Abstract
We pooled multiethnic data from four population-based studies and examined associations of menstrual and reproductive characteristics with breast cancer (BC) risk by tumor hormone receptor (HR) status [defined by estrogen receptor (ER) and progesterone receptor (PR)]. We estimated odds ratios and 95% confidence intervals using multivariable logistic regression, stratified by age (<50, ≥50 years) and ethnicity, for 5,186 HR+ (ER+ or PR+) cases, 1,365 HR− (ER− and PR−) cases and 7,480 controls. For HR+ BC, later menarche and earlier menopause were associated with lower risk in non-Hispanic whites (NHWs) and Hispanics, and higher parity and longer breast-feeding were associated with lower risk in Hispanics and Asian Americans, and suggestively in NHWs. Positive associations with later first full-term pregnancy (FTP), longer interval between menarche and first FTP and shorter time since last FTP were limited to younger Hispanics and Asian Americans. Except for nulliparity, reproductive characteristics were not associated with risk in African Americans. For HR− BC, lower risk was associated with later menarche, except in African Americans and older Asian Americans and with longer breast-feeding in Hispanics and Asian Americans only. In younger African Americans, HR− BC risk associated with higher parity (≥3 vs. 1 FTP) was increased fourfold in women who never breast-fed, but not in those with a breast-feeding history, suggesting that breast-feeding may mitigate the adverse effect of higher parity in younger African American women. Further work needs to evaluate why menstrual and reproductive risk factors vary in importance according to age and ethnicity.
Keywords: breast cancer, epidemiology, estrogen receptor status, progesterone receptor status, ethnicity, reproductive history
Introduction
Breast cancer incidence rates in the United States (U.S.) vary by ethnicity, with the highest rate (per 100,000) in non-Hispanic white (NHW) women (128.7), followed by African Americans (125.5), Hispanics (91.9) and Asian Americans/Pacific Islanders (90.7).1 The incidence of breast cancer defined by hormone receptor (HR) status also differs by ethnicity,2 with higher rates of HR negative [estrogen receptor (ER) negative and progesterone receptor (PR) negative)] breast cancer in African American and Hispanic women compared to NHW and Asian American women. The reasons underlying these ethnic disparities in breast cancer incidence overall and of specific subtypes are not understood, and ethnic differences in multiple factors likely contribute to the observed incidence patterns and differences by ethnicity.2
Menstrual and reproductive characteristics are well-established risk factors for breast cancer.3 Some associations differ by age,4 and there is growing evidence that associations differ by HR status5 or molecular tumor markers.6,7 Data on associations of menstrual and reproductive history with breast cancer defined by HR status come primarily from studies in NHW women. Few studies have reported on HR-specific associations in U.S. minority populations,8–16 and those that have, often lack sufficiently large or diverse study populations to examine associations of menstrual and reproductive characteristics with risk of specific breast cancer subtypes across multiple ethnic groups. We pooled case–control data for African American, Asian American, Hispanic and NHW women to evaluate associations of menstrual characteristics, pregnancy history and breast-feeding practices with risk of breast cancer defined by joint ER and PR status and differences in associations by age and ethnicity.
Materials and Methods
Study sample
The Breast Cancer Etiology in Minorities (BEM) Study harmonized interview and cancer registry data for participants in four population-based studies of female breast cancer described in more detail elsewhere17 and in the Supporting Information Methods. They include three case–control studies [the San Francisco Bay Area Breast Cancer Study (SFBCS), the 4-Corners Breast Cancer Study (4-CBCS) and the Los Angeles County Asian American Breast Cancer Study (AABCS)] and a family study that also included population controls [the Northern California site of the Breast Cancer Family Registry (NC-BCFR)]. A total of 9,234 women aged 18–79 years with a first primary invasive breast cancer and 7,767 control women without a history of breast cancer participated in these studies. The studies were approved by the Institutional Review Board of each participating institution, and study participants provided written informed consent.
In NC-BCFR, we limited the analysis to cases diagnosed from 1995 to 2003 for whom population controls were available, and excluded 325 cases who also participated in SFBCS, leaving 7,895 cases and 7,767 controls with interview data. Data on ER or PR status were available for 6,928 (88%) cases, including 5,457 cases diagnosed with HR+ (ER+ or PR+) breast cancer and 1,435 cases with HR− (ER− and PR−) breast cancer (Supporting Information Fig. S1).
Data collection and harmonization
For each study, trained interviewers administered structured questionnaires in English, Spanish, Cantonese or Mandarin at the participants’ home. The questionnaires assessed breast cancer risk factors in the reference year or exposure histories up to the reference year, defined as either the calendar year before diagnosis for cases, the calendar year before interview for controls in AABCS and NC-BCFR, or the calendar year before selection into the study for controls in SFBCS and 4-CBCS. Height and weight measured at interview and self-reported weight in the reference year were assessed for participants in the three case–control studies; for NC-BCFR participants, height and weight in the reference year were based on self-report. Data on ER and PR status were obtained from cancer registry records.
Study-specific data were harmonized to create derived variables using common definitions.17 Parity was defined as the number of full-term pregnancies (FTP). Lifetime breastfeeding was calculated summing duration of breast-feeding which was reported as a continuous measure for each live birth, except for NC-BCFR which assessed breast-feeding as a categorical measure (0, 1–12, 13–24, ≥25 months); for NC-BCFR, the midpoint of a category was assigned as the continuous value and 30 months was assigned to those reporting ≥25 months of breast-feeding. Given some evidence that breast-feeding may mitigate the adverse effect of higher parity in African American women,10,14,18 we also examined the joint association of parity and breast-feeding history. Women were classified as premenopausal if they still had menstrual periods or were pregnant, breast-feeding or perimenopausal during the reference year, and were under age 55 years. Women were classified as postmenopausal if they reported that prior to the reference year their periods had stopped naturally or due to surgery, medical treatment, or other reasons. Women who still had periods when they started using hormone therapy were classified as postmenopausal if they were ≥55 years of age; otherwise their menopausal status was classified as unknown. Body mass index (BMI) was calculated as self-reported weight (kg) in the reference year divided by measured or self-reported height (m) squared, and classified as <25.0, 25.0–29.9, or ≥30.0 kg/m2. Ethnicity was based on self-report and categorized as African American, Asian American (including Pacific Islanders), Hispanic (white or black and including Native Americans from 4-CBCS) or NHW.
Statistical analyses
We conducted separate analyses for HR+ and HR− breast cancer and stratified the analyses by age group (<50 vs. ≥50 years) and ethnicity. We used unconditional logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for menstrual variables (age at menarche, menopausal status, age at natural menopause) and pregnancy-related variables (nulliparity, age at first FTP, interval between menarche and first FTP, time since last FTP, parity, lifetime breast-feeding). All models were adjusted for age (continuous), study (AABCS, NC-BCFR, SFBCS, 4-CBCS) and year of diagnosis or selection/interview (1995–1998, 1999–2001, 2002–2007), and additionally for established breast cancer risk factors, including education (some high school or less, high school graduate, some college or vocational/technical school, college graduate or higher degree), family history of breast cancer among first-degree relatives (yes, no), personal history of benign breast disease (yes, no), alcohol consumption in reference year (none, <5, ≥5 drinks per week), BMI in reference year (<25.0, 25.0–29.9, ≥30.0 kg/m2), age menarche, age at first FTP, parity and breast-feeding, as noted in the footnotes of the tables. Analyses for all ethnicities combined were also adjusted for ethnicity, and analyses for all ages combined were additionally adjusted for menopausal status and an interaction term between BMI and menopausal status. Analyses of time since last FTP were limited to women ages <50 years. After excluding individuals with missing covariate data (153 HR+ cases, 45 HR− cases, 252 controls), the analyses were based on 5,304 HR+ cases, 1,390 HR− cases and 7,515 controls (Supporting Information Fig. S1).
Linear trends were assessed across ordinal values of categorical variables. We tested for differences in ORs by age or ethnicity by including interaction terms in the model. To assess differences in ORs between case groups (HR+ vs. HR-cases), we used polytomous regression models, and tested for differences in ORs between the two case groups using the Wald statistic p value. Forest plots were used to present ORs stratified by HR status, age and ethnicity. All statistical tests were two-sided, with p values <0.05 considered statistically significant. Statistical analyses were conducted using SAS version 9.4 software (SAS Institute, Inc., Cary, NC).
Data availability
The datasets used for the current study are available upon reasonable request from the corresponding author (E.M.J.), contingent upon approval by appropriate Institutional Review Boards and study Principal Investigators.
Results
Table 1 shows that compared to HR+ cases, HR− cases were more likely to be diagnosed at a younger age (mean: 50.5 years vs. 53.5 years) and with higher tumor grade (grade 3 or 4: 75% vs. 28%). Compared to controls, HR+ cases were more likely to have a college education, a first-degree family history of breast cancer, a personal history of benign breast disease and a BMI <25 kg/m2. HR− cases were more likely to have a first-degree family history of breast cancer.
Table 1.
