Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 May 1.
Published in final edited form as: Cancer Epidemiol Biomarkers Prev. 2015 Feb 25;24(5):798–809. doi: 10.1158/1055-9965.EPI-14-1136

Dense and Non-dense Mammographic Area and Risk of Breast Cancer by Age and Tumor Characteristics

Kimberly A Bertrand 1,2, Christopher G Scott 3, Rulla M Tamimi 1,2, Matthew R Jensen 3, V Shane Pankratz 3,*, Aaron D Norman 5, Daniel W Visscher 4, Fergus J Couch 5,6, John Shepherd 7, Yunn-Yi Chen 8, Bo Fan 7, Fang-Fang Wu 3, Lin Ma 9, Andrew H Beck 10, Steven R Cummings 11, Karla Kerlikowske 12, Celine M Vachon 5
PMCID: PMC4417380  NIHMSID: NIHMS666538  PMID: 25716949

Abstract

Background

Mammographic density (MD) is a strong breast cancer risk factor. We previously reported associations of percent MD with larger and node-positive tumors across all ages, and estrogen receptor (ER)-negative status among women ages <55 years. To provide insight into these associations, we examined the components of percent MD (dense area (DA) and non-dense area (NDA) with breast cancer subtypes.

Methods

Data were pooled from six studies including 4095 breast cancers and 8558 controls. DA and NDA were assessed from digitized film-screen mammograms and standardized across studies. Breast cancer odds by density phenotypes and age according to histopathological characteristics and receptor status were calculated using polytomous logistic regression.

Results

DA was associated with increased breast cancer risk [odds ratios (OR) for quartiles: 0.65, 1.00(Ref), 1.22, 1.55; p-trend <0.001] and NDA was associated with decreased risk [ORs for quartiles: 1.39, 1.00(Ref), 0.88, 0.72; p-trend <0.001] across all ages and invasive tumor characteristics. There were significant trends in the magnitude of associations of both DA and NDA with breast cancer by increasing tumor size (p-trend<0.001) but no differences by nodal status. Among women <55 years, DA was more strongly associated with increased risk of ER+ vs. ER− tumors [p-heterogeneity (het) = 0.02] while NDA was more strongly associated with decreased risk of ER− vs. ER+ tumors [p-het = 0.03].

Conclusions

DA and NDA have differential associations with ER+ vs. ER− tumors that vary by age.

Impact

DA and NDA are important to consider when developing age- and subtype-specific risk models.

Keywords: Mammographic density, Breast density, Breast cancer, Tumor subtypes, Mammography, Epidemiology

Introduction

Mammographic density (MD) represents the variability of breast tissue composition on the mammogram image. Radiographically, there are two main components of breast tissue: fat, which appears dark on a mammogram and is considered “non-dense”, and fibroglandular tissue (i.e., epithelial cells and connective tissue), which appears white and is defined as “dense” tissue (1). Women in the highest quartile of percent MD (i.e., proportion of dense fibroglandular tissue within the total area of the breast) have about 4 times the risk of developing breast cancer compared to women in the lowest quartile, even after adjusting for other known breast cancer risk factors (2). The biological mechanism by which MD increases breast cancer risk, however, remains largely unknown.

We reported percent MD to be a breast cancer risk factor across tumor characteristics and age groups (3). We noted stronger associations for tumors of large size and positive lymph nodes across all ages, and ER-negative status among women ages <55 years, suggesting high MD may play an important role in tumor aggressiveness, especially in younger women. Recent evidence from a large meta-analysis suggests that dense and non-dense area may be independently associated with breast cancer risk (4–7). Few previous studies have evaluated the possible differential associations of dense and non-dense breast area by breast cancer subtype or tumor characteristics. Therefore, we investigated the underlying associations of dense (fibroglandular) or non-dense (adipose) area, or both, with tumor characteristics. Understanding these differential associations could provide insight into the mechanism by which percent density influences risk.

Materials and Methods

Study populations

Participating studies included the Mayo Mammography Health Study (MMHS) (8, 9), Mayo Clinic Breast Cancer Study (MCBCS) (10, 11), Nurses’ Health Study (NHS) and NHSII (12–14), Mayo Clinic Mammography Study (MCMAM) (15), and the San Francisco Bay Area Breast Cancer SPORE and San Francisco Mammography Registry (SFMR) (16–18) (Table 1). Details of the study populations are in Supplementary Table S1 and described in our earlier report (3); the present analysis includes additional cases and controls, primarily from the SFMR, which were not available at the time of our earlier analysis. Incident breast cancer cases were identified by self-report, linkage to clinic and/or statewide tumor registries, or death certificates with further confirmation by medical record review. Controls were selected from the underlying cohorts (MMHS, NHS, NHSII, SFMR) or from the source population (MCBCS, MCMAM) and typically matched to cases on age, menopausal status, and year of examination (MMHS, MCMAM, SFMR), blood draw (NHS, NHSII) or diagnosis (MCBCS) as described previously (3). From all studies, we excluded breast cancer cases diagnosed within 6 months of mammography and their matched controls, to minimize prevalent cancers at time of mammography. Covariate information was obtained from medical record review (MCMAM), self-administered questionnaires (NHS, NHSII), or both (MMHS, MCBCS) prior to (NHS, NHSII) or at the time of (MSBCS, MMHS, MCMAM, SFMR) mammography. In total, these analyses included 4095 breast cancer cases and 8558 controls.

Table 1.

Baseline characteristics of study population by age.

Age <55
Age ≥55
Cases Controls Cases Controls
N 1884 4072 2211 4486
Standardized % mammographic density Median (IQR) 40.7 (30.2) 32 (30.8) 25 (26.1) 19 (23.2)
Standardized Dense Area cm2 Median (IQR) 51.9 (44.4) 42.2 (41.8) 41.1 (43.0) 31.6 (37.3)
Standardized Nondense Area cm2 Median (IQR) 79.6 (92.4) 98 (100.7) 130.7 (123.7) 138.6 (119.9)
Mean age at mammogram (SD) 47.2 (4.6) 47.3 (4.5) 64.9 (7.4) 65.1 (7.4)
Mean age at diagnosis (SD) 51.6 (5.5) -- 69 (7.6) --
Mean BMI (SD) 24.2 (6.5) 25.2 (6) 25.6 (7.8) 25.9 (5.5)
Body mass index categories, kg/m2
 <25 1072 (56.9%) 2352 (57.8%) 909 (41.1%) 2213 (49.3%)
 25–29 507 (26.9%) 1007 (24.7%) 701 (31.7%) 1422 (31.7%)
 30–34 157 (8.3%) 399 (9.8%) 319 (14.4%) 553 (12.3%)
 35+ 85 (4.5%) 275 (6.8%) 177 (8%) 275 (6.1%)
 Unknown 63 (3.3%) 39 (1%) 105 (4.7%) 23 (0.5%)
Menopausal status
 Premenopausal 1159 (61.5%) 2629 (64.6%) 16 (0.7%) 44 (1%)
 Postmenopausal 556 (29.5%) 1220 (30%) 2191 (99.1%) 4434 (98.8%)
 Unknown 169 (9%) 223 (5.5%) 4 (0.2%) 8 (0.2%)
Parity
 Nulliparous 419 (22.2%) 901 (22.1%) 330 (14.9%) 624 (13.9%)
 Parous 1315 (69.8%) 3046 (74.8%) 1675 (75.8%) 3526 (78.6%)
 Unknown 150 (8%) 125 (3.1%) 206 (9.3%) 336 (7.5%)
Postmenopausal hormone therapya
 Not current user 193 (41.1%) 498 (45.8%) 923 (51.6%) 2403 (61.9%)
 Current, estrogen 90 (19.1%) 256 (23.5%) 272 (15.2%) 561 (14.5%)
 Current, estrogen + progestin 155 (33%) 290 (26.7%) 373 (20.9%) 574 (14.8%)
 Unknown 32 (6.8%) 44 (4%) 220 (12.3%) 343 (8.8%)
Family history
 No 1467 (77.9%) 3588 (88.1%) 1644 (74.4%) 3745 (83.5%)
 Yes 315 (16.7%) 466 (11.4%) 463 (20.9%) 706 (15.7%)
 Unknown 102 (5.4%) 18 (0.4%) 104 (4.7%) 35 (0.8%)
Open in a new tab
a

Among postmenopausal women in MMHS, NHS, NHSII, and UCSF.

IQR: interquartile range

This study was approved by the Institutional Review Boards at Mayo Clinic, Brigham and Women’s Hospital, the University of California, San Francisco (UCSF), and the Connecticut Department of Public Health Human Investigations Committee. Informed consent was obtained or implied by return of questionnaires (NHS, NHSII).

Assessment of mammographic density

As described previously (3), dense area (DA) and non-dense area (NDA), were measured using two computer-assisted threshold techniques (Cumulus (19) and UCSF custom mammographic density software) (20) from digitized images of pre-diagnostic film screening mammograms of the craniocaudal view. Percent MD was calculated as the proportion of absolute DA over total breast area (DA + NDA). With the exception of NHS and NHSII, for which average DA and NDA of both breasts was used, DA and NDA were estimated from the contralateral breast for cases and corresponding side for matched controls.

We standardized PMD, DA and NDA measurements made within each study to remove variability in measurements due to reader (1, 21), time of density assessment, and age distributions of different study populations for pooled analyses (Supplementary Figure S1). We have previously described and applied this method to percent MD using a logit transformation (3). For absolute dense and non-dense areas, an appropriate transformation was selected via the Box-Cox procedure. For DA, the square root transformation was selected while the 4th root was selected for non-dense area. Briefly the following procedure was implemented on each measure after appropriate transformation. First, we focused on women without breast cancer and estimated study-specific linear age trends in the medians of transformed MD (TMD) values using quantile regression. Study-specific age trends were removed by computing the difference between each individual’s observed TMD and the age-predicted median TMD from the corresponding study set. Variability was standardized across studies by dividing the residuals within each study by the corresponding inter-quartile range (IQR), and then multiplying these re-scaled residual values by the IQR of the original residuals from all studies. This ensured that the variability in standardized TMD was consistent across studies, and roughly equivalent to the observed variability in TMD. Finally, we estimated an overall age by TMD trend from the original data, and added the age-predicted median TMD to the rescaled residuals from each individual. This reincorporated the known age trend in MD into the standardized TMD measurements (Supplementary Figure S2). These TMD values were back-transformed to the original scale for use in analyses. Of note, variability in the tails of the smoother and limited data under age 40 (n=68 controls), resulted in an apparent difference in the distribution for DA for the NHS2 study (Supplementary Figure 2).

