Background: While obesity is well-understood to increase breast cancer risk, the role of the neighborhood obesogenic environment, encompassing social and built environment attributes that influence body size, is poorly understood.

Methods: Using principal components factor analysis, five composite factors [neighborhood socioeconomic status (nSES), urban, mixed-land development, unhealthy food environment, parks] on the basis of geospatial data were developed to characterize the obesogenic environment for 48,247 postmenopausal women in the Multiethnic Cohort, residing predominately in Los Angeles County. We used Cox proportional hazards regression to examine the association between neighborhood obesogenic factors and breast cancer risk (n = 2,341 cases after 17 years of follow-up), adjusting for body mass index (BMI), weight gain since age 21, education, established risk factors, other neighborhood factors, and clustering by block group.

Results: Lower nSES was associated with lower breast cancer risk [quintile 1 vs. 5: HR, 0.79; 95% confidence interval (CI), 0.66–0.95], with a more pronounced association observed in Latinos (quintile 1 vs. 5: HR, 0.60; 95% CI, 0.43–0.85). More urban environments were associated with lower breast cancer risk in Japanese Americans (quintile 5 vs. 1: HR, 0.49; 95% CI, 0.26–0.90), and lower mixed-land development was associated with higher breast cancer risk in Latinos (quintile 1 vs. 5: HR, 1.46; 95% CI, 1.10–1.93).

Conclusions: Obesogenic neighborhood environment factors, especially nSES, urbanicity, and mixed-land development, were differentially and independently associated with breast cancer risk in this multiethnic population.

Impact: These findings highlight the need for additional studies of the driving contextual aspects of nSES that influence breast cancer risk. Cancer Epidemiol Biomarkers Prev; 26(4); 480–9. ©2017 AACR.

See all the articles in this CEBP Focus section, “Geospatial Approaches to Cancer Control and Population Sciences.”

Excess adiposity, measured by body mass index (BMI) or weight gain, is among the few modifiable risk factors for postmenopausal breast cancer (1–3). In a recent meta-analysis of 7 prospective studies, each 5-kg gain in weight since early adulthood increased the risk of postmenopausal breast cancer by 11%, at least among women who did not use hormone therapy (3). Excess adiposity has been shown to impact breast cancer risk differentially across race/ethnicity (4). In the Multiethnic Cohort (MEC), both BMI and adult weight gain were associated with a significantly higher risk of breast cancer overall, but the risks for Native Hawaiians and Japanese Americans with BMI > 30 kg/m2 were higher than for white, African American, and Latino women of the same BMI (5). Overall, each 5 kg of weight gained since age 21 conferred an additional 4% increase in the risk of breast cancer, and weight gain was more important than baseline BMI in its association with breast cancer when modeled simultaneously (5).

A missing piece in our understanding of the relation between adiposity and the development of breast cancer, and potentially in our ability to address prevention, is the contribution of neighborhood obesogenic environments; specifically, the social and man-made (“built”) physical attributes of an individual's residential surroundings relevant to energy balance that promote weight gain and obesity (6). As examples, greater adiposity has been associated with lower neighborhood socioeconomic status (nSES; refs. 7–10), proximity and density of grocery stores and supermarkets (11–13), fast food restaurants (12–15), and walkability (e.g., street connectivity, density of walkable destinations; refs. 16, 17). Walkability and availability of recreational facilities have been further associated with higher levels of physical activity (18–20). Despite the relationship of neighborhood environment to adiposity (reviewed in refs. 21–24) and the established role of excess adiposity as an important risk factor for postmenopausal breast cancer (5), to our knowledge, there have been no studies of associations between potentially obesogenic neighborhood characteristics and breast cancer risk or whether neighborhood obesogenic factors explain associations between BMI or weight gain and breast cancer risk. To quantify the influence of such neighborhoods on breast cancer risk independent of individual-level risk factors, we evaluated the association of a broad suite of neighborhood obesogenic factors with subsequent development of invasive breast cancer while accounting for prediagnostic BMI, weight gain, and established risk factors for breast cancer. Specifically, we examined 48,247 postmenopausal women in the MEC from 4 racial/ethnic groups living in California, who were followed prospectively for up to 17 years. Neighborhood characteristics were examined as factors derived from principal component factor analysis. We hypothesize that obesogenic neighborhood factors reflecting lower nSES, more urban environments, more unhealthy foods, lower mixed-land development, and fewer parks will be associated with increased breast cancer risk.

Study subjects

The MEC is a large prospective cohort study designed to investigate the role of lifestyle and dietary factors in the etiology of cancer among a multiethnic population of U.S. adults. Study recruitment and methods have been described in detail previously (25). Briefly, from 1993 through 1996, the MEC enrolled 215,521 men and women aged 45–75 years from 5 self-reported racial/ethnic groups (African Americans, Japanese Americans, Latinos, Native Hawaiians, and whites) residing in Hawaii or California (primarily Los Angeles County); the present analysis is limited to the California component of the cohort. At cohort entry, participants completed a 26-page mailed questionnaire with questions pertaining to demographic characteristics, anthropometrics, medical history, family history, physical activity, diet, reproductive history, mammography, and occupation. Participants were followed prospectively for diagnosis with incident, invasive breast cancer through routine linkage with regional member registries of the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) Program. Tumor characteristics included SEER summary stage and hormone receptor status [estrogen receptor (ER) and progesterone receptor (PR)]. Information regarding tumor HER2 status was not available for all cases. Participants were followed for vital status through linkages to the National Death Index and death certificate files in Hawaii and California. Female MEC participants eligible for the present study (n = 57,933; Supplementary Fig. S1) belong to 1 of 4 racial/ethnic groups (African American, Japanese American, Latino, or white), completed a baseline questionnaire while living in California, had no invasive breast cancer diagnosis prior to cohort entry as reported on baseline questionnaire or through linkage with the tumor registry, and had plausible dietary data (5). We excluded those who were premenopausal or had unknown menopausal status (n = 7,084), had missing or extreme BMI (<15, >50 kg/m2; n = 1,555), had residential addresses that could not be geocoded (n = 1,045), or had a date of cohort entry that coincided with the date of diagnosis and thus had zero follow-up (n = 2), providing an analytic cohort of 48,247 women.

Address history and geocoding

The MEC actively maintains accurate and up-to-date addresses on all participants via periodic mailings of newsletters and follow-up questionnaires, as well as linkages with databases (e.g., U.S. Postal Service). Residential baseline addresses of MEC participants in California were geocoded to latitude and longitude coordinates using parcel data and then street centerline data for those that failed to geocode to a parcel. Geocodes of participants' addresses (at baseline, 1993–1996) were linked to 7,947 1990 U.S. Census block groups, each including an average of 1,475 residents in Los Angeles County or 9 MEC participants, representing our neighborhood unit.

Social and built environment data

We characterized the neighborhood obesogenic environment on the basis of the California Neighborhoods Data System, an integrated system of census, business, farmers market, park, traffic, and other small-area contextual data (26). We operationalized nSES as a composite measure by principal component analysis of census block group data (27) for education, housing, employment, occupation, income, and poverty, as described previously (28). Built environment characteristics included: (i) population density (persons per square kilometer) and (ii) percentage of residents who commute by car/motorcycle (per census block group) based on the census data (27); the number of (iii) businesses, (iv) recreational facilities, and (v) parks; unhealthy retail and restaurant environment, that is, (vi) restaurant environment index (REI), the ratio of the number of fast food restaurants to other restaurants, and (vii) the retail food environment index (RFEI), the ratio of the number of convenience stores, liquor stores, and fast food restaurants to supermarkets and farmers' markets (29) based on business listings from Walls & Associates' National Establishment Time-Series Database from 1990–2008 (30) and a 3-year business activity window (1990–1993), farmers' markets listings from the California Department of Food and Agriculture (31), and park listings from NavTeq's NavStreets database (32); (viii) traffic density from the California Department of Transportation (2000; refs. 33, 34); and (ix) street connectivity (35) based on the gamma index, defined as the ratio of actual number of street segments to maximum possible number of intersections, based on the data from NavTeq's NavStreets database (32).

To more fully characterize a neighborhood's obesogenic environment, we derived factors from principal components factor analysis (36, 37) of 9 built environment characteristics modeled as continuous variables and quantified for block groups using all California residents (Supplementary Table S1). All block groups in California in 1990 with at least 15 residents and non-missing data for the 9 built environment variables were included (n = 20,874 1990 California block groups). For REI and RFEI, we set block groups with no restaurants or retail food outlets to zero and those with only unhealthy outlets to 2. The number of factors (components) retained for orthogonal varimax rotation was based on an eigenvalue criterion >1, screen plot analysis, and interpretability. Factor scores were linear combinations of the standardized scoring coefficients multiplied by the standardized built environment variables as continuous variables, only including the neighborhood variables with a loading >0.50 (38, 39). We identified 4 neighborhood factors: an “urban environment” factor, characterized by high population density (rotated factor loading of 0.79), high street connectivity (0.70), low % commute by car/motorcycle (−0.68), and high traffic density (0.51); a “mixed-land development” factor, characterized by high recreational facilities per population (0.90), and high businesses per population (0.89); an “unhealthy food outlets” factor, characterized by a high ratio of unhealthy to healthy restaurants (0.82) and a high ratio of unhealthy to healthy retail establishments (0.81); and a “parks” factor, characterized by high number of parks per population (0.95). The “urban environment” and “mixed-land development” neighborhood factors were categorized into quintiles on the basis of the Los Angeles County distributions; the “unhealthy food outlets” factor was categorized as none, some, or more; and the “parks” factor was categorized as none versus any. Each MEC participant was assigned the 4 neighborhood obesogenic factors for her baseline residential block group. In a separate analysis, we examined the associations of these neighborhood obesogenic factors with baseline BMI and found that each factor was associated with an increased odds of obesity, although associations varied across race/ethnicity (40).

Health characteristics and breast cancer risk factors at baseline

Self-reported height and weight were used to calculate BMI (kg/m2), which was categorized as underweight (<18.5 kg/m2), normal weight (18.5–24.9 kg/m2), overweight (25–29.9 kg/m2), or obese (≥30 kg/m2). Adult weight gain from age 21 was calculated as the difference between baseline weight and weight at age 21 categorized into quintiles according to the study participant distribution. Smoking exposure was based on smoking status and pack-years of smoking (never, current, former, missing) and pack-years of smoking (<20, ≥20 pack-years, unknown) as a combined variable. Physical activity was estimated as the number of hours per day spent engaging in moderate or vigorous activities, on average, in the previous year. Other risk factors for breast cancer included first-degree family history of breast cancer (yes, no, missing), age at menarche (≤12, 13–14, >14 years, missing), age at first live birth (no children, <20, 21–30, >30 years, missing), number of children (0, 1, 2–3, ≥4, missing), history of mammography (yes, no, missing), alcohol intake (non-drinker, drinker, missing), educational attainment (high school graduate or less, some college, college graduate, graduate and professional school, missing), and the use of hormone therapy (no estrogen use, past estrogen use, current estrogen use only, current estrogen use with past or current progesterone use, missing).

