Background:

We investigated the association between reproductive risk factors and breast cancer subtype in Black women. On the basis of the previous literature, we hypothesized that the relative prevalence of specific breast cancer subtypes might differ according to reproductive factors.

Methods:

We conducted a pooled analysis of 2,188 (591 premenopausal, 1,597 postmenopausal) Black women with a primary diagnosis of breast cancer from four studies in the southeastern United States. Breast cancers were classified by clinical subtype. Case-only polytomous logistic regression models were used to estimate ORs and 95% confidence intervals (CI) for HER2+ and triple-negative breast cancer (TNBC) status in relation to estrogen receptor–positive (ER+)/HER2 status (referent) for reproductive risk factors.

Results:

Relative to women who had ER+/HER2 tumors, women who were age 19–24 years at first birth (OR, 1.78; 95% CI, 1.22–2.59) were more likely to have TNBC. Parous women were less likely to be diagnosed with HER2+ breast cancer and more likely to be diagnosed with TNBC relative to ER+/HER2 breast cancer. Postmenopausal parous women who breastfed were less likely to have TNBC [OR, 0.65 (95% CI, 0.43–0.99)].

Conclusions:

This large pooled study of Black women with breast cancer revealed etiologic heterogeneity among breast cancer subtypes.

Impact:

Black parous women who do not breastfeed are more likely to be diagnosed with TNBC, which has a worse prognosis, than with ER+/HER2 breast cancer.

Racial differences in breast cancer occurrence are well documented. Current overall breast cancer incidence rates in the United States are higher among White women (130.8/100,000) than among Black women (126.7/100,000), although Black women have the highest incidence rates before age 40 years (1). Yet despite a somewhat lower overall lifetime risk of breast cancer, population-based data show that Black women experience 40% higher mortality from breast cancer than White women at every age and are more likely to be diagnosed with fast growing and late-stage breast cancer (1, 2).

Characterization of invasive breast cancers by IHC markers has identified distinct tumor subtypes, including estrogen receptor–positive (ER+) and/or progesterone receptor–positive (PR+), HER2-overexpressing (ER, PR, HER2+), and triple-negative breast cancer (TNBC; ER, PR, HER2). The proportional distribution of breast cancer subtypes tends to vary by race. White women have the highest rates of ER+ breast cancer and Black women have the highest rates of ER breast cancer in every age group (1, 3). Moreover, TNBC, a fast growing subtype of breast cancer, is more common among Black women than among White women and has a worse prognosis and limited treatment options compared with other subtypes (4). Less racial/ethnic variation has been observed in the HER2+ subtypes.

Accumulating epidemiologic evidence demonstrates that many of the established reproductive risk factors for breast cancer identified among White women, including early age at menarche, parity, and late age at first birth, may associate differently with breast cancer in Black women or by breast tumor subtype (5–8). Moreover, studies on the prevalence of breast cancer risk factors have shown that Black women, compared with their White counterparts, are more likely to report younger age at menarche, younger age at first full-term pregnancy and greater number of births, and lower rates of breastfeeding and use of hormone replacement therapy (HRT; ref. 9).

We utilized information from a large pooled dataset of Black women with a primary diagnosis of breast cancer from four studies in the southeastern United States to examine the distribution of ER, PR, and HER2 in breast tumors and whether reproductive risk factors associate differentially with different breast cancer subtypes.

All study protocols adhered to ethical guidelines and were approved by the Institutional Review Boards of Vanderbilt University Medical Center (Nashville, TN), Meharry Medical College (Nashville, TN), and/or University of South Florida (Tampa, FL), the respective statewide Departments of Health and cancer registries and, where applicable, participating local hospitals.

Southern Tri-State Breast Health Study

Breast cancer case ascertainment

The Southern Tri-State Breast Health Study (STSBHS) was a population-based, case-only study of incident breast cancer designed to investigate breast cancer genetic risk variants relevant for Black women. From May 2012 through June 2018, the statewide cancer registries of Tennessee (TN), South Carolina (SC), and Georgia (GA, also SEER registry) were utilized to identify Black women newly diagnosed with primary breast cancer between the ages of 25 and 75 years old, with no prior history of cancer other than nonmelanoma skin cancer, who had a residential telephone, spoke English and were able to provide consent to the study. Diagnosis dates differed by registry for the current analyses: TN—September 2009–March 2014, SC—September 2009–December 2012, and GA—January 2011–December 2012.

Initial approach letters were mailed to potential participants by the cancer registries, and to physicians of potential participants in SC only. Procedures for recruitment differed across the three states: passive approval in TN, active verbal approval in SC, and active written or verbal consent to release contact information in GA. Once recruited, breast cancer cases completed a telephone interview by trained interviewers or online questionnaire. Because TN did not require a consent to be contacted, the response rate to the interview was lower (54.7%), than among women in SC (70.5%), and GA (72.3%).

Information on demographic, anthropometric, medical, reproductive, lifestyle, and other reproductive risk factors was ascertained using a structured questionnaire. Black race was identified on the basis of self-report. At the end of the interview, women who indicated they agreed to provide a DNA sample were mailed a saliva sample collection kit along with a written informed consent seeking permission to use, store, and share biospecimens (saliva sample and tissue), to access tumor tissue and medical record/cancer registry records, and to allow future contact. Once the sample was received, women were sent a $15 incentive. Through June of 2018, the STSBHS had successfully recruited 1,164 Black women with incident invasive breast cancer (325 GA, 356 SC, 483 TN) for the interview.

Tumor marker ascertainment

From the cancer registry and pathology records of breast cancer cases, data were extracted related to disease diagnosis (e.g., types and results of diagnostic tests, histopathology, stage and grade of cancers, and ER/PR/HER2 status when available). On the basis of registry data and supplemented by pathology reports when available, tumor ER and PR status was available for 99% (n = 1,158) of women and HER2 status was available for 90% (n = 1,057) of women. Approximately 167 women (14%) with ER/PR status available from medical records/cancer registry had a rarer subtype (ER+/PR or ER/PR+) which required confirmation by hospital laboratories.

Southern Community Cohort Study

The Southern Community Cohort Study (SCCS) is an ongoing, prospective cohort study focused on health disparities with regards to race, sex, socioeconomic characteristics, and other factors (ref. 10); www.southerncommunitystudy.org). The cohort comprises nearly 86,000 residents of 12 southern states (AL, AR, FL, GA, KY, LA, MS, NC, SC, TN, VA, WV), 60% women and two-thirds Black, who were enrolled between 2002 and 2009, most (86%) at participating Community Health Centers (CHC), which provide primary health and preventive services to medically underserved and low-income populations. The remaining 14% of participants were enrolled through mail-based general population sampling in the same 12 states. Participants in the SCCS were required to speak English, be between 40 and 79 years old at enrollment, and not have undergone treatment for cancer during the past year. Upon enrollment into the SCCS, written informed consent was obtained from each participant.

Upon entry into the SCCS, participants were administered a baseline questionnaire (available at www.southerncommunitystudy.org), via computer-assisted personal interview for CHC participants and self-administered mailed questionnaire for general population participants. The questionnaire ascertained information about demographic, socioeconomic, and anthropometric characteristics, reproductive and menstrual history, personal and family medical history including use of oral contraceptives (OC) and HRT, tobacco and alcohol use, medication use, and other factors. Black race was identified on the basis of self-report.

