Jump to ContentJump to Main Navigation
Cancer Epidemiology and Prevention$

Michael Thun, Martha S. Linet, James R. Cerhan, Christopher A. Haiman, and David Schottenfeld

Print publication date: 2017

Print ISBN-13: 9780190238667

Published to Oxford Scholarship Online: December 2017

DOI: 10.1093/oso/9780190238667.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 19 October 2020

Breast Cancer

Breast Cancer

(p.861) 45 Breast Cancer
Cancer Epidemiology and Prevention

Louise A. Brinton

Mia M. Gaudet

Gretchen L. Gierach

Oxford University Press

Abstract and Keywords

Breast cancer is the most frequently diagnosed cancer in women worldwide, with annual estimates of 1.7 million newly diagnosed cases and 522,000 deaths. Although more breast cancers are diagnosed in economically developed than in developing countries, the reverse is true for mortality, reflecting limited screening and less effective treatments in the latter. Breast cancer incidence has been on the rise in the United States for many years, but in recent years this is restricted to certain subgroups, while internationally there have been continued generalized increases, likely reflecting adoption of more Westernized lifestyles. Breast cancer is widely recognized as being hormonally influenced, with most of the established risk factors believed to reflect the influence of cumulative exposure of the breast to stimulatory effects of ovarian hormones—leading to increased cellular proliferation, which in turn can result in genetic errors during cell division.

Keywords:   breast cancer, hormones, ovary, genetic, cell division


Breast cancer is the most frequently diagnosed cancer in women worldwide, with annual estimates of 1.7 million newly diagnosed cases and 522,000 deaths. Although more breast cancers are diagnosed in economically developed than developing countries, the reverse is true for mortality, reflecting limited screening and less effective treatments in such areas. Breast cancer incidence has been on the rise in the United States for many years. In recent years this is restricted to certain subgroups, while abroad there have been continued generalized increases, likely reflecting the adoption of more Westernized lifestyles. Breast cancer is widely recognized as being hormonally influenced, with most of the established risk factors believed to reflect the influence of cumulative exposure of the breast to the stimulatory effects of ovarian hormones—leading to increased cellular proliferation, which in turn can result in genetic errors during cell division. To better understand the role of endogenous hormones as etiologic factors, studies have taken advantage of recent laboratory advances, including efforts to measure hormone metabolites. Although epidemiologically breast cancer has been one of the most intensively studied tumors, only about half of the disease occurrence is explained by well-established risk factors. Most of the identified risk factors are not readily modifiable, leading to a need for additional research to better understand etiologic processes and to identify factors for which interventions might be feasible. During the past decade, there has been increasing appreciation that breast cancer is not one disease, but rather a collection of divergent diseases that have different biologic, clinical, and prognostic characteristics. Attention to such heterogeneity is essential for accurate risk prediction, as well as for advancing our understanding of biological processes and developing effective prevention modalities.

Normal Breast Development

The breast is composed of adipose and glandular tissue and the surrounding stroma. The glandular tissue is a network of ducts that branch from the nipple to the terminal duct lobular units (TDLUs) (Figure 45–1) (Ali and Coombes, 2002). The TDLU contains a small segment of terminal duct and a cluster of acini (or alveoli), which are the secretory units of the breast. The mammary epithelial cells lining the glandular tissue are composed of two main cellular lineages: luminal cells that surround a central lumen and myoepithelial cells that are located in a basal position adjacent to the basement membrane. The collagen-rich stroma includes the surrounding connective tissue, blood vessels, and lymph vessels (Russo and Russo, 2004).

Breast Cancer

Figure 45–1. Anatomy of the human mammary gland. (a) Anatomy of the human mammary gland. Each mammary gland contains 15–20 lobes, each lobe containing a series of branched ducts that drain into the nipple. (b) Each duct is lined with a layer of epithelial cells, responsible for milk production. These are surrounded by an outer layer of myoepithelial cells with contractile properties. The glandular ducts are embedded in fibroblast stroma. (c) This structure breaks down in breast cancer, resulting in an epithelial cell mass. (b) and (c) are immunostained using antibodies to the estrogen receptor showing that only a small proportion of epithelial cells are ER positive in the normal breast.

Source: Ali and Coombes, 2002.

The mammary epithelium undergoes remodeling during distinct developmental stages (Russo and Russo, 2004). A female newborn has primitive terminal end buds that form the network of primary and secondary branches, which develop during puberty under the influence of ovarian hormones. In the adult gland, tertiary branching and small alveolar bud formation occur during each menstrual cycle. Before the first pregnancy, the breast epithelium is thought to be particularly susceptible to carcinogens because the TDLUs are not fully differentiated. The breast during pregnancy undergoes extensive expansion of the TDLUs; late in pregnancy, the epithelial cells differentiate to allow for lactation. During weaning, the TDLUs revert back to an earlier pre-pregnancy state or completely disappear in a process called post-lactational involution. A similar process of involution also occurs as part of aging (Figueroa et al., 2014). Age-related involution appears to start around the third decade of age and accelerates during the menopausal transition (Russo and Russo, 2004).

Tumor Classification

Breast cancer is a heterogeneous disease at the morphological and molecular levels. The fourth edition of the World Health Organization (WHO) Classification of Tumors of the Breast defines 21 distinct histological types based on cell morphology, growth, and architecture patterns (World Health Organization, 2012). Approximately 95% of invasive breast cancers diagnosed in the United States are adenocarcinomas. They originate in the epithelial tissue of the TDLUs and infiltrate the surrounding stroma. The majority of adenocarcinomas (75%–80%) are classified as “invasive carcinoma of no special type” (NST) because they do not display sufficient histopathologic characteristics to warrant classification into the “special” subtypes of breast cancer. NST tumors prior to 2012 were labeled as “ductal carcinomas, not otherwise specified” (World Health Organization, 2003); however, the reference to “ductal” has been dropped because the term suggests that these tumors, as opposed to the other subtypes, specifically arise from the TDLUs. The “special” breast cancer subtypes are defined by their distinct morphology and clinical behavior. Lobular cancer, which comprises the largest special group and accounts for 5%–15% of all invasive cancers, usually displays a distinctive pattern of single-file cells that are not disruptive to the normal tissue architecture. Less frequently diagnosed histological subtypes include medullary (< 2% of diagnoses in the US), mucinous (1%–4%), and tubular (< 2%) subtypes, among others (each < 1% of diagnoses) (World Health Organization, 2012).

Histopathologic subtypes can be further classified by molecular characteristics. As the NST is the largest and most heterogeneous group of breast cancers (and these tumors have been the basis of most molecular studies), molecular characteristics are useful to discriminate unique clinical and biological subgroups, whereas the special histopathology tumor subtypes, with the exception of lobular cancers, tend to have more homogeneous molecular characteristics (Weigelt et al., 2010). Since the 1970s, breast tumors have been classified at the molecular level using immunohistochemical (IHC) stains for estrogen receptor (ER) and progesterone receptor (PR) status. Tumors that express ER (ER+) or PR (PR+), collectively referred to as hormone receptor positive (HR+) tumors, are likely to respond to endocrine therapy, and have been shown to have different risk factor profiles than ER-negative (ER–) and PR-negative (PR–) tumors (as discussed in more detail throughout the chapter). In the 1990s, human epidermal growth factor receptor-2 (HER2) status also began to be used to identify a subgroup of breast tumors with poor prognosis. The monoclonal antibody trastuzumab was developed as an adjuvant therapy to target the amplification and/or overexpression of the HER2 gene (also known as ERBB2).

Global gene expression profiling of tumors has revealed further heterogeneity at the molecular level. At least four “intrinsic” subtypes (luminal A, luminal B, HER2-enriched, and basal-like tumors) have been identified. With some important exceptions, the intrinsic subtypes recapitulate the historical subtypes defined by IHC staining for (p.862) ER, PR, and HER2 (Table 45–1). Although the subtypes differ in clinical presentation, response to therapy, and prognosis, the biological and etiological inferences are still an active area of research (Norum et al., 2014). In the United States, the luminal A subtype comprises the largest proportion of tumors (73%) (Howlader et al., 2014). The proportion of molecular subtypes differs in other countries due to differences in screening and distributions of age and other breast cancer risk factors (Carvalho et al., 2014). Luminal A tumors are usually low grade and are characterized at the molecular level by strong expression of the genes encoding ER (ESR1), PR (PGR), and other genes regulated by ER; by overexpression of cyclin D; by low expression of proliferation-associated genes; and by lack of expression of HER2. They have a relatively low mutation rate, but phosphatidyl inositol-4,5-bisphosphate 3-kinase catalytic subunit-α‎ (PIK3CA), GATA3, and mitogen-activated protein kinase-1 (MAPK1) are commonly mutated (Norum et al., 2014).

Table 45–1. Clinical and Molecular Features and Immunohistochemical Definitions for the Common Intrinsic Subtypes of Breast Cancer

Intrinsic Subtype


Gene Expression

Immuno- Histochemical Definition

Somatic Mutations

Chromosomal Aberrations

Histological Special Type




Expression of Other Genes

Luminal A



Some, high


Luminal epithelial genes, cyclin D1

ER+ and/or PR+, HER2–, Ki-67 low


Diploid with whole arm aberrations

Lobular, tubular, mucinous, neuroendocrine

Luminal B



Low or none

Some, overexpression


ER+ and/or PR+, Ki-67 high

TP53 and PIK3CA, RB1, MAPK

Whole arm aberrations and complex rearrangements

Lobular, micropapillary






Basal cytokeratins, EGFR

ER–, PR–, HER2– (triple negative)



Secretory, adenoid cystic, medullary, metaplastic, acinic cell

HER2 enriched





Genes on 17q22

ER–, HER2+ (amplified or overexpressed)


Focal high-level amplifications

Lobular, apocrine, micropapillary

Abbreviations: EGFR = epidermal growth factor receptor; ER = estrogen receptor; GATA3 = GATA binding protein 3; HER2 = human epidermal growth factor receptor; MAPK = mitogen-activated protein kinase; PIK3CA = phosphatidylinositol-4,5-bisphospate 3-kinase catalytic subunit α‎; PR = progesterone receptor; TP53 = tumor protein p53.

Similar to luminal A tumors, luminal B cancers express proliferation genes as well as ER and PR, albeit at much lower levels. They also are distinguished from luminal A tumors by overexpression of HER2, higher rates of proliferation, and higher grades at diagnosis (Ades et al., 2014). They are more often aneuploidic than luminal A tumors, and display mutations in TP53 and PIK3CA genes (Norum et al., 2014). The proportion of luminal B tumors is about 10% in the United States (Howlader et al., 2014).

Luminal A and basal-like subtypes are the most contrasted groups at every level, including etiology, clinical presentation, response to treatment, and prognosis. Basal-like tumors present with high grades, necrosis, prominent lymphocytic infiltrates, and pushing borders. On the molecular level, basal-like tumors share more similarity with tumors arising in the basal layer of the epidermis, such as squamous carcinomas of the lung or head and neck as well as epithelial ovarian tumors. They express cytokeratins, epidermal growth factor receptor (EGFR), and other genes commonly expressed in basal/myoepithelial cells, and do not express ER, PR, or HER2 (collectively referred to as “triple negative”). They are frequently aneuploidic with complex genomic rearrangements and have somatic mutations in TP53. BRCA1 mutation carriers are more likely to be diagnosed with basal-like tumors than other subtypes, suggesting a strong role of double-stranded DNA repair mechanisms in their development (Norum et al., 2014). In the United States, the proportion of basal-like cancer diagnosed is 12% (Howlader et al., 2014).

The proportion of HER2-enriched tumors is nearly 5% in the United States (Howlader et al., 2014). The HER2-enriched tumors are characterized by overexpression of HER2 and other genes in the (p.863) same chromosomal region. They are high grade at diagnosis and lack expression of ER and PR (Norum et al., 2014).

Although lobular cancers encompass all intrinsic molecular subtypes, luminal A tumors predominate (Dieci et al., 2014). The distinct histopathology of lobular cancer is the result of the loss of cellular membrane expression of E-cadherin (encoded by CDH1), which causes the dysregulation of cell-cell adhesion properties (McCart Reed et al., 2015).

The rare histological subtypes are more homogeneous at the transcriptomic level than invasive carcinoma of NST, and each subtype clusters into only one or two molecular subtypes (Weigelt et al., 2010). Tubular and mucinous carcinomas are typically characterized as luminal tumors. Adenoid cystic, medullary, and metaplastic carcinomas display the basal-like phenotype. Metaplastic tumors are also found in the recently characterized claudin-low subtype. Apocrine carcinomas cluster in the HER2-enriched and molecular apocrine subtypes. Micropapillary carcinomas display characteristics of luminal and HER2-enriched tumors (Weigelt et al., 2010).

While these tumor subgroups have been widely accepted in clinical and research settings, they are not highly reproducible (Bombonati and Sgroi, 2011). Gene expression definitions of subgroups are dependent on the analytical method used, although most clusters result in similar ability to predict outcomes. It is likely that broad biological processes (e.g., cell proliferation) are driving each of the tumor subtypes, and these processes can be characterized by the expression of a number of different genes. Furthermore, although a gene expression array of a minimized set of genes for breast cancer is commercially available (e.g., PAM50), its use currently is not widespread in clinical and most research settings. While surrogate definitions of these subtypes based on the results of IHC staining for ER, PR, HER2, and the basal markers are widely used in clinical and research settings, there are issues with accuracy due to problems with marker staining (Welsh et al., 2011) and precision (Jenkins et al., 2014). Future research that integrates gene expression data with other molecular profiling techniques (e.g., somatic mutations, copy number, epigenetic alterations) is likely to result in deeper understanding of subtypes in breast and other anatomical tissues (Dieci et al., 2014).

Intratumoral heterogeneity presents further complexities in the identification of the molecular subtype of a tumor, and itself might also account for variability in prognosis, treatment response, and factors related to tumor initiation, promotion, and progression (Norum et al., 2014). Intratumoral heterogeneity is the coexistence of subclones of cancer cells that differ in histology, genetic sequence, epigenetic patterns, or protein expression. The heterogeneity might occur spatially or temporally. Subpopulations of tumor cells develop when changes in the genetic sequence or epigenetic patterns provide a selective advantage in the surrounding microenvironment. It is unclear whether the establishment and maintenance of tumor heterogeneity are due, in part, to different cells of origin (Martelotto et al., 2014).

Cell of Origin

Two theories of breast carcinogenesis currently are debated in the literature (Visvader and Stingl, 2014): the sporadic clonal evolution model (stochastic) and the cancer stem cell (hierarchical) theories. The sporadic clonal evolution model posits that any breast epithelial cell can be the target of random mutations. The cells with advantageous genetic and epigenetic alterations are selected over time to contribute to tumorigenesis. In contrast, the cancer stem cell theory postulates that stem cells, which either gave rise to the tumor or were acquired by a subpopulation of cells within the tumor, maintain tumorigenesis through their capacity for self-renewal and differentiation. The remaining bulk of tumor cells have limited proliferative potential.

In support of the cancer stem cell theory, researchers have proved the existence of mammary stem cells with multilineage differentiation and self-renewing capabilities that are responsible for the dynamic nature of the mammary epithelium throughout the life course (Visvader and Stingl, 2014). Although the lineage of mammary stem cells in the adult breast is an active area of research, it currently is thought that the mammary stem cell compartment contains long-term and short-term repopulating cells. These cells give rise to committed progenitor cells for the myoepithelial and luminal (ductal and alveolar) epithelial lineages.

The epithelial cellular hierarchy appears to be critical to understanding the different cells of origin for each molecular subtype. There is strong evidence to support that the claudin-low molecular subtype is derived from mammary stem cells, while the basal-like subtype is derived from luminal progenitor cells. It is suspected, but unconfirmed at this time, that luminal A, luminal B, and HER2-enriched tumors develop from more mature luminal progenitor cells (Visvader and Stingl, 2014). It is expected that the deeper understanding of cellular hierarchy in the normal breast tissue will be an active area of research and will lead to a better understanding of the carcinogenesis of breast cancer molecular subtypes.

Precancers or Precursor Lesions

While about a quarter of diagnostic breast biopsies in the United States yield invasive diagnoses, the vast majority of biopsy-based diagnoses range from benign to preinvasive disease (Weaver et al., 2006). Benign lesions and in situ carcinomas (lobular carcinoma in situ [LCIS] and ductal carcinoma in situ [DCIS]) are a morphologically and biologically heterogeneous group of lesions associated with varying degrees of subsequent breast cancer risk (Morrow et al., 2015). A classification scheme for benign breast disease (BBD), proposed by Dupont and Page (1985) and endorsed by the College of American Pathologists (Fitzgibbons et al., 1998), categorizes BBD lesions into three clinically relevant groups: nonproliferative, proliferative without atypia, and atypical hyperplasia (Hartmann et al., 2005) (Figure 45–2).

Breast Cancer

Figure 45–2. Histopathological appearance of benign breast disease. Panel (a) shows nonproliferative fibrocystic changes: the architecture of the terminal duct lobular unit is distorted by the formation of microcysts, associated with interlobular fibrosis. Panel (b) shows proliferative hyperplasia without atypia. This is adenosis, a distinctive form of hyperplasia characterized by the proliferation of lobular acini, forming crowded gland-like structures. For comparison, a normal lobule is on the left side. Panel (c) also shows proliferative hyperplasia without atypia. This is moderate ductal hyperplasia, which is characterized by a duct that is partially distended by hyperplastic epithelium within the lumen. Panel (d) again shows proliferative hyperplasia without atypia, but this is florid ductal hyperplasia: the involved duct is greatly expanded by a crowded, jumbled-appearing epithelial proliferation. Panel (e) shows atypical ductal hyperplasia: these proliferations are characterized by a combination of architectural complexity with partially formed secondary lumens and mild nuclear hyperchromasia in the epithelial-cell population. Panel (f) shows atypical lobular hyperplasia: monotonous cells fill the lumens of partially distended acini in this terminal duct lobular unit.

Source: Hartmann et al. (2005).

Large epidemiologic cohorts of patients diagnosed with BBD have established a relation between the histologic classification of BBD and breast cancer risk (Morrow et al., 2015). In a retrospective cohort of 9087 women diagnosed with BBD at the Mayo Clinic who were followed for a median of 15 years, the relative risks (RRs) in comparison to the general population were 1.27 (95% confidence interval [CI]: 1.15–1.41) for nonproliferative lesions, 1.88 (1.66–2.12) for proliferative changes without atypia, and 4.24 (3.26–5.41) for atypia (Hartmann et al., 2005). Results from a nested case-control study of BBD and breast cancer risk in the Nurses’ Health Study (NHS) also demonstrated that risks were highest among those with atypical hyperplasia: compared with women who had nonproliferative lesions, the ORs associated with proliferative lesions without atypia and atypical hyperplasia diagnoses were 1.62 (95% CI: 1.21–2.18) and 4.04 (2.76–5.92), respectively (Collins et al., 2006).

Atypical epithelial hyperplasia encompasses two histologically distinct lesions—atypical ductal hyperplasia and atypical lobular hyperplasia—and both lesions are associated with approximately a 4-fold increased risk of breast cancer (Morrow et al., 2015). In a 2014 update from the Mayo BBD Cohort, 698 women with atypical hyperplasia were followed for an average of 12.5 years, and subsequent breast cancers occurred with a 2:1 ratio in the ipsilateral compared with the contralateral breast. This ipsilateral predominance was marked in the first 5 years, suggesting that atypical hyperplasia lesions are in fact cancer precursors in some women (Hartmann et al., 2014). However, most women with atypical hyperplasia do not develop breast cancer (Morrow et al., 2015). In the Mayo BBD Cohort, the 20-year cumulative incidence of DCIS or invasive breast cancer among women diagnosed with atypical hyperplasia was 21% (95% CI: 14%–28%) (Degnim et al., 2007). Thus, identifying factors that modify cancer risk associated with these lesions is of great interest. Younger age or premenopausal status at diagnosis of atypical hyperplasia, multiple foci of atypia (Hartmann et al., 2014), and reduced TDLU involution in women with BBD (Baer et al., 2009; Milanese et al., 2006) (see later discussion) have all been associated with increased breast cancer risk. Although a positive family history of breast cancer was initially reported to increase the RR of breast cancer among women with atypical hyperplasia (Dupont and Page, 1985), this has not been (p.864) confirmed in more recent investigations (Collins et al., 2006; Hartmann et al., 2014).

While atypical ductal and lobular hyperplasia have been recognized as breast cancer risk factors since the 1980s, flat epithelial atypia, an alteration of breast lobules, has more recently been recognized by the WHO Working Group on the Pathology and Genetics of Tumors of the Breast and Female Genital Organs (World Health Organization, 2003). The natural history of flat epithelial atypia is not well understood, but emerging epidemiologic evidence suggests that the breast cancer risk associated with flat epithelial atypia is not as high as that observed with other atypical breast lesions (Morrow et al., 2015). A recent study from the Mayo BBD Cohort found that flat epithelial atypia was accompanied by atypical hyperplasia about half of the time, and the RR of breast cancer associated with isolated flat epithelial atypia (RR = 2.75, 95% CI: 1.76–4.10) was comparable to the risk associated with proliferative disease without atypia (Said et al., 2014).

LCIS and DCIS are both associated with increased risk of subsequent breast cancer. DCIS is considered to be a non-obligate precursor lesion to invasive breast cancer (Sherman et al., 2014), whereas LCIS is generally thought to be a more general marker of risk (Morrow et al., 2015); however, molecular studies suggesting a clonal link with invasive lobular carcinoma (Venkitaraman, 2010) have renewed interest in LCIS as a non-obligate precursor lesion as well as a risk indicator. Whereas the incidence of DCIS has increased dramatically with mammographic screening, manifesting generally as clustered calcifications (Virnig et al., 2010), LCIS is typically an incidental finding in breast biopsies, as it lacks clinical manifestations, such as a lump or other changes to the breast, and may not be detectable by screening mammography (Morrow et al., 2015).

The RR of invasive breast cancer subsequent to a diagnosis of LCIS ranges from ~7–10, and the RRs from more contemporary studies are similar to those reported in the 1970s (Morrow et al., 2015). The lack of untreated cohorts with DCIS is a challenge for calculating the RR of invasive breast cancer associated with DCIS. Among women diagnosed with DCIS who are treated with excision only, the risk of subsequent invasive breast cancer appears to exceed that seen with a prior LCIS diagnosis (Morrow et al., 2015). In addition, the risk of invasive breast cancer is highest in the same breast as the initial DCIS diagnosis, and the cancer often arises in the same breast quadrant (Erbas et al., 2006). While the evidence for DCIS as an invasive precursor is strong, DCIS is not an obligate precursor; in autopsy studies, the prevalence of undiagnosed DCIS is estimated to be around 9% (Welch and Black, 1997). A recent systematic review has shown that DCIS and invasive breast cancer share similar risk factor associations, including elevated mammographic density, family history of breast cancer, and history of benign breast disease, supporting the idea that they share a common etiology (Virnig et al., 2010).

As these high-risk lesions are diagnosed more frequently in the current era of broad mammography screening programs, individualized risk prediction is needed. This includes the identification of molecular biomarkers that predict progression to invasive carcinoma, given that such efforts to date have been unsuccessful (Allred, 2011). Until predictors of invasive carcinoma risk are found, high-risk precursor lesions pose a dual problem of overdiagnosis (and over-treatment) among some women and failure of early detection (or undertreatment) among others.

Descriptive Epidemiology

Demographic Factors


Unlike cancers at other sites, which generally show a log-linear relationship with age, the age incidence curve for breast cancer is log-linear only until about age 50, after which time the slope begins to flatten out. The change point around age 50 is referred to as the “Clemmesen’s hook.” To account for the peculiar shape of the breast cancer age-specific rate curve, Pike et al. (1983) proposed a concept of “breast tissue age” to reflect biological rather than chronological age based on key reproductive events. Risk factors accelerated breast cancer incidence rates, while protective factors retarded rates, culminating in a pause or inflection near menopause. Others have refined the Pike model, but none of these models incorporates the concept of breast cancer heterogeneity, nor can these models fully explain the distinct (p.865) age-specific incidence rate patterns by histopathologic and/or molecular subtypes (e.g., the different patterns for ER– and ER+ cancers). Age-specific rates for ER– cancers rise rapidly until age 50 years, and then plateau, decline, or fall. In contrast, age-specific rates for ER+ cancers rise continuously with advancing age, though more slowly after 50 years of age.

