast Cancer Reduction
As indicated above, the PAR estimates the percentage reduction in disease burden that can be achieved on a population level with a reduction in the prevalence of a risk factor. The estimated benefit on an absolute scale either for the population or for an individual may be small (Petracci et al., 2011). For a woman with a diagnosis of breast cancer who is of normal weight, the contribution of being overweight or obese to her breast cancer has to be zero. Alternatively, for a woman with breast cancer who is overweight or obese, the chances that her weight contributed to the development of her breast cancer might be higher than the contribution
of overweight and obesity to breast cancer cases in the population as a whole. An uncommon exposure will usually have a small PAR because the percentage of all cases that is attributable to the exposure will be small. However, for a woman who has that exposure, reducing or eliminating it could substantially lower her risk and be very important to her individually. That is, a rare, high-risk exposure may have little impact on population rates of cancer, but it may be a quite important determinant of an exposed woman’s personal risk.
Recent efforts have tried to further clarify risks for both individual women and populations by developing models that estimate the absolute risk of breast cancer from relative risks and estimates of attributable risk. In a study by Petracci et al. (2011), the authors used data on Italian women to develop a model to predict breast cancer risk, making use of both nonmodifiable risk factors and the modifiable risk factors of BMI, alcohol consumption, and physical activity. Data from a cohort study were used to assess the potential impact on absolute breast cancer risk of reducing exposures to the modifiable risk factors. The projected 20-year absolute risk of breast cancer for 65-year-old women, for example, ranged from 6.5 to 18.6 percent, depending on their risk profiles. If these women optimized their BMI, alcohol consumption, and physical activity, the estimated 20-year absolute risks would be reduced to 4.9 and 14.1 percent, respectively (Petracci et al., 2011). Presentation of the absolute risk reductions along with estimates of relative risk and the PAR reduction that could maximally be achieved may be a useful approach to both individual counseling and public health decision making (Schwartz et al., 2006; Akl et al., 2011; Helzlsouer, 2011). It illustrates the well-known concept that small changes at the individual level can have a large impact at the population level (Rose, 1992).
Many of the established risk factors for breast cancer—age, sex, age at menarche and menopause, age at first full-term pregnancy—offer little or no opportunity to intervene. For a limited set of other risk factors, evidence suggests that action can be taken in ways that that have the potential to reduce risk for breast cancer for many women: eliminating unnecessary medical radiation throughout life, avoiding use of postmenopausal hormone therapy, avoiding active and passive smoking, reducing alcohol consumption, increasing physical activity, and minimizing weight gain. Chemoprevention may be an appropriate choice for some women.
For the many chemicals that are manufactured or generated as by-products of other processes, the committee found little basis in the human evidence it examined to point to avoiding or eliminating exposure as a specific strategy for reducing breast cancer risk. Exceptions were benzene,
1,3-butadiene, ethylene oxide, for which certain measures to control occupational exposures are already be in place. However, these chemicals can also be encountered by the general public (although likely at much lower exposure levels) through exposure to facilities emissions, tobacco smoke, and gasoline vapors and vehicular exhaust (benzene and 1,3-butadiene). While for other compounds that were reviewed, such as BPA, animal and mechanistic evidence may indicate breast cancer hazard is biologically plausible, given sufficient dosing, information to assess the magnitude of risk in humans is lacking or inadequate in human studies, posing a substantial challenge for gauging the extent to which an individual’s actions may reduce risk.
Even when action appears possible, most approaches to risk reduction come with potentially complex trade-offs. These trade-offs may be social or economic (e.g., the potential influence of earlier age at first birth on a woman’s education or employment), or they may be health related (e.g., moderate alcohol consumption increases breast cancer risk, but it may reduce risk of heart disease; tamoxifen reduces risk for breast cancer but increases risk for stroke and endometrial cancer). It is also important to keep in mind that what the committee has outlined in this chapter are areas where the evidence indicates that action is likely to reduce risk in an average population. The actual change in risk for any individual woman who takes such actions might range from very small to moderate.
Chapter 7 outlines the committee’s recommendations for further research to strengthen the knowledge base on breast cancer and, hopefully, to point to more and better opportunities to reduce risk for this disease.
Ahn, J., A. Schatzkin, J. V. Lacey, Jr., D. Albanes, R. Ballard-Barbash, K. F. Adams, V. Kipnis, T. Mouw, et al. 2007. Adiposity, adult weight change, and postmenopausal breast cancer risk. Arch Intern Med 167(19):2091–2102.
