Abstract
Background
The empirical dietary inflammation pattern score (EDIP), which measures the ability of the diet to regulate chronic inflammation, is associated with both higher adiposity and breast cancer (BC) risk. Mammographic density (MD) is an important risk factor for BC.
Objective
We examined the associations between EDIP and mammographic features overall and stratified by menopausal status, and assessed the extent to which these associations are mediated by adiposity.
Methods
We included 4145 participants without BC in the Nurses’ Health Study (NHS) and NHSII. Cumulative average EDIP was assessed by food frequency questionnaires every 4–6 y. We assessed MD parameters (percent MD, dense area, and nondense area) and V (measure of grayscale variation). MD parameters were square-root transformed. Multivariable-adjusted linear regression models were used to analyze the associations between EDIP score and MD parameters. Baron and Kenny’s regression method was used to assess the extent to which the associations of EDIP and mammographic traits were mediated by BMI.
Results
In multivariable-adjusted models, EDIP was significantly inversely associated with percent MD [top compared with bottom quartile, β = –0.57; 95% confidence interval (CI): –0.78, –0.36]. Additional adjustment for BMI attenuated the association (β = –0.15; 95% CI: –0.34, 0.03), with 68% (β = 0.68, 20; 95% CI: 0.54, 0.86) mediation via BMI. In addition, EDIP was positively associated with nondense area after adjusting for BMI and other covariates. No associations were observed for dense area and V measure. Results were similar when stratified by menopausal status.
Conclusions
EDIP score was inversely associated with percent MD and positively associated with nondense area, and these associations were largely mediated by BMI.
Keywords: empirical dietary inflammation pattern, EDIP score, dietary pattern, mammographic density, mammographic features, breast density, Nurses’ Health Study
Introduction
Inflammation is a hallmark of cancer and chronic inflammation has been found to be associated with multiple cancers including breast cancer [1]. Breast cancer is the most commonly diagnosed cancer and is the leading cause of cancer-related deaths among females [2]. In the United States, >300,000 new cases of breast cancer are diagnosed annually, accounting for ∼30% of female cancers [2]. Overexpression of certain proinflammation markers including IL6, C-reactive protein (CRP), and TNF-α may result in biological processes associated with carcinogenesis. For example, IL-6 and TNF-α have an important role in regulating estrogen synthesis in peripheral tissues including normal breast tissues [3], whereas increase in IL6 and TNF-α were shown to elevate aromatase activity within breast tissue that may lead to increased mammographic density [3], a risk factor for breast cancer. The empirical dietary inflammatory pattern (EDIP) score [4] is a food-based dietary index to assess the ability of the diet to regulate chronic inflammation. EDIP was derived from investigating the association between intakes of certain food groups and circulating inflammatory markers including (IL-6, CRP, and TNFαR2) [4].
Higher mammographic density is a well-established factor associated with risk factor for breast cancer. The dense part of the breast consists of epithelial tissue and stroma and appears bright on a mammogram, whereas fat tissue appears dark. Previous studies showed that women of the same age and BMI (kg/m2), with very dense (>75% density) breasts, have a 4–6-fold greater risk of breast cancer compared with women with little density (<5%–10%) or fatty breasts [5,6]. There is an interplay between inflammation and dense stromal tissue in the breast. Mammographic density reflects the proliferation of fibro-glandular (dense) tissue in the breast, relative to fatty (nondense) tissue, and inflammatory markers affect cellular proliferation [7,8], which may increase breast cancer risk. Also, previous findings showed that women with higher expression of proinflammatory markers (IL6, IL8, CPR, and TNF-α) in breast tissue had remarkably higher percent mammographic density than those having lower expression of proinflammatory markers [9]. The EDIP score, developed based on these circulating inflammatory markers is an obesogenic dietary pattern, associated with greater risk of higher adiposity and breast cancer [10,11]. In addition to mammographic density parameters, several studies have shown that mammographic texture features (V measure) are associated with breast cancer risk independent of mammographic density [[12], [13], [14], [15]].
Interestingly, although both BMI (among postmenopausal women) and mammographic density are associated with greater risk of breast cancer, BMI is inversely associated with mammographic density [16], whereas positively correlated with the fatty or nondense area of the breast [17]. The percentage of adipose tissue in the breast is a key determinant of mammographic density, and higher proportion of adipose tissue in the breast correlates with lower mammographic density [18]. Therefore, we hypothesize that a proinflammatory dietary pattern may influence mammographic features largely (although not entirely) via mediation by adiposity. Adiposity may also modify the association of diet-related inflammation and mammographic density, where the association may vary by levels of fibroglandular tissue compared with adipose tissue in the breast based on overall body size. Additional complexity in the relationships between diet-related inflammation, adiposity, and mammographic density is potential confounding by adiposity, which has been shown to be associated with the EDIP dietary pattern, while also directly impacting mammographic density differentially by age (menopausal status) [19,20].
To address these important yet complex relationships between diet-related inflammation, adiposity, and mammographic density, this study examined the associations between EDIP and mammographic features overall and stratified by menopausal status, and assessed the extent to which these associations are mediated by adiposity.
Methods
Study population
The study included women from the Nurses’ Health Study (NHS) and NHSII. NHS was initiated in 1976 when 121,700 registered nurses in the United States who were aged 30–55 y returned an initial mailed questionnaire with detailed information on their medical and reproductive histories. The information on BMI, reproductive history, age at menopause, menopausal hormone therapy (MHT) use, and any diagnoses of cancer or other disease were updated biennially using questionnaires. The response rate at each questionnaire cycle documented as >90% [21]. Every 2 y since 1976, follow-up questionnaires have been mailed to update exposure status and to identify new cases of cancer (and other diseases of interest) diagnosed within the previous 2-y period [22,23].
The study protocol was approved by the institutional review boards of the Brigham and Women’s Hospital and Harvard T.H. Chan School of Public Health, and those of participating registries as required.