Alternatively, the distinct ER– and ER+ age-specific incidence rate patterns suggest that breast cancer overall consists of not one type but rather a mixture of two main types of cancer, with a different incidence rate pattern for each type. The first type is premenopausal with peak occurrence early in life (similar to ER– cancers). The second type is postmenopausal with peak incidence later in life (similar to ER+ cancers). Clemmesen’s hook can then be viewed as the confluence or superimposition of falling early-onset ER– and rising late-onset ER+ breast cancers. Notably, and somewhat paradoxically, the falling ER– incidence rates near age 50 years suggest that menopause (and by implication, exposures that occur early in reproductive life) have a greater impact on ER– than ER+ cancers.

Further evidence for two main breast cancer types is provided by the corresponding age distribution at diagnosis for breast cancer overall, which demonstrates a bimodal pattern with modal or peak ages near 50 and 70 years (Anderson et al., 2014). Bimodality is of interest because it implies heterogeneity in an otherwise homogeneous population. ER– and ER+ cancers also have bimodal age distributions at diagnosis. ER– cancers have a bimodal pattern with a dominant early mode near age 50 years and minor mode near age 70 years, whereas ER+ cancers have a bimodal pattern with a dominant mode near age 70 years and minor mode near age 50 years.

In recent analyses that considered not only ER and PR but also HER2 (which has recently been incorporated into Surveillance Epidemiology and End Results [SEER] data), patients with triple negative, HR+/HER2+ and HR–/HER2+ breast cancers were 10%–30% less likely to be diagnosed at older ages compared with HR+/HER2– patients. Notably, incidence rates for HR+/HER2– tumors peaked at 75–79 years of age, while triple negative cancers peaked prior to age 70 years; in contrast, HER2–overexpressing tumors (both HR+/HER2+ and HR–/HER2+) showed less dramatic increases in incidence with age than either of the other tumor subtypes (Howlader et al., 2014).


Although breast cancer is predominately a female disease, it does occur rarely among men, who exhibit an incidence rate that is approximately 1/100th that of females. Recent investigations have demonstrated that risk factors among men appear relatively similar to those among women (Brinton et al., 2014a), including showing a strong relation with higher levels of endogenous estrogens (Brinton et al., 2015). Because of the rarity of breast cancer in males, this chapter will focus on the disease that occurs much more commonly among females, where there has been extensive study of patterns of disease and risk factors.

Race, Ethnicity, and Socioeconomic Status

Breast cancer incidence rates show significant heterogeneity by race and ethnicity. The latest available SEER-18 statistics in the United States (SEER, 2015) show that, in general, among non-Hispanics rates are higher for white than black women (respective incidence rates per 100,000 of 135.3 vs. 125.0), but the reverse is true at younger ages, where there is a crossover and higher rates among blacks (Jatoi and Anderson, 2010). Among Hispanics, there are also differential rates between whites and blacks (respective rates of 95.7 vs. 55.8). In addition, rates are relatively low among Asian/Pacific Islanders (94.9) and American Indian/Alaskan Natives (61.1).

(p.866) Racial and ethnic heterogeneity has also been observed for different subtypes of breast cancer, with non-Hispanic white women having the highest incidence rates of the HR+/HER2– subtype, and non-Hispanic black women having the highest rates of triple negative cancers. Compared with women with the HR+/HER2– subtype, triple negative patients are more likely to be non-Hispanic black and Hispanic, HR+/HER2+ patients are more likely to be Asian/Pacific Islander, and HR–/HER2+ patients are more likely to be non-Hispanic black, Asian/Pacific Islander, and Hispanic (Howlader et al., 2014) (Figure 45–3).

Breast Cancer

Figure 45–3. Average annual incidence of breast cancer per 100,000 women by age group for different subtypes of breast cancer. Age-specific incidence rates of breast cancer subtypes by race/ethnicity, Surveillance, Epidemiology, and End Results 18, excluding Alaska, 2010.

Abbreviations: API = Asian Pacific Islander; HER = human epidermal growth factor; HR = hormone receptor; NH = non-Hispanic.

Source: Howlader et al. (2014).

These ethnic differences do not appear to be explained by socioeconomic strata differences (Sineshaw et al., 2014). However, it is increasingly being recognized that distributions across racial and ethnic groups are complex, with considerable diversity of subtypes even within populations, such as Asians (Parise and Caggiano, 2014), as well as within other population subgroups.

Breast cancer incidence has been noted to be highest among single women and women of higher socioeconomic status, presumably reflecting, at least in part, the influence of more prevalent risk factors (such as fewer children/delays in childbearing and greater exposure to menopausal hormone therapy) and greater access to screening. Such differences in prevalence of risk factors may also partially explain the racial and ethnic differences.

Geographic Variation

The United States has one of the highest incidence rates of breast cancer in the world. Although there is some geographic variation within the country (Siegel et al., 2015), it is small in comparison to international variation, with much of the national fluctuations presumably due to differences in the prevalence of established breast cancer risk factors.

International Patterns of Incidence and Mortality

Breast cancer is the most frequently diagnosed cancer and the leading cause of cancer death worldwide, with an estimated 1.7 million cases diagnosed during 2012 (comprising 25% of all cancers) (Torre et al., 2015). More developed countries account for about one-half of all breast cancer cases.

Incidence rates vary more than 4-fold across world regions, with annual standardized incidence rates ranging from 27 per 100,000 in Middle Africa and Eastern Asia to 96 in Western Europe. Rates are generally high in North America, Australia/New Zealand, and northern and western Europe; intermediate in central and eastern Europe, and the Caribbean; and low in most of Africa and Asia (Figure 45–4).

Breast Cancer

Figure 45–4. Estimated age-standardized breast cancer incidence rates worldwide per 100,000 in 2012.

Source: Ferlay et al. (2015).

Breast cancer ranks as the fifth cause of death from cancer overall (522,000 deaths). While it is the most frequent cause of death in women in less developed regions (324,000 deaths, 14.3% of total), it is now the second cause of cancer death, after lung cancer, in more developed regions (198,000 deaths, 15.4%) (Ferlay et al., 2015). Mortality differences across countries are not as substantial as incidence differences because of the more favorable survival of breast cancer in developed regions, with rates ranging from 6 per 100,000 in eastern Asia to 20 per 100,000 in western Africa. In many developing areas, including Africa, large proportions of women present with late stage tumors, for which treatment modalities are ineffective (Brinton et al., 2014b); further, in many of these countries, treatments such as radiotherapy and chemotherapy are not always available. Thus, a major challenge for many developing countries is in educating women to seek medical assistance at the first signs of symptoms to assure that the most effective treatment modalities can be utilized. Limited screening in many countries (largely reflecting an absence of mammography facilities) is a further impediment to early detection.

Migrant Studies

An interesting facet of breast cancer is that incidence rates change when individuals migrate from low incidence countries to high incidence countries. Among Asian women, it has been documented that when they migrate to the United States that within two to three generations their rates increase to or become even greater than US whites (Hoover, 2012). This pattern of change supports a more important role for environmental than genetic factors.

Although initially women in the West were noted to have breast cancer incidence rates 4- to 6-fold higher than women in certain Asian countries, more recently—in response to the “Westernization” of certain countries—this difference dropped to 2-to 3-fold. In a cohort of women from Shanghai, it has been demonstrated that a 2.8-fold difference in incidence with US whites decreased to a 1.4-fold difference after adjustment for a variety of breast cancer risk factors (Linos et al., 2008).

Given changes in migration patterns to the United States, attention has also focused on how breast cancer incidence rates change among Hispanics when they leave their host countries to reside in the United States. Analyses indicate that breast cancer risk is 50% lower in foreign-born Hispanics than US-born Hispanics (John et al., 2005). Among long-term foreign-born residents, risk was lower among Hispanics who moved to the United States at age 20 or older and those who mainly spoke Spanish. Further, the difference between third- or higher-generation Hispanics and recent migrants from rural areas was approximately 6-fold in postmenopausal women and 4-fold in premenopausal women. Adjustment for breast cancer risk factors attenuated these relationships.

Temporal Trends

In the United States, incidence rates of breast cancer have been gradually increasing. Between 1980 and 1987, a time when there was increased uptake of mammography, breast cancer incidence rates increased rapidly (Figure 45–5), followed by a slower rate of increase between 1987 and 2002 (Siegel et al., 2015), when rates declined sharply—presumably in response to declining rates of menopausal hormone therapy (MHT) usage following publication of findings from the Women’s Health Initiative (WHI) clinical trial linking estrogen plus progestin therapy to increases in the development of breast cancer (Ravdin et al., 2007). Since that time, incidence rates have been more stable, although with some upswings for older (≥ 50 years) women, particularly blacks.

Breast Cancer

Figure 45–5. Trends in incidence of invasive breast cancer (age-adjusted US standard) in selected populations, 1975–2012, SEER-9. Cancer sites include invasive cases.

Recent projections (Rosenberg et al., 2015) indicate that the total number of new tumors in the United States should rise from 283,000 in 2011 to 444,000 by 2030. This largely reflects proportional increases in older women and for ER+ in situ cancers; in contrast, the proportion of ER– tumors is expected to decrease. Increasing rates of ER+ tumors are primarily thought to reflect changing prevalences in breast cancer risk factors, including greater delays in childbirth and more obesity; reasons for declining rates of ER– tumors are less clear.

Globally, breast cancer incidence has been continually increasing. Between 1990 and 2013, incidence rates have increased 17% globally: 46% in developing countries and 8% in developed countries (Global Burden of Disease Cancer, 2015). Some of the most rapid increases in incidence have occurred in South America, Africa, and Asia (Torre et al., 2015), although the absence of standardized cancer registration in many countries leads to questions as to the accuracy of estimates. These increases are believed to be the result of changing reproductive patterns, increasing obesity, decreasing physical activity, and some breast cancer screening activity. The convergence toward the risk profile of Western countries has resulted in a narrowing of the international gap in breast cancer incidence.

In the United States, mortality from breast cancer increased during the 1930s, remained relatively stable from the 1940s through the 1970s, increased 0.4% per year from 1975 to 1990, and then decreased 36% from 1990 to 2012. The decrease occurred in both younger and older women, but since 2007 the breast cancer death rate has been level among women younger than 50 years of age. These trends are difficult to interpret given that they reflect combined effects of changes in incidence (e.g., ER– rates have recently fallen), variations in screening practices, and effectiveness of treatment. In recent modeling efforts (p.867) (p.868) aimed at assessing the influence of screening and therapy on decreases in mortality from breast cancer, it was determined that for ER+ cancers adjuvant treatment made a higher relative contribution than screening, whereas for ER– cancers the contributions were similar (Munoz et al., 2014). Although ER– cancers were less likely than ER+ cancers to be screen detected, when they were screen detected they yielded a greater survival gain.

Most recently, breast cancer death rates declined in each racial/ethnic group, with the exception of stable rates among American Indians/Alaska Natives. However, breast cancer death rates were 42% higher in black than white women by 2012. This mortality difference is likely due to a combination of factors, including differences in stage at diagnosis, distribution of survival factors (e.g., obesity and other comorbidities), and tumor characteristics, as well as access to, compliance with, and response to treatments.

Mortality worldwide has also been increasing, largely mirroring incidence trends. In many developing countries, the presentation of late-stage tumors and the absence of effective treatments further portend a poor prognosis.


Breast cancer is the leading cause of cancer death in women in developed countries (Torre et al., 2015). Recent relative survival rates for US women diagnosed with breast cancer overall were 89% at 5 years after diagnosis, 83% after 10 years, and 78% after 15 years. Since 1975, the breast cancer 5-year relative survival rate has increased significantly for both black and white women; however, the most recent data continue to show a substantial racial difference. The racial disparity in survival reflects both later stage at diagnosis and poorer stage-specific survival in black women.

Among clinical factors, prognosis for aggressive tumors is most strongly influenced by the presence of metastases in the liver, lung, or brain, the number of metastatic sites, and tumor grade (Arpino et al., 2015). For early stage disease, age (< 40 years), lymphovascular invasion, positive or close margin status at surgery, and large tumor size are the strongest predictors of recurrence (Kent et al., 2015). Tumor grade, scored 1 to 3, is a summary of an assessment of tubule/gland formation, nuclear pleomorphism, and mitotic count, and is most relevant for NST tumors (World Health Organization, 2012). Recent molecular genetic studies have shown that grade is an independent predictor of prognosis for ER+ tumors, even after the inclusion of gene signatures, and that grade 1 and 3 tumors are likely two different diseases.

There are two main staging systems for cancer. The TNM staging is the most widely used system and incorporates the extent of cancer at the primary site (tumor, T), the regional nodes (nodes, N), and the spread to distant metastatic sites (M, commonly found in the liver, bones, lungs, or brain). In the fourth edition of the WHO Classification of Tumors of the Breast (World Health Organization, 2012), tumors measuring 2.0 cm or less with micrometastases to axillary nodes are classified as stage Ib rather than stage II. Stage II tumors are therefore those with macrometastases to axillary nodes and represent a poorer prognosis than stage Ib tumors. The SEER Summary Stage system is more simplified and is commonly used in reporting cancer registry data. According to this system, local stage tumors are confined to the breast (corresponding to stage I and some stage II cancers in the TNM staging system), regional stage tumors have spread to surrounding tissue or nearby lymph nodes (most stage II or III cancers, depending on size and lymph node involvement), and distant stage tumors have metastasized to distant organs or lymph nodes above the sternum (stages IIIc and IV). Five-year relative survival by stage is 99% for localized disease, 85% for regional disease, and 26% for distant-stage disease (American Cancer Society, 2014).

Prognosis of invasive tumors also varies by special histological types. In general, lobular cancer has a good prognosis, although approximately 20% of lobular cancer diagnoses are contralateral at diagnosis and a very rare subtype (< 1%), pleomorphic lobular carcinoma, is associated with a poor outcome. Mucinous and tubular breast cancers tend to be diagnosed with good prognostic factors and have life expectancies similar to comparably aged women in the general population (Dieci et al., 2014). Despite the difference in prognosis by ER status for invasive carcinomas of NST, medullary cancers diagnosed as either ER+ or ER– have an excellent prognosis.

By molecular subtype, the mortality hazard rates for basal-like breast cancer rise steeply the first 2 years after diagnosis and then drop precipitously. The short-term survival of non-basal-like breast cancer (p.869) is significantly better, due to inherent characteristics of the tumor and the availability of targeted drugs. However, the risk of relapse of luminal A breast cancer extends up to 20 years or more after diagnosis. After 8 years of diagnosis, the survival curves for basal-like and non-basal-like breast cancers cross over. Prognosis for women diagnosed with luminal B tumors is considerably worse than that for those diagnosed with luminal A tumors, in part due to the more aggressive nature of the tumor and poor response to chemotherapy and endocrine therapy (Ades et al., 2014). HER2-enriched cancers have historically been associated with poor prognoses; however, the advent of HER2-targeted therapies has improved the prognosis of patients with such malignancies (Verma et al., 2012).

Survival Factors

Identifying factors that prolong life expectancy after a breast cancer diagnosis is a relatively new area of research. The World Cancer Research Fund International Continuous Update Project recently reviewed the existing literature regarding the associations of diet, weight, and physical activity with prognosis. The scientific panel determined that limitations in the design and execution of the existing studies hindered specific recommendations for breast cancer survivors; however, they suggested promising associations for improved breast cancer–specific survival with maintaining a healthy body weight and being physically active (WCRF/AICR, 2014). Specifically, two meta-analyses of epidemiologic studies found obesity (compared to normal weight) to be associated with elevated breast cancer risks and breast cancer–specific mortality (RR = 1.2–1.3), with slightly higher risks for premenopausal women (Chan et al., 2014; Protani et al., 2010).

Of 17 epidemiologic studies included in the Continuous Update Project (WCRF/AICR, 2014), most found a statistically significant benefit of physical activity on breast cancer–specific death, ranging from 36%–67% reductions, with more beneficial associations with physical activity for older compared to younger women (Fontein et al., 2013). Eating foods containing fiber and soy and diets low in total and saturated fat were determined to have limited evidence for associations with all-cause mortality (WCRF/AICR, 2014). The limited data for other dietary factors, including fruit and vegetable intake, alcohol intake, and adult attained height, precluded any conclusions in the Continuous Update Project.

Identification of other factors related to breast cancer–specific mortality has been hindered by a variety of methodological issues, including small numbers of breast cancer cases, recruitment of breast cancer survivors a year or more after diagnosis, limited assessment of exposures before and after diagnosis, and incomplete control of tumor characteristics and treatments that are known to be associated with exposures under study.


Currently 6.3 million women worldwide are living with a personal history of breast cancer (Jemal et al., 2011). The number of survivors is expected to grow as the proportion of older women increases and early detection and treatment improve. After completion of initial treatment, breast cancer survivors face unique medical, psychological, and social issues compared to the general population. Beyond higher risk of disease recurrence and second primaries, survivors experience short-term and long-term pain, fatigue, depression, and lymphedema (Luctkar-Flude et al., 2015). Factors associated with survival have also been examined for their role in breast cancer survivorship. For example, physical activity has been shown to improve quality of life and decrease depression and fatigue (Fontein et al., 2013). Ongoing research is focusing on survivorship care guidelines for physicians to improve the overall health and quality of life of breast cancer survivors (Luctkar-Flude et al., 2015).

Environmental Risk Factors

There has been extensive study of a variety of factors as they relate to breast cancer risk. Risk factors have been identified from cohort as well as case-control studies, with the former being less dependent on the effects of recall biases, which can impact the assessment of some of the exposures of interest (e.g., induced abortion). With more recent interest in comparing risk factors across tumor subtypes, case-case studies have also been commonly employed, offering an efficient analytic approach to examine etiologic heterogeneity (Begg and Zhang, 1994). In such investigations, risk factor distributions for breast cancer (p.870) subtypes are typically compared with luminal A tumors and provide a general indication of the direction of risk factor associations, albeit without derivation of magnitudes of risk given the absence of a non-diseased comparison group.

The magnitude of association for many of the identified risk factors is shown in Table 45–2, along with the extent to which risk factors vary by tumor molecular subtypes. It has been difficult to determine the magnitudes of associations for some of the subtypes given limited numbers studied, although it is becoming inherently clear that there is tremendous etiologic heterogeneity.

Table 45–2. Relative Strengths of Associations According to Subtypes Defined by Molecular Markers

Risk Factor

All Breast Cancers—Magnitude of Risk

Luminal A

Luminal B

HER2 Overexpressing

Triple Negative

Younger age at menarche

RR = 1.6

< 12 vs. 15+ years






RR = 0.5

3+ vs. 0 births




Older age at first birth

RR = 3.0

35+ vs. < 20 years among parous women






RR = 0.5

2+ years vs. none among parous women


Older age at menopause

RR = 2.0

55+ vs. < 45 years





Obesity (premenopausal)

RR = 0.8

BMI 30+ vs. < 25




Obesity (postmenopausal)

RR = 1.3

BMI 30+ vs. < 25





Family history of breast cancer

RR = 2.0

Any first-degree relative vs. none





Alcohol use

RR = 1.6

2+ drinks/day vs. none





Oral contraceptive use

RR = 1.3

10+ years of use vs. none for early onset cancers




Menopausal hormone therapy use

RR = 2.0

10+ years of combination estrogen+progestin use





Pluses indicate the consistency of positive associations between the exposure and the tumor subtype, e.g., +++ indicates consistent evidence of a positive association, ++ indicates a probable positive association, and + indicates a possible positive association. Minuses indicate similar consistency of inverse associations. Abbreviations: HER2 = human epidermal growth factor receptor; RR = relative risk; BMI = body mass index.

Adapted from (Barnard et al., 2015)

Menstrual and Reproductive Factors

Menstrual Characteristics

It is well recognized that breast cancer risk is inversely associated with early ages at menarche and directly associated with later ages at natural menopause, presumably reflecting the importance of accumulating exposure to endogenous hormones. In results published from the Collaborative Group on Hormonal Factors in Breast Cancer (Collaborative Group on Hormonal Factors in Breast Cancer, 2012), breast cancer risk increased by a factor of 1.05 (95% CI: 1.04–1.06) for every younger year at menarche and independently by 1.03 (1.02–1.03) for every older year at menopause. In addition, conditioned on age, premenopausal women had a greater risk of breast cancer than postmenopausal women (RR at age 45–54: 1.43, 1.33–1.52).

Studies have also shown that women who experience an early artificial menopause due to a bilateral oophorectomy are at reduced risk, experiencing even lower risks than women undergoing natural menopause at comparable ages, reflecting a more precipitous decline in endogenous hormones among those undergoing ovarian ablation. One study found that women who had a bilateral oophorectomy prior to age 45 had an odds ratio (OR) of 0.59 (95% CI: 0.50–0.69) compared with premenopausal women who have not undergone premenopausal reproductive surgery; this risk reduction, however, was observed only for ER+ tumors, with no risk reduction seen for ER– tumors (Press et al., 2011). Hysterectomy with ovarian conservation also appears to result in some risk reduction, presumably due to this operation also having the potential to damage ovarian function (Gaudet et al., 2014b). In contrast, tubal sterilization does not to appear to impact breast cancer risk (Gaudet et al., 2013b).

Recent studies have focused on whether menstrual associations vary according to tumor subtypes, with some evidence that age at menarche is moderately inversely associated with triple negative breast cancers (Ambrosone et al., 2015; Islam et al., 2012; Li et al., 2013a; Tamimi et al., 2012) and probably inversely associated with luminal A cancers (Gaudet et al., 2011; Tamimi et al., 2012). Age at menarche relationships with luminal B and HER2 overexpressing tumors are less clear. In terms of age at menopause, the most consistent relations have been seen for luminal A tumors (Tamimi et al., 2012; Yang et al., 2011). There is also some evidence that all three menstrual parameters (age at menarche and menopause and type of menopause) show stronger associations with lobular tumors than other histologies (Collaborative Group on Hormonal Factors in Breast Cancer, 2012).

Reproductive Factors

Investigations have highlighted elevated risks associated with nulliparity and late ages at first birth, factors that appear to operate independently of each other. Nulliparous women are at approximately twice the risk of women who have had three or more births, while the relation with later ages at first birth is linear, with approximately a 2- to 3-fold increased risk for women with a first birth after age 35 as compared with prior to age 20. The effects of these reproductive parameters appear dependent on the pregnancy being full-term, with no apparent relations of risk for either miscarriages or induced abortions (Guo et al., 2015).

Although an increased risk is associated with later ages at first birth, some studies suggest that this relation is only observed approximately 5 or more years following a birth, with there being a transient increase in risk for short intervals since a last birth. This effect has been postulated to reflect the time required for pregnancy hormones to promote the progression of breast cells that have undergone early stages of malignant transformation. This is thought to at least partially explain why women diagnosed at young ages exhibit increased risks associated with parity (rather than nulliparity), as well as why women who delay a first birth until after the age of 30 demonstrate higher risks than nulliparous women.

Lower risks for multiparous versus nulliparous women have been most consistently seen for luminal A tumors (Ma et al., 2010; Tamimi et al., 2012; Yang et al., 2011). In contrast, a number of investigations have found parity unrelated or directly related to basal-like or triple-negative cancers (Li et al., 2013a; Ma et al., 2010; Palmer et al., 2014; Phipps et al., 2011a; Yang et al., 2011). Conflicting results have emerged regarding whether the relations of reproductive factors for HER2 overexpressing tumors are similar to or divergent from those observed for luminal tumors. This most likely reflects that studies have only recently focused on such subtyping, with many of the resultant studies involving relatively small numbers of women (Horn et al., 2014; Li et al., 2013a; Ma et al., 2010; Phipps et al., 2011a).

Increased risks related to older ages at first birth have been most consistently observed for luminal A tumors (Gaudet et al., 2011; Ma et al., 2010; Tamimi et al., 2012), with inconsistent associations observed for the other breast tumor subtypes.


Among parous women, breastfeeding, especially if long term, can result in risk reductions. The most comprehensive data for this derive from a large collaborative analysis, which showed the RR of breast cancer decreasing by 4.3% for every 12 months of breastfeeding, in addition to a decrease of 7.0% for each birth (Collaborative Group on Hormonal Factors in Breast Cancer, 2002).