Characteristics of hormone receptor positive cases, hormone receptor negative cases and controls
| HR+ cases (n = 5,304) | HR− cases (n = 1,390) | Controls (n = 7,515) | |
|---|---|---|---|
|
| |||
| Study | |||
| AABCS | 1,287 (24%) | 283 (20%) | 1,880 (25%) |
| NC-BCFR | 1,316 (25%) | 392 (28%) | 621 (8%) |
| SFBCS | 1,558 (29%) | 434 (31%) | 2,656 (35%) |
| 4-CBCS | 1,143 (22%) | 281 (20%) | 2,358 (31%) |
| Ethnicity | |||
| African American | 536 (10%) | 216 (16%) | 663 (9%) |
| Asian American | 1,802 (34%) | 420 (30%) | 1,973 (26%) |
| Hispanic | 1,322 (25%) | 418 (30%) | 2,336 (31%) |
| Non-Hispanic white | 1,644 (31%) | 336 (24%) | 2,543 (34%) |
| Year of diagnosis or selection/interview | |||
| 1995–1998 | 2,195 (41%) | 574 (41%) | 2,238 (30%) |
| 1999–2001 | 1,840 (35%) | 480 (35%) | 2,290 (30%) |
| 2002–2007 | 1,269 (24%) | 336 (24%) | 2,987 (40%) |
| Age at diagnosis or selection/interview (years) | |||
| <40 | 522 (10%) | 234 (17%) | 794 (11%) |
| 40–49 | 1,569 (30%) | 430 (31%) | 2,246 (30%) |
| 50–64 | 2,296 (43%) | 578 (42%) | 2,890 (38%) |
| ≥65 | 917 (17%) | 148 (11%) | 1,585 (21%) |
| Mean | 53.5 | 50.5 | 53.7 |
| Stage at diagnosis | |||
| Localized | 3,240 (61%) | 823 (59%) | |
| Regional | 1,807 (34%) | 495 (36%) | |
| Distant | 83 (2%) | 23 (2%) | |
| Missing | 174 (3%) | 49 (4%) | |
| Histologic grade | |||
| 1 | 1,037 (20%) | 27 (2%) | |
| 2 | 2,324 (44%) | 216 (16%) | |
| 3 | 1,361 (26%) | 960 (69%) | |
| 4 | 86 (2%) | 77 (6%) | |
| Missing | 496 (9%) | 110 (8%) | |
| Education | |||
| Some high school or less | 756 (14%) | 246 (18%) | 1,420 (19%) |
| High school graduate | 887 (17%) | 258 (19%) | 1,346 (18%) |
| Some college or vocational/technical school | 1,573 (30%) | 420 (30%) | 2,135 (28%) |
| College graduate or higher degree | 2,088 (39%) | 466 (34%) | 2,614 (35%) |
| First-degree family history of breast cancer | |||
| Yes | 1,105 (21%) | 237 (17%) | 870 (12%) |
| No | 4,199 (79%) | 1,153 (83%) | 6,645 (88%) |
| Personal history of benign breast disease | |||
| Yes | 1,084 (20%) | 217 (16%) | 1,082 (14%) |
| No | 4,220 (80%) | 1,173 (84%) | 6,433 (86%) |
| Alcohol consumption in reference year (drinks per week) | |||
| 0 | 3,371 (64%) | 922 (66%) | 4,714 (63%) |
| <5 | 1,024 (19%) | 273 (20%) | 1,640 (22%) |
| ≥5 | 909 (17%) | 195 (14%) | 1,161 (15%) |
| Body mass index (kg/m2) (premenopausal women) | |||
| <25 | 1,290 (64%) | 336 (54%) | 1,592 (57%) |
| 25–29.9 | 437 (22%) | 159 (25%) | 685 (24%) |
| ≥30 | 282 (14%) | 129 (21%) | 533 (19%) |
| Body mass index (kg/m2) (postmenopausal women) | |||
| <25 | 1,374 (44%) | 309 (43%) | 1,802 (41%) |
| 25–29.9 | 943 (30%) | 214 (30%) | 1,360 (31%) |
| ≥30 | 798 (26%) | 193 (27%) | 1,273 (29%) |
Abbreviations: ER, estrogen receptor; HR+, hormone receptor positive (ER+ or PR+); HR−, hormone receptor negative (ER− and PR−); PR, progesterone receptor.
Menstrual and reproductive characteristics of control women differed by ethnicity (Table 2). Mean age at menarche ranged from 12.7 years (NHWs and African Americans) to 12.9 years (Asian Americans), and mean age at natural menopause ranged from 46.0 years (Hispanics) to 49.6 years (Asian Americans). Nulliparity and low parity were more frequent in NHW and Asian American controls, whereas an early first FTP and a short interval between menarche and first FTP were more frequent in African Americans and Hispanics.
Table 2.
Menstrual and reproductive characteristics of control women, by ethnicity
| All controls | Non-Hispanic white controls | African American controls | Hispanic controls | Asian American controls | |
|---|---|---|---|---|---|
|
| |||||
| All controls | n = 7,515 | n = 2,543 | n = 663 | n = 2,336 | n = 1,973 |
| Age at menarche (years) | |||||
| <12 | 1,467 (20%) | 475 (19%) | 149 (22%) | 505 (22%) | 338 (17%) |
| 12 | 1,887 (25%) | 685 (27%) | 168 (25%) | 514 (22%) | 520 (26%) |
| 13 | 1,936 (26%) | 695 (27%) | 174 (26%) | 570 (24%) | 497 (25%) |
| ≥14 | 2,225 (30%) | 688 (27%) | 172 (26%) | 747 (32%) | 618 (31%) |
| Mean | 12.8 | 12.7 | 12.7 | 12.8 | 12.9 |
| Menopausal status | |||||
| Premenopausal | 2,810 (37%) | 781 (31%) | 195 (29%) | 796 (34%) | 1,038 (53%) |
| Postmenopausal | 4,435 (59%) | 1,683 (66%) | 431 (65%) | 1,414 (61%) | 907 (46%) |
| Unknown | 270 (4%) | 79 (3%) | 37 (6%) | 126 (5%) | 28 (1%) |
| Age at natural menopause (years)1 | |||||
| ≤45 | 1,048 (33%) | 456 (37%) | 83 (33%) | 410 (40%) | 99 (15%) |
| 46–50 | 1,052 (33%) | 375 (30%) | 68 (27%) | 332 (32%) | 277 (42%) |
| >50 | 1,069 (34%) | 409 (33%) | 103 (41%) | 280 (27%) | 277 (42%) |
| Mean | 47.2 | 46.9 | 47.7 | 46.0 | 49.6 |
| Nulliparity | |||||
| Nulliparous | 1,051 (14%) | 448 (18%) | 87 (13%) | 176 (8%) | 340 (17%) |
| Parous | 6,464 (86%) | 2,095 (82%) | 576 (87%) | 2,160 (92%) | 1,633 (83%) |
| Age at first FTP (years), parous women | |||||
| <20 | 1,274 (17%) | 293 (12%) | 246 (37%) | 644 (28%) | 91 (5%) |
| 20–24 | 2,373 (32%) | 886 (35%) | 212 (32%) | 849 (36%) | 426 (22%) |
| 25–29 | 1,742 (23%) | 571 (22%) | 76 (11%) | 445 (19%) | 650 (33%) |
| ≥30 | 1,075 (14%) | 345 (14%) | 42 (6%) | 222 (10%) | 466 (24%) |
| Mean | 24.3 | 24.6 | 21.3 | 22.7 | 26.9 |
| Interval between menarche and first FTP (years), parous women | |||||
| <8 | 1,745 (23%) | 450 (18%) | 288 (43%) | 828 (35%) | 179 (9%) |
| 8–10 | 1,393 (19%) | 522 (21%) | 127 (19%) | 501 (21%) | 243 (12%) |
| 11–14 | 1,586 (21%) | 563 (22%) | 89 (13%) | 453 (19%) | 481 (24%) |
| ≥15 | 1,740 (23%) | 560 (22%) | 72 (11%) | 378 (16%) | 730 (37%) |
| Mean | 11.5 | 11.8 | 8.6 | 9.9 | 14.1 |
| Parity, parous women2 | |||||
| 1 | 1,012 (13%) | 331 (13%) | 121 (18%) | 225 (10%) | 335 (17%) |
| 2 | 2,107 (28%) | 773 (30%) | 161 (24%) | 497 (21%) | 676 (34%) |
| 3 | 1,560 (21%) | 511 (20%) | 130 (20%) | 561 (24%) | 358 (18%) |
| ≥4 | 1,785 (24%) | 480 (19%) | 164 (25%) | 877 (38%) | 264 (13%) |
| Mean | 3.0 | 2.7 | 3.0 | 3.7 | 2.5 |
| Lifetime breast-feeding (months), parous women | |||||
| 0 | 2,373 (32%) | 620 (24%) | 308 (46%) | 646 (28%) | 799 (40%) |
| 1–12 | 2,251 (30%) | 840 (33%) | 162 (24%) | 705 (30%) | 544 (28%) |
| 13–24 | 872 (12%) | 321 (13%) | 54 (8%) | 330 (14%) | 167 (8%) |
| ≥25 | 968 (13%) | 314 (12%) | 52 (8%) | 479 (21%) | 123 (6%) |
| Mean | 11.9 | 11.3 | 8.1 | 17.1 | 7.3 |
| Parity and breast-feeding history | |||||
| Parity 1–2 | |||||
| Never | 1,290 (17%) | 361 (14%) | 169 (25%) | 263 (11%) | 497 (25%) |
| Ever | 1,829 (24%) | 743 (29%) | 113 (17%) | 459 (20%) | 514 (26%) |
| Parity ≥3 | |||||
| Never | 1,083 (14%) | 259 (10%) | 139 (21%) | 383 (16%) | 302 (15%) |
| Ever | 2,262 (30%) | 732 (29%) | 155 (23%) | 1,055 (45%) | 320 (16%) |
Abbreviation: FTP, full-term pregnancy.
