Background:

Obesity is associated with risk of aggressive prostate cancer. It is not known whether neighborhood obesogenic factors are independently associated with prostate cancer risk.

Methods:

Neighborhood socioeconomic status (nSES) and four neighborhood obesogenic environment factors (urbanicity, mixed-land development, unhealthy food environment, and parks) were assessed for associations with prostate cancer risk among 41,563 African American, Japanese American, Latino, and White males in the Multiethnic Cohort (MEC) Study, California site. Multivariable Cox proportional hazards regression was used to estimate HRs and 95% confidence intervals (CI) for nonaggressive and aggressive prostate cancer, adjusting for individual-level sociodemographic, behavioral, and prostate cancer risk factors. Analyses were stratified by race, ethnicity, and, among Latino males, nativity.

Results:

Males residing in low-SES, compared with high-SES, neighborhoods had lower risk of nonaggressive prostate cancer [lowest vs. highest quintile HR = 0.81; 95% confidence interval (CI) = 0.68–0.95, Ptrend 0.024], driven by a similar trend among foreign-born Latino males. Foreign-born Latino males in neighborhoods with low mixed-land development had increased risk of non-aggressive disease (lowest vs. highest quintile HR = 1.49; 95% CI = 1.07–2.09). For aggressive disease, the only association noted was between lower mixed-land development and lower risk among White males (Ptrend = 0.040).

Conclusions:

nSES and obesogenic environment factors were independently associated with prostate cancer risk; associations varied by race, ethnicity, nativity, and disease aggressiveness.

Impact:

Upstream structural and social determinants of health that contribute to neighborhood obesogenic characteristics likely impact prostate cancer risk differently across groups defined by race, ethnicity, and nativity and by disease aggressiveness.

This article is featured in Highlights of This Issue, p. 917

Prostate cancer is the most commonly diagnosed cancer among males in the United States, and accounts for more than 10% of all cancer deaths among males (1). Non-Hispanic Black males have a substantially disproportionate burden of prostate cancer; their incidence of prostate cancer is more than 75% higher than any other group defined by race and ethnicity and they are twice as likely than any other group to be diagnosed with aggressive prostate cancer (1–3). In 2021, the incidence rate (IR) of prostate cancer is estimated to be 171.6 (per 100,000) among non-Hispanic Black males, followed by non-Hispanic White (IR, 97.1), Hispanic (IR, 85.6), and Asian American/Pacific Islander males (IR, 53.8; ref. 1). Established risk factors for prostate cancer include age, family history, and genetic susceptibility (4, 5). While obesity has not been consistently associated with the risk of overall prostate cancer (6–8), compelling evidence shows males with obesity have a higher risk of being diagnosed with aggressive, high-grade prostate cancer compared to males without obesity (6, 7, 9, 10).

In addition to neighborhood socioeconomic status (nSES), specific attributes of neighborhood social and built environments have been hypothesized to promote obesity (11–13) and have recently been considered for their role in cancer risk (12). Moreover, the impact of neighborhood obesogenic environment factors on cancer may be particularly relevant for specific racial and ethnic groups that experience higher levels of obesity and obesity-related diseases (14). We previously showed that nine specific attributes of the neighborhood built and social environment (population density, commuting patterns, the number of businesses, number of recreational facilities, number of parks, the restaurant environment, the retail food environment, street connectivity, traffic density) associated with increased odds of obesity, but that associations differed across racial and ethnic groups (12). We applied principal components factor analysis to these attributes to yield four neighborhood obesogenic factors (urbanicity, mixed-land development, unhealthy food environment, and parks; ref. 13). In addition to nSES, two of these neighborhood obesogenic factors (mixed-land development and urbanicity) were shown to associate with breast cancer risk across groups defined by race and ethnicity (13). Thus, studies are needed that examine the race and ethnicity-specific impacts of the obesogenic environment on prostate cancer risk while considering that the impact of the obesogenic environment on prostate cancer risk may vary by racial and ethnic group and disease aggressiveness (6, 7, 9, 10).

In a case–control study, the San Francisco Bay Area Prostate Cancer Study, we examined the associations of specific measures of the social and built environment with prostate cancer risk, and reported that higher nSES was associated with increased risks of localized and advanced prostate cancer; yet, specific neighborhood social and built environment factors explained the association of nSES only with risk of localized disease (15). Two neighborhood obesogenic attributes included in that study (i.e., parks and restaurant food environment) were not associated with diagnosis of localized or advanced prostate cancer, but this study could not examine associations across groups defined by race and ethnicity (15).

Thus, in this large prospective analysis, we further examined associations between the neighborhood obesogenic environment and prostate cancer risk, accounting for individual-level sociodemographic, behavioral, and clinical factors, among African American, Japanese American, Latino, and White males in the California component of the Multiethnic Cohort (MEC) Study (16). Understanding whether the impact of the obesogenic environment on prostate cancer risk differs across groups defined by race and ethnicity has important implications for developing tailored neighborhood-level prevention strategies and cancer screening guidelines aimed at reducing prostate cancer disparities.

Study subjects

Methodologic details of the MEC have been described previously (16). Briefly, from 1993 through 1996, 215,831 males and females between 45 and 75 years of age from Hawai’i and California (primarily Los Angeles County) were enrolled into the MEC. Participants completed a baseline questionnaire including demographic characteristics, height and weight, medical history, family history of cancer, smoking, physical activity, medication use, alcohol consumption, and diet. Participants were followed prospectively for diagnosis with incident, invasive prostate cancer [International Classification of Disease for Oncology, 3rd Edition, (ICD-O-3) site codes C61.9] through routine linkage with state registries of the NCI's Surveillance, Epidemiology and End Results (SEER) Program through 2010.

For this study, eligible African American, Japanese American, Latino, and White participants completed a baseline questionnaire while living in California, had no prostate cancer diagnosis prior to cohort entry as reported on the baseline questionnaire or through linkage with the tumor registry, and had plausible dietary data (N = 43,223). Among Latino males, we derived a joint ethnicity/nativity variable (U.S.-born Latino and foreign-born Latino); one participant missing nativity status was excluded from relevant analyses. Other racial and ethnic groups (African American, Japanese American, and White) were predominantly U.S. born, so the joint variable was not applied to these groups. We excluded participants with missing body mass index (BMI), defined as weight in kg divided by height in m2 or extreme BMI (<15 or >50 kg/m2; n = 594), as well as those with residential addresses that could not be geocoded (n = 1,064), resulting in 41,563 male participants for the present analysis. During a median follow-up of 16.2 years (interquartile range Q1–Q3 = 6.6) from cohort entry, 4,650 males were diagnosed with invasive prostate cancer. This study was approved by the Institutional Review Boards at participating institutions.

Neighborhood environment

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 were linked to U.S. Census block groups (our neighborhood unit), which average 1,475 residents. The study population represents 6,370 census block groups in California, 5,252 of which are in Los Angeles County. Characterization of the neighborhood obesogenic environment was based on linkage of baseline geocodes to the California Neighborhoods Data System, an integrated system of census, business, farmers market, park, traffic, and other small-area contextual data (17). nSES was operationalized as a composite measure by principal component analysis of census block group data on education, housing, employment, occupation, income, and poverty, as described previously (18).

As described in our previous study (12), neighborhood obesogenic attributes included specific neighborhood built and social environment attributes for which existing literature reported associations with obesity: population density (persons per square kilometer); percentage of residents who commute by car/motorcycle (per census block group); the number of businesses, recreational facilities, and parks; restaurant environment (restaurant environment index, the ratio of the number of fast-food restaurants to other restaurants) and retail food environment (retail food environment index, ratio of the number of convenience stores, liquor stores, and fast-food restaurants to supermarkets and farmers’ markets); street connectivity and traffic density. For these nine obesogenic attributes modeled as continuous variables, neighborhood obesogenic factors were previously derived from principal component factor analysis (13, 19, 20). We identified four neighborhood obesogenic environment factors: mixed-land development (more businesses and recreational facilities), urban environment (high population density, low commute, high traffic density, high street connectivity), parks (more parks per population), and unhealthy food environment (high ratio of unhealthy to healthy restaurants and retail food outlets; refs. 12, 14). Factor loadings for each were published previously (12).

nSES, mixed-land development, and urban environment were categorized into quintiles based on their distributions across Los Angeles County block groups (95% of the California MEC sample resided in Los Angeles County at baseline); each individual in the study population was assigned a quintile based on their baseline residential block group. We confirmed sufficient variability in neighborhood measures across the study population; distributions of index scores for nSES and neighborhood obesogenic factors in Los Angeles County are very similar to those for the state of California. Parks were categorized as none (no parks) and some (any parks). Unhealthy food environment index was categorized as none (no fast-food restaurant and no unhealthy food outlet), some (less than the median of non-minimal factor index values), and more (greater than or equal to the median of nonminimal factor index values).

Individual-level factors

Individual-level factors included in the analyses (assessed at baseline), which have previously been associated with or hypothesized to be associated with prostate cancer risk, were: age at cohort entry (continuous), first-degree family history of prostate cancer (yes, no, or unknown), marital status (married/widowed or divorced/single/unknown), smoking status/pack-years (non-smoker, former smoker/<20 pack-years, former smoker/20+ years, former smoker/pack-years unknown, current smoker/<20 pack-years, current smoker/20+ pack-years, and current smoker/pack-years unknown), education (a proxy for individual-level SES categorized as high school graduate or less, some college, college graduate, graduate or professional school, and unknown) and history of diabetes (yes/no). We also considered potential mediators of associations between neighborhood obesogenic factors and prostate cancer risk: BMI [underweight (<18.5 kg/m2), normal weight (18.5–24.9 kg/m2), overweight (25–29.9 kg/m2), and obese (≥30 kg/m2)], energy intake in quintiles of kcal/day [quintile 1 (475.4–1411.6), quintile 2 (1411.6–1876.5), quintile 3 (1876.6–2393.6), quintile 4 (2393.7–3214.6), and quintile 5 (3215.0–8434.2)], and moderate/vigorous physical activity characterized as none [0.0 hours/day and quartile 1 [(0.11–0.32), quartile 2 (0.36–0.71), quartile 3 (0.82–1.43), and quartile 4 (1.54–13.29)]. Prostate cancer aggressiveness was defined as advanced stage (regional or distant stage at diagnosis) or grade III or IV. Of the 4,650 males in the study population diagnosed with prostate cancer, 2,888 were diagnosed with non-aggressive disease, 1,683 were diagnosed with aggressive disease, and 79 were missing data to define aggressiveness.

Statistical analyses

We used frequencies and percentages to describe the distribution of individual- and neighborhood-level characteristics of the study population according to race, ethnicity, and, among Latino males, according to joint ethnicity and nativity (21). We used χ2 tests of association to examine whether individual-level factors differed across categories of neighborhood-level factors.