Like other reproductive factors, these relations are most likely not consistent across tumor subtypes. Although inverse relations have been seen with extended breastfeeding for the two types of luminal tumors, several studies have indicated that breastfeeding relations may be strongest for basal-like or triple negative cancers (Gaudet et al., 2011; Li et al., 2013a; Palmer et al., 2014; Tamimi et al., 2012). In fact, the consistency of this observation has prompted the suggestion that breastfeeding is the one established protective factor for this otherwise aggressive form of breast cancer (Li et al., 2013a). Further, in one investigation it was noted that breastfeeding negated the high risk associated with parity observed for ER–/PR– tumors, suggesting that the higher incidence of such tumors in black women may in part be explained by higher parity and lower prevalence of breastfeeding (Palmer et al., 2011).

Exogenous Hormones

Hormonal Birth Control

The most conclusive evidence linking oral contraceptives to breast cancer risk derives from a large collaborative analysis, which found a 24% increase in risk (95% CI: 1.15–1.33) related to recent usage, which dissipated 10 or more years after discontinuation (Collaborative Group on Hormonal Factors in Breast Cancer, 1996). Relations in this analysis were somewhat stronger for women diagnosed with breast cancers prior to 35 years of age. Although many investigations have assessed whether different types of preparations could have varying effects on risk, it has been difficult to draw definitive conclusions, possibly due to frequent changes in prescribing patterns and difficulties in patients recalling patterns of usage. However, several recent investigations have shown elevated risks associated with combination oral contraceptives containing either high dosages of estrogens or ethynodiol diacetate, or triphasic preparations containing the progestins, levonorgestrel or norethindrone (Beaber et al., 2014a; Hunter et al., 2010), supporting the need for further investigations of risks by types of formulations used.

Recent analyses have focused on whether associations vary according to tumor subtypes. Some studies have suggested that oral (p.871) contraceptives may exert stronger effects for HR– or basal-like/triple negative cancers (Beaber et al., 2014b; Ma et al., 2010). However, another large investigation showed recent oral contraceptive usage similarly associated with ER+, ER–, and triple negative tumors, although the risk decline after cessation of use was not apparent for ER+ cancers until 15–19 years after cessation and for even longer intervals for ER– cancers (Bethea et al., 2015).

Some (Li et al., 2012), but not all (Strom et al., 2004), studies have found recent use of the injectable progestin-only contraceptive depot-medroxyprogesterone actetate (Depo-Provera) to be associated with increased breast cancer risk. Studies on levonorgestrel-releasing intrauterine devices (IUDs) have provided conflicting results. However, the most recent investigation on the topic (Soini et al., 2014) found some evidence for increased risk, of interest given the consistency of the finding with adverse effects noted for women who use combination estrogen plus progestin MHT (see the following).

Menopausal Hormone Therapy

Although the effects of MHT use have been the attention of epidemiologic studies for several decades, relationships to breast cancer risk only became widely accepted in 2002 when results from the WHI were published showing that women in the continuous combined estrogen-progestin therapy arm of this large clinical trial had 40% increases in their risk of breast cancer—an association that has persisted with extended follow-up (Chlebowski et al., 2013). The initial publication of these findings resulted in precipitous declines in usage nationwide, with subsequent decreases in the incidence of breast cancer linked to the decreased usage patterns. In the WHI trial, as well as in other investigations, there was evidence of substantial variation in hormone usage associations according to body mass, with much stronger relations seen among thin women, supporting a stronger effect of exogenous hormones in the presence of low levels of endogenous hormones. Studies have also demonstrated higher risks when hormone therapy is begun before or soon after menopause than after a longer gap (Beral et al., 2011).

In contrast, women in the estrogen-only arm of the trial did not demonstrate a breast cancer excess, which contrasted with numerous observational studies demonstrating increased risks following such usage (Bakken et al., 2011; Beral et al., 2011; Brinton et al., 2008; Calle et al., 2009). Reasons for this discrepancy have been widely discussed, with no immediate resolution. Attenuated risks may at least partially reflect the high rate of obesity among trial participants.

Relationships with MHT use tend to be strongest for luminal A breast cancers (Barnard et al., 2015). In one study, over a 2-fold increased risk for such tumors was observed among current estrogen plus progestin users of 15 of more years (Saxena et al., 2010). There is also some suggestion that associations may be stronger for lobular than ductal cancers. However, lobular cancers tend predominantly to be ER+, and it is unclear once this is accounted for whether histologic differences persist (Brinton et al., 2008).

Fertility Drugs

The relation of fertility drugs to breast cancer risk has been assessed in a number of studies, with recent investigations focusing on drugs used in the context of in vitro fertilization. In general, studies have provided reassuring results regarding effects on breast cancer risk, although a few studies have raised concern regarding slightly elevated risks associated with either higher drug dosages (Brinton et al., 2014c) or long intervals since initial exposures (Reigstad et al., 2015). A few investigations have noted higher fertility drug-related risks for nulliparous as compared with parous women, suggesting that associations may be more reflective of resistant infertility rather than of the drugs prescribed. The relation of fertility drugs to cancer risk is difficult to assess given the need to account for indications for usage; for breast cancer, studies have suggested that both endometriosis and polycystic ovary disease may be independent predictors, although with some discrepant results (Barry et al., 2014; Gottschau et al., 2015; Munksgaard and Blaakaer, 2011).


Women prescribed diethylstilbestrol (DES) during pregnancy to attempt to lower their risks of miscarriage have been found to have a modestly increased risk of breast cancer (RR = 1.27, 95% CI: 1.07–1.52) (Titus-Ernstoff et al., 2001). There is also some suggestion that daughters exposed in utero to DES may have a slightly elevated breast cancer risk; when attention focused on breast cancers that developed at older ages (≥ 40 years of age), the risk was significantly elevated (RR = 1.82, 95% CI: 1.04–3.18), supporting the need for further follow-up of such exposed women (Hoover et al., 2011).

Other Medications

It is well established that women with low bone mineral density and those with bone fractures are at decreased breast cancer risk, presumably due to their low levels of endogenous estrogens. This relation has been demonstrated to be stronger for ER+ than ER– cancers (Grenier et al., 2011). Bisphosphonates, prescribed to deter bone loss, have therefore been of interest with respect to breast cancer risk, although it has not always been possible to disentangle indications for usage from actual drug effects. Usage has been found in several studies to be inversely associated with breast cancer risk, although the findings are less convincing from randomized trials than they are from observational studies (Chlebowski et al., 2010).

There is growing interest in the role of inflammation in the etiology of breast cancer. In the only randomized clinical trial to date, the Women’s Health Study found that alternate-day use of low-dose aspirin versus placebo for an average of 10 years did not reduce breast cancer development (Cook et al., 2005). However, numerous observational studies have shown that regular users of non-steroidal anti-inflammatory drugs (NSAIDs) are at a reduced risk of breast cancer. One meta-analysis found NSAID use to be associated with a RR of 0.88 (95% CI: 0.84–0.93), with similar relations seen for aspirin and ibuprofen use (Takkouche et al., 2008). Given the potential adverse side effects of these medications, future efforts that consider the potential risk–benefit ratio may help to identify specific subgroups of women for whom targeted chemoprevention with aspirin and other NSAIDs may be appropriate (Bardia et al., 2016).

Metformin is another drug that has garnered recent attention in terms of its effects on reducing breast cancer risk. However, the association appears to be relatively modest and whether it is a causal relationship has been questioned (Gandini et al., 2014). Results of ongoing clinical trials to assess the efficacy of metformin in primary and secondary prevention will be important in further clarifying effects.

Statins act as a competitive inhibitor of hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase and have anti-proliferative, apoptotic, and anti-invasive properties. Several studies have indicated a reduction in overall breast cancer risk associated with usage, as well as earlier stages of disease at diagnosis, lower rates of recurrence, and lower breast cancer mortality. In one of the latest analyses, deriving from the WHI (Desai et al., 2015), an association with lower breast cancer stage at diagnosis was observed (RR = 0.80, 95% CI: 0.64–0.98), with stronger reductions seen for ER+ tumors, along with marginally lower risks of breast cancer mortality (RR = 0.59, 95% CI: 0.32–1.06).

Nutritional Factors

Despite hypotheses drawn from migrant and other ecological studies, there has been no conclusive evidence on the associations of diet with breast cancer risk, with the exception of alcohol intake (Vera-Ramirez et al., 2013) (see later discussion). More recent studies, however, are providing new support for possible associations of dietary fats, soy consumption, and fruit and vegetable intake with breast cancer risk. Although a recent meta-analysis of adult animal fat intake and breast cancer that involved more than 20,000 breast cancer cases concluded that there was no association (Alexander et al., 2010), findings from the NHS suggest that the timing of the exposure may be important given that a high-fat diet during adolescence was found associated with a moderate increase in premenopausal breast cancer risk (Linos et al., 2010). Soy consumption may (p.872) reduce breast cancer risk, supported in part by the historically low breast cancer rates among Asian women. A meta-analysis showed that soy intake was inversely associated with breast cancer risk in Asian but not Western populations, perhaps because Asian women start consuming soy products at an earlier age than women in Western populations (Dong and Qin, 2011). Further, there is growing evidence that high levels of fruit and vegetable consumption may reduce the risk of ER– breast cancer (Jung et al., 2013). The effect of diet on breast cancer risk remains an active area of research, with studies focusing on relations according to timing of exposure, specific dietary components (including metabolomics), and risks by tumor molecular subtype.

Alcohol is a causal risk factor for breast cancer according to the International Agency for Research on Cancer (IARC Working Group on the Evaluation of Carcinogenic Risks to Humans, 2010). Compared to non-drinkers of alcohol, intake of up to one alcoholic drink per day as an adult is associated with a 4% higher risk of breast cancer, while three or more drinks per day is associated with a 40%–50% higher risk (Seitz et al., 2012). The association is similar for risk of premenopausal and postmenopausal breast cancer. The strongest evidence for an association with alcohol is for luminal A breast cancers (Barnard et al., 2015). Evidence is weaker for triple negative breast cancers (Barnard et al., 2015); however, the magnitude of the association might actually be greater for risk of triple negative than luminal A breast cancer (Romieu et al., 2015). There is little to no association with risk of HER2 overexpressing breast cancer (Barnard et al., 2015; Romieu et al., 2015).

Anthropometric Factors


Epidemiologic studies have consistently demonstrated a positive association between height and breast cancer risk. In the Million Women Study cohort, the RR for invasive breast cancer was 1.17 (95% CI: 1.14–1.20) per 10 cm increase in height (Green et al., 2011). In addition, several studies have suggested that more rapid height growth during childhood and adolescence may influence breast cancer risk (Colditz et al., 2014). Two European cohorts have found significant increases in breast cancer risk associated with measured height growth in childhood. This includes a Danish study, which found an RR of 1.17 (95% CI: 1.09–1.25) per 6 cm increase (95% CI: 1.09–1.25) for growth from ages 8 to 14 years (Ahlgren et al., 2004), and a British study, which found RRs of 1.54 (95% CI: 1.13–2.09) per one standard deviation increase in height velocity for growth from ages 4 to 7 years and 1.29 (0.97–1.71) for growth from ages 11 to 15 years (De Stavola et al., 2004). The mechanisms underlying the association between height and breast cancer risk are complicated, as adult height is positively correlated with age at menarche (Luo et al., 2003), which in turn is inversely associated with breast cancer risk (Collaborative Group on Hormonal Factors in Breast Cancer, 2012). Genetic and environmental factors, including diet and levels of hormones and growth factors during childhood and adolescence, may link attained height to breast cancer risk (Crowe et al., 2011; Schernhammer et al., 2007).

Weight and Weight Change During Adulthood

An association between body mass index (BMI, calculated as weight in kilograms divided by height in meters2), a proxy of overall adiposity, and breast cancer risk is well established (Renehan et al., 2008; Suzuki et al., 2009; Vrieling et al., 2010) and seems to be modified by menopausal status and possibly by MHT use (WCRF/AICR, 2008). In a systematic literature review and dose–response meta-analysis of cohort studies published in 2008 by the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR), breast cancer risk decreased with higher BMI in premenopausal women (RR = 0.93 per 5 kg/m2 increase in BMI, 95% CI: 0.88–0.98), and higher BMI was noted as being a “probable” risk-reducing factor (WCRF/AICR, 2008). In contrast, postmenopausal breast cancer risk increased with increasing BMI (RR = 1.13 per 5 kg/m2 increase, 95% CI: 1.08–1.18), a relation that has been deemed a “convincing” breast cancer cause. In a recent secondary analysis of postmenopausal participants in the WHI clinical trials, where weight and height were measured at baseline and annually and mammography was required as part of the trial protocol, thus minimizing ascertainment bias, investigators found a strong linear dose–response relationship in which breast cancer risk progressively increased across BMI categories (Neuhouser et al., 2015). Compared with women with a BMI less than 25, women with a BMI higher than 35 had a 58% (95% CI: 1.40–1.79) increased risk of invasive breast cancer. Numerous observational studies have suggested that the positive association between obesity (i.e., BMI ≥ 30) and breast cancer among postmenopausal women may be limited to those who have never taken MHT (reviewed in Munsell et al., 2014; WCRF/AICR, 2008; and Kabat et al., 2015); however, a biologic mechanism to explain effect modification by MHT is not clear, and the WHI clinical trials report did not confirm this finding (Neuhouser et al., 2015).

Limited studies have evaluated the association between BMI and breast cancer among racially and ethnically diverse populations, though relationships seem to be similar for white and African American women by menopausal status (Amadou et al., 2013; Munsell et al., 2014). There is some evidence to suggest heterogeneity in the BMI risk association among Asian women for whom an elevation in premenopausal breast cancer risk associated with increased BMI has been observed (Amadou et al., 2013; Renehan et al., 2008). Among Hispanics, BMI associations with premenopausal breast cancer have been similar to those observed in non-Hispanic white populations (John et al., 2015; Munsell et al., 2014), whereas some reports in postmenopausal Hispanic women have been null (Muller et al., 2014; Sexton et al., 2011). Findings underscore the importance of adequately powered studies of adiposity and breast cancer in diverse populations.

Whereas adult weight gain appears to be unrelated to risk of premenopausal breast cancer (Keum et al., 2015; Keum et al., 2008), a statistically significant increased risk associated with adult weight gain has been reported for postmenopausal breast cancer in a recent meta-analysis of cohorts of non-users of MHT (RR = 1.11 for each 5 kg increase in weight [95% CI 1.08–1.13]) (Keum et al., 2015). Notably, in the WHI clinical trials, women with a baseline BMI < 25 who gained more than 5% of body weight during follow-up experienced an increased breast cancer risk (RR = 1.36, 95% CI: 1.11–1.65), but among women already overweight or obese at baseline there was no association of weight change (gain or loss) with breast cancer risk (Neuhouser et al., 2015). Weight control among normal weight women may therefore represent an effective strategy for breast cancer risk reduction among postmenopausal women, but this notion requires confirmation in clinical trials. Additional research on the role of weight loss in overweight and obese women is needed.

Stratification by menopausal status has also shed some light on the relationship between BMI and risk of intrinsic tumor subtypes (reviewed in Barnard et al., 2015). Premenopausal obesity appears to decrease the risk of luminal A tumors (Millikan et al., 2008; Yang et al., 2007) and possibly increases the risk of triple negative tumors (Chen et al., 2013; Gaudet et al., 2011; Kwan et al., 2009; Millikan et al., 2008; Yang et al., 2011). Among postmenopausal women, there is some evidence to suggest that associations for BMI (Canchola et al., 2012; Munsell et al., 2014; Neuhouser et al., 2015) and weight gain (Canchola et al., 2012; Suzuki et al., 2009; Vrieling et al., 2010) are stronger for ER+/PR+ as compared to ER–/PR– breast cancer, but no clear patterns of association between BMI and specific intrinsic breast cancer subtypes have emerged (Barnard et al., 2015).

Body Fat Distribution

Measures that reflect central adiposity and intra-abdominal visceral fat, such as waist circumference and waist-to-hip ratio (WHR), appear to be positively related to breast cancer risk irrespective of menopausal status, though associations are weak and are affected by adjustment for BMI (WCRF/AICR, 2008). In a pooled risk estimate from two large cohorts, a positive association between waist circumference and breast cancer risk among premenopausal women was borderline statistically (p.873) significant after accounting for overall obesity (RR = 1.12 per 8 cm increase, 95% CI: 1.00–1.25) and the RR for WHR was not statistically significant (WCRF/AICR, 2008). In a more recent report, waist circumference and WHR were not significantly associated overall with premenopausal breast cancer; however, both metrics were associated with increased risk of ER– breast cancer (RR = 2.75 for Q5 vs. Q1, 95% CI: 1.15–6.54) for waist circumference and 1.95 (95% CI: 1.10–3.46) for WHR (Harris et al., 2011), suggesting that abdominal adiposity may influence premenopausal breast cancer risk through sex hormone–independent pathways.

In the WICR/AICR meta-analysis involving four postmenopausal cohorts, the relation between greater waist circumference and increased breast cancer risk was attenuated with BMI adjustment (RR = 1.04, 95% CI: 1.00–1.06) and the RR for WHR was not statistically significant (WCRF/AICR, 2008). Consistent with these findings, a recent report concluded that waist circumference was not associated with postmenopausal breast cancer risk beyond its contribution to BMI (Gaudet et al., 2014a). In contrast, in the WHI observational cohort and randomized trials, a significant trend between waist circumference and breast cancer risk was unchanged after BMI adjustment (RR = 1.42 for Q5 vs. Q1, 95% CI: 1.31–1.53) (Kabat et al., 2015).

Several hypothesized biologic mechanisms may explain how adiposity influences breast cancer risk. Obesity may increase levels of circulating endogenous sex hormones owing to the conversion of androgens to estrogens by aromatase in the adipose tissue. Insulin and insulin-like growth factors and genetic predispositions to obesity and to specific body fat distributions have also been implicated (Calle and Kaaks, 2004).

Physical Activity

Epidemiologic evidence suggests that increased physical activity reduces breast cancer risk. In a review of 73 studies conducted worldwide through 2009, a 25% average risk reduction was found for the most physically active women as compared with the least active women (Lynch et al., 2011b). Associations varied depending on the intensity, type, and timing of activity. The strongest associations were observed for activity of moderate to vigorous intensity, recreational activity, and activity sustained across the life course (Lynch et al., 2011b).

Inverse associations with increased activity have been observed for both pre- and postmenopausal breast cancer risk, with more consistent associations noted among postmenopausal women (Lynch et al., 2011b; WCRF/AICR, 2008). Among postmenopausal participants in the E3N Cohort (the French component of the European Prospective Investigation into Cancer and Nutrition, EPIC), even modest levels of recent recreational activity (≥ 12 metabolic equivalent hours/week within the previous 4 years) were associated with significant reductions in breast cancer risk (Fournier et al., 2014). Among US postmenopausal participants in the Cancer Prevention Study-II (CPS-II) Cohort, ≥ 7 hours/week of walking (relative to ≤ 3 hours/week of walking) were associated with a significant reduction in breast cancer risk (Hildebrand et al., 2013). While physical activity has been associated with a lower risk of breast cancer irrespective of hormone receptor status (Phipps et al., 2011b), a recent meta-analysis of prospective studies found stronger inverse associations for ER– and PR– as compared with ER+/PR+ breast cancers (Wu et al., 2013). Few studies have evaluated physical activity and breast cancer risk among African American women, and results have been inconsistent; a recent prospective analysis of data from the Black Women’s Health Study found that high levels of recent vigorous exercise or brisk walking may reduce breast cancer incidence (Rosenberg et al., 2014).

Sedentary Behavior

While numerous studies have evaluated physical activity in relation to breast cancer, relatively fewer studies have considered the potential independent role of sedentary behavior on risk. Two recent meta-analyses demonstrated statistically significant increased risks of breast cancer (ranging from 8% to 17%) associated with sedentary behavior (Shen et al., 2014; Zhou et al., 2015). Additional studies are needed in diverse study populations to better understand relationships of sedentary behavior with risk of specific subtypes of breast cancer (Cohen et al., 2013).

The influence of both active and sedentary behavior on breast cancer risk is biologically plausible. Sedentary behavior, as objectively assessed using accelerometers, has been independently associated with breast cancer risk factors, including increased adiposity, elevated endogenous estrogens (Dallal et al., 2015), and biomarkers of insulin resistance and increased inflammation (Healy et al., 2011; Lynch et al., 2011a). These biological pathways are implicated in breast carcinogenesis and are also hypothesized mechanisms for the association between physical activity and breast cancer risk (Neilson et al., 2009).

Cigarette Smoking

Determining whether there is a causal relationship between active cigarette smoking and breast cancer risk has been controversial. There are strong biological data linking active smoking, particularly at young ages, with breast carcinogenesis. This includes the identification of 20 tobacco smoke compounds that induce mammary tumors in rodents and, in human breasts, the detection of tobacco metabolites, smoking-specific DNA adducts, and p53 mutation smoking-signatures (US Department of Health and Human Services, 2014). A meta-analysis of 15 published cohort studies, which included over 750,000 women, found increased risks of breast cancer, compared to never smokers, for both current (RR = 1.16, 95% CI: 1.11–1.20) as well as former (RR = 1.09, 95% CI: 1.06–1.13) smokers (Gaudet et al., 2013a). Parous smokers who started smoking before first birth were at higher risk than parous never smokers (RR = 1.24, 95% CI: 1.16–1.31). However, despite the large number of epidemiologic studies and consensus reviews of the literature on the topic, a 2014 US Surgeon General’s report concluded that “the evidence is suggestive but not sufficient to infer a causal relationship between active smoking and breast cancer” (US Department of Health and Human Services, 2014). This report noted several lingering epidemiological issues concerning the assessment of smoking and breast cancer risk, including possible confounding or effect modification by alcohol consumption (US Department of Health and Human Services, 2014).

Most studies have not found a link between exposure to secondhand smoke and breast cancer risk (Xue et al., 2011; Yang et al., 2013). However, some recent studies suggest that secondhand smoke may increase risk, particularly for premenopausal breast cancer (Dossus et al., 2014; Luo et al., 2011; Pirie et al., 2008).

Family and Personal History of Breast Cancer

Familial aggregation of breast cancer and twin studies (Lichtenstein et al., 2000) suggest that there is a considerable genetic component to breast cancer risk. Women with one or more first-degree female relatives with a history of breast cancer have a 1.5- to 2-fold higher risk of developing breast cancer themselves (Familial Breast Cancer Working Group, 2001), regardless of intrinsic subtype (Barnard et al., 2015). Breast cancer risk increases with the number of affected family members and when family members are diagnosed at a young age or with bilateral disease. These findings have stimulated much recent attention on the role of genetic alterations in the etiology of breast cancer, as discussed in more detail in the section “Host Factors” later in this chapter.

Women with an index breast cancer have between a 4%–15% lifetime risk of developing metachronous contralateral breast cancers (Bernstein et al., 2003). This equates to approximately a 2-fold increased risk of development of a second primary tumor in the contralateral breast compared with the risk of women in the general population developing a first primary breast cancer, who have an annual risk of 0.4%–0.5% (Kurian et al., 2009; Portschy et al., 2015). The estimated 10-year cumulative incidence of contralateral breast cancer for women aged 25–54 years at diagnosis of their index breast cancer is 4.6% (95% CI: 4.0–5.1%), which is further increased among women with a first-degree family history of breast cancer (8.6%, 95% (p.874) CI: 6.1–11.5%), or a history of bilateral breast cancer in a first-degree relative (15.6%, 95% CI: 8.5–28.5%) (Reiner et al., 2013).

Predisposing Medical Conditions

As previously discussed, obesity is a well-recognized risk factor for postmenopausal breast cancer, and it is therefore not surprising that metabolic syndrome has also been of concern. It appears, however, that obesity may be the most important component of this syndrome in affecting breast cancer risk, with little evidence that other components (including diabetes) (Tsilidis et al., 2015) have strong effects on risk. Risks, however, may vary by tumor subtype (e.g., a recent study demonstrated a high risk of diabetes associated with triple negative cancers; Garcia-Esquinas et al., 2015), thus warranting continued evaluation of subtype-specific associations.

A variety of other diseases have also been investigated with respect to breast cancer risk. Apart from the previously discussed inverse association with osteoporosis, most other diseases appear not to be strongly associated with breast cancer risk after adjustment for correlated risk factors, such as obesity. This includes histories of both hypertension and various thyroid diseases.