Breast cancer case ascertainment

Follow-up of the cohort for ascertainment of incident cancer was performed through annual linkage to state cancer registries in the study enrollment states. The current analysis is a case-only study nested within the SCCS, based on incident cases of breast cancer diagnosed among Black female SCCS participants after the date of enrollment. Breast cancers were defined by International Classification of Disease for Oncology Codes (ICD-O-3) C50.0-C50.6 and C50.8-C50.9. Through June of 2018, the SCCS had identified 788 Black women with incident invasive breast cancer.

Tumor marker ascertainment

Information on tumor ER, PR, and HER2 status was abstracted, where available, from the cancer registry data supplemented by available pathology reports and medical records. Tumor ER and PR status was available for 92% (n = 727) of women and HER2 status was available for 73% (n = 575) of women. Approximately 110 women (15%) with ER/PR status available from medical records/cancer registry had a rarer subtype (ER+/PR or ER/PR+) which required confirmation by hospital laboratories.

Nashville Breast Health Study

Breast cancer case ascertainment

The Nashville Breast Health Study (NBHS) was a population-based case-control study conducted in and around the Nashville Metropolitan area (11, 12). Breast cancer cases were identified through a rapid case-ascertainment system in conjunction with five local hospitals and the Tennessee Cancer Registry. Eligible cases were women diagnosed between 2001 and 2011 with primary breast cancer who were between the ages of 25 and 75 years and had no prior history of cancer other than nonmelanoma skin cancer. Written informed consent was obtained from participants. Trained interviewers conducted structured telephone interviews to ascertain information on demographic characteristics, known and potential breast cancer risk factors, reproductive and lifestyle factors. Black race was identified on the basis of self-report. Through June of 2018, the NBHS had successfully recruited 416 Black women with incident invasive breast cancer for the interview.

Tumor marker ascertainment

From the medical and pathology records of breast cancer cases, data were extracted related to disease diagnosis (e.g., types and results of diagnostic tests, histopathology, stage and grade of cancers, including ER/PR/HER2 status when available), and treatment. On the basis of registry data and supplemented by pathology reports when available, tumor ER and PR status was available for 81% (n = 335) of women and HER2 status was available for 62% (n = 259) of women. Approximately 54 women (15%) with ER/PR status available from medical records/cancer registry had a rarer subtype (ER+/PR or ER/PR+) requiring confirmation by hospital laboratories.

Black Women: Etiology and Survival of Triple-negative Breast Cancers Study

Breast cancer case ascertainment

The Black Women: Etiology and Survival of Triple-negative Breast Cancers Study (BEST) was a population-based case-only study conducted in Florida. Briefly, Black women ages 50 years or younger at diagnosis, diagnosed with primary incident breast cancer between 2009 and 2012, were identified through the statewide cancer registry (13). Written informed consent was obtained from participants. Participants completed a baseline questionnaire through completion of paper forms, with the option of completion through a phone interview. Data collected included demographic characteristics, suspected breast cancer risk factors, and reproductive risk factors. Black race was identified on the basis of self-report. Through June of 2018, BEST had successfully recruited 449 Black women with incident invasive breast cancer for the interview.

Tumor marker ascertainment

On the basis of registry data and supplemented by pathology reports when available, tumor ER and PR status was available for 94% (n = 420) of women and HER2 status was available for 84% (n = 375) of women. Approximately 47 women (11%) with ER/PR status available from medical records/cancer registry had a rarer subtype (ER+/PR or ER/PR+) that required confirmation by hospital laboratories.

Statistical analysis

Enrollment characteristics of the four study populations are presented in Table 1. The overall harmonized dataset, defining all variables identically across studies, consisted of cases from STSBHS (n = 1,164), SCCS (n = 788), NBHS (n = 416), and BEST (n = 449). Women were classified as postmenopausal if cessation of menstrual periods occurred at least 6 months before the date of breast cancer diagnosis. Women missing menopausal status (n = 83; STSBHS n = 8, SCCS n = 67, NBHS n = 0, BEST n = 8) or ER/PR/HER2 status (n = 468; STSBHS n = 116, SCCS n = 193, NBHS n = 163, BEST n = 74) were excluded, resulting in 2,188 Black women for analysis. Breastfeeding was only asked of parous women, was not available for NBHS and was only available for 30% of STSBHS participants.

Table 1.

Enrollment characteristics of the four study populations contributing to the pooled analysis.

STSBHSSCCSNBHSBEST
Calendar years of enrollment TN: 2009–2014 2002–2009; cohort follow-up for cancer through 2018 2001–2011 2009–2012 
 SC: 2009–2012    
 GA: 2011–2012    
State(s) of enrollment TN, SC, GA 12 Southeastern states (AL, AR, FL, GA, KY, LA, MS, NC, SC, TN, VA, WV) TN FL 
Age at enrollment 25–75y 40–79y (age at diagnosis 40–88y) 25–75y 20–50y 
Total number of Black women with invasive breast cancer 1,164 788 416 449 
Menopausal status at breast cancer diagnosis 1,156 721 416 441 
Postmenopausal 877 (75.9%) 675 (93.6%) 285 (68.5%) 164 (37.2%) 
Premenopausal 279 (24.1%) 46 (6.4%) 131 (31.5%) 277 (62.8%) 
Missing 67 
Total number with known breast tumor ER/PR/HER2 status 1,040 528 253 367 
ER+ or PR+/HER2 622 (59.8%) 326 (61.7%) 119 (47.0%) 175 (47.7%) 
ER+ or PR+/HER2+ 114 (11.0%) 49 (9.3%) 45 (17.8%) 54 (14.7%) 
ER or PR/HER2+ 54 (5.2%) 36 (6.8%) 21 (8.3%) 24 (6.5%) 
TNBC 250 (24.0%) 117 (22.2%) 68 (26.9%) 114 (31.1%) 
Missing 116 193 163 74 
STSBHSSCCSNBHSBEST
Calendar years of enrollment TN: 2009–2014 2002–2009; cohort follow-up for cancer through 2018 2001–2011 2009–2012 
 SC: 2009–2012    
 GA: 2011–2012    
State(s) of enrollment TN, SC, GA 12 Southeastern states (AL, AR, FL, GA, KY, LA, MS, NC, SC, TN, VA, WV) TN FL 
Age at enrollment 25–75y 40–79y (age at diagnosis 40–88y) 25–75y 20–50y 
Total number of Black women with invasive breast cancer 1,164 788 416 449 
Menopausal status at breast cancer diagnosis 1,156 721 416 441 
Postmenopausal 877 (75.9%) 675 (93.6%) 285 (68.5%) 164 (37.2%) 
Premenopausal 279 (24.1%) 46 (6.4%) 131 (31.5%) 277 (62.8%) 
Missing 67 
Total number with known breast tumor ER/PR/HER2 status 1,040 528 253 367 
ER+ or PR+/HER2 622 (59.8%) 326 (61.7%) 119 (47.0%) 175 (47.7%) 
ER+ or PR+/HER2+ 114 (11.0%) 49 (9.3%) 45 (17.8%) 54 (14.7%) 
ER or PR/HER2+ 54 (5.2%) 36 (6.8%) 21 (8.3%) 24 (6.5%) 
TNBC 250 (24.0%) 117 (22.2%) 68 (26.9%) 114 (31.1%) 
Missing 116 193 163 74 

Demographic and reproductive characteristics of participants with complete and incomplete menopausal and hormone receptor (HR) status presented in Supplementary Table S1 show that all reproductive characteristics with the exception of parity were comparable by completion status.