Akl, E. A., A. D. Oxman, J. Herrin, G. E. Vist, I. Terrenato, F. Sperati, C. Costiniuk, D. Blank, et al. 2011. Using alternative statistical formats for presenting risks and risk reductions. Cochrane Database Syst Rev (3):CD006776.
Alliance for Radiation Safety in Pediatric Imaging. 2008. Image gently. http://www.pedrad.org/associations/5364/ig/ (accessed June 23, 2011).
Ashar, B. H., T. N. Rice, and S. D. Sisson. 2007. Physicians’ understanding of the regulation of dietary supplements. Arch Intern Med 167(9):966–969.
Bakken, K., E. Alsaker, A. E. Eggen, and E. Lund. 2004. Hormone replacement therapy and incidence of hormone-dependent cancers in the Norwegian Women and Cancer study. Int J Cancer 112(1):130–134.
Baumann, B. M., E. H. Chen, A. M. Mills, L. Glaspey, N. M. Thompson, M. K. Jones, and M. C. Farner. 2011. Patient perceptions of computed tomographic imaging and their understanding of radiation risk and exposure. Ann Emerg Med 58(1):1–7.
Beral, V., G. Reeves, D. Bull, and J. Green. 2011. Breast cancer risk in relation to the interval between menopause and starting hormone therapy. J Natl Cancer Inst 103(4):296–305.
Berrington de Gonzalez, A., M. Mahesh, K. P. Kim, M. Bhargavan, R. Lewis, F. Mettler, and C. Land. 2009. Projected cancer risks from computed tomographic scans performed in the United States in 2007. Arch Intern Med 169(22):2071–2077.
Bogdanich, W. 2011. West Virginia hospital overradiated brain scan patients, records show. New York Times, March 5.
Bogdanich, W., and K. Rebelo. 2010. A pinpoint beam strays invisibly, harming instead of healing. New York Times, December 28.
Boyd, N. F., L. J. Martin, M. Yaffe, and S. Minkin. 2009. Mammographic density. Breast Cancer Res 11(Suppl 3):S4.
Boyd, N. F., L. J. Martin, M. Bronskill, M. J. Yaffe, N. Duric, and S. Minkin. 2010. Breast tissue composition and susceptibility to breast cancer. J Natl Cancer Inst 102(16): 1224–1237.
Brinton, L. A., D. Richesson, M. F. Leitzmann, G. L. Gierach, A. Schatzkin, T. Mouw, A. R. Hollenbeck, and J. V. Lacey, Jr. 2008. Menopausal hormone therapy and breast cancer risk in the NIH–AARP Diet and Health Study Cohort. Cancer Epidemiol Biomarkers Prev 17(11):3150–3160.
Brody, J. G., K. B. Moysich, O. Humblet, K. R. Attfield, G. P. Beehler, and R. A. Rudel. 2007. Environmental pollutants and breast cancer: Epidemiologic studies. Cancer 109(12 Suppl):2667–2711.
CalEPA (California Environmental Protection Agency). 2005. Proposed identification of environmental tobacco smoke as a toxic air contaminant. http://www.arb.ca.gov/regact/ets2006/ets2006.htm (accessed November 9, 2011).
California Breast Cancer Research Program. 2007. Identifying gaps in breast cancer research: Addressing disparities and the roles of the physical and social environment. http://cbcrp.org/sri/reports/identifyingGaps/index.php (accessed October 25, 2011).
Cantor, K. P., C. M. Villanueva, D. T. Silverman, J. D. Figueroa, F. X. Real, M. Garcia-Closas, N. Malats, S. Chanock, et al. 2010. Polymorphisms in GSTT1, GSTZ1, and CYP2E1, disinfection by-products, and risk of bladder cancer in Spain. Environ Health Perspect 118(11):1545–1550.
CDC (Centers for Disease Control and Prevention). 2011. Alcohol and public health– frequently asked questions. http://www.cdc.gov/alcohol/faqs.htm#heavyDrinking (accessed November 9, 2011).
Chen, W. Y., G. A. Colditz, B. Rosner, S. E. Hankinson, D. J. Hunter, J. E. Manson, M. J. Stampfer, W. C. Willett, et al. 2002. Use of postmenopausal hormones, alcohol, and risk for invasive breast cancer. Ann Intern Med 137(10):798–804.
Chlebowski, R. T., L. H. Kuller, R. L. Prentice, M. L. Stefanick, J. E. Manson, M. Gass, A. K. Aragaki, J. K. Ockene, et al. 2009. Breast cancer after use of estrogen plus progestin in postmenopausal women. N Engl J Med 360(6):573–587.