231 controls with missing age at menopause.
Number of full-term pregnancies.
Menstrual characteristics and risk of breast cancer by HR status, age and ethnicity
Age at menarche.
Among all women combined, older age at menarche was associated with reduced risk of HR+ and HRbreast cancer, and for each subtype associations differed by ethnicity (pheterogeneity < 0.01 and 0.01, respectively; Table 3). For later menarche (≥14 vs. <12 years), risk reductions ranged from 25% to 42%, and statistically significant inverse trends were found for Hispanics and NHWs (both subtypes), African Americans (HR− subtype only) and younger Asian Americans (both subtypes; Supporting Information Table S1 and Fig. S2a). In African Americans, age at menarche was not associated with HR+ breast cancer risk, whereas in older Asian Americans, later menarche was associated with increased risk of HR+ (ptrend = 0.01) and HR− (ptrend = 0.01) breast cancer, and heterogeneity by age was statistically significant for both subtypes.
Table 3.
Menstrual characteristics and breast cancer risk by hormone receptor status and ethnicity
| All ethnicities |
Non-Hispanic white |
African American |
Hispanic |
Asian American |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cs | Cn | OR (95% CI)1,2 | Cs | Cn | OR (95% CI)1 | Cs | Cn | OR (95% CI)1 | Cs | Cn | OR (95% CI)1 | Cs | Cn | OR (95% CI)1 | |
|
| |||||||||||||||
| Age at menarche 3,4 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| <12 | 1,124 | 1,467 | 1.00 | 333 | 475 | 1.0 | 120 | 149 | 1.0 | 339 | 505 | 1.0 | 332 | 338 | 1.0 |
| 12 | 1,376 | 1,887 | 0.94 (0.84–1.05) | 449 | 685 | 0.91 (0.72–1.14) | 135 | 168 | 1.22 (0.83–1.79) | 344 | 514 | 0.93 (0.75–1.16) | 448 | 520 | 0.95 (0.76–1.17) |
| 13 | 1,338 | 1,936 | 0.85 (0.76–0.95) | 468 | 695 | 0.94 (0.75–1.18) | 129 | 174 | 0.97 (0.66–1.43) | 298 | 570 | 0.73 (0.58–0.91) | 443 | 497 | 0.90 (0.72–1.12) |
| ≥14 | 1,466 | 2,225 | 0.87 (0.77–0.97) | 394 | 688 | 0.75 (0.59–0.94) | 152 | 172 | 1.23 (0.84–1.81) | 341 | 747 | 0.69 (0.55–0.85) | 579 | 618 | 1.04 (0.84–1.29) |
| p trend | <0.01 | 0.02 | 0.53 | <0.01 | 0.68 | ||||||||||
| Per year | 0.97 (0.95–0.99) | 0.94 (0.89–0.98) | 1.03 (0.96–1.10) | 0.93 (0.88–0.97) | 1.01 (0.97–1.06) | ||||||||||
| p heterogeneity by ethnicity <0.01 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| <12 | 317 | 1,467 | 1.0 | 72 | 475 | 1.0 | 60 | 149 | 1.0 | 115 | 505 | 1.0 | 70 | 338 | 1.0 |
| 12 | 363 | 1,887 | 0.90 (0.75–1.07) | 103 | 685 | 0.80 (0.54–1.18) | 61 | 168 | 1.13 (0.71–1.82) | 91 | 514 | 0.68 (0.49–0.95) | 108 | 520 | 1.08 (0.75–1.56) |
| 13 | 354 | 1,936 | 0.83 (0.69–0.99) | 98 | 695 | 0.79 (0.53–1.18) | 49 | 174 | 0.67 (0.40–1.12) | 103 | 570 | 0.72 (0.52–0.99) | 104 | 497 | 1.03 (0.71–1.50) |
| ≥14 | 356 | 2,225 | 0.79 (0.66–0.94) | 63 | 688 | 0.58 (0.38–0.88) | 46 | 172 | 0.61 (0.36–1.03) | 109 | 747 | 0.66 (0.48–0.90) | 138 | 618 | 1.14 (0.79–1.63) |
| p trend | <0.01 | 0.01 | 0.02 | 0.02 | 0.55 | ||||||||||
| Per year | 0.96 (0.92–0.99) | 0.87 (0.80–0.96) | 0.88 (0.80–0.97) | 0.93 (0.87–0.99) | 1.05 (0.98–1.12) | ||||||||||
| p heterogeneity by ethnicity = 0.01 | |||||||||||||||
| p heterogeneity by subtype | 0.33 | 0.17 | 0.04 | 0.80 | 0.85 | ||||||||||
| Menopausal status 5,6 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| Premenopausal | 2,009 | 2,810 | 1.0 | 499 | 781 | 1.0 | 178 | 195 | 1.0 | 478 | 796 | 1.0 | 854 | 1,038 | 1.0 |
| Postmenopausal | 3,115 | 4,435 | 1.01 (0.89–1.14) | 1,085 | 1,683 | 1.09 (0.85–1.40) | 334 | 431 | 0.91 (0.58–1.40) | 776 | 1,414 | 0.91 (0.71–1.16) | 920 | 907 | 0.88 (0.71–1.11) |
| p heterogeneity by ethnicity <0.01 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| Premenopausal | 624 | 2,810 | 1.0 | 153 | 781 | 1.0 | 82 | 195 | 1.0 | 197 | 796 | 1.0 | 192 | 1,038 | 1.0 |
| Postmenopausal | 716 | 4,435 | 1.14 (0.95–1.38) | 176 | 1,683 | 0.86 (0.56–1.33) | 126 | 431 | 1.26 (0.71–2.21) | 194 | 1,414 | 0.89 (0.62–1.28) | 220 | 907 | 1.45 (1.01–2.09) |
| p heterogeneity by ethnicity <0.01 | |||||||||||||||
| p heterogeneity by subtype | 0.10 | 0.49 | 0.17 | 0.83 | <0.01 | ||||||||||
| Age at natural menopause (years) 5 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| ≤45 | 501 | 1,048 | 1.0 | 181 | 456 | 1.0 | 73 | 83 | 1.0 | 166 | 410 | 1.0 | 81 | 99 | 1.0 |
| 46–50 | 775 | 1,052 | 1.34 (1.15–1.56) | 254 | 375 | 1.56 (1.19–2.05) | 54 | 68 | 0.89 (0.52–1.53) | 181 | 332 | 1.28 (0.96–1.72) | 286 | 277 | 1.26 (0.87–1.84) |
| ≥51 | 871 | 1,069 | 1.58 (1.35–1.84) | 309 | 409 | 1.94 (1.48–2.54) | 79 | 103 | 0.82 (0.50–1.33) | 178 | 280 | 1.55 (1.15–2.08) | 305 | 277 | 1.28 (0.87–1.87) |
| p trend | <0.01 | <0.01 | 0.41 | <0.01 | 0.31 | ||||||||||
| Per year | 1.03 (1.02–1.04) | 1.04 (1.02–1.05) | 0.98 (0.95–1.01) | 1.03 (1.02–1.05) | 1.03 (1.00–1.06) | ||||||||||
| p heterogeneity by ethnicity = 0.05 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| ≤45 | 126 | 1,048 | 1.0 | 37 | 456 | 1.0 | 17 | 83 | 1.0 | 46 | 410 | 1.0 | 26 | 99 | 1.0 |
| 46–50 | 186 | 1,052 | 1.30 (1.00–1.68) | 47 | 375 | 1.31 (0.79–2.19) | 21 | 68 | 1.21 (0.54–2.74) | 58 | 332 | 1.46 (0.94–2.26) | 60 | 277 | 0.73 (0.40–1.31) |
| ≥51 | 152 | 1,069 | 1.22 (0.93–1.60) | 35 | 409 | 1.24 (0.73–2.11) | 21 | 103 | 1.02 (0.47–2.22) | 38 | 280 | 1.17 (0.72–1.89) | 58 | 277 | 0.78 (0.43–1.43) |
| p trend | 0.16 | 0.41 | 0.98 | 0.46 | 0.59 | ||||||||||
| Per year | 1.02 (1.00–1.04) | 1.01 (0.98–1.04) | 1.00 (0.96–1.05) | 1.03 (1.00–1.06) | 1.00 (0.95–1.05) | ||||||||||
| p heterogeneity by ethnicity = 0.93 | |||||||||||||||
| p heterogeneity by subtype | 0.06 | 0.09 | 0.75 | 0.29 | 0.25 | ||||||||||
Covariates were categorized as follows: Study (AABCS, NC-BCFR, SFBCS, 4-CBCS); year of diagnosis or selection/interview (1995–1998, 1999–2001, 2002–2007); ethnicity (non-Hispanic white, African American, Hispanic, Asian American); education (some high school or less, high school graduate, some college or vocational/technical school, college graduate or higher degree); first-degree family history of breast cancer (yes, no); personal history of benign breast disease (yes, no); alcohol consumption in reference year (none, <4, ≥5 drinks per week); BMI in reference year (<25.0, 25.0–29.9, ≥30.0 kg/m2); age at menarche (<12, 12, 13, ≥14 years); menopausal status (premenopausal, postmenopausal, unknown); age at first FTP (nulliparous, <20, 20–24, 25–29, ≥30 years); parity (nulliparous, 1, 2, 3, ≥4 FTPs); lifetime breast-feeding (nulliparous, 0, 1–12, 13–24, ≥25 months).