The distribution of tumors by HR status differed by study site (Table 1). Women recruited from NBHS and BEST had a lower proportion of ER+ or PR+/HER2 tumors compared with women recruited from other studies (48% vs. 60%), while the proportion of TNBC was higher in women recruited from BEST than women recruited from other studies (31% vs. 24%). Similar findings were evident when stratified by menopausal status. For analysis, breast cancers were classified by subtype on the basis of tumor ER/PR/HER2 status as (i) ER+ or PR+ and HER2, (ii) HER2+ (regardless of ER/PR status due to the small percentages of women with ER or PR shown in Table 1), and (iii) TNBC; ER/PR/HER2.

Crude frequency distributions of demographic, reproductive, and other categorical variables were examined by ER/PR/HER2 status. Case-only polytomous unconditional logistic regression models were used to estimate ORs and 95% confidence intervals (CI) for HER2+ and TNBC diagnosis in relation to ER+/HER2 diagnosis (referent) for reproductive factors overall and by menopausal status at the time of breast cancer diagnosis. This case-only approach allows for testing the hypothesis that the strength of association with each independent factor is different for the three case groups defined by HR status (14). We observed suggestive between-study heterogeneity (P < 0.20) for the associations of two reproductive risk factors with clinical subtype based on Cochran Q statistic. Because tests of heterogeneity are known to lack statistical power, we utilized random effects polytomous regression models with adjustment for study site and other covariates (15). Additional covariates included age at breast cancer diagnosis (5-year age groups), family history of breast cancer (yes/no), annual household income (<$15,000, $15,000-$24,999, $25,000-$49,999, $50,000+), age at menarche (<13, 13, 14+ years), OC use (never/ever), and age at first birth (<19, 19–24, 25+ years). Analyses among postmenopausal women were additionally adjusted for age at menopause (<45, 45–50, 50+ years). Ordered categorical trend tests across levels of exposure were conducted. Significance was determined utilizing two-sided tests with a 0.05 nominal significance level. Between-study heterogeneity was examined by performing a random effects meta-analysis using STATA/IC 12 statistical software (StataCorp LP). All other statistical analyses were performed using SAS software version 9.4 (SAS Institute Inc).

Table 2 presents the distribution of demographic characteristics of women with breast cancer by HR and HER2 status, overall and by menopausal status. Among the 2,188 women for whom tumor ER, PR, and HER2 status were all known, 1,242 (57%) were ER+ or PR+ and HER2, 397 (18%) were HER2+, and 549 (25%) were triple negative (ER/PR/HER2). Age at diagnosis, family history of breast cancer and income differed by HR status among women overall and by menopausal status.

Table 2.

Demographic characteristics of breast cancer cases by tumor ER, PR, and HER2 status in the pooled dataset, overall and by menopausal status at time of diagnosis.

OverallPremenopausalPostmenopausal
ER+/HER2HER2+TNBCER+/HER2HER2+TNBCER+/HER2HER2+TNBC
(n = 1242)(n = 397)(n = 549)(n = 308)(n = 128)(n = 155)(n = 934)(n = 269)(n = 394)
Characteristic%%%%%%%%%
Age at diagnosis 
 20–39 years 8.9 11.8 11.8 27.6 31.3 30.3 2.7 2.6 4.6 
 40–44 years 8.6 11.6 12.6 23.0 28.1 23.9 3.8 3.7 8.1 
 45–49 years 14.6 16.4 19.9 30.5 26.6 31.6 9.3 11.5 15.2 
 50–54 years 13.5 17.6 15.9 15.3 11.7 11.6 12.8 20.4 17.5 
 55–59 years 16.3 15.4 16.4 3.6 2.3 2.6 20.6 21.6 21.8 
 60–64 years 15.7 13.4 11.7 — — — 20.9 19.7 16.3 
 65–69 years 13.4 9.3 8.7 — — — 17.8 13.8 12.2 
 70–88 years 9.1 4.5 3.1 — — — 12.1 6.7 4.3 
Family history of breast cancer 
 No 74.5 77.7 68.4 68.5 72.7 66.9 76.5 80.2 69.0 
 Yes 25.5 22.3 31.6 31.5 27.3 33.1 23.5 19.8 31.0 
Education 
 ≤High school 38.4 36.0 35.3 22.4 25.4 31.2 43.6 41.0 37.0 
 Vocational school/ 33.5 33.0 38.5 39.0 31.7 37.6 31.7 33.6 38.8 
 Some college 
 College+ 28.2 31.0 26.2 38.6 42.9 31.2 24.7 25.4 24.2 
Income 
 <$15,000 33.2 34.8 34.7 21.9 25.2 30.8 36.9 39.4 36.2 
 $15,000–$24,999 20.2 20.7 17.6 16.5 19.5 19.2 21.4 21.2 16.9 
 $25,000–$49,999 22.5 16.2 19.4 26.6 20.3 17.8 21.1 14.3 20.1 
 $50,000+ 24.1 28.3 28.3 35.0 35.0 32.2 20.6 25.1 26.8 
BMI (kg/m2
 <25 15.2 13.5 15.8 18.8 16.5 16.9 14.0 12.1 15.4 
 25–<30 28.7 29.6 27.6 30.8 28.4 27.3 27.9 30.2 27.7 
 30–<35 26.2 29.9 23.3 25.7 26.0 24.7 26.4 31.7 22.8 
 35+ 29.9 27.0 33.3 24.7 29.1 31.1 31.7 26.0 34.1 
OverallPremenopausalPostmenopausal
ER+/HER2HER2+TNBCER+/HER2HER2+TNBCER+/HER2HER2+TNBC
(n = 1242)(n = 397)(n = 549)(n = 308)(n = 128)(n = 155)(n = 934)(n = 269)(n = 394)
Characteristic%%%%%%%%%
Age at diagnosis 
 20–39 years 8.9 11.8 11.8 27.6 31.3 30.3 2.7 2.6 4.6 
 40–44 years 8.6 11.6 12.6 23.0 28.1 23.9 3.8 3.7 8.1 
 45–49 years 14.6 16.4 19.9 30.5 26.6 31.6 9.3 11.5 15.2 
 50–54 years 13.5 17.6 15.9 15.3 11.7 11.6 12.8 20.4 17.5 
 55–59 years 16.3 15.4 16.4 3.6 2.3 2.6 20.6 21.6 21.8 
 60–64 years 15.7 13.4 11.7 — — — 20.9 19.7 16.3 
 65–69 years 13.4 9.3 8.7 — — — 17.8 13.8 12.2 
 70–88 years 9.1 4.5 3.1 — — — 12.1 6.7 4.3 
Family history of breast cancer 
 No 74.5 77.7 68.4 68.5 72.7 66.9 76.5 80.2 69.0 
 Yes 25.5 22.3 31.6 31.5 27.3 33.1 23.5 19.8 31.0 
Education 
 ≤High school 38.4 36.0 35.3 22.4 25.4 31.2 43.6 41.0 37.0 
 Vocational school/ 33.5 33.0 38.5 39.0 31.7 37.6 31.7 33.6 38.8 
 Some college 
 College+ 28.2 31.0 26.2 38.6 42.9 31.2 24.7 25.4 24.2 
Income 
 <$15,000 33.2 34.8 34.7 21.9 25.2 30.8 36.9 39.4 36.2 
 $15,000–$24,999 20.2 20.7 17.6 16.5 19.5 19.2 21.4 21.2 16.9 
 $25,000–$49,999 22.5 16.2 19.4 26.6 20.3 17.8 21.1 14.3 20.1 
 $50,000+ 24.1 28.3 28.3 35.0 35.0 32.2 20.6 25.1 26.8 
BMI (kg/m2
 <25 15.2 13.5 15.8 18.8 16.5 16.9 14.0 12.1 15.4 
 25–<30 28.7 29.6 27.6 30.8 28.4 27.3 27.9 30.2 27.7 
 30–<35 26.2 29.9 23.3 25.7 26.0 24.7 26.4 31.7 22.8 
 35+ 29.9 27.0 33.3 24.7 29.1 31.1 31.7 26.0 34.1 