Early Life Factors

Recent attention has focused on the relations of early life exposures on subsequent breast cancer risk, including in utero exposures. Numerous studies have evaluated such relations, with one meta-analysis documenting elevated risks associated with higher birth weights (RR = 1.15, 95% CI: 1.08–2.21), longer birth lengths (1.28, 1.11–1.48), and more advanced maternal (1.13, 1.02–1.25) and paternal (1.12, 1.05–1.19) ages, and decreased risks associated with maternal pre-eclampsia (0.48, 0.30–0.78) and twin membership (0.93, 0.87–1.00) (Xue and Michels, 2007). Additional studies are needed to elucidate potential underlying mechanisms for these associations, which could include exposure to maternal endogenous sex hormones and growth factors, germ cell mutations, formation of cancer stem cells, or other genetic/epigenetic events.


Elevated risks of breast cancer have been observed for atomic bomb survivors, particularly among those exposed at young ages (Kaiser et al., 2012), suggesting a stronger influence of radiation on undifferentiated breast cells. Notably, for age at exposure of 25 years and attained age of 70 years, the excess RR for 1 Gy of radiation exposure has been estimated to be 1.2 (90% CI: 0.72–2.1), which is 30% higher than that associated with exposure at age 55 years (Jacob and Stram, 2013).

High-dose radiation to the chest has also been related to elevated breast cancer risks. This includes those who have received radiation therapy for childhood cancers, for whom breast cancer risk has been demonstrated to be directly related to the dose of the radiation received (Morton et al., 2014). In one analysis, in which the median delivered dose of radiation was 14 GY to a large volume (whole-lung field), the standardized incidence ratio (SIR) of breast cancer was 43.6 (95% CI: 27.2–70.3) and the cumulative incidence by age 50 years was 30% (95% CI: 25–34) (Moskowitz et al., 2014).

Occupational Exposures

As increasing numbers of women enter the workforce, occupational exposures are becoming more of a concern with respect to breast cancer risk. Of primary interest have been the effects of exposures to polycyclic aromatic amines, plasticizers, and solvents. The number of studies evaluating such effects has been modest, and many have involved small numbers of exposed subjects. Although a few studies have suggested potential adverse effects of some endocrine disruptors and solvents, the magnitude of associated risk has usually been modest and results have been far from conclusive (Brophy et al., 2012; Peplonska et al., 2010; Villeneuve et al., 2011). There has also been concern regarding potential effects on breast cancer risk of exposure to high levels of ethylene oxide among women employed in commercial sterilization facilities (Mikoczy et al., 2011).

A number of studies have focused on whether occupations that involve night shift work (a surrogate for exposure to night at light with subsequent melatonin suppression) could increase breast cancer risk. A recent meta-analysis of 28 studies (15 on shift work, 7 on short sleep duration, and 3 on flight attendants) found significantly elevated breast cancer risks among women employed in such occupations (RR = 1.14, 95% CI: 1.08–1.21) (He et al., 2015). Within selected case-control studies, a dose-response analysis showed that each 10-year increment of shift work was associated with a 16% elevation in the risk of breast cancer. Concern has also been expressed regarding potential effects of electromagnetic fields (EMFs) given their interference with the production of melatonin by the pineal gland. However, a large meta-analysis involving 15 studies (Chen et al., 2010) and a recent study among Chinese textile workers exposed to EMFs (Li et al., 2013c) failed to support any substantial risk alterations, even among heavily exposed subjects.

Other Environmental Factors

Extensive attention has focused recently on a wide variety of agents known as endocrine disruptors given their capacity to elicit hormonal responses. These include organochlorine pesticides and polycyclic aromatic hydrocarbons obtained through food intake or air pollution, as well as exposures to bisphenol A and phthalates contained in plastic containers. In addition, there are a variety of naturally occurring chemicals that mimic or inhibit endogenous hormones that have been of concern, including genistein, resveratrol, and zearolenone. Although this remains a controversial topic, the most recent investigations provide no or only limited evidence for detrimental effects on breast cancer risk (Ingber et al., 2013; Liu et al., 2015; Macon and Fenton, 2013). It is clear, however, that additional research on the topic will be undertaken given public concerns and the biologic plausibility of an association.

Other factors that have been suggested to affect risk include the use of hair dyes (Takkouche et al., 2005), bras (Chen et al., 2014), underarm deodorants (Mirick et al., 2002), and breast implants (Lipworth et al., 2009). Studies to date, however, have provided no conclusive evidence of any substantial link of risk with any of these exposures.

Host Factors

Endogenous Hormones

The evidence for a role of endogenous hormones in breast cancer etiology is compelling (see chapter 22 for further details) and may explain a number of identified risk factors. The below discussion focuses primarily on the epidemiologic evidence from prospective studies linking endogenous hormones to breast cancer risk.

Estrogens and Estrogen Metabolites

Prospective studies have consistently demonstrated that breast cancer risk increases with elevated levels of endogenous estrogens (Key et al., 2002, 2013). Pooled analyses of cohorts relating endogenous sex hormones to pre- and postmenopausal breast cancer risk were published in 2002 and 2013, respectively. These analyses estimated RRs of 2 and 1.4 in pre- and postmenopausal breast cancer risk, comparing women in the highest versus the lowest quintiles of circulating total estradiol. Among premenopausal women, the ORs for estrogens were larger for ER+ as compared to ER– tumors, but differences by ER status were not statistically significant. Results from more recent cohorts of postmenopausal breast cancer risk have been consistent with the 2002 pooled analysis, demonstrating about a 2- to 3-fold increase in risk comparing top to bottom quintiles (Hankinson and Eliassen, 2007). Furthermore, a single measure of estradiol was recently found (p.875) to predict risk of postmenopausal breast cancer for up to 16–20 years (Zhang et al., 2013). Among postmenopausal women, significant heterogeneity has been observed by HR status, with the strongest relations for estradiol being observed for HR+ tumors (Farhat et al., 2011; James et al., 2011; Zhang et al., 2013).

Endogenous estrogens are thought to increase breast cancer risk by either stimulating cell proliferation or by being metabolized to genotoxic and/or mutagenic metabolites (Yager and Davidson, 2006). The parent estrogens, estrone and estradiol, can be metabolized along three pathways via irreversible hydroxylation at the C-2, C-4, or C-16 position of the steroid ring. Relatively few epidemiologic studies have evaluated the effects of estrogen metabolites on breast cancer risk, in part because their comprehensive study has only recently become technically feasible with highly reproducible and sensitive liquid chromatography-tandem mass spectroscopy (LC-MS/MS) methods, which concurrently measure 15 estrogens and estrogen metabolites in serum (Fuhrman et al., 2014) and urine (Falk et al., 2008). Early epidemiologic studies used direct enzyme immunoassays to measure only two metabolites, 2-hydroxyestrone and 16α‎-hydroxyestrone, based on the hypothesis that an elevated ratio of 2-hydroxyestrone to 16α‎-hydroxyestrone is a biomarker of reduced breast cancer risk. Results from these early studies were generally inconsistent and not statistically significant (Ziegler et al., 2015). Four prospective studies using LC-MS/MS methods have demonstrated that increased levels of parent estrogens are consistently associated with increased risks of postmenopausal breast cancer (Dallal et al., 2014; Falk et al., 2013; Fuhrman et al., 2012; Moore et al., 2016). Each of these studies also suggested reductions in risk associated with elevated ratios of 2-pathway metabolites relative to parent estrogens. Another study has noted that urinary levels of 2- and 4-hydroxylation pathway metabolites are inversely, albeit not significantly, associated with premenopausal breast cancer risk (Eliassen et al., 2011). Additional studies are needed to address whether estrogen associations vary by hormone receptor status of the tumors.


After menopause, a major source of estrogen is adipose tissue, where aromatase activity results in the conversion of androgens to estrogens (Siiteri, 1987). Thus, androgens are hypothesized to influence breast cancer risk directly, by influencing cellular growth and proliferation, or indirectly, through their conversion to estrogens (Liao and Dickson, 2002). In pooled analyses, circulating levels of androgens have been found to be associated with significant elevations in pre- and postmenopausal breast cancer risk. Testosterone levels have been associated with RRs of 1.3 and 2.2 for pre- and postmenopausal women, respectively, comparing top to bottom quintiles (Key et al., 2002, 2013). Among premenopausal women, the RRs for androgens seem to be somewhat stronger for ER+ as compared to ER– tumors (Key et al., 2013). Among postmenopausal women, higher testosterone levels have been linked to elevated risks of HR+ breast cancers (Zhang et al., 2013); in contrast, there is some evidence to suggest that higher testosterone levels are associated with a lower risk of HR– breast cancer (Farhat et al., 2011; Zhang et al., 2013). In addition, adjustment for estradiol seems to attenuate risk estimates for testosterone, supporting the idea that the relation between androgens and postmenopausal breast cancer risk is due at least in part to the peripheral conversion of androgens to estrogens (Hankinson and Eliassen, 2007).

Estrogens and Androgens as Mediators of Other Risk Factors

In addition to assessing the effects of estrogens on breast cancer risk, a number of studies have attempted to assess the extent to which sex hormone concentrations might explain observed associations with anthropometric and lifestyle factors (Key et al., 2011; Shafrir et al., 2014). Specifically, higher levels of both endogenous estrogens and androgens have been observed among obese women (Key et al., 2011; Shafrir et al., 2014), smokers of 15+ cigarettes per day (Key et al., 2011), and alcohol drinkers of greater than 15–20 grams per day (Key et al., 2011; Shafrir et al., 2014), suggesting that sex hormones may mediate the effects of these factors on postmenopausal breast cancer risk. On the other hand, studies have not identified strong relationships between reproductive breast cancer risk factors and multiple circulating sex hormones (Key et al., 2011; Shafrir et al., 2014). Future work is needed to determine the association between sex hormones earlier in life with breast cancer risk factors.


Progesterone plays a key role in breast development, and experimental data suggest that progesterone metabolites may both decrease and increase breast cancer risk (Wiebe, 2006; Wiebe et al., 2010). In premenopausal women, the ovary is the primary source of progesterone, with smaller amounts produced in adipose tissue. Recent studies suggest that local production of progesterone metabolites may also occur in the breast (Wiebe and Lewis, 2003). However, population-based prospective studies of progesterone levels and their relation to pre- and postmenopausal breast cancer risk have been largely null (Key et al., 2013; Missmer et al., 2004). Prior studies have been limited to commercially available radioimmunoassays with inadequate sensitivity and poor specificity. In particular, low levels found in postmenopausal women may have limited the ability to detect modest associations for this hormone. Further, available commercial assays measure progesterone or 17-hydroxyprogesterone, and not other progesterone metabolites, which may also play a role in breast carcinogenesis. The recent development of an LC-MS/MS assay that measures concentrations of pregnenolone (a progesterone precursor), progesterone, and three progesterone metabolites may help to inform future work in this area (Trabert et al., 2015).


Prolactin also plays an important role in breast development and lactation (Clevenger et al., 2003). Though studies have provided conflicting results, two recent cohort studies have demonstrated a positive association between elevated prolactin levels and increased breast cancer risk, primarily in postmenopausal women (Tikk et al., 2014; Tworoger et al., 2013). In the NHS/NHSII studies, in which nearly 2900 cases were diagnosed over 20 years of follow-up and in which a sample of women provided two blood samples 10 years apart, there was an elevated risk for higher proximate (< 10 years before diagnosis) prolactin levels, but not for distant (10+ years) levels (Tworoger et al., 2013), a finding consistent with laboratory evidence that prolactin may influence tumor promotion (Clevenger et al., 2003). In addition, results were strongest for postmenopausal women and for ER+ breast cancer (Tworoger et al., 2013). Although the prolactin immunoassay measures all isoforms, which are thought to have varying biologic activity, a recent report from NHS/NHSII showed similar risk associations for prolactin levels measured by a sensitive bioassay (Tworoger et al., 2015), suggesting that future studies may use the less expensive and higher throughput immunoassay to accurately characterize breast cancer risk.

Anti-Müllerian Hormone

Anti-Müllerian hormone (AMH), also called Müllerian inhibiting substance (MIS), is a peptide hormone produced by small, growing follicles in the ovary (Visser et al., 2012). AMH levels are thought to peak in the mid-twenties and decline thereafter, becoming virtually undetectable about 5–6 years before menopause (Lie Fong et al., 2012; Sowers et al., 2008). To date, two prospective studies have demonstrated strong, positive associations between elevated AMH levels and increased breast cancer risk among premenopausal women (Dorgan et al., 2009; Nichols et al., 2015a). In light of conflicting evidence from laboratory-based studies, which have suggested an inverse relationship between AMH and breast tumor development (Gupta et al., 2005), additional studies exploring associations overall and by intrinsic breast cancer subtype are warranted.

Insulin/Insulin-like Growth Factors Axis


Insulin is a growth factor for a wide range of tissues, including the breast, and plays a significant role in normal breast development (Rosenfeld, (p.876) 2003). In addition, insulin increases the bioavailability of estrogen by down-regulating hepatic sex hormone–binding protein synthesis and can sensitize the ER to estrogen, thereby enhancing its activity. In the WHI, the largest study to date with prospectively collected specimens from fasting women not using MHT, high levels of insulin were associated with a 2-fold increase in breast cancer risk compared with the lowest level, independent of estradiol (Gunter et al., 2009). Smaller prospective cohort studies of C-peptide, a marker of insulin levels measured in non-fasting blood samples, have found mixed results; however, there is some consistency for a positive association among older women or women who are not taking MHT (Cust et al., 2009; Eliassen et al., 2007; Gaudet et al., 2013c; Verheus et al., 2006). These findings are consistent with a role of hyperinsulinemia in breast carcinogenesis.

Insulin-Like Growth Factors

Insulin-like growth factor (IGF)-I is a peptide hormone related to insulin and is also thought to play an important role in normal breast development and breast carcinogenesis (Pollak, 2008). In rodent models, the IGF system interacts with hormonal and non-hormonal pathways to influence breast development over time, including during puberty, pregnancy, and the post-lactational involution that accompanies weaning (Dearth et al., 2010; Rowzee et al., 2008; Su and Cheng, 2004). In women, increased levels of IGF-I are thought to increase breast cancer risk through the induction of cell proliferation and the inhibition of apoptosis in the human breast epithelium (Pollak, 2008). Associations between circulating IGF-I concentrations and breast cancer risk have been robust. A pooled analysis of data from approximately 20,000 women included in 17 prospective studies in 12 countries concluded that elevated IGF-I was significantly associated with a modest 1.3-fold increase in breast cancer risk that did not vary markedly by menopausal status (Breast Cancer Collaborative et al., 2010); the association was restricted to ER+ tumors. Although IGF-I bioavailability is modulated by IGF binding proteins (IGFBPs) (Pollak, 2008), in the pooled analysis the positive association between IGF-I and breast cancer risk was not substantially altered when adjusting for IGFBP-3 (Breast Cancer Collaborative Group et al., 2010).

Other Obesity-Related Biomarkers

Given the recognized importance of obesity in the etiology of breast cancer, numerous studies have attempted to assess the role of various mediating markers. This has included assessing circulating levels of IGFs and C-peptide (elaborated earlier) as well as adiponectin and C-reactive protein (CRP). Adiponectin levels have not been convincingly related to risk (Gaudet et al., 2013c; Gunter et al., 2015), but there is emerging evidence that high levels of CRP may be associated with elevated breast cancer risk (Gaudet et al., 2013c; Gunter et al., 2015; Wang et al., 2015). A recent meta-analysis showed a modestly increased risk among women in the highest versus lowest quintile of CRP (RR = 1.26, 95% CI: 1.07–1.49) (Wang et al., 2015). Given that CRP is a non-specific marker of inflammation, additional studies employing more precise biomarkers of inflammation should be pursued.

Germline Genetic Susceptibility

The breast cancer susceptibility genes BRCA1 and BRCA2 contain rare mutations (minor allele frequency [MAF] < 0.01) that confer a very high risk of breast cancer (RRs of >4 for carriers versus non-carriers). By age 70, women with a BRCA1 mutation have a 65% (95% CI: 44%–78%) increased risk of breast cancer, whereas those with a BRCA2 mutation have a 45% (95% CI: 31%–56%) risk (Antoniou et al., 2003). Additional genetic variants with MAF ranging from 0.005–0.05 and more moderate breast cancer risk (RR of 2–4) have been identified for checkpoint kinase 2 (CHEK2), ataxia telangiectasia (ATM), and neurofibromatosis (NF1), among others. PALB2 (partner and localizer of BRCA2), phosphatase and tensin homolog 2 (PTEN), seminal threonine kinase 11 (STK11), and cadherin 1 (CDH1) also contain genetic variants that are associated with moderate or high risk, but additional large studies are necessary to precisely estimate their penetrance (Easton et al., 2015). Combined, these rare, highly penetrant mutations and uncommon, moderately penetrant genetic variants account for only ~35% of all familial aggregation of breast cancer risk and less than 10% of breast cancers in the general population (Easton, 1999; Melchor and Benitez, 2013). The remaining breast cancer heritability is likely explained by unknown rare, highly penetrant mutations together with a polygenetic component (Antoniou et al., 2002).

Genome-wide association studies (GWAS) of breast cancer risk have led to the identification of 167 genome-wide significant breast cancer susceptibility loci in populations predominantly of European origin (Oncoarray, 2016; Michailidou et al., 2017). Of the identified loci, the strongest and most consistent association with breast cancer risk is for the minor allele of rs2981582 (MAF = 0.38, RR = 1.26, 95% CI: 1.23, 1.30, p-value = 2 × 10–76), located in intron 2 of the fibroblast growth factor receptor 2 (FGFR2) gene. The magnitude of the associations for the other identified breast cancer susceptibility loci has ranged from 1.03 to 1.26 and 0.76 to 0.97 with MAF as low as 3% in women of European ancestry (Oncoarray, 2016). In populations of different genetic ancestry, single nucleotide polymorphisms (SNPs) that best represent the risk association and the MAF of the SNPs might differ due to differences in genetic architecture; however, the majority of SNPs in black and Asian populations have associations in the same direction as previous reports (Chen et al., 2011; Feng et al., 2014; Zheng et al., 2013). Populations with different genetic ancestry also are informative for fine mapping to identify better markers of the biologically functional alleles and for identification of new loci (Cai et al., 2014; Chen et al., 2011; Feng et al., 2014).

Etiologic heterogeneity clearly exists for associations of known variants and different molecular subtypes of breast cancer. Approximately a third of the variants have stronger associations with ER+ than ER– tumors (e.g., FGFR2), while another third of breast cancer susceptibility alleles have only been associated with ER+ breast cancers (e.g., COX11). The GWAS conducted in mutation carriers of BRCA1, who are more likely to be diagnosed with ER– breast cancer, demonstrated the potential to uncover additional loci (e.g., 19p13) for ER– breast cancer (Antoniou et al., 2010).

Fine mapping and understanding the genetic mechanisms and molecular pathways of the identified loci affecting breast carcinogenesis is an active area of research (Fachal and Dunning, 2015). Thus far, the implicated pathways include DNA damage recognition and repair, apoptosis, ER signaling, formation of mammary glands, telomere length, insulin signaling, tumor progression and metastatic disease, and epigenetic changes (Fachal and Dunning, 2015). Genetic variation associated with breast cancer risk factors and intermediate markers of risk (see later discussion), including TDLU involution and mammographic density, might also be informative to determine functionality of loci (Stone et al., 2015). Identification of gene–environment interactions could provide mechanistic actions of susceptibility loci, yet to date no gene–environment interactions have been confirmed.

These common SNPs explain an additional 18% of familial aggregation of breast cancer (Michailidou et al., 2017). It is estimated that another ~10% of familial risk can be explained by more than 1000 loci that have yet to be identified (Michailidou et al., 2013). The characterization of the remaining breast cancer heritability will require the continued collaborative efforts of consortia to capture common variants (with very small effect sizes) and uncommon variants (with low MAF), to conduct genome- and exome-wide DNA sequencing and pathway-based approaches, and to identify gene-gene and gene-environment interactions.

Intermediate Markers of Risk

Mammographic Density

Mammographic density reflects the tissue composition of the breast: adipose tissue is radiolucent and fibroglandular tissue is dense. Although high mammographic density is related to some breast cancer risk factors such as nulliparity, a positive family history of breast cancer, and MHT use, numerous studies over the past four decades have consistently demonstrated that high mammographic density is a strong and independent breast cancer risk factor, conferring RRs of 4- to 5-fold when comparing women with high versus low mammographic (p.877) density (McCormack and dos Santos Silva, 2006; Santen and Mansel, 2005) (Figure 45–6). Various methods have been developed to measure mammographic density, including visual examination of mammograms to estimate dense tissue in the clinic (e.g., the American College of Radiology Breast Imaging-Reporting and Data System [BI-RADS]) (Sickles et al., 2013) or quantitative computer-assisted methods, which are more reliable (Yaffe, 2008). In a systematic review and meta-analysis of 42 studies, high mammographic density was a strong indicator of breast cancer risk irrespective of the modality used to measure mammographic density (McCormack and dos Santos Silva, 2006).

Breast Cancer

Figure 45–6. Risk of breast cancer according to percent mammographic breast density. Shown on the y-axis are the combined relative risks (95% confidence intervals) from a meta-analysis of percent breast density and its association with incident breast cancer (McCormack and dos Santos Silva, 2006). Representative mammograms of varying degrees of percent breast density from participants in the BREAST Stamp Project are shown (Gierach et al., 2014).

Adapted from Santen and Mansel (2005).

Dense areas on a mammogram, which appear white, may be associated with delays in breast cancer diagnosis due to non-visualization of tumors (which also appear white); however, high mammographic density is related to long-term prospective increases in tumor incidence, independent of its effects on detection (McCormack and dos Santos Silva, 2006). Intuitively, one might hypothesize that the absolute amount of dense tissue is driving the risk association. However, numerous studies have demonstrated that the proportion of dense tissue relative to total breast area (i.e., percent mammographic density) is also strongly predictive of risk, implying that characteristics of non-dense tissue are also important (Pettersson et al., 2014). Understanding factors that affect mammographic density and their underlying mechanisms are important, yet understudied, research questions.

Proposed biologic explanations for the strong positive relationship between mammographic density and breast cancer risk include hypotheses that mammographic density reflects the cumulative effects of breast cancer risk factors, differences in the number of cells that are susceptible to malignant change, and characteristics of the microenvironment of dense breast tissue that are more conducive to carcinogenesis (Martin and Boyd, 2008). A recently published conceptual model adapted from Martin and Boyd (2008) illustrated potential mechanisms linking genetic and non-genetic risk factors to mammographic density and breast cancer risk (Sun et al., 2013). Epidemiologic factors strongly associated with mammographic density explain only ~20%–30% of the variance in mammographic density, whereas data from twin studies suggest that up to 67% of the variation in density may be attributed to common genetic factors (Boyd et al., 2002; Ursin et al., 2009). Findings from ongoing consortial efforts aimed at identifying genes associated with density phenotypes (Brand et al., 2015; Lindstrom et al., 2011, 2014; Stone et al., 2015; Vachon et al., 2015, 2012) may serve as a powerful reflection of the cumulative interplay of numerous genetic and environmental breast cancer risk factors over time.

Morphometric studies of breast tissue obtained from biopsies, mastectomies, and autopsies have found that high mammographic density is negatively associated with fat and positively associated with epithelium and stroma (Ginsburg et al., 2008; Sun et al., 2013). Higher mammographic density is also associated with a greater proportion of TDLUs (Ghosh et al., 2010a; Gierach et al., 2015), the specific microanatomic structures from which nearly all breast cancers arise (Russo et al., 2000). As women age, fibroglandular (epithelial and stromal) tissue comprises a declining percentage of total breast tissue, whereas the percentage of fat increases (Gertig et al., 1999; Ginsburg et al., 2008; Sun et al., 2014). Persistence of elevated density and TDLUs with aging may be an indicator of elevated risk (see later discussion for further details regarding the interplay between TDLU involution and mammographic density). Little is known about the relation between mammographic density and protein expression in breast tissue, including the role of growth factors or stromal matrix proteins, although it has been suggested that protein expression of tissue inhibitor metalloprotease 3 and IGF-I in breast tissue may be positively related to mammographic density (Guo et al., 2001). A recent analysis relating gene expression profiling signatures of normal tissues, adjacent to the tumors, indicated that mammographic density reflects broad transcriptional changes, including changes in both epithelia- and stroma-derived signaling (Sun et al., 2013).