Adjusted for age at diagnosis or selection/interview (continuous), study, year of diagnosis or selection/interview, education, first-degree family history of breast cancer, personal history of benign breast disease, alcohol consumption, BMI, age at first FTP, parity and breast-feeding.
Additionally adjusted for ethnicity.
Additionally adjusted for menopausal status and interaction between menopausal status and BMI.
Additionally adjusted for age at first FTP, parity and breast-feeding.
Additionally adjusted for age at menarche, age at first FTP, parity and breast-feeding.
Excludes 180 HR+ cases, 50 HR− cases and 270 controls with unknown menopausal status.
Abbreviations: BMI, body mass index; CI, confidence interval; Cs, cases; Cn, controls; ER, estrogen receptor; FTP, full-term pregnancy; HR+, hormone receptor positive (ER+ or PR+); HR−, hormone receptor negative (ER− and PR−); OR, odds ratio; PR, progesterone receptor.
Menopausal status and age at natural menopause.
Menopausal status was not associated with risk of HR+ or HR− breast cancer in any ethnicity group, except Asian Americans, in whom postmenopausal women were at increased risk of HR− breast cancer (Table 3, Supporting Information Fig. S2b). In women with natural menopause, a positive association with later menopause was found for HR+ breast cancer only (pheteroogeneity = 0.06). The association differed by ethnicity (pheterogeneity = 0.05), with increased risks associated with later menopause (≥51 vs. ≤45 years) observed in NHWs (94%, ptrend < 0.01) and Hispanics (55%, ptrend < 0.01) only (Table 3, Supporting Information Fig. S2c).
Pregnancy-related characteristics and risk of breast cancer by HR status, age and ethnicity
Nulliparity.
Nulliparous women were at increased risk of HR+ breast cancer only (pheterogeneity < 0.01), and the association differed by ethnicity (pheterogeneity < 0.01). The increase in risk was borderline in NHWs (19%), and ranged from 48% to 85% in the other ethnicity groups (Table 4). Positive associations were found in younger and older Hispanic and Asian American women, and in older, but not younger African American women (Supporting Information Table S1 and Fig. S2d).
Table 4.
Pregnancies and breast-feeding and risk of breast cancer by hormone receptor status and ethnicity
| All ethnicities |
Non-Hispanic white |
African American |
Hispanic |
Asian American |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cs | Cn | OR (95% CI)1,2 | Cs | Cn | OR (95% CI)1 | Cs | Cn | OR (95% CI)1 | Cs | Cn | OR (95% CI)2 | Cs | Cn | OR (95% CI)2 | |
|
| |||||||||||||||
| Nulliparity | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| Parous | 4,204 | 6,464 | 1.0 | 1,298 | 2,095 | 1.0 | 431 | 576 | 1.0 | 1,148 | 2,160 | 1.0 | 1,327 | 1,633 | 1.0 |
| Nulliparous | 1,100 | 1,051 | 1.46 (1.32–1.62) | 346 | 448 | 1.19 (0.97–1.45) | 105 | 87 | 1.48 (1.03–2.13) | 174 | 176 | 1.57 (1.21–2.02) | 475 | 340 | 1.85 (1.56–2.19) |
| p heterogeneity by ethnicity <0.01 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| Parous | 1,170 | 6,464 | 1.0 | 263 | 2,095 | 1.0 | 188 | 576 | 1.0 | 373 | 2,160 | 1.0 | 346 | 1,633 | 1.0 |
| Nulliparous | 220 | 1,051 | 1.02 (0.86–1.22) | 73 | 448 | 1.00 (0.68–1.45) | 28 | 87 | 0.78 (0.45–1.35) | 45 | 176 | 1.22 (0.82–1.81) | 74 | 340 | 1.13 (0.83–1.53) |
| p heterogeneity by ethnicity = 0.29 | |||||||||||||||
| p heterogeneity by subtype | <0.01 | 0.24 | 0.04 | 0.08 | <0.01 | ||||||||||
| Age at first FTP (years), parous women 3 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| <20 | 694 | 1,274 | 1.0 | 178 | 293 | 1.0 | 174 | 246 | 1.0 | 279 | 644 | 1.0 | 63 | 91 | 1.0 |
| 20–24 | 1,414 | 2,373 | 1.04 (0.91–1.17) | 517 | 886 | 1.07 (0.83–1.40) | 152 | 212 | 1.04 (0.74–1.46) | 440 | 849 | 1.14 (0.93–1.40) | 305 | 426 | 0.95 (0.65–1.40) |
| 25–29 | 1,231 | 1,742 | 1.13 (0.98–1.30) | 372 | 571 | 1.02 (0.76–1.36) | 61 | 76 | 1.06 (0.66–1.70) | 249 | 445 | 1.11 (0.87–1.42) | 549 | 650 | 1.14 (0.77–1.68) |
| ≥30 | 865 | 1,075 | 1.15 (0.98–1.35) | 231 | 345 | 0.99 (0.70–1.40) | 44 | 42 | 1.60 (0.90–2.83) | 180 | 222 | 1.38 (1.02–1.86) | 410 | 466 | 1.14 (0.76–1.72) |
| p trend | 0.045 | 0.78 | 0.2 | 0.07 | 0.17 | ||||||||||
| p heterogeneity by ethnicity = 0.17 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| <20 | 264 | 1,274 | 1.0 | 50 | 293 | 1.0 | 83 | 246 | 1.0 | 111 | 644 | 1.0 | 20 | 91 | 1.0 |
| 20–24 | 391 | 2,373 | 0.84 (0.69–1.01) | 90 | 886 | 0.66 (0.42–1.03) | 63 | 212 | 0.96 (0.61–1.51) | 152 | 849 | 0.98 (0.73–1.31) | 86 | 426 | 0.87 (0.48–1.59) |
| 25–29 | 323 | 1,742 | 0.91 (0.73–1.12) | 78 | 571 | 0.61 (0.36–1.01) | 26 | 76 | 0.97 (0.51–1.83) | 77 | 445 | 0.91 (0.64–1.31) | 142 | 650 | 1.01 (0.55–1.84) |
| ≥30 | 192 | 1,075 | 0.74 (0.57–0.96) | 45 | 345 | 0.49 (0.27–0.92) | 16 | 42 | 0.83 (0.38–1.83) | 33 | 222 | 0.68 (0.41–1.11) | 98 | 466 | 0.85 (0.45–1.61) |
| p trend | 0.08 | 0.03 | 0.67 | 0.18 | 0.78 | ||||||||||
| p heterogeneity by ethnicity = 0.89 | |||||||||||||||
| p heterogeneity by subtype | <0.01 | 0.11 | 0.50 | 0.01 | 0.35 | ||||||||||
| Interval between menarche and first FTP (years), parous women 4 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| <8 | 899 | 1,745 | 1.0 | 239 | 450 | 1.0 | 208 | 288 | 1.0 | 342 | 828 | 1.0 | 110 | 179 | 1.0 |
| 8–10 | 885 | 1,393 | 1.17 (1.03–1.33) | 304 | 522 | 1.11 (0.86–1.44) | 88 | 127 | 0.98 (0.67–1.44) | 288 | 501 | 1.23 (0.99–1.54) | 205 | 243 | 1.26 (0.90–1.77) |
| 11–14 | 1,057 | 1,586 | 1.12 (0.98–1.28) | 363 | 563 | 1.12 (0.86–1.46) | 68 | 89 | 1.06 (0.68–1.66) | 245 | 453 | 1.02 (0.80–1.28) | 381 | 481 | 1.18 (0.85–1.62) |
| ≥15 | 1,363 | 1,740 | 1.21 (1.05–1.39) | 392 | 560 | 1.15 (0.86–1.54) | 67 | 72 | 1.33 (0.83–2.15) | 273 | 378 | 1.24 (0.96–1.60) | 631 | 730 | 1.31 (0.94–1.82) |
| p trend | 0.02 | 0.38 | 0.3 | 0.26 | 0.2 | ||||||||||
| p heterogeneity by ethnicity = 0.51 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| <8 | 319 | 1,745 | 1.0 | 61 | 450 | 1.0 | 92 | 288 | 1.0 | 129 | 828 | 1.0 | 37 | 179 | 1.0 |
| 8–10 | 242 | 1,393 | 0.97 (0.80–1.18) | 55 | 522 | 0.82 (0.52–1.30) | 33 | 127 | 0.79 (0.47–1.33) | 95 | 501 | 1.08 (0.78–1.50) | 59 | 243 | 1.24 (0.74–2.06) |
| 11–14 | 274 | 1,586 | 0.94 (0.76–1.15) | 65 | 563 | 0.76 (0.47–1.23) | 34 | 89 | 1.08 (0.60–1.92) | 81 | 453 | 0.96 (0.68–1.36) | 94 | 481 | 0.98 (0.60–1.61) |
| ≥15 | 335 | 1,740 | 0.90 (0.72–1.12) | 82 | 560 | 0.64 (0.37–1.09) | 29 | 72 | 0.95 (0.50–1.79) | 68 | 378 | 0.91 (0.61–1.36) | 156 | 730 | 1.00 (0.60–1.67) |
| p trend | 0.34 | 0.1 | 0.97 | 0.59 | 0.66 | ||||||||||
| p heterogeneity by ethnicity = 0.77 | |||||||||||||||