Analyses were designed to test our hypothesis that neighborhood obesogenic factors (lower mixed-land development, lower urbanicity, fewer parks, and more unhealthy food) were associated with greater risk of prostate cancer, particularly aggressive prostate cancer, through individual-level mediators related to obesity (BMI, energy intake, and physical activity). A causal diagram of these hypotheses and potential confounders is in Supplementary Fig. S1.

Multivariable Cox proportional hazards regression was used to estimate hazard ratios (HR) and corresponding 95% confidence intervals (CI) for associations of nSES and each neighborhood obesogenic factor with prostate cancer risk. All neighborhood obesogenic factors were included together in three separate multilevel models (marginal competing risk) assessing overall, nonaggressive, and aggressive prostate cancer. Minimally adjusted models included neighborhood obesogenic factors, nSES, age at cohort entry, and clustering at the census block group. Fully adjusted models were also adjusted for individual-level factors known to impact prostate cancer risk (i.e., family history and marital status) and potential individual-level and neighborhood-level confounders (i.e., nSES, smoking status/pack-years, education, diabetes); see causal diagram, Supplementary Fig. S1). Adjustment for clustering by census block group was done with a robust sandwich estimator. We tested for significant random effects with Cox frailty models and found that the random effect by block group was not significant. Thus, we kept the multivariable models adjusted for clustering by census block group as the more parsimonious approach. Reference groups for each neighborhood factor were as follows: nSES, quintile 5; mixed-land development, quintile 5; urban, quintile 1; parks, “none”; unhealthy food environment, “none.”

Tests of trends across categories of neighborhood obesogenic factors were performed by modeling each neighborhood variable as an ordinal parameter, with the P value for the association of the ordinal variable with risk reported as P for trend. Models were stratified according to groups defined by race and ethnicity and, among Latino males, nativity. Tests for heterogeneity by (i) race and ethnicity (African American, Japanese American, White) and (ii) nativity (U.S.- and foreign-born Latino) and (iii) race, ethnicity, and nativity (African American, Japanese American, U.S.-born Latino, foreign-born Latino, White) for models assessing nSES and each obesogenic environment factor were performed by including multiplicative interaction terms in the model.

To account for changes in address over time before diagnosis, we conducted a sensitivity analysis using a time-dependent measure of nSES based on residential histories during follow-up. To account for the effects of PSA screening use, we conducted a sensitivity analysis adjusting for receipt of screening. To account for potential lag time bias, we conducted a sensitivity analysis censoring individuals diagnosed within the first 7 years of the study. To account for potential misclassification bias in our definition of aggressive prostate cancer that includes consideration of both stage and grade, we conducted a sensitivity analysis of the associations between neighborhood obesogenic factors and risk of prostate cancer according to aggressiveness, with aggressiveness defined with stage and grade separately. We examined models additionally adjusted for individual-level factors hypothesized to mediate associations between neighborhood obesogenic factors and prostate cancer risk (BMI, energy intake, and physical activity).

All P values presented are two-sided. Analyses were conducted using SAS (version 9.4).

Data availability statement

The Multiethnic Cohort investigators and institutions affirm their intention to share the research data consistent with all relevant NIH resource/data sharing policies. Data requests should be submitted through MEC online data request system at https://www.uhcancercenter.org/for-researchers/mec-data-sharing.

The study population of 41,563 males was diverse with respect to race, ethnicity, and nativity: 25% African American, 14% Japanese American, 24% U.S.-born Latino, 25% foreign-born Latino, and 12% White males (Table 1). Table 1 shows the distribution of neighborhood obesogenic factors according to race, ethnicity, and nativity. Japanese American, U.S.-born Latino, and White males more frequently resided in high SES neighborhoods and neighborhoods with more mixed-land development. African American and foreign-born Latino males were more likely to reside in low SES and more urban neighborhoods. The distribution of individual-level factors according to race, ethnicity, and nativity is shown in Table 2. The median age at the time of cohort entry was 61.5 years. The majority of African American (69.7%), U.S.-born Latino (75.0%), foreign-born Latino (75.0%), and White males (65.3%) had a BMI categorized as overweight or obese (BMI ≥25 kg/m2) at baseline. χ2 tests of association most frequently showed unequal distribution of individual-level factors across categories of nSES, and least frequently according to unhealthy food environment (Supplementary Table S1). Results of tests of association varied across groups defined by race, ethnicity, and nativity.

Table 1.

Neighborhood obesogenic factors according to race, ethnicity, and nativity among California males, MEC, 1993–1996.

AllAfrican AmericanJapanese AmericanLatino, US-bornLatino, foreign-bornWhite
n (%)n (%)n (%)n (%)n (%)n (%)
Total population 41,563 10,593 5,869 9,884 10,314 4,902 
nSESa 
 Quintile 1 (low) 9,601 (23.1) 3,616 (34.1) 293 (5.0) 1,582 (16.0) 3,654 (35.4) 455 (9.3) 
 Quintile 2 10,551 (25.4) 3,032 (28.6) 718 (12.2) 2,675 (27.1) 3,361 (32.6) 765 (15.6) 
 Quintile 3 8,356 (20.1) 1,726 (16.3) 1,426 (24.3) 2,432 (24.6) 1,792 (17.4) 980 (20.0) 
 Quintile 4 7,582 (18.2) 1,651 (15.6) 1,764 (30.1) 1,902 (19.2) 986 (9.6) 1,279 (26.1) 
 Quintile 5 (highest) 5,473 (13.2) 568 (5.4) 1,668 (28.4) 1,293 (13.1) 521 (5.1) 1,423 (29.0) 
Mixed-land developmenta 
 Quintile 1 (low) 11,665 (28.1) 3,456 (32.6) 1,000 (17.0) 2,809 (28.4) 3,540 (34.3) 860 (17.5) 
 Quintile 2 9,767 (23.5) 3,048 (28.8) 1,115 (19.0) 2,241 (22.7) 2,447 (23.7) 916 (18.7) 
 Quintile 3 7,522 (18.1) 1,938 (18.3) 1,183 (20.2) 1,688 (17.1) 1,737 (16.8) 975 (19.9) 
 Quintile 4 7,260 (17.5) 1,288 (12.2) 1,403 (23.9) 1,872 (18.9) 1,502 (14.6) 1,195 (24.4) 
 Quintile 5 (high) 5,349 (12.9) 863 (8.1) 1,168 (19.9) 1,274 (12.9) 1,088 (10.5) 956 (19.5) 
Urban environmenta 
 Quintile 1 (low) 10,346 (24.9) 1,485 (14.0) 1,986 (33.8) 3,197 (32.3) 1,839 (17.8) 1,839 (37.5) 
 Quintile 2 8,596 (20.7) 1,816 (17.1) 1,636 (27.9) 2,351 (23.8) 1,832 (17.8) 961 (19.6) 
 Quintile 3 7,781 (18.7) 2,123 (20.0) 1,018 (17.3) 1,914 (19.4) 1,910 (18.5) 816 (16.6) 
 Quintile 4 7,471 (18.0) 2,815 (26.6) 653 (11.1) 1,330 (13.5) 1,982 (19.2) 691 (14.1) 
 Quintile 5 (high) 7,369 (17.7) 2,354 (22.2) 576 (9.8) 1,092 (11.0) 2,751 (26.7) 595 (12.1) 
Parksb 
 Some 34,007 (81.8) 9,085 (85.8) 4,444 (75.7) 7,883 (79.8) 8,796 (85.3) 3,798 (77.5) 
 None 7,556 (18.2) 1,508 (14.2) 1,425 (24.3) 2,001 (20.2) 1,518 (14.7) 1,104 (22.5) 
Unhealthy food environmentc 
 None 23,464 (56.5) 6,190 (58.4) 3,191 (54.4) 5,655 (57.2) 5,535 (53.7) 2,892 (59.0) 
 Some 11,775 (28.3) 2,848 (26.9) 1,653 (28.2) 2,748 (27.8) 3,249 (31.5) 1,277 (26.1) 
 More 6,324 (15.2) 1,555 (14.7) 1,025 (17.5) 1,481 (15.0) 1,530 (14.8) 733 (15.0) 
AllAfrican AmericanJapanese AmericanLatino, US-bornLatino, foreign-bornWhite
n (%)n (%)n (%)n (%)n (%)n (%)
Total population 41,563 10,593 5,869 9,884 10,314 4,902 
nSESa 
 Quintile 1 (low) 9,601 (23.1) 3,616 (34.1) 293 (5.0) 1,582 (16.0) 3,654 (35.4) 455 (9.3) 
 Quintile 2 10,551 (25.4) 3,032 (28.6) 718 (12.2) 2,675 (27.1) 3,361 (32.6) 765 (15.6) 
 Quintile 3 8,356 (20.1) 1,726 (16.3) 1,426 (24.3) 2,432 (24.6) 1,792 (17.4) 980 (20.0) 
 Quintile 4 7,582 (18.2) 1,651 (15.6) 1,764 (30.1) 1,902 (19.2) 986 (9.6) 1,279 (26.1) 
 Quintile 5 (highest) 5,473 (13.2) 568 (5.4) 1,668 (28.4) 1,293 (13.1) 521 (5.1) 1,423 (29.0) 
Mixed-land developmenta 
 Quintile 1 (low) 11,665 (28.1) 3,456 (32.6) 1,000 (17.0) 2,809 (28.4) 3,540 (34.3) 860 (17.5) 
 Quintile 2 9,767 (23.5) 3,048 (28.8) 1,115 (19.0) 2,241 (22.7) 2,447 (23.7) 916 (18.7) 
 Quintile 3 7,522 (18.1) 1,938 (18.3) 1,183 (20.2) 1,688 (17.1) 1,737 (16.8) 975 (19.9) 
 Quintile 4 7,260 (17.5) 1,288 (12.2) 1,403 (23.9) 1,872 (18.9) 1,502 (14.6) 1,195 (24.4) 
 Quintile 5 (high) 5,349 (12.9) 863 (8.1) 1,168 (19.9) 1,274 (12.9) 1,088 (10.5) 956 (19.5) 
Urban environmenta 
 Quintile 1 (low) 10,346 (24.9) 1,485 (14.0) 1,986 (33.8) 3,197 (32.3) 1,839 (17.8) 1,839 (37.5) 
 Quintile 2 8,596 (20.7) 1,816 (17.1) 1,636 (27.9) 2,351 (23.8) 1,832 (17.8) 961 (19.6) 
 Quintile 3 7,781 (18.7) 2,123 (20.0) 1,018 (17.3) 1,914 (19.4) 1,910 (18.5) 816 (16.6) 
 Quintile 4 7,471 (18.0) 2,815 (26.6) 653 (11.1) 1,330 (13.5) 1,982 (19.2) 691 (14.1) 
 Quintile 5 (high) 7,369 (17.7) 2,354 (22.2) 576 (9.8) 1,092 (11.0) 2,751 (26.7) 595 (12.1) 
Parksb 
 Some 34,007 (81.8) 9,085 (85.8) 4,444 (75.7) 7,883 (79.8) 8,796 (85.3) 3,798 (77.5) 
 None 7,556 (18.2) 1,508 (14.2) 1,425 (24.3) 2,001 (20.2) 1,518 (14.7) 1,104 (22.5) 
Unhealthy food environmentc 
 None 23,464 (56.5) 6,190 (58.4) 3,191 (54.4) 5,655 (57.2) 5,535 (53.7) 2,892 (59.0) 
 Some 11,775 (28.3) 2,848 (26.9) 1,653 (28.2) 2,748 (27.8) 3,249 (31.5) 1,277 (26.1) 
 More 6,324 (15.2) 1,555 (14.7) 1,025 (17.5) 1,481 (15.0) 1,530 (14.8) 733 (15.0) 

Abbreviation: SES, socioeconomic status.

aQuintiles for nSES, mixed-land development, and urban environment were determined according to the state distribution of each measure.

bParks (per population) was categorized as None (no park to minimal factor index value of −0.11229) and Some (factor index value of −0.08191 to 9.04815).

cThe unhealthy food environment index (high ratios of unhealthy to healthy restaurants and retail food outlets) was categorized as None (no fast-food restaurant and no unhealthy food outlet to minimal factor index value of −0.51148), Some (factor index value of −0.48073 to 0.64072), and More (factor index value of 0.65687 to 9.62594).

Table 2.

Individual-level factors according to race, ethnicity, and nativity among California males, MEC, 1993–1996.

AllAfrican AmericanJapanese AmericanLatino, US-bornLatino, foreign-bornWhite
n (%)an (%)n (%)n (%)n (%)n (%)
Total population 41,563 10,593 5,869 9,884 10,314 4,902 
Age at cohort entry (mean, std) 61.5 (8.2) 61.9 (8.8) 62.5 (8.8) 61.9 (7.5) 59.5 (7.7)   
Family history 
 No 34,136 (82.1) 8,322 (78.6) 5,118 (87.2) 8,309 (84.1) 8,287 (80.3) 4,100 (83.6) 
 Yes 2,822 (6.8) 935 (8.8) 345 (5.9) 692 (7.0) 446 (4.3) 403 (8.2) 
History of diabetes 
 No 35,350 (85.1) 8,920 (84.2) 5,133 (87.5) 8,051 (81.5) 8,806 (85.4) 4,439 (90.6) 
 Yes 6,213 (14.9) 1,673 (15.8) 736 (12.5) 1,833 (18.5) 1,508 (14.6) 463 (9.4) 
Marital status 
 Married 30,939 (74.4) 6,573 (62.1) 4,669 (79.6) 7,661 (77.5) 8,512 (82.5) 3,523 (71.9) 
 Separated/divorced/widowed 7,542 (18.1) 3,083 (29.1) 615 (10.5) 1,680 (17.0) 1,328 (12.9) 836 (17.1) 
 Single 2,671 (6.4) 780 (7.4) 559 (9.5) 472 (4.8) 353 (3.4) 507 (10.3) 
Smoking status and pack-years 
 Never-smoker 11,936 (28.7) 2,502 (23.6) 1,617 (27.6) 2,907 (29.4) 3,358 (32.6) 1,552 (31.7) 
 Former, less than 20 years 14,677 (35.3) 3,452 (32.6) 2,202 (37.5) 3,902 (39.5) 3,695 (35.8) 1,426 (29.1) 
 Former, 20 or more years 4,844 (11.7) 1,240 (11.7) 1,081 (18.4) 955 (9.7) 612 (5.9) 956 (19.5) 
 Former, pack-years unknown 1,341 (3.2) 382 (3.6) 90 (1.5) 281 (2.8) 489 (4.7) 99 (2.0) 
 Current, less than 20 years 4,527 (10.9) 1,610 (15.2) 346 (5.9) 998 (10.1) 1,314 (12.7) 258 (5.3) 
 Current, 20 or more years 3,393 (8.2) 1,208 (11.4) 468 (8.0) 693 (7.0) 479 (4.6) 545 (11.1) 
 Current, pack-years unknown 220 (0.5) 94 (0.9) 11 (0.2) 37 (0.4) 62 (0.6) 16 (0.3) 
Energy intake quintilesb 
 Quintile 1 (least) 7,899 (19.0) 2,739 (25.9) 972 (16.6) 1,831 (18.5) 1,494 (14.5) 863 (17.6) 
 Quintile 2 7,974 (19.2) 2,050 (19.4) 1,438 (24.5) 1,753 (17.7) 1,595 (15.5) 1,138 (23.2) 
 Quintile 3 8,011 (19.3) 1,903 (18.0) 1,413 (24.1) 1,885 (19.1) 1,756 (17.0) 1,054 (21.5) 
 Quintile 4 7,994 (19.2) 1,761 (16.6) 1,183 (20.2) 1,988 (20.1) 2,116 (20.5) 945 (19.3) 
 Quintile 5 (most) 7,954 (19.1) 1,688 (15.9) 619 (10.5) 2,138 (21.6) 2,896 (28.1) 613 (12.5) 
Physical activity quartilesc 
 No activity 3,220 (7.7) 929 (8.8) 179 (3.0) 515 (5.2) 1,283 (12.4) 314 (6.4) 
 Quartile 1 (least) 7,615 (18.3) 2,138 (20.2) 898 (15.3) 1,674 (16.9) 2,165 (21.0) 739 (15.1) 
 Quartile 2 11,837 (28.5) 3,169 (29.9) 1,704 (29.0) 2,875 (29.1) 2,542 (24.6) 1,547 (31.6) 
 Quartile 3 9,169 (22.1) 2,160 (20.4) 1,545 (26.3) 2,316 (23.4) 2,010 (19.5) 1,138 (23.2) 
 Quartile 4 (most) 8,791 (21.2) 1,896 (17.9) 1,491 (25.4) 2,365 (23.9) 1,937 (18.8) 1,102 (22.5) 
Education 
 High school or less 8,148 (19.6) 892 (8.4) 88 (1.5) 1,287 (13.0) 5,558 (53.9) 323 (6.6) 
 Some college 12,191 (29.3) 3,514 (33.2) 1,375 (23.4) 4,173 (42.2) 1,976 (19.2) 1,153 (23.5) 
 College graduate 11,832 (28.5) 3,788 (35.8) 2,045 (34.8) 2,868 (29.0) 1,662 (16.1) 1,468 (29.9) 
 Graduate or professional school 8,813 (21.2) 2,275 (21.5) 2,311 (39.4) 1,438 (14.5) 883 (8.6) 1,906 (38.9) 
BMId 
 Underweight 240 (0.6) 93 (0.9) 53 (0.9) 29 (0.3) 49 (0.5) 16 (0.3) 
 Normal 12,894 (31.0) 3,114 (29.4) 3,114 (53.1) 2,443 (24.7) 2,538 (24.6) 1,684 (34.4) 
 Overweight 20,594 (49.5) 5,031 (47.5) 2,364 (40.3) 5,155 (52.2) 5,689 (55.2) 2,355 (48.0) 
 Obese 7,835 (18.9) 2,355 (22.2) 338 (5.8) 2,257 (22.8) 2,038 (19.8) 847 (17.3) 
PSA screening 
 No 14,958 (36.0) 3,234 (30, 52,538 (43.2) 3,754 (38.0) 3,950 (38.3) 1,481 (30.2) 
 Yes 14,577 (35.1) 3,723 (35.1) 2,300 (39.2) 3,728 (37.7) 2,635 (25.5) 2,191 (44.7) 
AllAfrican AmericanJapanese AmericanLatino, US-bornLatino, foreign-bornWhite
n (%)an (%)n (%)n (%)n (%)n (%)
Total population 41,563 10,593 5,869 9,884 10,314 4,902 
Age at cohort entry (mean, std) 61.5 (8.2) 61.9 (8.8) 62.5 (8.8) 61.9 (7.5) 59.5 (7.7)   
Family history 
 No 34,136 (82.1) 8,322 (78.6) 5,118 (87.2) 8,309 (84.1) 8,287 (80.3) 4,100 (83.6) 
 Yes 2,822 (6.8) 935 (8.8) 345 (5.9) 692 (7.0) 446 (4.3) 403 (8.2) 
History of diabetes 
 No 35,350 (85.1) 8,920 (84.2) 5,133 (87.5) 8,051 (81.5) 8,806 (85.4) 4,439 (90.6) 
 Yes 6,213 (14.9) 1,673 (15.8) 736 (12.5) 1,833 (18.5) 1,508 (14.6) 463 (9.4) 
Marital status 
 Married 30,939 (74.4) 6,573 (62.1) 4,669 (79.6) 7,661 (77.5) 8,512 (82.5) 3,523 (71.9) 
 Separated/divorced/widowed 7,542 (18.1) 3,083 (29.1) 615 (10.5) 1,680 (17.0) 1,328 (12.9) 836 (17.1) 
 Single 2,671 (6.4) 780 (7.4) 559 (9.5) 472 (4.8) 353 (3.4) 507 (10.3) 
Smoking status and pack-years 
 Never-smoker 11,936 (28.7) 2,502 (23.6) 1,617 (27.6) 2,907 (29.4) 3,358 (32.6) 1,552 (31.7) 
 Former, less than 20 years 14,677 (35.3) 3,452 (32.6) 2,202 (37.5) 3,902 (39.5) 3,695 (35.8) 1,426 (29.1) 
 Former, 20 or more years 4,844 (11.7) 1,240 (11.7) 1,081 (18.4) 955 (9.7) 612 (5.9) 956 (19.5) 
 Former, pack-years unknown 1,341 (3.2) 382 (3.6) 90 (1.5) 281 (2.8) 489 (4.7) 99 (2.0) 
 Current, less than 20 years 4,527 (10.9) 1,610 (15.2) 346 (5.9) 998 (10.1) 1,314 (12.7) 258 (5.3) 
 Current, 20 or more years 3,393 (8.2) 1,208 (11.4) 468 (8.0) 693 (7.0) 479 (4.6) 545 (11.1) 
 Current, pack-years unknown 220 (0.5) 94 (0.9) 11 (0.2) 37 (0.4) 62 (0.6) 16 (0.3) 
Energy intake quintilesb 
 Quintile 1 (least) 7,899 (19.0) 2,739 (25.9) 972 (16.6) 1,831 (18.5) 1,494 (14.5) 863 (17.6) 
 Quintile 2 7,974 (19.2) 2,050 (19.4) 1,438 (24.5) 1,753 (17.7) 1,595 (15.5) 1,138 (23.2) 
 Quintile 3 8,011 (19.3) 1,903 (18.0) 1,413 (24.1) 1,885 (19.1) 1,756 (17.0) 1,054 (21.5) 
 Quintile 4 7,994 (19.2) 1,761 (16.6) 1,183 (20.2) 1,988 (20.1) 2,116 (20.5) 945 (19.3) 
 Quintile 5 (most) 7,954 (19.1) 1,688 (15.9) 619 (10.5) 2,138 (21.6) 2,896 (28.1) 613 (12.5) 
Physical activity quartilesc 
 No activity 3,220 (7.7) 929 (8.8) 179 (3.0) 515 (5.2) 1,283 (12.4) 314 (6.4) 
 Quartile 1 (least) 7,615 (18.3) 2,138 (20.2) 898 (15.3) 1,674 (16.9) 2,165 (21.0) 739 (15.1) 
 Quartile 2 11,837 (28.5) 3,169 (29.9) 1,704 (29.0) 2,875 (29.1) 2,542 (24.6) 1,547 (31.6) 
 Quartile 3 9,169 (22.1) 2,160 (20.4) 1,545 (26.3) 2,316 (23.4) 2,010 (19.5) 1,138 (23.2) 
 Quartile 4 (most) 8,791 (21.2) 1,896 (17.9) 1,491 (25.4) 2,365 (23.9) 1,937 (18.8) 1,102 (22.5) 
Education 
 High school or less 8,148 (19.6) 892 (8.4) 88 (1.5) 1,287 (13.0) 5,558 (53.9) 323 (6.6) 
 Some college 12,191 (29.3) 3,514 (33.2) 1,375 (23.4) 4,173 (42.2) 1,976 (19.2) 1,153 (23.5) 
 College graduate 11,832 (28.5) 3,788 (35.8) 2,045 (34.8) 2,868 (29.0) 1,662 (16.1) 1,468 (29.9) 
 Graduate or professional school 8,813 (21.2) 2,275 (21.5) 2,311 (39.4) 1,438 (14.5) 883 (8.6) 1,906 (38.9) 
BMId 
 Underweight 240 (0.6) 93 (0.9) 53 (0.9) 29 (0.3) 49 (0.5) 16 (0.3) 
 Normal 12,894 (31.0) 3,114 (29.4) 3,114 (53.1) 2,443 (24.7) 2,538 (24.6) 1,684 (34.4) 
 Overweight 20,594 (49.5) 5,031 (47.5) 2,364 (40.3) 5,155 (52.2) 5,689 (55.2) 2,355 (48.0) 
 Obese 7,835 (18.9) 2,355 (22.2) 338 (5.8) 2,257 (22.8) 2,038 (19.8) 847 (17.3) 
PSA screening 
 No 14,958 (36.0) 3,234 (30, 52,538 (43.2) 3,754 (38.0) 3,950 (38.3) 1,481 (30.2) 
 Yes 14,577 (35.1) 3,723 (35.1) 2,300 (39.2) 3,728 (37.7) 2,635 (25.5) 2,191 (44.7) 

Abbreviations: BMI, body mass index; std, standard deviation.

aColumn percentages may not add up to 100% due to missing data.

bEnergy intake in quintiles of kcal/day according to the distribution of intake among the study population [quintile 1 (475.4–1,411.6), quintile 2 (1,411.6–1,876.5), quintile 3 (1,876.6–2,393.6), quintile 4 (2,393.7–3,214.6), and quintile 5 (3,215.0–8,434.2)].

cPhysical activity categorizes as none (0.0 hours/day) and quartiles of hours/day according to the distribution of activity among the study population [quartile 1 (0.11–0.32), quartile 2 (0.36–0.71), quartile 3 (0.82–1.43), and quartile 4 (1.54–13.29)].

dBMI [underweight (<18.5 kg/m2), normal weight (18.5–24.9 kg/m2), overweight (25–29.9 kg/m2), and obese (≥30 kg/m2)].

Table 3 shows associations between neighborhood obesogenic factors and overall prostate cancer risk. There were trends in associations between lower nSES and lower prostate cancer risk among all groups combined (Ptrend = 0.016) as well as U.S.-born Latino and foreign-born Latino males (Ptrend = 0.028 and 0.008, respectively). In addition, compared with the highest quintile of mixed-land development (quintile 5), residents of neighborhoods in quintiles 4 and 1 had higher prostate cancer risk (quintile 4, HR = 1.45, 95% CI = 1.10–1.90; quintile 1, HR = 1.33, 95% CI = 1.03–1.72).

Table 3.

Associations between baseline neighborhood obesogenic environment factors and risk of overall prostate cancer (nonaggressive and aggressive combined) according to race, ethnicity, and nativity, MEC, 1993–2010.

Alla,bAfrican AmericanJapanese AmericanLatino, US-bornLatino, foreign-bornWhite
CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)
nSES 
 Quintile 5 (high) 1,032 Reference 525 Reference 16 Reference 159 Reference 282 Reference 50 Reference 
 Quintile 4 1,214 0.91 (0.81–1.01) 515 0.96 (0.75–1.22) 55 1.01 (0.79–1.28) 282 0.95 (0.77–1.16) 284 0.64 (0.47–0.87) 78 0.85 (0.67–1.07) 
 Quintile 3 927 0.91 (0.81–1.02) 292 1.01 (0.78–1.30) 134 1.15 (0.88–1.50) 242 0.83 (0.67–1.02) 159 0.66 (0.49–0.88) 100 0.96 (0.74–1.23) 
 Quintile 2 850 0.92 (0.81–1.03) 278 1.07 (0.83–1.37) 152 0.92 (0.64–1.32) 213 0.85 (0.69–1.06) 84 0.65 (0.49–0.87) 123 1.01 (0.74–1.38) 
 Quintile 1 (low) 627 0.81 (0.70–0.93) 104 0.93 (0.72–1.22) 131 0.67 (0.36–1.25) 153 0.73 (0.55–0.97) 72 0.56 (0.41–0.77) 167 1.13 (0.74–1.70) 
Ptrend  0.016  0.853  0.800  0.028  0.008  0.643 
Mixed-land development 
 Quintile 5 (high) 1,318 Reference 561 Reference 82 Reference 309 Reference 292 Reference 74 Reference 
 Quintile 4 1,132 1.01 (0.91–1.13) 486 0.94 (0.76–1.16) 104 0.85 (0.64–1.13) 252 0.84 (0.67–1.05) 201 1.45 (1.10–1.90) 89 1.15 (0.88–1.48) 
 Quintile 3 820 0.94 (0.85–1.05) 304 0.89 (0.73–1.09) 91 0.82 (0.62–1.09) 175 0.93 (0.75–1.16) 142 1.13 (0.85–1.50) 108 1.00 (0.77–1.30) 
 Quintile 2 803 1.02 (0.92–1.14) 213 0.95 (0.79–1.14) 110 1.07 (0.82–1.41) 169 1.06 (0.86–1.30) 163 1.23 (0.94–1.61) 148 0.90 (0.67–1.19) 
 Quintile 1 (low) 577 1.03 (0.93–1.15) 150 0.96 (0.79–1.16) 101 0.87 (0.63–1.20) 144 1.10 (0.89–1.36) 83 1.33 (1.03–1.72) 99 0.74 (0.54–1.03) 
Ptrend  0.428  0.981  0.976  0.050  0.212  0.024 
Urban 
 Quintile 1 (low) 1,111 Reference 247 Reference 166 Reference 336 Reference 169 Reference 193 Reference 
 Quintile 2 970 1.02 (0.93–1.12) 341 1.20 (1.00–1.44) 136 0.90 (0.70–1.16) 243 0.97 (0.81–1.14) 147 0.92 (0.73–1.16) 103 1.11 (0.87–1.42) 
 Quintile 3 934 1.08 (0.98–1.19) 352 1.07 (0.90–1.28) 98 1.11 (0.85–1.47) 215 1.10 (0.91–1.31) 183 1.09 (0.87–1.36) 86 1.07 (0.81–1.41) 
 Quintile 4 821 0.94 (0.84–1.05) 421 0.97 (0.81–1.18) 48 0.90 (0.63–1.28) 136 0.97 (0.78–1.22) 151 0.86 (0.67–1.10) 65 0.93 (0.68–1.26) 
 Quintile 5 (high) 814 1.06 (0.94–1.19) 353 1.09 (0.89–1.34) 40 0.93 (0.61–1.40) 119 1.13 (0.88–1.46) 231 1.00 (0.78–1.29) 71 1.29 (0.91–1.83) 
Ptrend  0.846  0.791  0.898  0.444  0.854  0.477 
Parks 
 Some 3,807 0.97 (0.90–1.05) 1,446 0.92 (0.80–1.07) 371 0.98 (0.77–1.26) 832 0.92 (0.79–1.08) 754 1.09 (0.90–1.33) 404 0.97 (0.78–1.21) 
 None 843 Reference 268 Reference 117 Reference 217 Reference 127 Reference 114 Reference 
Unhealthy food environment 
 None 2,647 Reference 1,017 Reference 276 Reference 598 Reference 456 Reference 300 Reference 
 Some 1,324 1.05 (0.98–1.13) 478 1.09 (0.97–1.22) 139 0.97 (0.78–1.21) 295 1.09 (0.94–1.26) 278 1.05 (0.90–1.22) 134 0.92 (0.74–1.15) 
 More 679 0.99 (0.91–1.09) 219 0.90 (0.77–1.06) 73 0.77 (0.58–1.01) 156 1.06 (0.88–1.28) 147 1.17 (0.96–1.41) 84 1.03 (0.80–1.34) 
Ptrend  0.727  0.551  0.085  0.366  0.132  0.997 
Alla,bAfrican AmericanJapanese AmericanLatino, US-bornLatino, foreign-bornWhite
CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)
nSES 
 Quintile 5 (high) 1,032 Reference 525 Reference 16 Reference 159 Reference 282 Reference 50 Reference 
 Quintile 4 1,214 0.91 (0.81–1.01) 515 0.96 (0.75–1.22) 55 1.01 (0.79–1.28) 282 0.95 (0.77–1.16) 284 0.64 (0.47–0.87) 78 0.85 (0.67–1.07) 
 Quintile 3 927 0.91 (0.81–1.02) 292 1.01 (0.78–1.30) 134 1.15 (0.88–1.50) 242 0.83 (0.67–1.02) 159 0.66 (0.49–0.88) 100 0.96 (0.74–1.23) 
 Quintile 2 850 0.92 (0.81–1.03) 278 1.07 (0.83–1.37) 152 0.92 (0.64–1.32) 213 0.85 (0.69–1.06) 84 0.65 (0.49–0.87) 123 1.01 (0.74–1.38) 
 Quintile 1 (low) 627 0.81 (0.70–0.93) 104 0.93 (0.72–1.22) 131 0.67 (0.36–1.25) 153 0.73 (0.55–0.97) 72 0.56 (0.41–0.77) 167 1.13 (0.74–1.70) 
Ptrend  0.016  0.853  0.800  0.028  0.008  0.643 
Mixed-land development 
 Quintile 5 (high) 1,318 Reference 561 Reference 82 Reference 309 Reference 292 Reference 74 Reference 
 Quintile 4 1,132 1.01 (0.91–1.13) 486 0.94 (0.76–1.16) 104 0.85 (0.64–1.13) 252 0.84 (0.67–1.05) 201 1.45 (1.10–1.90) 89 1.15 (0.88–1.48) 
 Quintile 3 820 0.94 (0.85–1.05) 304 0.89 (0.73–1.09) 91 0.82 (0.62–1.09) 175 0.93 (0.75–1.16) 142 1.13 (0.85–1.50) 108 1.00 (0.77–1.30) 
 Quintile 2 803 1.02 (0.92–1.14) 213 0.95 (0.79–1.14) 110 1.07 (0.82–1.41) 169 1.06 (0.86–1.30) 163 1.23 (0.94–1.61) 148 0.90 (0.67–1.19) 
 Quintile 1 (low) 577 1.03 (0.93–1.15) 150 0.96 (0.79–1.16) 101 0.87 (0.63–1.20) 144 1.10 (0.89–1.36) 83 1.33 (1.03–1.72) 99 0.74 (0.54–1.03) 
Ptrend  0.428  0.981  0.976  0.050  0.212  0.024 
Urban 
 Quintile 1 (low) 1,111 Reference 247 Reference 166 Reference 336 Reference 169 Reference 193 Reference 
 Quintile 2 970 1.02 (0.93–1.12) 341 1.20 (1.00–1.44) 136 0.90 (0.70–1.16) 243 0.97 (0.81–1.14) 147 0.92 (0.73–1.16) 103 1.11 (0.87–1.42) 
 Quintile 3 934 1.08 (0.98–1.19) 352 1.07 (0.90–1.28) 98 1.11 (0.85–1.47) 215 1.10 (0.91–1.31) 183 1.09 (0.87–1.36) 86 1.07 (0.81–1.41) 
 Quintile 4 821 0.94 (0.84–1.05) 421 0.97 (0.81–1.18) 48 0.90 (0.63–1.28) 136 0.97 (0.78–1.22) 151 0.86 (0.67–1.10) 65 0.93 (0.68–1.26) 
 Quintile 5 (high) 814 1.06 (0.94–1.19) 353 1.09 (0.89–1.34) 40 0.93 (0.61–1.40) 119 1.13 (0.88–1.46) 231 1.00 (0.78–1.29) 71 1.29 (0.91–1.83) 
Ptrend  0.846  0.791  0.898  0.444  0.854  0.477 
Parks 
 Some 3,807 0.97 (0.90–1.05) 1,446 0.92 (0.80–1.07) 371 0.98 (0.77–1.26) 832 0.92 (0.79–1.08) 754 1.09 (0.90–1.33) 404 0.97 (0.78–1.21) 
 None 843 Reference 268 Reference 117 Reference 217 Reference 127 Reference 114 Reference 
Unhealthy food environment 
 None 2,647 Reference 1,017 Reference 276 Reference 598 Reference 456 Reference 300 Reference 
 Some 1,324 1.05 (0.98–1.13) 478 1.09 (0.97–1.22) 139 0.97 (0.78–1.21) 295 1.09 (0.94–1.26) 278 1.05 (0.90–1.22) 134 0.92 (0.74–1.15) 
 More 679 0.99 (0.91–1.09) 219 0.90 (0.77–1.06) 73 0.77 (0.58–1.01) 156 1.06 (0.88–1.28) 147 1.17 (0.96–1.41) 84 1.03 (0.80–1.34) 
Ptrend  0.727  0.551  0.085  0.366  0.132  0.997 

Note: Bolded estimates indicate P < 0.05.

Abbreviations: CI, confidence interval; HR, hazard ratio; SES, socioeconomic status; US, United States.

aModels include all neighborhood obesogenic factors and nSES and are additionally adjusted for age at cohort entry, family history of prostate cancer, marital status, smoking status and years smoked, education, history of diabetes, and clustering at the Census block group.

bFor models assessing neighborhood factors among all races/ethnicities and each stratum of racial/ethnic group, P values for heterogeneity across disease aggressiveness were all P > 0.05.

Associations between neighborhood obesogenic environment factors and nonaggressive prostate cancer risk within fully adjusted models are shown in Table 4. Among all males, risk of nonaggressive prostate cancer decreased with lower nSES, among foreign-born Latino males (Ptrend = 0.014). In addition, foreign-born Latino males residing in neighborhoods with the highest mixed-land development had lower risk of non-aggressive prostate cancer compared to the lower four quintiles, although a trend in the association between mixed-land development and risk was not detected. Results from these fully adjusted models were very similar to those from minimally adjusted models (Supplementary Table S2A).

Table 4.

Associations between baseline neighborhood obesogenic environment factors and risk of nonaggressive prostate cancer according to race, ethnicity, and nativity, MEC, 1993–2010.

Alla,b,cAfrican AmericanJapanese AmericanLatino, US bornLatino, foreign bornWhite
CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)
nSES 
 Quintile 5 (high) 375 Reference 67 Reference 74 Reference 95 Reference 47 Reference 92 Reference 
 Quintile 4 546 1.02 (0.89–1.17) 191 1.07 (0.80–1.43) 93 1.14 (0.85–1.53) 134 1.01 (0.78–1.31) 56 0.66 (0.45–0.98) 72 0.94 (0.69–1.29) 
 Quintile 3 586 1.03 (0.89–1.18) 179 1.02 (0.77–1.36) 82 1.27 (0.92–1.76) 155 0.95 (0.73–1.23) 106 0.71 (0.50–1.01) 64 1.15 (0.81–1.62) 
 Quintile 2 768 1.05 (0.91–1.21) 328 1.18 (0.89–1.58) 29 0.87 (0.55–1.39) 178 0.94 (0.72–1.23) 182 0.69 (0.49–0.97) 51 1.27 (0.84–1.93) 
 Quintile 1 (low) 613 0.91 (0.76–1.07) 323 1.06 (0.78–1.44) 10 0.75 (0.36–1.58) 89 0.77 (0.55–1.10) 161 0.54 (0.37–0.79) 30 1.27 (0.71–2.24) 
Ptrend  0.434  0.647  0.968  0.208  0.014  0.209 
Mixed-land development 
 Quintile 5 (high) 343 Reference 96 Reference 59 Reference 79 Reference 49 Reference 60 Reference 
 Quintile 4 509 1.06 (0.92–1.23) 136 0.91 (0.70–1.18) 70 0.96 (0.67–1.38) 112 1.02 (0.76–1.36) 106 1.60 (1.13–2.28) 85 1.06 (0.75–1.49) 
 Quintile 3 524 1.01 (0.88–1.16) 206 0.93 (0.73–1.18) 50 0.81 (0.55–1.20) 114 1.14 (0.86–1.52) 88 1.20 (0.83–1.74) 66 1.02 (0.72–1.44) 
 Quintile 2 689 1.05 (0.91–1.20) 307 0.93 (0.73–1.17) 59 1.09 (0.75–1.59) 149 1.14 (0.87–1.51) 120 1.30 (0.92–1.83) 54 0.89 (0.61–1.30) 
 Quintile 1 (low) 823 1.08 (0.94–1.24) 343 0.92 (0.72–1.17) 50 0.90 (0.60–1.36) 197 1.28 (0.97–1.69) 189 1.49 (1.07–2.09) 44 0.74 (0.49–1.14) 
Ptrend  0.363  0.652  0.886  0.043  0.151  0.125 
Urban 
 Quintile 1 (low) 698 Reference 164 Reference 96 Reference 218 Reference 110 Reference 110 Reference 
 Quintile 2 605 1.02 (0.91–1.14) 224 1.19 (0.96–1.48) 75 0.84 (0.61–1.17) 149 0.91 (0.73–1.13) 100 0.97 (0.74–1.27) 57 1.12 (0.81–1.55) 
 Quintile 3 582 1.08 (0.96–1.22) 220 1.01 (0.81–1.27) 64 1.34 (0.97–1.85) 134 1.05 (0.84–1.31) 108 1.03 (0.78–1.37) 56 1.25 (0.87–1.79) 
 Quintile 4 520 0.98 (0.86–1.12) 265 0.94 (0.74–1.19) 30 1.10 (0.72–1.67) 82 0.93 (0.70–1.24) 100 0.93 (0.69–1.27) 43 1.10 (0.73–1.66) 
 Quintile 5 (high) 483 1.05 (0.91–1.22) 215 1.02 (0.79–1.33) 23 1.06 (0.62–1.79) 68 1.06 (0.76–1.47) 134 0.98 (0.72–1.34) 43 1.46 (0.92–2.32) 
Ptrend  0.685  0.481  0.269  0.845  0.837  0.166 
Parks 
 None 2,353 0.94 (0.86–1.04) 914 0.93 (0.78–1.10) 217 0.90 (0.68–1.18) 516 0.94 (0.78–1.13) 471 1.11 (0.87–1.41) 235 0.81 (0.61–1.08) 
 Some 535 Reference 174 Reference 71 Reference 135 Reference 81 Reference 74 Reference 
Unhealthy food environment 
 None 1,650 Reference 639 Reference 163 Reference 383 Reference 292 Reference 173 Reference 
 Some 807 1.05 (0.96–1.14) 304 1.11 (0.97–1.28) 85 1.03 (0.78–1.34) 169 0.99 (0.83–1.20) 169 1.04 (0.86–1.26) 80 0.92 (0.68–1.24) 
 More 431 1.02 (0.92–1.15) 145 0.95 (0.78–1.15) 40 0.71 (0.49–1.01) 99 1.06 (0.84–1.34) 91 1.17 (0.92–1.49) 56 1.18 (0.85–1.64) 
Ptrend  0.491  0.983  0.096  0.688  0.227  0.490 
Alla,b,cAfrican AmericanJapanese AmericanLatino, US bornLatino, foreign bornWhite
CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)
nSES 
 Quintile 5 (high) 375 Reference 67 Reference 74 Reference 95 Reference 47 Reference 92 Reference 
 Quintile 4 546 1.02 (0.89–1.17) 191 1.07 (0.80–1.43) 93 1.14 (0.85–1.53) 134 1.01 (0.78–1.31) 56 0.66 (0.45–0.98) 72 0.94 (0.69–1.29) 
 Quintile 3 586 1.03 (0.89–1.18) 179 1.02 (0.77–1.36) 82 1.27 (0.92–1.76) 155 0.95 (0.73–1.23) 106 0.71 (0.50–1.01) 64 1.15 (0.81–1.62) 
 Quintile 2 768 1.05 (0.91–1.21) 328 1.18 (0.89–1.58) 29 0.87 (0.55–1.39) 178 0.94 (0.72–1.23) 182 0.69 (0.49–0.97) 51 1.27 (0.84–1.93) 
 Quintile 1 (low) 613 0.91 (0.76–1.07) 323 1.06 (0.78–1.44) 10 0.75 (0.36–1.58) 89 0.77 (0.55–1.10) 161 0.54 (0.37–0.79) 30 1.27 (0.71–2.24) 
Ptrend  0.434  0.647  0.968  0.208  0.014  0.209 
Mixed-land development 
 Quintile 5 (high) 343 Reference 96 Reference 59 Reference 79 Reference 49 Reference 60 Reference 
 Quintile 4 509 1.06 (0.92–1.23) 136 0.91 (0.70–1.18) 70 0.96 (0.67–1.38) 112 1.02 (0.76–1.36) 106 1.60 (1.13–2.28) 85 1.06 (0.75–1.49) 
 Quintile 3 524 1.01 (0.88–1.16) 206 0.93 (0.73–1.18) 50 0.81 (0.55–1.20) 114 1.14 (0.86–1.52) 88 1.20 (0.83–1.74) 66 1.02 (0.72–1.44) 
 Quintile 2 689 1.05 (0.91–1.20) 307 0.93 (0.73–1.17) 59 1.09 (0.75–1.59) 149 1.14 (0.87–1.51) 120 1.30 (0.92–1.83) 54 0.89 (0.61–1.30) 
 Quintile 1 (low) 823 1.08 (0.94–1.24) 343 0.92 (0.72–1.17) 50 0.90 (0.60–1.36) 197 1.28 (0.97–1.69) 189 1.49 (1.07–2.09) 44 0.74 (0.49–1.14) 
Ptrend  0.363  0.652  0.886  0.043  0.151  0.125 
Urban 
 Quintile 1 (low) 698 Reference 164 Reference 96 Reference 218 Reference 110 Reference 110 Reference 
 Quintile 2 605 1.02 (0.91–1.14) 224 1.19 (0.96–1.48) 75 0.84 (0.61–1.17) 149 0.91 (0.73–1.13) 100 0.97 (0.74–1.27) 57 1.12 (0.81–1.55) 
 Quintile 3 582 1.08 (0.96–1.22) 220 1.01 (0.81–1.27) 64 1.34 (0.97–1.85) 134 1.05 (0.84–1.31) 108 1.03 (0.78–1.37) 56 1.25 (0.87–1.79) 
 Quintile 4 520 0.98 (0.86–1.12) 265 0.94 (0.74–1.19) 30 1.10 (0.72–1.67) 82 0.93 (0.70–1.24) 100 0.93 (0.69–1.27) 43 1.10 (0.73–1.66) 
 Quintile 5 (high) 483 1.05 (0.91–1.22) 215 1.02 (0.79–1.33) 23 1.06 (0.62–1.79) 68 1.06 (0.76–1.47) 134 0.98 (0.72–1.34) 43 1.46 (0.92–2.32) 
Ptrend  0.685  0.481  0.269  0.845  0.837  0.166 
Parks 
 None 2,353 0.94 (0.86–1.04) 914 0.93 (0.78–1.10) 217 0.90 (0.68–1.18) 516 0.94 (0.78–1.13) 471 1.11 (0.87–1.41) 235 0.81 (0.61–1.08) 
 Some 535 Reference 174 Reference 71 Reference 135 Reference 81 Reference 74 Reference 
Unhealthy food environment 
 None 1,650 Reference 639 Reference 163 Reference 383 Reference 292 Reference 173 Reference 
 Some 807 1.05 (0.96–1.14) 304 1.11 (0.97–1.28) 85 1.03 (0.78–1.34) 169 0.99 (0.83–1.20) 169 1.04 (0.86–1.26) 80 0.92 (0.68–1.24) 
 More 431 1.02 (0.92–1.15) 145 0.95 (0.78–1.15) 40 0.71 (0.49–1.01) 99 1.06 (0.84–1.34) 91 1.17 (0.92–1.49) 56 1.18 (0.85–1.64) 
Ptrend  0.491  0.983  0.096  0.688  0.227  0.490 

Note: Bolded estimates indicate P < 0.05.

Abbreviations: CI, confidence interval; HR, hazard ratio; SES, socioeconomic status; US, United States.

aNonaggressive prostate cancer was defined as local stage at diagnosis or grade I or II.

bModels include all neighborhood obesogenic factors and neighborhood SES and are additionally adjusted for age at cohort entry, family history of prostate cancer, marital status, smoking status and years smoked, education, history of diabetes, and clustering at the Census block group.

cFor models assessing nSES and all neighborhood factors, P values for heterogeneity across groups defined by race, ethnicity, and nativity were >0.1.

Associations between neighborhood obesogenic factors and risk of aggressive prostate cancer risk within fully adjusted models are shown in Table 5. There were no significant associations between nSES and risk of aggressive prostate cancer in any group. There was a trend in the association of lower mixed-land development and lower risk of prostate cancer among White males (Ptrend = 0.040), although none of the quintile risk estimates were significantly different from that of the reference group. Results from these fully adjusted models were very similar to those from minimally adjusted models (Supplementary Table S2B).

Table 5.

Associations between baseline neighborhood obesogenic environment factors and risk of aggressive prostate cancer according to race, ethnicity, and nativity, MEC, 1993–2010.

Alla,b,cAfrican AmericanJapanese AmericanLatino, US bornLatino, foreign bornWhite
CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)
nSES 
 Quintile 5 (high) 243 Reference 36 Reference 54 Reference 57 Reference 23 Reference 73 Reference 
 Quintile 4 292 0.82 (0.69–0.98) 83 0.81 (0.51–1.29) 57 0.94 (0.62–1.42) 77 0.93 (0.65–1.31) 27 0.63 (0.36–1.11) 48 0.82 (0.57–1.16) 
 Quintile 3 326 0.88 (0.73–1.05) 109 1.10 (0.70–1.73) 52 1.18 (0.77–1.81) 81 0.77 (0.53–1.10) 49 0.64 (0.38–1.09) 35 0.89 (0.60–1.33) 
 Quintile 2 427 0.90 (0.75–1.09) 178 1.03 (0.65–1.63) 25 1.22 (0.70–2.14) 98 0.81 (0.57–1.16) 101 0.74 (0.44–1.24) 25 0.93 (0.58–1.52) 
 Quintile 1 (low) 395 0.86 (0.69–1.07) 186 0.90 (0.56–1.45) 0.73 (0.28–1.89) 68 0.86 (0.55–1.37) 117 0.71 (0.41–1.24) 18 1.22 (0.61–2.43) 
Ptrend  0.540  0.905  0.565  0.345  0.754  0.897 
Mixed-land development 
 Quintile 5 (high) 221 Reference 46 Reference 41 Reference 62 Reference 33 Reference 39 Reference 
 Quintile 4 279 0.91 (0.76–1.08) 75 1.01 (0.69–1.48) 37 0.72 (0.44–1.17) 54 0.61 (0.43–0.88) 53 1.13 (0.73–1.75) 60 1.16 (0.80–1.69) 
 Quintile 3 286 0.86 (0.72–1.02) 94 0.86 (0.59–1.25) 40 0.92 (0.60–1.41) 58 0.72 (0.50–1.02) 53 1.00 (0.65–1.52) 41 0.87 (0.58–1.31) 
 Quintile 2 421 0.99 (0.84–1.17) 170 1.06 (0.75–1.50) 44 1.12 (0.73–1.71) 98 0.95 (0.69–1.32) 77 1.09 (0.72–1.64) 32 0.79 (0.50–1.24) 
 Quintile 1 (low) 476 0.97 (0.82–1.14) 207 1.11 (0.79–1.56) 32 0.85 (0.53–1.38) 109 0.92 (0.65–1.29) 101 1.04 (0.70–1.56) 27 0.69 (0.41–1.15) 
Ptrend  0.709  0.295  0.704  0.401  0.968  0.040 
Urban 
 Quintile 1 (low) 396 Reference 77 Reference 68 Reference 112 Reference 56 Reference 83 Reference 
 Quintile 2 349 1.06 (0.91–1.23) 110 1.25 (0.92–1.70) 61 0.97 (0.64–1.45) 91 1.18 (0.89–1.55) 44 0.81 (0.54–1.22) 43 1.08 (0.75–1.58) 
 Quintile 3 340 1.13 (0.96–1.32) 128 1.27 (0.93–1.73) 32 0.83 (0.52–1.31) 79 1.30 (0.96–1.77) 71 1.24 (0.84–1.83) 30 0.88 (0.56–1.36) 
 Quintile 4 283 0.92 (0.77–1.09) 149 1.12 (0.82–1.54) 16 0.64 (0.36–1.13) 50 1.13 (0.78–1.64) 50 0.81 (0.53–1.25) 18 0.57 (0.34–0.97) 
 Quintile 5 (high) 315 1.13 (0.94–1.37) 128 1.29 (0.92–1.82) 17 0.78 (0.40–1.52) 49 1.39 (0.91–2.13) 96 1.13 (0.73–1.75) 25 1.01 (0.57–1.81) 
Ptrend  0.598  0.374  0.169  0.146  0.587  0.270 
Parks 
 Some 1,385 1.00 (0.88–1.15) 501 0.87 (0.69–1.11) 150 1.16 (0.76–1.75) 300 0.87 (0.67–1.12) 274 1.11 (0.79–1.54) 160 1.26 (0.89–1.78) 
 None 298 Reference 91 Reference 44 Reference 81 Reference 43 Reference 39 Reference 
Unhealthy food environment 
 None 952 Reference 357 Reference 109 Reference 207 Reference 158 Reference 121 Reference 
 Some 493 1.08 (0.97–1.21) 164 1.06 (0.87–1.29) 53 0.94 (0.66–1.35) 118 1.24 (0.98–1.57) 106 1.11 (0.87–1.43) 52 0.95 (0.68–1.34) 
 More 238 0.97 (0.83–1.13) 71 0.85 (0.66–1.10) 32 0.90 (0.59–1.37) 56 1.09 (0.80–1.48) 53 1.21 (0.87–1.67) 26 0.84 (0.55–1.28) 
Ptrend  0.905  0.429  0.599  0.280  0.221  0.421 
Alla,b,cAfrican AmericanJapanese AmericanLatino, US bornLatino, foreign bornWhite
CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)CasesHR (95% CI)
nSES 
 Quintile 5 (high) 243 Reference 36 Reference 54 Reference 57 Reference 23 Reference 73 Reference 
 Quintile 4 292 0.82 (0.69–0.98) 83 0.81 (0.51–1.29) 57 0.94 (0.62–1.42) 77 0.93 (0.65–1.31) 27 0.63 (0.36–1.11) 48 0.82 (0.57–1.16) 
 Quintile 3 326 0.88 (0.73–1.05) 109 1.10 (0.70–1.73) 52 1.18 (0.77–1.81) 81 0.77 (0.53–1.10) 49 0.64 (0.38–1.09) 35 0.89 (0.60–1.33) 
 Quintile 2 427 0.90 (0.75–1.09) 178 1.03 (0.65–1.63) 25 1.22 (0.70–2.14) 98 0.81 (0.57–1.16) 101 0.74 (0.44–1.24) 25 0.93 (0.58–1.52) 
 Quintile 1 (low) 395 0.86 (0.69–1.07) 186 0.90 (0.56–1.45) 0.73 (0.28–1.89) 68 0.86 (0.55–1.37) 117 0.71 (0.41–1.24) 18 1.22 (0.61–2.43) 
Ptrend  0.540  0.905  0.565  0.345  0.754  0.897 
Mixed-land development 
 Quintile 5 (high) 221 Reference 46 Reference 41 Reference 62 Reference 33 Reference 39 Reference 
 Quintile 4 279 0.91 (0.76–1.08) 75 1.01 (0.69–1.48) 37 0.72 (0.44–1.17) 54 0.61 (0.43–0.88) 53 1.13 (0.73–1.75) 60 1.16 (0.80–1.69) 
 Quintile 3 286 0.86 (0.72–1.02) 94 0.86 (0.59–1.25) 40 0.92 (0.60–1.41) 58 0.72 (0.50–1.02) 53 1.00 (0.65–1.52) 41 0.87 (0.58–1.31) 
 Quintile 2 421 0.99 (0.84–1.17) 170 1.06 (0.75–1.50) 44 1.12 (0.73–1.71) 98 0.95 (0.69–1.32) 77 1.09 (0.72–1.64) 32 0.79 (0.50–1.24) 
 Quintile 1 (low) 476 0.97 (0.82–1.14) 207 1.11 (0.79–1.56) 32 0.85 (0.53–1.38) 109 0.92 (0.65–1.29) 101 1.04 (0.70–1.56) 27 0.69 (0.41–1.15) 
Ptrend  0.709  0.295  0.704  0.401  0.968  0.040 
Urban 
 Quintile 1 (low) 396 Reference 77 Reference 68 Reference 112 Reference 56 Reference 83 Reference 
 Quintile 2 349 1.06 (0.91–1.23) 110 1.25 (0.92–1.70) 61 0.97 (0.64–1.45) 91 1.18 (0.89–1.55) 44 0.81 (0.54–1.22) 43 1.08 (0.75–1.58) 
 Quintile 3 340 1.13 (0.96–1.32) 128 1.27 (0.93–1.73) 32 0.83 (0.52–1.31) 79 1.30 (0.96–1.77) 71 1.24 (0.84–1.83) 30 0.88 (0.56–1.36) 
 Quintile 4 283 0.92 (0.77–1.09) 149 1.12 (0.82–1.54) 16 0.64 (0.36–1.13) 50 1.13 (0.78–1.64) 50 0.81 (0.53–1.25) 18 0.57 (0.34–0.97) 
 Quintile 5 (high) 315 1.13 (0.94–1.37) 128 1.29 (0.92–1.82) 17 0.78 (0.40–1.52) 49 1.39 (0.91–2.13) 96 1.13 (0.73–1.75) 25 1.01 (0.57–1.81) 
Ptrend  0.598  0.374  0.169  0.146  0.587  0.270 
Parks 
 Some 1,385 1.00 (0.88–1.15) 501 0.87 (0.69–1.11) 150 1.16 (0.76–1.75) 300 0.87 (0.67–1.12) 274 1.11 (0.79–1.54) 160 1.26 (0.89–1.78) 
 None 298 Reference 91 Reference 44 Reference 81 Reference 43 Reference 39 Reference 
Unhealthy food environment 
 None 952 Reference 357 Reference 109 Reference 207 Reference 158 Reference 121 Reference 
 Some 493 1.08 (0.97–1.21) 164 1.06 (0.87–1.29) 53 0.94 (0.66–1.35) 118 1.24 (0.98–1.57) 106 1.11 (0.87–1.43) 52 0.95 (0.68–1.34) 
 More 238 0.97 (0.83–1.13) 71 0.85 (0.66–1.10) 32 0.90 (0.59–1.37) 56 1.09 (0.80–1.48) 53 1.21 (0.87–1.67) 26 0.84 (0.55–1.28) 
Ptrend  0.905  0.429  0.599  0.280  0.221  0.421 

Note: Bolded estimates indicate P < 0.05.

Abbreviations: CI, confidence interval; HR, hazard ratio; SES, socioeconomic status; US, United States.

aAggressive prostate cancer defined as regional or distant stage or grade III or IV.

bModels include all neighborhood obesogenic factors and nSES and are additionally adjusted for age at cohort entry, family history of prostate cancer, marital status, smoking status and years smoked, education, history of diabetes, and clustering at the Census block group.

cFor models assessing nSES and all neighborhood factors, P values for heterogeneity across groups defined by race, ethnicity, and nativity were >0.1.

Models adjusting for putative individual-level mediators of the obesogenic environment (BMI, energy intake, and physical activity) did not alter effect sizes of observed associations between neighborhood nSES or obesogenic factors and risk of either nonaggressive or aggressive prostate cancer (Supplementary Table S3). Sensitivity analyses accounting for use of PSA screening and changes in address over time showed similar patterns and effect sizes compared with our main analyses. Sensitivity analyses assessing associations between nSES and prostate cancer risk (overall, nonaggressive, and aggressive) using a time-dependent nSES variable did not indicate substantial bias due to our use of baseline neighborhood factors (Supplementary Table S4). Sensitivity analyses testing for lag time bias did not differ substantially from the main analysis (Supplementary Table S5). Sensitivity analyses assessing associations between neighborhood obesogenic factors and risk of prostate cancer according to aggressiveness, with aggressiveness defined with stage and grade separately are presented in Supplementary Table S6. We found that our results with defining aggressiveness using both stage and grade more closely approximate results defining aggressiveness with grade only. However, there are substantially fewer cases available for analysis with aggressiveness defined by stage only, which may be contributing to the small number of differences noted between the stage only and stage and grade sets of analyses.

This study is the first prospective assessment of the role of neighborhood obesogenic factors and prostate cancer risk. In addition, this is the first study, to our knowledge, to assess a wide-ranging set of multilevel prostate cancer risk factors by race, ethnicity, and nativity and to assess differences in these associations by disease aggressiveness. We observed varying associations between neighborhood obesogenic factors and prostate cancer risk across race and ethnicity, across nativity among Latino males, and across disease aggressiveness. We hypothesized that these associations would be dependent upon individual-level factors related to obesity (BMI, caloric intake, and physical activity), but these were not significant mediators. It is possible that upstream factors of the neighborhood environment contributed to observed associations between neighborhood obesogenic factors and prostate cancer risk.

Like others, we considered race and ethnicity as sociopolitical constructs reflecting hierarchies that are established and maintained by society, structures, and systems of racism that differentially allocate power and resources. Residential segregation and resultant neighborhood disinvestment have resulted in the segregation of minoritized racial and ethnic groups into spaces that have aspects of social and built environments, like obesogenic factors, that may be detrimental to health. These segregated spaces, however, may also serve as enclaves within which residents’ health benefits from other factors of the social environment related to coethnic support, positive social and cultural norms, and protection from discrimination. Assessing the impact of neighborhood factors on cancer risk is, therefore, improved by diverse study populations and analyses that examine variation in impacts across racial and ethnic groups.

Higher nSES has been associated with higher prostate cancer risk, independent of disease aggressiveness or individual-level SES (22–27). Some studies assessed this association across racial and ethnic groups where data allowed, with results differing across studies (22, 25–27). Our results showed a positive association only among foreign-born Latino males with non-aggressive disease, suggesting that the association of nSES with risk may be confounded by additional unmeasured factors. Indeed, our results illustrated the importance of considering additional sociodemographic characteristics within racial, ethnic, and nativity groups. Ours was also one of few studies that examined the independent associations of individual-level SES and nSES with prostate cancer risk (23, 24), and confirmed that, at least for foreign-born Latino males, the association of nSES with risk was independent of individual-level SES (in this case, individual-level education) as well as other individual-level prostate cancer risk factors. Frequently, use of PSA screening is offered as a rationale for observations of associations between higher SES and diagnosis of prostate cancer; however, our sensitivity analysis including PSA screening use did not change our results. An important distinction between our study and previous studies was our study's prospective design using address data at baseline (previous studies utilized addresses at diagnosis among cases and matched controls) as well as important prostate cancer risk factors. Results of sensitivity analyses using time-dependent measures of nSES based on residential histories during follow-up were not different from main analyses using nSES at baseline.

We found that BMI, physical activity, and energy intake did not alter the associations observed between neighborhood factors and prostate cancer risk. This result could have been influenced by the specific measures we have for these behaviors. BMI may be a relatively imprecise measure of body size in studies of cancer risk; a recent study, for example, suggested that while associations with BMI and prostate cancer are weak, these associations might be more evident for waist circumferences (28). Energy intake does not measure the quality of individuals’ diets; however, an additional analysis including a measure of diet quality (the Alternative Healthy Eating Index-2010) did not change results. MEC survey questions on physical activity were detailed and captured both time spent and intensity of physical activity, but real-time, active monitoring of physical activity (e.g., through wearable devices) could confer additional accuracy in the measurement of physical activity. Nevertheless, our reported observation warrants consideration of an alternative hypothesis by which obesogenic factors (each representing a different domain of the built environment) impacted prostate cancer risk through mechanisms other than individual body size, diet, or physical activity. It is possible that these obesogenic factors were confounded by other domains of the social and built neighborhood environment or interacted with other domains to impact prostate cancer risk.

Our study adds to growing evidence for the need to assess interactions between neighborhood factors and individual-level sociodemographic factors. A previous study of these obesogenic factors and postmenopausal breast cancer risk also observed varying associations of these neighborhood-level factors by race and ethnicity (13).