Statistical analyses

Associations of neighborhood obesogenic factors, weight gain and BMI with breast cancer were examined using multivariable Cox proportional hazards regression with adjustment for clustering by block group and individual-level breast cancer risk factors. Age from cohort entry to breast cancer diagnosis, death, or end of study follow-up (December 31, 2010) was used as the time metric. Models of obesogenic neighborhood factors were adjusted for individual risk factors and neighborhood clusters, as well as for other neighborhood obesogenic factors. Additional analyses included models for hormone receptor–positive (ER+/PR+) or negative (ER/PR) and early-stage (I–III) or late-stage (IV) disease breast tumors. There was no violation of the proportional hazards assumption as evidenced by the Schoenfeld residuals. Neighborhood or individual-level factors significantly associated with breast cancer (P < 0.05) in age- and race/ethnicity-adjusted models were included in fully-adjusted models (see Tables 2 and 3 for adjustments). Reference categories for the neighborhood factors were chosen to reflect their positive associations with obesity (e.g., higher obesity with lower SES, more urban areas, more unhealthy food, lower mixed-land development, and fewer parks). Trend tests were performed by entering the variable as an ordinal parameter representing categories in the corresponding models. We also tested for spatial autocorrelation with Moran's I and found no evidence of it. As a sensitivity analysis, gamma frailty models were fit with block group as a random-effect (41). As the random effect term was not statistically significant and the confidence intervals (CI) for the neighborhood obesogenic factors did not change, we present the fixed-effects models. All P values presented are 2-sided. A P value threshold <0.05 was used to determine statistical significance. Cox proportional hazards regression analyses were conducted using SAS (version 9.3) and gamma frailty models were conducted using R (version 3.2.2).

Overall, the 48,247 women were diverse with respect to race/ethnicity, with 35% African American and 39% Latino (Table 1). The median age at the time of cohort entry was 62 years [interquartile range (IQR), 56–68]. The prevalence of excess adiposity (being overweight or obese as defined by BMI ≥ 25 kg/m2) at baseline varied substantially across race/ethnicity with a relatively higher proportion of overweight/obesity in African Americans (75%), Latinos (69%), and whites (53%) than among Japanese Americans (25%; Table 1). Similarly, a relatively higher proportion of African Americans (66%), Latinos (57%), and whites (50%) than Japanese Americans (23%) reported having gained at least 11.3 kg (25 pounds) since age 21. Compared with other racial/ethnic groups, Japanese Americans and whites were more likely to live in neighborhoods characterized by high SES (quintile 5) and more mixed-land development (quintile 5); conversely, African Americans and Latinos were more likely to live in more urban neighborhoods (quintile 5; Supplementary Table S2). During a median follow-up of 16.6 years (IQR, 14.5–17.3) from cohort entry, 2,341 women were diagnosed with invasive breast cancer; specifically, 1,542 (66%) ER+/PR+, 361 (15%) ER/PR, and 438 (19%) unknown hormone receptor status tumors. Most tumors (n = 1,564; 67%) were diagnosed at an early (localized) stage, whereas 741 (31%) were at late-stage (regional or distant) and 36 were unstaged (2%).

Table 1.

Distributions of breast cancer risk factors by race/ethnicity, postmenopausal women residing in California at baseline (n = 48,247), MEC, 1993–2010