Measurement of mammographic density and texture feature
We collected film-screening mammograms from eligible women in the nested case-control study. From the mammograms collected, the cranio-caudal views of both breasts were digitized at 261 μm/pixel with a Lumysis 85 laser film scanner (Lumysis) or a VIDAR CAD PRO Advantage scanner (VIDAR Systems Corporation) (using a comparable resolution of 150 dots per inch and 12-bit depth inch and 12-bit depth). The correlation between percent mammographic density from the 2 scanners was very high (r = 0.88) [24]. We used computer-assisted thresholding to measure dense area, nondense area (dense area subtracted from the total area), and percent mammographic density calculated as dense area divided by the total breast area (Cumulus software, University of Toronto) [25,26]. Analyses of these measurements were based on values averaged across left and right breasts. Mammographic density of the right and left breast is highly correlated [27]. A detailed description of mammographic density measurement in the NHS and NHSII participants was reported in an earlier publication [24]. The mammogram reader was blinded to the subject's case/control status and biomarker data. Two observers read the mammograms from NHS participants in 2 batches. For NHSII, a single observer read mammograms in 3 batches. There was high reproducibility within each batch [25]. We included controls with measured mammographic density features N = 4342 (NHS = 2514 and NHSII = 1828). A texture summary measure called “V” captures grayscale variation in mammograms and was shown to be a stronger predictor of breast cancer risk than percent mammographic density [28]. The summary measure of texture feature “V” is an automated measure that captures grayscale variation on a mammogram. The algorithm and the method have been described previously [28]. For a subset N = 3252 (NHS = 2183 and NHSII = 1069) of women, texture feature “V” was measured.
Diet assessment and calculation of the EDIP score
Food frequency questionnaires (FFQs) data are updated prospectively every 4 y in the NHS (since 1980) and NHSII (since 1991) [[29], [30], [31]]. With each questionnaire, participants quantified how frequently they consumed a standard portion of a specific food item during the previous y, using 9 categories ranging from, “never or less than once per month,” to “more than 6 times daily.” The EDIP score is a food-based dietary index for assessing dietary inflammatory potential, which is determined by the circulating concentrations of CRP, IL-6, and TNFαR2. EDIP is a weighted sum of 18 food groups from FFQs derived in the NHS [4] and validated in NHSII and the Health Professionals Follow-Up Study [4,31,32]. The component food groups comprising the EDIP score that are associated with an increase in concentrations of inflammatory markers are processed meat, red meat, organ meat, nondark fish and other seafood such as shrimp and lobster, vegetables other than green leafy vegetables/dark yellow vegetables (e.g., carrots, yam/sweet potatoes), refined grains, high-energy beverages (carbonated beverages with sugar, fruit drinks), low-energy beverages (low-energy cola/other carbonated beverages), and tomatoes. The component food groups that are inversely associated with concentrations of inflammatory markers are beer, wine, tea, coffee, dark yellow vegetables (e.g., carrots, yellow squash, sweet potatoes), green leafy vegetables, snacks (including wholegrain snacks and dark chocolate), fruit juice, and pizza [4]. Dietary intake with a higher EDIP score reflects proinflammatory potential, whereas food intake with a lower EDIP score reflects anti-inflammatory potential. Cumulative average EDIP scores were calculated for N = 4280 women (NHS = 2488 and NHSII = 1792) using data from all FFQs preceding the mammogram date. Consistent with previous work, we calculated continuous EDIP score for each participant at each questionnaire cycle [[33], [34], [35]] and adjusted for total energy intake using the residual method [36]. The cumulative average EDIP score was computed in each questionnaire cycle by averaging the EDIP from baseline through to that follow-up cycle (median 8, range 5–8 FFQs) and was categorized into quartiles.
The study participant flowchart with reasons for exclusion is given in Figure 1.
FIGURE 1.
Participant flowchart.
Covariates
We selected covariates based on our previous and ongoing work examining the role of dietary patterns on circulating cardiometabolic health markers and on breast cancer risk [11,33,37,38]. Weight and height were reported on study questionnaires; self-reported weight and height were previously validated in the NHSII [39]. Height was obtained at enrollment. Weight (and other covariates) was updated every 2 y, starting from 1976 (NHS) or 1991 (NHSII). BMI was evaluated as cumulative average BMI over the available values before and including the participants’ weight at the time of mammography. Percent mammographic density is inversely associated with BMI [40] and positively associated with breast cancer risk [41]. Therefore, it is necessary to adjust for BMI in any study including mammographic features. The following information on covariates was obtained from the questionnaire closest to the time of mammogram collection: age (y), MHT use (postmenopausal and never used MHT, postmenopausal and formerly used MHT, and postmenopausal and currently using MHT), age at menopause (y), age at menarche (y), waist-to-hip-ratio (WHR), age at first birth and parity (nulliparous, 1–2 children and <25 y old, 1–2 children and 25+, ≥3 children and <5 y old, ≥3 children and 25+ y old), physical activity [metabolic equivalent of task (MET)-h/wk), history of benign breast diseases (no, yes), family history of breast cancer (no, yes), and diabetes status (no, yes)]. Women were considered postmenopausal if they reported any of the following: no menstrual periods within the previous 12 mo (i.e., natural menopause), bilateral oophorectomy (i.e., surgical menopause), or hysterectomy with 1 or more intact ovaries. Also, women were considered postmenopausal if they were ≥54 y and smoked or ≥56 y and were a nonsmoker [42,43].
Statistical analysis
Descriptive statistics for established determinants of mammographic density parameters and/or EDIP score among breast cancer controls according to quartiles of energy-adjusted cumulative average EDIP score are shown in Table 1. The distributions of dense area, nondense area, and percent mammographic density were skewed. Square-root transformation yielded normal distributions for each of these parameters. We used linear regression models with 95% confidence intervals (CIs) for the following: 1) the association between energy-adjusted cumulative average EDIP score with square-root-transformed mammographic density parameters and V measure, adjusted for the potential confounders (described in detail in the covariates section above) except for BMI (i.e., the total association between the exposure and outcome—direct effect) and 2) the same association with additional adjustment for BMI (that is, the association between the exposure and outcome not through BMI indirect effect). Finally, to more formally quantify the extent to which the associations between EDIP score and mammographic features are mediated by BMI (continuous), we performed mediation analyses based on Baron and Kenny's 4-step regression approach [44] and the bootstrapping method was conducted to test the statistical significance of the mediation and compute 95% CIs [45]. We used the bootstrapping technique implemented in the R package “mediation” [46].
TABLE 1.