Assessment of tumor characteristics among cases

Information on tumor type, histology, grade, nodal involvement, tumor size, and ER, PR, and HER2 status was obtained from state-wide Surveillance Epidemiology and End Results programs (SFMR), pathology reports (NHS, NHSII), state and clinic cancer registries (MMHS, MCMAM, MCBCS), and medical records (MMHS, MCMAM, MCBCS). For 313 cases in NHS, 52 cases in NHS II, and 194 cases in MCMAM with missing receptor data on pathology reports, receptor status was obtained from immunohistochemical staining performed on paraffin sections of the tumor tissue microarray (TMA) according to a standard protocol (22). A proportion of cases (18%; N=624 cases ranging from 8% to 34% across studies) were still missing HER2 status after incorporating the TMA data and were excluded from HER2 analyses. Another 2% were cases with borderline HER2 results (2+ without available FISH) and not used in analyses.

Statistical analyses

We categorized the standardized DA and NDA measurements into quartiles based on the control distribution across studies. We fit polytomous (multinomial) logistic regression models to estimate odds ratios (OR) and 95% confidence intervals (CI) for the associations of DA and NDA with risk of breast cancer overall as well as with breast cancer defined by tumor type (invasive or DCIS), histologic type (ductal or lobular), grade (well-differentiated, moderately differentiated, or poorly differentiated), tumor size (<1.1 cm, 1.1–2.0 cm, or 2.1+ cm), involvement of lymph nodes (positive or negative), and receptor status (ER+ or ER−; PR+ or PR−; HER2+ or HER2−). We also updated our prior analyses of percent MD (3) with this larger sample size for comparison purposes; percent MD categories were 0–10%, 11–25% (reference), 26%–50%, and 51%+.

We pooled data across the six studies and adjusted for study site, age (continuous), and body mass index (BMI, continuous) in multivariable models. We further considered potential confounding by parity (nulliparous, parous, or unknown) and first-degree family history of breast cancer (yes, no, or unknown) by evaluating the magnitude of change in ORs observed after including each potential confounder individually in the model. Postmenopausal hormone therapy (current estrogen alone, current estrogen plus progesterone, never/former, or unknown) was also evaluated as a confounder among postmenopausal women in the subset of studies for which this information was available (MMHS, NHS, NHSII, and SFMR). Addition of these variables to the models did not substantially change risk estimates and were not included in final models. In secondary analyses, we considered models that mutually adjusted for continuous measures of square root DA and NDA.

Because our previous findings suggested differences of percent MD and tumor characteristics primarily for younger women, we stratified by ages <55 years vs. ≥55 years only. We evaluated whether the associations between DA or NDA and breast cancer differed by specific tumor characteristics, both overall and within age groups, using polytomous logistic regression models (Pheterogeneity). For subtypes with a natural ordering, including tumor size and grade, tests of trend (Ptrend) across categories were used to assess significance. Formal tests of interaction (Page-interaction) assessed the significance of differential DA and NDA associations with each of the breast cancer characteristics and subtype by age groups.

Prior to pooling data across the six studies, study specific estimates were obtained by fitting separate models for each study and assessing individual associations between MD and each tumor subtype. We assessed the statistical significance of differences in associations by study site by testing for interactions between study group and DA or NDA category in the pooled analysis and, in general, found no evidence of differences across study (P-values >0.09) other than as noted in results below.

Analyses were performed using SAS software (version 9.3, SAS Institute, Cary, NC). All statistical tests were two-sided and P-values <0.05 were considered statistically significant.

Results

Overall, mean age at mammogram was 57 years among both cases and controls. Median time to diagnosis at mammogram was 4.1 years (interquartile range: 2.3–6.0) for cases. DA and NDA were not strongly correlated in the combined study population (r=0.07 based on continuous measure) or across individual study populations (correlations ranged from 0.06 (NHS) to 0.29 (NHS2)). Among both cases and controls, median percent MD and DA was lower while NDA was higher in women ages ≥55 years than women <55 years. Further, within each age group, median percent MD and DA was higher among cases vs. controls. Median NDA was lower among cases vs. controls in women ages <55 years but similar in cases and controls ≥55 years (Table 1). DCIS was more common among women ages <55 years (15.8%) vs. women ≥55 years (11.7%), while, among invasive cancers, more aggressive tumor characteristics were evident in women <55 years at mammography compared to women ≥55 years (Table 2).

Table 2.

Distribution (%) of breast cancer cases from six studies by age and tumor characteristics

Age < 55 Age ≥55
N % N %
Controls 4072 68.4 4486 67
Cases 1884 31.6 2211 33
 Invasive 1579 83.8 1944 87.9
 In situ 297 15.8 259 11.7
 Unknown 8 0.4 8 0.4
Tumor characteristics
Histology
 Ductal 1277 80.9 1437 73.9
 Lobular 156 9.9 265 13.6
 Mixed 88 5.6 133 6.8
 Unknown/other 58 3.7 109 5.6
Histologic Grade
 Well differentiated 393 24.9 622 32
 Moderately differentiated 605 38.3 739 38
 Poorly differentiated 447 28.3 373 19.2
 Unknown 134 8.5 210 10.8
Tumor size
 0.1–1.0 cm 488 30.9 701 36.1
 1.1.–2.0 cm 633 40.1 744 38.3
 2.1+ cm 409 25.9 435 22.4
 Unknown 49 3.1 64 3.3
Involvement of lymph nodes
 Negative 1054 66.8 1323 68.1
 Positive 445 28.2 422 21.7
 Unknown 80 5.1 199 10.2
Estrogen Receptor status
 Negative 289 18.3 279 14.4
 Positive 1236 78.3 1581 81.3
 Borderline/Unknown 54 3.4 84 4.3
Progesterone Receptor status
 Negative 407 25.8 476 24.5
 Positive 1114 70.6 1383 71.1
 Borderline/Unknown 58 3.7 85 4.4
HER2 status
 Negative 1092 69.2 1268 65.2
 Positive 231 14.6 223 11.5
 Borderline/Unknown 256 16.2 453 23.3
Open in a new tab

In general, results of our updated analyses for percent MD were consistent with our previous report which included a large subset of these data (3) and are presented in Tables 3 and 4. However, our earlier report stratified age into three categories, instead of two as shown here (Table 4). Consistent with our earlier analyses, we found significant positive associations between percent MD and breast cancer risk. Briefly, with the addition of new cases (mostly invasive) and controls, we found similar or stronger associations than what we previously reported. In the updated analyses, we continue to observe stronger associations with increasing tumor size, positive nodal status, and lobular (vs. ductal) cancer)(Phet<0.02) across age groups (Table 3). Among women <55 years, there were stronger associations with node positive vs. node negative tumors (Table 4). Of note, the associations of percent MD with ER-negative vs. ER-positive tumors are not statistically significantly different across the two age groups examined here, <55 vs. 55+ (Page-interaction =0.12). However, when we analyzed by the original three age groups, the age-interaction remains (Page-interaction =0.048) suggesting it is partially driven by differential associations across the older age groups (Data not shown). Below, we focus on results for DA and NDA.

Table 3.

Associations of categoriesa of percent density, dense area, and non dense area with breast cancer overall and by morphological subtypes