All women (n = 48,247)aAfrican American (n = 16,703)aJapanese American (n = 5,063)aLatino (n = 18,752)aWhite (n = 7,729)a
n (%)n (%)n (%)n (%)n (%)
Age at baseline, y 
 45–49 2,992 (6.2) 1,339 (8.0) 199 (3.9) 1,062 (5.7) 392 (5.1) 
 50–54 5,972 (12.4) 2,152 (12.9) 492 (9.7) 2,490 (13.3) 838 (10.8) 
 55–59 10,133 (21.0) 2,848 (17.1) 880 (17.4) 4,806 (25.6) 1,599 (20.7) 
 60–64 10,365 (21.5) 2,762 (16.5) 1,031 (20.4) 4,851 (25.9) 1,721 (22.3) 
 65–69 9,458 (19.6) 3,703 (22.2) 1,061 (21.0) 3,124 (16.7) 1,570 (20.3) 
 70–74 8,038 (16.7) 3,350 (20.1) 1,151 (22.7) 2,142 (11.4) 1,395 (18.0) 
 75+ 1,289 (2.7) 549 (3.3) 249 (4.9) 277 (1.5) 214 (2.8) 
BMI at baseline, kg/m2 
 Underweight, <18.5 868 (1.8) 171 (1.0) 354 (7.0) 177 (0.9) 166 (2.1) 
 Normal, 18.5–24.9 16,612 (34.4) 4,047 (24.2) 3,437 (67.9) 5,662 (30.2) 3,466 (44.8) 
 Overweight, 25.0–29.9 17,378 (36.0) 6,289 (37.7) 1,056 (20.9) 7,557 (40.3) 2,476 (32.0) 
 Obese, ≥30.0 13,389 (27.8) 6,196 (37.1) 216 (4.3) 5,356 (28.6) 1,621 (21.0) 
Adult weight gain since age 21,b,c kg 
 Quintile 1: −130.5 to <5.4 8,732 (18.1) 1,623 (9.7) 2,354 (46.5) 2,808 (15.0) 1,947 (25.2) 
 Quintile 2: 5.4 to <11.3 8,104 (16.8) 1,985 (11.9) 1,349 (26.6) 3,254 (17.4) 1,516 (19.6) 
 Quintile 3: 11.3 to <17.6 9,318 (19.3) 3,018 (18.1) 775 (15.3) 4,008 (21.4) 1,517 (19.6) 
 Quintile 4: 17.6 to <24.8 8,678 (18.0) 3,522 (21.1) 286 (5.6) 3,673 (19.6) 1,197 (15.5) 
 Quintile 5: 25.2 to <90.0 8,804 (18.2) 4,515 (27.0) 86 (1.7) 3,037 (16.2) 1,166 (15.1) 
Education 
 High school graduate or less 26,022 (53.9) 7,057 (42.2) 1,977 (39.0) 13,770 (73.4) 3,218 (41.6) 
 Some college/technical school 13,511 (28.0) 5,904 (35.3) 1,838 (36.3) 3,258 (17.4) 2,511 (32.5) 
 College graduate 4,155 (8.6) 1,821 (10.9) 761 (15.0) 647 (3.5) 926 (12.0) 
 Graduated and professional school 3,855 (8.0) 1,713 (10.3) 442 (8.7) 710 (3.8) 990 (12.8) 
Smoking status 
 Never smoker 25,784 (53.4) 7,266 (43.5) 3,403 (67.2) 11,594 (61.8) 3,521 (45.6) 
 Former smoker 14,335 (29.7) 5,864 (35.1) 1,196 (23.6) 4,479 (23.9) 2,796 (36.2) 
 Current smoker 6,858 (14.2) 3,334 (20.0) 415 (8.2) 1,806 (9.6) 1,303 (16.9) 
Alcohol intake      
 Non-drinker 29,123 (60.4) 10,172 (60.9) 3,653 (72.2) 11,838 (63.1) 3,460 (44.8) 
 Drinker 17,062 (35.4) 5,831 (34.9) 1,215 (24.0) 6,140 (32.7) 3,876 (50.1) 
Moderate or vigorous physical activity,c h/d 
 0 3,840 (8.0) 972 (5.8) 119 (2.4) 2,436 (13.0) 313 (4.0) 
 Quartile 1: <0.4 7,953 (16.5) 2,871 (17.2) 707 (14.0) 3,507 (18.7) 868 (11.2) 
 Quartile 2: 0.4 to <0.7 11,901 (24.7) 4,845 (29.0) 1,335 (26.4) 4,044 (21.6) 1,677 (21.7) 
 Quartile 3: 0.7 to <1.2 10,589 (21.9) 3,721 (22.3) 1,246 (24.6) 3,730 (19.9) 1,892 (24.5) 
 Quartile 4: 1.2 to <13.3 12,153 (25.2) 3,591 (21.5) 1,575 (31.1) 4,143 (22.1) 2,844 (36.8) 
Energy,c kcal/d 
 Quintile 1: 417.4 to <1,152.8 9,237 (19.1) 3,933 (23.5) 905 (17.9) 2,999 (16.0) 1,400 (18.1) 
 Quintile 2: 1,152.8 to <1,533.2 9,237 (19.1) 3,186 (19.1) 1,227 (24.2) 3,044 (16.2) 1,780 (23.0) 
 Quintile 3: 1,533.2 to <1,947.0 9,237 (19.1) 3,036 (18.2) 1,224 (24.2) 3,299 (17.6) 1,678 (21.7) 
 Quintile 4: 1,947.0 to <2,606.9 9,237 (19.1) 2,959 (17.7) 1,003 (19.8) 3,745 (20.0) 1,530 (19.8) 
 Quintile 5: 2,606.9 to <7,211.3 9,237 (19.1) 2,889 (17.3) 509 (10.1) 4,891 (26.1) 948 (12.3) 
Ever had mammogram 
 Yes 5,192 (10.8) 1,293 (7.7) 578 (11.4) 2,596 (13.8) 725 (9.4) 
 No 41,384 (85.8) 14,784 (88.5) 4,356 (86.0) 15,371 (82.0) 6,873 (88.9) 
Family history of breast cancerd 
 Yes 4,975 (10.3) 1,003 (13.0) 1,832 (11.0) 516 (10.2) 1,624 (8.7) 
 No 39,380 (81.6) 6,269 (81.1) 13,498 (80.8) 4,286 (84.7) 15,327 (81.7) 
Age at first birth, y 
 Nulliparous 5,691 (11.8) 2,261 (13.5) 800 (15.8) 1,528 (8.1) 1,102 (14.3) 
 <20 17,466 (36.2) 7,442 (44.6) 354 (7.0) 7,468 (39.8) 2,202 (28.5) 
 21–30 21,009 (43.5) 5,706 (34.2) 3,288 (64.9) 8,262 (44.1) 3,753 (48.6) 
 >30 2,665 (5.5) 668 (4.0) 516 (10.2) 960 (5.1) 521 (6.7) 
Age at menarche, y 
 <12 22,666 (47.0) 7,988 (47.8) 2,251 (44.5) 8,625 (46.0) 3,802 (49.2) 
 13–14 18,382 (38.1) 6,211 (37.2) 2,029 (40.1) 7,147 (38.1) 2,995 (38.8) 
 >14 6,344 (13.1) 2,139 (12.8) 713 (14.1) 2,646 (14.1) 846 (10.9) 
Parity 
 Nulliparous 5,691 (11.8) 2,261 (13.5) 800 (15.8) 1,528 (8.1) 1,102 (14.3) 
 1 5,432 (11.3) 2,651 (15.9) 581 (11.5) 1,271 (6.8) 929 (12.0) 
 2–3 18,106 (37.5) 5,825 (34.9) 2,784 (55.0) 5,811 (31.0) 3,686 (47.7) 
 ≥4 18,342 (38.0) 5,735 (34.3) 868 (17.1) 9,805 (52.3) 1,934 (25.0) 
Menopause status 
 Natural menopause 26,584 (55.1) 7,611 (45.6) 3,354 (66.2) 11,121 (59.3) 4,498 (58.2) 
 Oophorectomy 7,685 (15.9) 3,198 (19.1) 718 (14.2) 2,395 (12.8) 1,374 (17.8) 
 Hysterectomy 9,441 (19.6) 4,153 (24.9) 552 (10.9) 3,402 (18.1) 1,334 (17.3) 
Hormone replacement therapy 
 Never estrogen use 23,367 (48.4) 8,752 (52.4) 2,385 (47.1) 9,303 (49.6) 2,927 (37.9) 
 Past estrogen use 9,519 (19.7) 3,672 (22.0) 737 (14.6) 3,562 (19.0) 1,548 (20.0) 
 Current unopposed estrogen use 6,312 (13.1) 2,037 (12.2) 768 (15.2) 2,177 (11.6) 1,330 (17.2) 
 Current estrogen–progesterone use 6,083 (12.6) 1,231 (7.4) 1,008 (19.9) 2,183 (11.6) 1,661 (21.5) 
All women (n = 48,247)aAfrican American (n = 16,703)aJapanese American (n = 5,063)aLatino (n = 18,752)aWhite (n = 7,729)a
n (%)n (%)n (%)n (%)n (%)
Age at baseline, y 
 45–49 2,992 (6.2) 1,339 (8.0) 199 (3.9) 1,062 (5.7) 392 (5.1) 
 50–54 5,972 (12.4) 2,152 (12.9) 492 (9.7) 2,490 (13.3) 838 (10.8) 
 55–59 10,133 (21.0) 2,848 (17.1) 880 (17.4) 4,806 (25.6) 1,599 (20.7) 
 60–64 10,365 (21.5) 2,762 (16.5) 1,031 (20.4) 4,851 (25.9) 1,721 (22.3) 
 65–69 9,458 (19.6) 3,703 (22.2) 1,061 (21.0) 3,124 (16.7) 1,570 (20.3) 
 70–74 8,038 (16.7) 3,350 (20.1) 1,151 (22.7) 2,142 (11.4) 1,395 (18.0) 
 75+ 1,289 (2.7) 549 (3.3) 249 (4.9) 277 (1.5) 214 (2.8) 
BMI at baseline, kg/m2 
 Underweight, <18.5 868 (1.8) 171 (1.0) 354 (7.0) 177 (0.9) 166 (2.1) 
 Normal, 18.5–24.9 16,612 (34.4) 4,047 (24.2) 3,437 (67.9) 5,662 (30.2) 3,466 (44.8) 
 Overweight, 25.0–29.9 17,378 (36.0) 6,289 (37.7) 1,056 (20.9) 7,557 (40.3) 2,476 (32.0) 
 Obese, ≥30.0 13,389 (27.8) 6,196 (37.1) 216 (4.3) 5,356 (28.6) 1,621 (21.0) 
Adult weight gain since age 21,b,c kg 
 Quintile 1: −130.5 to <5.4 8,732 (18.1) 1,623 (9.7) 2,354 (46.5) 2,808 (15.0) 1,947 (25.2) 
 Quintile 2: 5.4 to <11.3 8,104 (16.8) 1,985 (11.9) 1,349 (26.6) 3,254 (17.4) 1,516 (19.6) 
 Quintile 3: 11.3 to <17.6 9,318 (19.3) 3,018 (18.1) 775 (15.3) 4,008 (21.4) 1,517 (19.6) 
 Quintile 4: 17.6 to <24.8 8,678 (18.0) 3,522 (21.1) 286 (5.6) 3,673 (19.6) 1,197 (15.5) 
 Quintile 5: 25.2 to <90.0 8,804 (18.2) 4,515 (27.0) 86 (1.7) 3,037 (16.2) 1,166 (15.1) 
Education 
 High school graduate or less 26,022 (53.9) 7,057 (42.2) 1,977 (39.0) 13,770 (73.4) 3,218 (41.6) 
 Some college/technical school 13,511 (28.0) 5,904 (35.3) 1,838 (36.3) 3,258 (17.4) 2,511 (32.5) 
 College graduate 4,155 (8.6) 1,821 (10.9) 761 (15.0) 647 (3.5) 926 (12.0) 
 Graduated and professional school 3,855 (8.0) 1,713 (10.3) 442 (8.7) 710 (3.8) 990 (12.8) 
Smoking status 
 Never smoker 25,784 (53.4) 7,266 (43.5) 3,403 (67.2) 11,594 (61.8) 3,521 (45.6) 
 Former smoker 14,335 (29.7) 5,864 (35.1) 1,196 (23.6) 4,479 (23.9) 2,796 (36.2) 
 Current smoker 6,858 (14.2) 3,334 (20.0) 415 (8.2) 1,806 (9.6) 1,303 (16.9) 
Alcohol intake      
 Non-drinker 29,123 (60.4) 10,172 (60.9) 3,653 (72.2) 11,838 (63.1) 3,460 (44.8) 
 Drinker 17,062 (35.4) 5,831 (34.9) 1,215 (24.0) 6,140 (32.7) 3,876 (50.1) 
Moderate or vigorous physical activity,c h/d 
 0 3,840 (8.0) 972 (5.8) 119 (2.4) 2,436 (13.0) 313 (4.0) 
 Quartile 1: <0.4 7,953 (16.5) 2,871 (17.2) 707 (14.0) 3,507 (18.7) 868 (11.2) 
 Quartile 2: 0.4 to <0.7 11,901 (24.7) 4,845 (29.0) 1,335 (26.4) 4,044 (21.6) 1,677 (21.7) 
 Quartile 3: 0.7 to <1.2 10,589 (21.9) 3,721 (22.3) 1,246 (24.6) 3,730 (19.9) 1,892 (24.5) 
 Quartile 4: 1.2 to <13.3 12,153 (25.2) 3,591 (21.5) 1,575 (31.1) 4,143 (22.1) 2,844 (36.8) 
Energy,c kcal/d 
 Quintile 1: 417.4 to <1,152.8 9,237 (19.1) 3,933 (23.5) 905 (17.9) 2,999 (16.0) 1,400 (18.1) 
 Quintile 2: 1,152.8 to <1,533.2 9,237 (19.1) 3,186 (19.1) 1,227 (24.2) 3,044 (16.2) 1,780 (23.0) 
 Quintile 3: 1,533.2 to <1,947.0 9,237 (19.1) 3,036 (18.2) 1,224 (24.2) 3,299 (17.6) 1,678 (21.7) 
 Quintile 4: 1,947.0 to <2,606.9 9,237 (19.1) 2,959 (17.7) 1,003 (19.8) 3,745 (20.0) 1,530 (19.8) 
 Quintile 5: 2,606.9 to <7,211.3 9,237 (19.1) 2,889 (17.3) 509 (10.1) 4,891 (26.1) 948 (12.3) 
Ever had mammogram 
 Yes 5,192 (10.8) 1,293 (7.7) 578 (11.4) 2,596 (13.8) 725 (9.4) 
 No 41,384 (85.8) 14,784 (88.5) 4,356 (86.0) 15,371 (82.0) 6,873 (88.9) 
Family history of breast cancerd 
 Yes 4,975 (10.3) 1,003 (13.0) 1,832 (11.0) 516 (10.2) 1,624 (8.7) 
 No 39,380 (81.6) 6,269 (81.1) 13,498 (80.8) 4,286 (84.7) 15,327 (81.7) 
Age at first birth, y 
 Nulliparous 5,691 (11.8) 2,261 (13.5) 800 (15.8) 1,528 (8.1) 1,102 (14.3) 
 <20 17,466 (36.2) 7,442 (44.6) 354 (7.0) 7,468 (39.8) 2,202 (28.5) 
 21–30 21,009 (43.5) 5,706 (34.2) 3,288 (64.9) 8,262 (44.1) 3,753 (48.6) 
 >30 2,665 (5.5) 668 (4.0) 516 (10.2) 960 (5.1) 521 (6.7) 
Age at menarche, y 
 <12 22,666 (47.0) 7,988 (47.8) 2,251 (44.5) 8,625 (46.0) 3,802 (49.2) 
 13–14 18,382 (38.1) 6,211 (37.2) 2,029 (40.1) 7,147 (38.1) 2,995 (38.8) 
 >14 6,344 (13.1) 2,139 (12.8) 713 (14.1) 2,646 (14.1) 846 (10.9) 
Parity 
 Nulliparous 5,691 (11.8) 2,261 (13.5) 800 (15.8) 1,528 (8.1) 1,102 (14.3) 
 1 5,432 (11.3) 2,651 (15.9) 581 (11.5) 1,271 (6.8) 929 (12.0) 
 2–3 18,106 (37.5) 5,825 (34.9) 2,784 (55.0) 5,811 (31.0) 3,686 (47.7) 
 ≥4 18,342 (38.0) 5,735 (34.3) 868 (17.1) 9,805 (52.3) 1,934 (25.0) 
Menopause status 
 Natural menopause 26,584 (55.1) 7,611 (45.6) 3,354 (66.2) 11,121 (59.3) 4,498 (58.2) 
 Oophorectomy 7,685 (15.9) 3,198 (19.1) 718 (14.2) 2,395 (12.8) 1,374 (17.8) 
 Hysterectomy 9,441 (19.6) 4,153 (24.9) 552 (10.9) 3,402 (18.1) 1,334 (17.3) 
Hormone replacement therapy 
 Never estrogen use 23,367 (48.4) 8,752 (52.4) 2,385 (47.1) 9,303 (49.6) 2,927 (37.9) 
 Past estrogen use 9,519 (19.7) 3,672 (22.0) 737 (14.6) 3,562 (19.0) 1,548 (20.0) 
 Current unopposed estrogen use 6,312 (13.1) 2,037 (12.2) 768 (15.2) 2,177 (11.6) 1,330 (17.2) 
 Current estrogen–progesterone use 6,083 (12.6) 1,231 (7.4) 1,008 (19.9) 2,183 (11.6) 1,661 (21.5) 

aNs may not add to racial/ethnic totals and percentages may not add to 100 due to missing.

bDifference between baseline weight and weight at age 21.

cQuintiles or quartiles based on study population distribution.

dAt least one of mother of sisters had breast cancer.