Although the mechanisms by which mammographic density influences breast cancer risk are unknown, mammographic density is hypothesized to reflect cumulative exposure to some endogenous growth factors and hormones. Circulating IGF-I is a promising etiological marker, as it is related to increased cellular proliferation and has been positively associated in several studies with both mammographic density (Martin and Boyd, 2008) and breast cancer risk (Breast Cancer Collaborative et al., 2010). In addition, mammographic density appears to be modified by exogenous hormonal agents and increases with exposure to MHT (Boyd et al., 2011). Conversely, mammographic density decreases with exposure to tamoxifen, a selective estrogen receptor modulator (SERM), in both high-risk women and breast cancer patients (Cuzick et al., 2011; Kim et al., 2012; Ko et al., 2013; Li et al., 2013b; Nyante et al., 2015). Accordingly, hormonal mechanisms may mediate both mammographic density and mammographic density–related breast cancer risk, although data relating endogenous estrogen levels to mammographic density are inconsistent (Becker and Kaaks, 2009; Martin and Boyd, 2008).

Mammographically dense breasts are clinically important because women with such breasts are more likely to be diagnosed with interval cancers that are missed by routine mammography (Boyd et al., 2007). Furthermore, tumors that develop in dense as compared with non-dense breasts are larger and more frequently node positive (Yaghjyan (p.878) et al., 2011). Public awareness and concerns about mammographic density have recently risen to the forefront as a growing number of US states have passed laws requiring radiologists to inform patients about their mammographic density (Price et al., 2013). High mammographic density seems to increase risk for both ER+ and ER– cancers (Bertrand et al., 2015; Huo et al., 2014). Therefore, understanding the association between elevated mammographic density and breast cancer provides an opportunity to increase our knowledge of the etiology of ER-negative cancers. In addition to its association with incidence, several studies have found that high mammographic density is a risk factor for breast cancer recurrence (Huo et al., 2014). Data also suggest that increases and decreases in mammographic density relate to corresponding changes in breast cancer risk (Cuzick et al., 2011; Kerlikowske et al., 2007; Work et al., 2014). Taken together, these results suggest that mammographic density might represent a causal intermediate in carcinogenesis; therefore, lowering density could reduce breast cancer risk.

Terminal Duct Lobular Unit Involution

As described earlier, TDLUs are the anatomical structures of the breast from which most breast carcinomas originate (Russo et al., 2000), representing the at-risk epithelium. TDLU involution, a normal process of aging, is characterized by a reduction in the number and size of TDLUs and their secretory substructures, called acini (Russo et al., 2000). Like mammographic density, TDLU involution is thought to reflect the cumulative effects of breast cancer risk factors across the life course (Figueroa et al., 2014); this hypothesis is closely aligned with the concept of “breast tissue aging” as proposed by Pike et al. over three decades ago (Pike et al., 1983). As mentioned earlier, studies of two large cohorts of women with benign breast disease (BBD) have found that women with reduced TDLU involution are at increased breast cancer risk (Baer et al., 2009; Figueroa et al., 2016; Milanese et al., 2006). Data from the Mayo BBD Cohort also suggest that both elevated mammographic density and benign breast tissue demonstrating reduced TDLU involution are independent risk factors for the development of breast cancer (Ghosh et al., 2010b). Furthermore, having a combination of no involution and dense breasts was associated with even greater breast cancer risk (RR = 4.08, 95% CI: 1.72–9.68). These findings indicate that visual representations of the breast tissue architecture on both microscopic (i.e., involution) and macroscopic (i.e., mammographic density) scales potentially represent clinically useful intermediate endpoints (Gierach et al., 2010).

Identification of factors that are associated with TDLU involution may reveal underlying biological pathways related to breast cancer risk. Recent studies have suggested a role in TDLU involution of circulating IGFs (Horne et al., 2016) and sex steroid hormones (Khodr et al., 2014; Oh et al. 2016). These findings suggest that elevated levels of endogenous hormones and growth factors may increase breastwcancer risk potentially through delaying TDLU involution. Future GWAS of TDLU involution may provide clues into genetic determinants that may influence histologic changes in the breast that could also put women at higher risk of breast cancer.

Assessing lobules in thin histological sections of human breast tissues is challenging and requires expertise in breast histomorphometry. To date, findings relating TDLU involution to breast cancer risk have largely been based on non-quantitative measures of involution, in which the extent of involution was classified visually by a breast pathologist (Baer et al., 2009; Ghosh et al., 2010b; Milanese et al., 2006). In a recent report, objective and reproducible quantitative measures of TDLUs that are inversely associated with TDLU involution (i.e., TDLU count, median TDLU span, and median acini count per TDLU) were significantly related to breast cancer risk factors, including age, menopause, and parity, among women who donated normal breast tissues (Figueroa et al., 2014) (Figure 45–7). These same standardized measures also predicted breast cancer risk when applied to the Mayo BBD Cohort (Figueroa et al., 2016). In the future, computer-assisted image analysis methods (Beck et al., 2011; McKian et al., 2009; Rosebrock et al., 2013) may allow for more precise and high-throughput measurement of microscopic features in breast tissues for use as intermediate endpoints in studies of breast cancer etiology, prevention, and progression.

Breast Cancer

Figure 45–7. Terminal duct lobular unit. (TDLU) involution assessment. Quantitative measures associated with reduced levels of TDLU involution were assessed from digitized images of H&E stained tissue sections. (a) A digital H&E section with multiple TDLUs (0.75×). (b) Representative TDLUs for which the longest TDLU span was measured in microns using a digital ruler (4.27×). A representative acinus was circled and indicated with an arrow.

Source: Figueroa et al. (2014).

Risk Prediction

Over the past three decades, numerous statistical models have been developed to assess breast cancer risk in populations over time (i.e., 5-year, 10-year, or lifetime risks) (Amir et al., 2010). Risk prediction models have utility in planning intervention trials, estimating population burden of disease, clinical decision-making and creating risk–benefit indices, targeted screening, and deciding on risk-reduction strategies. One of the most widely studied risk prediction models, the NCI Breast Cancer Risk Assessment Tool (“Gail model”), includes a combination of known risk factors: age, age at menarche, age at first live birth, number of prior breast biopsies, history of atypical hyperplasia, and number of first-degree relatives with breast cancer (Costantino et al., 1999). The Gail model has been tested in large populations of white women and has been shown to provide accurate estimates of breast cancer risk (Costantino et al., 1999; Rockhill et al., 2001). This model has also been validated for Asian and Pacific Islander women (Matsuno et al., 2011) and performs well in black women, though it may underestimate risk among black women with prior breast biopsies (Gail et al., 2007). Further validation of the model for Hispanic women and other racial/ethnic populations is needed https://www.ncbi.nlm.nih.gov/pubmed/28003316. In addition, recent studies suggest that the Gail model may have limited discriminatory accuracy, particularly for higher-risk populations, such as women enrolled in family history clinics, perhaps because the model does not capture age at cancer onset among first-degree relatives or second-degree family (p.879) histories (Amir et al., 2010; Howell et al., 2014). The Tyrer-Cuzick International Breast Cancer Intervention Study (IBIS) model (Tyrer et al., 2004), which includes more detailed family history and MHT use, among other risk factors, performs well among women considered to be at average or above-average risk (Quante et al., 2012).

Although existing models provide accurate estimates of lifetime risk at the population level, individualized risk prediction is poor (Amir et al., 2010). A commonly used measure of discriminatory power is the concordance (c)-statistic, which represents the area under the receiver operating characteristics curve. C-statistics for most breast cancer risk models hover around 0.6, indicating that the risk prediction model is accurate only 60% of the time (Cummings et al., 2009). Thus, there is interest in adding risk factors to the current models, including endogenous hormone levels (Tworoger et al., 2014), proliferative benign biopsy diagnoses (Tice et al., 2015), lifestyle factors, mammographic density, and genetic polymorphisms (described in further detail later), in the hope of improving their discriminatory accuracy. For example, several studies (Barlow et al., 2006; Cecchini et al., 2012; Chen et al., 2006; Tice et al., 2005) have found that adding mammographic density to the Gail model improves breast cancer risk prediction modestly, and efforts to incorporate mammographic density in newer risk models are ongoing (Tice et al., 2008, 2015; Vachon et al., 2015; Warwick et al., 2014). The minimal improvement in risk prediction with the addition of mammographic density has been consistent across study populations. However, the potential gains in risk prediction that might be realized by using automated, quantitative measures of density obtained through full-field digital mammography or other emerging technologies have not been fully explored (Gierach et al., 2013). In addition, studies characterizing mammographic density and breast cancer risk in non-white populations are limited. Finally, elevated mammographic density may produce its strongest effect among young women who are below the age of initiation of mammographic screening, but who might benefit from preventive interventions (Boyd et al., 2007). Evaluating density without exposing young women to ionizing radiation is critical, and these approaches have not yet been implemented in clinical practice.

The contribution of genetic polymorphisms to risk prediction models continues to evolve. The polygenetic model of genetic susceptibility assumes a wide distribution of risk in the population (Ponder et al., 2005). Initially, the addition of seven genome-wide significant genetic polymorphisms to the Gail model only slightly improved its performance, although the improved discriminatory accuracy was less than that of the addition of mammographic density (Gail, 2008). As additional genome-wide significant genetic variants are identified, it is expected that an aggregate polygenetic risk score of these variants will improve the performance of risk prediction models (Garcia-Closas et al., 2014). In contrast, screening of high-risk families for known mutations responsible for autosomal dominant familial syndromes has become an integral component of the practice of preventive oncology (Weitzel et al., 2011). For instance, risk assessment tools for breast and ovarian cancer (e.g., BOADICEA and BRCAPRO) have been developed to predict BRCA1/2 mutation carrier probability (Fischer et al., 2013). Women with a strong family history of cancer should be offered counseling to determine if genetic testing is appropriate.

Opportunities for Prevention

Lifestyle Modification

The identification of potentially modifiable established or suspected risk factors for breast cancer, such as alcohol consumption, cigarette smoking, physical activity, and certain anthropometric factors, provides opportunities for risk-reducing interventions among women at both average and high risk for breast cancer. Other established modifiable risk factors reviewed previously in this chapter include ionizing radiation exposure, combination MHT use, and possibly mammographic density. With respect to lifestyle factors, in a recent report on postmenopausal participants in EPIC, researchers developed a Healthy Lifestyle Index constructed from five modifiable factors (diet, physical activity, smoking, alcohol consumption, and anthropometry) and estimated that with each point added to a woman’s Healthy Lifestyle Index score, breast cancer risk fell by 3% (McKenzie et al., 2015). Although risk estimates associated with modifiable factors are modest in magnitude, even small changes at the individual level could result in substantial changes in breast cancer incidence at the population level (Colditz and Bohlke, 2014).

In 2012, the American Cancer Society (ACS) published nutrition and physical activity guidelines based on expert consensus and scientific evidence (Kushi et al., 2012). To reduce breast cancer risk, the ACS offered the following recommendations: “to engage in regular, intentional physical activity; to minimize lifetime weight gain through the combination of caloric restriction (in part by consuming a diet rich in vegetables and fruits) and regular physical activity; and to avoid or limit intake of alcoholic beverages” (Kushi et al., 2012). The WCRF/AICR expert report provides similar recommendations for breast cancer prevention (Norat et al., 2014). In the WHI, women who most closely followed the ACS cancer prevention guidelines had a significantly 22% lower risk of breast cancer as compared with women who had the lowest adherence (Thomson et al., 2014). Public health programs helping women to maintain and engage in healthy behaviors could aid in breast cancer prevention. Furthermore, emerging evidence that early life exposures may influence a woman’s lifetime risk of breast cancer suggests that breast cancer prevention efforts may have the greatest impact when initiated at a young age and continued over the life span (Colditz and Bohlke, 2014). The ACS guidelines recognize that individual choices occur within a community context and thus offer strategies for the introduction of cancer prevention guidelines in the community, which will require action at local, state, and national levels (Kushi et al., 2012).

Although it is not necessarily feasible or desirable to alter reproductive behaviors related to breast cancer risk, such as the timing and number of pregnancies, breastfeeding should be encouraged for many reasons, including its association with reduced risk of more aggressive molecular tumor subtypes (described earlier in this chapter). Breastfeeding may prove to be an effective prevention strategy, particularly for black women, who tend to have higher rates of triple negative breast cancer and lower rates of breastfeeding (Palmer et al., 2014).


Among women who are at increased risk of breast cancer, SERMs have been demonstrated to effectively reduce breast cancer risk (Chlebowski, 2014). A recent meta-analysis of nine randomized prevention trials comparing four SERMs with placebo (or with tamoxifen in one study) found that the overall reduction in risk of in situ or invasive breast cancer was 38% (Cuzick et al., 2013). Thromboembolic events were significantly increased with all SERMs (Cuzick et al., 2013). Randomized trials of the SERM tamoxifen showed a 44% reduction in the incidence of ER+ invasive breast cancer (Cuzick et al., 2013), and this protective effect persisted with extended follow-up after cessation of therapy (Cuzick et al., 2013, 2015). In the Study of Tamoxifen and Raloxifen (STAR) trial, tamoxifen was significantly superior to the SERM raloxifene in preventing invasive breast cancer with long-term follow-up; however, raloxifene was associated with fewer side effects (Vogel et al., 2010). Statistically significant reductions in breast cancer risk have also been observed for aromatase inhibitors (Cuzick et al., 2014; Goss et al., 2011), which are only suitable for use in postmenopausal women and have not been approved by the US Food and Drug Administration (FDA) for breast cancer prevention.

In 2013, both the US Preventive Services Task Force (USPSTF) (Moyer and Force, 2013) and the American Society of Clinical Oncology (ASCO) (Visvanathan et al., 2013) issued updated guidelines regarding breast cancer chemoprevention. The USPSTF recommended that clinicians engage in shared, informed decision-making with women at elevated breast cancer risk regarding chemoprevention (Moyer and Force, 2013). For women 35 years and older who are at elevated risk for breast cancer and at low risk for adverse medication effects (Freedman et al., 2011), the USPSTF guidelines state that clinicians should offer to prescribe risk-reducing medications such as tamoxifen or raloxifene. For women who are not at elevated breast (p.880) cancer risk, the USPSTF recommended against routine chemoprevention (Moyer and Force, 2013). The ASCO guidelines state that risk reduction medication “should be discussed as an option” for women 35 years and older who are at increased breast cancer risk, defined as a 5-year risk > 1.66% or those with LCIS (Visvanathan et al., 2013).

Despite the strong scientific evidence showing the clinical effectiveness of these risk-reducing medications, population-based studies indicate that the uptake of SERMs remains low in the United States, largely due to concerns about treatment side effects and a perceived unfavorable balance between benefits and harms (Bambhroliya et al., 2015; Cuzick et al., 2013). Furthermore, among those deciding to take tamoxifen for primary prevention, substantial non-adherence to the recommended 5-year course of tamoxifen has been reported (Nichols et al., 2015b). Thus, identifying early predictors of tamoxifen response would have particular value, providing encouragement for responders to continue treatment and indicating the need to consider alternatives among non-responders. Data from the IBIS-I chemoprevention trial suggest that declines in mammographic density after 12–18 months of tamoxifen therapy are an excellent predictor of response in the prevention setting (Cuzick et al., 2011). Further evaluation of mammographic density decline as a surrogate marker for other preventive interventions, along with increased understanding of the mechanisms associated with density change, may advance breast cancer prevention efforts. In addition, interventions that reduce mammographic density may in turn improve sensitivity of screening mammography.

With any preventive intervention, careful consideration of the risks and benefits is needed to identify women most likely to benefit. It is estimated that over half of all breast cancers could be prevented through a combination of healthy behaviors and chemoprevention (Colditz and Bohlke, 2014).

Screening and Other Preventive Strategies

In addition to chemoprevention and screening for earlier disease detection (discussed in Chapter 63), prophylactic surgery is increasingly being considered as a cancer prevention strategy. In high-risk women and BRCA1/2 mutation carriers, prophylactic bilateral mastectomy has been shown to reduce the incidence of breast cancer by approximately 85%–100% and mortality by 81%–100% (Nelson et al., 2014). In such women, bilateral oophorectomies have also been shown to reduce breast cancer risk by approximately 50%. Contralateral mastectomies are also increasingly being performed among women diagnosed with unilateral breast cancer given demonstrated breast cancer risk reductions; however, data regarding improvements in disease-specific survival are inconsistent, possibly owing to incomplete control for confounding factors (Davies et al., 2015). Alternate risk-reducing strategies such as adjuvant endocrine therapy, which has been shown to reduce not only contralateral breast cancer risk but also recurrence and death (Early Breast Cancer Trialists’ Collaborative Group et al., 2011; Gierach et al., 2017), will be important to consider as clinicians engage in shared, informed decision-making with patients regarding the best course of treatment for unilateral breast cancer.

Future Research Directions

Although slightly more than half of all breast cancers can be explained on the basis of identified risk factors (Pfeiffer et al., 2013), there remain many breast cancers with undefined causes. This may reflect failures to identify all etiologic factors, to assess interactive effects among breast cancer risk factors, to completely understand the biologic intricacies of these factors, or to appreciate differences across breast cancer subtypes. Additional research is needed along all such avenues.

Given the extensive insights that have been gained in recent times regarding the heterogeneity of breast cancers, it is also clear that a major emphasis in future research will be on reliable classification of distinct tumor subtypes and for evaluating risk factors for these subtypes. With more consistent classification of molecular markers such as HER2, it will be important to further assess tumor subgroups defined by this and other markers. Further, basal-like tumors appear especially unique as compared to the other tumors and recent efforts are confirming this through molecular marker studies. Finally, projects such as The Cancer Genome Atlas (TCGA) that are using gene expression techniques to classify tumors into newly defined subgroups will likely provide understandings of etiologic, clinical, and prognostic diversity of breast cancers that go beyond current classification schemes.

Another area of major interest is in identifying windows of particular susceptibility for breast cancer risk. This includes factors that affect the initial development of breast cells (including hormonal exposures received in utero or immediately postnatally), as well as factors that may have special effects on less differentiated breast cells that are found prior to menarche, during adolescence, or prior to a first pregnancy, including cigarette smoking, alcohol consumption, and body size. There may also be exposures that operate differently within the context of altered endogenous hormones; for instance, there are suggestions that the relations of MHT use may depend on a woman’s body size, as well as when they are initiated with respect to the menopausal transition. Other factors may exhibit similar complex interactions, although identification of such patterns will undoubtedly be dependent on having sufficiently large and long-term studies to detect combined effects.

Recent studies that have focused on intermediates of breast cancer risk, such as mammographic densities and TDLUs, have provided clues to the origins of breast cancer, allowing us to better understand how risk factors operate at early stages and how they are mediated by a variety of biologic parameters. Although we have derived some clues as to the underlying biologic mechanisms for these cellular changes, it is clear that future continued research efforts are needed, including expanding on the biologic markers studied, assessing interactive effects (e.g., interactive effects of hormonal and immunologic markers), and identifying markers that may be useful for incorporation into new risk-prediction models and primary prevention and screening strategies.

In addition to better understanding biologic processes in breast cancer, it will also be important to understand the global distribution of breast cancer, given that rates are changing quite rapidly in many parts of the world. The types of breast cancers occurring in various regions may be distinct, in part due to different lifestyle factors, but also due to genetic profiles and interactions of genetic and environmental factors. Thus, an appreciation of the diversity of tumors and risk factor profiles within different regions will be important for developing effective and sustainable prevention strategies and for assuring that treatment modalities are best designed to meet changing needs. It will also be essential, as treatment programs become better developed in such regions, that more attention focus on survivorship issues; this will be equally important in developed countries given the relatively good prognosis of breast tumors when diagnosed early and the fact that many women face years of living with the repercussions of having been diagnosed with breast cancers.


Bibliography references:

Ades F, Zardavas D, Bozovic-Spasojevic I, et al. 2014. Luminal B breast cancer: molecular characterization, clinical management, and future perspectives. J Clin Oncol, 32(25), 2794–2803. PMID: 25049332.

Ahlgren M, Melbye M, Wohlfahrt J, and Sorensen TI. 2004. Growth patterns and the risk of breast cancer in women. N Engl J Med, 351(16), 1619–1626. PMID: 15483280.

Alexander DD, Morimoto LM, Mink PJ, and Lowe KA. 2010. Summary and meta-analysis of prospective studies of animal fat intake and breast cancer. Nutr Res Rev, 23(1), 169–179. PMID: 20181297.

Ali S, and Coombes RC, 2002. Endocrine-responsive breast cancer and strategies for combating resistance. Nat Rev Cancer, 2(2), 101–112. PMID: 12635173.

Allred DC. 2011. Molecular biomarkers of risk in premalignancy and breast cancer prevention. Cancer Prev Res (Phila), 4(12), 1947–1952. PMID: 22144468.

Amadou A, Ferrari P, Muwonge R, et al. 2013. Overweight, obesity and risk of premenopausal breast cancer according to ethnicity: a systematic review and dose-response meta-analysis. Obes Rev, 14(8), 665–678. PMID: 23615120.

Ambrosone CB, Zirpoli G, Hong CC, et al. 2015. Important role of menarche in development of estrogen receptor-negative breast cancer in African American women. J Natl Cancer Inst, 107(9), djv172. PMID: 26085483.

American Cancer Society. 2014. Breast Cancer Facts and Figures 2013–2014.

(p.881) Amir E, Freedman OC, Seruga B, and Evans DG. 2010. Assessing women at high risk of breast cancer: a review of risk assessment models. J Natl Cancer Inst, 102(10), 680–691. PMID: 20427433.

Anderson WF, Rosenberg PS, Prat A, Perou CM, and Sherman ME. 2014. How many etiological subtypes of breast cancer: two, three, four, or more? J Natl Cancer Inst, 106(8). PMCID: PMC4148600.

Antoniou A, Pharoah PD, Narod S, et al. 2003. Average risks of breast and ovarian cancer associated with BRCA1 or BRCA2 mutations detected in case Series unselected for family history: a combined analysis of 22 studies. Am J Hum Genet, 72(5), 1117–1130. PMCID: PMC1180265.

Antoniou AC, Pharoah PD, McMullan G, et al. 2002. A comprehensive model for familial breast cancer incorporating BRCA1, BRCA2 and other genes. Br J Cancer, 86(1), 76–83. PMCID: PMC2746531.

Antoniou AC, Wang X, Fredericksen ZS, et al. 2010. A locus on 19p13 modifies risk of breast cancer in BRCA1 mutation carriers and is associated with hormone receptor-negative breast cancer in the general population. Nat Genet, 42(10), 885–892. PMCID: PMC3130795.

Arpino G, Milano M, and De Placido S. 2015. Features of aggressive breast cancer. Breast, 24(5), 594-600. PMID: 26144637.

Baer HJ, Collins LC, Connolly JL, et al. 2009. Lobule type and subsequent breast cancer risk: results from the Nurses’ Health Studies. Cancer, 115(7), 1404–1411. PMCID: PMC2661011.

Bakken K, Fournier A, Lund E, et al. 2011. Menopausal hormone therapy and breast cancer risk: impact of different treatments. The European Prospective Investigation into Cancer and Nutrition. Int J Cancer, 128(1), 144–156. PMID: 20232395.

Bambhroliya A, Chavez-MacGregor M, and Brewster AM. 2015. Barriers to the use of breast cancer risk reduction therapies. J Natl Comp Cancer Network, 13(7), 927–935. PMID: 26150584.

Bardia A, Keenan TE, Ebbert JO, et al. 2016. Personalizing aspirin use for targeted breast cancer chemoprevention in postmenopausal women. Mayo Clin Proc, 91(1), 71–80. PMCID: PMC4807132.

Barlow WE, White E, Ballard-Barbash R, et al. 2006. Prospective breast cancer risk prediction model for women undergoing screening mammography. J Natl Cancer Inst, 98(17), 1204–1214. PMID: 16954473.

Barnard ME, Boeke CE, and Tamimi RM. 2015. Established breast cancer risk factors and risk of intrinsic tumor subtypes. Biochim Biophys Acta, 1856(1), 73–85. PMID: 26071880.

Barry JA, Azizia MM, and Hardiman PJ. 2014. Risk of endometrial, ovarian and breast cancer in women with polycystic ovary syndrome: a systematic review and meta-analysis. Hum Reprod Update, 20(5), 748–758. PMCID: PMC4326303.

Beaber EF, Buist DS, Barlow WE, et al. 2014a. Recent oral contraceptive use by formulation and breast cancer risk among women 20 to 49 years of age. Cancer Res, 74(15), 4078–4089. PMCID: PMC4154499.