Chuang, S. C., V. Gallo, D. Michaud, K. Overvad, A. Tjonneland, F. Clavel-Chapelon, I. Romieu, K. Straif, et al. 2011. Exposure to environmental tobacco smoke in childhood and incidence of cancer in adulthood in never smokers in the european prospective investigation into cancer and nutrition. Cancer Causes Control 22(3):487–494.
Cirigliano, M. 2007. Bioidentical hormone therapy: A review of the evidence. J Womens Health (Larchmt) 16(5):600–631.
Clarke, C. A., D. M. Purdie, and S. L. Glaser. 2006. Population attributable risk of breast cancer in white women associated with immediately modifiable risk factors. BMC Cancer 6:170.
Cohen, P. A. 2009. American roulette—contaminated dietary supplements. N Engl J Med 361(16):1523–1525.
Collaborative Group on Hormonal Factors in Breast Cancer. 1996. 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(9017):1713–1727.
Collaborative Group on Hormonal Factors in Breast Cancer. 1997. 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. Lancet 350(9084):1047–1059.
Collaborative Group on Hormonal Factors in Breast Cancer. 2002. Alcohol, tobacco and breast cancer—collaborative reanalysis of individual data from 53 epidemiological studies, including 58,515 women with breast cancer and 95,067 women without the disease. Br J Cancer 87(11):1234–1245.
Collishaw, N. C., N. F. Boyd, K. P. Cantor, S. K. Hammond, K. C. Johnson, J. Millar, A. B. Miller, M. Miller, J. R. Palmer, A. G. Salmon, and F. Turcotte. 2009. Canadian expert panel on tobacco smoke and breast cancer risk. OTRU Special Report Series. Toronto, Canada: Ontario Tobacco Research Unit. http://www.otru.org/pdf/special/expert_panel_tobacco_breast_cancer.pdf (accessed October 29, 2011).
Cummings, S. R., J. A. Tice, S. Bauer, W. S. Browner, J. Cuzick, E. Ziv, V. Vogel, J. Shepherd, et al. 2009. Prevention of breast cancer in postmenopausal women: Approaches to estimating and reducing risk. J Natl Cancer Inst 101(6):384–398.
Cuzick, J. 2008. Assessing risk for breast cancer. Breast Cancer Res 10(Suppl 4):S13.
Cuzick, J., A. DeCensi, B. Arun, P. H. Brown, M. Castiglione, B. Dunn, J. F. Forbes, A. Glaus, et al. 2011. Preventive therapy for breast cancer: A consensus statement. Lancet Oncol 12(5):496–503.
Davidson, N. E., and K. J. Helzlsouer. 2002. Good news about oral contraceptives. N Engl J Med 346(26):2078–2079.
DeRoo, L. A., P. Cummings, and B. A. Mueller. 2011. Smoking before the first pregnancy and the risk of breast cancer: A meta-analysis. Am J Epidemiol 174(4):390-402.
Don, S. 2011. Pediatric digital radiography summit overview: State of confusion. Pediatr Radiol 41(5):567–572.
Eliassen, A. H., G. A. Colditz, B. Rosner, W. C. Willett, and S. E. Hankinson. 2006. Adult weight change and risk of postmenopausal breast cancer. JAMA 296(2):193–201.
Endocrine Society. 2006. Position statement: Bioidentical hormones. http://www.endo-society.org/advocacy/policy/upload/BH_Position_Statement_final_10_25_06_w_Header.pdf (accessed November 8, 2011).
EPA (Environmental Protection Agency). 2005. Guidelines for carcinogen risk assessment. EPA/630/P-03/001F. http://www.epa.gov/raf/publications/pdfs/CANCER_GUIDELINES_FINAL_3-25-05.PDF (accessed November 17, 2011).
EPA. 2011. Integrated risk information system (IRIS). http://www.epa.gov/ncea/iris/index.html (accessed October 28, 2011).
Fazel, R., H. M. Krumholz, Y. Wang, J. S. Ross, J. Chen, H. H. Ting, N. D. Shah, K. Nasir, et al. 2009. Exposure to low-dose ionizing radiation from medical imaging procedures. N Engl J Med 361(9):849–857.
FDA (Food and Drug Administration). 2005. FDA authority over cosmetics. http://www.fda.gov/Cosmetics/GuidanceComplianceRegulatoryInformation/ucm074162.htm (accessed November 17, 2011).
FDA. 2009. Consumer information: Overview of dietary supplements. http://www.fda.gov/Food/DietarySupplements/ConsumerInformation/ucm110417.htm (accessed November 9, 2011).