Women living in the lowest versus highest SES neighborhoods had 21% lower risk of breast cancer (quintile 1 vs. 5: HR, 0.79; 95% CI, 0.66–0.95; Ptrend = 0.02), after adjustment for baseline BMI, weight gain, education, breast cancer risk factors, and other neighborhood obesogenic factors (Table 2). This nSES–breast cancer association was significant for Latinos (Ptrend = 0.01) and suggestive for whites (Ptrend = 0.19) but not evident for African Americans and Japanese Americans (Ptrends> 0.80). The nSES–breast cancer associations appeared somewhat stronger for hormone receptor–positive tumors, compared with all tumors. Risk of hormone receptor–positive breast cancer among women living in the lowest versus the highest SES neighborhoods was 27% lower (quintile 1 vs. 5: HR, 0.73; 95% CI, 0.58–0.92). Overall associations were fairly similar with and without adjustment for weight gain and BMI.

Table 2.

Associations of neighborhood SES, neighborhood obesogenic factors, and risk of invasive breast cancer and hormone receptor–positive subtype by race/ethnicity, postmenopausal women residing in California at baseline (n = 48,247), MEC, 1993–2010

All women(n = 48,247)African American(n = 16,703)Japanese American (n = 5,063)Latino (n = 18,752)White (n = 7,729)
Neighborhood measureHRa (95% Cl)HRa (95% Cl)HRa (95% Cl)HRa (95% Cl)HRa (95%Cl)
Overall 
 Neighborhood SESb 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 0.92 (0.79–1.07) 0.88 (0.61–1.27) 0.88 (0.63–1.21) 0.95 (0.71–1.27) 1.01 (0.80–1.28) 
  Quintile 3 0.88 (0.76–1.03) 1.12 (0.79–1.58) 0.88 (0.63–1.25) 0.78 (0.59–1.04) 0.81 (0.61–1.07) 
  Quintile 2 0.89 (0.76–1.05) 1.03 (0.73–1.46) 0.73 (0.44–1.20) 0.83 (0.62–1.11) 0.89 (0.63–1.25) 
  Quintile 1: low 0.79 (0.66–0.95) 0.94 (0.65–1.36) 1.73 (0.86–3.48) 0.60 (0.43–0.85) 0.80 (0.50–1.29) 
  Ptrendc 0.02 0.95 0.83 0.01 0.19 
 Urban environmentd 
  Quintile 1: low 1.00 1.00 1.00 1.00 1.00 
  Quintile 2 1.08 (0.95–1.23) 1.03 (0.80–1.34) 0.96 (0.70–1.34) 1.03 (0.82–1.28) 1.36 (1.07–1.73) 
  Quintile 3 1.03 (0.89–1.18) 1.13 (0.88–1.47) 0.95 (0.65–1.39) 0.97 (0.76–1.24) 1.07 (0.8–1.44) 
  Quintile 4 1.00 (0.86–1.15) 1.06 (0.82–1.37) 0.66 (0.41–1.08) 0.94 (0.72–1.22) 1.24 (0.91–1.69) 
  Quintile 5: high 1.02 (0.87–1.20) 1.10 (0.84–1.45) 0.49 (0.26–0.90) 1.07 (0.80–1.42) 1.13 (0.76–1.70) 
  Ptrendc 0.87 0.56 0.03 0.98 0.45 
 Unhealthy foode 
  None 1.00 1.00 1.00 1.00 1.00 
  Some 0.96 (0.84–1.09) 1.03 (0.82–1.28) 0.85 (0.58–1.26) 0.93 (0.74–1.17) 0.93 (0.69–1.25) 
  More 1.10 (1.00–1.21) 1.15 (0.97–1.35) 1.08 (0.82–1.42) 1.06 (0.89–1.27) 1.04 (0.84–1.29) 
  Ptrendc 0.07 0.11 0.65 0.55 0.79 
 Mixed-land developmentf 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 1.07 (0.92–1.25) 1.16 (0.84–1.59) 1.08 (0.76–1.56) 1.38 (1.04–1.83) 0.81 (0.61–1.09) 
  Quintile 3 1.02 (0.88–1.19) 1.16 (0.86–1.56) 1.00 (0.66–1.51) 1.29 (0.96–1.73) 0.76 (0.57–1.01) 
  Quintile 2 1.02 (0.88–1.18) 1.19 (0.89–1.58) 0.96 (0.66–1.41) 1.15 (0.86–1.53) 0.88 (0.67–1.14) 
  Quintile 1: low 1.11 (0.95–1.29) 1.24 (0.93–1.66) 0.78 (0.5–1.22) 1.46 (1.10–1.93) 0.87 (0.64–1.18) 
  Ptrendc 0.36 0.18 0.24 0.10 0.49 
 Parkg 
  Some 1.00 1.00 1.00 1.00 1.00 
  None 1.03 (0.92–1.15) 0.93 (0.75–1.15) 1.05 (0.78–1.42) 1.00 (0.83–1.21) 1.12 (0.88–1.44) 
  Ptrendc 0.66 0.52 0.76 0.97 0.36 
ER+/PR+ subtype 
 Neighborhood SESb 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 0.89 (0.74–1.06) 0.85 (0.53–1.38) 0.82 (0.55–1.21) 0.87 (0.61–1.25) 0.99 (0.76–1.29) 
  Quintile 3 0.82 (0.68–0.98) 1.05 (0.65–1.68) 0.84 (0.57–1.26) 0.65 (0.45–0.93) 0.75 (0.54–1.03) 
  Quintile 2 0.81 (0.67–0.99) 0.99 (0.62–1.59) 0.44 (0.23–0.86) 0.68 (0.47–0.99) 0.88 (0.59–1.30) 
  Quintile 1: low 0.73 (0.58–0.92) 0.92 (0.56–1.53) 1.20 (0.54–2.66) 0.52 (0.34–0.80) 0.77 (0.45–1.31) 
  Ptrendc 0.01 0.97 0.20 <0.01 0.18 
 Urband 
  Quintile 1: low 1.00 1.00 1.00 1.00 1.00 
  Quintile 2 1.20 (1.03–1.39) 1.11 (0.81–1.54) 1.20 (0.83–1.73) 1.15 (0.88–1.51) 1.43 (1.09–1.88) 
  Quintile 3 1.12 (0.95–1.33) 1.26 (0.92–1.75) 0.97 (0.60–1.56) 1.12 (0.84–1.50) 1.20 (0.86–1.69) 
  Quintile 4 1.04 (0.86–1.24) 1.19 (0.85–1.65) 0.82 (0.47–1.42) 0.86 (0.61–1.20) 1.32 (0.91–1.91) 
  Quintile 5: high 1.06 (0.87–1.29) 1.27 (0.91–1.79) 0.55 (0.27–1.09) 1.04 (0.73–1.49) 1.09 (0.69–1.72) 
  Ptrendc 0.96 0.22 0.14 0.66 0.41 
 Unhealthy foode 
  None 1.00 1.00 1.00 1.00 1.00 
  Some 0.95 (0.81–1.12) 1.02 (0.77–1.34) 0.94 (0.60–1.47) 0.96 (0.73–1.28) 0.86 (0.60–1.23) 
  More 1.09 (0.97–1.23) 1.05 (0.86–1.3) 1.14 (0.82–1.59) 1.07 (0.86–1.34) 1.08 (0.84–1.38) 
  Ptrendc 0.19 0.63 0.47 0.57 0.61 
 Mixed-land developmentf 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 1.07 (0.90–1.27) 1.18 (0.82–1.71) 1.13 (0.72–1.76) 1.50 (1.04–2.16) 0.75 (0.55–1.03) 
  Quintile 3 1.00 (0.83–1.20) 1.07 (0.75–1.5) 1.18 (0.71–1.96) 1.37 (0.94–2.00) 0.67 (0.48–0.94) 
  Quintile 2 0.98 (0.82–1.17) 0.96 (0.68–1.35) 1.12 (0.70–1.78) 1.29 (0.89–1.87) 0.84 (0.62–1.15) 
  Quintile 1: low 1.04 (0.87–1.25) 1.03 (0.74–1.44) 0.95 (0.56–1.61) 1.49 (1.02–2.16) 0.85 (0.60–1.19) 
  Ptrendc 0.97 0.57 0.94 0.20 0.47 
 Parkg 
  Some 1.00 1.00 1.00 1.00 1.00 
  None 1.06 (0.93–1.22) 1.00 (0.77–1.3) 1.14 (0.79–1.64) 1.07 (0.84–1.36) 1.10 (0.82–1.47) 
  Ptrendc 0.39 1.00 0.50 0.57 0.55 
All women(n = 48,247)African American(n = 16,703)Japanese American (n = 5,063)Latino (n = 18,752)White (n = 7,729)
Neighborhood measureHRa (95% Cl)HRa (95% Cl)HRa (95% Cl)HRa (95% Cl)HRa (95%Cl)
Overall 
 Neighborhood SESb 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 0.92 (0.79–1.07) 0.88 (0.61–1.27) 0.88 (0.63–1.21) 0.95 (0.71–1.27) 1.01 (0.80–1.28) 
  Quintile 3 0.88 (0.76–1.03) 1.12 (0.79–1.58) 0.88 (0.63–1.25) 0.78 (0.59–1.04) 0.81 (0.61–1.07) 
  Quintile 2 0.89 (0.76–1.05) 1.03 (0.73–1.46) 0.73 (0.44–1.20) 0.83 (0.62–1.11) 0.89 (0.63–1.25) 
  Quintile 1: low 0.79 (0.66–0.95) 0.94 (0.65–1.36) 1.73 (0.86–3.48) 0.60 (0.43–0.85) 0.80 (0.50–1.29) 
  Ptrendc 0.02 0.95 0.83 0.01 0.19 
 Urban environmentd 
  Quintile 1: low 1.00 1.00 1.00 1.00 1.00 
  Quintile 2 1.08 (0.95–1.23) 1.03 (0.80–1.34) 0.96 (0.70–1.34) 1.03 (0.82–1.28) 1.36 (1.07–1.73) 
  Quintile 3 1.03 (0.89–1.18) 1.13 (0.88–1.47) 0.95 (0.65–1.39) 0.97 (0.76–1.24) 1.07 (0.8–1.44) 
  Quintile 4 1.00 (0.86–1.15) 1.06 (0.82–1.37) 0.66 (0.41–1.08) 0.94 (0.72–1.22) 1.24 (0.91–1.69) 
  Quintile 5: high 1.02 (0.87–1.20) 1.10 (0.84–1.45) 0.49 (0.26–0.90) 1.07 (0.80–1.42) 1.13 (0.76–1.70) 
  Ptrendc 0.87 0.56 0.03 0.98 0.45 
 Unhealthy foode 
  None 1.00 1.00 1.00 1.00 1.00 
  Some 0.96 (0.84–1.09) 1.03 (0.82–1.28) 0.85 (0.58–1.26) 0.93 (0.74–1.17) 0.93 (0.69–1.25) 
  More 1.10 (1.00–1.21) 1.15 (0.97–1.35) 1.08 (0.82–1.42) 1.06 (0.89–1.27) 1.04 (0.84–1.29) 
  Ptrendc 0.07 0.11 0.65 0.55 0.79 
 Mixed-land developmentf 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 1.07 (0.92–1.25) 1.16 (0.84–1.59) 1.08 (0.76–1.56) 1.38 (1.04–1.83) 0.81 (0.61–1.09) 
  Quintile 3 1.02 (0.88–1.19) 1.16 (0.86–1.56) 1.00 (0.66–1.51) 1.29 (0.96–1.73) 0.76 (0.57–1.01) 
  Quintile 2 1.02 (0.88–1.18) 1.19 (0.89–1.58) 0.96 (0.66–1.41) 1.15 (0.86–1.53) 0.88 (0.67–1.14) 
  Quintile 1: low 1.11 (0.95–1.29) 1.24 (0.93–1.66) 0.78 (0.5–1.22) 1.46 (1.10–1.93) 0.87 (0.64–1.18) 
  Ptrendc 0.36 0.18 0.24 0.10 0.49 
 Parkg 
  Some 1.00 1.00 1.00 1.00 1.00 
  None 1.03 (0.92–1.15) 0.93 (0.75–1.15) 1.05 (0.78–1.42) 1.00 (0.83–1.21) 1.12 (0.88–1.44) 
  Ptrendc 0.66 0.52 0.76 0.97 0.36 
ER+/PR+ subtype 
 Neighborhood SESb 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 0.89 (0.74–1.06) 0.85 (0.53–1.38) 0.82 (0.55–1.21) 0.87 (0.61–1.25) 0.99 (0.76–1.29) 
  Quintile 3 0.82 (0.68–0.98) 1.05 (0.65–1.68) 0.84 (0.57–1.26) 0.65 (0.45–0.93) 0.75 (0.54–1.03) 
  Quintile 2 0.81 (0.67–0.99) 0.99 (0.62–1.59) 0.44 (0.23–0.86) 0.68 (0.47–0.99) 0.88 (0.59–1.30) 
  Quintile 1: low 0.73 (0.58–0.92) 0.92 (0.56–1.53) 1.20 (0.54–2.66) 0.52 (0.34–0.80) 0.77 (0.45–1.31) 
  Ptrendc 0.01 0.97 0.20 <0.01 0.18 
 Urband 
  Quintile 1: low 1.00 1.00 1.00 1.00 1.00 
  Quintile 2 1.20 (1.03–1.39) 1.11 (0.81–1.54) 1.20 (0.83–1.73) 1.15 (0.88–1.51) 1.43 (1.09–1.88) 
  Quintile 3 1.12 (0.95–1.33) 1.26 (0.92–1.75) 0.97 (0.60–1.56) 1.12 (0.84–1.50) 1.20 (0.86–1.69) 
  Quintile 4 1.04 (0.86–1.24) 1.19 (0.85–1.65) 0.82 (0.47–1.42) 0.86 (0.61–1.20) 1.32 (0.91–1.91) 
  Quintile 5: high 1.06 (0.87–1.29) 1.27 (0.91–1.79) 0.55 (0.27–1.09) 1.04 (0.73–1.49) 1.09 (0.69–1.72) 
  Ptrendc 0.96 0.22 0.14 0.66 0.41 
 Unhealthy foode 
  None 1.00 1.00 1.00 1.00 1.00 
  Some 0.95 (0.81–1.12) 1.02 (0.77–1.34) 0.94 (0.60–1.47) 0.96 (0.73–1.28) 0.86 (0.60–1.23) 
  More 1.09 (0.97–1.23) 1.05 (0.86–1.3) 1.14 (0.82–1.59) 1.07 (0.86–1.34) 1.08 (0.84–1.38) 
  Ptrendc 0.19 0.63 0.47 0.57 0.61 
 Mixed-land developmentf 
  Quintile 5: high 1.00 1.00 1.00 1.00 1.00 
  Quintile 4 1.07 (0.90–1.27) 1.18 (0.82–1.71) 1.13 (0.72–1.76) 1.50 (1.04–2.16) 0.75 (0.55–1.03) 
  Quintile 3 1.00 (0.83–1.20) 1.07 (0.75–1.5) 1.18 (0.71–1.96) 1.37 (0.94–2.00) 0.67 (0.48–0.94) 
  Quintile 2 0.98 (0.82–1.17) 0.96 (0.68–1.35) 1.12 (0.70–1.78) 1.29 (0.89–1.87) 0.84 (0.62–1.15) 
  Quintile 1: low 1.04 (0.87–1.25) 1.03 (0.74–1.44) 0.95 (0.56–1.61) 1.49 (1.02–2.16) 0.85 (0.60–1.19) 
  Ptrendc 0.97 0.57 0.94 0.20 0.47 
 Parkg 
  Some 1.00 1.00 1.00 1.00 1.00 
  None 1.06 (0.93–1.22) 1.00 (0.77–1.3) 1.14 (0.79–1.64) 1.07 (0.84–1.36) 1.10 (0.82–1.47) 
  Ptrendc 0.39 1.00 0.50 0.57 0.55 