Cohort characteristics according to quartiles EDIP in Nurses' Health Study and Nurses' Health Study II combined (N = 4145) among controls.
| Characteristics | Quartile of cumulative average (EDIP score)1 |
|||
|---|---|---|---|---|
| Q1 N = 1036 (–7.99 to <–0.68) |
Q2 N = 1036 (≥–0.68 to ≤–0.10) |
Q3 N = 1037 (>–0.10 to ≤0.46) |
Q4 N = 1036 (>0.46 to 4.24) |
|
| Age (y), mean (SD) | 52.8 (8.6) | 52.4 (8.7) | 53.4 (9.1) | 53.2 (9.3) |
| Menopausal status, N (%) | ||||
| Premenopausal | 531 (51.3) | 531 (51.3) | 482 (46.5) | 459 (44.3) |
| Postmenopausal | 505 (48.7) | 505 (48.7) | 555 (53.5) | 577 (55.7) |
| MHT status, N (%) | ||||
| Postmenopausal never used MHT | 145 (14.0) | 142 (13.7) | 164 (15.8) | 171 (16.5) |
| Postmenopausal formerly used MHT | 92 (8.9) | 95 (9.2) | 99 (9.5) | 109 (10.5) |
| Postmenopausal currently using MHT | 169 (16.3) | 169 (16.3) | 185 (17.8) | 186 (18.0) |
| Age at menopause2 (y), mean (SD) | 49.0 (4.6) | 48.4 (4.9) | 48.7 (4.9) | 48.4 (5.2) |
| BMI (kg/m2), mean (SD) | 24.8 (4.5) | 25.3 (4.7) | 26.0 (5.3) | 27.6 (6.2) |
| <25 | 641 (61.9) | 583 (56.3) | 548 (52.8) | 435 (42.0) |
| 25–29.9 | 278 (26.8) | 302 (29.2) | 290 (28.0) | 289 (27.9) |
| ≥30 | 117 (11.3) | 151 (14.6) | 199 (19.2) | 312 (30.1) |
| Waist-hip-ratio, mean (SD) | 0.78 (0.07) | 0.79 (0.7) | 0.80 (0.13) | 0.80 (0.08) |
| <0.85 | 607 (58.6) | 600 (57.9) | 556 (53.6) | 518 (50.0) |
| ≥0.85 | 91 (8.8) | 113 (10.9) | 152 (14.7) | 167 (16.1) |
| Age at menarche (y), mean (SD) | 12.4 (1.4) | 12.6 (1.4) | 12.5 (1.4) | 12.5 (1.4) |
| Age at first birth/parity, N (%) | ||||
| Nulliparous | 119 (11.5) | 114 (11.00) | 109 (10.5) | 100 (9.7) |
| 1 or 2 children and <25 y old | 150 (14.5) | 112 (10.8) | 145 (14.0) | 173 (16.7) |
| 1 or 2 children and 25+ y old | 304 (29.3) | 305 (29.4) | 262 (25.3) | 265 (25.6) |
| 3+ children and <25 y old | 242 (23.4) | 292 (28.2) | 294 (28.4) | 290 (28.0) |
| 3+ children and 25+ y old | 213 (20.6) | 202 (19.5) | 218 (21.0) | 203 (19.6) |
| Physical activity, mean (SD), MET-h/wk | 19.6 (23.0) | 16.5 (19.5) | 18.5 (32.3) | 14.5 (16.7) |
| Benign breast diseases, N (%) | ||||
| No | 819 (79.1) | 823 (79.4) | 844 (81.4) | 826 (79.7) |
| Yes | 217 (20.9) | 213 (20.6) | 193 (18.6) | 210 (20.3) |
| Family history of breast cancer, N (%) | ||||
| No | 930 (89.8) | 918 (88.6) | 917 (88.4) | 928 (89.6) |
| Yes | 106 (10.2) | 118 (11.4) | 120 (11.6) | 108 (10.4) |
Abbreviations: EDIP, empirical dietary inflammation pattern; MET, metabolic equivalent of task; MHT, menopausal hormone therapy.
Energy-adjusted EDIP score 1-cycle before mammogram.
Among postmenopausal women.
As a supplementary analyses, we evaluated the association of EDIP score with mammographic density parameters and V measure separately by menopausal status, by BMI categories [normal BMI (<25 kg/m2), obese and overweight (≥25 kg/m2)], and by WHR categories [normal WHR (<0.85), high WHR (≥0.85)]. Additionally, we evaluated the association between EDIP score per SD as a continuous value with mammographic density parameters and V measure. All P value tests for trend were conducted using the Wald test, where the medians of the quartiles were modeled as ordinal variables. Statistical significance in all the analyses was assessed at 0.05 level. All analyses were performed using R version 4.1.0.
Results
Baseline characteristics
Baseline characteristics for the 4145 NHS/NHSII women without breast cancer are presented by quartile of cumulative average EDIP score in Table 1. The median age at baseline mammogram was 51 y (interquartile range = 14). Of these, 2003 women were premenopausal and 2142 were postmenopausal at the time of their mammogram. Overall, women with a higher EDIP score had a higher BMI than those with a lower score (mean BMI for highest quartile compared with lowest: 27.6 kg/m2 compared with 24.8 kg/m2). Similarly, the proportion of women with higher WHR (≥0.85) was greater among those with greater EDIP score (16%) than women with lower EDIP score (8.8%) Additionally, women with higher EDIP score were less physically active than those with the lower EDIP score (mean physical activity MET-h/wk for highest quartile compared with lowest: 19.6 MET-h/wk compared with 14.5 MET-h/wk). There were no substantial differences in other determinants of mammographic density parameters and texture feature across quartiles of EDIP score.
EDIP score and mammographic density parameters
Table 2 shows the associations between quartiles of energy-adjusted cumulative average EDIP score, and mammographic density features (i.e., V measure, and square-root-transformed mammographic density parameters including percent mammographic density, dense area, and nondense area) with the lowest quartile considered as the reference category. The distribution statistics of the square-root-transformed mammographic density parameters are presented in Supplemental Table 1. We observed a statistically significant inverse association between EDIP score and percent mammographic density in the adjusted model (Q4 compared with Q1, β = –0.57; 95% CI: –0.78, –0.36; P-trend < 0.0001). Additional adjustment for BMI substantially attenuated the association and the results were no longer statistically significant (Q4 compared with Q1; β = –0.15; 95% CI: –0.34, 0.03; P-trend = 0.08). We observed no association between EDIP score and dense area before and after additionally adjustment for BMI. Finally, a strong positive association between EDIP score and nondense area was detected (Q4 compared with Q1, β = 1.56; 95% CI: 1.18, 1.93; P-trend < 0.0001). After further adjustment for BMI, this association was attenuated but remained statistically significant overall (Q4 compared with Q1, β = 0.41; 95% CI: 0.13,.69; P-trend = 0.01). When considering EDIP score as a continuous variable (per SD increments), the results remained unchanged compared with EDIP as a categorical variable (Supplemental Table 2).
TABLE 2.