FDA. 2010. Initiative to reduce unnecessary radiation exposure from medical imaging. http://www.fda.gov/radiation-emittingproducts/radiationsafety/radiationdosereduction/ucm199904.htm (accessed November 17, 2011).
Fennell, T. R., J. P. MacNeela, R. W. Morris, M. Watson, C. L. Thompson, and D. A. Bell. 2000. Hemoglobin adducts from acrylonitrile and ethylene oxide in cigarette smokers: Effects of glutathione S-transferase T1-null and M1-null genotypes. Cancer Epidemiol Biomarkers Prev 9(7):705–712.
Fisher, B., J. P. Costantino, D. L. Wickerham, R. S. Cecchini, W. M. Cronin, A. Robidoux, T. B. Bevers, M. T. Kavanah, et al. 2005. Tamoxifen for the prevention of breast cancer: Current status of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 97(22):1652–1662.
Flegal, K. M., M. D. Carroll, C. L. Ogden, and L. R. Curtin. 2010. Prevalence and trends in obesity among US adults, 1999–2008. JAMA 303(3):235–241.
Freedman, A. N., B. Yu, M. H. Gail, J. P. Costantino, B. I. Graubard, V. G. Vogel, G. L. Anderson, and W. McCaskill-Stevens. 2011. Benefit/risk assessment for breast cancer chemoprevention with raloxifene or tamoxifen for women age 50 years or older. J Clin Oncol 29(17):2327–2333.
Friedenreich, C. M. 2010. The role of physical activity in breast cancer etiology. Semin Oncol 37(3):297–302.
Friedenreich, C. M., C. G. Woolcott, A. McTiernan, R. Ballard-Barbash, R. F. Brant, F. Z. Stanczyk, T. Terry, N. F. Boyd, et al. 2010. Alberta physical activity and breast cancer prevention trial: Sex hormone changes in a year-long exercise intervention among postmenopausal women. J Clin Oncol 28(9):1458–1466.
Friedenreich, C. M., C. G. Woolcott, A. McTiernan, T. Terry, R. Brant, R. Ballard-Barbash, M. L. Irwin, C. A. Jones, et al. 2011. Adiposity changes after a 1-year aerobic exercise intervention among postmenopausal women: A randomized controlled trial. Int J Obes (Lond) 35(3):427-435.
Gail, M. H., L. A. Brinton, D. P. Byar, D. K. Corle, S. B. Green, C. Schairer, and J. J. Mulvihill. 1989. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst 81(24):1879–1886.
GAO (Government Accountability Office). 2005. Options exist to improve EPA’s ability to assess health risks and manage its Chemical Review Program. GAO-05-458. Washington, DC: GAO. http://www.gao.gov/new.items/d05458.pdf (accessed November 9, 2011).
GAO. 2006. Actions are needed to improve the effectiveness of EPA’s Chemical Review Program. GAO-06-1032T. Washington, DC: GAO. http://www.gao.gov/new.items/d061032t.pdf (accessed November 9, 2011).
GAO. 2007. Comparison of U.S. and recently enacted European Union approaches to protect against the risks of toxic chemicals. GAO-07-825. Washington, DC: GAO. http://www.gao.gov/new.items/d07825.pdf (accessed November 9, 2011).
GAO. 2009a. Options for enhancing the effectiveness of the Toxic Substances Control Act. GAO-09-428T. Washington, DC: GAO. http://www.gao.gov/new.items/d09428t.pdf (accessed November 9, 2011).
GAO. 2009b. Biomonitoring: EPA needs to coordinate its research strategy and clarify its authority to obtain biomonitoring data. GAO-09-353. Washington, DC: GAO. http://www.gao.gov/new.items/d09353.pdf (accessed October 29, 2011).
Gapstur, S. M., J. D. Potter, T. A. Sellers, and A. R. Folsom. 1992. Increased risk of breast cancer with alcohol consumption in postmenopausal women. Am J Epidemiol 136(10):1221–1231.
Gibbons, J., and S. Laha. 1999. Water purification systems: A comparative analysis based on the occurrence of disinfection by-products. Environ Pollut 106(3):425–428.
Goss, P. E., J. N. Ingle, J. E. Ales-Martinez, A. M. Cheung, R. T. Chlebowski, J. Wactawski-Wende, A. McTiernan, J. Robbins, et al. 2011. Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med 364(25):2381–2391.
Gray, J. 2010. State of the evidence: The connection between breast cancer and the environment, 6th ed. San Francisco, CA: Breast Cancer Fund.
Gronbaek, M. 2009. The positive and negative health effects of alcohol—and the public health implications. J Intern Med 265(4):407–420.