Table 3 presents the distributions of reproductive characteristics of women with breast cancer by HR and HER2 status, overall and stratified by menopausal status. Most variables differed by HR status among women overall including age at menarche, OC use, parity, age at first birth, breastfeeding, and age at menopause (for postmenopausal women). Notably, when restricted to postmenopausal women, women with ER+/HER2 tumors were less likely to have used hormone therapy or OCs and were more likely to have breastfed for more than a year compared with women with the other two subtypes.

Table 3.

Reproductive characteristics of breast cancer cases by tumor ER, PR, and HER2 status in the pooled dataset, overall and by menopausal status at time of diagnosis.

OverallPremenopausalPostmenopausal
ER+/HER2HER2+TNBCER+/HER2HER2+TNBCER+/HER2HER2+TNBC
(n = 1242)(n = 397)(n = 549)(n = 308)(n = 128)(n = 155)(n = 934)(n = 269)(n = 394)
CharacteristicN (%)N (%)N (%)N (%)N (%)N (%)N (%)N (%)N (%)
Hormone therapy 
 Never 941 (76.5) 301 (77.2) 409 (75.3) 288 (93.5) 120 (94.5) 146 (95.4) 653 (70.8) 181 (68.8) 263 (67.4) 
 Ever 289 (23.5) 89 (22.8) 134 (24.7) 20 (6.5) 7 (5.5) 7 (4.6) 269 (29.2) 82 (31.2) 127 (32.6) 
Age at menarche 
 <13 years 570 (46.6) 201 (51.7) 276 (51.1) 158 (51.5) 69 (54.7) 83 (54.6) 412 (45.0) 132 (50.2) 193 (49.7) 
 13 years 331 (27.1) 78 (20.0) 128 (23.7) 81 (26.4) 22 (17.5) 34 (22.4) 250 (27.3) 56 (21.3) 94 (24.2) 
 14+ years 322 (26.3) 110 (28.3) 136 (25.2) 68 (22.1) 35 (27.8) 35 (23.0) 254 (27.7) 75 (28.5) 101 (26.1) 
Oral contraceptive 
 Never 356 (28.9) 102 (26.0) 113 (20.7) 51 (16.6) 27 (21.1) 25 (16.2) 305 (32.9) 75 (28.4) 88 (22.5) 
 Ever 878 (71.1) 290 (74.0) 433 (79.3) 257 (83.4) 101 (78.9) 129 (83.8) 621 (67.1) 189 (71.6) 304 (77.5) 
Parity 
 Nulliparous 158 (12.9) 67 (17.2) 50 (9.2) 49 (16.2) 25 (20.0) 17 (11.2) 109 (11.8) 42 (15.8) 33 (8.5) 
 1 230 (18.7) 73 (18.7) 88 (16.3) 68 (22.4) 37 (29.6) 29 (19.1) 162 (17.5) 36 (13.6) 59 (15.2) 
 2 341 (27.8) 113 (29.0) 174 (32.1) 93 (30.7) 25 (20.0) 59 (38.8) 248 (26.8) 88 (33.2) 115 (29.5) 
 3 237 (19.3) 80 (20.5) 121 (22.4) 53 (17.5) 25 (20.0) 25 (16.4) 184 (19.9) 55 (20.8) 96 (24.7) 
 4+ 262 (21.3) 57 (14.6) 108 (20.0) 40 (13.2) 13 (10.4) 22 (14.5) 222 (24.0) 44 (16.6) 86 (22.1) 
Age at first birth 
 Nulliparous 158 (12.9) 67 (17.2) 50 (9.3) 49 (16.2) 25 (20.0) 17 (11.3) 109 (11.8) 42 (15.8) 33 (8.5) 
 ≤18 years 356 (29.0) 93 (23.9) 151 (28.0) 60 (19.8) 22 (17.6) 33 (21.8) 296 (32.0) 71 (26.8) 118 (30.4) 
 19–24 431 (35.1) 144 (36.9) 234 (43.4) 95 (31.3) 41 (32.8) 66 (43.7) 336 (36.4) 103 (38.9) 168 (43.3) 
 25+ years 282 (23.0) 86 (22.0) 104 (19.3) 99 (32.7) 37 (29.6) 35 (23.2) 183 (19.8) 49 (18.5) 69 (17.8) 
Breastfeedinga 
 Never 296 (56.8) 98 (61.6) 170 (64.6) 63 (47.4) 33 (55.9) 35 (50.7) 233 (60.1) 65 (65.0) 135 (69.6) 
 0–12 months 170 (32.6) 48 (30.2) 79 (30.1) 59 (44.3) 19 (32.2) 29 (42.0) 111 (28.6) 29 (29.0) 50 (25.8) 
 13+month 55 (10.6) 13 (8.2) 14 (5.3) 11 (8.3) 7 (11.9) 5 (7.3) 44 (11.3) 6 (6.0) 9 (4.6) 
Age at menopause 
 <45 years — — — — — — 421 (48.6) 107 (41.6) 193 (52.2) 
 45–49 years — — — — — — 210 (24.2) 73 (28.4) 98 (26.5) 
 50+ years — — — — — — 236 (27.2) 77 (30.0) 79 (21.3) 
OverallPremenopausalPostmenopausal
ER+/HER2HER2+TNBCER+/HER2HER2+TNBCER+/HER2HER2+TNBC
(n = 1242)(n = 397)(n = 549)(n = 308)(n = 128)(n = 155)(n = 934)(n = 269)(n = 394)
CharacteristicN (%)N (%)N (%)N (%)N (%)N (%)N (%)N (%)N (%)
Hormone therapy 
 Never 941 (76.5) 301 (77.2) 409 (75.3) 288 (93.5) 120 (94.5) 146 (95.4) 653 (70.8) 181 (68.8) 263 (67.4) 
 Ever 289 (23.5) 89 (22.8) 134 (24.7) 20 (6.5) 7 (5.5) 7 (4.6) 269 (29.2) 82 (31.2) 127 (32.6) 
Age at menarche 
 <13 years 570 (46.6) 201 (51.7) 276 (51.1) 158 (51.5) 69 (54.7) 83 (54.6) 412 (45.0) 132 (50.2) 193 (49.7) 
 13 years 331 (27.1) 78 (20.0) 128 (23.7) 81 (26.4) 22 (17.5) 34 (22.4) 250 (27.3) 56 (21.3) 94 (24.2) 
 14+ years 322 (26.3) 110 (28.3) 136 (25.2) 68 (22.1) 35 (27.8) 35 (23.0) 254 (27.7) 75 (28.5) 101 (26.1) 
Oral contraceptive 
 Never 356 (28.9) 102 (26.0) 113 (20.7) 51 (16.6) 27 (21.1) 25 (16.2) 305 (32.9) 75 (28.4) 88 (22.5) 
 Ever 878 (71.1) 290 (74.0) 433 (79.3) 257 (83.4) 101 (78.9) 129 (83.8) 621 (67.1) 189 (71.6) 304 (77.5) 
Parity 
 Nulliparous 158 (12.9) 67 (17.2) 50 (9.2) 49 (16.2) 25 (20.0) 17 (11.2) 109 (11.8) 42 (15.8) 33 (8.5) 
 1 230 (18.7) 73 (18.7) 88 (16.3) 68 (22.4) 37 (29.6) 29 (19.1) 162 (17.5) 36 (13.6) 59 (15.2) 
 2 341 (27.8) 113 (29.0) 174 (32.1) 93 (30.7) 25 (20.0) 59 (38.8) 248 (26.8) 88 (33.2) 115 (29.5) 
 3 237 (19.3) 80 (20.5) 121 (22.4) 53 (17.5) 25 (20.0) 25 (16.4) 184 (19.9) 55 (20.8) 96 (24.7) 
 4+ 262 (21.3) 57 (14.6) 108 (20.0) 40 (13.2) 13 (10.4) 22 (14.5) 222 (24.0) 44 (16.6) 86 (22.1) 
Age at first birth 
 Nulliparous 158 (12.9) 67 (17.2) 50 (9.3) 49 (16.2) 25 (20.0) 17 (11.3) 109 (11.8) 42 (15.8) 33 (8.5) 
 ≤18 years 356 (29.0) 93 (23.9) 151 (28.0) 60 (19.8) 22 (17.6) 33 (21.8) 296 (32.0) 71 (26.8) 118 (30.4) 
 19–24 431 (35.1) 144 (36.9) 234 (43.4) 95 (31.3) 41 (32.8) 66 (43.7) 336 (36.4) 103 (38.9) 168 (43.3) 
 25+ years 282 (23.0) 86 (22.0) 104 (19.3) 99 (32.7) 37 (29.6) 35 (23.2) 183 (19.8) 49 (18.5) 69 (17.8) 
Breastfeedinga 
 Never 296 (56.8) 98 (61.6) 170 (64.6) 63 (47.4) 33 (55.9) 35 (50.7) 233 (60.1) 65 (65.0) 135 (69.6) 
 0–12 months 170 (32.6) 48 (30.2) 79 (30.1) 59 (44.3) 19 (32.2) 29 (42.0) 111 (28.6) 29 (29.0) 50 (25.8) 
 13+month 55 (10.6) 13 (8.2) 14 (5.3) 11 (8.3) 7 (11.9) 5 (7.3) 44 (11.3) 6 (6.0) 9 (4.6) 
Age at menopause 
 <45 years — — — — — — 421 (48.6) 107 (41.6) 193 (52.2) 
 45–49 years — — — — — — 210 (24.2) 73 (28.4) 98 (26.5) 
 50+ years — — — — — — 236 (27.2) 77 (30.0) 79 (21.3) 