Association between quartiles of EDIP score, square-root-transformed mammographic density parameters, and V measure among controls, Nurses’ Health Study and Nurses’ Health Study II combined (N = 4145).
| All women | Quartile cumulative average EDIP score1 |
P value of trend2 | Prop. mediated, 95% (CI) |
|||
|---|---|---|---|---|---|---|
| Q1 N = 1036 |
Q2 N = 1036 |
Q3 N = 1036 |
Q4 N = 1036 |
(P value)3 | ||
| Square-root mammographic percent density (%) | ||||||
| Mean, SD | 5.6 (1.8) | 5.4 (1.9) | 5.2 (1.9) | 5.0 (1.9) | ||
| Multivariable adjusted4 | Ref. | –0.29 (−0.49 to –0.08) | –0.45 (–0.65 to –0.24) | –0.57 (–0.78 to –0.36) | <0.0001 | |
| Multivariable-adjusted + BMI | Ref. | –0.20 (–0.38 to –0.01) | –0.27 (–0.46 to –0.09) | –0.15 (–0.34 to 0.03) | 0.08 | 0.68 (0.54–0.86) (P < 0.0001) |
| Square-root dense area (cm2) | ||||||
| Mean, SD | 6.1 (2.0) | 6.1 (2.2) | 6.0 (2.1) | 5.9 (2.2) | ||
| Multivariable adjusted4 | Ref. | –0.01 (–0.27 to 0.26) | –0.25 (–0.51 to 0.01) | –0.15 (–0.41 to 0.11) | 0.12 | |
| Multivariable-adjusted + BMI | Ref. | 0.02 (–0.24 to 0.28) | –0.20 (–0.46 to 0.06) | –0.02 (–0.29 to 0.24) | 0.54 | — |
| Square-root nondense area (cm2) | ||||||
| Mean, SD | 9.3 (3.0) | 9.7 (3.3) | 10.1 (3.3) | 10.7 (3.6) | ||
| Multivariable adjusted4 | Ref. | 0.77 (0.39 to 1.14) | 0.90 (0.53 to 1.27) | 1.56 (1.18 to 1.93) | <0.0001 | |
| Multivariable-adjusted + BMI | Ref. | 0.52 (0.24 to 0.79) | 0.43 (0.15 to 0.70) | 0.41 (0.13 to 0.69) | 0.01 | 0.70 (0.60–0.82) (P < 0.0001) |
| All women5 | ||||||
| V measure | ||||||
| Mean, SD | 0.001 (0.97) | –0.03 (0.98) | –0.12 (1.00) | –0.18 (0.97) | ||
| Multivariable adjusted4 | Ref. | –0.01 (–0.13 to 0.11) | –0.07 (–0.18 to 0.04) | –0.07 (–0.19 to 0.04) | 0.15 | |
| Multivariable-adjusted + BMI | Ref. | 0.01 (–0.10 to 0.12) | –0.01 (–0.12 to 0.10) | 0.06 (–0.05 to 0.17) | 0.35 | — |
Abbreviations: CI, confidence interval; EDIP, Empirical dietary inflammation pattern; V measure, measure of grayscale variation.
Energy-adjusted EDIP score 1-cycle before mammogram.
P value; tests for trend were conducted using the Wald test, where the medians of the quartiles were modeled continuously.
Mediation analysis was performed based on Baron and Kenny's 4-step regression approach and the bootstrapping method was conducted to test the statistical significance of the mediation. EDIP score considered continuous.
Adjusted for age at mammogram (continuous), the batch of mammography density reading (categorical), menopausal status (premenopausal, postmenopausal), age at menarche (continuous), age at first birth/parity (nulliparous; 1 or 2 children and <25 y old; 1 or 2 children and 25+ y old; 3+ children and <25 y old; 3+ children and 25+ y old), physical activity (continuous), benign breast diseases (yes, no), and family history of breast cancer (yes, no).
Among women with V measure (N = 3204).
We observed that 68% of the association between EDIP score with mammographic percent density (β = 0.68; 95% CI: 0.54, 0.86; P < 0.0001) and 70% of the association with nondense area (β = 0.70; 95% CI: 0.60, 0.82; P < 0.0001) were mediated by BMI.
When analyses were stratified by menopausal status, approximately similar patterns of results were seen for pre- and postmenopausal women (Supplemental Tables 3 and 4) as for the overall analyses.
Additional analyses stratified by BMI category (normal and overweight/obese) showed similar results for women with healthy BMI (BMI <25 kg/m2) as in the overall analyses (Supplemental Table 5). However, the associations between EDIP score with mammographic percent density and nondense area were slightly stronger among women with BMI ≥25 kg/m2 than women with healthy BMI.
When stratifying analyses by WHR categories (WHR < 0.85 and WHR ≥ 0.85), similar results as for the overall analyses were observed for WHR (<0.85) (Supplemental Table 6). Additionally, no association between EDIP score and mammographic percent density or nondense area was detected among women with high WHR.
Overall, mammographic percent density was inversely associated with BMI (Supplemental Figure 1A), with dense area and V measure demonstrating a less strong inverse association (Supplemental Figure 1B and C). However, nondense area, which represents adipose tissue in the breast, was positively associated with BMI (Supplemental Figure 1C).
Discussion
In this large and well-characterized cohort of 4145 women without breast cancer who participated in NHS and NHSII, we found a statistically significant inverse association between energy-adjusted cumulative average EDIP score and percent mammographic density. Results were attenuated and were no longer statistically significant after additionally adjusting for BMI. A statistically significant positive association between EDIP score and nondense area was also observed, Importantly, we found that BMI significantly mediated the association between EDIP score with mammographic percent density and with mammographic nondense area, providing important insights regarding the complex interplay of BMI, EDIP, and mammographic density.
The association between diet and mammographic density has been previously investigated; however, these studies focused on specific food items (e.g., milk and alcohol), dietary intakes (e.g., cholesterol, saturated fat, total dairy intake, and vitamins), and/or specific dietary patterns (Western and Mediterranean diets). Vachon et al. [47], in a cohort of 1508 women (premenopausal n = 283, postmenopausal n = 1225) who participated in the Minnesota Breast Cancer Family Study, found positive associations between mammographic density and fat, cholesterol intake, various meat, dairy food groups, different vitamins (vitamins C and E), and alcohol intake; however, overall dietary pattern was not assessed. In a cross-sectional study of 3548 Spanish women, Castelló et al. [48] found that women with higher adherence to a Western dietary pattern were more likely to have higher mammographic density than those with low adherence. However, no association between the Mediterranean dietary pattern and mammographic density was observed. A potential limitation of both mentioned studies was the use of a more limited, subjective measurement of mammographic density (by 1 radiologist), which is subject to bias [47,48]. The mechanisms by which diet may affect mammographic density are not known. Previous studies showed that specific dietary factors and macronutrients (e.g., animal fat, carbohydrate, and dairy intake) could potentially alter percent mammographic density [47] and dense area [49] through their impact on sex hormones. The role of plasma lipoproteins as carriers for steroid hormones could potentially lead to an elevation in hormone levels within breast tissue. This could subsequently result in increased mammographic dense area [50].