Gunzerath, L., V. Faden, S. Zakhari, and K. Warren. 2004. National Institute on Alcohol Abuse and Alcoholism report on moderate drinking. Alcohol Clin Exp Res 28(6): 829–847.
Ha, M., K. Mabuchi, A. J. Sigurdson, D. M. Freedman, M. S. Linet, M. M. Doody, and M. Hauptmann. 2007. Smoking cigarettes before first childbirth and risk of breast cancer. Am J Epidemiol 166(1):55–61.
Hammond, E. C., I. J. Selikoff, and H. Seidman. 1979. Asbestos exposure, cigarette smoking and death rates. Ann N Y Acad Sci 330:473–490.
Hankinson, S. E. 2005–2006. Endogenous hormones and risk of breast cancer in postmenopausal women. Breast Dis 24:3–15.
Harvie, M., A. Howell, R. A. Vierkant, N. Kumar, J. R. Cerhan, L. E. Kelemen, A. R. Folsom, and T. A. Sellers. 2005. Association of gain and loss of weight before and after menopause with risk of postmenopausal breast cancer in the Iowa Women’s Health Study. Cancer Epidemiol Biomarkers Prev 14(3):656–661.
Helzlsouer, K. J. 2011. The numbers game: The risky business of projecting risk. J Natl Cancer Inst 103(13):992–993.
HHS (Department of Health and Human Services). 2006. The health consequences of involuntary exposure to tobacco smoke: A report of the Surgeon General. Atlanta, GA: Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
HHS. 2008. 2008 Physical activity guidelines for Americans. http://www.health.gov/paguidelines/guidelines/default.aspx (accessed August 9, 2011).
Horn-Ross, P. L., A. J. Canchola, D. W. West, S. L. Stewart, L. Bernstein, D. Deapen, R. Pinder, R. K. Ross, et al. 2004. Patterns of alcohol consumption and breast cancer risk in the California Teachers Study cohort. Cancer Epidemiol Biomarkers Prev 13(3):405–411.
Huang, Z., S. E. Hankinson, G. A. Colditz, M. J. Stampfer, D. J. Hunter, J. E. Manson, C. H. Hennekens, B. Rosner, et al. 1997. Dual effects of weight and weight gain on breast cancer risk. JAMA 278(17):1407–1411.
Hunter, D. J., G. A. Colditz, S. E. Hankinson, S. Malspeis, D. Spiegelman, W. Chen, M. J. Stampfer, and W. C. Willett. 2010. Oral contraceptive use and breast cancer: A prospective study of young women. Cancer Epidemiol Biomarkers Prev 19(10):2496–2502.
IARC (International Agency for Research on Cancer). 2010. Alcohol consumption and ethyl carbamate. IARC monographs on the evaluation of carcinogenic risks to humans. Vol. 96. Lyon, France: IARC.
IARC. 2011. IARC monographs on the evaluation of carcinogenic risks to humans. http://monographs.iarc.fr/ (accessed October 28, 2011).
Key, T., P. Appleby, I. Barnes, and G. Reeves. 2002. Endogenous sex hormones and breast cancer in postmenopausal women: Reanalysis of nine prospective studies. J Natl Cancer Inst 94(8):606–616.
Klatsky, A. L. 2009. Alcohol and cardiovascular diseases. Expert Rev Cardiovasc Ther 7(5): 499–506.
La Vecchia, C. 2001. Epidemiology of ovarian cancer: A summary review. Eur J Cancer Prev 10(2):125–129.
LaCroix, A. Z., R. T. Chlebowski, J. E. Manson, A. K. Aragaki, K. C. Johnson, L. Martin, K. L. Margolis, M. L. Stefanick, et al. 2011. Health outcomes after stopping conjugated equine estrogens among postmenopausal women with prior hysterectomy: A randomized controlled trial. JAMA 305(13):1305–1314.
Lee, C. I., A. H. Haims, E. P. Monico, J. A. Brink, and H. P. Forman. 2004. Diagnostic CT scans: Assessment of patient, physician, and radiologist awareness of radiation dose and possible risks. Radiology 231(2):393–398.
Luo, J., K. L. Margolis, J. Wactawski-Wende, K. Horn, C. Messina, M. L. Stefanick, H. A. Tindle, E. Tong, et al. 2011. Association of active and passive smoking with risk of breast cancer among postmenopausal women: A prospective cohort study. BMJ 342:d1016.
Marchbanks, P. A., J. A. McDonald, H. G. Wilson, S. G. Folger, M. G. Mandel, J. R. Daling, L. Bernstein, K. E. Malone, et al. 2002. Oral contraceptives and the risk of breast cancer. N Engl J Med 346(26):2025–2032.