aAmong parous; NBHS was excluded from the breastfeeding analysis and breastfeeding was only available for 30% of STSBHS.

Table 4 presents the multivariable polytomous regression results of the case-only comparison of HER2+ and TNBC versus ER+/HER2 (referent), overall and stratified by menopausal status. Relative to women who had ER+/HER2 tumors, women who were age 13 at menarche or age 18 or younger at first birth appeared less likely to be HER2+. Compared with all nulliparous women, parous women appeared to be less likely to have HER2+ [OR, 0.70 (95% CI, 0.50–0.98)] and more likely to have TNBC tumors [OR, 1.44 (95% CI, 1.01–2.05)] relative to ER+/HER2 tumors. Similar patterns were seen for postmenopausal women. Relative to women who had ER+/HER2 tumors, women who were age 19–24 years at first birth (OR, 1.78; 95% CI, 1.22–2.59) were more likely to have TNBC. Postmenopausal parous women who breastfed were less likely to have TNBC [OR, 0.65 (95% CI, 0.43–0.99). In addition, among postmenopausal women, those who had used hormone therapy [OR, 1.32 (95% CI, 0.98–1.78)] or OCs [OR, 1.31 (95% CI, 0.97–1.78)] were more likely to have TNBC as opposed to ER+/HER2 tumors. Results were less precise among premenopausal women.

Table 4.

Case-only polytomous regression analysis of the association between reproductive risk factors and HER2+ and TNBC compared with ER+ breast cancer, overall and by menopausal status at time of diagnosis.