PERCENT DENSITY DENSE AREA NON DENSE AREA
No. cases No. controls OR (95% CI)b No. cases No. controls OR (95% CI)b No. cases No. controls OR (95% CI)b
Overall breast cancer
  Category 1 517 1784 0.63 (0.56, 0.72) 618 2139 0.65 (0.58, 0.74) 1280 2139 1.39 (1.25, 1.55)
  Category 2 (REF) 1015 2597 1.00 (REF) 910 2140 1.00 (REF) 983 2140 1.00 (REF)
  Category 3 1626 2890 1.62 (1.47, 1.79) 1134 2139 1.22 (1.10, 1.37) 943 2139 0.88 (0.78, 0.98)
  Category 4 937 1287 2.34 (2.07, 2.65) 1433 2140 1.55 (1.40, 1.73) 889 2140 0.72 (0.63, 0.81)
Invasiveness
 In situ
  Category 1 51 1784 0.51 (0.37, 0.72) 73 2139 0.56 (0.42, 0.75) 184 2139 1.19 (0.94, 1.51)
  Category 2 (REF) 140 2597 1.00 (REF) 135 2140 1.00 (REF) 143 2140 1.00 (REF)
  Category 3 242 2890 1.58 (1.26, 1.98) 167 2139 1.18 (0.93, 1.50) 124 2139 0.85 (0.66, 1.10)
  Category 4 123 1287 1.88 (1.42, 2.49) 181 2140 1.29 (1.02, 1.63) 105 2140 0.75 (0.56, 1.00)
 Invasive
  Category 1 461 1784 0.64 (0.56, 0.73) 542 2139 0.67 (0.59, 0.76) 1089 2139 1.42 (1.27, 1.60)
  Category 2 (REF) 875 2597 1.00 (REF) 770 2140 1.00 (REF) 836 2140 1.00 (REF)
  Category 3 1377 2890 1.62 (1.45, 1.80) 961 2139 1.23 (1.10, 1.38) 818 2139 0.89 (0.79, 1.00)
  Category 4 810 1287 2.40 (2.11, 2.74) 1250 2140 1.61 (1.44, 1.80) 780 2140 0.71 (0.62, 0.81)
p-het 0.18 0.25 0.40
Histologyc
 Ductal
  Category 1 356 1784 0.67 (0.57, 0.77) 433 2139 0.72 (0.63, 0.83) 838 2139 1.38 (1.22, 1.56)
  Category 2 (REF) 665 2597 1.00 (REF) 581 2140 1.00 (REF) 650 2140 1.00 (REF)
  Category 3 1073 2890 1.62 (1.44, 1.82) 749 2139 1.27 (1.12, 1.44) 635 2139 0.90 (0.79, 1.03)
  Category 4 620 1287 2.32 (2.01, 2.67) 951 2140 1.61 (1.42, 1.82) 591 2140 0.71 (0.61, 0.82)
 Lobular
  Category 1 49 1784 0.51 (0.36, 0.72) 51 2139 0.49 (0.35, 0.70) 135 2139 1.59 (1.21, 2.10)
  Category 2 (REF) 114 2597 1.00 (REF) 96 2140 1.00 (REF) 98 2140 1.00 (REF)
  Category 3 148 2890 1.44 (1.11, 1.87) 107 2139 1.13 (0.85, 1.50) 95 2139 0.82 (0.61, 1.11)
  Category 4 110 1287 3.00 (2.23, 4.04) 167 2140 1.83 (1.41, 2.37) 93 2140 0.65 (0.47, 0.89)
p-het 0.02 0.03 0.50
Histologic grade
 Well differentiated
  Category 1 139 1784 0.62 (0.50, 0.77) 157 2139 0.70 (0.56, 0.87) 284 2139 1.25 (1.03, 1.51)
  Category 2 (REF) 274 2597 1.00 (REF) 215 2140 1.00 (REF) 247 2140 1.00 (REF)
  Category 3 379 2890 1.44 (1.22, 1.71) 283 2139 1.35 (1.11, 1.63) 246 2139 0.91 (0.75, 1.10)
  Category 4 223 1287 2.23 (1.81, 2.74) 360 2140 1.75 (1.45, 2.10) 238 2140 0.77 (0.62, 0.95)
 Moderately differentiated
  Category 1 177 1784 0.65 (0.53, 0.79) 202 2139 0.69 (0.57, 0.84) 398 2139 1.37 (1.16, 1.62)
  Category 2 (REF) 327 2597 1.00 (REF) 280 2140 1.00 (REF) 313 2140 1.00 (REF)
  Category 3 523 2890 1.64 (1.40, 1.91) 353 2139 1.25 (1.05, 1.48) 329 2139 0.96 (0.81, 1.14)
  Category 4 317 1287 2.53 (2.10, 3.04) 509 2140 1.81 (1.54, 2.13) 304 2140 0.74 (0.61, 0.90)
 Poorly differentiated
  Category 1 99 1784 0.65 (0.50, 0.84) 120 2139 0.62 (0.48, 0.79) 279 2139 1.55 (1.27, 1.90)
  Category 2 (REF) 184 2597 1.00 (REF) 188 2140 1.00 (REF) 195 2140 1.00 (REF)
  Category 3 335 2890 1.79 (1.47, 2.17) 221 2139 1.12 (0.91, 1.38) 170 2139 0.80 (0.64, 0.99)
  Category 4 202 1287 2.62 (2.08, 3.30) 291 2140 1.45 (1.19, 1.77) 176 2140 0.65 (0.51, 0.83)
p-het 0.68 0.77 0.76
Tumor size
 <1.1 cm
  Category 1 219 1784 0.90 (0.75, 1.09) 244 2139 0.90 (0.74, 1.08) 319 2139 1.19 (0.99, 1.42)
  Category 2 (REF) 318 2597 1.00 (REF) 262 2140 1.00 (REF) 282 2140 1.00 (REF)
  Category 3 429 2890 1.33 (1.13, 1.56) 314 2139 1.19 (1.00, 1.42) 280 2139 0.94 (0.78, 1.12)
  Category 4 223 1287 1.70 (1.39, 2.07) 369 2140 1.42 (1.20, 1.69) 308 2140 0.93 (0.77, 1.14)
 1.1 – 2.0 cm
  Category 1 149 1784 0.52 (0.43, 0.64) 188 2139 0.57 (0.47, 0.69) 463 2139 1.62 (1.37, 1.90)
  Category 2 (REF) 349 2597 1.00 (REF) 320 2140 1.00 (REF) 311 2140 1.00 (REF)
  Category 3 547 2890 1.60 (1.37, 1.86) 388 2139 1.20 (1.02, 1.41) 328 2139 0.96 (0.81, 1.14)
  Category 4 332 1287 2.45 (2.05, 2.93) 481 2140 1.50 (1.29, 1.76) 275 2140 0.68 (0.56, 0.82)
 2.1+ cm
  Category 1 74 1784 0.42 (0.32, 0.56) 91 2139 0.51 (0.39, 0.66) 273 2139 1.47 (1.21, 1.79)
  Category 2 (REF) 189 2597 1.00 (REF) 165 2140 1.00 (REF) 214 2140 1.00 (REF)
  Category 3 349 2890 2.02 (1.67, 2.45) 224 2139 1.32 (1.07, 1.64) 186 2139 0.74 (0.60, 0.92)
  Category 4 232 1287 3.60 (2.88, 4.51) 364 2140 2.13 (1.75, 2.60) 171 2140 0.49 (0.39, 0.63)
p-het <0.001 <0.001 <0.001
Involvement of lymph nodes
 Negative
  Category 1 314 1784 0.64 (0.55, 0.75) 370 2139 0.68 (0.59, 0.79) 722 2139 1.37 (1.20, 1.57)
  Category 2 (REF) 609 2597 1.00 (REF) 520 2140 1.00 (REF) 565 2140 1.00 (REF)
  Category 3 908 2890 1.50 (1.33, 1.69) 658 2139 1.24 (1.08, 1.41) 568 2139 0.93 (0.81, 1.07)
  Category 4 546 1287 2.25 (1.95, 2.61) 829 2140 1.58 (1.39, 1.79) 522 2140 0.73 (0.63, 0.85)
 Positive
  Category 1 89 1784 0.56 (0.43, 0.74) 117 2139 0.61 (0.48, 0.78) 300 2139 1.61 (1.32, 1.96)
  Category 2 (REF) 189 2597 1.00 (REF) 186 2140 1.00 (REF) 202 2140 1.00 (REF)
  Category 3 369 2890 1.96 (1.62, 2.37) 230 2139 1.18 (0.96, 1.45) 191 2139 0.86 (0.70, 1.07)
  Category 4 220 1287 2.89 (2.31, 3.61) 334 2140 1.70 (1.40, 2.07) 174 2140 0.63 (0.49, 0.80)
p-het 0.01 0.39 0.11
ER status
 Negative
  Category 1 69 1784 0.7 (0.51, 0.95) 88 2139 0.57 (0.43, 0.75) 209 2139 1.66 (1.31, 2.10)
  Category 2 (REF) 126 2597 1.00 (REF) 151 2140 1.00 (REF) 136 2140 1.00 (REF)
  Category 3 236 2890 1.81 (1.43, 2.27) 153 2139 0.96 (0.76, 1.21) 113 2139 0.76 (0.59, 0.99)
  Category 4 137 1287 2.49 (1.89, 3.26) 176 2140 1.11 (0.88, 1.40) 110 2140 0.61 (0.45, 0.81)
 Positive
  Category 1 373 1784 0.63 (0.54, 0.72) 431 2139 0.69 (0.60, 0.79) 837 2139 1.38 (1.22, 1.56)
  Category 2 (REF) 716 2597 1.00 (REF) 589 2140 1.00 (REF) 667 2140 1.00 (REF)
  Category 3 1082 2890 1.57 (1.40, 1.76) 764 2139 1.28 (1.13, 1.45) 664 2139 0.90 (0.79, 1.03)
  Category 4 646 1287 2.40 (2.09, 2.76) 1033 2140 1.74 (1.54, 1.97) 649 2140 0.73 (0.64, 0.85)
p-het 0.64 0.003 0.05
PR status
 Negative
  Category 1 119 1784 0.67 (0.53, 0.86) 137 2139 0.59 (0.47, 0.73) 277 2139 1.41 (1.16, 1.72)
  Category 2 (REF) 220 2597 1.00 (REF) 221 2140 1.00 (REF) 215 2140 1.00 (REF)
  Category 3 349 2890 1.61 (1.34, 1.93) 245 2139 1.08 (0.89, 1.31) 191 2139 0.81 (0.66, 1.01)
  Category 4 195 1287 2.23 (1.79, 2.78) 280 2140 1.24 (1.03, 1.50) 200 2140 0.72 (0.57, 0.90)
 Positive
  Category 1 324 1784 0.63 (0.54, 0.73) 381 2139 0.70 (0.60, 0.82) 762 2139 1.42 (1.24, 1.61)
  Category 2 (REF) 619 2597 1.00 (REF) 515 2140 1.00 (REF) 588 2140 1.00 (REF)
  Category 3 971 2890 1.61 (1.43, 1.82) 670 2139 1.28 (1.12, 1.47) 590 2139 0.91 (0.79, 1.04)
  Category 4 583 1287 2.47 (2.14, 2.86) 931 2140 1.79 (1.58, 2.04) 557 2140 0.71 (0.61, 0.83)
p-het 0.70 0.007 0.74
HER2 status
 Negative
  Category 1 325 1784 0.66 (0.57, 0.77) 355 2139 0.65 (0.56, 0.76) 686 2139 1.36 (1.19, 1.56)
  Category 2 (REF) 591 2597 1.00 (REF) 517 2140 1.00 (REF) 543 2140 1.00 (REF)
  Category 3 898 2890 1.56 (1.38, 1.76) 613 2139 1.17 (1.03, 1.34) 566 2139 0.96 (0.83, 1.10)
  Category 4 546 1287 2.40 (2.07, 2.78) 875 2140 1.68 (1.48, 1.91) 565 2140 0.81 (0.70, 0.95)
 Positive
  Category 1 52 1784 0.70 (0.49, 0.99) 63 2139 0.67 (0.48, 0.93) 154 2139 1.37 (1.06, 1.77)
  Category 2 (REF) 92 2597 1.00 (REF) 92 2140 1.00 (REF) 121 2140 1.00 (REF)
  Category 3 204 2890 2.16 (1.67, 2.80) 141 2139 1.46 (1.11, 1.92) 93 2139 0.72 (0.54, 0.95)
  Category 4 106 1287 2.67 (1.96, 3.64) 158 2140 1.64 (1.25, 2.14) 86 2140 0.54 (0.39, 0.75)
p-het 0.07 0.21 0.04
Open in a new tab
a

Categories are 1: 0–10%, 2: 11–25%, 3: 26–50%, and 4: 51%+ for PMD and quartiles for DA and NDA.

b

adjusted for study site, age, BMI.

c

mixed and other histology categories are excluded

p-het: test for heterogeneity in association by subtype

Table 4.