NOTE: Values in bold represent P < 0.05.

aModels [invasive breast cancer overall and hormone receptor–positive (ER+/PR+) subtype] adjusted for age at entry, clustering effect of block group, race/ethnicity for all women, risk factors (family history of breast cancer, age at menarche, age at first live birth, number of children, hormone replacement therapy, ever had mammography, alcohol use, physical activity, education), baseline BMI and adult weight change, neighborhood SES, and all neighborhood obesogenic factors in table.

bComposite measure of 7 indicator variables for U.S. census block groups (Liu education index, proportion with a blue collar job, proportion older than age 16 in the workforce without a job, median household income, percent below 200% of the poverty line, median rent, median house value); categories based on distribution for block groups in Los Angeles County.

cTrend analysis excluding underweight and missing values. Trend variable was assigned ordinal value (i.e., 1, 2, 3) corresponding to categories.

dComposite measure of high population density, low commute by car, high traffic density, high street connectivity.

eComposite measure of high number of fast food restaurants and unhealthy retail food outlets. Unhealthy food environment index was categorized into none (no fast food restaurant and no unhealthy food outlet/minimal factor index value), some (factor index value greater than minimal and less than the median of non-minimal values), more (factor index value greater than or equal to the median of non-minimal values).

fComposite measure of high number of businesses and high number of recreational facilities.

gNumber of parks categorized into none (no park/minimal factor index value) and any (≥1 parks/factor index value greater than minimal).

The nSES–breast cancer association was slightly attenuated with sequential modeling of individual risk factors (data not shown). For example among Latinos, the estimated minimally adjusted (age, race/ethnicity, clustering by block group) HR for the lowest versus highest nSES quintile of 0.55 (95% CI, 0.41–0.73) was attenuated 13% (HR, 0.62; 95% CI, 0.47–0.83) with adjustment for reproductive factors and hormone replacement therapy, an additional 3% (HR, 0.64; 95% CI, 0.48–0.85) with further adjustment for education, and an additional 2% (HR, 0.65; 95% CI, 0.49–0.88) with further adjustment for mammography screening, family history of breast cancer, alcohol intake, physical activity, energy intake, and smoking. Conversely, further adjustment for baseline BMI and weight change since age 21 slightly strengthened the association (HR, 0.64; 95% CI, 0.48–0.86), reflecting the positive association between adult weight change and breast cancer risk. Additional adjustment for the other neighborhood obesogenic factors attenuated the HR by 6% (HR, 0.60; 95% CI, 0.43–0.85).

Neighborhood obesogenic factors for unhealthy food outlets, mixed-land development, and urban environments were differentially associated with breast cancer risk in this multiethnic population, after adjustment for individual and other neighborhood obesogenic factors (Table 2). Women living in neighborhoods with more unhealthy foods outlets versus none had 10% higher breast cancer risk (HR, 1.10; 95% CI, 1.00–1.21; Ptrend = 0.07). Living in neighborhoods characterized by the lowest versus highest quintile of mixed-land development was associated with higher breast cancer risk among Latinos (quintile 1 vs. 5: HR, 1.46; 95% CI, 1.10- 1.93; Ptrend = 0.10). In contrast, Japanese Americans living in more urban environments had significantly lower breast cancer risk (quintile 1 vs. 5: HR, 0.49; 95% CI, 0.26–0.90; Ptrend = 0.03). Overall, associations were attenuated in analyses limited to hormone receptor–positive tumors, except for nSES- and mixed-land development–breast cancer associations among Latinos, with evidence of a threshold effect of 50% higher risk for mixed-land development (quintile 4 vs. 5: HR, 1.50; 95% CI, 1.04–2.16; quintile 1 vs. 5: HR, 1.49; 95% CI, 1.02–2.16). Associations (overall and by tumor subtype) were fairly similar with and without adjustment for weight gain and BMI.

The associations between weight gain since age 21 and breast cancer risk persisted with adjustment for neighborhood obesogenic factors (Table 3). For African Americans and whites, who reported gaining at least 25 versus <5 kg since age 21, the risk of breast cancer was 85% and 60% higher, respectively, with adjustment for age, BMI, education, and other established risk factors (HR, 1.85; 95% CI, 1.30–2.61; Ptrend < 0.01; and HR, 1.60; 95% CI, 1.06–2.41; Ptrend = 0.08, respectively). For Japanese Americans and Latinos, associations of weight gain with breast cancer trended in the same direction. Further adjustment for nSES and neighborhood obesogenic factors did not strengthen nor attenuate these associations. Associations were generally higher for hormone receptor–positive than -negative subtypes while adjusting for individual and neighborhood factors (i.e., weight gain of >25 vs. <5 kg among all women: HR, 1.76; 95% CI, 1.40–2.22; Ptrend< 0.01 for ER+/PR+; and HR, 1.42; 95% CI, 0.85–2.36; Ptrend = 0.12 for ER−/PR), and these associations also persisted after further adjustment for neighborhood obesogenic factors (data not shown).

Table 3.