| p heterogeneity by subtype | 0.02 | 0.06 | 0.87 | 0.22 | 0.32 | ||||||||||
| Parity (number of FTPs), parous women 4 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| 1 | 819 | 1,012 | 1.0 | 212 | 331 | 1.0 | 102 | 121 | 1.0 | 194 | 225 | 1.0 | 311 | 335 | 1.0 |
| 2 | 1,619 | 2,107 | 0.96 (0.85–1.08) | 566 | 773 | 1.17 (0.91–1.50) | 129 | 161 | 0.94 (0.62–1.42) | 343 | 497 | 0.86 (0.66–1.12) | 581 | 676 | 0.89 (0.72–1.09) |
| 3 | 905 | 1,560 | 0.76 (0.66–0.87) | 275 | 511 | 0.83 (0.62–1.12) | 95 | 130 | 1.03 (0.66–1.61) | 267 | 561 | 0.60 (0.45–0.80) | 268 | 358 | 0.75 (0.58–0.96) |
| ≥4 | 861 | 1,785 | 0.67 (0.58–0.78) | 245 | 480 | 0.81 (0.60–1.10) | 105 | 164 | 1.03 (0.65–1.63) | 344 | 877 | 0.55 (0.41–0.74) | 167 | 264 | 0.57 (0.42–0.77) |
| p trend | <0.01 | 0.01 | 0.79 | <0.01 | <0.01 | ||||||||||
| Per FTP | 0.89 (0.86–0.91) | 0.92 (0.86–0.98) | 0.99 (0.91–1.08) | 0.86 (0.82–0.91) | 0.87 (0.81–0.93) | ||||||||||
| p heterogeneity by ethnicity <0.01 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| 1 | 223 | 1,012 | 1.0 | 48 | 331 | 1.0 | 46 | 121 | 1.0 | 45 | 225 | 1.0 | 84 | 335 | 1.0 |
| 2 | 422 | 2,107 | 0.91 (0.75–1.10) | 103 | 773 | 0.99 (0.63–1.56) | 59 | 161 | 0.73 (0.43–1.22) | 118 | 497 | 1.06 (0.70–1.61) | 142 | 676 | 0.74 (0.53–1.04) |
| 3 | 289 | 1,560 | 0.88 (0.71–1.09) | 75 | 511 | 1.07 (0.65–1.76) | 47 | 130 | 0.89 (0.50–1.57) | 90 | 561 | 0.79 (0.51–1.22) | 77 | 358 | 0.67 (0.45–1.00) |
| ≥4 | 236 | 1,785 | 0.65 (0.51–0.82) | 37 | 480 | 0.60 (0.34–1.05) | 36 | 164 | 0.60 (0.33–1.10) | 120 | 877 | 0.83 (0.52–1.30) | 43 | 264 | 0.50 (0.30–0.81) |
| p trend | <0.01 | 0.1 | 0.17 | 0.17 | <0.01 | ||||||||||
| Per FTP | 0.92 (0.88–0.97) | 0.87 (0.76–0.99) | 0.93 (0.82–1.05) | 0.97 (0.90–1.04) | 0.86 (0.77–0.97) | ||||||||||
| p heterogeneity by ethnicity = 0.68 | |||||||||||||||
| p heterogeneity by subtype | 0.57 | 0.93 | 0.31 | 0.048 | 0.98 | ||||||||||
| Lifetime breast-feeding (months), parous women 5 | |||||||||||||||
| HR+ breast cancer | |||||||||||||||
| 0 | 1,732 | 2,373 | 1.0 | 442 | 620 | 1.0 | 236 | 308 | 1.0 | 432 | 646 | 1.0 | 622 | 799 | 1.0 |
| 1–12 | 1,538 | 2,251 | 0.99 (0.90–1.10) | 530 | 840 | 0.88 (0.72–1.07) | 120 | 162 | 0.89 (0.63–1.24) | 398 | 705 | 0.85 (0.70–1.03) | 490 | 544 | 0.82 (0.68–1.0) |
| 13–24 | 463 | 872 | 0.88 (0.76–1.01) | 169 | 321 | 0.75 (0.56–1.00) | 38 | 54 | 0.95 (0.56–1.62) | 151 | 330 | 0.79 (0.61–1.03) | 105 | 167 | 0.63 (0.47–0.86) |
| ≥25 | 471 | 968 | 0.86 (0.74–1.00) | 157 | 314 | 0.88 (0.65–1.19) | 37 | 52 | 1.06 (0.62–1.80) | 167 | 479 | 0.69 (0.54–0.89) | 110 | 123 | 0.76 (0.54–1.07) |
| p trend | 0.02 | 0.17 | 0.96 | <0.01 | <0.01 | ||||||||||
| Per 12 months | 0.96 (0.93–0.99) | 0.99 (0.92–1.06) | 0.99 (0.88–1.11) | 0.95 (0.91–0.99) | 0.92 (0.86–0.99) | ||||||||||
| p heterogeneity by ethnicity <0.01 | |||||||||||||||
| HR− breast cancer | |||||||||||||||
| 0 | 484 | 2,373 | 1.0 | 80 | 620 | 1.0 | 111 | 308 | 1.0 | 125 | 646 | 1.0 | 168 | 799 | 1.0 |
| 1–12 | 432 | 2,251 | 1.00 (0.86–1.17) | 110 | 840 | 0.94 (0.65–1.36) | 52 | 162 | 0.95 (0.61–1.48) | 140 | 705 | 0.87 (0.65–1.16) | 130 | 544 | 0.75 (0.55–1.02) |
| 13–24 | 113 | 872 | 0.77 (0.61–0.97) | 34 | 321 | 0.89 (0.52–1.51) | 13 | 54 | 0.80 (0.39–1.63) | 46 | 330 | 0.69 (0.46–1.03) | 20 | 167 | 0.43 (0.25–0.74) |
| ≥25 | 141 | 968 | 0.88 (0.69–1.11) | 39 | 314 | 1.17 (0.69–1.99) | 12 | 52 | 0.92 (0.42–2.01) | 62 | 479 | 0.71 (0.48–1.04) | 28 | 123 | 0.58 (0.34–0.99) |
| p trend | 0.09 | 0.72 | 0.59 | 0.04 | <0.01 | ||||||||||
| Per 12 months | 0.92 (0.88–0.97) | 1.00 (0.88–1.14) | 0.88 (0.74–1.06) | 0.93 (0.87–1.00) | 0.79 (0.68–0.91) | ||||||||||
| p heterogeneity by ethnicity = 0.72 | |||||||||||||||
| p heterogeneity by subtype | 0.64 | 0.61 | 0.29 | 0.81 | 0.12 | ||||||||||
Covariates were categorized as follows: Study (AABCS, NC-BCFR, SFBCS, 4-CBCS); year of diagnosis or selection/interview (1995–1998, 1999–2001, 2002–2007); ethnicity (non-Hispanic white, African American, Hispanic, Asian American); education (some high school or less, high school graduate, some college or vocational/technical school, college graduate or higher degree); first-degree family history of breast cancer (yes, no); personal history of benign breast disease (yes, no); alcohol consumption in reference year (none, <4, ≥5 drinks per week); BMI in reference year (<25.0, 25.0–29.9, ≥30.0 kg/m2); age at menarche (<12, 12, 13, >14 years); menopausal status (premenopausal, postmenopausal, unknown); age at first FTP (nulliparous, <20, 20–24, 25–29, >30 years); parity (nulliparous, 1, 2, 3, ≥4 FTPs); lifetime breast-feeding (nulliparous, 0, 1–12, 13–24, ≥25 months).
Adjusted for age at diagnosis or selection/interview (continuous), study, year of diagnosis or selection/interview, education, first-degree family history of breast cancer, personal history of benign breast disease, age at menarche, menopausal status, alcohol consumption, BMI and interaction between menopausal status and BMI.
Additionally adjusted for ethnicity.
Additionally adjusted for parity and breast-feeding.
Additionally adjusted for age at first FTP and breast-feeding.
Additionally adjusted for age at first FTP and parity.
Abbreviations: BMI, body mass index; CI, confidence interval; Cs, cases; Cn, controls; ER, estrogen receptor; FTP, full-term pregnancy; HR+, hormone receptor positive (ER+ or PR+); HR−, hormone receptor negative (ER− and PR−); OR, odds ratio; PR, progesterone receptor.
Age at first FTP.
A first FTP at a later age was associated with increased risk of HR+ breast cancer only (pheterogeneity < 0.01; Table 4) and limited to younger women (Supporting Information Table S1 and Fig. S2e). The association differed by ethnicity (pheterogeneity < 0.01), and alate first FTP (≥ 30 vs. <20 years) was associated with increased risk in young Hispanics (75%, ptrend = 0.02) and young Asian Americans (78%, ptrend = 0.04) only.
Interval between menarche and first FTP.
A positive association with a longer interval was found for HR+ breast cancer only (pheterogeneity = 0.02; Table 4), and limited to younger women (Supporting Information Table S1 and Fig. S2f). The association differed by ethnicity (pheterogeneity = 0.01). Risk associated with a longer interval (≥15 vs. <8 years) was increased in young Hispanics only (73%, ptrend = 0.03), and suggestively in young Asian Americans (61%, ptrend = 0.10).
Time since last FTP.
Positive associations with shorter time since last FTP were observed for HR+ breast cancer only, and limited to younger Hispanics (ptrend = 0.02) and Asian Americans (ptrend = 0.04)(Supporting Information Table S1 and Fig. S2g).
Parity.