In this study, we sought to focus analyses on a specific dietary pattern (EDIP) and, furthermore, incorporate BMI into analyses given the established association between BMI and EDIP, and BMI and mammographic density. Specifically, higher long-term adherence to EDIP has been associated with higher BMI [37,51] in multiple studies as well as with long-term weight gain [11]. Additionally, BMI has been shown to mediate the association of EDIP with several health outcomes including type 2 diabetes [52] and endometrial cancer [53], in alignment with the attenuation of the association between EDIP and mammographic density measures in this study. In a study of 709 premenopausal women who participated in NHSII, Garzia et al. [54] investigated the association between adolescent and early adulthood dietary inflammatory potential (a proinflammatory dietary pattern and the Alternative Health Eating Index anti-inflammatory dietary pattern) with mammographic density. Similar to our findings, Garzia et al. [54] observed that as adolescent proinflammatory dietary pattern score increased, mammographic percent density decreased (P-trend = 0.005), and nondense area increased (P-trend < 0.0001). However, the associations were no longer significant when adjusted for BMI at mammogram nor in the multivariable-adjusted models. These results show that the association of proinflammatory diets associated with lower mammographic percent density may be largely affected by adiposity.
Understanding the association between mammographic density features and BMI is critical. Mammographic percent density (calculated by dividing the dense area by the total breast area) is inversely associated with BMI (Supplemental Figure 1A); however, nondense area (which represents the adipose tissue in the breast) is positively associated with BMI (Supplemental Figure 1C). It is well established that elevated mammographic percent density is an important risk factor for breast cancer [55]. Accumulating evidence consistently indicates a positive association between increased increase BMI/adiposity and postmenopausal breast cancer [56,57]. However, in premenopausal women, a higher BMI has been reported to be unrelated to breast cancer or associated with reduced breast cancer risk [57,58]. Furthermore, it has been shown that mammographic percent density and BMI are inversely associated and act as confounders to each other’s effects [57]. Similar to the results for mammographic dense area, we observed no association between EDIP score and V measure after adjusting for BMI. Evidence indicates that texture and mammographic density parameters (percent density and dense area) both contribute to predicting future breast cancer risk [15].
Strengths of our study are the availability of high-quality information on covariates, outcome, and the main exposure of interest (diet). Mammographic density was measured using a quantitative measure (Cumulus has been the gold standard for quantitative density measurement for many years now); and several validation studies have demonstrated this method’s high reproducibility [[59], [60], [61], [62]]). Additionally, having detailed information on lifestyle, demographic, and reproductive factors within both cohorts enabled us to account for potential confounding effects of lifestyle factors.
There were also limitations in our study. Although we had a comprehensive information on determinants of mammographic density and EDIP score in this study that allowed for adjustment for potential confounders, the potential for residual confounding especially with respect to adiposity is a concern. Also, in this study we used the self-reported dietary intake (diet assessment by FFQ) and covariates data, which are prone to information bias. However, the information bias is most likely nondifferential because women were not aware of their mammographic feature measurements. If anything, our estimates could therefore be diluted. Additionally, previous studies in these cohorts that evaluated the relative validity of FFQ data have shown reasonably good correlations between FFQs and diet records, which suggests that dietary intake is generally well measured in our cohorts [4,31,32]. Given the limitation of BMI as a measure of adiposity, future studies to investigate the influence of EDIP in the context of more objective body composition and fat distribution measures are warranted. Another study limitation is that the associations were based on a single measurement of mammographic density. Because mammographic density is a dynamic trait that typically declines with increasing age [27,63], a single mammogram might not be reflective of the woman’s lifetime mammographic density pattern. However, studies have suggested that a single mammographic density measure can predict breast cancer risk for ≤10 y in both pre- and postmenopausal women [64]. Furthermore, the distributions of mammographic density parameters (including percent mammographic density, dense area, and nondense area) were skewed. To address this, we applied a square-root transformation of these parameters. However, this transformation somewhat limited our ability to interpret the result. Despite this, previous research conducted on the same cohort (NHS and NHSII) demonstrated that the direction and relative magnitude of associations remained consistent across models regardless of whether they used untransformed or square-root-transformed density parameters [65]. Finally, the study population consists largely of White nurses, potentially limiting generalizability, and although additional studies in more diverse populations are needed to further confirm our results, we have successfully applied the EDIP score in cohorts with higher racial diversity including the Women’s Health Initiative [37].
In conclusion, this large analysis showed that a proinflammatory dietary pattern (based on higher cumulative average EDIP score) was inversely associated with mammographic percent density and positively associated with nondense area after adjusting for covariates. Additional adjustment for BMI substantially attenuated the associations reflecting our primary hypothesis of mediation, which we formally confirmed. No associations between EDIP score and nondense area or V texture feature measure were observed before and after additional adjustment for BMI. Future studies to dissect the role of diet-related inflammation in the context of body composition and fat distribution are necessary to further understand the complex interplay of EDIP, adiposity, and mammographic density as a risk factor for breast cancer.
Author contributions
The authors’ responsibilities were as follows – SA: research design, analyses, writing of the manuscript, critical input into the writing of the manuscript, and reading and approval of the final manuscript; RMT, DGS, FKT: conceptualization, project supervision, and critical input into the writing of the manuscript, and reading and approval of the final manuscript; and all authors: made substantial contributions to the interpretation of data, revisions to the manuscript for important intellectual content, writing—original draft, and review and approval of the final version.
Funding
This work was supported by Ohio State Division of Medical Oncology Pilot Grant (SA, FKT, DGS), U01CA260352 (RMT, DGS), National Cancer Institute at the National Institutes of Health [CA131332, CA124865, CA175080 to RMT, UM1 CA186107 and P01 CA087969, to MS, U01 CA200464 to JH], Avon Foundation for Women, Susan G. Komen for the Cure (to RMT), and Breast Cancer Research Foundation and Ramon Areces Foundation (to AR-N). NHSII cohort infrastructure grant [U01 CA176726].
Data availability
The data that support the findings of this study are available from Nurses’ Health Study, but restrictions apply to the availability of these data, which were used under license for this study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission.
Code availability
The underlying code for this study is not publicly available but may be made available to qualified researchers on reasonable request from the corresponding author.
Conflict of interest
PS received salary from Merck & Co., Inc. However, this had no role in the study design, data collection, analyses, and data interoperation, in writing the paper or in the decision to submit the paper for publication. All other authors report no conflicts of interest.