Associations of body size and adult weight gain and risk of invasive breast cancer and hormone receptor subtypes by race/ethnicity, postmenopausal women residing in California at baseline (n = 48,247), MEC, 1993–2010

All women(n = 48,247)African American (n = 16,703)Japanese American (n = 5,063)Latino (n = 18,752)White (n = 7,729)
HR (95% Cl)HR (95% Cl)HR (95% Cl)HR (95% Cl)HR (95% Cl)
Independent of individual-level factorsa 
 BMI at baseline, kg/m2 
  Underweight 0.83 (0.57–1.21) 1.00 (0.44–2.27) 0.61 (0.31–1.21) 1.19 (0.53–2.69) 1.00 (0.50–1.98) 
  Normal 1.00 1.00 1.00 1.00 1.00 
  Overweight 1.01 (0.89–1.14) 0.85 (0.69–1.05) 1.28 (0.86–1.90) 1.11 (0.90–1.38) 0.93 (0.71–1.22) 
  Obese 0.98 (0.84–1.14) 0.83 (0.65–1.05) 0.97 (0.41–2.34) 1.20 (0.93–1.55) 0.84 (0.57–1.22) 
  Ptrendb 0.80 0.15 0.52 0.17 0.36 
 Adult weight gain since age 21,c kg 
  −130.5 to <5.4 1.00 1.00 1.00 1.00 1.00 
  5.4 to <11.3 1.07 (0.92–1.25) 1.23 (0.87–1.74) 1.04 (0.76–1.41) 0.99 (0.75–1.31) 1.02 (0.77–1.34) 
  11.3 to <17.6 1.17 (1.00–1.37) 1.22 (0.87–1.71) 1.34 (0.87–2.06) 1.18 (0.88–1.56) 0.97 (0.71–1.33) 
  17.6 to <24.8 1.35 (1.13–1.61) 1.56 (1.11–2.17) 1.76 (0.96–3.26) 1.11 (0.81–1.52) 1.44 (1.01–2.04) 
  25.2 to <90.0 1.56 (1.29–1.88) 1.85 (1.30–2.61) 1.51 (0.51–4.44) 1.32 (0.94–1.83) 1.60 (1.06–2.41) 
  Ptrendb <0.01 <0.01 0.08 0.05 0.08 
Independent of individual-level + neighborhood-level factorsd 
 BMI at baseline, kg/m2 
  Underweight 0.83 (0.57–1.21) 0.99 (0.43–2.24) 0.61 (0.31–1.2) 1.17 (0.52–2.65) 0.98 (0.49–1.95) 
  Normal 1.00 1.00 1.00 1.00 1.00 
  Overweight 1.01 (0.89–1.15) 0.85 (0.69–1.04) 1.30 (0.88–1.94) 1.13 (0.91–1.41) 0.94 (0.72–1.23) 
  Obese 0.99 (0.85–1.16) 0.83 (0.65–1.05) 1.02 (0.42–2.47) 1.23 (0.96–1.59) 0.84 (0.57–1.22) 
  Ptrendb 0.91 0.15 0.46 0.12 0.37 
 Adult weight gain since age 21,c kg 
  −130.5 to <5.4 1.00 1.00 1.00 1.00 1.00 
  5.4 to <11.3 1.07 (0.92–1.25) 1.24 (0.88–1.74) 1.02 (0.75–1.4) 1.00 (0.75–1.32) 1.00 (0.76–1.32) 
  11.3 to <17.6 1.17 (0.99–1.37) 1.23 (0.88–1.72) 1.32 (0.86–2.03) 1.18 (0.88–1.56) 0.97 (0.71–1.33) 
  17.6 to <24.8 1.34 (1.13–1.60) 1.55 (1.11–2.16) 1.79 (0.96–3.32) 1.10 (0.81–1.51) 1.41 (0.99–2.00) 
  25.2 to <90.0 1.56 (1.29–1.88) 1.84 (1.30–2.60) 1.59 (0.53–4.72) 1.30 (0.94–1.82) 1.58 (1.05–2.39) 
  Ptrendb <0.01 <0.01 0.08 0.05 0.09 
All women(n = 48,247)African American (n = 16,703)Japanese American (n = 5,063)Latino (n = 18,752)White (n = 7,729)
HR (95% Cl)HR (95% Cl)HR (95% Cl)HR (95% Cl)HR (95% Cl)
Independent of individual-level factorsa 
 BMI at baseline, kg/m2 
  Underweight 0.83 (0.57–1.21) 1.00 (0.44–2.27) 0.61 (0.31–1.21) 1.19 (0.53–2.69) 1.00 (0.50–1.98) 
  Normal 1.00 1.00 1.00 1.00 1.00 
  Overweight 1.01 (0.89–1.14) 0.85 (0.69–1.05) 1.28 (0.86–1.90) 1.11 (0.90–1.38) 0.93 (0.71–1.22) 
  Obese 0.98 (0.84–1.14) 0.83 (0.65–1.05) 0.97 (0.41–2.34) 1.20 (0.93–1.55) 0.84 (0.57–1.22) 
  Ptrendb 0.80 0.15 0.52 0.17 0.36 
 Adult weight gain since age 21,c kg 
  −130.5 to <5.4 1.00 1.00 1.00 1.00 1.00 
  5.4 to <11.3 1.07 (0.92–1.25) 1.23 (0.87–1.74) 1.04 (0.76–1.41) 0.99 (0.75–1.31) 1.02 (0.77–1.34) 
  11.3 to <17.6 1.17 (1.00–1.37) 1.22 (0.87–1.71) 1.34 (0.87–2.06) 1.18 (0.88–1.56) 0.97 (0.71–1.33) 
  17.6 to <24.8 1.35 (1.13–1.61) 1.56 (1.11–2.17) 1.76 (0.96–3.26) 1.11 (0.81–1.52) 1.44 (1.01–2.04) 
  25.2 to <90.0 1.56 (1.29–1.88) 1.85 (1.30–2.61) 1.51 (0.51–4.44) 1.32 (0.94–1.83) 1.60 (1.06–2.41) 
  Ptrendb <0.01 <0.01 0.08 0.05 0.08 
Independent of individual-level + neighborhood-level factorsd 
 BMI at baseline, kg/m2 
  Underweight 0.83 (0.57–1.21) 0.99 (0.43–2.24) 0.61 (0.31–1.2) 1.17 (0.52–2.65) 0.98 (0.49–1.95) 
  Normal 1.00 1.00 1.00 1.00 1.00 
  Overweight 1.01 (0.89–1.15) 0.85 (0.69–1.04) 1.30 (0.88–1.94) 1.13 (0.91–1.41) 0.94 (0.72–1.23) 
  Obese 0.99 (0.85–1.16) 0.83 (0.65–1.05) 1.02 (0.42–2.47) 1.23 (0.96–1.59) 0.84 (0.57–1.22) 
  Ptrendb 0.91 0.15 0.46 0.12 0.37 
 Adult weight gain since age 21,c kg 
  −130.5 to <5.4 1.00 1.00 1.00 1.00 1.00 
  5.4 to <11.3 1.07 (0.92–1.25) 1.24 (0.88–1.74) 1.02 (0.75–1.4) 1.00 (0.75–1.32) 1.00 (0.76–1.32) 
  11.3 to <17.6 1.17 (0.99–1.37) 1.23 (0.88–1.72) 1.32 (0.86–2.03) 1.18 (0.88–1.56) 0.97 (0.71–1.33) 
  17.6 to <24.8 1.34 (1.13–1.60) 1.55 (1.11–2.16) 1.79 (0.96–3.32) 1.10 (0.81–1.51) 1.41 (0.99–2.00) 
  25.2 to <90.0 1.56 (1.29–1.88) 1.84 (1.30–2.60) 1.59 (0.53–4.72) 1.30 (0.94–1.82) 1.58 (1.05–2.39) 
  Ptrendb <0.01 <0.01 0.08 0.05 0.09 

NOTE: Values in bold represent P < 0.05.

aAdjusted for individual-level factors: age at entry, race/ethnicity for all women, family history of breast cancer, age at menarche, age at first live birth, number of children, hormone replacement therapy, ever had mammography, alcohol use, physical activity, and education, baseline BMI and adult weight change.

bTrend analysis excluding underweight and missing values. Trend variable was assigned ordinal value (i.e., 1, 2, 3) corresponding to categories.

cDifference between baseline weight and weight at age 21 categorized into quintiles based on study population distribution.

dAdjusted for individual-level factors with further adjustment for neighborhood SES and all neighborhood obesogenic factors (urban environment, unhealthy food, mixed-land development, parks).

This study, to our knowledge, is the first assessment of the role of the neighborhood obesogenic environment and risk of postmenopausal breast cancer. Neighborhood SES was the strongest contextual characteristic associated with breast cancer risk independent of weight gain and established risk factors for breast cancer. Living in the lowest versus highest SES neighborhoods was associated with 21% lower breast cancer risk among all women and a more pronounced, 40% lower risk among Latinos. Having more compared with no unhealthy food retail establishments was associated with 10% higher risk overall. Notable associations for the other composite built environment factors included residence in the highest versus lowest quintile of urban neighborhoods was associated with a 51% lower risk among Japanese Americans and lowest versus highest quintile of mixed-land development was associated with a 46% higher breast cancer risk in Latinos. The higher breast cancer risk associated with weight gain since age 21 persisted after adjustment for neighborhood obesogenic factors.

The higher incidence of breast cancer among high SES neighborhoods is well described in population-based studies across multiple racial/ethnic groups (reviewed in ref. 42), with associations primarily observed in hormone receptor–positive tumors (43). In general, the magnitude of association has been reported to be greater in Latinos and Asians/Pacific Islanders than whites and African Americans (28, 44). Only a few studies have been able to examine nSES (mostly as a composite measure of block group census data) while accounting for individual-level SES (mostly measuring education) and established reproductive, hormonal, and behavioral risk factors for breast cancer (45–49). In 2 prior population-based case–control studies in Wisconsin (45) and Massachusetts (46), both including largely non-Hispanic, postmenopausal white women, living in the highest versus lowest nSES quintile was associated with 20% to 30% higher odds of breast cancer independent of reproductive history, BMI, alcohol intake, and mammography, as well as urban environments. In the Black Women's Health Study of African American women, aged 21–69 years from all regions in the United States., higher nSES was associated with increased ER+ breast cancer risk (Ptrend = 0.02; n = 621 after 14 years of follow-up), but associations were slightly attenuated after adjustment for parity and age at first birth (Ptrend = 0.14) and further attenuated after additional adjustment for education, reproductive history, including lactation, BMI, vigorous exercise, alcohol intake, and mammography (Ptrend = 0.59; ref. 48). Limited number of cases and geographical variation in nSES may account for the lack of association in the other prospective studies of mostly white, postmenopausal women in Washington state (n = 856 cases; ref. 47) and Maryland (n = 272; ref. 49).

Individual-level risk factors did not explain the nSES–breast cancer risk associations in our prospective study. Significant nSES–breast cancer risk associations remained among all women after adjustment for BMI, adult weight gain, education, established reproductive and hormonal risk factors, behavioral risk factors (alcohol intake, vigorous activity), and mammography. The lower nSES–lower breast cancer risk associations were unlikely to be related to access to mammographic screening, as analyses were adjusted for ever having had a mammogram, similar to other studies (45, 46) and analyses specific for early- or late-stage breast cancer did not yield different results. Residual confounding related to individual-level SES is likely unaccounted for in our analysis that lacked detailed data on income, wealth, or occupation. However, education is a stable measure that is typically attained relatively early in life and that has been associated more consistently with breast cancer incidence than has income (50).

A notable contribution of this study, conducted using the California component of the MEC, is the examination of neighborhood effects for four distinct racial/ethnic groups. Differential associations across racial/ethnic groups were evident with a more pronounced lower nSES–lower breast cancer risk among Latinos and more urban environments–lower breast cancer risk among Japanese Americans. In general, established breast cancer risk factors have less of an influence on breast cancer risk in Latino (51) and Japanese American (52) compared with white women, which may account for the persistent association among these racial/ethnic groups while accounting for individual-level breast cancer risk factors. These associations may also reflect early-life exposures that are more prevalent in Latinos living in lower SES neighborhoods or Japanese Americans living in more urban environments such as exposures related to hygiene and infectious diseases (53). It is plausible that in low nSES environments, the development and maintenance of closer ethnic ties, stronger cultural mores and ethnically distinct resources for Japanese Americans and Latinos may have beneficial effects on breast cancer risk. Further studies are needed to understand the racial/ethnic-specific contextual role of the neighborhood on breast cancer risk (54).

More unhealthy food environments among all women and lower mixed-land development among Latinos were associated with higher breast cancer risk while accounting for body size and individual breast cancer risk factors. It is plausible that the higher breast cancer risk associated with more unhealthy food outlets and less mixed-land development operates via mechanisms involving physical or social stressors. Exposure to psychological stress has been implicated in breast cancer development (55). Further studies are needed that incorporate neighborhood characteristics associated with physiological stress [e.g., neighborhood safety/disorder (56, 57), neighborhood stability (58)], especially as chronic physiological stress can lead to the accumulation of excess body fat (59), and whether associations with breast cancer risk differ by race/ethnicity.

Strengths of our study include large sample size allowing for race/ethnicity-specific associations, data on prediagnostic BMI and adult weight gain, extensive questionnaire data on individual-level breast cancer risk factors, and 17 years of follow-up. However, there were limitations to our study. We characterize neighborhoods using census data, as opposed to detailed audits or self-reported perceived neighborhood data, potentially misclassifying neighborhoods attributes (60). However, census block group boundaries have been shown to correlate well with perceived neighborhoods (61). While the models we present did not account for length of stay at residence or change in environment (due to changes in neighborhood factors or relocations), we did conduct time-dependent analysis that accounted for change in obesogenic neighborhood factors and observed similar associations of lower nSES and lower breast cancer risk (data not shown). Furthermore, among participants who changed residences, more than 57% had little or no change in nSES level, whereas 27% increased and 16% decreased in nSES levels. We could not address the contextual influences of non-residential environments like workplace (62), geographic health care access (63), crime/disorder (64, 65), and psychosocial stress (66) or contextual exposures during early adulthood. We were not able to examine early life exposures (67) or breast cancer subtypes defined jointly by ER, PR, and HER2. We did not adjust for multiple comparisons; some of the reported differential associations across racial/ethnic groups may be due to chance.

This detailed assessment of the association between the neighborhood obesogenic environment and breast cancer risk adds to the evidence of a strong, contextual influence of nSES on breast cancer development. It suggests that there are additional operating exposures associated with breast cancer risk beyond body size, adult weight gain, and known reproductive and behavioral risk factors. Further research is needed to delineate the unique contextual aspects of high nSES, more urban environments, and less mixed-land development neighborhoods that place residents of specific racial/ethnic groups at higher risk for breast cancer or promote an unhealthy breast cancer risk profile over the life course, perhaps with a focus on early life exposures.

C.A. Clarke is a Population Health Scientist at GRAIL, Inc. No potential conflicts of interest were disclosed by the other authors.

Conception and design: C.A. Clarke, I. Cheng

Development of methodology: C.A. Clarke, S. Shariff-Marco, Y.B. Shvetsov

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C.A. Clarke, K.R. Monroe, L.N. Kolonel, L. Le Marchand, L.R. Wilkens

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Conroy, J. Yang, S. Shariff-Marco, Y.B. Shvetsov, A. Hertz, K.R. Monroe, L.R. Wilkens, I. Cheng

Writing, review, and/or revision of the manuscript: S. Conroy, C.A. Clarke, J. Yang, S. Shariff-Marco, Y.B. Shvetsov, S.-Y. Park, C.L. Albright, K.R. Monroe, L.N. Kolonel, L. Le Marchand, L.R. Wilkens, S.L. Gomez, I. Cheng

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Hertz, L. Le Marchand

Study supervision: I. Cheng

This work was supported by National Cancer Institute grant R01 CA154644. The MEC was supported by National Cancer Institute grant U01 CA164973. The development of the California Neighborhoods Data System was supported by National Cancer Institute grant R03 CA117324 and by a Rapid Response Surveillance Study from the SEER program under a modification to contract N01-PC-35136.