Pooled associations of categoriesa of percent density, dense area, and non dense area for morphological subtypes of invasive breast cancer by age

PERCENT DENSITY DENSE AREA NONDENSE AREA
Age < 55 Age ≥55 Age < 55 Age ≥55 Age < 55 Age ≥55
No. cases No.
controls		 OR (95% CI)b No. cases No.
controls		 OR (95% CI)b No. cases No.
controls		 OR (95% CI)b No. cases No.
controls		 OR (95% CI)b No. cases No.
controls		 OR (95% CI)b No.
cases		 No.
controls		 OR (95% CI)b
Overall breast cancer
  Category 1 109 530 0.60 (0.47, 0.77) 408 1254 0.61 (0.53, 0.71) 191 759 0.65 (0.53, 0.80) 427 1380 0.64 (0.55, 0.75) 822 1357 1.44 (1.24, 1.66) 458 782 1.33 (1.13, 1.56)
  Category 2 (REF) 316 1016 1.00 (REF) 699 1581 1.00 (REF) 376 948 1.00 (REF) 534 1192 1.00 (REF) 456 1087 1.00 (REF) 527 1053 1.00 (REF)
  Category 3 809 1604 1.68 (1.44, 1.97) 817 1286 1.60 (1.41, 1.83) 553 1135 1.22 (1.04, 1.43) 581 1004 1.26 (1.09, 1.46) 362 878 0.96 (0.81, 1.14) 581 1261 0.82 (0.71, 0.96)
  Category 4 650 922 2.40 (2.02, 2.86) 287 365 2.16 (1.79, 2.61) 764 1230 1.56 (1.34, 1.81) 669 910 1.58 (1.37, 1.83) 244 750 0.70 (0.57, 0.86) 645 1390 0.67 (0.57, 0.79)
Invasiveness
 In situ
  Category 1 15 530 0.64 (0.35, 1.18) 36 1254 0.43 (0.29, 0.64) 33 759 0.79 (0.50, 1.24) 40 1380 0.42 (0.28, 0.62) 129 1357 1.14 (0.84, 1.54) 55 782 1.21 (0.82, 1.78)
  Category 2 (REF) 44 1016 1.00 (REF) 96 1581 1.00 (REF) 56 948 1.00 (REF) 79 1192 1.00 (REF) 81 1087 1.00 (REF) 62 1053 1.00 (REF)
  Category 3 138 1604 1.96 (1.37, 2.81) 104 1286 1.40 (1.04, 1.89) 91 1135 1.33 (0.94, 1.88) 76 1004 1.10 (0.79, 1.54) 58 878 0.86 (0.60, 1.23) 66 1261 0.86 (0.60, 1.24)
  Category 4 100 922 2.38 (1.60, 3.54) 23 365 1.31 (0.80, 2.13) 117 1230 1.60 (1.15, 2.24) 64 910 1.06 (0.75, 1.49) 29 750 0.57 (0.35, 0.91) 76 1390 0.81 (0.55, 1.19)
 Invasive
  Category 1 92 530 0.58 (0.45, 0.76) 369 1254 0.63 (0.54, 0.74) 158 759 0.63 (0.51, 0.79) 384 1380 0.68 (0.58, 0.80) 688 1357 1.49 (1.28, 1.74) 401 782 1.35 (1.14, 1.59)
  Category 2 (REF) 272 1016 1.00 (REF) 603 1581 1.00 (REF) 317 948 1.00 (REF) 453 1192 1.00 (REF) 374 1087 1.00 (REF) 462 1053 1.00 (REF)
  Category 3 667 1604 1.62 (1.37, 1.92) 710 1286 1.63 (1.42, 1.87) 458 1135 1.19 (1.01, 1.42) 503 1004 1.29 (1.10, 1.51) 304 878 0.99 (0.82, 1.19) 514 1261 0.82 (0.70, 0.96)
  Category 4 548 922 2.39 (1.99, 2.88) 262 365 2.27 (1.87, 2.76) 646 1230 1.56 (1.33, 1.83) 604 910 1.68 (1.44, 1.95) 213 750 0.72 (0.58, 0.90) 567 1390 0.65 (0.55, 0.77)
p-het 0.55 0.08 0.78 0.03 0.40 0.58
Histologyc
 Ductal
  Category 1 82 530 0.65 (0.49, 0.86) 274 1254 0.65 (0.54, 0.77) 135 759 0.66 (0.52, 0.83) 298 1380 0.75 (0.63, 0.90) 546 1357 1.47 (1.24, 1.73) 292 782 1.29 (1.07, 1.55)
  Category 2 (REF) 216 1016 1.00 (REF) 449 1581 1.00 (REF) 260 948 1.00 (REF) 321 1192 1.00 (REF) 302 1087 1.00 (REF) 348 1053 1.00 (REF)
  Category 3 544 1604 1.67 (1.39, 2.00) 529 1286 1.61 (1.38, 1.88) 374 1135 1.19 (0.99, 1.42) 375 1004 1.35 (1.14, 1.61) 253 878 1.02 (0.84, 1.24) 382 1261 0.83 (0.69, 0.98)
  Category 4 435 922 2.41 (1.97, 2.95) 185 365 2.12 (1.71, 2.63) 508 1230 1.50 (1.26, 1.78) 443 910 1.75 (1.47, 2.07) 176 750 0.74 (0.58, 0.93) 415 1390 0.65 (0.53, 0.78)
 Lobular
  Category 1 5 530 0.29 (0.11, 0.78) 44 1254 0.53 (0.36, 0.78) 12 759 0.59 (0.29, 1.18) 39 1380 0.45 (0.30, 0.68) 75 1357 1.55 (1.03, 2.33) 60 782 1.55 (1.06, 2.26)
  Category 2 (REF) 27 1016 1.00 (REF) 87 1581 1.00 (REF) 26 948 1.00 (REF) 70 1192 1.00 (REF) 39 1087 1.00 (REF) 59 1053 1.00 (REF)
  Category 3 57 1604 1.43 (0.89, 2.30) 91 1286 1.47 (1.08, 2.01) 42 1135 1.31 (0.80, 2.17) 65 1004 1.08 (0.76, 1.54) 22 878 0.65 (0.38, 1.12) 73 1261 0.90 (0.62, 1.29)
  Category 4 67 922 3.14 (1.91, 5.17) 43 365 2.73 (1.83, 4.08) 76 1230 2.22 (1.41, 3.51) 91 910 1.68 (1.21, 2.32) 20 750 0.58 (0.32, 1.08) 73 1390 0.64 (0.43, 0.95)
p-het 0.07 0.27 0.18 0.08 0.31 0.79
Histologic grade
 Well differentiated
  Category 1 19 530 0.48 (0.29, 0.82) 120 1254 0.62 (0.49, 0.80) 39 759 0.72 (0.48, 1.07) 118 1380 0.67 (0.52, 0.87) 170 1357 1.31 (1.00, 1.71) 114 782 1.16 (0.89, 1.52)
  Category 2 (REF) 72 1016 1.00 (REF) 202 1581 1.00 (REF) 72 948 1.00 (REF) 143 1192 1.00 (REF) 103 1087 1.00 (REF) 144 1053 1.00 (REF)
  Category 3 159 1604 1.41 (1.05, 1.90) 220 1286 1.48 (1.20, 1.83) 111 1135 1.29 (0.95, 1.76) 172 1004 1.41 (1.11, 1.80) 75 878 0.91 (0.66, 1.25) 171 1261 0.90 (0.71, 1.15)
  Category 4 143 922 2.31 (1.67, 3.18) 80 365 2.02 (1.51, 2.72) 171 1230 1.86 (1.39, 2.48) 189 910 1.70 (1.34, 2.16) 45 750 0.62 (0.41, 0.92) 193 1390 0.78 (0.60, 1.01)
 Moderately differentiated
  Category 1 44 530 0.77 (0.53, 1.13) 133 1254 0.59 (0.47, 0.75) 65 759 0.74 (0.54, 1.03) 137 1380 0.65 (0.51, 0.83) 254 1357 1.50 (1.19, 1.89) 144 782 1.26 (0.98, 1.61)
  Category 2 (REF) 97 1016 1.00 (REF) 230 1581 1.00 (REF) 111 948 1.00 (REF) 169 1192 1.00 (REF) 137 1087 1.00 (REF) 176 1053 1.00 (REF)
  Category 3 254 1604 1.74 (1.35, 2.25) 269 1286 1.61 (1.32, 1.97) 167 1135 1.25 (0.97, 1.62) 186 1004 1.27 (1.01, 1.60) 126 878 1.13 (0.87, 1.47) 203 1261 0.86 (0.68, 1.07)
  Category 4 210 922 2.64 (2.00, 3.48) 107 365 2.43 (1.85, 3.18) 262 1230 1.81 (1.43, 2.30) 247 910 1.85 (1.49, 2.30) 88 750 0.82 (0.60, 1.12) 216 1390 0.65 (0.51, 0.83)
 Poorly differentiated
  Category 1 23 530 0.53 (0.32, 0.86) 76 1254 0.65 (0.48, 0.89) 39 759 0.52 (0.35, 0.77) 81 1380 0.67 (0.49, 0.91) 197 1357 1.47 (1.14, 1.90) 82 782 1.63 (1.18, 2.25)
  Category 2 (REF) 71 1016 1.00 (REF) 113 1581 1.00 (REF) 95 948 1.00 (REF) 93 1192 1.00 (REF) 109 1087 1.00 (REF) 86 1053 1.00 (REF)
  Category 3 202 1604 1.87 (1.40, 2.50) 133 1286 1.69 (1.29, 2.22) 142 1135 1.24 (0.94, 1.64) 79 1004 0.97 (0.71, 1.33) 76 878 0.85 (0.62, 1.16) 94 1261 0.77 (0.56, 1.05)
  Category 4 151 922 2.53 (1.84, 3.49) 51 365 2.49 (1.73, 3.59) 171 1230 1.36 (1.04, 1.78) 120 910 1.59 (1.19, 2.12) 65 750 0.73 (0.51, 1.05) 111 1390 0.56 (0.40, 0.79)
p-het 0.48 0.91 0.36 0.40 0.70 0.15
Tumor size
 <1.1 cm
  Category 1 46 530 0.94 (0.65, 1.37) 173 1254 0.85 (0.68, 1.06) 70 759 0.94 (0.68, 1.30) 174 1380 0.86 (0.68, 1.08) 195 1357 1.28 (1.00, 1.65) 124 782 1.09 (0.85, 1.41)
  Category 2 (REF) 96 1016 1.00 (REF) 222 1581 1.00 (REF) 98 948 1.00 (REF) 164 1192 1.00 (REF) 114 1087 1.00 (REF) 168 1053 1.00 (REF)
  Category 3 196 1604 1.25 (0.96, 1.63) 233 1286 1.40 (1.14, 1.72) 134 1135 1.13 (0.86, 1.49) 180 1004 1.27 (1.01, 1.60) 97 878 1.12 (0.84, 1.50) 183 1261 0.84 (0.67, 1.06)
  Category 4 150 922 1.63 (1.21, 2.17) 73 365 1.62 (1.20, 2.18) 186 1230 1.47 (1.14, 1.91) 183 910 1.41 (1.12, 1.78) 82 750 1.14 (0.82, 1.59) 226 1390 0.79 (0.62, 1.01)
 1.1 – 2.0 cm
  Category 1 32 530 0.50 (0.33, 0.76) 117 1254 0.51 (0.40, 0.65) 57 759 0.53 (0.39, 0.74) 131 1380 0.58 (0.45, 0.74) 288 1357 1.68 (1.34, 2.10) 175 782 1.56 (1.23, 1.98)