Maskarinec, G., L. Meng, and L. N. Kolonel. 1998. Alcohol intake, body weight, and mortality in a multiethnic prospective cohort. Epidemiology 9(6):654–661.
McTiernan, A., S. S. Tworoger, K. B. Rajan, Y. Yasui, B. Sorenson, C. M. Ulrich, J. Chubak, F. Z. Stanczyk, et al. 2004a. Effect of exercise on serum androgens in postmenopausal women: A 12-month randomized clinical trial. Cancer Epidemiol Biomarkers Prev 13(7): 1099–1105.
McTiernan, A., S. S. Tworoger, C. M. Ulrich, Y. Yasui, M. L. Irwin, K. B. Rajan, B. Sorensen, R. E. Rudolph, et al. 2004b. Effect of exercise on serum estrogens in postmenopausal women: A 12-month randomized clinical trial. Cancer Res 64(8):2923–2928.
Mezzetti, M., C. La Vecchia, A. Decarli, P. Boyle, R. Talamini, and S. Franceschi. 1998. Population attributable risk for breast cancer: Diet, nutrition, and physical exercise. J Natl Cancer Inst 90(5):389–394.
Million Women Study Collaborators. 2003. Breast cancer and hormone-replacement therapy in the Million Women Study. Lancet 362(9382):419–427.
Monninkhof, E. M., S. G. Elias, F. A. Vlems, I. van der Tweel, A. J. Schuit, D. W. Voskuil, and F. E. van Leeuwen. 2007. Physical activity and breast cancer: A systematic review. Epidemiology 18(1):137–157.
Monninkhof, E. M., M. J. Velthuis, P. H. Peeters, J. W. Twisk, and A. J. Schuit. 2009. Effect of exercise on postmenopausal sex hormone levels and role of body fat: A randomized controlled trial. J Clin Oncol 27(27):4492–4499.
Moreno, M. A. 2011. Advice for patients: Decreasing unnecessary radiation exposure for children. Arch Pediatr Adolesc Med 165(5):480.
NCI (National Cancer Institute). 2010. SEER cancer statistics review, 1975–2007. Edited by S. F. Altekruse, C. L. Kosary, M. Krapcho, N. Neyman, R. Aminou, W. Waldron, J. Ruhl, N. Howlader, Z. Tatalovich, H. Cho, A. Mariotto, M. P. Eisner, D. R. Lewis, K. Cronin, H. S. Chen, E. J. Feuer, D. G. Stinchcomb, and B. K. Edwards. Bethesda, MD:
NCI. http://seer.cancer.gov/csr/1975_2007/ (accessed January 6, 2011).
NCI. 2011a. Breast cancer risk assessment tool. http://www.cancer.gov/bcrisktool/ (accessed December 21, 2011).
NCI. 2011b. Breast cancer risk assessment tool: About the tool. http://www.cancer.gov/bcrisktool/about-tool.aspx (accessed December 21, 2011).
Nelson, H. D., R. Fu, J. C. Griffin, P. Nygren, M. E. Smith, and L. Humphrey. 2009. Systematic review: Comparative effectiveness of medications to reduce risk for primary breast cancer. Ann Intern Med 151(10):703–715, W-226–W-735.
Neutel, C. I., and H. Morrison. 2010. Could recent decreases in breast cancer incidence really be due to lower HRT use? Trends in attributable risk for modifiable breast cancer risk factors in Canadian women. Can J Public Health 101(5):405–409.
NTP (National Toxicology Program). 2007a. Toxicology and carcinogenesis studies of dibromoacetic acid (CAS No. 631-64-1) in F344/N rats and B6C3F1 mice (drinking water studies). NTP TR 537. Research Triangle Park, NC: NTP. http://ntp.niehs.nih.gov/files/537_FINAL_web.pdf (accessed December 21, 2011).
NTP. 2007b. Toxicology studies of dichloroacetic acid (CAS No. 79-43-6) in genetically modified mice. NTP GMM 11. Research Triangle Park, NC: NTP. http://www.epa.gov/iris/toxreviews/0654tr.pdf (accessed November 8, 2011).
NTP. 2011. 12th report on carcinogens (RoC). http://ntp.niehs.nih.gov/index.cfm?objectid=03C9AF75-E1BF-FF40-DBA9EC0928DF8B15 (accessed November 9, 2011).
Palmer, J. R., L. L. Adams-Campbell, D. A. Boggs, L. A. Wise, and L. Rosenberg. 2007. A prospective study of body size and breast cancer in black women. Cancer Epidemiol Biomarkers Prev 16(9):1795–1802.
Petracci, E., A. Decarli, C. Schairer, R. M. Pfeiffer, D. Pee, G. Masala, D. Palli, and M. H. Gail. 2011. Risk factor modification and projections of absolute breast cancer risk. J Natl Cancer Inst 103(13):1037–1048.
Physical Activity Guidelines Advisory Committee. 2008. Physical activity guidelines advisory committee report, 2008. Washington, DC: HHS. http://www.health.gov/paguidelines/committeereport.aspx (accessed August 9, 2011).
Preston, D. L., E. Ron, S. Tokuoka, S. Funamoto, N. Nishi, M. Soda, K. Mabuchi, and K. Kodama. 2007. Solid cancer incidence in atomic bomb survivors: 1958–1998. Radiat Res 168(1):1–64.
Ravdin, P. M., K. A. Cronin, N. Howlader, C. D. Berg, R. T. Chlebowski, E. J. Feuer, B. K. Edwards, and D. A. Berry. 2007. The decrease in breast-cancer incidence in 2003 in the United States. N Engl J Med 356(16):1670–1674.
Reeves, G. K., V. Beral, J. Green, T. Gathani, and D. Bull. 2006. Hormonal therapy for menopause and breast-cancer risk by histological type: A cohort study and meta-analysis. Lancet Oncol 7(11):910–918.
Reynolds, P., S. Hurley, D. E. Goldberg, H. Anton-Culver, L. Bernstein, D. Deapen, P. L. Horn-Ross, D. Peel, et al. 2004. Active smoking, household passive smoking, and breast cancer: Evidence from the California Teachers Study. J Natl Cancer Inst 96(1):29–37.
Reynolds, P., D. Goldberg, S. Hurley, D. O. Nelson, J. Largent, K. D. Henderson, and L. Bernstein. 2009. Passive smoking and risk of breast cancer in the California Teachers Study. Cancer Epidemiol Biomarkers Prev 18(12):3389–3398.
Richardson, S. D., M. J. Plewa, E. D. Wagner, R. Schoeny, and D. M. Demarini. 2007. Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: A review and roadmap for research. Mutat Res 636(1–3):178–242.
Ronksley, P. E., S. E. Brien, B. J. Turner, K. J. Mukamal, and W. A. Ghali. 2011. Association of alcohol consumption with selected cardiovascular disease outcomes: A systematic review and meta-analysis. BMJ 342:d671.
Rose, G. 1992. Strategies of prevention: The individual and the population. In Coronary heart disease epidemiology: From aetiology to public health. Edited by M. Marmot and P. Elliott. Oxford, UK: Oxford University Press. Pp. 311–324.
Rothman, K. J., and S. Greenland, eds. 1998. Modern epidemiology, 2nd ed. Philadelphia, PA: Lippincott–Raven.
Rudel, R. A., K. R. Attfield, J. N. Schifano, and J. G. Brody. 2007. Chemicals causing mammary gland tumors in animals signal new directions for epidemiology, chemicals testing, and risk assessment for breast cancer prevention. Cancer 109(12 Suppl):2635–2666.
Rudel, R. A., J. M. Gray, C. L. Engel, T. W. Rawsthorne, R. E. Dodson, J. M. Ackerman, J. Rizzo, J. L. Nudelman, et al. 2011. Food packaging and bisphenol A and bis(2-ethyhexyl) phthalate exposure: Findings from a dietary intervention. Environ Health Perspect 119(7):914–920.
Santen, R. J., D. C. Allred, S. P. Ardoin, D. F. Archer, N. Boyd, G. D. Braunstein, H. G. Burger, G. A. Colditz, et al. 2010. Postmenopausal hormone therapy: An Endocrine Society scientific statement. J Clin Endocrinol Metab 95(7 Suppl 1):s1–s66.
Schoeny, R. 2010. Disinfection by-products: A question of balance. Environ Health Perspect 118(11):A466–A467.
Schwartz, L. M., S. Woloshin, E. L. Dvorin, and H. G. Welch. 2006. Ratio measures in leading medical journals: Structured review of accessibility of underlying absolute risks. BMJ 333(7581):1248.
Secretan, B., K. Straif, R. Baan, Y. Grosse, F. El Ghissassi, V. Bouvard, L. Benbrahim-Tallaa, N. Guha, et al. 2009. A review of human carcinogens—Part E: Tobacco, areca nut, alcohol, coal smoke, and salted fish. Lancet Oncol 10(11):1033–1034.
Shuryak, I., R. K. Sachs, and D. J. Brenner. 2010. Cancer risks after radiation exposure in middle age. J Natl Cancer Inst 102(21):1628–1636.
Smith-Bindman, R. 2010. Is computed tomography safe? N Engl J Med 363(1):1–4.
Smith-Bindman, R. 2011. Ionizing radiation exposure to the U.S. population, with a focus on radiation from medical imaging. Paper commissioned by the Committee on Breast Cancer and the Environment (see Appendix F; available at http://www.nap.edu/catalog.php?record_id=13263).
Smith-Warner, S. A., D. Spiegelman, S. S. Yaun, P. A. van den Brandt, A. R. Folsom, R. A. Goldbohm, S. Graham, L. Holmberg, et al. 1998. Alcohol and breast cancer in women: A pooled analysis of cohort studies. JAMA 279(7):535–540.
Sovak, M., A. L. Seligson, M. Konas, M. Hajduch, M. Dolezal, M. Machala, and R. Nagourney. 2002. Herbal composition PC-SPES for management of prostate cancer: Identification of active principles. J Natl Cancer Inst 94(17):1275–1281.
Taylor, H., and R. Leftman, eds. 2002. Widespread ignorance of regulation and labeling of vitamins, minerals, and food supplements. Health Care News 2(23):1–5. http://www.harrisinteractive.com/news/newsletters/healthnews/HI_HealthCareNews2002Vol2_Iss23.pdf (accessed August 9, 2011).
Teras, L. R., M. Goodman, A. V. Patel, W. R. Diver, W. D. Flanders, and H. S. Feigelson. 2011. Weight loss and postmenopausal breast cancer in a prospective cohort of overweight and obese U.S. women. Cancer Causes Control 22(4):573–579.
Tseng, M., C. R. Weinberg, D. M. Umbach, and M. P. Longnecker. 1999. Calculation of population attributable risk for alcohol and breast cancer (United States). Cancer Causes Control 10(2):119–123.
Tyrer, J., S. W. Duffy, and J. Cuzick. 2004. A breast cancer prediction model incorporating familial and personal risk factors. Stat Med 23:1111–1130.
USPSTF (U.S. Preventive Services Task Force). 2002. Chemoprevention of breast cancer: Recommendations and rationale. http://www.uspreventiveservicestaskforce.org/uspstf/uspsbrpv.htm (accessed August 9, 2011).
USPSTF. 2005. Hormone therapy for the prevention of chronic conditions in postmenopausal women: Recommendation statement. http://www.uspreventiveservicestaskforce.org/uspstf/uspspmho.htm (accessed August 9, 2011).
Vogel, V. G., J. P. Costantino, D. L. Wickerham, W. M. Cronin, R. S. Cecchini, J. N. Atkins, T. B. Bevers, L. Fehrenbacher, et al. 2010. Update of the National Surgical Adjuvant Breast and Bowel Project Study of Tamoxifen and Raloxifene (STAR) P-2 Trial: Preventing breast cancer. Cancer Prev Res (Phila) 3(6):696–706.
WCRF/AICR (World Cancer Research Fund/American Institute for Cancer Research). 2007. Food, nutrition, physical activity, and the prevention of cancer: A global perspective. Washington DC: AICR.
WCRF/AICR. 2008. The associations between food, nutrition and physical activity and the risk of breast cancer. WCRF/AICR systematic literature review continuous update report. http://dietandcancerreport.org/downloads/cu/cu_breast_cancer_report_2008.pdf (accessed October 29, 2011).
WCRF/AICR. 2010. Food, nutrition and physical activity and the prevention of breast cancer. WCRF/AICR continuous update report summary. http://dietandcancerreport.org/downloads/cu/cu_breast_cancer_summary_2008.pdf (accessed October 29, 2011).
White, J. 2002. PC-SPES—a lesson for future dietary supplement research. J Natl Cancer Inst 94(17):1261–1263.
Winzer, B. M., D. C. Whiteman, M. M. Reeves, and J. D. Paratz. 2011. Physical activity and cancer prevention: A systematic review of clinical trials. Cancer Causes Control 22(6):811–826.
Writing Group for the Women’s Health Initiative Investigators. 2002. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: Principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288(3):321–333.
Xue, F., W. C. Willett, B. A. Rosner, S. E. Hankinson, and K. B. Michels. 2011. Cigarette smoking and the incidence of breast cancer. Arch Intern Med 171(2):125–133.
Yang, C. Z., S. I. Yaniger, V. C. Jordan, D. J. Klein, and G. D. Bittner. 2011. Most plastic products release estrogenic chemicals: A potential health problem that can be solved. Environ Health Perspect 119(7):989–996.
This page intentionally left blank.