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

1.
Eliassen
AH
,
Colditz
GA
,
Rosner
B
,
Willett
WC
,
Hankinson
SE
. 
Adult weight change and risk of postmenopausal breast cancer
.
JAMA
2006
;
296
:
193
201
.
2.
Han
X
,
Stevens
J
,
Truesdale
KP
,
Bradshaw
PT
,
Kucharska-Newton
A
,
Prizment
AE
, et al
Body mass index at early adulthood, subsequent weight change and cancer incidence and mortality
.
Int J Cancer
2014
;
135
:
2900
9
.
3.
Keum
N
,
Greenwood
DC
,
Lee
DH
,
Kim
R
,
Aune
D
,
Ju
W
, et al
Adult weight gain and adiposity-related cancers: a dose-response meta-analysis of prospective observational studies
.
J Natl Cancer Inst
2015
;
107
.
pii: djv088
.
4.
Bandera
EV
,
Maskarinec
G
,
Romieu
I
,
John
EM
. 
Racial and ethnic disparities in the impact of obesity on breast cancer risk and survival: a global perspective
.
Adv Nutr
2015
;
6
:
803
19
.
5.
White
KK
,
Park
SY
,
Kolonel
LN
,
Henderson
BE
,
Wilkens
LR
. 
Body size and breast cancer risk: the Multiethnic Cohort
.
Int J Cancer
2012
;
131
:
E705
16
.
6.
Swinburn
B
,
Egger
G
,
Raza
F
. 
Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity
.
Prev Med
1999
;
29
:
563
70
.
7.
Mujahid
MS
,
Diez Roux
AV
,
Borrell
LN
,
Nieto
FJ
. 
Cross-sectional and longitudinal associations of BMI with socioeconomic characteristics
.
Obes Res
2005
;
13
:
1412
21
.
8.
Powell-Wiley
TM
,
Cooper-McCann
R
,
Ayers
C
,
Berrigan
D
,
Lian
M
,
McClurkin
M
, et al
Change in neighborhood socioeconomic status and weight gain: Dallas Heart Study
.
Am J Prev Med
2015
;
49
:
72
9
.
9.
Powell-Wiley
TM
,
Ayers
C
,
Agyemang
P
,
Leonard
T
,
Berrigan
D
,
Ballard-Barbash
R
, et al
Neighborhood-level socioeconomic deprivation predicts weight gain in a multi-ethnic population: longitudinal data from the Dallas Heart Study
.
Prev Med
2014
;
66
:
22
7
.
10.
Coogan
PF
,
Cozier
YC
,
Krishnan
S
,
Wise
LA
,
Adams-Campbell
LL
,
Rosenberg
L
, et al
Neighborhood socioeconomic status in relation to 10-year weight gain in the Black Women's Health Study
.
Obesity
2010
;
18
:
2064
5
.
11.
Yan
R
,
Bastian
ND
,
Griffin
PM
. 
Association of food environment and food retailers with obesity in US adults
.
Health Place
2015
;
33
:
19
24
.
12.
Holsten
JE
. 
Obesity and the community food environment: a systematic review
.
Public Health Nutr
2009
;
12
:
397
405
.
13.
Spence
JC
,
Cutumisu
N
,
Edwards
J
,
Raine
KD
,
Smoyer-Tomic
K
. 
Relation between local food environments and obesity among adults
.
BMC Public Health
2009
;
9
:
192
.
14.
Pruchno
R
,
Wilson-Genderson
M
,
Gupta
AK
. 
Neighborhood food environment and obesity in community-dwelling older adults: individual and neighborhood effects
.
Am J Public Health
2014
;
104
:
924
9
.
15.
Li
F
,
Harmer
P
,
Cardinal
BJ
,
Bosworth
M
,
Johnson-Shelton
D
. 
Obesity and the built environment: does the density of neighborhood fast-food outlets matter?
Am J Health Promot
2009
;
23
:
203
9
.
16.
Muller-Riemenschneider
F
,
Pereira
G
,
Villanueva
K
,
Christian
H
,
Knuiman
M
,
Giles-Corti
B
, et al
Neighborhood walkability and cardiometabolic risk factors in Australian adults: an observational study
.
BMC Public Health
2013
;
13
:
755
.
17.
Glazier
RH
,
Creatore
MI
,
Weyman
JT
,
Fazli
G
,
Matheson
FI
,
Gozdyra
P
, et al
Density, destinations or both? A comparison of measures of walkability in relation to transportation behaviors, obesity and diabetes in Toronto, Canada
.
PLoS One
2014
;
9
:
e85295
.
18.
Diez Roux
AV
,
Evenson
KR
,
McGinn
AP
,
Brown
DG
,
Moore
L
,
Brines
S
, et al
Availability of recreational resources and physical activity in adults
.
Am J Public Health
2007
;
97
:
493
9
.
19.
Sallis
JF
,
Hovell
MF
,
Hofstetter
CR
,
Elder
JP
,
Hackley
M
,
Caspersen
CJ
, et al
Distance between homes and exercise facilities related to frequency of exercise among San Diego residents
.
Public Health Rep
1990
;
105
:
179
85
.
20.
McCormack
GR
,
Giles-Corti
B
,
Bulsara
M
. 
The relationship between destination proximity, destination mix and physical activity behaviors
.
Prev Med
2008
;
46
:
33
40
.
21.
Mackenbach
JD
,
Rutter
H
,
Compernolle
S
,
Glonti
K
,
Oppert
JM
,
Charreire
H
, et al
Obesogenic environments: a systematic review of the association between the physical environment and adult weight status, the SPOTLIGHT project
.
BMC Public Health
2014
;
14
:
233
.
22.
Leal
C
,
Chaix
B
. 
The influence of geographic life environments on cardiometabolic risk factors: a systematic review, a methodological assessment and a research agenda
.
Obes Rev
2011
;
12
:
217
30
.
23.
Feng
J
,
Glass
TA
,
Curriero
FC
,
Stewart
WF
,
Schwartz
BS
. 
The built environment and obesity: a systematic review of the epidemiologic evidence
.
Health Place
2010
;
16
:
175
90
.
24.
Papas
MA
,
Alberg
AJ
,
Ewing
R
,
Helzlsouer
KJ
,
Gary
TL
,
Klassen
AC
. 
The built environment and obesity
.
Epidemiol Rev
2007
;
29
:
129
43
.
25.
Kolonel
LN
,
Henderson
BE
,
Hankin
JH
,
Nomura
AM
,
Wilkens
LR
,
Pike
MC
, et al
A multiethnic cohort in Hawaii and Los Angeles: baseline characteristics
.
Am J Epidemiol
2000
;
151
:
346
57
.
26.
Gomez
SL
,
Glaser
SL
,
McClure
LA
,
Shema
SJ
,
Kealey
M
,
Keegan
TH
, et al
The California Neighborhoods Data System: a new resource for examining the impact of neighborhood characteristics on cancer incidence and outcomes in populations
.
Cancer Causes Control
2011
;
22
:
631
47
.
27.
U.S. Census Bureau. 
Census 2000 Summary File 3
. Washington, DC:
U.S. Census Bureau
; 
2002
.
28.
Yost
K
,
Perkins
C
,
Cohen
R
,
Morris
C
,
Wright
W
. 
Socioeconomic status and breast cancer incidence in California for different race/ethnic groups
.
Cancer Causes Control
2001
;
12
:
703
11
.
29.
Babey SH, Diamant AL, Hastert TA, Harvey S, Goldstein H, Fluornoy Ret al 
Designed for disease: the link between local food environments and obesity and diabetes
.
California Center for Public Health Advocacy, PolicyLink, and the UCLA Center for Health Policy Research
.
April
2008
.
30.
Walls & Associates
.
National Establishment Time-Series (NETS) Database, 2009.
Oakland, CA
: Walls & Associates; 
2008
.
31.
California Department of Food and Agriculture
.
California certified farmers' market database
.
Sacramento, CA
:
California Department of Food and Agriculture
; 
2010
32.
NavTeq
.
NAVSTREETS Street Data Reference Manual v3.7. 1 July 2010
.
Chicago, IL
:
NavTeq
.
33.
California Department of Transportation
.
Highway performance and monitoring system
; 
2004
.
34.
Gunier
RB
,
Hertz
A
,
Von Behren
J
,
Reynolds
P
. 
Traffic density in California: socioeconomic and ethnic differences among potentially exposed children
.
J Expo Anal Environ Epidemiol
2003
;
13
:
240
6
.
35.
Berrigan
D
,
Pickle
LW
,
Dill
J
. 
Associations between street connectivity and active transportation
.
Int J Health Geogr
2010
;
9
:
20
.
36.
Edefonti
V
,
Randi
G
,
La Vecchia
C
,
Ferraroni
M
,
Decarli
A
. 
Dietary patterns and breast cancer: a review with focus on methodological issues
.
Nutr Rev
2009
;
67
:
297
314
.
37.
Johnson
R
,
Wichern
D
.
Applied multivariate statistical analysis
.
Englewood Cliffs, NJ
:
Prentice-Hall, Inc.
; 
1982
.
38.
Hatcher
L
.
A step-by-step approach to using the SAS system for factor analysis and structural equation modeling
.
Cary, NC
:
SAS Institute, Inc
; 
1994
.
39.
Hirsch
JA
,
Moore
KA
,
Barrientos-Gutierrez
T
,
Brines
SJ
,
Zagorski
MA
,
Rodriguez
DA
, et al
Built environment change and change in BMI and waist circumference: Multi-ethnic Study of Atherosclerosis
.
Obesity
2014
;
22
:
2450
7
.
40.
Shariff-Marco
S
,
Tai
L
,
Yang
J
,
Canchola
AJ
,
Clark
CR
,
Shvetsov
YB
, et al
Development of neighborhood obesogenic factors and their applicability to obesity-related studies in diverse populations: The Multiethnic Cohort Study
. In: Proceedings of the
Conference on Geospatial Approaches to Cancer Control and Population Sciences
; 2016 Sep 12–14;
Bethesda, MD
. Bethesda, MD:
National Cancer Institute, Division of Cancer Control and Population Science
; 
2016
.
41.
Therneau
TM
,
Grambsch
PM
,
Pankratz
VS
. 
Penalized survival models and frailty
.
J Comput Graph Stat
2003
;
12
:
156
75
.
42.
Akinyemiju
TF
,
Genkinger
JM
,
Farhat
M
,
Wilson
A
,
Gary-Webb
TL
,
Tehranifar
P
. 
Residential environment and breast cancer incidence and mortality: a systematic review and meta-analysis
.
BMC Cancer
2015
;
15
:
191
.
43.
Akinyemiju
TF
,
Pisu
M
,
Waterbor
JW
,
Altekruse
SF
. 
Socioeconomic status and incidence of breast cancer by hormone receptor subtype
.
Springerplus
2015
;
4
:
508
.
44.
Krieger
N
,
Chen
JT
,
Waterman
PD
,
Rehkopf
DH
,
Yin
R
,
Coull
BA
. 
Race/ethnicity and changing US socioeconomic gradients in breast cancer incidence: California and Massachusetts, 1978–2002 (United States)
.
Cancer Causes Control
2006
;
17
:
217
26
.
45.
Robert
SA
,
Strombom
I
,
Trentham-Dietz
A
,
Hampton
JM
,
McElroy
JA
,
Newcomb
PA
, et al
Socioeconomic risk factors for breast cancer: distinguishing individual- and community-level effects
.
Epidemiology
2004
;
15
:
442
50
.
46.
Webster
TF
,
Hoffman
K
,
Weinberg
J
,
Vieira
V
,
Aschengrau
A
. 
Community- and individual-level socioeconomic status and breast cancer risk: multilevel modeling on Cape Cod, Massachusetts
.
Environ Health Perspect
2008
;
116
:
1125
9
.
47.
Hastert
TA
,
Beresford
SA
,
Sheppard
L
,
White
E
. 
Disparities in cancer incidence and mortality by area-level socioeconomic status: a multilevel analysis
.
J Epidemiol Community Health
2015
;
69
:
168
76
.
48.
Palmer
JR
,
Boggs
DA
,
Wise
LA
,
Adams-Campbell
LL
,
Rosenberg
L
. 
Individual and neighborhood socioeconomic status in relation to breast cancer incidence in African-American women
.
Am J Epidemiol
2012
;
176
:
1141
6
.
49.
Torio
CM
,
Klassen
AC
,
Curriero
FC
,
Caballero
B
,
Helzlsouer
K
. 
The modifying effect of social class on the relationship between body mass index and breast cancer incidence
.
Am J Public Health
2010
;
100
:
146
51
.
50.
Clegg
LX
,
Reichman
ME
,
Miller
BA
,
Hankey
BF
,
Singh
GK
,
Lin
YD
, et al
Impact of socioeconomic status on cancer incidence and stage at diagnosis: selected findings from the surveillance, epidemiology, and end results: National Longitudinal Mortality Study
.
Cancer Causes Control
2009
;
20
:
417
35
.
51.
Hines
LM
,
Risendal
B
,
Slattery
ML
,
Baumgartner
KB
,
Giuliano
AR
,
Sweeney
C
, et al
Comparative analysis of breast cancer risk factors among Hispanic and non-Hispanic white women
.
Cancer
2010
;
116
:
3215
23
.
52.
Pike
MC
,
Kolonel
LN
,
Henderson
BE
,
Wilkens
LR
,
Hankin
JH
,
Feigelson
HS
, et al
Breast cancer in a multiethnic cohort in Hawaii and Los Angeles: risk factor-adjusted incidence in Japanese equals and in Hawaiians exceeds that in whites
.
Cancer Epidemiol Biomarkers Prev
2002
;
11
:
795
800
.
53.
Dowd
JB
,
Zajacova
A
,
Aiello
A
. 
Early origins of health disparities: burden of infection, health, and socioeconomic status in U.S. children
.
Soc Sci Med
2009
;
68
:
699
707
.
54.
Gomez
SL
,
Shariff-Marco
S
,
DeRouen
M
,
Keegan
TH
,
Yen
IH
,
Mujahid
M
, et al
The impact of neighborhood social and built environment factors across the cancer continuum: Current research, methodological considerations, and future directions
.
Cancer
2015
;
121
:
2314
30
.
55.
Antonova
L
,
Aronson
K
,
Mueller
CR
. 
Stress and breast cancer: from epidemiology to molecular biology
.
Breast Cancer Res
2011
;
13
:
208
.
56.
Barrington
WE
,
Stafford
M
,
Hamer
M
,
Beresford
SA
,
Koepsell
T
,
Steptoe
A
. 
Neighborhood socioeconomic deprivation, perceived neighborhood factors, and cortisol responses to induced stress among healthy adults
.
Health Place
2014
;
27
:
120
6
.
57.
Matthews
SA
,
Yang
TC
. 
Exploring the role of the built and social neighborhood environment in moderating stress and health
.
Ann Behav Med
2010
;
39
:
170
83
.
58.
Boardman
JD
. 
Stress and physical health: the role of neighborhoods as mediating and moderating mechanisms
.
Soc Sci Med
2004
;
58
:
2473
83
.
59.
Kyrou
I
,
Chrousos
GP
,
Tsigos
C
. 
Stress, visceral obesity, and metabolic complications
.
Ann N Y Acad Sci
2006
;
1083
:
77
110
.
60.
Kwan
MP
. 
From place-based to people-based exposure measures
.
Soc Sci Med
2009
;
69
:
1311
3
.
61.
Diez Roux
AV
. 
Neighborhoods and health: where are we and were do we go from here?
Rev Epidemiol Sante Publique
2007
;
55
:
13
21
.
62.
Barrington
WE
,
Beresford
SA
,
Koepsell
TD
,
Duncan
GE
,
Moudon
AV
. 
Worksite neighborhood and obesogenic behaviors: findings among employees in the promoting activity and changes in eating (PACE) trial
.
Am J Prev Med
2015
;
48
:
31
41
.
63.
Khan-Gates
JA
,
Ersek
JL
,
Eberth
JM
,
Adams
SA
,
Pruitt
SL
. 
Geographic access to mammography and its relationship to breast cancer screening and stage at diagnosis: a systematic review
.
Womens Health Issues
2015
;
25
:
482
93
.
64.
Glass
TA
,
Rasmussen
MD
,
Schwartz
BS
. 
Neighborhoods and obesity in older adults: the Baltimore memory study
.
Am J Prev Med
2006
;
31
:
455
63
.
65.
Burdette
AM
,
Hill
TD
. 
An examination of processes linking perceived neighborhood disorder and obesity
.
Soc Sci Med
2008
;
67
:
38
46
.
66.
Williams
DR
,
Mohammed
SA
,
Shields
AE
. 
Understanding and effectively addressing breast cancer in African American women: Unpacking the social context
.
Cancer
2016
;
122
:
2138
49
.
67.
Colditz
GA
,
Frazier
AL
. 
Models of breast cancer show that risk is set by events of early life: prevention efforts must shift focus
.
Cancer Epidemiol Biomarkers Prev
1995
;
4
:
567
71
.