Beaber EF, Malone KE, Tang MT, et al. 2014b. Oral contraceptives and breast cancer risk overall and by molecular subtype among young women. Cancer Epidemiol Biomarkers Prev, 23(5), 755–764. PMCID: PMC4032363.

Beck AH, Sangoi AR, Leung S, et al. 2011. Systematic analysis of breast cancer morphology uncovers stromal features associated with survival. Sci Transl Med, 3(108), 108ra113. PMID: 22072638.

Becker S, and Kaaks R. 2009. Exogenous and endogenous hormones, mammographic density and breast cancer risk: can mammographic density be considered an intermediate marker of risk? Recent Results Cancer Res, 181, 135–157. PMID: 19213565.

Begg C, and Zhang Z. 1994. Statistical analysis of molecular epidemiology studies employing case-series. Cancer Epidemiol Biomarkers Prev, 3(2), 173–175. PMID: 8049640.

Beral V, Reeves G, Bull D, Green J, and Million Women Study C. 2011. Breast cancer risk in relation to the interval between menopause and starting hormone therapy. J Natl Cancer Inst, 103(4), 296–305. PMCID: PMC3039726.

Bernstein JL, Lapinski RH, Thakore SS, Doucette JT, and Thompson WD. 2003. The descriptive epidemiology of second primary breast cancer. Epidemiology, 14(5), 552–558. PMID: 14501270.

Bertrand KA, Scott CG, Tamimi RM, et al. 2015. Dense and nondense mammographic area and risk of breast cancer by age and tumor characteristics. Cancer Epidemiol Biomarkers Prev, 24(5), 798–809. PMCID: PMC4417380.

Bethea TN, Rosenberg L, Hong CC et al. 2015. A case-control analysis of oral contraceptive use and breast cancer subtypes in the African American Breast Cancer Epidemiology and Risk Consortium. Breast Cancer Res, 17, 22. PMCID: PMC4358874.

Bombonati A, and Sgroi DC. 2011. The molecular pathology of breast cancer progression. J Pathol, 223(2), 307–317. PMCID: PMC3069504.

Boyd N, Dite G, Stone J, et al. 2002. Heritability of mammographic density, a risk factor for breast cancer. N Engl J Med, 347, 886–894. PMID: 12239257.

Boyd N, Guo H, Martin L, et al. 2007. Mammographic density and the risk and detection of breast cancer. N Engl J Med, 356(3), 227-236. PMID: 17229950.

Boyd NF, Melnichouk O, Martin LJ, et al. 2011. Mammographic density, response to hormones, and breast cancer risk. J Clin Oncol, 29(22), 2985–2992. PMID: 21709206.

Brand JS, Li J, Humphreys K, et al. 2015. Identification of two novel mammographic density loci at 6Q25.1. Breast Cancer Res, 17(1), 75. PMCID: PMC4501298.

Breast Cancer Collaborative Group, Key TJ, Appleby PN, Reeves GK, and Roddam AW. 2010. Insulin-like growth factor 1 (IGF1), IGF binding protein 3 (IGFBP3), and breast cancer risk: pooled individual data analysis of 17 prospective studies. Lancet Oncol, 11(6), 530–542. PMCID: PMC3113287.

Brinton LA, Cook MB, McCormack V, et al. 2014a. Anthropometric and hormonal risk factors for male breast cancer: male breast cancer pooling project results. J Natl Cancer Inst, 106(3), djt465. PMCID: PMC3975166.

Brinton LA, Figueroa JD, Awuah B, et al. 2014b. Breast cancer in Sub-Saharan Africa: opportunities for prevention. Breast Cancer Res Treat, 144(3), 467–478. PMCID: PMC4023680.

Brinton LA, Key TJ, Kolonel LN, et al. 2015. Prediagnostic sex steroid hormones in relation to male breast cancer risk. J Clin Oncol, 33(18), 2041–2050. PMCID: PMC4461805.

Brinton LA, Richesson D, Leitzmann MF, et al. 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. PMCID: PMC2755180.

Brinton LA, Scoccia B, Moghissi KS, et al. 2014c. Long-term relationship of ovulation-stimulating drugs to breast cancer risk. Cancer Epidemiol Biomarkers Prev, 23(4), 584–593. PMCID: PMC3979528.

Brophy JT, Keith MM, Watterson A, et al. 2012. Breast cancer risk in relation to occupations with exposure to carcinogens and endocrine disruptors: a Canadian case-control study. Environ Health, 11, 87. PMCID: PMC3533941.

Cai Q, Zhang B, Sung H, et al. 2014. Genome-wide association analysis in East Asians identifies breast cancer susceptibility loci at 1q32.1, 5q14.3 and 15q26.1. Nat Genet, 46(8), 886–890. PMCID: PMC4127632.

Calle EE, Feigelson HS, Hildebrand JS, et al. 2009. Postmenopausal hormone use and breast cancer associations differ by hormone regimen and histologic subtype. Cancer, 115(5), 936–945. PMID: 19156895.

Calle EE, and Kaaks R. 2004. Overweight, obesity and cancer: epidemiological evidence and proposed mechanisms. Nat Rev Cancer, 4(8), 579–591. PMID: 15286738.

Canchola AJ, Anton-Culver H, Bernstein L, et al. 2012. Body size and the risk of postmenopausal breast cancer subtypes in the California Teachers Study cohort. Cancer Causes Control, 23(3), 473–485. PMCID: PMC3366039.

Carvalho FM, Bacchi LM, Pincerato KM, Van de Rijn M, and Bacchi CE. 2014. Geographic differences in the distribution of molecular subtypes of breast cancer in Brazil. BMC Womens Health, 14, 102. PMCID: PMC4153008.

Cecchini R, Costantino JP, Cauley JA, et al. 2012. Baseline mammographic breast density and the risk of invasive breast cancer in postmenopausal women participating in the NSABP Study of Tamoxifen and Raloxifene (STAR). Cancer Prev Res (Phila), 5(11), 1321–1329. PMCID: PMC4131535.

Chan DS, Vieira AR, Aune D, et al. 2014. Body mass index and survival in women with breast cancer: systematic literature review and meta-analysis of 82 follow-up studies. Ann Oncol, 25(10): 1901–1914. PMCID: PMC4176449.

Chen C, Ma X, Zhong M, and Yu Z. 2010. Extremely low-frequency electromagnetic fields exposure and female breast cancer risk: a meta-analysis based on 24,338 cases and 60,628 controls. Breast Cancer Res Treat, 123(2), 569–576. PMID: 20145992.

Chen F, Chen GK, Millikan RC, et al. 2011. Fine-mapping of breast cancer susceptibility loci characterizes genetic risk in African Americans. Hum Mol Genet, 20(22), 4491–4503. PMCID: PMC3196889.

Chen FY, Ou HY, Wang SM, et al. 2013. Associations between body mass index and molecular subtypes as well as other clinical characteristics of breast cancer in Chinese women. Ther Clin Risk Manag, 9, 131–137. PMCID: PMC3617914.

Chen J, Pee D, Ayyagari R, et al. 2006. Projecting absolute invasive breast cancer risk in white women with a model that includes mammographic density. J Natl Cancer Inst, 98(17), 1215–1226. PMID: 16954474.

Chen L, Malone KE, and Li CI. 2014. Bra wearing not associated with breast cancer risk: a population-based case-control study. Cancer Epidemiol Biomarkers Prev, 23(10), 2181–2185. PMCID: PMC4184992.

Chlebowski RT. 2014. Current concepts in breast cancer chemoprevention. Pol Arch Med Wewn, 124(4), 191–199. PMID: 24618912.

(p.882) Chlebowski RT, Chen Z, Cauley JA, et al. 2010. Oral bisphosphonate use and breast cancer incidence in postmenopausal women. J Clin Oncol, 28(22), 3582–3590. PMCID: PMC2917313.

Chlebowski RT, Manson JE, Anderson GL, et al. 2013. Estrogen plus progestin and breast cancer incidence and mortality in the Women’s Health Initiative Observational Study. J Natl Cancer Inst, 105(8), 526–535. PMCID: PMC3691942.

Clevenger CV, Furth PA, Hankinson SE, and Schuler LA. 2003. The role of prolactin in mammary carcinoma. Endocr Rev, 24(1), 1–27. PMCID: PMC1698952.

Cohen SS, Matthews CE, Bradshaw PT, et al. 2013. Sedentary behavior, physical activity, and likelihood of breast cancer among Black and White women: a report from the Southern Community Cohort Study. Cancer Prev Res (Phila), 6(6), 566–576. PMCID: PMC3703619.

Colditz GA, and Bohlke K. 2014. Priorities for the primary prevention of breast cancer. CA Cancer J Clin, 64(3), 186–194. PMCID: PMC24647877.

Colditz GA, Bohlke K, and Berkey CS. 2014. Breast cancer risk accumulation starts early: prevention must also. Breast Cancer Res Treat, 145(3), 567–579. PMCID: PMC4079839.

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. PMID: 8656904.

Collaborative Group on Hormonal Factors in Breast Cancer. 2002. Breast cancer and breastfeeding: collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50302 women with breast cancer and 96973 women without the disease. Lancet, 360(9328), 187–195. PMID: 12133652.

Collaborative Group on Hormonal Factors in Breast Cancer. 2012. Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol, 13(11), 1141–1151. PMCID: PMC3488186.

Collins LC, Baer HJ, Tamimi RM et al. 2006. The influence of family history on breast cancer risk in women with biopsy-confirmed benign breast disease: results from the Nurses’ Health Study. Cancer, 107(6), 1240–1247. PMID: 16902983.

Cook NR, Lee IM, Gaziano JM, et al. 2005. Low-dose aspirin in the primary prevention of cancer: the Women’s Health Study: a randomized controlled trial. JAMA, 294(1), 47–55. PMID: 15998890.

Costantino JP, Gail MH, Pee D, et al. 1999. Validation studies for models projecting the risk of invasive and total breast cancer incidence. J Natl Cancer Inst, 91(18), 1541–1548. PMID: 10491430.

Crowe FL, Key TJ, Allen NE, et al. 2011. A cross-sectional analysis of the associations between adult height, BMI and serum concentrations of IGF-I and IGFBP-1 -2 and -3 in the European Prospective Investigation into Cancer and Nutrition (EPIC). Ann Hum Biol, 38(2), 194–202. PMID: 20731527.

Cummings SR, Tice JA, Bauer S, et al. 2009. Prevention of breast cancer in postmenopausal women: approaches to estimating and reducing risk. J Natl Cancer Inst, 101(6), 384–398. PMCID: PMC2720698.

Cust AE, Stocks T, Lukanova A, et al. 2009. The influence of overweight and insulin resistance on breast cancer risk and tumour stage at diagnosis: a prospective study. Breast Cancer Res Treat, 113(3), 567–576. PMID: 18330696.

Cuzick J, Sestak I, Bonanni B, et al. 2013. Selective oestrogen receptor modulators in prevention of breast cancer: an updated meta-analysis of individual participant data. Lancet, 381(9880), 1827–1834. PMCID: PMC3671272.

Cuzick J, Sestak I, Cawthorn S, et al. 2015. Tamoxifen for prevention of breast cancer: extended long-term follow-up of the IBIS-I breast cancer prevention trial. Lancet Oncol, 16(1), 67–75. PMCID: PMC4772450.

Cuzick J, Sestak I, Forbes JF, et al. 2014. Anastrozole for prevention of breast cancer in high-risk postmenopausal women (IBIS-II): an international, double-blind, randomised placebo-controlled trial. Lancet, 383(9922), 1041–1048. PMID: 24333009.

Cuzick J, Warwick J, Pinney E, et al. 2011. Tamoxifen-induced reduction in mammographic density and breast cancer risk reduction: a nested case–control study. J Natl Cancer Inst, 103, 744–752. PMID: 21483019.

Dallal CM, Brinton LA, Matthews CE, et al. 2016. Association of active and sedentary behaviors with postmenopausal estrogen metabolism. Med Sci Sports Exerc, 48(3), 439–448. PMCID: PMC4861700.

Dallal CM, Tice JA, Buist DS, et al. 2014. Estrogen metabolism and breast cancer risk among postmenopausal women: a case-cohort study within B~FIT. Carcinogenesis, 35(2), 346–355. PMCID: PMC3908751.

Davies KR, Cantor SB, and Brewster AM. 2015. Better contralateral breast cancer risk estimation and alternative options to contralateral prophylactic mastectomy. Int J Womens Health, 7, 181–187. PMCID: PMC4324540.

De Stavola BL, dos Santos Silva I, McCormack V, et al. 2004. Childhood growth and breast cancer. Am J Epidemiol, 159(7), 671–682. PMID: 15033645.

Dearth RK, Delgado DA, Hiney JK, et al. 2010. Parity-induced decrease in systemic growth hormone alters mammary gland signaling: a potential role in pregnancy protection from breast cancer. Cancer Prev Res (Phila), 3(3), 312–321. PMCID: PMC2833223.

Degnim AC, Visscher DW, Berman HK, et al. 2007. Stratification of breast cancer risk in women with atypia: a Mayo cohort study. J Clin Oncol, 25(19), 2671–2677. PMID: 17563394.

Desai P, Lehman A, Chlebowski RT, et al. 2015. Statins and breast cancer stage and mortality in the Women’s Health Initiative. Cancer Causes Control, 26(4), 529–539. PMID: 25736184.

Dieci MV, Orvieto E, Dominici M, Conte P, and Guarneri V. 2014. Rare breast cancer subtypes: histological, molecular, and clinical peculiarities. Oncologist, 19(8), 805–813. PMCID: PMC4122475.

Dong JY, and Qin LQ. 2011. Soy isoflavones consumption and risk of breast cancer incidence or recurrence: a meta-analysis of prospective studies. Breast Cancer Res Treat, 125(2), 315–323. PMID: 21113655.

Dorgan JF, Stanczyk FZ, Egleston BL, et al. 2009. Prospective case-control study of serum mullerian inhibiting substance and breast cancer risk. J Natl Cancer Inst, 101(21), 1501–1509. PMCID: PMC2773186.

Dossus L, Boutron-Ruault MC, Kaaks R, et al. 2014. Active and passive cigarette smoking and breast cancer risk: results from the EPIC cohort. Int J Cancer, 134(8), 1871–1888. PMID: 24590452.

Dupont WD, and Page DL. 1985. Risk factors for breast cancer in women with proliferative breast disease. N Engl J Med, 312(3), 146–151. PMID: 3965932.

Early Breast Cancer Trialists’ Collaborative Group, Davies C, Godwin J et al. 2011. Relevance of breast cancer hormone receptors and other factors to the efficacy of adjuvant tamoxifen: patient-level meta-analysis of randomised trials. Lancet, 378(9793), 771–784. PMCID: PMC3163848.

Easton DF. 1999. How many more breast cancer predisposition genes are there? Breast Cancer Res, 1(1), 14–17. PMCID: PMC138504.

Easton DF, Pharoah PD, Antoniou AC, et al. 2015. Gene-panel sequencing and the prediction of breast-cancer risk. N Engl J Med, 372(23), 2243–2257. PMCID: PMC4610139.

Eliassen AH, Spiegelman D, Xu X, et al. 2011. Urinary estrogens and estrogen metabolites and subsequent risk of breast cancer among premenopausal women. Cancer Res, 72(3), 696–706. PMCID: PMC3271178.

Eliassen AH, Tworoger SS, Mantzoros CS, Pollak MN, and Hankinson SE. 2007. Circulating insulin and c-peptide levels and risk of breast cancer among predominately premenopausal women. Cancer Epidemiol Biomarkers Prev, 16(1), 161–164. PMID: 17220346.

Erbas B, Provenzano E, Armes J, and Gertig D. 2006. The natural history of ductal carcinoma in situ of the breast: a review. Breast Cancer Res Treat, 97(2), 135–144. PMID: 16319971.

Fachal L, and Dunning AM. 2015. From candidate gene studies to GWAS and post-GWAS analyses in breast cancer. Curr Opin Genet Dev, 30, 32–41. PMID: 25727315.

Falk RT, Brinton LA, Dorgan JF, et al. 2013. Relationship of serum estrogens and estrogen metabolites to postmenopausal breast cancer risk: a nested case-control study. Breast Cancer Res, 15(2), R34. PMCID: PMC4053199.

Falk RT, Xu X, Keefer L, Veenstra TD, and Ziegler RG. 2008. A liquid chromatography-mass spectrometry method for the simultaneous measurement of 15 urinary estrogens and estrogen metabolites: assay reproducibility and interindividual variability. Cancer Epidemiol Biomarkers Prev, 17(12), 3411–3418. PMCID: PMC4158914.

Familial Breast Cancer Working Group. 2001. Familial breast cancer: collaborative reanalysis of individual data from 52 epidemiological studies including 58,209 women with breast cancer and 101,986 women without the disease. Lancet, 358(9291), 1389–1399. PMID: 11705483.

Farhat GN, Cummings SR, Chlebowski RT, et al. 2011. Sex hormone levels and risks of estrogen receptor-negative and estrogen receptor-positive breast cancers. J Natl Cancer Inst, 103(7), 562–570. PMCID: PMC3071353.

Feng Y, Stram DO, Rhie SK, et al. 2014. A comprehensive examination of breast cancer risk loci in African American women. Hum Mol Genet, 23(20), 5518–5526. PMCID: PMC4168823.

Ferlay J, Soerjomataram I, Dikshit R, et al. 2015. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer, 136(5), E359–386. PMID: 25220842.

Figueroa JD, Pfeiffer RM, Brinton LA, et al. 2016. Standardized measures of lobular involution and subsequent breast cancer risk among women with benign breast disease among women with benign breast disease; a nested case-control study. Breast Cancer Res, 159(1), 13-172. PMID: 27488681.

Figueroa JD, Pfeiffer RM, Patel DA, et al. 2014. Terminal duct lobular unit involution of the normal breast: implications for breast cancer etiology. J Natl Cancer Inst, 106(10), dju286. PMCID: PMC4200067.

(p.883) Fischer C, Kuchenbacker K, Engel C, et al. 2013. Evaluating the performance of the breast cancer genetic risk models BOADICEA, IBIS, BRCAPRO and Claus for predicting BRCA1/2 mutation carrier probabilities: a study based on 7352 families from the German Hereditary Breast and Ovarian Cancer Consortium. J Med Genet, 50(6), 360–367. PMID: 23564750.

Fitzgibbons PL, Henson DE, and Hutter RV. 1998. Benign breast changes and the risk for subsequent breast cancer: an update of the 1985 consensus statement. Cancer Committee of the College of American Pathologists. Arch Pathol Lab Med, 122(12), 1053–1055. PMID: 98708652.

Fontein DB, de Glas NA, Duijm M, et al. 2013. Age and the effect of physical activity on breast cancer survival: a systematic review. Cancer Treat Rev, 39(8), 958–965. PMID: 23608116.

Fournier A, Dos Santos G, Guillas G, et al. 2014. Recent recreational physical activity and breast cancer risk in postmenopausal women in the E3N cohort. Cancer Epidemiol Biomarkers Prev, 23(9), 1893–1902. PMID: 25114017.

Freedman AN, Yu B, Gail MH, et al. 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. PMCID: PMC3107748.

Fuhrman BJ, Schairer C, Gail MH, et al. 2012. Estrogen metabolism and risk of breast cancer in postmenopausal women. J Natl Cancer Inst, 104(4), 326–339. PMCID: PMC3283536.

Fuhrman BJ, Xu X, Falk RT, et al. 2014. Assay reproducibility and interindividual variation for 15 serum estrogens and estrogen metabolites measured by liquid chromatography-tandem mass spectrometry. Cancer Epidemiol Biomarkers Prev, 23(12), 2649–2657. PMCID: PMC4289148.

Gail MH. 2008. Discriminatory accuracy from single-nucleotide polymorphisms in models to predict breast cancer risk. J Natl Cancer Inst, 100(14), 1037–1041. PMCID: PMC2528005.

Gail MH, Costantino JP, Pee D, et al. 2007. Projecting individualized absolute invasive breast cancer risk in African American women. J Natl Cancer Inst, 99(23), 1782–1792. PMID: 18042936.

Gandini S, Puntoni M, Heckman-Stoddard BM, et al. 2014. Metformin and cancer risk and mortality: a systematic review and meta-analysis taking into account biases and confounders. Cancer Prev Res (Phila), 7(9), 867–885. PMCID: PMC4154969.

Garcia-Closas M, Gunsoy NB, and Chatterjee N. 2014. Combined associations of genetic and environmental risk factors: implications for prevention of breast cancer. J Natl Cancer Inst, 106(11). PMCID: PMC4271030.

Garcia-Esquinas E, Guino E, Castano-Vinyals G, et al. 2016. Association of diabetes and diabetes treatment with incidence of breast cancer. Acta Diabetol, 53(1), 99–107. PMCID: PMC4749645.

Gaudet MM, Carter BD, Patel AV, et al. 2014a. Waist circumference, body mass index, and postmenopausal breast cancer incidence in the Cancer Prevention Study-II Nutrition Cohort. Cancer Causes Control, 25(6), 737–745. PMID: 24715420.

Gaudet MM, Gapstur SM, Sun J, et al. 2013a. Active smoking and breast cancer risk: original cohort data and meta-analysis. J Natl Cancer Inst, 105(8), 515–525. PMID: 23449445.

Gaudet MM, Gapstur SM, Sun J, et al. 2014b. Oophorectomy and hysterectomy and cancer incidence in the Cancer Prevention Study-II Nutrition Cohort. Obstet Gynecol, 123(6), 1247–1255. PMID: 24807324.

Gaudet MM, Patel AV, Sun J, Teras LR, and Gapstur SM. 2013b. Tubal sterilization and breast cancer incidence: results from the cancer prevention study II nutrition cohort and meta-analysis. Am J Epidemiol, 177(6), 492–499. PMID: 23416445.

Gaudet MM, Patel AV, Teras LR, et al. 2013c. Obesity-related markers and breast cancer in CPS-II Nutrition Cohort. Int J Mol Epidemiol Genet, 4(3), 156–166. PMCID: PMC3773567.

Gaudet MM, Press MF, Haile RW, et al. 2011. Risk factors by molecular subtypes of breast cancer across a population-based study of women 56 years or younger. Breast Cancer Res Treat, 130(2), 587–597. PMCID: PMC3721192.

Gertig DM, Stillman IE, Byrne C et al. 1999. Association of age and reproductive factors with benign breast tissue composition. Cancer Epidemiol Biomarkers Prev, 8(10), 873–879. PMID: 10548315.

Ghosh K, Hartmann LC, Reynolds C, et al. 2010a. Association between mammographic density and age-related lobular involution of the breast. J Clin Oncol, 28(13), 2207–2212. PMCID: PMC2860438.

Ghosh K, Vachon CM, Pankratz VS, et al. 2010b. Independent association of lobular involution and mammographic breast density with breast cancer risk. J Natl Cancer Inst, 102(22), 1716–1723. PMCID: PMC2982810.

Gierach GL, Brinton LA, and Sherman ME. 2010. Lobular involution, mammographic density, and breast cancer risk: visualizing the future? J Natl Cancer Inst, 102(22), 1685–1687. PMCID: PMC2982811.

Gierach GL, Curtis RE, Pfeiffer RM, et al. 2017. Association of adjuvant tamoxifen and aromatase inhibitor therapy with contralateral breast cancer risk among US women with breast cancer in a general community setting. JAMA Oncol, 3(2):186–193. PMID: 27711920.

Gierach GL, Geller BM, Shepherd JA, Patel DA, Vacek PM, Weaver DL, et al. 2014. Comparison of mammographic density assessed as volumes and areas among women undergoing diagnostic image-guided breast biopsy. Cancer Epidemiol Biomarkers Prev, 23(11), 2338–2348.

Gierach GL, Patel DA, Pfeiffer RM, et al. 2016. Relationship of terminal duct lobular unit involution of the breast with area and volume mammographic densities. Cancer Prev Res, 9(2), 149–158. PMCID: PMC4784982.

Gierach GL, Yang XR, Figueroa JD, and Sherman ME. 2013. Emerging concepts in breast cancer risk prediction. Curr Obstet Gynecol Rep, 2(1), 43–52. PMCID: PMC4262133.

Ginsburg OM, Martin LJ, and Boyd NF. 2008. Mammographic density, lobular involution, and risk of breast cancer. Br J Cancer, 99, 1369–1374. PMCID: PMC2579686.