Acknowledgments
We would like to acknowledge the contribution to this study from central cancer registries supported through the Centers for Disease Control and Prevention’s National Program of Cancer Registries (NPCR) and/or the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) Program. Central registries may also be supported by state agencies, universities, and cancer centers. Participating central cancer registries include the following: Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, Florida, Georgia, Hawaii, Idaho, Indiana, Iowa, Kentucky, Louisiana, Massachusetts, Maine, Maryland, Michigan, Mississippi, Montana, Nebraska, Nevada, New Hampshire, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Puerto Rico, Rhode Island, Seattle SEER Registry, South Carolina, Tennessee, Texas, Utah, Virginia, West Virginia, and Wyoming.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.tjnut.2024.09.009.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
References
- 1.Mantovani A., Allavena P., Sica A., Balkwill F. Cancer-related inflammation. Nature. 2008;454(7203):436–444. doi: 10.1038/nature07205. [DOI] [PubMed] [Google Scholar]
- 2.Siegel R.L., Miller K.D., Wagle N.S., Jemal A. Cancer statistics. CA Cancer J. Clin. 2023;73(1):17–48. doi: 10.3322/caac.21763. 2023. [DOI] [PubMed] [Google Scholar]
- 3.Purohit A., Newman S.P., Reed M.J. The role of cytokines in regulating estrogen synthesis: implications for the etiology of breast cancer. Breast Cancer Res. 2002;4(2):65–69. doi: 10.1186/bcr425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tabung F.K., Smith-Warner S.A., Chavarro J.E., Wu K., Fuchs C.S., Hu F.B., et al. Development and validation of an empirical dietary inflammatory index. J. Nutr. 2016;146(8):1560–1570. doi: 10.3945/jn.115.228718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Byng J.W., Boyd N.F., Little L., Lockwood G., Fishell E., Jong R.A., et al. Symmetry of projection in the quantitative analysis of mammographic images. Eur. J. Cancer Prev. 1996;5(5):319–327. doi: 10.1097/00008469-199610000-00003. [DOI] [PubMed] [Google Scholar]
- 6.McCormack V.A., dos Santos Silva I. Breast density and parenchymal patterns as markers of breast cancer risk: a meta-analysis. Cancer Epidemiol. Biomarkers Prev. 2006;15(6):1159–1169. doi: 10.1158/1055-9965.EPI-06-0034. [DOI] [PubMed] [Google Scholar]
- 7.Giles E.D., Jindal S., Wellberg E.A., Schedin T., Anderson S.M., Thor A.D., et al. Metformin inhibits stromal aromatase expression and tumor progression in a rodent model of postmenopausal breast cancer. Breast Cancer Res. 2018;20(1):50. doi: 10.1186/s13058-018-0974-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Wellberg E.A., Kabos P., Gillen A.E., Jacobsen B.M., Brechbuhl H.M., Johnson S.J., et al. FGFR1 underlies obesity-associated progression of estrogen receptor-positive breast cancer after estrogen deprivation. JCI Insight. 2018;3(14) doi: 10.1172/jci.insight.120594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hanna M., Dumas I., Orain M., Jacob S., Têtu B., Sanschagrin F., et al. Association between expression of inflammatory markers in normal breast tissue and mammographic density among premenopausal and postmenopausal women. Menopause. 2017;24(5):524–535. doi: 10.1097/GME.0000000000000794. [DOI] [PubMed] [Google Scholar]
- 10.Jin Q., Shi N., Lee D.H., Rexrode K.M., Manson J.E., Balasubramanian R., et al. Hyperinsulinemic and pro-inflammatory dietary patterns and metabolomic profiles are associated with increased risk of total and site-specific cancers among postmenopausal women. Cancers (Basel) 2023;15(6):1756. doi: 10.3390/cancers15061756. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Tabung F.K., Satija A., Fung T.T., Clinton S.K., Giovannucci E.L. Long-term change in both dietary insulinemic and inflammatory potential is associated with weight gain in adult women and men. J. Nutr. 2019;149(5):804–815. doi: 10.1093/jn/nxy319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Malkov S., Shepherd J.A., Scott C.G., Tamimi R.M., Ma L., Bertrand K.A., et al. Mammographic texture and risk of breast cancer by tumor type and estrogen receptor status. Breast Cancer Res. 2016;18(1):122. doi: 10.1186/s13058-016-0778-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Manduca A., Carston M.J., Heine J.J., Scott C.G., Pankratz V.S., Brandt K.R., et al. Texture features from mammographic images and risk of breast cancer. Cancer Epidemiol. Biomarkers Prev. 2009;18(3):837–845. doi: 10.1158/1055-9965.EPI-08-0631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Nielsen M., Vachon C.M., Scott C.G., Chernoff K., Karemore G., Karssemeijer N., et al. Mammographic texture resemblance generalizes as an independent risk factor for breast cancer. Breast Cancer Res. 2014;16(2):R37. doi: 10.1186/bcr3641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wanders J.O.P., van Gils C.H., Karssemeijer N., Holland K., Kallenberg M., Peeters P.H.M., et al. The combined effect of mammographic texture and density on breast cancer risk: a cohort study. Breast Cancer Res. 2018;20(1):36. doi: 10.1186/s13058-018-0961-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Shieh Y., Scott C.G., Jensen M.R., Norman A.D., Bertrand K.A., Pankratz V.S., et al. Body mass index, mammographic density, and breast cancer risk by estrogen receptor subtype. Breast Cancer Res. 2019;21(1):48. doi: 10.1186/s13058-019-1129-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Pettersson A., Hankinson S.E., Willett W.C., Lagiou P., Trichopoulos D., Tamimi R.M. Nondense mammographic area and risk of breast cancer. Breast Cancer Res. 2011;13(5):R100. doi: 10.1186/bcr3041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hollenbeck S., Keely P., Seewaldt V. Mammographic density: intersection of science, the law, and clinical practice. Am. Soc. Clin. Oncol. Educ. Book. Ann. Meet. 2013;33(1) doi: 10.1200/EdBook_AM.2013.33.e63. [DOI] [PubMed] [Google Scholar]