OverallPremenopausalPostmenopausal
HER+TNBCHER2+TNBCHER2+TNBC
CharacteristicORa (95% CI)ORa (95% CI)ORa (95% CI)ORa (95% CI)ORa,b (95% CI)ORa,b (95% CI)
Hormone therapy 
 Never 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Ever 1.08 (0.80–1.45) 1.30 (0.99–1.69) 0.88 (0.35–2.22) 0.69 (0.27–1.75) 1.08 (0.77–1.52) 1.32 (0.98–1.78) 
Age at menarche 
 <13 years 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 13 years 0.67 (0.49–0.91) 0.81 (0.63–1.06) 0.68 (0.39–1.20) 0.81 (0.48–1.36) 0.60 (0.41–0.88) 0.76 (0.55–1.04) 
 14+ years 0.99 (0.75–1.31) 0.87 (0.67–1.13) 1.24 (0.73–2.11) 0.99 (0.59–1.67) 0.82 (0.58–1.18) 0.79 (0.58–1.09) 
Ptrend 0.65 0.23 0.64 0.83 0.17 0.11 
Oral contraceptive 
 Never 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Ever 1.11 (0.84–1.48) 1.22 (0.94–1.59) 0.68 (0.38–1.21) 0.93 (0.52–1.67) 1.24 (0.88–1.75) 1.31 (0.97–1.78) 
Parity 
 Nulliparous 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Parous 0.70 (0.50–0.98) 1.44 (1.01–2.05) 0.86 (0.48–1.53) 1.78 (0.92–3.44) 0.65 (0.42–1.01) 1.29 (0.83–2.00) 
Age at first birth 
 Nulliparous 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 ≤18 years 0.60 (0.41–0.90) 1.39 (0.94–2.07) 0.81 (0.39–1.71) 1.96 (0.90–4.26) 0.55 (0.33–0.91) 1.22 (0.75–1.98) 
 19–24 0.79 (0.54–1.13) 1.78 (1.22–2.59) 0.89 (0.46–1.71) 2.49 (1.24–5.03) 0.76 (0.48–1.23) 1.50 (0.94–2.39) 
 25+ years 0.68 (0.46–1.01) 1.04 (0.69–1.57) 0.86 (0.45–1.67) 1.05 (0.50–2.22) 0.62 (0.36–1.05) 1.04 (0.62–1.74) 
Ptrend 0.32 0.81 0.72 0.84 0.44 0.76 
Breastfeedingc 
 Never 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Ever 0.94 (0.64–1.39) 0.78 (0.56–1.09) 0.69 (0.35–1.36) 1.04 (0.54–2.00) 0.99 (0.59–1.65) 0.65 (0.43–0.99) 
Age at menopause 
 <45 years — — — — 1.0 (Ref) 1.0 (Ref) 
 45–49 years — — — — 1.34 (0.92–1.95) 1.08 (0.78–1.49) 
 50+ years — — — — 1.46 (1.00–2.14) 0.97 (0.69–1.37) 
Ptrend     0.04 0.97 
OverallPremenopausalPostmenopausal
HER+TNBCHER2+TNBCHER2+TNBC
CharacteristicORa (95% CI)ORa (95% CI)ORa (95% CI)ORa (95% CI)ORa,b (95% CI)ORa,b (95% CI)
Hormone therapy 
 Never 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Ever 1.08 (0.80–1.45) 1.30 (0.99–1.69) 0.88 (0.35–2.22) 0.69 (0.27–1.75) 1.08 (0.77–1.52) 1.32 (0.98–1.78) 
Age at menarche 
 <13 years 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 13 years 0.67 (0.49–0.91) 0.81 (0.63–1.06) 0.68 (0.39–1.20) 0.81 (0.48–1.36) 0.60 (0.41–0.88) 0.76 (0.55–1.04) 
 14+ years 0.99 (0.75–1.31) 0.87 (0.67–1.13) 1.24 (0.73–2.11) 0.99 (0.59–1.67) 0.82 (0.58–1.18) 0.79 (0.58–1.09) 
Ptrend 0.65 0.23 0.64 0.83 0.17 0.11 
Oral contraceptive 
 Never 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Ever 1.11 (0.84–1.48) 1.22 (0.94–1.59) 0.68 (0.38–1.21) 0.93 (0.52–1.67) 1.24 (0.88–1.75) 1.31 (0.97–1.78) 
Parity 
 Nulliparous 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Parous 0.70 (0.50–0.98) 1.44 (1.01–2.05) 0.86 (0.48–1.53) 1.78 (0.92–3.44) 0.65 (0.42–1.01) 1.29 (0.83–2.00) 
Age at first birth 
 Nulliparous 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 ≤18 years 0.60 (0.41–0.90) 1.39 (0.94–2.07) 0.81 (0.39–1.71) 1.96 (0.90–4.26) 0.55 (0.33–0.91) 1.22 (0.75–1.98) 
 19–24 0.79 (0.54–1.13) 1.78 (1.22–2.59) 0.89 (0.46–1.71) 2.49 (1.24–5.03) 0.76 (0.48–1.23) 1.50 (0.94–2.39) 
 25+ years 0.68 (0.46–1.01) 1.04 (0.69–1.57) 0.86 (0.45–1.67) 1.05 (0.50–2.22) 0.62 (0.36–1.05) 1.04 (0.62–1.74) 
Ptrend 0.32 0.81 0.72 0.84 0.44 0.76 
Breastfeedingc 
 Never 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 1.0 (Ref) 
 Ever 0.94 (0.64–1.39) 0.78 (0.56–1.09) 0.69 (0.35–1.36) 1.04 (0.54–2.00) 0.99 (0.59–1.65) 0.65 (0.43–0.99) 
Age at menopause 
 <45 years — — — — 1.0 (Ref) 1.0 (Ref) 
 45–49 years — — — — 1.34 (0.92–1.95) 1.08 (0.78–1.49) 
 50+ years — — — — 1.46 (1.00–2.14) 0.97 (0.69–1.37) 
Ptrend     0.04 0.97 

aModels are adjusted for study site (STSBHS, SCCS, NBHS, BEST), age at diagnosis (5-years age groups), family history of breast cancer (yes/no), annual household income (<$15,000, $15,000–$24,999, $25,000–$49,999, $50,000+), age at menarche (<13, 13, 14+ years), oral contraceptive use (never/ever), and age at first birth (nulliparous, ≤18, 19–24, 25+ years).

bAdditionally adjusted for age at menopause (<45, 45–49, 50+ years).

cAmong parous; NBHS was excluded from the breastfeeding analysis and breastfeeding was only available for 30% of STSBHS.

Our examination of reproductive and hormonal characteristics among Black women in the southeastern United States suggests that some of their well-established associations with breast cancer, in particular a protective effect of parity, may not be consistent across specific subtypes of cancer among Black women. Our observation that parous women were less likely to be diagnosed with HER2+ breast cancer and more likely to be diagnosed with TNBC relative to ER+/HER2 breast cancer builds upon a growing number of studies that have identified parity as an important factor that may increase the risk of developing ER breast cancer, including HER2-enriched breast cancer or TNBC (3, 7, 8, 16), in contrast to the inverse association well documented for ER+ tumors. In particular, our findings are consistent with the report from a case-only analysis by ER status of 2,354 Black women with breast cancer in the African American Breast Cancer Epidemiology and Risk (AMBER) consortium that showed etiologic heterogeneity for age at menarche, parity (vs. nulliparity) and age at first birth (5). In that study, compared with nulliparous women, parous women had an approximately 2-fold higher likelihood of developing ER compared with ER+ cancer, while age at first full-term birth ≥25 years was associated with a 25% reduction in risk of ER relative to ER+ cancer.

The observed enhanced risk associated with parity for TNBC relative to ER+/HER2 breast cancer in our study suggests that HR tumors may have a different etiology compared with HR+ tumors. This points to a need for additional research to identify other potential risk factors for TNBC beyond the hormone-related factors that have traditionally been deemed most important in breast cancer etiology. It has been postulated that immune dysregulation may play a particularly important role in enhancing TNBC risk among Black women and among young women, including exposures at young ages such as nutrition, psychosocial stress, and toxins (17, 18).

Findings of distinct reproductive profiles of different tumor subtypes are of particular importance for Black women, who have the highest incidence rate of TNBC (2), the subtype most likely to present with high-grade disease. TNBCs account for about 15% of all breast cancers, but the percentage is higher among younger and premenopausal women; among Black women with breast cancer, including in our study, approximately one-fourth had TNBC. Moreover, Black women are more likely to have higher parity and less likely to breastfeed, as compared to White women (19). In general, lower rates of breastfeeding occur among women who are low income, Black, overweight, unmarried, and less educated (19). Although the rate of exclusive breastfeeding at 6 months among Black mothers has increased from 11% in 2009, a 2015 (20) study showed that only 36% of Black mothers were exclusively breastfeeding at 3 months and only 17% at 6 months, and there remains a disparity between Black and White mothers. Many barriers to initiating and continuing breastfeeding have been reported among Black women, including personal beliefs or knowledge of breastfeeding, lack of social support, and employment (21), but additional research is needed. If in fact breastfeeding is a lifestyle factor confirmed to reduce the likelihood of developing TNBC relative to ER+ breast cancer, then Black women should be prioritized for intervention and promotion efforts to enhance uptake of breastfeeding.

The postpartum increase in the inflammatory state of breast tissue has been implicated as a possible link between parity and the development of ER breast cancer (22–24). It has been postulated that prolonged lactation may lead to a more gradual involution period, resulting in less breast tissue inflammation and thus reduced risk of ER breast cancer resulting from parity. In our study, while parity was estimated to be associated with greater likelihood of TNBC compared with ER+/HER2 breast cancer, breastfeeding among parous women decreased the likelihood of developing TNBC relative to ER+/HER2 breast cancer, and may represent a modifiable lifestyle behavior that could contribute to lower likelihood of developing a particular subtype of breast cancer common among Black women which implies poor prognosis. Several other studies have reported increased risk of ER breast cancer and TNBC among parous women who did not breastfeed or have reported attenuation of associations between increased parity and the incidence of ER breast cancers among women who breasted (3, 7, 8, 16, 25). In the recent report from the AMBER consortium (5), women who were parous but did not breastfeed had the highest ORs for basal-like breast cancer, classified on the basis of six IHC biomarkers (ER, PR, HER2, Ki67, EGFR, and CK5/6; ref. 26), confirming their previous results based on clinical markers showing that parous women who did not breastfeed had increased risk for ER and TNBC subtypes.

We observed a 30% increased odds of having TNBC, as compared with ER+/HER2 breast cancer, among postmenopausal women who ever used OCs or HRT. Although we were unable to classify hormone replacement or OC use by recency of use or by formulation, the observed direction of association is consistent with previous findings from the Black Women's Health study (27) in which OC use was associated with increased risk of ER breast cancer in Black women. In contrast, in the AMBER consortium, no significant heterogeneity was observed for OC use (5) by subtype. No other factors showed significant heterogeneity by breast tumor subtype in our study, with the exception of a reduced risk for HER2+ cancer among women age 13 years at menarche (vs. <13).

The current pooled study is one of the largest to compare breast cancer subtypes among Black women stratified by menopausal status, and the case-only approach allowed for examination of etiologic heterogeneity by subgroups defined by tumor HR and HER2 status. An additional strength is the study's location in the southeastern United States which includes states with among the highest racial disparities for breast cancer mortality in the country (28). Limitations include the small number of women who were premenopausal at the time of breast cancer diagnosis, and the earlier years of enrollment for the SCCS and NBHS when HER2 was less likely to be available. Missing data on tumor subtype resulted in a relatively small number of women with HER2+ cancer, which may have limited our power to detect associations in that subgroup. Moreover, two of the four studies (NBHS and STSBHS) in our analysis did not include complete data on breastfeeding; therefore, the associations of other reproductive and hormonal factors with breast cancer were not adjusted for breastfeeding in the multivariable regression models. Finally, our studies relied on a three-marker IHC panel for classification of breast tumors; it has been demonstrated that this panel has high agreement with the clinical record, but that additional biomarkers are required for accurate classification of HER2-enriched cancers (29).

In conclusion, we observed etiologic heterogeneity among breast cancer subtypes among Black women. Our findings are consistent with accumulating evidence that traditional reproductive protective factors, such as younger age at first birth and parity, may in fact be associated with increased likelihood of developing TNBC as opposed to ER+ breast cancer. Most important, breastfeeding may represent a modifiable lifestyle behavior to reduce likelihood of developing TNBC relative to ER+/HER2 breast cancer among Black parous women.

A. Beeghly-Fadiel reports grants from NIH U24 MD010722 PMHDC Pilot Study and NIH U54 CA163072 MVTCP outside the submitted work. S.-A. Dujon reports grants from NCI during the conduct of the study. J. Davis reports grants from NCI, Health Resources and Services Administration, Vanderbilt-Ingram Cancer Center, Bankhead Coley Granting agency, American Cancer Society, and Anne Potter Wilson Chair funds during the conduct of the study. M. Fortune reports grants from Health Resources and Services Administration (HRSA) during the conduct of the study. K. Benson reports grants from Health Resources and Services Administration during the conduct of the study. X.-O. Shu reports grants from NCI during the conduct of the study. W.J. Blot reports grants from NIH during the conduct of the study. No disclosures were reported by the other authors.

M. Sanderson: Conceptualization, resources, formal analysis, supervision, funding acquisition, investigation, writing–original draft, writing–review and editing. T. Pal: Conceptualization, resources, supervision, funding acquisition, investigation, writing–review and editing. A. Beeghly-Fadiel: Data curation, writing–review and editing. M.K. Fadden: Formal analysis, methodology, writing–review and editing. S.-A. Dujon: Methodology, writing–review and editing. C. Clinton: Methodology, writing–review and editing. C. Jimenez: Methodology, writing–review and editing. J. Davis: Methodology, writing–review and editing. M. Fortune: Methodology, writing–review and editing. J. Thompson: Methodology, writing–review and editing. K. Benson: Methodology, writing–review and editing. N. Conley: Methodology, writing–review and editing. S. Reid: Methodology, writing–review and editing. A. Tezak: Methodology, writing–review and editing. X.-O. Shu: Methodology, writing–review and editing. W. Zheng: Conceptualization, resources, supervision, funding acquisition, investigation, writing–review and editing. W.J. Blot: Conceptualization, resources, supervision, funding acquisition, investigation, writing–review and editing. L. Lipworth: Conceptualization, formal analysis, supervision, investigation, writing–original draft, writing–review and editing.