Increasing number of FTPs was associated with reduced risk of HR+ (ptrend < 0.01) and HR− (ptrend < 0.01) breast cancer (Table 4). For HR+ breast cancer, the association differed by ethnicity (pheterogeneity < 0.01), with lower risks found in all ethnicity groups, except African Americans. Risk reductions per FTP were 14% for Hispanic, 13% for Asian American and 8% for NHW women. In Hispanics and Asian Americans, significant inverse trends were found for younger and older women (Supporting Information Table S1 and Fig. S2h), with risk reductions per FTP ranging from 12% to 20%. For HR− breast cancer, risk was reduced by 50% in Asian Americans with high parity (≥4 vs. 1 FTP; ptrend < 0.01; Table 4), but only in older women (51%, ptrend < 0.01).
Breast-feeding.
Longer lifetime breast-feeding was associated with reduced risk of HR+ breast cancer (ptrend = 0.02), and the association differed by ethnicity (pheterogeneity < 0.01; Table 4). Inverse trends were found in younger (ptrend = 0.02) and older (ptrend = 0.04) Hispanics, whereas in Asian Americans, inverse trends were borderline in younger (ptrend = 0.05) and older (ptrend = 0.08) women (Supporting Information Table 1 and Fig. S2i). In NHW and African American women, breast-feeding was not associated with HR+ breast cancer risk, neither in younger or older women. For HR− breast cancer (Table 4), risk reductions associated with longer breastfeeding were found in Hispanics (ptrend = 0.04) and Asian Americans (ptrend < 0.01)only.
Joint association of parity and breast-feeding in African American women.
In younger African American women, risk of HR− breast cancer was increased in those with higher parity (≥3 vs. 1 FTP) who never breast-fed (OR = 4.59, 95% CI = 1.69–12.5; pheterogeneity < 0.01; Table 5). This association was unique to African Americans and not seen in other ethnicity groups (pheterogeneity ≥ 0.05, data not shown).
Table 5.
Parity and breast-feeding and risk of HR− breast cancer among African American women, by age
| African American |
||||||
|---|---|---|---|---|---|---|
| <50 years | ≥50 years | |||||
|
| ||||||
| Cs | Cn | OR (95% CI)1 | Cs | Cn | OR (95% CI)1 | |
| Nulliparity | ||||||
| Nulliparous | 15 | 40 | 1.0 | 13 | 47 | 1.0 |
| Parous | 79 | 186 | 1.55 (0.68–3.51) | 109 | 390 | 1.00 (0.48–2.11) |
| Nulliparity and breast-feeding history | ||||||
| Nulliparous | 15 | 40 | 1.0 | 13 | 47 | 1.0 |
| Parous, never | 44 | 99 | 1.76 (0.74–4.23) | 67 | 209 | 1.07 (0.49–2.32) |
| Parous, ever | 35 | 87 | 1.36 (0.57–3.29) | 42 | 181 | 0.94 (0.42–2.07) |
| Parity | ||||||
| 1 | 24 | 54 | 1.0 | 22 | 67 | 1.0 |
| 2 | 26 | 73 | 1.00 (0.46–2.17) | 33 | 88 | 0.64 (0.31–1.33) |
| ≥3 | 29 | 59 | 1.81 (0.73–4.51) | 54 | 235 | 0.55 (0.28–1.08) |
| p trend | 0.23 | 0.09 | ||||
| Parity and breast-feeding history | ||||||
| Parity 1–2 | ||||||
| Never | 24 | 75 | 1.0 | 35 | 94 | 1.0 |
| Ever | 26 | 52 | 1.41 (0.63–3.15) | 20 | 61 | 1.24 (0.61–2.54) |
| Parity ≥3 | ||||||
| Never | 20 | 24 | 4.59 (1.69–12.5) | 32 | 115 | 0.92 (0.46–1.81) |
| Ever | 9 | 35 | 0.62 (0.21–1.84) | 22 | 120 | 0.64 (0.31–1.32) |
| p heterogeneity by parity and breast-feeding history | <0.01 | 0.26 | ||||
Adjusted for age at diagnosis or selection/interview (continuous), study, year of diagnosis or selection/interview, education, first-degree family history of breast cancer, personal history of benign breast disease, age at menarche, alcohol consumption and BMI.
Discussion
In this large population-based pooled dataset, we observed notable differences by ethnicity in menstrual and reproductive patterns and associations with risk of breast cancer defined by tumor HR status. Associations with HR+ breast cancer were generally consistent between Hispanic and Asian American women and largely in agreement with the literature predominantly from studies in NHW women.5,19–24 For NHWs, associations with reproductive characteristics were weaker, but in the expected directions, whereas in African American women, most menstrual and reproductive characteristics were not associated with HR+ breast cancer risk. For HR− breast cancer, inverse associations in the expected direction were observed for age at menarche and breast-feeding, although there were some differences by ethnicity. In younger African Americans, risk associated with higher parity was increased in the absence of breast-feeding.
Data are sparse on HR-specific associations with menstrual and reproductive characteristics in Hispanic8 and Asian American9 women. More data are available for African American women,10–12,16 including from the African American Breast Cancer Epidemiology and Risk (AMBER) Consortium, the largest pooled analysis for African American women to date.14,15 Thus, the present study contributes important new information on ethnic differences in HR-specific associations with menstrual and reproductive characteristics, which were defined in the same manner, and based on identical cut points for direct comparison.
Consistent with a meta-analysis25 and recent studies,5,19,20,26,27 we found lower risks of HR+ and HRbreast cancer associated with older age at menarche, although for African Americans, we found an inverse association with HR− breast cancer only, whereas other studies in African American women reported inverse associations for both subtypes.13,15,16 For Asian Americans, we found inverse associations with HR+ and HR− breast cancer in younger women only. The positive associations with later menarche in older women were unexpected. A study of Chinese women ages 20–70 years also reported inverse associations for both subtypes, but did not stratify the analyses by age or menopausal status.28 In Hispanic women, we found that later menarche was associated with lower risk of both subtypes. 4-CBCS previously reported borderline reduced risk associated with later menarche for ER+ breast cancer only,8 whereas a Latin American study found an inverse association for ER− breast cancer only.29 The present pooled analysis was based on a much larger sample size of Hispanic women, and our findings add to the evidence for Hispanic women that early menarche is likely an important risk factor for HR− breast cancer, a subtype with few known risk factors.
Later menopause has been associated with higher breast cancer risk, and consistent with other studies,25,30,31 we found a positive association for HR+ breast cancer only, although limited to NHW and Hispanic women. Late menarche and early menopause have been related to lower lifetime exposure to sex hormones, and thus lower breast cancer risk. We saw such a risk pattern most consistently in Hispanic women.
The present results emphasize the importance of the timing of pregnancies in relation to age at diagnosis and age at menarche, although there were some differences in associations by ethnicity. Consistent with studies primarily in NHW women,5,19–21,24,26 we found a positive association with older age at first FTP for HR+ breast cancer, but only in younger Hispanics and suggestively in younger Asian Americans. Unlike AMBER,15 we found no association in African American women. For HR− breast cancer, we found no association with age at first FTP, consistent with the findings from AMBER,15 whereas the Black Women’s Health Study reported an increased risk of ER− subtype for women ages <45 years who had a first FTP at a young age.16 For younger NHW women, we also found a trend of increasing risk of HR− breast cancer with younger age at first FTP. A very large cancer registry study from Denmark found no association of age at first live birth with risk ER− breast cancer for women ages <50 years,24 but risk was increased for ER− HER2− breast cancer in women with a first birth before age 20 years and multiple live births. Thus, the effect of early pregnancies on risk of ER− subtypes warrants further investigation.
Our finding of a positive association between longer interval between menarche and first FTP and risk of HR+ breast cancer is consistent with studies primarily in NHW women.21,32,33 The association, however, was limited to younger Hispanic women and suggestively to younger Asian American women. Unlike AMBER,15 we found no association in African American women.
Previous studies have shown that a recent FTP is associated with a short-term increase in breast cancer risk that wanes after about 10 years,5,24,34 although a recent pooled analysis of prospective data from women under age 55 years suggests that in parous compared to nulliparous women the risk of ER+ breast cancer remains elevated for more than 20 years.35 In agreement with these findings, in women ages <50 years, we also found that a recent pregnancy (<3 vs. ≥20 years since last FTP) was associated with increased risk, but only for HR+ subtype and limited to Hispanic and Asian American women. In contrast, the association with a recent pregnancy did not differ by subtype in the Danish cancer registry study.24
Unlike our findings for Hispanic and Asian American women, associations with parity and breast-feeding were weak and nonsignificant in NHW women. Other studies in predominantly NHW women also reported relatively small risk reductions of 9%36 and 8%5 associated with high parity (≥4 vs. 1 birth) and a lack of association with breast-feeding.36 For African American women, we found no associations between parity and breast-feeding and risk of HR+ breast cancer, consistent with the findings from AMBER,14 suggesting that associations with parity and breast-feeding may be different for African American women compared to other ethnicity groups.