  Category 2 (REF) 110 1016 1.00 (REF) 239 1581 1.00 (REF) 136 948 1.00 (REF) 184 1192 1.00 (REF) 139 1087 1.00 (REF) 172 1053 1.00 (REF)
  Category 3 271 1604 1.63 (1.28, 2.07) 276 1286 1.59 (1.31, 1.93) 189 1135 1.15 (0.91, 1.46) 199 1004 1.26 (1.01, 1.57) 128 878 1.12 (0.86, 1.45) 200 1261 0.86 (0.69, 1.08)
  Category 4 220 922 2.39 (1.83, 3.11) 112 365 2.43 (1.87, 3.17) 251 1230 1.42 (1.13, 1.78) 230 910 1.59 (1.28, 1.97) 78 750 0.71 (0.52, 0.99) 197 1390 0.61 (0.47, 0.78)
 2.1+ cm
  Category 1 12 530 0.28 (0.15, 0.54) 62 1254 0.44 (0.32, 0.61) 28 759 0.49 (0.31, 0.77) 63 1380 0.51 (0.36, 0.71) 183 1357 1.47 (1.13, 1.91) 90 782 1.42 (1.05, 1.92)
  Category 2 (REF) 60 1016 1.00 (REF) 129 1581 1.00 (REF) 71 948 1.00 (REF) 94 1192 1.00 (REF) 107 1087 1.00 (REF) 107 1053 1.00 (REF)
  Category 3 176 1604 2.10 (1.53, 2.88) 173 1286 2.00 (1.56, 2.56) 120 1135 1.40 (1.03, 1.91) 104 1004 1.28 (0.95, 1.71) 70 878 0.74 (0.54, 1.03) 116 1261 0.75 (0.56, 1.00)
  Category 4 161 922 3.77 (2.69, 5.28) 71 365 3.24 (2.33, 4.51) 190 1230 2.02 (1.51, 2.70) 174 910 2.30 (1.76, 3.01) 49 750 0.45 (0.30, 0.67) 122 1390 0.48 (0.35, 0.66)
p-het <0.001 <0.001 0.002 <0.001 0.001 0.006
Involvement of lymph nodes
 Negative
  Category 1 64 530 0.57 (0.42, 0.77) 250 1254 0.65 (0.54, 0.78) 110 759 0.66 (0.52, 0.86) 260 1380 0.68 (0.56, 0.82) 458 1357 1.53 (1.28, 1.83) 264 782 1.24 (1.02, 1.50)
  Category 2 (REF) 198 1016 1.00 (REF) 411 1581 1.00 (REF) 211 948 1.00 (REF) 309 1192 1.00 (REF) 240 1087 1.00 (REF) 325 1053 1.00 (REF)
  Category 3 427 1604 1.41 (1.17, 1.71) 481 1286 1.58 (1.35, 1.85) 309 1135 1.21 (0.99, 1.47) 349 1004 1.29 (1.08, 1.54) 210 878 1.08 (0.87, 1.33) 358 1261 0.84 (0.70, 1.00)
  Category 4 365 922 2.16 (1.75, 2.66) 181 365 2.23 (1.79, 2.78) 424 1230 1.54 (1.28, 1.86) 405 910 1.64 (1.38, 1.95) 146 750 0.81 (0.63, 1.04) 376 1390 0.64 (0.53, 0.78)
 Positive
  Category 1 23 530 0.62 (0.37, 1.02) 66 1254 0.51 (0.38, 0.70) 38 759 0.53 (0.35, 0.78) 79 1380 0.66 (0.48, 0.91) 195 1357 1.47 (1.13, 1.90) 105 782 1.81 (1.34, 2.45)
  Category 2 (REF) 59 1016 1.00 (REF) 130 1581 1.00 (REF) 91 948 1.00 (REF) 95 1192 1.00 (REF) 109 1087 1.00 (REF) 93 1053 1.00 (REF)
  Category 3 205 1604 2.36 (1.73, 3.22) 164 1286 1.75 (1.36, 2.24) 123 1135 1.11 (0.83, 1.48) 107 1004 1.28 (0.96, 1.72) 84 878 0.91 (0.67, 1.24) 107 1261 0.85 (0.63, 1.14)
  Category 4 158 922 3.40 (2.43, 4.76) 62 365 2.48 (1.77, 3.48) 193 1230 1.60 (1.22, 2.09) 141 910 1.83 (1.39, 2.42) 57 750 0.60 (0.41, 0.87) 117 1390 0.61 (0.44, 0.84)
p-het 0.02 0.28 0.56 0.78 0.54 0.06
ER status
 Negative
  Category 1 14 530 0.53 (0.29, 1.00) 55 1254 0.73 (0.51, 1.04) 28 759 0.47 (0.30, 0.74) 60 1380 0.65 (0.45, 0.92) 139 1357 1.72 (1.26, 2.34) 70 782 1.64 (1.15, 2.35)
  Category 2 (REF) 42 1016 1.00 (REF) 84 1581 1.00 (REF) 74 948 1.00 (REF) 77 1192 1.00 (REF) 70 1087 1.00 (REF) 66 1053 1.00 (REF)
  Category 3 130 1604 2.14 (1.48, 3.09) 106 1286 1.63 (1.21, 2.21) 88 1135 0.97 (0.70, 1.35) 65 1004 0.95 (0.67, 1.34) 47 878 0.76 (0.51, 1.12) 66 1261 0.79 (0.55, 1.13)
  Category 4 103 922 3.16 (2.13, 4.70) 34 365 1.86 (1.21, 2.86) 99 1230 1.01 (0.74, 1.39) 77 910 1.24 (0.89, 1.73) 33 750 0.48 (0.30, 0.77) 77 1390 0.67 (0.46, 0.99)
 Positive
  Category 1 76 530 0.60 (0.45, 0.80) 297 1254 0.61 (0.51, 0.72) 125 759 0.68 (0.54, 0.87) 306 1380 0.68 (0.57, 0.82) 527 1357 1.46 (1.23, 1.74) 310 782 1.27 (1.06, 1.53)
  Category 2 (REF) 221 1016 1.00 (REF) 495 1581 1.00 (REF) 234 948 1.00 (REF) 355 1192 1.00 (REF) 287 1087 1.00 (REF) 380 1053 1.00 (REF)
  Category 3 509 1604 1.51 (1.26, 1.82) 573 1286 1.62 (1.40, 1.88) 348 1135 1.23 (1.02, 1.48) 416 1004 1.36 (1.15, 1.61) 247 878 1.06 (0.87, 1.30) 417 1261 0.81 (0.68, 0.96)
  Category 4 430 922 2.30 (1.88, 2.81) 216 365 2.32 (1.89, 2.86) 529 1230 1.73 (1.45, 2.07) 504 910 1.79 (1.52, 2.11) 175 750 0.80 (0.63, 1.01) 474 1390 0.65 (0.54, 0.77)
p-het 0.22 0.47 0.02 0.09 0.02 0.48
PR status
 Negative
  Category 1 23 530 0.55 (0.33, 0.89) 96 1254 0.69 (0.53, 0.92) 39 759 0.51 (0.34, 0.75) 98 1380 0.63 (0.47, 0.83) 181 1357 1.57 (1.20, 2.06) 96 782 1.25 (0.94, 1.68)
  Category 2 (REF) 72 1016 1.00 (REF) 148 1581 1.00 (REF) 96 948 1.00 (REF) 125 1192 1.00 (REF) 96 1087 1.00 (REF) 119 1053 1.00 (REF)
  Category 3 180 1604 1.70 (1.27, 2.28) 169 1286 1.54 (1.21, 1.96) 131 1135 1.13 (0.86, 1.50) 114 1004 1.03 (0.79, 1.36) 74 878 0.91 (0.66, 1.26) 117 1261 0.76 (0.58, 1.00)
  Category 4 132 922 2.29 (1.66, 3.17) 63 365 2.13 (1.53, 2.96) 141 1230 1.13 (0.86, 1.49) 139 910 1.38 (1.06, 1.79) 56 750 0.69 (0.47, 1.01) 144 1390 0.69 (0.51, 0.92)
 Positive
  Category 1 67 530 0.61 (0.45, 0.83) 257 1254 0.61 (0.51, 0.73) 112 759 0.69 (0.53, 0.88) 269 1380 0.70 (0.58, 0.85) 479 1357 1.46 (1.23, 1.75) 283 782 1.35 (1.11, 1.63)
  Category 2 (REF) 188 1016 1.00 (REF) 431 1581 1.00 (REF) 209 948 1.00 (REF) 306 1192 1.00 (REF) 261 1087 1.00 (REF) 327 1053 1.00 (REF)
  Category 3 462 1604 1.61 (1.33, 1.96) 509 1286 1.64 (1.41, 1.92) 303 1135 1.19 (0.98, 1.46) 367 1004 1.39 (1.17, 1.66) 222 878 1.05 (0.85, 1.29) 368 1261 0.82 (0.69, 0.98)
  Category 4 397 922 2.49 (2.02, 3.08) 186 365 2.28 (1.83, 2.83) 490 1230 1.79 (1.49, 2.15) 441 910 1.82 (1.53, 2.16) 152 750 0.75 (0.59, 0.96) 405 1390 0.63 (0.52, 0.77)
p-het 0.73 0.64 0.007 0.15 0.68 0.73
HER2 status
 Negative
  Category 1 69 530 0.63 (0.46, 0.85) 256 1254 0.64 (0.53, 0.77) 106 759 0.61 (0.47, 0.79) 249 1380 0.66 (0.55, 0.80) 451 1357 1.39 (1.17, 1.67) 235 782 1.27 (1.03, 1.55)
  Category 2 (REF) 186 1016 1.00 (REF) 405 1581 1.00 (REF) 221 948 1.00 (REF) 296 1192 1.00 (REF) 262 1087 1.00 (REF) 281 1053 1.00 (REF)
  Category 3 455 1604 1.64 (1.35, 1.99) 443 1286 1.53 (1.30, 1.80) 294 1135 1.10 (0.91, 1.34) 319 1004 1.26 (1.05, 1.51) 219 878 1.03 (0.84, 1.26) 347 1261 0.91 (0.75, 1.09)
  Category 4 382 922 2.52 (2.04, 3.13) 164 365 2.12 (1.69, 2.65) 471 1230 1.63 (1.35, 1.95) 404 910 1.75 (1.47, 2.09) 160 750 0.79 (0.62, 1.01) 405 1390 0.77 (0.63, 0.93)
 Positive
  Category 1 10 530 0.54 (0.26, 1.12) 42 1254 0.71 (0.47, 1.07) 20 759 0.68 (0.39, 1.19) 43 1380 0.63 (0.42, 0.96) 106 1357 1.48 (1.06, 2.08) 48 782 1.24 (0.83, 1.84)
  Category 2 (REF) 31 1016 1.00 (REF) 61 1581 1.00 (REF) 38 948 1.00 (REF) 54 1192 1.00 (REF) 59 1087 1.00 (REF) 62 1053 1.00 (REF)
  Category 3 115 1604 2.43 (1.60, 3.67) 89 1286 2.00 (1.42, 2.82) 82 1135 1.79 (1.21, 2.66) 59 1004 1.24 (0.84, 1.81) 40 878 0.83 (0.54, 1.26) 53 1261 0.64 (0.44, 0.94)
  Category 4 75 922 2.82 (1.78, 4.45) 31 365 2.51 (1.58, 3.99) 91 1230 1.82 (1.23, 2.69) 67 910 1.55 (1.07, 2.24) 26 750 0.54 (0.32, 0.91) 60 1390 0.50 (0.33, 0.75)
p-het 0.11 0.54 0.07 0.92 0.38 0.16
Open in a new tab
a