Global Burden of Disease Cancer. 2015. The global burden of cancer 2013. JAMA Oncol, 1(4), 505–527. PMCID: PMC4500822.

Goss PE, Ingle JN, Ales-Martinez JE, et al. 2011. Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med, 364(25), 2381–2391. PMID: 21639806.

Gottschau M, Kjaer SK, Jensen A, Munk C, and Mellemkjaer L. 2015. Risk of cancer among women with polycystic ovary syndrome: a Danish cohort study. Gynecol Oncol, 136(1), 99–103. PMID: 25451694.

Green J, Cairns BJ, Casabonne D, et al. 2011. Height and cancer incidence in the Million Women Study: prospective cohort, and meta-analysis of prospective studies of height and total cancer risk. Lancet Oncol, 12(8), 785–794. PMCID: PMC3148429.

Grenier D, Cooke AL, Lix L, et al. 2011. Bone mineral density and risk of postmenopausal breast cancer. Breast Cancer Res Treat, 126(3), 679–686. PMID: 20838879.

Gunter MJ, Hoover DR, Yu H, et al. 2009. Insulin, insulin-like growth factor-I, and risk of breast cancer in postmenopausal women. J Natl Cancer Inst, 101(1), 48–60. PMCID: PMC2639294.

Gunter MJ, Wang T, Cushman M, et al. 2015. Circulating adipokines and inflammatory markers and postmenopausal breast cancer risk. J Natl Cancer Inst, 107(9), djv169. PMCID: PMC4651104.

Guo J, Huang Y, Yang L, et al. 2015. Association between abortion and breast cancer: an updated systematic review and meta-analysis based on prospective studies. Cancer Causes Control, 26(6), 811–819. PMID: 25779378.

Guo YP, Martin LJ, Hanna W, et al. 2001. Growth factors and stromal matrix proteins associated with mammographic densities. Cancer Epidemiol Biomarkers Prev, 10(3), 243–248. PMID: 11303594.

Gupta V, Carey JL, Kawakubo H, et al. 2005. Mullerian inhibiting substance suppresses tumor growth in the C3(1)T antigen transgenic mouse mammary carcinoma model. Proc Natl Acad Sci U S A, 102(9), 3219–3224. PMCID: PMC552936.

Hankinson SE, and Eliassen AH. 2007. Endogenous estrogen, testosterone and progesterone levels in relation to breast cancer risk. J Steroid Biochem Mol Biol, 106(1–5), 24–30. PMCID: PMC2715949.

Harris HR, Willett WC, Terry KL, and Michels KB. 2011. Body fat distribution and risk of premenopausal breast cancer in the Nurses’ Health Study II. J Natl Cancer Inst, 103(3), 273–278. PMCID: PMC3107569.

Hartmann LC, Radisky DC, Frost MH et al. 2014. Understanding the premalignant potential of atypical hyperplasia through its natural history: a longitudinal cohort study. Cancer Prev Res (Phila), 7(2), 211–217. PMCID: PMC4167687.

Hartmann LC, Sellers TA, Frost MH, et al. 2005. Benign breast disease and the risk of breast cancer. N Engl J Med, 353(3), 229–237. PMID: 16034008.

He C, Anand ST, Ebell MH, Vena JE, and Robb SW. 2015. Circadian disrupting exposures and breast cancer risk: a meta-analysis. Int Arch Occup Environ Health, 88(5), 533–547. PMID: 25261318.

Healy GN, Matthews CE, Dunstan DW, Winkler EA, and Owen N. 2011. Sedentary time and cardio-metabolic biomarkers in US adults: NHANES 2003–06. Eur Heart J, 32(5), 590–597. PMCID: PMC3634159.

Hildebrand JS, Gapstur SM, Campbell PT, Gaudet MM, and Patel AV. 2013. Recreational physical activity and leisure-time sitting in relation to postmenopausal breast cancer risk. Cancer Epidemiol Biomarkers Prev, 22(10), 1906–1912. PMID: 24097200.

Hoover RN. 2012. That recognised risk factors can explain past and present international differences in breast cancer incidence: misconceptions 5. Br J Cancer, 107(3), 408–410. PMCID: PMC3405203.

Hoover RN, Hyer M, Pfeiffer RM, et al. 2011. Adverse health outcomes in women exposed in utero to diethylstilbestrol. N Engl J Med, 365(14), 1304–1314. PMID: 21991952.

Horn J, Opdahl S, Engstrom MJ, et al. 2014. Reproductive history and the risk of molecular breast cancer subtypes in a prospective study of Norwegian women. Cancer Causes Control, 25(7), 881–889. PMID: 24789514.

Horne HN, Sherman ME, Pfeiffer RM et al. 2016. Circulating insulin-like growth factor-I, insulin-like growth factor binding protein-3 and (p.884) terminal duct lobular unit involution of the breast: a cross-sectional study of women with benign breast disease. Breast Cancer Res, 18(1), 24. PMCID: PMC4758090.

Howell A, Anderson AS, Clarke RB, et al. 2014. Risk determination and prevention of breast cancer. Breast Cancer Res, 16(5), 446. PMCID: PMC4303126.

Howlader N, Altekruse SF, Li CI, et al. 2014. US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status. J Natl Cancer Inst, 106(5), dju055. PMCID: PMC459–552.

Hunter DJ, Colditz GA, Hankinson SE, et al. 2010. Oral contraceptive use and breast cancer: a prospective study of young women. Cancer Epidemiol Biomarkers Prev, 19(10), 2496–2502. PMCID: PMC3055790.

Huo CW, Chew GL, Britt KL, et al. 2014. Mammographic density: a review on the current understanding of its association with breast cancer. Breast Cancer Res Treat, 144(3), 479–502. PMID: 24615497.

IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. 2010. Alcohol consumption and ethyl carbamate. IARC Monogr Eval Carcinog Risks Hum, 96, 3–1383. PMID: 21735939.

Ingber SZ, Buser MC, Pohl HR, et al. 2013. DDT/DDE and breast cancer: a meta-analysis. Regul Toxicol Pharmacol, 67(3), 421–433. PMID: 24021539.

Islam T, Matsuo K, Ito H, et al. 2012. Reproductive and hormonal risk factors for luminal, HER2-overexpressing, and triple-negative breast cancer in Japanese women. Ann Oncol, 23(9), 2435–2441. PMID: 22328736.

Jacob P, and Stram DO. 2013. Late health effects of radiation exposure: new statistical, epidemiological, and biological approaches. Int J Radiat Biol, 89(8), 673–683. PMID: 23402349.

James RE, Lukanova A, Dossus L, et al. 2011. Postmenopausal serum sex steroids and risk of hormone receptor-positive and -negative breast cancer: a nested case-control study. Cancer Prev Red (Phila), 4(10), 1626–1635. PMID: 21813404.

Jatoi I, and Anderson WF. 2010. Qualitative age interactions in breast cancer studies: a mini-review. Future Oncol, 6(11), 1781–1788. PMID: 21142663.

Jemal A, Bray F, Center MM, et al. 2011. Global cancer statistics. CA Cancer J Clin, 61(2), 69–90. PMID: 21296855.

Jenkins EO, Deal AM, Anders CK, et al. 2014. Age-specific changes in intrinsic breast cancer subtypes: a focus on older women. Oncologist, 19(10), 1076–1083. PMCID: PMC4200998.

John EM, Phipps AI, Davis A, and Koo J. 2005. Migration history, acculturation, and breast cancer risk in Hispanic women. Cancer Epidemiol Biomarkers Prev, 14(12), 2905–2913. PMID: 16365008.

John EM, Sangaramoorthy M, Hines LM, et al. 2015. Overall and abdominal adiposity and premenopausal breast cancer risk among Hispanic women: the breast cancer health disparities study. Cancer Epidemiol Biomarkers Prev, 24(1), 138–147. PMCID: PMC4294975.

Jung S, Spiegelman D, Baglietto L, et al. 2013. Fruit and vegetable intake and risk of breast cancer by hormone receptor status. J Natl Cancer Inst, 105(3), 219–236. PMCID: PMC3593764.

Kabat GC, Xue X, Kamensky V, et al. 2015. Risk of breast, endometrial, colorectal, and renal cancers in postmenopausal women in association with a body shape index and other anthropometric measures. Cancer Causes Control, 26(2), 219–229. PMID: 25430815.

Kaiser JC, Jacob P, Meckbach R, and Cullings HM. 2012. Breast cancer risk in atomic bomb survivors from multi-model inference with incidence data 1958–1998. Radiat Environ Biophys, 51(1), 1–14. PMID: 21947564.

Kent C, Horton J, Blitzblau R, and Koontz BF. 2015. Whose disease will recur after mastectomy for early stage, node-negative breast cancer? A systematic review. Clin Breast Cancer, 15(6), 403–412. PMID: 26198331.

Kerlikowske K, Ichikawa L, Miglioretti DL, et al. 2007. Longitudinal measurement of clinical mammographic breast density to improve estimation of breast cancer risk. J Natl Cancer Inst, 99(5), 386–395. PMID: 17341730.

Keum N, Greenwood DC, Lee DH, et al. 2015. Adult weight gain and adiposity-related cancers: a dose-response meta-analysis of prospective observational studies. J Natl Cancer Inst, 107(2), djv088. PMID: 25757865.

Key T, Appleby P, Barnes I, et al. 2002. Endogenous sex hormones and breast cancer in postmenopausal women: reanalysis of nine prospective studies. J Natl Cancer Inst, 94(8), 606–616. PMID: 11959894.

Key TJ, Appleby PN, Reeves GK, et al. 2011. Circulating sex hormones and breast cancer risk factors in postmenopausal women: reanalysis of 13 studies. Br J Cancer, 105(5), 709–722. PMCID: PMC3188939.

Key TJ, Appleby PN, Reeves GK, et al. 2013. Sex hormones and risk of breast cancer in premenopausal women: a collaborative reanalysis of individual participant data from seven prospective studies. Lancet Oncol, 14(10), 1009–1019. PMID: 23890780.

Khodr ZG, Sherman ME, Pfeiffer RM, et al. 2014. Circulating sex hormones and terminal duct lobular unit involution of the normal breast. Cancer Epidemiol Biomarkers Prev, 23(12), 2765–2773. PMCID: PMC4333643.

Kim J, Han W, Moon H-G, et al. 2012. Breast density change as a predictive surrogate for response to adjuvant endocrine therapy in hormone receptor positive breast cancer. Breast Cancer Res, 14(4), R102. PMCID: PMC3680951.

Ko KL, Shin IS, You JY, et al. 2013. Adjuvant tamoxifen-induced mammographic breast density reduction as a predictor for recurrence in estrogen receptor-positive premenopausal breast cancer patients. Breast Cancer Res Treat, 142(3), 559–567. PMID: 24233999.

Kurian AW, McClure LA, John EM, et al. 2009. Second primary breast cancer occurrence according to hormone receptor status. J Natl Cancer Inst, 101(15), 1058–1065. PMCID: PMC2720990.

Kushi LH, Doyle C, McCullough M, et al. 2012. American Cancer Society Guidelines on nutrition and physical activity for cancer prevention: reducing the risk of cancer with healthy food choices and physical activity. CA Cancer J Clin, 62(1), 30–67. PMID: 22237782.

Kwan ML, Kushi LH, Weltzien E, et al. 2009. Epidemiology of breast cancer subtypes in two prospective cohort studies of breast cancer survivors. Breast Cancer Res, 11(3), R31. PMID: 19463150.

Li CI, Beaber EF, Tang MT, et al. 2012. Effect of depo-medroxyprogesterone acetate on breast cancer risk among women 20 to 44 years of age. Cancer Res, 72(8), 2028–2035. PMCID: PMC3328650.

Li CI, Beaber EF, Tang MT, et al. 2013a. Reproductive factors and risk of estrogen receptor positive, triple-negative, and HER2-neu overexpressing breast cancer among women 20–44 years of age. Breast Cancer Res Treat, 137(2), 579–587. PMCID: PMC3547981.

Li J, Humphreys K, Eriksson L, et al. 2013b. Mammographic density reduction is a prognostic marker of response to adjuvant tamoxifen therapy in postmenopausal patients with breast cancer. J Clin Oncol, 31(18), 2249–2256. PMCID: PMC3677838.

Li W, Ray RM, Thomas DB, et al. 2013c. Occupational exposure to magnetic fields and breast cancer among women textile workers in Shanghai, China. Am J Epidemiol, 178(7), 1038–1045. PMCID: PMC3783098.

Liao DJ, and Dickson RB. 2002. Roles of androgens in the development, growth, and carcinogenesis of the mammary gland. J Steroid Biochem Mol Biol, 80(2), 175–189. PMID: 11897502.

Lichtenstein P, Holm NV, Verkasalo PK, et al. 2000. Environmental and heritable factors in the causation of cancer: analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med, 343(2), 78–85. PMID: 10891514.

Lie Fong S, Visser JA, Welt CK, et al. 2012. Serum anti-mullerian hormone levels in healthy females: a nomogram ranging from infancy to adulthood. J Clin Endocrinol Metab, 97(12), 4650–4655. PMCID: PMC3683801.

Lindstrom S, Thompson DJ, Paterson AD, et al. 2014. Genome-wide association study identifies multiple loci associated with both mammographic density and breast cancer risk. Nat Commun, 5, 5303. PMCID: PMC4320806.

Lindstrom S, Vachon CM, Li J, et al. 2011. Common variants in ZNF365 are associated with both mammographic density and breast cancer risk. Nat Genet, 43(3), 185–187. PMCID: PMC3076615.

Linos E, Spanos D, Rosner BA, et al. 2008. Effects of reproductive and demographic changes on breast cancer incidence in China: a modeling analysis. J Natl Cancer Inst, 100(19), 1352–1360. PMCID: PMC2556703.

Linos E, Willett WC, Cho E, and Frazier L. 2010. Adolescent diet in relation to breast cancer risk among premenopausal women. Cancer Epidemiol Biomarkers Prev, 19(3), 689–696. PMCID: PMC2866323.

Lipworth L, Tarone RE, Friis S, et al. 2009. Cancer among Scandinavian women with cosmetic breast implants: a pooled long-term follow-up study. Int J Cancer, 124(2), 490–493. PMID: 19003966.

Liu R, Nelson DO, Hurley S, Hertz A, and Reynolds P. 2015. Residential exposure to estrogen disrupting hazardous air pollutants and breast cancer risk: the California Teachers Study. Epidemiology, 26(3), 365–373. PMID: 25760782.

Luctkar-Flude M, Aiken A, McColl MA, and Tranmer J. 2015. A comprehensive framework and key guideline recommendations for the provision of evidence-based breast cancer survivorship care within the primary care setting. Fam Pract, 32(2), 129–140. PMID: 25500746.

Luo J, Margolis KL, Wactawski-Wende J, et al. 2011. Association of active and passive smoking with risk of breast cancer among postmenopausal women: a prospective cohort study. BMJ, 342, d1016. PMCID: PMC3047002.

Luo ZC, Cheung YB, He Q, Albertsson-Wikland K, and Karlberg J. 2003. Growth in early life and its relation to pubertal growth. Epidemiology, 14(1), 65–73. PMID: 12500048.

Lynch BM, Friedenreich CM, Winkler EA, et al. 2011a. Associations of objectively assessed physical activity and sedentary time with biomarkers of breast cancer risk in postmenopausal women: findings from NHANES (2003–2006). Breast Cancer Res Treat, 130(1), 183–194. PMID: 21553294.

Lynch BM, Neilson HK, and Friedenreich CM. 2011b. Physical activity and breast cancer prevention. Recent Results Cancer Res, 186, 13–42. PMID: 21113759.

Ma H, Wang Y, Sullivan-Halley J, et al. 2010. Use of four biomarkers to evaluate the risk of breast cancer subtypes in the women’s contraceptive and reproductive experiences study. Cancer Res, 70(2), 575–587. PMCID: PMC2807992.

(p.885) Macon MB, and Fenton SE. 2013. Endocrine disruptors and the breast: early life effects and later life disease. J Mammary Gland Biol Neoplasia, 18(1), 43–61. PMCID: PMC3682794.

Martelotto LG, Ng CK, Piscuoglio S, Weigelt B, and Reis-Filho JS. 2014. Breast cancer intra-tumor heterogeneity. Breast Cancer Res, 16(3), 210. PMCID: PMC4053234.

Martin LJ, and Boyd NF. 2008. Mammographic density: potential mechanisms of breast cancer risk associated with mammographic density: hypotheses based on epidemiological evidence. Breast Cancer Res, 10(1), 201. PMCID: PMC2374950.

Matsuno RK, Costantino JP, Ziegler RG, et al. 2011. Projecting individualized absolute invasive breast cancer risk in Asian and Pacific Islander American women. J Natl Cancer Inst, 103(12), 951–961. PMCID: PMC3119648.

McCart Reed AE, Kutasovic JR, Lakhani SR, and Simpson PT. 2015. Invasive lobular carcinoma of the breast: morphology, biomarkers and ‘omics.’ Breast Cancer Res, 17, 12. PMCID: PMC4310190.

McCormack VA, and dos Santos Silva I. 2006. Breast density and parenchymal patterns as markers of breast cancer risk: a meta- analysis. Cancer Epidemiol Biomarkers Prev, 15(6), 1159–1169. PMID: 16775176.

McKenzie F, Ferrari P, Freisling H, et al. 2015. Healthy lifestyle and risk of breast cancer among postmenopausal women in the European Prospective Investigation into Cancer and Nutrition cohort study. Int J Cancer, 136(11), 2640–2648. PMID: 25379993.

McKian KP, Reynolds CA, Visscher DW, et al. 2009. Novel breast tissue feature strongly associated with risk of breast cancer. J Clin Oncol, 27(35), 5893–5898. PMCID: PMC2793038.

Melchor L, and Benitez J. 2013. The complex genetic landscape of familial breast cancer. Hum Genet, 132(8), 845–863. PMID: 23552954.

Michailidou K, Hall P, Gonzalez-Neira A, et al. 2013. Large-scale genotyping identifies 41 new loci associated with breast cancer risk. Nat Genet, 45(4), 353–361, 361e351–352. PMCID: PMC3771688.

Michailidou K, Lindstrom S, Dennis J, et al. 2017. Association analysis identifies 65 new breast cancer risk loci and predicts target genes. Nat Genet, [under review].

Mikoczy Z, Tinnerberg H, Bjork J, and Albin M. 2011. Cancer incidence and mortality in Swedish sterilant workers exposed to ethylene oxide: updated cohort study findings 1972–2006. Int J Environ Res Public Health, 8(6), 2009–2019. PMCID: PMC3138010.

Milanese TR, Hartmann LC, Sellers TA, et al. 2006. Age-related lobular involution and risk of breast cancer. J Natl Cancer Inst, 98(22), 1600–1607. PMID: 17105983.

Millikan R, Newman B, Tse C, et al. 2008. Epidemiology of basal-like breast cancer. Breast Cancer Res Treat, 109, 123–139. PMCID: PMC2443103.

Mirick DK, Davis S, and Thomas DB. 2002. Antiperspirant use and the risk of breast cancer. J Natl Cancer Inst, 94(20), 1578–1580. PMID: 12381712.

Missmer SA, Eliassen AH, Barbieri RL, and Hankinson SE. 2004. Endogenous estrogen, androgen, and progesterone concentrations and breast cancer risk among postmenopausal women. J Natl Cancer Inst, 96(24), 1856–1865. PMID: 15601642.

Moore SC, Matthews CE, Ou Shu X, et al. 2016. Endogenous estrogens, estrogen metabolites, and breast cancer risk in postmenopausal Chinese women. J Natl Cancer Inst, 108(10), djw103. PMCID: PMC4893973.

Morrow M, Schnitt SJ, and Norton L. 2015. Current management of lesions associated with an increased risk of breast cancer. Nat Rev Clin Oncol, 12(4), 227–238. PMID: 25622978.

Morton LM, Onel K, Curtis RE, Hungate EA, and Armstrong GT. 2014. The rising incidence of second cancers: patterns of occurrence and identification of risk factors for children and adults. Am Soc Clin Oncol Educ Book, e57–67. PMID: 24857148.

Moskowitz CS, Chou JF, Wolden SL, et al. 2014. Breast cancer after chest radiation therapy for childhood cancer. J Clin Oncol, 32(21), 2217–2223. PMCID: PMC4100937.

Moyer VA, and Force USPST. 2013. Medications to decrease the risk for breast cancer in women: recommendations from the U.S. Preventive Services Task Force recommendation statement. Ann Intern Med, 159(10), 698–708. PMID: 24061412.

Muller C, Wallaschofski H, Brabant G, et al. 2014. The association between IGF-I/IGFBP-3 and subclinical end points: epidemiology faces the limits. J Clin Endocrinol Metab, 99(8), 2804–2812. PMID: 24823454.

Munksgaard PS, and Blaakaer J. 2011. The association between endometriosis and gynecological cancers and breast cancer: a review of epidemiological data. Gynecol Oncol, 123(1), 157–163. PMID: 21742370.

Munoz D, Near AM, van Ravesteyn NT, et al. 2014. Effects of screening and systemic adjuvant therapy on ER-specific US breast cancer mortality. J Natl Cancer Inst, 106(11), dju289. PMCID: PMC4271026.

Munsell MF, Sprague BL, Berry DA, Chisholm G, and Trentham-Dietz A. 2014. Body mass index and breast cancer risk according to postmenopausal estrogen-progestin use and hormone receptor status. Epidemiol Rev, 36, 114–136. PMCID: PMC3873844.

Neilson HK, Friedenreich CM, Brockton NT, and Millikan RC. 2009. Physical activity and postmenopausal breast cancer: proposed biologic mechanisms and areas for future research. Cancer Epidemiol Biomarkers Prev, 18(1), 11–27. PMID: 19124476.

Nelson HD, Pappas M, Zakher B, et al. 2014. Risk assessment, genetic counseling, and genetic testing for BRCA-related cancer in women: a systematic review to update the U.S. Preventive Services Task Force recommendation. Ann Intern Med, 160(4), 255–266. PMID: 24366442.

Neuhouser ML, Aragaki AK, Prentice RL, et al. 2015. Overweight, obesity, and postmenopausal invasive breast cancer risk: a secondary analysis of the women’s health initiative randomized clinical trials. JAMA Oncol, 1(5), 611–621. PMID: 26182172.

Nichols HB, Baird DD, Stanczyk FZ, et al. 2015a. Anti-mullerian hormone concentrations in premenopausal women and breast cancer risk. Cancer Prev Res (Phila), 8(6), 528–534. PMCID: PMC4452465.

Nichols HB, DeRoo LA, Scharf DR, and Sandler DP. 2015b. Risk-benefit profiles of women using tamoxifen for chemoprevention. J Natl Cancer Inst 107(1), dju354. PMCID: PMC4296193.

Norat T, Aune D, Chan D, and Romaguera D. 2014. Fruits and vegetables: updating the epidemiologic evidence for the WCRF/AICR lifestyle recommendations for cancer prevention. Cancer Treat Res, 159, 35–50. PMID: 24114473.

Norum JH, Andersen K, and Sorlie T. 2014. Lessons learned from the intrinsic subtypes of breast cancer in the quest for precision therapy. Br J Surg, 101(8), 925–938. PMID: 24849143.

Nyante SJ, Sherman ME, Pfeiffer RM, et al. 2015. Prognostic significance of mammographic density change after initiation of tamoxifen for ER-positive breast cancer. J Natl Cancer Inst, 107(3), dju425. PMCID: PMC4334825.

Oh H, Khodr ZG, Sherman ME, et al. 2016. Relation of serum estrogen metabolites with terminal duct lobular unit involution among women undergoing diagnostic image-guided breast biopsy. Horm Cancer, 7(5-6), 305–315. PMCID: PMC5088060.

Oncoarray. 2016. Latest oncoarray data relation. Palmer JR, Boggs DA, Wise LA, et al. 2011. Parity and lactation in to estrogen receptor negative breast cancer in African American women. Cancer Epidemiol Biomarkers Prev, 20(9), 1883–1891. PMCID: PMC317917.

Palmer JR, Viscidi E, Troester MA, et al. 2014. Parity, lactation, and breast cancer subtypes in African American women: results from the AMBER Consortium. J Natl Cancer Inst, 106(10). PMCID: PMC4271113.

Parise C, and Caggiano V. 2014. Disparities in the risk of the ER/PR/HER2 breast cancer subtypes among Asian Americans in California. Cancer Epidemiol, 38(5), 556–562. PMID: 25172158.