Associations of obesogenic factors and prostate cancer risk differ somewhat between U.S.-born and foreign-born Latino males, with more and stronger associations evident among foreign-born Latino males. In fact, associations for obesogenic factors among foreign-born Latino males mirror those among all males (all races, ethnicities, and nativity) combined, suggesting that associations among foreign-born Latino males were driving those observed among all males combined. Among Latino males, lower mixed-land development was associated with increased risk of nonaggressive disease while greater mixed-land development was associated with greater risk of aggressive disease. We were unable to explain these opposing associations across disease aggressiveness. It is possible an unmeasured confounder would explain these observations.

This study was novel in the prospective, multilevel assessment of neighborhood obesogenic factors and prostate cancer risk, in its assessment of risk across groups defined by race, ethnicity, and nativity, and assessment of both individual-level (education) and neighborhood-level SES, while accounting for known prostate cancer risk factors. Nevertheless, some limitations are noted. In some groups defined by race, ethnicity, and nativity, the number of men diagnosed with prostate cancer were relatively low, which may have limited our power to observe significant risk differences across categories of neighborhood factors. As mentioned previously, our results suggested that unmeasured neighborhood factors, upstream of the obesogenic factors examined, could have impacted prostate cancer risk. Racial residential segregation and resultant neighborhood disinvestment of low income or non-White neighborhoods, for example, may impact both the presence of an obesogenic environment and prostate cancer through mechanisms not directly related to obesity. Future studies of prostate cancer risk in diverse populations should consider such upstream neighborhood environment factors. In addition, measures used to define the neighborhood obesogenic environment utilized secondary data that did not capture individuals’ lived experiences within their neighborhoods. Lived experiences may differ for individuals across race, ethnicity, and nativity status in ways that impact prostate cancer (29–31). Thus, ongoing research seeking to capture the nature of individuals’ lived experiences as they relate to cancer risk is warranted.

Conclusion

In summary, this study demonstrated variations in associations between neighborhood obesogenic factors and prostate cancer risk across race, ethnicity, nativity, and disease aggressiveness and suggested that obesogenic factors impact prostate cancer risk independently of individuals’ BMI, energy intake, or level of physical activity.

S. Shariff-Marco reports grants from NCI during the conduct of the study. L. Le Marchand reports grants from NCI during the conduct of the study. S.L. Gomez reports grants from NCI during the conduct of the study. I. Cheng reports grants from NCI during the conduct of the study. No disclosures were reported by the other authors.

M.C. DeRouen: Investigation, visualization, methodology, writing–original draft, writing–review and editing. L. Tao: Investigation, methodology, writing–original draft. S. Shariff-Marco: Conceptualization, investigation, methodology, writing–review and editing. J. Yang: Formal analysis, visualization, methodology, writing–review and editing. Y.B. Shvetsov: Investigation, writing–review and editing. S.-Y. Park: Investigation, writing–review and editing. C.L. Albright: Investigation, writing–review and editing. K.R. Monroe: Investigation, writing–review and editing. L. Le Marchand: Resources, investigation, writing–review and editing. L.R. Wilkens: Investigation, writing–review and editing. S.L. Gomez: Investigation, methodology, writing–review and editing. I. Cheng: Conceptualization, supervision, funding acquisition, investigation, methodology, writing–review and editing.

This work was supported by NCI grant R01 CA154644 (PI: I. Cheng; Support: M.C. DeRouen, L. Tao, J. Yang, S. Shariff-Marco, Y.B. Shvetsov, C.L. Albright, K.R. Monroe, L.R. Wilkens, and S.L. Gomez. The MEC was supported by NCI grant U01 CA164973 (MPIs: L. Le Marchand, C. Haiman, and L.R. Wilkens; Support: S.-Y. Park, K.R. Monroe, I. Cheng, J. Yang, S. Shariff-Marco). The development of the California Neighborhoods Data System was supported by NCI grant R03 CA117324 (PI: S.L. Gomez) and by a Rapid Response Surveillance Study from the SEER program under a modification to contract N01-PC-35136 (PI: S.L. Gomez).

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.
American Cancer Society
.
Cancer facts and figures 2021
.
Atlanta
:
American Cancer Society
;
2021
. Available from: https://www.cancer.org/content/dam/cancer-org/research/cancer-facts-and-statistics/annual-cancer-facts-and-figures/2021/cancer-facts-and-figures-2021.pdf.
2.
Chornokur
G
,
Dalton
K
,
Borysova
ME
,
Kumar
NB
.
Disparities at presentation, diagnosis, treatment, and survival in African American men, affected by prostate cancer
.
Prostate
2011
;
71
:
985
97
.
3.
Pietro
GD
,
Chornokur
G
,
Kumar
NB
,
Davis
C
,
Park
JY
.
Racial differences in the diagnosis and treatment of prostate cancer
.
Int Neurourol J
2016
;
20
:
S112
119
.
4.
Grönberg
H
.
Prostate cancer epidemiology
.
Lancet
2003
;
361
:
859
64
.
5.
Park
S-Y
,
Haiman
CA
,
Cheng
I
,
Park
SL
,
Wilkens
LR
,
Kolonel
LN
, et al
.
Racial/ethnic differences in lifestyle-related factors and prostate cancer risk: the Multiethnic Cohort Study
.
Cancer Causes Control
2015
;
26
:
1507
15
.
6.
Møller
H
,
Roswall
N
,
Van Hemelrijck
M
,
Larsen
SB
,
Cuzick
J
,
Holmberg
L
, et al
.
Prostate cancer incidence, clinical stage and survival in relation to obesity: a prospective cohort study in Denmark
.
Int J Cancer
2015
;
136
:
1940
7
.
7.
Zhang
X
,
Zhou
G
,
Sun
B
,
Zhao
G
,
Liu
D
,
Sun
J
, et al
.
Impact of obesity upon prostate cancer-associated mortality: a meta-analysis of 17 cohort studies
.
Oncol Lett
2015
;
9
:
1307
12
.
8.
Wilson
KM
,
Giovannucci
EL
,
Mucci
LA
.
Lifestyle and dietary factors in the prevention of lethal prostate cancer
.
Asian J Androl
2012
;
14
:
365
74
.
9.
Wright
ME
,
Chang
S-C
,
Schatzkin
A
,
Albanes
D
,
Kipnis
V
,
Mouw
T
, et al
.
Prospective study of adiposity and weight change in relation to prostate cancer incidence and mortality
.
Cancer
2007
;
109
:
675
84
.
10.
Allott
EH
,
Masko
EM
,
Freedland
SJ
.
Obesity and prostate cancer: weighing the evidence
.
Eur Urol
2013
;
63
:
800
9
.
11.
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
.
12.
Conroy
SM
,
Shariff-Marco
S
,
Yang
J
,
Hertz
A
,
Cockburn
M
,
Shvetsov
YB
, et al
.
Characterizing the neighborhood obesogenic environment in the multiethnic cohort: a multi-level infrastructure for cancer health disparities research
.
Cancer Causes Control
2018
;
29
:
167
83
.
13.
Conroy
SM
,
Clarke
CA
,
Yang
J
,
Shariff-Marco
S
,
Shvetsov
YB
,
Park
S-Y
, et al
.
Contextual impact of neighborhood obesogenic factors on postmenopausal breast cancer: the multiethnic cohort
.
Cancer Epidemiol Biomarkers Prev
2017
;
26
:
480
9
.
14.
Lovasi
GS
,
Hutson
MA
,
Guerra
M
,
Neckerman
KM
.
Built environments and obesity in disadvantaged populations
.
Epidemiol Rev
2009
;
31
:
7
20
.
15.
DeRouen
MC
,
Schupp
CW
,
Yang
J
,
Koo
J
,
Hertz
A
,
Shariff-Marco
S
, et al
.
Impact of individual and neighborhood factors on socioeconomic disparities in localized and advanced prostate cancer risk
.
Cancer Causes Control
2018
;
29
:
951
66
.
16.
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
.
17.
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
.
18.
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
.
19.
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
.
20.
Johnson
RA
,
Wichern
DW
.
Applied multivariate statistical analysis
.
Upper Saddle River, NJ: Prentice Hall
;
2002
.
21.
Neighborhood poverty | National Equity Atlas
. Available from: https://nationalequityatlas.org/indicators/Neighborhood_poverty#/.
22.
Cheng
I
,
Witte
JS
,
McClure
LA
,
Shema
SJ
,
Cockburn
MG
,
John
EM
, et al
.
Socioeconomic status and prostate cancer incidence and mortality rates among the diverse population of California
.
Cancer Causes Control
2009
;
20
:
1431
40
.
23.
Hastert
TA
,
Beresford
SAA
,
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
.
24.
DeRouen
MC
,
Schupp
CW
,
Koo
J
,
Yang
J
,
Hertz
A
,
Shariff-Marco
S
, et al
.
Impact of individual and neighborhood factors on disparities in prostate cancer survival
.
Cancer Epidemiol
2018
;
53
:
1
11
.
25.
Yin
D
,
Morris
C
,
Allen
M
,
Cress
R
,
Bates
J
,
Liu
L
.
Does socioeconomic disparity in cancer incidence vary across racial/ethnic groups?
Cancer Causes Control
2010
;
21
:
1721
30
.
26.
Singh
GK
,
Jemal
A
.
Socioeconomic and racial/ethnic disparities in cancer mortality, incidence, and survival in the United States, 1950–2014: over six decades of changing patterns and widening inequalities
.
J Environ Public Health
2017
;
2017
:
2819372
.
27.
Krieger
N
,
Quesenberry
C
,
Peng
T
,
Horn-Ross
P
,
Stewart
S
,
Brown
S
, et al
.
Social class, race/ethnicity, and incidence of breast, cervix, colon, lung, and prostate cancer among Asian, Black, Hispanic, and White residents of the San Francisco Bay Area, 1988–92 (United States)
.
Cancer Causes Control
1999
;
10
:
525
37
.
28.
Choi
JB
,
Myong
J-P
,
Lee
Y
,
Kim
I
,
Kim
JH
,
Hong
S-H
, et al
.
Does increased body mass index lead to elevated prostate cancer risk? It depends on waist circumference
.
BMC Cancer
2020
;
20
:
589
.
29.
Kwan
M-P
.
From place-based to people-based exposure measures
.
Soc Sci Med
2009
;
69
:
1311
3
.
30.
Kwan
M-P
.
The limits of the neighborhood effect: contextual uncertainties in geographic, environmental health, and social science research
.
Ann Am Assoc Geogr
2018
;
108
:
1482
90
.
31.
Ford
CL
,
Airhihenbuwa
CO
.
Critical race theory, race equity, and public health: toward antiracism praxis
.
Am J Public Health
2010
;
100
:
S30
5
.

Supplementary data