Categories are 1: 0–10%, 2: 11–25%, 3: 26–50%, and 4: 51%+ for PMD and quartiles for DA and NDA.

b

adjusted for study site, age, BMI.

c

mixed and other histology categories are excluded

p-het: test for heterogeneity in association by subtype

Overall and invasive breast cancer and DCIS

Overall, DA was significantly positively associated with breast cancer risk while NDA was significantly inversely associated with breast cancer risk (Table 3) and across age groups (Table 4). Specifically, the ORs for overall breast cancer associated with DA were: Q1 vs. Q2, 0.65; Q3 vs. Q2, 1.22; Q4 vs. Q2, 1.55 (p-trend <0.001) and the ORs for overall breast cancer associated with NDA were: Q1 vs. Q2, 1.39; Q3 vs. Q2, 0.88; Q4 vs. Q2, 0.72 (p-trend <0.001). For DA, associations were similar by age; for NDA, however, the interaction with age was statistically significant (Page-interaction <0.01) although the differences in associations by age were not clinically meaningful: <55 years (OR for Q1 vs. Q2, 1.44; 95% CI, 1.24–1.66) compared to those ≥55 years (corresponding OR, 1.33; 95% CI: 1.13–1.56) (Table 4).

DA was significantly positively associated with both invasive breast cancer and DCIS across all age groups (Tables 3 and 4; Figure 1). Among women ≥55 years, this association was stronger for invasive tumors than DCIS (Pheterogeneity = 0.03; Table 4; Figure 1); however, there was no evidence of a significant interaction between age and DA for associations with tumor type (Page-interaction = 0.41). Again, even though a statistically significant association was seen by age (Page-interaction =0.02), NDA was significantly inversely associated with risk of both invasive breast cancer and DCIS among both younger and older women (Tables 3 and 4; Figure 2).

Figure 1. Associations of categorical dense area (DA) for breast cancer tumor type and selected tumor characteristics of invasive breast cancer, by age.

Figure 1

Open in a new tab

dds ratios and 95% confidence intervals, adjusted for age, body mass index, and study, are shown for quartiles of DA. A) Tumor type, B) Tumor size, C) ER status, D) PR status.

Figure 2. Associations of categorical non dense area (NDA) for breast cancer tumor type and selected tumor characteristics of invasive breast cancer, by age.

Figure 2

Open in a new tab

Odds ratios and 95% confidence intervals, adjusted for age, body mass index, and study, are shown for quartiles of NDA. A) Tumor type, B) Tumor size, C) ER status, D) PR status.

Grade, invasive histology, size and nodal status

DA was significantly positively associated with all invasive tumor characteristics evaluated while NDA was significantly inversely associated with these characteristics (Tables 3 and 4). While there were no differences in the magnitude of associations of DA or NDA with tumor histology, grade, or nodal involvement, we did observe heterogeneity of associations with tumor size. Specifically, DA was positively associated with invasive tumors of all sizes; however, stronger positive associations of DA and breast cancer were noted for larger tumors ≥2.1 cm compared to smaller tumors across all ages (Ptrend <0.01) (Figure 1). For example, the overall ORs comparing women in Q4 of DA vs. Q2 were 1.42, 1.50, and 2.13 for tumors <1.1 cm, 1.1–2.0 cm, ≥2.1 cm, respectively (Table 3) and findings were similar among ages <55 and ≥55 years (Table 4; Figure 1; Page-interaction =0.91). The opposite trend was observed for associations of NDA with tumor size, with a stronger inverse association noted for larger tumors compared to smaller tumors across age groups (Ptrend <0.01), with the strongest associations most apparent for tumors 1.1–2.0 cm and 2.1+ cm in women ages <55 and ≥55 years (Table 4; Figure 2). This trend was also similar across age (Page-interaction =0.30).

ER, PR and HER2 receptor status

Among women of all ages, stronger associations of DA were noted for ER+ and PR+ tumors compared to hormone receptor negative tumors (Pheterogeneity <0.01). Although there was no significant evidence of differences by age (Page-interaction >0.38), among women <55 years, stronger associations were observed for ER+ (OR for Q4 vs. Q2, 1.73; 95% CI, 1.45–2.07) vs. ER− (corresponding OR, 1.01; 95% CI, 0.74–1.39) (Pheterogeneity =0.02; Table 4; Figure 1) and PR+ (OR for Q4 vs. Q2, 1.79; 95% CI, 1.49–2.15) vs. PR− (corresponding OR, 1.13; 95% CI, 0.86, 1.49) (Pheterogeneity =0.01; Table 4; Figure 1). Similarly, although not significantly different by age group (Page-interaction = 0.08), among women <55 years, NDA was more strongly inversely associated with ER− tumors (OR for Q4 vs. Q2, 0.48; 95% CI, 0.30, 0.77) than with ER+ tumors (corresponding OR, 0.8; 95% CI, 0.63–1.01) (Pheterogeneity =0.03; Table 4; Figure 2). In contrast, among women ages ≥55, DA and NDA were similarly associated with tumors defined by ER or PR status (Pheterogeneity >0.08; Table 4; Figure 2). Finally, DA and NDA were similarly associated with tumors defined by HER2 status (Tables 3 and 4).

Results were not materially changed in models that included mutual adjustment for DA and NDA (Data not shown). Finally, there was little evidence of differences across study (majority of P-values >0.09). Between study heterogeneity was noted, however, for associations of DA with overall breast cancer (P=0.02) and tumor histology (P=0.04), suggesting caution when interpreting these results.

Discussion

In this large study, the positive associations between percent MD and breast cancer overall and by tumor characteristics were similar or stronger than in our first paper based on a subset of these data (23). In analyses of DA and NDA, we found that DA was significantly associated with increased breast cancer risk and NDA was significantly associated with decreased risk and that these were independent risk factors for breast cancer. Further, statistically significant associations of the absolute DA and NDA measures with breast cancer were apparent for all tumor characteristics evaluated. Our findings suggest greater magnitude of association for DA with ER+ vs. ER− disease and PR+ vs. PRv disease and stronger associations of NDA with ER− vs. ER+ disease in women <55 years. We also observed significant positive and inverse trends for associations of DA and NDA, respectively, with tumor size across all ages.

Our findings of opposing associations of DA and NDA with breast cancer risk generally agree with most of the existing literature in this area, including a recent large meta-analysis that included several of the studies here (7). However, while the meta-analysis found that associations for NDA were attenuated in many studies upon adjustment for absolute DA (7), we did not observe attenuation in mutually adjusted models, possibly because correlations between DA and NDA were low (0.06–0.29) and similar across studies or because we adjusted for BMI, which is a surrogate for NDA. We conclude DA and absolute NDA are independent risk factors associated with breast cancer risk.

Few previous studies reported associations of absolute DA or NDA with breast cancer according to specific tumor characteristics. Consistent with our findings, in 601 cases and 667 controls from the Multiethnic Cohort, absolute DA was associated with both invasive breast cancer and DCIS (24); although in the present analysis, there was suggestion of a stronger association for invasive cancers vs. DCIS among women ≥55 years. Also in the Multiethnic Cohort, stronger associations of DA with ER+/PR+ vs. ER−/PR− tumors were observed (25). Like us, Eriksson et al. (26) reported stronger associations of absolute DA with ER+ vs. ER− tumors (p-value =0.065) and with PR+ vs. PR− (p-value = 0.099) in a case-only study of 110 breast cancer patients. Positive associations of absolute DA with ER+ vs. ER− tumors and larger vs. smaller tumors were also observed in recent UK case-control study (27). Similar to our findings, a case-only study among postmenopausal women (n=286) reported a non-significant positive trend of DA with tumor size and a non-significant trend of NDA with tumor size as well as significant positive associations between DA and ER and PR positivity (28). Our study is among the first to comprehensively explore associations of absolute NDA with breast tumor characteristics and, to our knowledge, is the largest to date. Current hypotheses to explain associations between increased MD and breast cancer risk have been reviewed recently (29) and include the higher amount of fibroglandular tissue “at risk” of transformation into cancer (30) and the increased epithelial and fibroblast cellular activity and interaction between stroma and epithelium in dense tissue (31, 32) as well as hormonal mechanisms, including the influence of sex steroid hormones and growth factors on density and breast cancer risk (33). Evaluating associations by tumor characteristics can provide insight into these hypothesized mechanisms. If the mechanism of action were purely through hormonal influences, then we might expect to observe associations of DA with ER+ tumors only; however, we observed significant positive associations of DA with both ER+ and ER− tumors, although the magnitude of association was greater for ER+ tumors among women <55 years. Moreover, we observed strong inverse associations of NDA with ER− tumors in this age group, independent of DA. Our findings of independent associations of DA and NDA with breast cancer risk across tumor characteristics suggest that several causal pathways may play a role in associations with risk. Petterson and Tamimi (34) propose several mechanisms by which breast fat (nondense area) may lead to reduced risk of breast cancer, including the possible direct effect of adipose tissue on normal breast development, indirect effects of adipose tissue in regard to the endocrine environment of the breast, or via lobular involution, which is positively correlated with NDA and inversely associated with breast cancer risk (35). On the other hand, some studies have suggested breast fat as a risk factor for breast cancer (6, 36).

As in our previously published analysis based on a subset of these data (23), we found that percent MD was more strongly associated with risk of ER-negative breast cancer than with ER-positive breast cancer among women < 55 years of age. Our current findings of the MD area phenotypes further suggest that the positive association observed between percent MD and ER-negative disease among women <55 years is driven by the inverse association of non-dense area with ER-negative disease in this group, rather than by a positive association with absolute dense breast area. Based on the results of our analyses and considering the current body of published literature on this topic, it appears that breast density (including percent and area measures) plays an important role in tumor aggressiveness, especially in younger women, giving differential associations observed with respect to tumor size, nodal status as well as ER-status. In light of the lack of significant age-interaction, however, we cannot discount an association of MD phenotypes with tumor aggressiveness among older women.

Limitations of the study have been described (3) and include variation in study design and populations, use of clinical pathology as opposed to central pathology review; changes in diagnostic criteria over time that may influence tumor characteristics and receptor status, in particular, and generalizability of results primarily to Caucasian women. Even with 4000 cases, power to detect age-interactions remained limited. Detection bias is also a potential limitation, given that extent of breast density may make earlier tumors more difficult to detect on screening mammogram (37). While we were not able to evaluate the influence of detection bias directly in this analysis due to the lack of high-quality data regarding interval vs. screen-detected cancers for most included studies, in the Breast Cancer Surveillance Consortium, Kerlikowske et al. reported that higher breast density in premenopausal women was more strongly related to aggressive tumors and that this finding persisted in analyses restricted to screen-detected cases only (38). We did find evidence of study heterogeneity for the analyses of DA with overall breast cancer and by invasive vs. in situ status, so these results should be cautiously interpreted. However, our associations of these absolute measures with overall breast cancer were consistent with the literature. Finally, this study relied on digitized film mammograms vs. more contemporary full field digital mammograms.

Strengths of this pooled analysis include the large sample size with mammograms available years prior to the cancer (for cases), standardized estimates of NDA and DA, detailed information on covariates and tumor characteristics from pathology reports, supplemented with information from TMAs, and screening mammograms assessed in a generally systematic fashion.

In summary, we found that percent MD and absolute dense breast area were associated with increased breast cancer risk while non-dense area was associated with decreased risk across all ages and invasive tumor characteristics. Among women <55 years, dense area was more strongly associated with an increased risk for ER+ vs. ER− tumors [p-heterogeneity (het) = 0.02] while non-dense area was more strongly associated with a decreased risk for ER− vs. ER+ tumors [p-het = 0.03]. Dense area was similarly associated with increased risk (and non-dense area decreased risk) of both node-positive and node-negative tumors, while significant trends in the magnitude of these associations were observed with increasing tumor size.

Our results suggest DA is positively associated (and NDA, inversely associated) with breast cancer across tumor characteristics. Further, these results suggest differential associations for these phenotypes with ER+ vs. ER− tumors, particularly in younger women. As such, DA and NDA may be important to consider when developing age- and subtype-specific risk models for breast cancer. Further research is warranted to clarify the possible differential associations of DA and NDA on breast cancer risk according to tumor characteristics.

Supplementary Material

1
2
3
4

Acknowledgments

Financial Support:

This work was supported in part by the National Institutes of Health, National Cancer Institute (NCI) grants listed below:

  • R01 CA140286 R01 CA128931 and R01 CA97396 to Dr. Vachon

  • R01 CA124865 and R01 CA131332 to Dr. Tamimi;

  • P50 CA116201 to Dr. Ingle (Dr. Couch has Project on SPORE);

  • R01 CA140286, P50 CA58207, U01 CA63740 and P01 CA154292 to Dr. Kerlikowske;

  • R01 CA116167 to Dr. Couch;

  • P01CA087969 and UM1 CA186107 to Dr. Stampfer (supporting NHS); and

  • R01 CA050385, UM1 CA176726 and Breast Cancer Research Foundation to Dr. Willett (supporting NHS2).

Additional support was provided by the Simeon J. Fortin Charitable Foundation, Bank of America, N.A., Co-Trustee to Dr. Bertrand; Department of Defense (DAMD 17-00-1-033) to Dr. Vachon;

We would like to thank the participants and staff of all the studies for their valuable contributions as well as the following state cancer registries for their help: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, MN, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, WI, WY. We also thank Fang Fang Wu who provided percent mammographic density estimates on all Mayo studies. The authors assume full responsibility for analyses and interpretation of these data.

Abbreviations

MD

Mammographic density

MMHS

Mayo Mammography Health Study

MCBCS

Mayo Clinic Breast Cancer Study

NHS and NHSII

Nurses’ Health Study

MCMAM

Mayo Clinic Mammography Study

SFMR

San Francisco Bay Area Breast Cancer SPORE and San Francisco Mammography Registry

SEER

Surveillance Epidemiology and End Results

BMI

body mass index

OR

odds ratios

CI

confidence intervals

Footnotes

Disclosure of Potential Conflicts of Interest:

The authors declare that they have no conflicts of interests to disclose.