Peplonska B, Stewart P, Szeszenia-Dabrowska N, et al. 2010. Occupational exposure to organic solvents and breast cancer in women. Occup Environ Med, 67(11), 722–729. PMCID: PMC2891315.

Pettersson A, Graff RE, Ursin G, et al. 2014. Mammographic density phenotypes and risk of breast cancer: a meta-analysis. J Natl Cancer Inst, 106(5), dju078. PMCID: PMC4568991.

Pfeiffer RM, Park Y, Kreimer AR, et al. 2013. Risk prediction for breast, endometrial, and ovarian cancer in white women aged 50 y or older: derivation and validation from population-based cohort studies. PLoS Med, 10(7), e1001492. PMCID: PMC3728034.

Phipps AI, Buist DS, Malone KE, et al. 2011a. Reproductive history and risk of three breast cancer subtypes defined by three biomarkers. Cancer Causes Control, 22(3), 399–405. PMCID: PMC3042513.

Phipps AI, Chlebowski RT, Prentice R, et al. 2011b. Body size, physical activity, and risk of triple-negative and estrogen receptor-positive breast cancer. Cancer Epidemiol Biomarkers Prev, 20(3), 454–463. PMCID: PMC3064558.

Pike MC, Krailo MD, Henderson BE, Casagrande JT, and Hoel DG. 1983. ‘Hormonal’ risk factors, ‘breast tissue age’ and the age-incidence of breast cancer. Nature, 303(5920), 767–770. PMID: 6866078.

Pirie K, Beral V, Peto R, et al. 2008. Passive smoking and breast cancer in never smokers: prospective study and meta-analysis. Int J Epidemiol, 37(5), 1069–1079. PMID: 18544575.

Pollak M. 2008. Insulin and insulin-like growth factor signalling in neoplasia. Nat Rev Cancer, 8(12), 915–928. PMID: 19029956.

Ponder BA, Antoniou A, Dunning A, Easton DF, and Pharoah PD. 2005. Polygenic inherited predisposition to breast cancer. Cold Spring Harb Symp Quant Biol, 70, 35–41. PMID: 16869736.

Portschy PR, Abbott AM, Burke EE, et al. 2015. Perceptions of contralateral breast cancer risk: a prospective, longitudinal study. Ann Surg Oncol, 22(12), 3846–3852. PMID: 25762480.

Press DJ, Sullivan-Halley J, Ursin G, et al. 2011. Breast cancer risk and ovariectomy, hysterectomy, and tubal sterilization in the women’s contraceptive (p.886) and reproductive experiences study. Am J Epidemiol, 173(1), 38–47. PMCID: PMC3025644.

Price ER, Hargreaves J, Lipson JA, et al. 2013. The California breast density information group: a collaborative response to the issues of breast density, breast cancer risk, and breast density notification legislation. Radiology, 269(3), 887–892. PMID: 24023072.

Protani M, Coory M, and Martin JH. 2010. Effect of obesity on survival of women with breast cancer: systematic review and meta-analysis. Breast Cancer Res Treat, 123(3), 627–635. PMID: 20571870.

Quante AS, Whittemore AS, Shriver T, Strauch K, and Terry MB. 2012. Breast cancer risk assessment across the risk continuum: genetic and nongenetic risk factors contributing to differential model performance. Breast Cancer Res, 14(6), R144. PMCID: PMC4053132.

Ravdin PM, Cronin KA, Howlader N, et al. 2007. The decrease in breast-cancer incidence in 2003 in the United States. N Engl J Med, 356(16), 1670–1674. PMID: 17442911.

Reigstad MM, Larsen IK, Myklebust TA, et al. 2015. Risk of breast cancer following fertility treatment--a registry based cohort study of parous women in Norway. Int J Cancer, 136(5), 1140–1148. PMCID: PMC4268160.

Reiner AS, John EM, Brooks JD, et al. 2013. Risk of asynchronous contralateral breast cancer in noncarriers of BRCA1 and BRCA2 mutations with a family history of breast cancer: a report from the Women’s Environmental Cancer and Radiation Epidemiology Study. J Clin Oncol, 31(4), 433–439. PMCID: PMC3731919.

Renehan AG, Tyson M, Egger M, Heller RF, and Zwahlen M. 2008. Body-mass index and incidence of cancer: a systematic review and meta-analysis of prospective observational studies. Lancet, 371(9612), 569–578. PMID: 18280327.

Rockhill B, Spiegelman D, Byrne C, Hunter DJ, and Colditz GA. 2001. Validation of the Gail et al. model of breast cancer risk prediction and implications for chemoprevention. J Natl Cancer Inst, 93(5), 358–366. PMID: 11238697.

Romieu I, Scoccianti C, Chajes V, et al. 2015. Alcohol intake and breast cancer in the European prospective investigation into cancer and nutrition. Int J Cancer, 137(8), 1921–1930. PMID: 25677034.

Rosebrock A, Caban JJ, Figueroa J, et al. 2013. Quantitative analysis of TDLUs using adaptive morphological shape techniques. Proc SPIE Int Soc Opt Eng, vol.8676. PMCID: PMC4339025.

Rosenberg L, Palmer JR, Bethea TN, et al. 2014. A prospective study of physical activity and breast cancer incidence in African-American women. Cancer Epidemiol Biomarkers Prev, 23(11), 2522–2531. PMCID: PMC4221421.

Rosenberg PS, Barker KA, and Anderson WF. 2015. Estrogen receptor status and the future burden of invasive and in situ breast cancers in the United States. J Natl Cancer Inst, 107(9), djv159. PMID: 26063794.

Rosenfeld RG. 2003. Insulin-like growth factors and the basis of growth. N Engl J Med, 349(23), 2184–2186. PMID: 14657423.

Rowzee AM, Lazzarino DA, Rota L, Sun Z, and Wood TL. 2008. IGF ligand and receptor regulation of mammary development. J Mammary Gland Biol Neoplasia, 13(4), 361–370. PMCID: PMC2665296.

Russo J, Hu YF, Yang X, and Russo IH. 2000. Developmental, cellular, and molecular basis of human breast cancer. J Natl Cancer Inst Monogr 27, 17–37. PMID: 10963618.

Russo J, and Russo IH. 2004. Development of the human breast. Maturitas, 49(1), 2–15. PMID: 15351091.

Said SM, Rizzo W, Degnim AC, et al. 2014. Risk of developing breast cancer in patients with flat epithelial atypia (FEA) in benign breast biopsies. Modern Pathology, 27 (Supp 2s) USCAP 103, 79A.

Santen RJ, and Mansel R. 2005. Benign breast disorders. N Engl J Med, 353(3), 275–285. PMID: 16034013.

Saxena T, Lee E, Henderson KD, et al. 2010. Menopausal hormone therapy and subsequent risk of specific invasive breast cancer subtypes in the California Teachers Study. Cancer Epidemiol Biomarkers Prev, 19(9), 2366–2378. PMCID: PMC2936672.

Schernhammer ES, Tworoger SS, Eliassen AH, et al. 2007. Body shape throughout life and correlations with IGFs and GH. Endocr Relat Cancer, 14(3), 721–732. PMID: 17914102.

SEER. 2015. Surveillance, Epidemiology, and End Results (SEER) Program Populations (1969–2013) (www.seer.cancer.gov/popdata), National Cancer Institute, DCCPS, Surveillance Research Program, Surveillance Systems Branch, released January 2015.

Seitz HK, Pelucchi C, Bagnardi V, and La Vecchia C. 2012. Epidemiology and pathophysiology of alcohol and breast cancer: update 2012. Alcohol Alcohol, 47(3), 204–212. PMID: 22459019.

Sexton KR, Franzini L, Day RS, et al. 2011. A review of body size and breast cancer risk in Hispanic and African American women. Cancer, 117(23), 5271–5281. PMID: 21598244.

Shafrir AL, Zhang X, Poole EM, Hankinson SE, and Tworoger SS. 2014. The association of reproductive and lifestyle factors with a score of multiple endogenous hormones. Horm Cancer, 5(5), 324–335. PMCID: PMC4167477.

Shen D, Mao W, Liu T, et al. 2014. Sedentary behavior and incident cancer: a meta-analysis of prospective studies. PLoS One, 9(8), e105709. PMCID: PMC4143275.

Sherman ME, Mies C, and Gierach GL. 2014. Opportunities for molecular epidemiological research on ductal carcinoma in-situ and breast carcinogenesis: interdisciplinary approaches. Breast Dis, 34(3), 105–116. PMCID: PMC4558907.

Sickles EA, D’Orsi CJ, Bassett LW, et al. 2013. ACR BiRADS Atlas, Breast Imaging Reporting and Data System. Reston, VA: American College of Radiology.

Siegel RL, Miller KD, and Jemal A. 2015. Cancer statistics, 2015. CA Cancer J Clin, 65(1), 5–29. PMID: 25559415.

Siiteri PK. 1987. Adipose tissue as a source of hormones. Am J Clin Nutr, 45(1), 277–282. PMID: 3541569.

Sineshaw HM, Gaudet M, Ward EM, et al. 2014. Association of race/ethnicity, socioeconomic status, and breast cancer subtypes in the National Cancer Data Base (2010–2011). Breast Cancer Res Treat, 145(3), 753–763. PMID: 24794028.

Soini T, Hurskainen R, Grenman S, et al. 2014. Cancer risk in women using the levonorgestrel-releasing intrauterine system in Finland. Obstet Gynecol, 124(2 Pt 1), 292–299. PMID: 25004338.

Sowers MR, Eyvazzadeh AD, McConnell D, et al. 2008. Anti-mullerian hormone and inhibin B in the definition of ovarian aging and the menopause transition. J Clin Endocrinol Metab, 93(9), 3478–3483. PMCID: PMC2567855.

Stone J, Thompson DJ, Dos Santos Silva I, et al. 2015. Novel associations between common breast cancer susceptibility variants and risk-predicting mammographic density measures. Cancer Res, 75(12), 2457–2467. PMCID: PMC4470785.

Strom BL, Berlin JA, Weber AL, et al. 2004. Absence of an effect of injectable and implantable progestin-only contraceptives on subsequent risk of breast cancer. Contraception, 69(5), 353–360. PMID: 15105056.

Su HY, and Cheng WT. 2004. Increased milk yield in transgenic mice expressing insulin-like growth factor 1. Anim Biotechnol, 15(1), 9–19. PMID: 15248597.

Sun X, Gierach GL, Sandhu R, et al. 2013. Relationship of mammographic density and gene expression: analysis of normal breast tissue surrounding breast cancer. Clin Cancer Res, 19(18), 4972–4982. PMCID: PMC4073678.

Sun X, Sandhu R, Figueroa JD, et al. 2014. Benign breast tissue composition in breast cancer patients: association with risk factors, clinical variables, and gene expression. Cancer Epidemiol Biomarkers Prev, 23(12), 2810–2818. PMCID: PMC4278568.

Suzuki R, Orsini N, Saji S, Key TJ, and Wolk A. 2009. Body weight and incidence of breast cancer defined by estrogen and progesterone receptor status: a meta-analysis. Int J Cancer, 124(3), 698–712. PMID: 18988226.

Takkouche B, Etminan M, and Montes-Martinez A. 2005. Personal use of hair dyes and risk of cancer: a meta-analysis. JAMA, 293(20), 2516–2525. PMID: 15914752.

Takkouche B, Regueira-Mendez C, and Etminan M. 2008. Breast cancer and use of nonsteroidal anti-inflammatory drugs: a meta-analysis. J Natl Cancer Inst, 100(20), 1439–1447. PMID: 18840819.

Tamimi RM, Colditz GA, Hazra A, et al. 2012. Traditional breast cancer risk factors in relation to molecular subtypes of breast cancer. Breast Cancer Res Treat, 131(1), 159–167. PMCID: PMC3237947.

Thomson CA, McCullough ML, Wertheim BC, et al. 2014. Nutrition and physical activity cancer prevention guidelines, cancer risk, and mortality in the women’s health initiative. Cancer Prev Res (Phila), 7(1), 42–53. PMCID: PMC4090781.

Tice JA, Cummings SR, Smith-Bindman R, et al. 2008. Using clinical factors and mammographic breast density to estimate breast cancer risk: development and validation of a new predictive model. Ann Intern Med, 148(5), 337–347. PMCID: PMC2674327.

Tice JA, Cummings SR, Ziv E, and Kerlikowske K. 2005. Mammographic breast density and the Gail model for breast cancer risk prediction in a screening population. Breast Cancer Res Treat, 94(2), 115–122. PMID: 16261410.

Tice JA, Miglioretti DL, Li CS, et al. 2015. Breast density and benign breast disease: risk assessment to identify women at high risk of breast cancer. J Clin Oncol, 33(28), 3137–3143. PMCID: PMC4582144.

Tikk K, Sookthai D, Johnson T, et al. 2014. Circulating prolactin and breast cancer risk among pre- and postmenopausal women in the EPIC cohort. Ann Oncol, 25(7), 1422–1428. PMID: 24718887.

(p.887) Titus-Ernstoff L, Hatch EE, Hoover RN, et al. 2001. Long-term cancer risk in women given diethylstilbestrol (DES) during pregnancy. Br J Cancer, 84(1), 126–133. PMCID: PMC2363605.

Torre LA, Bray F, Siegel RL, et al. 2015. Global cancer statistics, 2012. CA Cancer J Clin, 65(2), 87–108. PMID: 25651787.

Trabert B, Falk RT, Stanczyk FZ, et al. 2015. Reproducibility of a novel assay to measure serum progesterone pathway metabolites using liquid chromatography-tandem mass spectrometry. Horm Molec Biol Clin Invest, 23(3), 79–84. PMCID: PMC4966666.

Tsilidis KK, Kasimis JC, Lopez DS, Ntzani EE, and Ioannidis JP. 2015. Type 2 diabetes and cancer: umbrella review of meta-analyses of observational studies. BMJ, 350, g7607. PMID: 25555821.

Tworoger SS, Eliassen AH, Zhang X, et al. 2013. A 20-year prospective study of plasma prolactin as a risk marker of breast cancer development. Cancer Res, 73(15), 4810–4819. PMCID: PMC3738582.

Tworoger SS, Rice MS, Rosner BA, et al. 2015. Bioactive prolactin levels and risk of breast cancer: a nested case-control study. Cancer Epidemiol Biomarkers Prev, 24(1), 73–80. PMCID: PMC4294963.

Tworoger SS, Zhang X, Eliassen AH, et al. 2014. Inclusion of endogenous hormone levels in risk prediction models of postmenopausal breast cancer. J Clin Oncol, 32(28), 3111–3117. PMCID: PMC4171356.

Tyrer J, Duffy SW, and Cuzick J. 2004. A breast cancer prediction model incorporating familial and personal risk factors. Stat Med, 23(7), 1111–1130. PMID: 15057881.

US Department of Health and Human Services. 2014. The Health Consequences of Smoking: 50 Years of Progress: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.

Ursin G, Lillie EO, Lee E, et al. 2009. The relative importance of genetics and environment on mammographic dnsity. Cancer Epidemiol Biomarkers Prev, 18(1), 102–112. PMID: 19124487.

Vachon CM, Pankratz VS, Scott CG, et al. 2015. The contributions of breast density and common genetic variation to breast cancer risk. J Natl Cancer Inst, 107(5). PMID: 25745020.

Vachon CM, Scott CG, Fasching PA, et al. 2012. Common breast cancer susceptibility variants in LSP1 and RAD51L1 are associated with mammographic density measures that predict breast cancer risk. Cancer Epidemiol Biomarkers Prev, 21(7), 1156–1166. PMCID: PMC3569092.

Venkitaraman R. 2010. Lobular neoplasia of the breast. Breast J, 16(5), 519–528. PMID: 20701604.

Vera-Ramirez L, Ramirez-Tortosa MC, Sanchez-Rovira P, et al. 2013. Impact of diet on breast cancer risk: a review of experimental and observational studies. Crit Rev Food Sci Nutr, 53(1), 49–75. PMID: 23035920.

Verheus M, Peeters PH, Rinaldi S, et al. 2006. Serum C-peptide levels and breast cancer risk: results from the European Prospective Investigation into Cancer and Nutrition (EPIC). Int J Cancer, 119(3), 659–667. PMID: 16572422.

Verma S, Miles D, Gianni L, et al. 2012. Trastuzumab emtansine for HER2-positive advanced breast cancer. N Engl J Med, 367(19), 1783–1791. PMID: 23020162.

Villeneuve S, Fevotte J, Anger A, et al. 2011. Breast cancer risk by occupation and industry: analysis of the CECILE study, a population-based case-control study in France. Am J Ind Med, 54(7), 499–509. PMID: 21472744.

Virnig BA, Wang SY, Shamilyan T, Kane RL, and Tuttle TM. 2010. Ductal carcinoma in situ: risk factors and impact of screening. J Natl Cancer Inst Monogr, 2010(41), 113–116. PMID: 20956813.

Visser JA, Schipper I, Laven JS, and Themmen AP. 2012. Anti-Mullerian hormone: an ovarian reserve marker in primary ovarian insufficiency. Nat Rev Endocrinol, 8(6), 331–341. PMID: 22231848.

Visvader JE, and Stingl J. 2014. Mammary stem cells and the differentiation hierarchy: current status and perspectives. Genes Dev, 28(11), 1143–1158. PMCID: PMC4052761.

Visvanathan K, Hurley P, Bantug E, et al. 2013. Use of pharmacologic interventions for breast cancer risk reduction: American Society of Clinical Oncology clinical practice guideline. J Clin Oncol, 31(23), 2942–2962. PMID: 23835710.

Vogel VG, Costantino JP, Wickerham DL, 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. PMCID: PMC2935331.

Vrieling A, Buck K, Kaaks R, and Chang-Claude J. 2010. Adult weight gain in relation to breast cancer risk by estrogen and progesterone receptor status: a meta-analysis. Breast Cancer Res Treat, 123(3), 641–649. PMID: 20711809.

Wang J, Lee IM, Tworoger SS, et al. 2015. Plasma C-reactive protein and risk of breast cancer in two prospective studies and a meta-analysis. Cancer Epidemiol Biomarkers Prev, 24(8), 1199–1206. PMCD: PMC4526401.

Warwick J, Birke H, Stone J, et al. 2014. Mammographic breast density refines Tyrer-Cuzick estimates of breast cancer risk in high-risk women: findings from the placebo arm of the International Breast Cancer Intervention Study I. Breast Cancer Res, 16(5), 451. PMCID: PMC4303130.

WCRF/AICR. 2008. WCRF/AICR Systematic Literature Review Continuous Update Report: The Associations Between Food, Nutrition and Physical Activity and the Risk of Breast Cancer. London: Imperial College.

WCRF/AICR. 2014. Diet, Nutrition, Physical Activity, and Breast Cancer Survivors. World Cancer Research Fund International / American Institute for Cancer Research Continuous Update Project Report. London: World Cancer Research Fund International.

Weaver DL, Rosenberg RD, Barlow WE, et al. 2006. Pathologic findings from the Breast Cancer Surveillance Consortium: population-based outcomes in women undergoing biopsy after screening mammography. Cancer, 106(4), 732–742. PMID: 16411214.

Weigelt B, Geyer FC, and Reis-Filho JS. 2010. Histological types of breast cancer: how special are they? Mol Oncol, 4(3), 192–208. PMID: 20452298.

Weitzel JN, Blazer KR, MacDonald DJ, Culver JO, and Offit K. 2011. Genetics, genomics, and cancer risk assessment: state of the art and future directions in the era of personalized medicine. CA Cancer J Clin, 61(5), 327–359. PMCID: PMC3346864.

Welch HG, and Black WC. 1997. Using autopsy series to estimate the disease “reservoir” for ductal carcinoma in situ of the breast: how much more breast cancer can we find? Ann Intern Med, 127(11), 1023–1028. PMID: 9412284.

Welsh AW, Moeder CB, Kumar S, et al. 2011. Standardization of estrogen receptor measurement in breast cancer suggests false-negative results are a function of threshold intensity rather than percentage of positive cells. J Clin Oncol, 29(22), 2978–2984. PMCID: PMC3157961.

Wiebe JP. 2006. Progesterone metabolites in breast cancer. Endocr Relat Cancer, 13(3), 717–738. PMID: 16954427.

Wiebe JP, Beausoleil M, Zhang G, and Cialacu V. 2010. Opposing actions of the progesterone metabolites, 5alpha-dihydroprogesterone (5alphaP) and 3alpha-dihydroprogesterone (3alphaHP) on mitosis, apoptosis, and expression of Bcl-2, Bax and p21 in human breast cell lines. J Steroid Biochem Mol Biol, 118(1–2), 125–132. PMID: 19931389.

Wiebe JP, and Lewis MJ. 2003. Activity and expression of progesterone metabolizing 5alpha-reductase, 20alpha-hydroxysteroid oxidoreductase and 3alpha(beta)-hydroxysteroid oxidoreductases in tumorigenic (MCF-7, MDA-MB-231, T-47D) and nontumorigenic (MCF-10A) human breast cancer cells. BMC Cancer, 3, 9. PMCID: PMC154104.

Work ME, Reimers LL, Quante AS, et al. 2014. Changes in mammographic density over time in breast cancer cases and women at high risk for breast cancer. Int J Cancer, 135(7), 1740–1744. PMCID: PMC4107003.

World Health Organization. 2003. World Health Organization Classification of Tumors. 3rd ed. Lyon: IARC Press.

World Health Organization. 2012. World Health Organization Classification of Tumors of the Breast. 4th ed. Lyon: IARC Press.

Wu Y, Zhang D, and Kang S. 2013. Physical activity and risk of breast cancer: a meta-analysis of prospective studies. Breast Cancer Res Treat, 137(3), 869–882. PMID: 23274845.

Xue F, and Michels KB. 2007. Intrauterine factors and risk of breast cancer: a systematic review and meta-analysis of current evidence. Lancet Oncol, 8(12), 1088–1100. PMID: 18054879.

Xue F, Willett WC, Rosner BA, Hankinson SE, and Michels KB. 2011. Cigarette smoking and the incidence of breast cancer. Arch Intern Med, 171(2), 125–133. PMCID: PMC3131146.

Yaffe M. 2008. Mammographic density: measurement of mammographic density. Breast Cancer Res, 10(3), 209. PMCID: PMC2481498.

Yager JD, and Davidson NE. 2006. Estrogen carcinogenesis in breast cancer. N Engl J Med, 354(3), 270–282. PMID: 16421368.

Yaghjyan L, Colditz GA, Collins LC, et al. 2011. Mammographic breast density and subsequent risk of breast cancer in postmenopausal women according to tumor characteristics. J Natl Cancer Inst, 103(15), 1179–1189. PMCID: PMC3149043.

Yang XR, Chang-Claude J, Goode EL, et al. 2011. Associations of breast cancer risk factors with tumor subtypes: a pooled analysis from the Breast Cancer Association Consortium studies. J Natl Cancer Inst, 103(3), 250–263. PMCID: PMC3107570.

Yang XR, Sherman ME, Rimm DL, et al. 2007. Differences in risk factors for breast cancer molecular subtypes in a population-based study. Cancer Epidemiol Biomarkers Prev, 16(3), 439–443. PMID: 17372238.

(p.888) Yang Y, Zhang F, Skrip L, Wang Y, and Liu S. 2013. Lack of an association between passive smoking and incidence of female breast cancer in non-smokers: evidence from 10 prospective cohort studies. PLoS One, 8(10), e77029. PMCID: PMC3800073.

Zhang X, Tworoger SS, Eliassen AH, and Hankinson SE. 2013. Postmenopausal plasma sex hormone levels and breast cancer risk over 20 years of follow-up. Breast Cancer Res Treat, 137(3), 883–892. PMCID: PMC3582409.

Zheng W, Zhang B, Cai Q, et al. 2013. Common genetic determinants of breast-cancer risk in East Asian women: a collaborative study of 23 637 breast cancer cases and 25 579 controls. Hum Mol Genet, 22(12), 2539–2550. PMCID: PMC3658167.

Zhou Y, Zhao H, and Peng C. 2015. Association of sedentary behavior with the risk of breast cancer in women: update meta-analysis of observational studies. Ann Epidemiol, 25(9), 687–697. PMID: 26099193.

Ziegler RG, Fuhrman BJ, Moore SC, and Matthews CE. 2015. Epidemiologic studies of estrogen metabolism and breast cancer. Steroids, 99(Pt A), 67–75. PMID: 25725255.