- 19.Burton A., Maskarinec G., Perez-Gomez B., Vachon C., Miao H., Lajous M., et al. Mammographic density and ageing: a collaborative pooled analysis of cross-sectional data from 22 countries worldwide. PLOS Med. 2017;14(6) doi: 10.1371/journal.pmed.1002335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hart V., Reeves K.W., Sturgeon S.R., Reich N.G., Sievert L.L., Kerlikowske K., et al. The effect of change in body mass index on volumetric measures of mammographic density. Cancer Epidemiol. Biomarkers Prev. 2015;24(11):1724–1730. doi: 10.1158/1055-9965.EPI-15-0330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hankinson S.E., Willett W.C., Manson J.E., Hunter D.J., Colditz G.A., Stampfer M.J., et al. Alcohol, height, and adiposity in relation to estrogen and prolactin levels in postmenopausal women. J. Natl Cancer Inst. 1995;87(17):1297–1302. doi: 10.1093/jnci/87.17.1297. [DOI] [PubMed] [Google Scholar]
- 22.Colditz G.A., Hankinson S.E. The Nurses' Health Study: lifestyle and health among women. Nat. Rev. Cancer. 2005;5(5):388–396. doi: 10.1038/nrc1608. [DOI] [PubMed] [Google Scholar]
- 23.Tamimi R.M., Byrne C., Colditz G.A., Hankinson S.E. Endogenous hormone levels, mammographic density, and subsequent risk of breast cancer in postmenopausal women. J. Natl Cancer Inst. 2007;99(15):1178–1187. doi: 10.1093/jnci/djm062. [DOI] [PubMed] [Google Scholar]
- 24.Rice M.S., Rosner B.A., Tamimi R.M. Percent mammographic density prediction: development of a model in the nurses' health studies. Cancer Causes Control. 2017;28(7):677–684. doi: 10.1007/s10552-017-0898-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rice M.S., Bertrand K.A., VanderWeele T.J., Rosner B.A., Liao X., Adami H.O., et al. Mammographic density and breast cancer risk: a mediation analysis. Breast Cancer Res. 2016;18(1):94. doi: 10.1186/s13058-016-0750-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rosner B., Cook N., Portman R., Daniels S., Falkner B. Determination of blood pressure percentiles in normal-weight children: some methodological issues. Am. J. Epidemiol. 2008;167(6):653–666. doi: 10.1093/aje/kwm348. [DOI] [PubMed] [Google Scholar]
- 27.Boyd N., Martin L., Stone J., Little L., Minkin S., Yaffe M. A longitudinal study of the effects of menopause on mammographic features. Cancer Epidemiol. Biomarkers Prev. 2002;11(10 Pt 1):1048–1053. [PubMed] [Google Scholar]
- 28.Heine J.J., Scott C.G., Sellers T.A., Brandt K.R., Serie D.J., Wu F.F., et al. A novel automated mammographic density measure and breast cancer risk. J. Natl Cancer Inst. 2012;104(13):1028–1037. doi: 10.1093/jnci/djs254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Feskanich D., Rimm E.B., Giovannucci E.L., Colditz G.A., Stampfer M.J., Litin L.B., et al. Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J. Am. Diet. Assoc. 1993;93(7):790–796. doi: 10.1016/0002-8223(93)91754-e. [DOI] [PubMed] [Google Scholar]
- 30.Rimm E.B., Giovannucci E.L., Stampfer M.J., Colditz G.A., Litin L.B., Willett W.C. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. Am. J. Epidemiol. 1992;135(10):1114–1126. doi: 10.1093/oxfordjournals.aje.a116211. discussion 1127–1136. [DOI] [PubMed] [Google Scholar]
- 31.Willett W.C., Sampson L., Stampfer M.J., Rosner B., Bain C., Witschi J., et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am. J. Epidemiol. 1985;122(1):51–65. doi: 10.1093/oxfordjournals.aje.a114086. [DOI] [PubMed] [Google Scholar]
- 32.Tabung F.K., Smith-Warner S.A., Chavarro J.E., Fung T.T., Hu F.B., Willett W.C., et al. An empirical dietary inflammatory pattern score enhances prediction of circulating inflammatory biomarkers in adults. J. Nutr. 2017;147(8):1567–1577. doi: 10.3945/jn.117.248377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Asad S., Damicis A., Heng Y.J., Kananen K., Collier K.A., Adams E.J., et al. Association of body mass index and inflammatory dietary pattern with breast cancer pathologic and genomic immunophenotype in the nurses' health study. Breast Cancer Res. 2022;24(1):78. doi: 10.1186/s13058-022-01573-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ibrahim M.K., Wilechansky R.M., Challa P.K., Zhang X., Giovannucci E., Stampfer M., et al. The empirical dietary inflammatory pattern score and the risk of nonalcoholic fatty liver disease and cirrhosis. Hepatol. Commun. 2023;7(10) doi: 10.1097/HC9.0000000000000263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Sparks J.A., Barbhaiya M., Tedeschi S.K., Leatherwood C.L., Tabung F.K., Speyer C.B., et al. Inflammatory dietary pattern and risk of developing rheumatoid arthritis in women. Clin. Rheumatol. 2019;38(1):243–250. doi: 10.1007/s10067-018-4261-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Willett W.C., Howe G.R., Kushi L.H. Adjustment for total energy intake in epidemiologic studies. Am. J. Clin. Nutr. 1997;65(4 Suppl):1220S–1228S. doi: 10.1093/ajcn/65.4.1220S. discussion 1229S–1231S. [DOI] [PubMed] [Google Scholar]
- 37.Tabung F.K., Giovannucci E.L., Giulianini F., Liang L., Chandler P.D., Balasubramanian R., et al. An empirical dietary inflammatory pattern score is associated with circulating inflammatory biomarkers in a multi-ethnic population of postmenopausal women in the United States. J. Nutr. 2018;148(5):771–780. doi: 10.1093/jn/nxy031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zhang Z., Tabung F.K., Jin Q., Curran G., Irvin V.L., Shannon J., et al. Diet-driven inflammation and insulinemia and risk of interval breast cancer. Nutr. Cancer. 2022;74(9):3179–3193. doi: 10.1080/01635581.2022.2063350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Troy L.M., Hunter D.J., Manson J.E., Colditz G.A., Stampfer M.J., Willett W.C. The validity of recalled weight among younger women. Int. J. Obes. Relat. Metab. Disord. 1995;19(8):570–572. [PubMed] [Google Scholar]
- 40.Baglietto L., Krishnan K., Stone J., Apicella C., Southey M.C., English D.R., et al. Associations of mammographic dense and nondense areas and body mass index with risk of breast cancer. Am. J. Epidemiol. 2014;179(4):475–483. doi: 10.1093/aje/kwt260. [DOI] [PubMed] [Google Scholar]
- 41.Soguel L., Durocher F., Tchernof A., Diorio C. Adiposity, breast density, and breast cancer risk: epidemiological and biological considerations. Eur. J. Cancer Prev. 2017;26(6):511–520. doi: 10.1097/CEJ.0000000000000310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Stampfer M.J., Colditz G.A., Willett W.C., Manson J.E., Rosner B., Speizer F.E., et al. Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up from the nurses' health study. N. Engl. J. Med. 1991;325(11):756–762. doi: 10.1056/NEJM199109123251102. [DOI] [PubMed] [Google Scholar]
- 43.Tamimi R.M., Hankinson S.E., Chen W.Y., Rosner B., Colditz G.A. Combined estrogen and testosterone use and risk of breast cancer in postmenopausal women. Arch. Intern. Med. 2006;166(14):1483–1489. doi: 10.1001/archinte.166.14.1483. [DOI] [PubMed] [Google Scholar]