The STSBHS was supported by grants through the NCI (U54CA163069, U54CA163072, and R03CA192214) and the National Center for Advancing Translational Sciences (UL1TR000445). S.-A. Dujon, C. Clinton, J. Davis, and J. Thompson were supported by the NCI (U54CA163069). N. Conley was supported by the NCI (R25CA102209). S. Reid was supported by the NCI (T32CA160056). C. Jimenez, M. Fortune, and K. Benson were supported by the Health Resources and Services Administration (D34HP16299). The Southern Community Cohort Study (SCCS) was supported by the NCI (R01CA092447, U01CA202979, and P50CA098131). SCCS data collection performed by the Survey and Biospecimen Shared Resource which is supported in part by the Vanderbilt-Ingram Cancer Center (P30CA068485). The NBHS was supported by the NCI (R01CA100374). The BEST Study was supported by grants through the Bankhead Coley Granting agency (IBG10-34199), the American Cancer Society (RSG-11-268-01-CPPB), the NCI (R01CA204819), and the National Center for Advancing Translational Sciences (UL1TR000445). Additional support was provided by the NCI (R01CA202981) and by Zheng's Anne Potter Wilson Chair funds.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
DeSantis
CE
,
Ma
J
,
Gaudet
MM
,
Newman
LA
,
Miller
KD
,
Goding Sauer
A
, et al
Breast cancer statistics, 2019
.
CA Cancer J Clin
2019
;
69
:
438
51
.
2.
Howlader
N
,
Altekruse
SF
,
Li
CI
,
Chen
VW
,
Clarke
CA
,
Ries
LA
, et al
US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status
.
J Natl Cancer Inst
2014
;
106
:
dju055
.
3.
Millikan
RC
,
Newman
B
,
Tse
CK
,
Moorman
PG
,
Conway
K
,
Dressler
LG
, et al
Epidemiology of basal-like breast cancer
.
Breast Cancer Res Treat
2008
;
109
:
123
39
.
4.
Dietze
EC
,
Sistrunk
C
,
Miranda-Carboni
G
,
O'Regan
R
,
Seewaldt
VL
. 
Triple-negative breast cancer in African-American women: disparities versus biology
.
Nat Rev Cancer
2015
;
15
:
248
54
.
5.
Benefield
HC
,
Zirpoli
GR
,
Allott
EH
,
Shan
Y
,
Hurson
AN
,
Omilian
AR
, et al
Epidemiology of basal-like and luminal breast cancers among Black women in the AMBER consortium
.
Cancer Epidemiol Biomarkers Prev
2021
;
30
:
71
9
.
6.
Ambrosone
CB
,
Zirpoli
G
,
Ruszczyk
M
,
Shankar
J
,
Hong
CC
,
McIlwain
D
, et al
Parity and breastfeeding among African-American women: differential effects on breast cancer risk by estrogen receptor status in the Women's Circle of Health Study
.
Cancer Causes Control
2014
;
25
:
259
65
.
7.
Palmer
JR
,
Boggs
DA
,
Wise
LA
,
Ambrosone
CB
,
Adams-Campbell
LL
,
Rosenberg
L
. 
Parity and lactation in relation to estrogen receptor negative breast cancer in African American women
.
Cancer Epidemiol Biomarkers Prev
2011
;
20
:
1883
91
.
8.
Palmer
JR
,
Viscidi
E
,
Troester
MA
,
Hong
C-C
,
Schedin
P
,
Bethea
TN
, et al
Parity, lactation, and breast cancer subtypes in African American women: results from the AMBER Consortium
.
J Natl Cancer Inst
2014
;
106
:
dju237
.
9.
Anstey
EH
,
Shoemaker
ML
,
Barrera
CM
,
O'Neil
ME
,
Verma
AB
,
Holman
DM
. 
Breastfeeding and breast cancer risk reduction: implications for Black mothers
.
Am J Prev Med
2017
;
53
:
S40
6
.
10.
Signorello
LB
,
Hargreaves
MK
,
Blot
WJ
. 
The Southern Community Cohort Study: investigating health disparities
.
J Health Care Poor Underserved
2010
;
21
:
26
37
.
11.
Cui
Y
,
Deming-Halverson
SL
,
Beeghly-Fadiel
A
,
Lipworth
L
,
Shrubsole
MJ
,
Fair
AM
, et al
Interactions of hormone replacement therapy, body weight, and bilateral oophorectomy in breast cancer risk
.
Clin Cancer Res
2014
;
20
:
1169
78
.
12.
Cui
Y
,
Deming-Halverson
SL
,
Shrubsole
MJ
,
Beeghly-Fadiel
A
,
Fair
AM
,
Sanderson
M
, et al
Associations of hormone-related factors with breast cancer risk according to hormone receptor status among white and African American women
.
Clin Breast Cancer
2014
;
14
:
417
25
.
13.
Pal
T
,
Bonner
D
,
Cragun
D
,
Monteiro
AN
,
Phelan
C
,
Servais
L
, et al
A high frequency of BRCA mutations in young black women with breast cancer residing in Florida
.
Cancer
2015
;
121
:
4173
80
.
14.
Begg
CB
,
Zhang
ZF
. 
Statistical analysis of molecular epidemiology studies employing case-series
.
Cancer Epidemiol Biomarkers Prev
1994
;
3
;
173
5
.
15.
Friedenreich
CM
. 
Methods for pooled analyses of epidemiologic studies
.
Epidemiology
1993
;
4
:
295
302
.
16.
Shinde
SS
,
Forman
MR
,
Kuerer
HM
,
Yan
K
,
Peintinger
F
,
Hunt
KK
, et al
Higher parity and shorter breastfeeding duration: association with triple-negative phenotype of breast cancer
.
Cancer
2010
;
116
:
4933
43
.
17.
Bagby
SP
,
Martin
D
,
Chung
ST
,
Rajapakse
N
. 
From the outside in: biological mechanisms linking social and environmental exposures to chronic disease and to health disparities
.
Am J Public Health
2019
;
109
:
S56
63
.
18.
Ogony
JW
,
Radisky
DC
,
Ruddy
KJ
,
Goodison
S
,
Wickland
DP
,
Egan
KM
, et al
Immune responses and risk of triple-negative breast cancer: implications for higher rates among African American women
.
Cancer Prev Res
2020
;
13
:
901
10
.
19.
Jones
KM
,
Power
ML
,
Queenan
JT
,
Schulkin
J
. 
Racial and ethnic disparities in breastfeeding
.
Breastfeed Med
2015
;
10
:
186
96
.
20.
Beauregard
JL
,
Hamner
HC
,
Chen
J
,
Avila-Rodriguez
W
,
Elam-Evans
LD
,
Perrine
CG
. 
Racial disparities in breastfeeding initiation and duration among U.S. infants born in 2015
.
MMWR Morb Mortal Wkly Rep
2019
;
68
:
745
8
.
21.
Sayres
S
,
Visentin
L
. 
Breastfeeding: uncovering barriers and offering solutions
.
Curr Opin Pediatr
2018
;
30
:
591
6
.
22.
McDaniel
SM
,
Rumer
KK
,
Biroc
SL
,
Metz
RP
,
Singh
M
,
Porter
W
, et al
Remodeling of the mammary microenvironment after lactation promotes breast tumor cell metastasis
.
Am J Pathol
2006
;
168
:
608
20
.
23.
Schedin
P
. 
Pregnancy-associated breast cancer and metastasis
.
Nat Rev Cancer
2006
;
6
:
281
91
.
24.
Lyons
TR
,
Schedin
PJ
,
Borges
VF
. 
Pregnancy and breast cancer: when they collide
.
J Mammary Gland Biol Neoplasia
2009
;
14
:
87
98
.
25.
Warner
ET
,
Colditz
GA
,
Palmer
JR
,
Partridge
AH
,
Rosner
BA
,
Tamimi
RM
. 
Reproductive factors and risk of premenopausal breast cancer by age at diagnosis: are there differences before and after age 40?
Breast Cancer Res Treat
2013
;
142
:
165
75
.
26.
Allott
EH
,
Geradts
J
,
Cohen
SM
,
Khoury
T
,
Zirpoli
GR
,
Bshara
W
, et al
Frequency of breast cancer subtypes among African American women in the AMBER consortium
.
Breast Cancer Res
2018
;
20
:
12
.
27.
Rosenberg
L
,
Boggs
DA
,
Wise
LA
,
Adams-Campbell
LL
,
Palmer
JR
. 
Oral contraceptive use and estrogen/progesterone receptor-negative breast cancer among African American women
.
Cancer Epidemiol Biomarkers Prev
2010
;
19
:
2073
9
.
28.
DeSantis
CE
,
Ma
J
,
Goding Sauer
A
,
Newman
LA
,
Jemal
A
. 
Breast cancer statistics, 2017, racial disparity in mortality by state
.
CA Cancer J Clin
2017
;
67
:
439
48
.
29.
Allott
EH
,
Cohen
SM
,
Geradts
J
,
Sun
X
,
Khoury
T
,
Bshara
W
, et al
Performance of three-biomarker immunohistochemistry for intrinsic breast cancer subtyping in the AMBER Consortium
.
Cancer Epidemiol Biomarkers Prev
2016
;
25
:
470
8
.