References

  • 1.Yaffe MJ. Mammographic density. Measurement of mammographic density. Breast Cancer Res. 2008;10:209. doi: 10.1186/bcr2102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.McCormack VA, dos Santos Silva I. Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis. Cancer Epidemiol Biomarkers Prev. 2006;15:1159–69. doi: 10.1158/1055-9965.EPI-06-0034. [DOI] [PubMed] [Google Scholar]
  • 3.Bertrand KA, Tamimi RM, Scott CG, Jensen MR, Pankratz VS, Visscher D, et al. Mammographic density and risk of breast cancer by age and tumor characteristics. Breast Cancer Res. 2013;15:R104. doi: 10.1186/bcr3570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Pettersson A, Hankinson SE, Willett WC, Lagiou P, Trichopoulos D, Tamimi RM. Nondense mammographic area and risk of breast cancer. Breast Cancer Res. 2011;13:R100. doi: 10.1186/bcr3041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Stone J, Ding J, Warren RM, Duffy SW, Hopper JL. Using mammographic density to predict breast cancer risk: dense area or percentage dense area. Breast Cancer Res. 2010;12:R97. doi: 10.1186/bcr2778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Lokate M, Peeters PH, Peelen LM, Haars G, Veldhuis WB, van Gils CH. Mammographic density and breast cancer risk: the role of the fat surrounding the fibroglandular tissue. Breast Cancer Res. 2011;13:R103. doi: 10.1186/bcr3044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Pettersson A, Graff RE, Ursin G, Santos Silva ID, McCormack V, Baglietto L, et al. Mammographic Density Phenotypes and Risk of Breast Cancer: A Meta-analysis. J Natl Cancer Inst. 2014;106 doi: 10.1093/jnci/dju078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Heine JJ, Scott CG, Sellers TA, Brandt KR, Serie DJ, Wu FF, et al. A novel automated mammographic density measure and breast cancer risk. J Natl Cancer Inst. 2012;104:1028–37. doi: 10.1093/jnci/djs254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Olson JE, Sellers TA, Scott CG, Schueler BA, Brandt KR, Serie DJ, et al. The influence of mammogram acquisition on the mammographic density and breast cancer association in the Mayo Mammography Health Study Cohort. Breast Cancer Res. 2012;14:R147. doi: 10.1186/bcr3357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kelemen LE, Couch FJ, Ahmed S, Dunning AM, Pharoah PD, Easton DF, et al. Genetic variation in stromal proteins decorin and lumican with breast cancer: investigations in two case-control studies. Breast Cancer Res. 2008;10:R98. doi: 10.1186/bcr2201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Wang X, Goode EL, Fredericksen ZS, Vierkant RA, Pankratz VS, Liu-Mares W, et al. Association of genetic variation in genes implicated in the beta-catenin destruction complex with risk of breast cancer. Cancer Epidemiol Biomarkers Prev. 2008;17:2101–8. doi: 10.1158/1055-9965.EPI-08-0134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Tamimi RM, Hankinson SE, Colditz GA, Byrne C. Endogenous sex hormone levels and mammographic density among postmenopausal women. Cancer Epidemiol Biomarkers Prev. 2005;14:2641–7. doi: 10.1158/1055-9965.EPI-05-0558. [DOI] [PubMed] [Google Scholar]
  • 13.Tworoger SS, Sluss P, Hankinson SE. Association between plasma prolactin concentrations and risk of breast cancer among predominately premenopausal women. Cancer Res. 2006;66:2476–82. doi: 10.1158/0008-5472.CAN-05-3369. [DOI] [PubMed] [Google Scholar]
  • 14.Colditz GA, Hankinson SE. The Nurses’ Health Study: lifestyle and health among women. Nat Rev Cancer. 2005;5:388–96. doi: 10.1038/nrc1608. [DOI] [PubMed] [Google Scholar]
  • 15.Vachon CM, Brandt KR, Ghosh K, Scott CG, Maloney SD, Carston MJ, et al. Mammographic breast density as a general marker of breast cancer risk. Cancer Epidemiol Biomarkers Prev. 2007;16:43–9. doi: 10.1158/1055-9965.EPI-06-0738. [DOI] [PubMed] [Google Scholar]
  • 16.Kerlikowske K, Carney PA, Geller B, Mandelson MT, Taplin SH, Malvin K, et al. Performance of screening mammography among women with and without a first-degree relative with breast cancer. Ann Intern Med. 2000;133:855–63. doi: 10.7326/0003-4819-133-11-200012050-00009. [DOI] [PubMed] [Google Scholar]
  • 17.Kerlikowske K, Shepherd J, Creasman J, Tice JA, Ziv E, Cummings SR. Are breast density and bone mineral density independent risk factors for breast cancer? J Natl Cancer Inst. 2005;97:368–74. doi: 10.1093/jnci/dji056. [DOI] [PubMed] [Google Scholar]
  • 18.Ziv E, Tice J, Smith-Bindman R, Shepherd J, Cummings S, Kerlikowske K. Mammographic density and estrogen receptor status of breast cancer. Cancer Epidemiol Biomarkers Prev. 2004;13:2090–5. [PubMed] [Google Scholar]
  • 19.Boyd NF, Stone J, Martin LJ, Jong R, Fishell E, Yaffe M, et al. The association of breast mitogens with mammographic densities. Br J Cancer. 2002;87:876–82. doi: 10.1038/sj.bjc.6600537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Shepherd JA, Kerlikowske K, Ma L, Duewer F, Fan B, Wang J, et al. Volume of mammographic density and risk of breast cancer. Cancer Epidemiol Biomarkers Prev. 2011;20:1473–82. doi: 10.1158/1055-9965.EPI-10-1150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Prevrhal S, Shepherd JA, Smith-Bindman R, Cummings SR, Kerlikowske K. Accuracy of mammographic breast density analysis: results of formal operator training. Cancer Epidemiol Biomarkers Prev. 2002;11:1389–93. [PubMed] [Google Scholar]
  • 22.Tamimi RM, Baer HJ, Marotti J, Galan M, Galaburda L, Fu Y, et al. Comparison of molecular phenotypes of ductal carcinoma in situ and invasive breast cancer. Breast Cancer Res. 2008;10:R67. doi: 10.1186/bcr2128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bertrand KA, Tamimi RM, Scott CG, Jensen MR, Pankratz VS, Visscher D, et al. Mammographic density and risk of breast cancer by age and tumor characteristics. Breast Cancer Res. 2013;15:R104. doi: 10.1186/bcr3570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Gill JK, Maskarinec G, Pagano I, Kolonel LN. The association of mammographic density with ductal carcinoma in situ of the breast: the Multiethnic Cohort. Breast Cancer Res. 2006;8:R30. doi: 10.1186/bcr1507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Conroy SM, Pagano I, Kolonel LN, Maskarinec G. Mammographic density and hormone receptor expression in breast cancer: the Multiethnic Cohort Study. Cancer Epidemiol. 2011;35:448–52. doi: 10.1016/j.canep.2010.11.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Eriksson L, Hall P, Czene K, Dos Santos Silva I, McCormack V, Bergh J, et al. Mammographic density and molecular subtypes of breast cancer. Br J Cancer. 2012;107:18–23. doi: 10.1038/bjc.2012.234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Ding J, Warren R, Girling A, Thompson D, Easton D. Mammographic density, estrogen receptor status and other breast cancer tumor characteristics. Breast J. 2010;16:279–89. doi: 10.1111/j.1524-4741.2010.00907.x. [DOI] [PubMed] [Google Scholar]
  • 28.Ghosh K, Brandt KR, Sellers TA, Reynolds C, Scott CG, Maloney SD, et al. Association of mammographic density with the pathology of subsequent breast cancer among postmenopausal women. Cancer Epidemiol Biomarkers Prev. 2008;17:872–9. doi: 10.1158/1055-9965.EPI-07-0559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Pettersson A, Tamimi R. Breast density and breast cancer risk: understanding of biology and risk. Curr Epidemiol Rep. 2014;1:120–9. [Google Scholar]
  • 30.Trichopoulos D, Lipman RD. Mammary gland mass and breast cancer risk. Epidemiology. 1992;3:523–6. doi: 10.1097/00001648-199211000-00011. [DOI] [PubMed] [Google Scholar]
  • 31.Li T, Sun L, Miller N, Nicklee T, Woo J, Hulse-Smith L, et al. The association of measured breast tissue characteristics with mammographic density and other risk factors for breast cancer. Cancer Epidemiol Biomarkers Prev. 2005;14:343–9. doi: 10.1158/1055-9965.EPI-04-0490. [DOI] [PubMed] [Google Scholar]
  • 32.Guo YP, Martin LJ, Hanna W, Banerjee D, Miller N, Fishell E, et al. Growth factors and stromal matrix proteins associated with mammographic densities. Cancer Epidemiol Biomarkers Prev. 2001;10:243–8. [PubMed] [Google Scholar]
  • 33.Martin LJ, Boyd NF. Mammographic density. Potential mechanisms of breast cancer risk associated with mammographic density: hypotheses based on epidemiological evidence. Breast Cancer Res. 2008;10:201. doi: 10.1186/bcr1831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Pettersson A, Tamimi RM. Breast fat and breast cancer. Breast Cancer Res Treat. 2012;135:321–3. doi: 10.1007/s10549-012-2186-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ghosh K, Hartmann LC, Reynolds C, Visscher DW, Brandt KR, Vierkant RA, et al. Association between mammographic density and age-related lobular involution of the breast. J Clin Oncol. 2010;28:2207–12. doi: 10.1200/JCO.2009.23.4120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Beer AE, Billingham RE. Adipose tissue, a neglected factor in aetiology of breast cancer? Lancet. 1978;2:296. doi: 10.1016/s0140-6736(78)91694-x. [DOI] [PubMed] [Google Scholar]
  • 37.Boyd NF, Guo H, Martin LJ, Sun L, Stone J, Fishell E, et al. Mammographic density and the risk and detection of breast cancer. N Engl J Med. 2007;356:227–36. doi: 10.1056/NEJMoa062790. [DOI] [PubMed] [Google Scholar]
  • 38.Kerlikowske K, Cook AJ, Buist DS, Cummings SR, Vachon C, Vacek P, et al. Breast cancer risk by breast density, menopause, and postmenopausal hormone therapy use. J Clin Oncol. 2010;28:3830–7. doi: 10.1200/JCO.2009.26.4770. [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

1
NIHMS666538-supplement-1.pdf (37.2KB, pdf)
2
NIHMS666538-supplement-2.docx (14.4KB, docx)
3
NIHMS666538-supplement-3.pptx (1.6MB, pptx)
4
NIHMS666538-supplement-4.pptx (1.6MB, pptx)

RESOURCES