Although based on small numbers, risk of HR− breast cancer associated with higher parity (≥3 vs. 1) was increased fourfold in younger African Americans who never breast-fed, but not in those with a history of breast-feeding. This finding adds to the growing evidence that breast-feeding may mitigate the adverse effects of higher parity, as we and others have previously reported for ER−,14 ER− PR−,10,37,38 triple negative17 and basal-like18,36 breast cancer. This fi nding is particularly important for primary prevention, as few risk factors have been identified for ER− breast cancer which is more frequently diagnosed among African American and Hispanic women39,40 and has worse survival than ER+ breast cancer.40–42 Palmer et al. suggested that higher parity and lower prevalence of breast-feeding may contribute to the higher incidence of ER− breast cancer in African American women.14 In the present study, a pattern of high parity (≥3 FTPs) and no breast-feeding history that was twice as common in African American controls (21%) compared to NHWs (10%). Some studies in predominantly NHW women also reported positive associations between higher parity and ERsubtypes in younger women.5,24,38 In the present study, however, the increased risk of HR− breast cancer associated with higher parity, in the absence of breast-feeding, was limited to younger African American women.
The associations between menstrual and reproductive characteristics and risk of HR+ breast cancer tended to be stronger for women under age 50 years, consistent with other studies,16,43 suggesting that these factors may play a more important role in breast cancer development in younger than older women. Early-life reproductive events in particular appear to be important determinants of breast cancer risk.44 We found that the timing of key reproductive events differed substantially by ethnicity and early menarche, early first FTP and a short interval between these two events were most common in African American controls. However, in the present analysis, these reproductive variables were not associated with risk of HR+ breast cancer in African American women, unlike AMBER which reported reduced risk associated with later menarche and shorter interval between menarche and first birth.15 Our sample size for African American women may have not been large enough to detect significant associations.
Menstrual and reproductive characteristics may influence breast cancer risk through effects on sex hormones,45–47 hormonally mediated changes in breast structure,48 or other mechanisms. The time interval between menarche and first FTP is thought to be a window of heightened breast cancer susceptibility when breast tissue is relatively undifferentiated and potentially more susceptible to carcinogens, whereas a first FTP at an early age is thought to induce protective differentiation of mammary cells in the terminal duct lobular unit, with additional pregnancies and breast-feeding inducing further differentiation, thereby protecting breast tissue from carcinogenic transformation.48 Studies comparing breast tissue of nulliparous and parous women have shown differences in expression of genes, some of which are involved in cell differentiation, regulation of proliferation and cell growth and other processes.49,50 Parity-related molecular changes appear to be preserved in women with ER+ breast cancer, but disrupted in those with ER− tumors.51 It is not known whether these mechanisms differ by ethnicity and contribute to the lack of associations of pregnancies and breast-feeding with HR+ breast cancer among African American women.
Some limitations need to be considered when interpreting the present results. Tumor subtype was based on medical records and most breast cancer cases included in the present pooled analysis were diagnosed before HER2 status was routinely collected by cancer registries. We therefore defined subtype by ER and PR status only, as information on HER2 status was available only for a subset of cases in our pooled dataset.17 It is reassuring that in a pooled analysis of prospective data associations with parity, age at first birth and interval between menarche and first birth were similar for tumors defined by joint ER and PR status or joint ER, PR and HER2 status.5 On the other hand, Anderson et al. showed that an adverse effect of the joint variable of parity and age at first live birth only emerged when HER2 status in addition to ER status was considered,24 with an increased risk of ER− HER2− subtype noted for women with multiple pregnancies before age 20 years. These findings underline the importance of considering more refined molecular subtypes than hormone receptor status only and the need for large sample sizes to study the less common breast cancer subtypes. In the present study, the menstrual and reproductive characteristics were assessed by questionnaire, which is susceptible to recall bias. For 325 women who participated in both SFBCS and NC-BCFR around the same time and completed similar questionnaires on reproductive history, we found high reproducibility for age at first FTP and parity, with high proportions of cases who provided identical answers (89 and 95%, respectively) and no differences for younger and older women. However, reproducibility was somewhat lower for age at menarche (74%), which could have introduced some exposure misclassification. The extent of misclassification observed in our study is consistent with what was observed in another study of European women.4 Finally, even though the pooled dataset was large and diverse, with 68% of study participants from ethnic minority populations, analyses of HR− breast cancer, especially those stratified by age, were based on relatively small case numbers in each ethnicity group.
Despite these limitations, the present pooled analysis has several important strengths, including case ascertainment through population-based cancer registries and the inclusion of population controls which permitted the estimation of HR-specific relative risks associated with menstrual and reproductive characteristics, unlike some studies that were limited to case–case comparisons.27,52 With nearly 1,400 HR− cases, the present study is among those with the largest numbers of women diagnosed with this breast cancer subtype. The study population had substantial ethnic diversity and adequate sample size to conduct separate analyses for younger and older women.
To the best of our knowledge, this is the first large-scale investigation of HR-specific associations with menstrual and reproductive characteristics in multiple U.S. minority groups. Our analyses revealed consistent associations in Hispanic and Asian American women, and add to the evidence that pregnancies and breast-feeding may not protect against the development of HR+ breast cancer among African American women. We further found among African American women under age 50 years a suggestive positive association between parity and HR− breast cancer that may be mitigated by breast-feeding, as other studies have reported. The present findings emphasize the importance of studying breast cancer risk factors in diverse populations and warrant further work to understand why menstrual and reproductive risk factors vary in importance according to age and ethnicity.
Supplementary Material
Wha’s new?
To examine how breast cancer risk varies among ethnic groups, these authors pooled data from 4 population-based studies. They analyzed the relationship between menstrual and reproductive characteristics and breast cancer risk by hormone receptor status. In non-Hispanic whites, Hispanics, and Asian-Americans, associations were as expected. Characteristics such as later onset of menstruation, earlier menopause, higher parity, and longer breastfeeding were associated with lower risk of HR+ cancer in these groups. Among African Americans, however, most menstrual and reproductive characteristics showed no association with breast cancer risk, in contrast to a previous study. They did detect an increased risk for HR− cancer in African-American women with higher parity who never breast-fed.
Acknowledgements
The Breast Cancer Etiology in Minorities Study was funded by grant R03 CA199343 (to E.M.J.) from the National Cancer Institute. The San Francisco Bay Area Breast Cancer Study was supported by grants R01 CA063446 (to E.M.J.) and R01 CA077305 (to E.M.J.) from the National Cancer Institute, grant DAMD17-96-1-6071 (to E.M.J.) from the U.S. Department of Defense and grant 7PB-0068 (to E.M.J.) from the California Breast Cancer Research Program. The Northern California Breast Cancer Family Registry was funded by grant UM1 CA164920 (E.M.J.) from the National Cancer Institute. The content of this manuscript does not necessarily reflect the views or policies of the National Cancer Institute or any of the collaborating centers in the Breast Cancer Family Registry (BCFR), nor does mention of trade names, commercial products or organizations imply endorsement by the U.S. Government or the BCFR. The collection of cancer incidence data used in these studies was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program under contract HHSN261201000036C awarded to the Cancer Prevention Institute of California; and the Centers for Disease Control and Prevention’s National Program of Cancer Registries, under agreement #1U58 DP000807-01 awarded to the Public Health Institute. The Los Angeles County Asian American Breast Cancer Study was funded by the California Breast Cancer Research Program grants 1RB-0287, 3PB-0120, 5PB-0018 and 10PB0038 (to A.H.W.). The 4-Corners Breast Cancer Study was funded by grants R01 CA078682 (to M.L.S.), R01 CA078762 (to K.B.B.), R01 CA078552 (to T.B.) and R01 CA078802 (to A.R.G.) from the National Cancer Institute. The research also was supported by the Utah Cancer Registry, which is funded by contract N01-PC-67000 from the National Cancer Institute, with additional support from the State of Utah Department of Health, the New Mexico Tumor Registry and the Arizona and Colorado cancer registries, funded by the Centers for Disease Control and Prevention National Program of Cancer Registries and additional state support. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the National Cancer Institute or endorsement by the State of California Department of Public Health, the National Cancer Institute and the Centers for Disease Control and Prevention or their Contractors and Subcontractors.
Abbreviations:
- 4-CBCS
4-Corners Breast Cancer Study
- AABCS
Los Angeles County Asian American Breast Cancer Study
- AMBER
African American Breast Cancer Epidemiology and Risk
- BEM
Breast Cancer Etiology in Minorities
- BMI
body mass index
- CI
confidence interval
- ER
estrogen receptor
- FTP
full-term pregnancy
- HR
hormone receptor
- HT
hormone therapy
- NC-BCFR
Northern California Breast Cancer Family Registry
- NHW
non-Hispanic white
- OR
odds ratio
- PR
progesterone receptor
- SFBCS
San Francisco Bay Area Breast Cancer Study
- U.S.
United States
Footnotes
Conflict of interest
All authors declare they have no conflicts of interest.
Additional Supporting Information may be found in the online version of this article.
References
- 1.American Cancer Society. Breast Cancer Facts & Figures 2017–2018. New York, NY: American Cancer Society, 2017. [Google Scholar]
- 2.Kohler BA, Sherman RL, Howlader N, et al. Annual report to the nation on the status of Cancer, 1975–2011, featuring incidence of Breast Cancer subtypes by race/ethnicity, poverty, and state. J Natl Cancer Inst 2015;107:djv048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kelsey JL, Gammon MD, John EM. Reproductive factors and breast cancer. Epidemiol Rev 1993;15: 36–47. [DOI] [PubMed] [Google Scholar]
- 4.Clavel-Chapelon F, Gerber M. Reproductive factors and breast cancer risk. Do they differ according to age at diagnosis? Breast Cancer Res Treat 2002;72:107–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gaudet MM, Gierach GL, Carter BD, et al. Pooled analysis of nine cohorts reveals Breast Cancer risk factors by tumor molecular subtype. Cancer Res 2018;78:6011–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Tamimi RM, Colditz GA, Hazra A, et al. Traditional breast cancer risk factors in relation to molecular subtypes of breast cancer. Breast Cancer Res Treat 2012;131:159–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sisti JS, Collins LC, Beck AH, et al. Reproductive risk factors in relation to molecular subtypes of breast cancer: results from the nurses’ health studies. Int J Cancer 2016;138:2346–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sweeney C, Baumgartner KB, Byers T, et al. Reproductive history in relation to breast cancer risk among Hispanic and non-Hispanic white women. Cancer Causes Control 2008;19:391–401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wu AH, Vigen C, Lee E, et al. Traditional Breast Cancer risk factors in Filipina Americans compared with Chinese and Japanese Americans in Los Angeles County. Cancer Epidemiol Biomarkers Prev 2016;25:1572–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Palmer JR, Boggs DA, Wise LA, et al. Parity and lactation in relation to estrogen receptor negative breast cancer in African American women. Cancer Epidemiol Biomarkers Prev 2011;20:1883–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Warner ET, Tamimi RM, Boggs DA, et al. Estrogen receptor positive tumors: do reproductive factors explain differences in incidence between black and white women? Cancer Causes Control 2013;24:731–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Cui Y, Deming-Halverson SL, Shrubsole MJ, 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ambrosone CB, Zirpoli GR, Bovbjerg DH, et al. Associations between estrogen receptor-negative breast cancer and timing of reproductive events differ between African American and European American women. Cancer Epidemiol Biomarkers Prev 2014;23:1115–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Palmer JR, Viscidi E, Troester MA, et al. Parity, lactation, and breast cancer subtypes in African American women: results from the AMBER consortium. J Natl Cancer Inst 2014;106:dju237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ambrosone CB, Zirpoli G, Hong CC, et al. Important role of menarche in development of estrogen receptor-negative Breast Cancer in African American women. J Natl Cancer Inst 2015; 107:djv172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bertrand KA, Bethea TN, Adams-Campbell LL, et al. Differential patterns of risk factors for early-onset Breast Cancer by ER status in African American women. Cancer Epidemiol Biomarkers Prev 2017;26:270–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.John EM, Hines LM, Phipps AI, et al. Reproductive history, breast-feeding and risk of triple negative breast cancer: the Breast Cancer etiology in minorities (BEM) study. Int J Cancer 2018;142: 2273–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Millikan RC, Newman B, Tse CK, et al. Epidemiology of basal-like breast cancer. Breast Cancer Res Treat 2008;109:123–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Yang XR, Chang-Claude J, Goode EL, et al. Associations of breast cancer risk factors with tumor subtypes: a pooled analysis from the Breast Cancer association consortium studies. J Natl Cancer Inst 2011;103:250–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ma H, Bernstein L, Pike MC, et al. Reproductive factors and breast cancer risk according to joint estrogen and progesterone receptor status: a meta-analysis of epidemiological studies. Breast Cancer Res 2006;8:R43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ritte R, Tikk K, Lukanova A, et al. Reproductive factors and risk of hormone receptor positive and negative breast cancer: a cohort study. BMC Cancer 2013;13:584. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lambertini M, Santoro L, Del Mastro L, et al. Reproductive behaviors and risk of developing breast cancer according to tumor subtype: a systematic review and meta-analysis of epidemiological studies. Cancer Treat Rev 2016;49:65–76. [DOI] [PubMed] [Google Scholar]
- 23.Ellingjord-Dale M, Vos L, Tretli S, et al. Parity, hormones and breast cancer subtypes - results from a large nested case-control study in a national screening program. Breast Cancer Res 2017;19:10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Anderson WF, Pfeiffer RM, Wohlfahrt J, et al. Associations of parity-related reproductive histories with ER+/− and HER2+/− receptor-specific breast cancer aetiology. Int J Epidemiol 2017; 46:373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Collaborative Group on Hormonal Factors in Breast Cancer. Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol 2012; 13:1141–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Althuis MD, Fergenbaum JH, Garcia-Closas M, et al. Etiology of hormone receptor-defined breast cancer: a systematic review of the literature. Cancer Epidemiol Biomarkers Prev 2004;13: 1558–68. [PubMed] [Google Scholar]
- 27.Anderson KN, Schwab RB, Martinez ME. Reproductive risk factors and breast cancer subtypes: a review of the literature. Breast Cancer Res Treat 2014;144:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bao PP, Shu XO, Gao YT, et al. Association of hormone-related characteristics and breast cancer risk by estrogen receptor/progesterone receptor status in the shanghai breast cancer study. Am J Epidemiol 2011;174:661–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Romieu I, Biessy C, Carayol M, et al. Reproductive factors and molecular subtypes of breast cancer among premenopausal women in Latin America: the PRECAM A study. Sci Rep 2018;8: 13109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Setiawan VW, Monroe KR, Wilkens LR, et al. Breast cancer risk factors defined by estrogen and progesterone receptor status: the multiethnic cohort study. Am J Epidemiol 2009; 169:1251–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Colditz GA, Rosner BA, Chen WY, et al. Risk factors for breast cancer according to estrogen and progesterone receptor status. J Natl Cancer Inst 2004;96:218–28. [DOI] [PubMed] [Google Scholar]
- 32.Cl Li, Malone KE Daling JR, et al. Timing of menarche and first full-term birth in relation to breast cancer risk. Am J Epidemiol 2008; 167: 230–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Warner ET, Colditz GA, Palmer JR, et al. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Albrektsen G, Heuch I, Hansen S, et al. Breast cancer risk by age at birth, time since birth and time intervals between births: exploring interaction effects. Br J Cancer 2005;92: 167–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Nichols HB, Schoemaker MJ, Cai J, et al. Breast Cancer risk after recent childbirth: a pooled analysis of 15 prospective studies. Ann Intern Med 2018;170:22–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Fortner RT, Sisti J, Chai B, et al. Parity, breastfeeding, and breast cancer risk by hormone receptor status and molecular phenotype: results from the Nurses’ health studies. Breast Cancer Res 2019;21:40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Lord SJ, Bernstein L, Johnson KA, et al. Breast cancer risk and hormone receptor status in older women by parity, age of first birth, and breastfeeding: a case-control study. Cancer Epidemiol Biomarkers Prev 2008;17:1723–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Work ME, John EM, Andrulis IL, et al. Reproductive risk factors and oestrogen/progesterone receptor-negative breast cancer in the Breast Cancer family registry. Br J Cancer 2014;110:1367–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Chen L, Li CI. Racial disparities in breast cancer diagnosis and treatment by hormone receptor and HER2 status. Cancer Epidemiol Biomarkers Prev 2015;24:1666–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Howlader N, Cronin KA, Kurian AW, et al. Differences in Breast Cancer survival by molecular subtypes in the United States. Cancer Epidemiol Biomarkers Prev 2018;27:619–26. [DOI] [PubMed] [Google Scholar]
- 41.Ma H, Lu Y, Malone KE, et al. Mortality risk of black women and white women with invasive breast cancer by hormone receptors, HER2, and p53 status. BMC Cancer 2013; 13:225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Tao L, Gomez SL, Keegan TH, et al. Breast Cancer mortality in African-American and non-Hispanic white women by molecular subtype and stage at diagnosis: a population-based study. Cancer Epidemiol Biomarkers Prev 2015;24:1039–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Clavel-Chapelon F, Group ENE. Differential effects of reproductive factors on the risk of pre- and postmenopausal breast cancer. Results from a large cohort of French women. Br J Cancer 2002;86: 723–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Colditz GA, Bohlke K, Berkey CS. Breast cancer risk accumulation starts early: prevention must also. Breast Cancer Res Treat 2014;145:567–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Key TJ, Pike MC. The role of oestrogens and progestagens in the epidemiology and prevention of breast cancer. Eur J Cancer Clin Oncol 1988;24:29–43. [DOI] [PubMed] [Google Scholar]
- 46.Endogenous H, Breast Cancer Collaborative G, Key TJ, et al. Sex hormones and risk of breast cancer in premenopausal women: a collaborative reanalysis of individual participant data from seven prospective studies. Lancet Oncol 2013;14: 1009–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Key T, Appleby P, Barnes I, et al. Endogenous sex hormones and breast cancer in postmenopausal women: reanalysis of nine prospective studies. J Natl Cancer Inst 2002;94: 606–16. [DOI] [PubMed] [Google Scholar]
- 48.Russo J, Moral R, Balogh GA, et al. The protective role of pregnancy in breast cancer. Breast Cancer Res 2005;7:131–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Belitskaya-Levy I, Zeleniuch-Jacquotte A, Russo J, et al. Characterization of a genomic signature of pregnancy identified in the breast. Cancer Prev Res (Phila) 2011;4:1457–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Santucci-Pereira J, Zeleniuch-Jacquotte A, Afanasyeva Y, et al. Genomic signature of parity in the breast of premenopausal women. Breast Cancer Res 2019;21:46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Rotunno M, Sun X, Figueroa J, et al. Parity-related molecular signatures and breast cancer subtypes by estrogen receptor status. Breast Cancer Res 2014;16:R74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Brouckaert O, Rudolph A, Laenen A, et al. Reproductive profiles and risk of breast cancer subtypes: a multi-center case-only study. Breast Cancer Res 2017;19:119. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used for the current study are available upon reasonable request from the corresponding author (E.M.J.), contingent upon approval by appropriate Institutional Review Boards and study Principal Investigators.