- 44.Baron R.M., Kenny D.A. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 1986;51(6):1173–1182. doi: 10.1037//0022-3514.51.6.1173. [DOI] [PubMed] [Google Scholar]
- 45.Mackinnon D.P., Lockwood C.M., Williams J. Confidence limits for the indirect effect: distribution of the product and resampling methods. Multivariate Behav. Res. 2004;39(1):99. doi: 10.1207/s15327906mbr3901_4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Tingley D., Yamamoto T., Hirose K., Keele L., Imai K. mediation: R Package for causal mediation analysis. J. Stat. Softw. 2014;59(5):1–38. [Google Scholar]
- 47.Vachon C.M., Kushi L.H., Cerhan J.R., Kuni C.C., Sellers T.A. Association of diet and mammographic breast density in the Minnesota breast cancer family cohort. Cancer Epidemiol. Biomarkers Prev. 2000;9(2):151–160. [PubMed] [Google Scholar]
- 48.Castelló A., Ascunce N., Salas-Trejo D., Vidal C., Sanchez-Contador C., Santamariña C., et al. Association between Western and Mediterranean dietary patterns and mammographic density. Obstet. Gynecol. 2016;128(3):574–581. doi: 10.1097/AOG.0000000000001589. [DOI] [PubMed] [Google Scholar]
- 49.Boyd N.F., Greenberg C., Lockwood G., Little L., Martin L., Byng J., et al. Effects at two years of a low-fat, high-carbohydrate diet on radiologic features of the breast: results from a randomized trial, Canadian Diet and Breast Cancer Prevention Study Group. J. Natl Cancer Inst. 1997;89(7):488–496. doi: 10.1093/jnci/89.7.488. [DOI] [PubMed] [Google Scholar]
- 50.Boyd N.F., Connelly P., Byng J., Yaffe M., Draper H., Little L., et al. Plasma lipids, lipoproteins, and mammographic densities. Cancer Epidemiol. Biomarkers Prev. 1995;4(7):727–733. [PubMed] [Google Scholar]
- 51.Soltani S., Moslehi N., Hosseini-Esfahani F., Vafa M. The association between empirical dietary inflammatory pattern and metabolic phenotypes in overweight/obese adults. Int. J. Endocrinol. Metab. 2018;16(2) doi: 10.5812/ijem.60048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Lee D.H., Li J., Li Y., Liu G., Wu K., Bhupathiraju S., et al. Dietary inflammatory and insulinemic potential and risk of type 2 diabetes: results from three prospective U.S. cohort studies. Diabetes Care. 2020;43(11):2675–2683. doi: 10.2337/dc20-0815. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Romanos-Nanclares A., Tabung F.K., Sinnott J.A., Trabert B., De Vivo I., Playdon M.C., et al. Inflammatory and insulinemic dietary patterns and risk of endometrial cancer among US women. J. Natl Cancer Inst. 2023;115(3):311–321. doi: 10.1093/jnci/djac229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Garzia N.A., Cushing-Haugen K., Kensler T.W., Tamimi R.M., Harris H.R. Adolescent and early adulthood inflammation-associated dietary patterns in relation to premenopausal mammographic density. Breast Cancer Res. 2021;23(1):71. doi: 10.1186/s13058-021-01449-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Bissell M.C.S., Kerlikowske K., Sprague B.L., Tice J.A., Gard C.C., Tossas K.Y., et al. Breast cancer population attributable risk proportions associated with body mass index and breast density by race/ethnicity and menopausal status. Cancer Epidemiol. Biomarkers Prev. 2020;29(10):2048–2056. doi: 10.1158/1055-9965.EPI-20-0358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Neuhouser M.L., Aragaki A.K., Prentice R.L., Manson J.E., Chlebowski R., Carty C.L., et al. Overweight, obesity, and postmenopausal invasive breast cancer risk: a secondary analysis of the women's health initiative randomized clinical trials. JAMA Oncol. 2015;1(5):611–621. doi: 10.1001/jamaoncol.2015.1546. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Tran T.X.M., Moon S.G., Kim S., Park B. Association of the interaction between mammographic breast density, body mass index, and menopausal status with breast cancer risk among Korean Women. JAMA Netw. Open. 2021;4(12) doi: 10.1001/jamanetworkopen.2021.39161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Premenopausal Breast Cancer Collaborative Group. Schoemaker M.J., Nichols H.B., Wright L.B., Brook M.N., Jones M.E. Association of body mass index and age with subsequent breast cancer risk in premenopausal women. JAMA Oncol. 2018;4(11) doi: 10.1001/jamaoncol.2018.1771. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Boyd N.F., Guo H., Martin L.J., Sun L., Stone J., Fishell E., et al. Mammographic density and the risk and detection of breast cancer. N. Engl. J. Med. 2007;356(3):227–236. doi: 10.1056/NEJMoa062790. [DOI] [PubMed] [Google Scholar]
- 60.Boyd N.F., Lockwood G.A., Byng J.W., Tritchler D.L., Yaffe M.J. Mammographic densities and breast cancer risk. Cancer Epidemiol. Biomarkers Prev. 1998;7(12):33–1144. [PubMed] [Google Scholar]
- 61.Destounis S., Arieno A., Morgan R., Roberts C., Chan A. Qualitative versus quantitative mammographic breast density assessment: applications for the US and abroad. Diagnostics (Basel) 2017;7(2):30. doi: 10.3390/diagnostics7020030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 62.Harvey J.A., Bovbjerg V.E. Quantitative assessment of mammographic breast density: relationship with breast cancer risk. Radiology. 2004;230(1):29–41. doi: 10.1148/radiol.2301020870. [DOI] [PubMed] [Google Scholar]
- 63.Azam S., Sjölander A., Eriksson M., Gabrielson M., Czene K., Hall P. Determinants of mammographic density change. JNCI Cancer Spectr. 2019;3(1):pkz004. doi: 10.1093/jncics/pkz004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Byrne C., Schairer C., Wolfe J., Parekh N., Salane M., Brinton L.A., et al. Mammographic features and breast cancer risk: effects with time, age, and menopause status. J. Natl Cancer Inst. 1995;87(21):1622–1629. doi: 10.1093/jnci/87.21.1622. [DOI] [PubMed] [Google Scholar]
- 65.Ahern T.P., Hankinson S.E., Willett W.C., Pollak M.N., Eliassen A.H., Tamimi RM. Plasma C-peptide, mammographic breast density, and risk of invasive breast cancer, Cancer Epidemiol. Biomarkers Prev. 2013;22(10):1786–1796. doi: 10.1158/1055-9965.EPI-13-0375. [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 data that support the findings of this study are available from Nurses’ Health Study, but restrictions apply to the availability of these data, which were used under license for this study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission.