Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Nov 21.
Published in final edited form as: Health Place. 2015 Nov 21;36:162–172. doi: 10.1016/j.healthplace.2015.10.003

Impact of neighborhoods and body size on survival after breast cancer diagnosis

Salma Shariff-Marco a,b,c,*,1, Scarlett L Gomez a,b,c,1, Meera Sangaramoorthy a,1, Juan Yang a,1, Jocelyn Koo a,1, Andrew Hertz a,1, Esther M John a,b,c,1, Iona Cheng a,c,1, Theresa HM Keegan d,2
PMCID: PMC4684167  NIHMSID: NIHMS740029  PMID: 26606455

Abstract

With data from the Neighborhoods and Breast Cancer Study, we examined the associations between body size, social and built environments, and survival following breast cancer diagnosis among 4347 women in the San Francisco Bay Area. Lower neighborhood socioeconomic status and greater neighborhood crowding were associated with higher waist-to-hip ratio (WHR). After mutual adjustment, WHR, but not neighborhood characteristics, was positively associated with overall mortality and marginally with breast cancer-specific mortality. Our findings suggest that WHR is an important modifiable prognostic factor for breast cancer survivors. Future WHR interventions should account for neighborhood characteristics that may influence WHR.

Keywords: Body size, Waist-to-hip ratio, Breast cancer survival, Neighborhood

1. Introduction

With the growing number of breast cancer survivors in the United States, it is important to identify modifiable factors that contribute to better survival after breast cancer diagnosis (American Cancer Society, 2012). Prior studies have shown that lifestyle factors, including physical activity and body size, influence survival (Vance et al, 2011; Hauner et al, 2011; Protani et al, 2010; Carmichael and Bates, 2004; Chen et al, 2010; Caan et al, 2008; Conroy et al, 2011; Kwan et al, 2012, 2014). Neighborhood social and built environment factors may be associated with body size and ultimately with survival through several pathways, including material deprivation, health behaviors (healthy eating, physical activity) and access to resources (Feng et al., 2010; Northridge et al., 2003; Diez Roux and Mair, 2010; Yen et al., 2009; Meijer et al., 2012; Krieger, 2001; Gomez et al., 2015). Few studies, however, have examined associations between body size and survival among racially/ethnically diverse groups (Conroy et al., 2011; Kwan et al., 2012, 2014), and no studies have assessed how neighborhood factors are associated with body size and survival among women diagnosed with breast cancer.

Obesity has been consistently associated with worse overall (Hauner et al., 2011) and breast cancer-specific (Protani et al., 2010; Chen et al., 2010; Caan et al., 2008; Kwan et al., 2012, 2014) survival, with no variation by race/ethnicity (Conroy et al., 2011; Kwan et al., 2012). While body mass index (BMI) has been the most commonly studied indicator of body size, weight change (Vance et al., 2011; Chen et al., 2010; Caan et al., 2008) and waist-to-hip ratio (WHR), a measure of body fat distribution that reflects both adipose tissue and muscle mass (Molarius and Seidell, 1998), have also been considered. Although the findings for weight gain have been mixed (Vance et al., 2011; Chen et al., 2010; Caan et al., 2008), associations between larger WHR and worse survival after breast cancer diagnosis have been noted in two (Protani et al., 2010; Kwan et al., 2014) of three studies that examined these associations (Protani et al., 2010; Chen et al., 2010; Kwan et al., 2014).

We used data from the Neighborhoods and Breast Cancer (NABC) Study to examine the association of body size with survival after breast cancer diagnosis among a racially/ethnically diverse cohort of women with breast cancer. We also assessed the associations of neighborhood characteristics with body size and survival.

2. Materials and methods

2.1. Subjects

Breast cancer cases in the NABC Study, described in more detail elsewhere (Shariff-Marco et al., 2014; Keegan et al., 2014), were identified through the Greater Bay Area Cancer Registry and participated in the San Francisco Bay Area Breast Cancer Study (SFBCS), a case-control study in African American (AA), Hispanic, and non-Hispanic white (NHW) women that included breast cancer cases aged 35–79 years and diagnosed between 1995 and 2002 (John et al., 2003, 2005), or in the Northern California site of the Breast Cancer Family Registry (NC-BCFR), a multiethnic family study that included breast cancer cases aged 18–64 years and diagnosed between 1995 and 2009 (John et al., 2004, 2007). Cases were screened by telephone to assess study eligibility, with 84% and 83% participation among those contacted in SFBCS and NC-BCFR, respectively. Eligible cases completed an in-person interview (n = 2258 (88%) in SFBCS; and n= 3631 (77%) in NC-BCFR as of September 2009).

We limited the analysis to 5237 women diagnosed with a first primary invasive breast cancer between 1995 and 2008 who completed the interview themselves. We excluded cases for the following reasons: NC-BCFR duplicate cases who also participated in SFBCS (n= 339), no geocodeable address (n= 198) or follow-up information (n= 25) from the cancer registry, a prior cancer (n= 259), Native American or mixed race/ethnicity (n= 11), or unknown BMI (n= 58). The final analysis included 4347 cases interviewed on average 21.0 months (SD=11.1 months) after diagnosis. Mean follow-up after interview was 7.4 years. Study participants provided written informed consent and all protocols were approved by the Institutional Review Board of the Cancer Prevention Institute of California.

2.2. Data collection

In both studies, professional interviewers conducted in-person interviews at the participants’ homes in English, Spanish, or Chinese using similarly structured questionnaires which facilitated data harmonization and pooling for analysis. In both studies, the reference year was defined as the calendar year prior to diagnosis. Data were collected on age at diagnosis, race/ethnicity, education, first-degree family history of breast cancer, personal history of benign breast disease, years since last pregnancy, history of oral contraceptive use, history of menopausal hormone therapy use, alcohol intake during the reference year (Block et al., 1986, 1990), and recent (during the 3 years prior to diagnosis) recreational physical activity (hours per week) (Bernstein et al., 1994; John et al., 2003; Yang et al., 2003; Dallal et al., 2007; West-Wright et al., 2009; Keegan et al., 2014). In SFBCS, recreational physical activity was assessed using an approach developed by Dr. Leslie Bernstein that asked participants to list all episodes of sports and exercise in which they engaged (Bernstein et al., 1994); other studies of breast cancer have observed inverse associations with physical activity using a similar approach (John et al., 2003; Yang et al., 2003). In NC-BCFR, the questions on recreational physical activity were modeled after the approach used in the California Teachers Study where participants were asked to list hours per week that they spent doing moderate and strenuous physical activities (Dallal et al., 2007; West-Wright et al., 2009). Assessment and harmonization of recreational physical activity for these two studies has been previously reported in detail (Keegan et al., 2014).

Both studies assessed self-reported weight in the reference year (i.e., pre-diagnosis weight) and adult height. NC-BCFR also assessed self-reported weight at interview, whereas SFBCS measured weight and height at interview. For women who declined the measurements, self-reported height was used for the BMI calculation. Pre-diagnosis BMI (kg/m2) was calculated as weight (kg) in the reference year divided by height (m) and was categorized according to World Health Organization cut points (underweight: ≤ 18.5 kg/m2; normal weight: 18.6–24.9; overweight: 25.0–29.9; obese: ≥ 30.0) (World Health Organization, 2000). Percent weight change (kg) was calculated as the difference between weight at interview and weight in the reference year divided by weight in the reference year; percent weight change was categorized based on previously published work with the following distribution of total cases: decrease (≥ 2%), stable (± 1%), moderate increase (2–10%), and large increase (> 10%) (Bradshaw et al., 2012). Waist and hip circumferences were measured at interview in SFBCS only (n= 1916 cases). WHR was calculated as waist circumference (cm) divided by hip circumference (cm) measured at interview, and as done in prior studies WHR was categorized according to the quartile distribution among all cases (John et al., 2013, 2011; Kwan et al., 2014; Protani et al. 2010).

For each case, we obtained cancer registry information on year of diagnosis, ICD-O-3 tumor histologic subtype, histological grade, estrogen receptor (ER) and progesterone receptor (PR) status, AJCC stage, time to first and second subsequent tumors, first-course treatment, marital status, and vital status (routinely determined by the cancer registry through hospital follow-up and database linkages) as of December 31, 2009, and, for the deceased, the underlying cause of death (California Cancer Registry, 2009). Using cause of death information for breast cancer from cancer registries has been validated previously (Hu et al. 2013).

2.3. Neighborhood social and built environment characteristics

Data on neighborhood characteristics were obtained from the California Neighborhoods Data System (Gomez et al., 2011). We examined a broad suite of social and built environment factors to better understand which specific factors are contributing to body size and survival after breast cancer. Residential address at the time of diagnosis was geocoded to latitude and longitude coordinates and then assigned a 2000 Census block group (representing an average of 1500 residents with a range of 600–3000 residents). For 2% of cases, we geocoded their address at time of interview as their address at time of diagnosis was incomplete or not geocodeable (e.g., PO Box). For neighborhood-level socioeconomic status (nSES), we used a previously validated composite measure of seven SES indicators from Census data at the level of block group (Yost et al., 2001). In addition to population density (persons/square meter), neighborhood density was characterized at the block group level by urban/rural status (Reynolds et al., 2005) and percentage of occupied housing units with more than one occupant per room (crowding). Urban/rural status is derived from census defined Urbanized Areas (population ≥ 50,000) and Urban Clusters (population between 2500 and 50,000) (see footnotes of tables). Street connectivity was measured using Gamma, the ratio of actual number of street segments to maximum possible number of intersections, with a higher ratio indicating more street connectivity/ walkability (Berrigan et al., 2010). Data on traffic counts from the California Department of Transportation (California Department of Transportation, 2004) were used to obtain traffic density within a 500-meter buffer of each residence, using methods described previously (Gunier et al., 2003). Other neighborhood social factors include percentage of total housing units that are not single family dwellings (i.e., structures with more than 2 units), percentage of foreign-born residents, and percentage of linguistically isolated households (US Census Bureau, 2002). Quintiles/quartiles cut-points were based on distributions among the study cases with the exception of neighborhood SES and population density which were based on statewide distributions.

We derived information on neighborhood amenities including business listings from Walls and Associates’ National Establishment Time-Series Database from 1990 to 2008 (Walls and Associates, 2008), and farmers’ markets listings in 2010 from the California Department of Food and Agriculture (California Department of Food and Agriculture, 2010). Using ArcGIS software, neighborhood amenities within a 1600-meter network distance (Thornton et al., 2011) from residence at diagnosis were averaged over a 4 year window-one year before diagnosis, during the year of diagnosis, and two years after diagnosis. For the small proportion of cases diagnosed in 2007 and 2008 (2%) for whom we did not have 4 years of business data, we averaged over a 2 or 3 year window, depending on data availability. The average number of recreational facilities included places where recreational activities could take place. The Restaurant Environment Index is the ratio of the average number of fast food restaurants to other restaurants, and the Retail Food Environment Index (California Center for Public Health Advocacy et al., 2008) is the ratio of the average number of convenience stores, liquor stores, and fast food restaurants to supermarkets and farmers’ markets. Quintiles/quartiles cut-points for these measures were based on distributions among the study cases, with the exception of the Restaurant Environment Index, Retail Food Environment Index and number of farmer’s markets (see footnotes of tables).

2.4. Statistical analysis

We examined the association between body size (BMI, % weight change, WHR) with overall and breast cancer (BC)-specific mortality using stage- and study-stratified Cox proportional hazards regression to calculate hazard ratios (HR) and 95% confidence intervals (CI). Our base hazard regression models were adjusted for age at diagnosis, year of diagnosis (calendar year), study, and race/ethnicity. Subsequently, we adjusted for tumor characteristics, treatment, and personal factors associated with survival. We performed stratified analyses by age at diagnosis (< 50 or ≥ 50 years), ER status (ER+, ER−, unknown), and race/ ethnicity (NHW, AA, Hispanic, Asian American). Tests for heterogeneity across strata were conducted using likelihood ratio tests comparing models with and without an interaction term between body size measures and the stratified variable; no significant interactions were found (data not shown).

Because WHR was the only body size measure significantly associated with mortality, we examined the relationship between WHR (> median vs. ≤ median) and neighborhood factors, using logistic regression to calculate odds ratios (OR) and 95% CIs. Neighborhood factors that were associated with WHR and/or survival were included in the multivariable Cox regression models. Tests for linear trend were used to evaluate associations between mortality and increasing ordinal categories of body size and neighborhood characteristics (Liu, 2007). We also tested for interactions between nSES and WHR and found no statistically significant interactions (data not shown). All models included cluster adjustment for census block groups, as there were insufficient numbers of cases within each block group to warrant multilevel modeling; of the 1371 block groups in the WHR analysis, over 70% had only one case. The sandwich estimator of the covariance structure, applied to Cox proportional hazards regression models, accounted for any intracluster dependence and yielded robust standard error estimates even under model misspecification (Lin and Wei, 1989). Analyses were conducted using SAS (version 9.3, Cary, NC). We also tested for spatial autocorrelation (using Moran’s I) in the multivariable Cox regression models with deviance residuals from our fully-adjusted regression models using ArcGIS -ESRI (version 10.1, Redding, CA) and found no evidence of it.

For deceased women, survival time was measured in days from the date of diagnosis to the date of death of any cause for overall mortality and to the date of death from breast cancer for BC-specific mortality. We used left truncation at the date of interview to adjust for the time from diagnosis to interview. For BC-specific mortality, women who died from other causes were censored at the time of death. Women alive at the study end date (December 31, 2009) were censored at the earlier of the two—the study end date or the date of last follow-up (i.e., last known contact) which was obtained from the California Cancer Registry in October 2011. The proportional hazards assumption was tested for WHR and neighborhood variables using significance tests of interactions with the time scale, and visual examination of scaled Schoenfeld residual plots; there was no evidence that these variables violated the assumption of proportional hazards.

3. Results

The case cohort was comprised of women from diverse racial/ ethnic backgrounds (Table 1). A majority of women were diagnosed with breast cancer at age 45 years or older (73%), or at an early stage (AJCC stage I and II) (88%). The subset of women with WHR measures had similar distributions for most characteristics as the full case cohort, with a few exceptions (Table 1). In the WHR subset, higher proportions of women identified as Hispanic (52%), or reported being physically inactive (49%), and a higher proportion of deaths was due to non-breast cancer causes (45%). For the total case cohort, most women were overweight or obese in the reference year (57%) and did not experience a weight change (51%); in the subset with WHR data, half the women had a WHR > 0.82 (Table 2).

Table 1.

Characteristics of breast cancer patients with body mass index and waist-to-hip ratio measures, Neighborhoods and Breast Cancer Study, 1995–2008 (N = 4347).

Body mass index
(N= 4347)
 Waist-to-hip ratio
(N= 1916)
N % N %
Age at diagnosis (years)
  < 35 298 6.9% 0 0.0%
  35–44 862 19.8% 384 20.0%
  45–54 1439 33.1% 593 31.0%
  55–64 1222 28.2% 480 25.1%
  ≥ 65 526 12.1% 459 24.0%
AJCC stage at diagnosis
  I 1899 43.7% 867 45.3%
  II 1930 44.4% 864 45.1%
  III 306 7.0% 104 5.4%
  IV 79 1.8% 26 1.4%
  Unknown 133 3.1% 55 2.9%
Tumor estrogen and progesterone receptor status
  ER−PR− 921 21.2% 380 19.8%
  ER+ or PR+ 3003 69.1% 1339 69.9%
  Unknown 423 9.7% 197 10.3%
Subsequent primary tumor
  No 3722 85.6% 1598 83.4%
  Yes 625 14.4% 318 16.6%
Type of surgery
  None 89 2.0% 34 1.8%
  Lumpectomy 2375 54.6% 1054 55.0%
  Mastectomy 1882 43.3% 828 43.2%
  Unknown 1 0.0% 0 0.0%
Chemotherapy
  No 1895 43.6% 956 49.9%
  Yes 2400 55.2% 937 48.9%
  Unknown 52 1.2% 23 1.2%
Radiation
  No 1788 41.1% 780 40.7%
  Yes 2559 58.9% 1136 59.3%
Study recruitment
  San Francisco Bay Area Breast Cancer Study (SFBCS) 2075 47.7% 1916 100.0%
  Northern California site of the Breast Cancer Family Registry (NC-BCFR) 2272 52.3% 0 0.0%
Race/ethnicity
  African American 975 22.4% 424 22.1%
  Asian American 667 15.3% 1 0.1%
  Hispanic 1646 37.9% 1003 52.3%
  Non-Hispanic white 1059 24.4% 488 25.5%
Education
  Less than high school 825 19.0% 471 24.6%
  High school graduate 772 17.8% 421 22.0%
  Some college 1403 32.3% 574 30.0%
  College graduate or post graduate 1318 30.3% 424 22.1%
  Unknown 29 0.7% 26 1.4%
Marital status
  Single 794 18.3% 292 15.2%
  Married 2609 60.0% 1123 58.6%
  Separated or divorced 547 12.6% 234 12.2%
  Widowed 293 6.7% 208 10.9%
  Unknown 104 2.4% 59 3.1%
History of benign breast disease
  No 3572 82.2% 1523 79.5%
  Yes 772 17.8% 390 20.4%
  Unknown 3 0.1% 3 0.2%
Menopausal status
  Premenopausal 1516 34.9% 628 32.8%
  Postmenopausal 2560 58.9% 1141 59.6%
  Unknown 271 6.2% 147 7.7%
History of menopausal hormone therapy use
  Never 2849 65.5% 1124 58.7%
  Former 715 16.4% 290 15.1%
  Current 783 18.0% 502 26.2%
Alcohol consumption in reference year (g/day)
  0 2650 61.0% 1017 53.1%
  < 5 644 14.8% 440 23.0%
  5–9.9 344 7.9% 110 5.7%
  10–14.9 264 6.1% 124 6.5%
  ≥ 15 429 9.9% 225 11.7%
  Unknown 16 0.4% 0 0.0%
Recent recreational physical activity (hrs/wk) (quartiles)a
  None 1444 33.2% 943 49.2%
  Q1: 0.01–1.92 721 16.6% 263 13.7%
  Q2: 1.93–3.00 772 17.8% 189 9.9%
  Q3: 3.01–6.38 676 15.6% 266 13.9%
  Q4: 46.39 730 16.8% 255 13.3%
  Unknown 4 0.1% 0 0.0%
Vital status through December 31, 2009
  Alive 3452 79.4% 1427 74.5%
  Deceased 895 20.6% 489 25.5%
  % of deaths due to breast cancerb 560 62.6% 267 54.6%
Open in a new tab
a

Based on the quartile distribution among all cases in study population with non-zero values.

b

Percentages calculated using deceased cases as denominators.

Table 2.

Body size associations with survival after breast cancer diagnosis: hazard ratios (HR) with 95% confidence intervals (CI), adjusting for clinical and individual-level characteristics, Neighborhoods and Breast Cancer Study, 1995–2008 (N = 4347).

Total case cohort
 Overall mortality
 Breast Cancer-Specific Mortality
No. cases % No. of deaths % HRa 95% CI HRb 95% CI No. of deaths % HRa 95% CI HRb 95% CI
Pre-diagnosis BMI (kg/m2)
    ≤ 18.5: Underweight 78 2% 18 2% 1.43 0.92–2.23 1.55 0.97–2.48 9 2% 0.92 0.48–1.76 0.94 0.48–1.84
    18.6–24.9: Normal 1795 41% 323 36% 1.00 1.00 226 40% 1.00 1.00
    25–29.9: Overweight 1264 29% 261 29% 1.08 0.91–1.28 1.05 0.88–1.26 154 28% 1.02 0.82–1.27 1.03 0.82–1.29
    ≥ 30: Obese 1210 28% 293 33% 1.21 1.02–1.43 1.11 0.92–1.33 171 31% 1.13 0.91–1.40 1.08 0.86–1.36
     p trendc 0.03 0.23 0.26 0.42
% Weight changed
    Decreased, ≥ 2% 604 14% 174 21% 1.29 0.98–1.69 1.23 0.92–1.63 87 17% 1.34 0.96–1.88 1.35 0.94–1.95
    Stable, ± 1% 2112 51% 339 42% 1.00 1.00 247 48% 1.00 1.00
    Increased, 2–10% 784 19% 154 19% 0.84 0.63–1.12 0.88 0.65–1.20 81 16% 0.98 0.67–1.42 1.09 0.73–1.63
    Increased, > 10% 668 16% 148 18% 0.88 0.66–1.18 0.89 0.65–1.21 99 19% 1.09 0.75–1.59 1.19 0.79–1.79
     p trende 0.54 0.61 0.35 0.25
Waist-to-hip ratio (quartiles)f
    Q1: ≤ 0.77 482 25% 96 20% 1.00 1.00 59 22% 1.00 1.00
    Q2: 0.78–0.82 484 25% 109 22% 1.15 0.86–1.53 1.20 0.89–1.62 65 24% 1.22 0.84–1.76 1.29 0.87–1.90
    Q3: 0.83–0.86 479 25% 130 27% 1.48 1.07–1.83 1.33 0.99–1.78 72 27% 1.42 0.99–2.03 1.39 0.94–2.07
    Q4: ≥ 0.87 471 25% 154 31% 1.68 1.28–2.22 1.65 1.20–2.26 71 27% 1.44 0.99–2.09 1.62 1.06–2.48
     p trend <0.01 <0.01 0.04 0.03
Open in a new tab
a

Adjusted for age at diagnosis (continuous), year of diagnosis (continuous calendar year), race/ethnicity (non-Hispanic white, African American, Hispanic, Asian American), clustering by block group, and stratified by study (SFBCS, NC-BCFR) and AJCC stage (I, II, III, IV, unknown) except WHR models which were not stratified by study as all women were from SFBCS.

b

Adjusted for age at diagnosis (continuous), year of diagnosis (continuous calendar year), race/ethnicity (non-Hispanic white, African American, Hispanic, Asian American), histology (ductal, lobular, other), histological grade (1, 2, 3 or 4,unknown), ERPR status (ER−PR−, ER+ or PR+, unknown), first subsequent primary tumor (no, yes), time to first subsequent primary tumor (months, continuous), type of surgery (none, lumpectomy, mastectomy, unknown), chemotherapy (no, yes, unknown), radiation (no, yes), marital status (single, married, separated/divorced, widowed, unknown), education (less than high school (HS), HS graduate, vocational/technical school or some college, college graduate or graduate school, unknown), history of benign breast disease (no, yes, unknown), menopausal status (premenopausal, postmenopausal, unknown), age at menarche (< 12,12,13, ≥ 14), number of full-term pregnancies (0,1, 2, 3, ≥ 4), months of breastfeeding (nulliparous, 0, < 12,12–23, ≥ 24, unknown), years since last full-term pregnancy (< 2,2–4, ≥ 5, unknown), history of hormonal contraception use (never, ever, unknown), history of menopausal hormone therapy use (never, former, current, unknown), recent recreational physical activity (0, Q1, Q2, Q3, Q4), alcohol consumption in grams/day (0, < 5,5–9,10–14, ≥ 15, unknown), and clustering by block group, and stratified by study (SFBCS, NC-BCFR) and AJCC stage (I, II, III, IV, unknown) except WHR models which were not stratified by study as all women were from SFBCS. If the main effect variable is % weight change or waist-to-hip ratio, the models are further adjusted for pre-diagnosis BMI (underweight, normal, overweight, obese).

c

Trend excludes those with BMI ≤ 18.5 (underweight).

d

Percent change between pre- (reference year) and post-diagnosis (interview) weight.

e

Trend excludes those with decreased weight change.

f

Sample size for these analyses is 1916.

3.1. Body size and survival

3.1.1. Body mass index (BMI)

Women who were obese (versus normal weight) in the year before diagnosis had higher overall mortality in base (HR=1.21, 95% CI = 1.02–1.43, p-trend=0.03), but not in the fully-adjusted regression models. No association with pre-diagnosis BMI and BC-specific mortality was observed (Table 2).

3.1.2. Percent weight change

No associations with percent weight change were observed for overall mortality or BC-specific mortality (Table 2).

3.1.3. Waist-to-hip ratio (WHR)

Compared to women in the lowest WHR quartile, those with higher WHRs had higher overall mortality in the fully adjusted model (quartile 3: HR=1.33, 95% CI=0.99–1.78; quartile 4: HR=1.65, 95% CI =1.20–2.26, p-trend < 0.01). Similar associations were observed for BC-specific mortality (highest vs. lowest quartile: HR=1.62, 95% CI= 1.06–2.48, p-trend=0.03) (Table 2).

We also examined how subsets of the covariates in the fully adjusted model impacted the hazard ratio (HR) among women with Q4 versus Q1 WHR, and found that the driving factor is treatment, in particular, surgery (data not shown).

3.2. Neighborhood associations with WHR

Of women with WHR measures, the majority resided in neighborhoods of higher SES (62%) and higher population density (68%) (Table 3). In fully-adjusted models, only nSES, crowding, and Restaurant Environment Index remained significantly associated with higher WHR. Residing in lower SES neighborhoods was associated with over two times the odds of having higher WHRs (lowest vs. highest nSES: OR=2.54, 95% CI=1.26–5.11, p-trend < 0.01). Similar associations were observed for neighborhoods with more crowded housing (highest vs. lowest quartile: OR=1.70, 95% CI=1.02–2.82, p-trend=0.04). Lack of fast food was suggestively associated with lower WHR (No fast food restaurants vs. <median ratio of fast food restaurants to other restaurants OR=0.70, 95% CI=0.48–1.02).

Table 3.

Association of neighborhood characteristics with high waist-to-hip ratioa: multivariable odds ratios (OR) with 95% confidence intervals (CI), Neighborhoods and Breast Cancer Study, 1995–2008 (N =;1916).

N % Model 1: OR, 95% CIb Model 2: OR, 95% CIc
Neighborhood socioeconomic status (SES) (quintiles)d
    Q5: > 0.84 (high SES) 694 36.2% 1.00 1.00
    Q4: 0.23–0.84 493 25.7% 1.47 1.15–1.88 1.17 0.88–1.56
    Q3: −0.30−0.22 358 18.7% 2.11 1.59–2.80 1.45 1.01–2.07
    Q2: −0.90 to −0.31 277 14.5% 2.45 1.80–3.34 1.63 1.07–2.48
    Q1: <−0.90 (low SES) 94 4.9% 4.25 2.40–7.55 2.54 1.26–5.11
     p trend < 0.01 < 0.01
Population density (persons/square meter) (quartiles)d
    Q1: < 0.00108 239 12.5% 1.00 1.00
    Q2: 0.00108–0.00256 380 19.8% 1.13 0.80–1.61 1.01 0.68–1. 51
    Q3: 0.00257–0.00428 561 29.3% 1.45 1.04–2.02 1.05 0.69–1.58
    Q4: > 0.00428 736 38.4% 2.09 1.52–2.89 1.20 0.74–1.94
     p trend < 0.01 0.46
Percentage of non-single family units (quartiles)e
    Q1: < 3.6 479 25.0% 1.00 1.00
    Q2: 3.6–23.3 468 24.4% 1.27 0.97–1.66 1.07 0.81–1.43
    Q3: 23.4–51.8 475 24.8% 1.69 1.29–2.21 1.11 0.81–1.52
    Q4: > 51.8 494 25.8% 1.73 1.32–2.28 1.10 0.78–1.56
     p trend < 0.01 0.58
Percentage of occupied housing units with more than one occupant per room (crowding) (quartiles)e
    Q1: < 3.29 450 23.5% 1.00 1.00
    Q2: 3.30–9.80 481 25.1% 1.53 1.16–2.03 1.17 0.86–1.60
    Q3: 9.81–21.11 472 24.6% 1.93 1.45–2.57 1.34 0.91–1.96
    Q4: > 21.11 513 26.8% 2.95 2.20–3.95 1.70 1.02–2.82
     p trend < 0.01 0.04
Percentage of foreign-born residents (quartiles)e
    Q1: < 15.9 463 24.2% 1.00 1.00
    Q2: 15.9–26.3 495 25.8% 1.34 1.02–1.76 1.00 0.72–1.39
    Q3: 26.4–41.5 473 24.7% 1.66 1.24–2.21 1.06 0.71–1.58
    Q4: 441.5 485 25.3% 1.89 1.42–2.52 0.99 0.61–1.63
     p trend < 0.01 0.97
Percentage of linguistically isolated (quartiles) e
    Q1: < 3.01 457 23.9% 1.00 1.00
    Q2: 3.01–7.32 500 26.1% 1.49 1.13–1.96 1.12 0.81–1.55
    Q3: 7.33–13.96 468 24.4% 1.76 1.32–2.34 0.99 0.66–1. 4 8
    Q4: > 13.96 491 25.6% 2.16 1.63–2.87 0.91 0.56–1. 4 9
     p trend < 0.01 0.64
Block Group-level Gamma (quartiles) e
    Q1: < 0.40 443 23.1% 1.00 1.00
    Q2: 0.40–0.43 475 24.8% 1.36 1.03–1.79 1.09 0.79–1. 4 8
    Q3: 0.44–0.48 485 25.3% 1.72 1.30–2.28 1.27 0.91–1.78
    Q4: > 0.48 513 26.8% 1.89 1.43–2.48 1.22 0.84–1.78
     p trend < 0.01 0.21
Traffic density within 500 m of residence (vehicle miles traveled per square mile) (quartiles) e
    Q1: < 31,280 426 22.2% 1.00 1.00
    Q2: 31,281–60,581 462 24.1% 1.36 1.03–1.81 1.15 0.84–1. 57
    Q3: 60,582–99,608 478 24.9% 1.43 1.07–1.90 1.03 0.73–1. 4 4
    Q4: > 99,608 490 25.6% 1.69 1.28–2.24 1.09 0.75–1.59
    Unknown 60 3.1% 0.89 0.51–1.53 1.04 0.56–1.93
     p trendf < 0.01 0.93
Restaurant Environment Index within 1600 m of residenceg
    0 (No fast-food restaurants) 492 25.7% 0.54 0.41–0.70 0.70 0.48–1.02
    M1:< 0.11 625 32.6% 1.00 1.00
    M2:> 0.11 665 34.7% 0.98 0.78–1.23 0.99 0.75–1. 31
    No restaurants 134 7.0% 0.53 0.36–0.78 1.10 0.58–2.09
     p trendh < 0.01 0.04
Retail Food Environment Index within 1600 meters of residencei
    0 233 12.2% 1.00 1.00
    < 1 1,186 61.9% 1.79 1.30–2.45 1.05 0.68–1.63
    ≥ 1 346 18.1% 1.48 1.03–2.12 1.12 0.71–1.75
    No retail food outlets 151 7.9% 0.69 0.43–1.10 0.62 0.35–1.10
     p trendj 0.11 0.06
Number of total businesses within 1600 m (quartiles) e
    Q1: < 68 424 22.1% 1.00 1.00
    Q2: 681–132 460 24.0% 1.27 0.97–1.66 0.79 0.55–1.14
    Q3: 133–258 484 25.3% 1.55 1.18–2.03 0.84 0.54–1.28
    Q4: > 258 548 28.6% 1.70 1.29–2.25 0.92 0.53–1.59
   ptrend < 0.01 0.86
Number of farmer’s markets within 1600 m
    0 1389 72.5% 1.00 1.00
    1–2 483 25.2% 1.25 1.00–1.56 0.96 0.72–1. 27
    3+ 44 2.3% 1.35 0.72–2.41 0.93 0.46–1.87
     p trend 0.04 0.85
Number of recreational facilities within 1600 m (quartiles) e
    Q1: < 2 552 28.8% 1.00 1.00
    Q2: 2–3 484 25.3% 1.16 0.89–1.50 0.97 0.71–1.32
    Q3: 4–7 488 25.5% 1.08 0.83–1.40 0.82 0.58–1.16
    Q4: > 7 392 20.5% 1.34 1.02–1.77 0.94 0.60–1.45
     p trend 0.07 0.48
Urban/Rural Statusk
    Metropolitan urban 515 26.9% 1.00 1.00
    Metropolitan suburban 1,104 57.6% 0.68 0.55–0.85 1.15 0.81–1. 6 4
    City 282 14.7% 0.54 0.39–0.74 1.01 0.62–1.65
    Town/Rural 15 0.8% 0.44 0.12–1.58 1.06 0.29–3.85
Open in a new tab
a

Waist-to-hip ratio was categorized into high (>median) vs. low (≤ median).

b

Adjusted for age at diagnosis (continuous), study (SFBCS, NC-BCFR), race/ethnicity (non-Hispanic white, African American, Hispanic, Asian American), AJCC stage (I, II, III, IV, unknown), and clustering by block group.

c

Adjusted for all covariates in Model 1 and all neighborhood variables shown in the table, and clustering by block group.

d

Based on the quintile/quartile distribution for block groups in California.

e

Based on the quartile distribution among all study cases.

f

Does not include unknown category.

g

For the Restaurant Environment index, 0 indicates a neighborhood with no fast food restaurants; for neighborhoods with fast food restaurants, we used the median value of the ratio of fast foods to other restaurants to split the sample into those living in neighborhoods with relatively fewer fast foods to other restaurants (M1) and those living in neighborhoods with relatively more fast foods to other restaurants (M2). M2 includes those who have a numerator value > 0 and a denominator=0.

h

Does not include no restaurants category.

i

For the Retail Food Environment Index, 0 indicates that the neighborhood has no unhealthy food outlets, a ratio of < 1 indicates that there are fewer unhealthy food outlets compared to healthy food outlets, where as a ratio greater than 1 indicates that there are more unhealthy food outlets compared to healthy ones.

j

Does not include no retail food outlets category.

k

Urban/rural status is derived from census defined Urbanized Areas (population ≥ 50,000) and Urban Clusters (population between 2500 and 50,000). Classification is performed at the Block level. Blocks are then aggregated to Block Groups according to the dominant classification (by population). Blocks in Urbanized Areas with population ≥ 1,000,000 are classified as Metropolitan. Those blocks are further classified based on population density with the highest quartile being classified as Metropolitan Urban and the remaining as Metropolitan Suburban. The remaining Blocks in Urbanized Areas (population between 50,000 and 1,000,000) are classified as City. Blocks in Urban Clusters (population between 2500 and 50,000) and not in the lowest quartile of population density are classified as Town. The remaining blocks are classified as Rural (in Urban Clusters and the lowest quartile of population density or in neither Urbanized Areas nor Urban Clusters).

3.3. WHR, neighborhood, and survival

While we observed associations between specific neighborhood characteristics and overall mortality in base models (see Table 4, Model 1), no associations remained in models that additionally adjusted for tumor, treatment and personal characteristics, as well as all other neighborhood characteristics (Table 4, Models 2 and 3). WHR remained associated with higher overall mortality in models adjusting for neighborhood characteristics (highest vs. lowest quartile: HR =1.64, 95% CI=1.19–2.25; p-trend < 0.01). Results were similar for BC-specific mortality (highest vs. lowest quartile: HR=1.63, 95% CI =1.05–2.53; p-trend=0.03).

Table 4.

Association of waist-to-hip ratio, neighborhood features and survival after breast cancer diagnosis: multivariable hazard ratios (HR) with 95% confidence intervals (CI), Neighborhoods and Breast Cancer Study, 1995–2008 (N=1916).

Overall mortality
 Breast Cancer-Specific Mortality
Model 1: HR 95% CIa Model 2: HR 95% CIb Model 3: HR 95% CIc Model 1: HR 95%
CIa 		Model 2: HR 95%
CIb 		Model 3: HR 95% CIc
Waist-to-hip ratio (quartiles)
    Q1: ≤ 0.77 1.00 1.00 1.00 1.00 1.00 1.00
    Q2: 0.78–0.82 1.15 0.86–1.53 1.14 0.86–1.53 1.21 0.90–1.64 1.22 0.84–1.76 1.21 0.84–1.75 1.32 0.89–1. 97
    Q3: 0.83–0.86 1.40 1.07–1.83 1.34 1.02–1.76 1.32 0.98–1.77 1.42 0.99–2.03 1.39 0.96–2.02 1.43 0.95–2.16
    Q4: ≥ 0.87 1.68 1.28–2.22 1.62 1.23–2.13 1.64 1.19–2.25 1.44 0.99–2.09 1.41 0.96–2.06 1.63 1.05–2.53
     p trend < 0.01 < 0.01 < 0.01 0.04 0.06 0.03
Neighborhood socioeconomic status (SES) (quintiles)d
    Q1: <−0.90 (low SES) 1.73 1.17–2.55 1.29 0.78–2.14 1.15 0.71–1.88 1.27 0.73–2.20 0.89 0.42–1.85 0.83 0.41–1. 6 8
    Q2: −0.90 to −0.31 1.42 1.05–1.91 1.17 0.81–1.70 1.02 0.69–1.49 1.27 0.85–1.88 1.03 0.61–1.75 1.02 0.58–1.79
    Q3: −0.30–0.22 1.18 0.90–1.54 1.00 0.72–1.39 0.99 0.71–1.39 1.00 0.70–1.44 0.87 0.54–1.39 0.94 0.59–1.52
    Q4: 0.23–0.84 1.00 0.78–1.29 0.87 0.66–1.16 0.79 0.59–1.04 0.92 0.65–1.29 0.80 0.54–1.20 0.79 0.52–1.18
    Q5: > 0.84 (high SES) 1.00 1.00 1.00 1.00 1.00 1.00
     p trend < 0.01 0.21 0.47 0.21 0.93 0.99
Percentage of occupied housing units with more than one occupant per room (crowding) (quartiles)e
    Q1: < 3.29 1.00 1.00 1.00 1.00 1.00 1.00
    Q2: 3.30–9.80 0.97 0.75–1.26 1.21 0.91–1.62 0.97 0.73–1.27 0.88 0.61–1.27 0.85 0.59–1.23 1.02 0.70–1. 4 8
    Q3: 9.81–21.11 1.09 0.84–1.42 1.23 0.86–1.76 0.93 0.67–1.30 0.89 0.62–1.26 0.79 0.53–1.19 0.79 0.50–1.25
    Q4: > 21.11 1.30 1.01–1.68 1.07 0.70–1.66 1.12 0.78–1.61 1.31 0.94–1.83 1.24 0.80–1.92 1.27 0.78–2.05
     p trend 0.03 0.73 0.61 0.11 0.42 0.49
Percentage of linguistically isolated (quartiles)e
    Q1:< 3.01 1.00 1.00 1.00 1.00 1.00 1.00
    Q2:3.01–7.32 1.20 0.92–1.57 1.21 0.91–1.62 1.15 0.85–1.56 0.86 0.59–1.25 0.94 0.62–1.41 0.90 0.59–1.38
    Q3:7.33–13.96 1.31 1.00–1.72 1.23 0.86–1.76 1.20 0.82–1.75 1.28 0.89–1.83 1.34 0.84–2.16 1.34 0.80–2.24
    Q4: > 13.96 1.41 1.08–1.86 1.07 0.70–1.66 1.01 0.63–1.59 1.17 0.81–1.68 0.97 0.55–1.72 0.87 0.47–1.60
   ptrend 0.01 0.73 0.94 0.14 0.75 0.95
Traffic density within 500 meters of residence (vehicle miles traveled per square mile) (quartiles)e
    Q1:80 1.00 1.00 1.00 1.00 1.00 1.00
    Q2: 31,281–60,581 1.00 0.76–1.31 0.94 0.70–1.24 0.93 0.69–1.25 0.87 0.59–1.28 0.87 0.58–1.28 0.89 0.58–1.36
    Q3: 60,582–99,608 1.31 1.01–1.69 1.18 0.88–1.57 1.15 0.85–1.56 1.42 1.02–1.97 1.36 0.94–1.97 1.32 0.89–1.95
    Q4:> 99,608 1.25 0.97–1.61 1.11 0.81–1.50 1.07 0.77–1.47 1.14 0.81–1.63 1.08 0.71–1.64 1.03 0.65–1.63
    Unknown 1.14 0.69–1.89 1.16 0.69–1.98 1.42 0.82–2.45 0.79 0.34–1.84 0.83 0.35–1.98 1.01 0.39–2.67
   ptrendf 0.02 0.22 0.36 0.09 0.22 0.34
Restaurant Environment Index within 1600 meters of residence
    0 0.90 0.71–1.14 1.15 0.88–1.51 1.13 0.85–1.51 0.91 0.66–1.25 1.10 0.76–1.58 1.04 0.71–1.54
    M1: −0.11 1.00 1.00 1.00 1.00 1.00 1.00
    M2:> 0.11 1.13 0.91–1.41 1.21 0.96–1.53 1.17 0.92–1.49 0.97 0.72–1.31 1.01 0.73–1.40 1.00 0.71–1.40
    No Restaurants 0.94 0.66–1.33 1.25 0.80–1.95 1.13 0.70–1.81 0.92 0.56–1.51 1.22 0.68–2.18 1.02 0.55–1. 91
     p trendg 0.04 0.51 0.67 0.72 0.60 0.80
Number of farmers’ markets within 1600 meters
    0 1.00 1.00 1.00 1.00 1.00 1.00
    1–2 1.17 0.96–1.40 1.13 0.91–1.40 1.09 0.88–1.36 1.15 0.88–1.50 1.13 0.84–1.50 1.17 0.86–1.58
    ≥ 3 1.46 0.87–2.44 1.33 0.74–2.41 1.21 0.62–2.36 1.05 0.52–2.11 1.03 0.47–2.26 0.91 0.35–2.36
     p trend 0.05 0.20 0.39 0.38 0.51 0.52
Open in a new tab
a

Adjusted for age at diagnosis (continuous), year of diagnosis (continuous calendar year), race/ethnicity (non-Hispanic white, African American, Hispanic, Asian American), clustering by block group, and stratified by AJCC stage (I, II, III, IV, unknown).

b

Adjusted for all variables in Model 1 including all neighborhood variables shown in the table, waist-to-hip ratio, clustering by block group, and stratified by AJCC stage (I, II, III, IV, unknown).

c

Adjusted for all variables in Model 2 and histology (ductal, lobular, other), histological grade (1, 2, 3 or 4, unknown), ERPR status (ER−PR−, ER+ or PR+, unknown), first subsequent primary tumor (no, yes), time to first subsequent primary tumor (months, continuous), type of surgery (none, lumpectomy, mastectomy, unknown), chemotherapy (no, yes, unknown), radiation (no, yes), marital status (single, married, separated/divorced, widowed, unknown), education (less than high school (HS), HS graduate, vocational/technical school or some college, college graduate or graduate school, unknown), history of benign breast disease (no, yes, unknown), menopausal status (premenopausal, postmenopausal, unknown), age at menarche (< 12,12,13, ≥ 14), number of full-term pregnancies (0,1, 2, 3, ≥ 4), months of breastfeeding (nulliparous, 0, < 12,12–23, ≥ 24, unknown), years since last full-term pregnancy (< 2,2–4, ≥ 5, unknown), history of hormonal contraception use (never, ever, unknown), history of menopausal hormone therapy use (never, former, current, unknown), recent recreational physical activity (0, Q1, Q2, Q3, Q4), alcohol consumption in grams/day (0, < 5, 5–9,10–14, ≥ 15, unknown), pre-diagnosis BMI (underweight, normal, overweight, obese), clustering by block group, and stratified by study (SFBCS, NC-BCFR) and AJCC stage (I, II, III, IV, unknown).

d

Based on the quintile distribution for block groups in California.

e

Based on the quartile distribution among all study cases.

f

Does not include unknown category.

g

Does not include no restaurants category.

4. Discussion

In this study of racial/ethnically diverse women with breast cancer and data on clinical and tumor characteristics, personal factors, and social and built environment neighborhood characteristics, we found that WHR was independently associated with both overall and BC-specific mortality. These findings are consistent with prior studies (Protani et al., 2010; Kwan et al., 2014). In addition, like our study, others also did not observe that body size/ survival associations varied by race/ethnicity (Conroy et al., 2011; Kwan et al., 2012). Furthermore, we found that lower nSES and more household crowding, were associated with higher WHR, but not with survival, after adjustment for tumor, treatment, and personal characteristics and other neighborhood characteristics. Our WHR findings contribute to the growing literature on WHR as an important, modifiable prognostic factor that can be intervened upon (e.g., diet and/or physical activity programs or more regular follow-up for recurrence or other comorbidities).

Our findings of higher WHR associated with higher overall and BC-specific mortality are consistent with two of three studies that examined these associations (Protani et al., 2010; Kwan et al., 2014; Chen et al., 2010). A meta-analysis found that higher WHR was associated with higher BC-specific mortality (pooled HR across 4 studies=1.31; 95% CI=1.14–1.50) (Protani et al., 2010). The California Breast Cancer Survivorship Consortium, which included data from SFBCS and 5 other studies, also showed that higher WHR was associated with higher risk of both overall (among all women, African Americans, and Asian Americans) and breast cancer-specific (among Asian Americans) mortality (Kwan et al., 2014). Our finding of a borderline association between percent weight loss (≥ 2%) and mortality is consistent with prior studies; however, the more modest association in our study compared to others may be due to variability in the timing of post-diagnosis weight measurement as well as the influence of treatment, such as chemotherapy, on weight across studies (Vance et al., 2011; Chen et al., 2010; Caan et al., 2008). Conversely, we found that BMI in the year before breast cancer diagnosis, unlike in most prior studies (Hauner et al., 2011; Protani et al., 2010; Chen et al., 2010; Caan et al., 2008; Conroy et al., 2011; Kwan et al., 2012, 2014), was not associated with survival after adjusting for tumor characteristics, treatment and personal factors. While BMI is the most commonly used body size measure, some evidence suggests that it may not be the best measure, particularly in multiethnic populations (Protani et al., 2010; Kwan et al., 2014; Boeing, 2013). Our finding of an association with WHR illustrates the importance of considering multiple measures of body size to assess associations with survival among diverse racial/ethnic populations of breast cancer patients.

This is the first study to demonstrate that social and built environment factors were associated with WHR among women with breast cancer. We demonstrated that lower nSES, and more household crowding were associated with higher WHR while the lack of fast food restaurants was suggestively associated with lower WHR. Similar associations for nSES and restaurants environment have been shown in prior studies among non-cancer populations (Keller et al., 2013; Xu et al., 2013,; haix et al., 2008). However, in analyses that considered the associations of neighborhood characteristics and WHR with survival, only WHR remained associated with survival. To assess whether WHR attenuated associations between neighborhood characteristics and survival, we modeled neighborhood factors without WHR and did not find significant associations, suggesting that WHR was not mediating associations between neighborhood and mortality (data not shown). Furthermore, the lack of association with nSES once we accounted for other neighborhood factors may also have resulted from interactions with other neighborhood characteristics, as found previously for nSES and ethnic enclaves (Keegan et al., 2010). Although statistical power was limited to detect such interactions in this study and there was no evidence of multi-collinearity in the fully adjusted Cox regression models, distributions of the other neighborhood characteristics by nSES suggest that participants living in low-SES neighborhoods were also living in neighborhoods with higher traffic density, more crowding and more linguistically isolated households (data not shown). Future research is needed to determine the potential pathways through which neighborhood features (e.g., SES, housing and food environment) and individual factors (e.g. body size) may contribute to survival after breast cancer diagnosis.

Our study is subject to some limitations. As weight was a self-reported measure, it may be sensitive to inaccurate recall; however, correlation of self-reported and measured weight in a subset of participants who had both measures was very high (r=0.84). Other covariates such as alcohol consumption and physical activity also were self-reported and were not validated though these measures have been extensively used in prior studies (Block G et al., 1986, 1990; Bernstein et al., 1994; John et al., 2003; Yang et al., 2003; Dallal et al., 2007; West-Wright et al., 2009; Keegan et al., 2014). Also, because WHR was only available for participants in the SFBCS study, the statistical power for the WHR analysis was limited and our findings may not be generalizable to Asian Americans. However, we did not find race/ethnicity to modify the body size and survival findings, in agreement with two other studies (Conroy et al., 2011; Kwan et al., 2012). Due to data availability, some of our neighborhood measures were based on more contemporary data (e.g., farmers markets) that may not reflect neighborhoods prior to this time. We did not have perceived or audit neighborhood measures to assess quality and use by study participants. Lastly, while our study sample is representative of women with breast cancer in the San Francisco Bay Area, the findings may not be generalizable to other geographic regions across the country. Despite these limitations, this study considers social and built environment features at a small geography using objective measures from secondary data sources for the study of breast cancer survival. Additional strengths include a population-based design, high response rates from participants, and a racially/ethnically diverse study population. A large number of prognostic factors from both interview and clinical sources were considered and bias due to differential follow-up was minimized by linkage to population-based cancer registries and death registry records.

Our findings indicate that future research on modifiable prognostic factors after breast cancer diagnosis should consider body size measures beyond BMI, such as WHR, which may better characterize distribution of adiposity among diverse groups of women (Protani et al., 2010; Kwan et al., 2014; Boeing, 2013). We also found that certain neighborhood characteristics were associated with WHR. These findings could be used, along with WHR, to identify a priority subgroup of breast cancer survivors that might benefit from lifestyle interventions or increased medical surveillance that aim to improve their WHR and survival after diagnosis. Interventions aimed at improving WHR need to take into consideration neighborhood characteristics that can influence WHR and provide tailored resources and strategies that leverage neighborhood resources or overcome deficits.

Acknowledgements

The authors wanted to acknowledge Clayton W. Schupp and Myles Cockburn for their contributions to the Neighborhoods and Breast Cancer Study. This work was supported by National Cancer Institute funds from R21CA133255 (T.H.M.K) and R01CA140058 (S. L.G). The Breast Cancer Family Registry (BCFR) was supported by grant UM1 CA164920 from the USA National Cancer Institute. The content of this manuscript does not necessarily reflect the views or policies of the National Cancer Institute or any of the collaborating centers in the BCFR, nor does mention of trade names, commercial products, or organizations imply endorsement by the USA Government or the BCFR. The San Francisco Bay Area Breast Cancer Study was supported by funds (E.M.J.) from National Cancer Institute grants R01 CA63446 and R01 CA77305; U.S. Department of Defense Breast Cancer Research Program grant DAMD17-96-1-6071; and California Breast Cancer Research Program (CBCRP) grants 4JB-1106 and 7PB-0068. The collection of cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program under contract HHSN261201000140C awarded to the Cancer Prevention Institute of California, contract HHSN261201000035C awarded to the University of Southern California, and contract HHSN261201000034C awarded to the Public Health Institute; and the Centers for Disease Control and Prevention’s National Program of Cancer Registries, under agreement U58DP003862-01 awarded to the California Department of Public Health. The ideas and opinions expressed herein are those of the author(s) and endorsement by the State of California, Department of Public Health the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors is not intended nor should be inferred.

Financial disclosure

Dr. Scarlett Gomez reports receiving funding support from Genentech unrelated to this manuscript.

Appendix. Distribution of waist-to-hip ratio (WHR) by body mass index (BMI)1, Neighborhoods and Breast Cancer Study, 1995–2008 (N=1916)

Waist-to-hip ratio Body mass index
Underweight (≤ 18.5) Normal (18.6–24.9) Overweight (25–29.9) Obese (≥ 30) Total
≤ Median (≤ 0.82) 73.68% 72.73% 45.47% 29.61% 50.42%
>Median (>0.82) 26.32% 27.27% 54.53% 70.39% 49.58%
Total 19 682 607 608 1916
Open in a new tab
1

Waist and hip circumferences were measured at interview in SFBCS only.

Footnotes

Conflict of Interest

All authors have no potential conflicts of interest to disclose.

Contributor Information

Scarlett L. Gomez, Email: Scarlett.Gomez@cpic.org.

Meera Sangaramoorthy, Email: Meera.Sangaramoorthy@cpic.org.

Juan Yang, Email: Juan.Yang@cpic.org.

Jocelyn Koo, Email: Jocelyn.Koo@cpic.org.

Andrew Hertz, Email: Andrew.Hertz@cpic.org.

Esther M. John, Email: Esther.John@cpic.org.

Iona Cheng, Email: Iona.Cheng@cpic.org.

Theresa H.M. Keegan, Email: tkeegan@ucdavis.edu.

References

  1. American Cancer Society. Cancer Treatment and Survivorship Facts and Figures 2012–2013. Atlanta, GA: American Cancer Society; 2012. [Google Scholar]
  2. Bernstein L, Henderson BE, Hanisch R, Sullivan-Halley J, Ross RK. Physical exercise and reduced risk of breast cancer in young women. JNCI. 1994;86:1403–1408. doi: 10.1093/jnci/86.18.1403. [DOI] [PubMed] [Google Scholar]
  3. Berrigan D, Pickle LW, Dill J. Associations between street connectivity and active transportation. Int. J. Health Geogr. 2010;9:20. doi: 10.1186/1476-072X-9-20. http://dx.doi.org/10.1186/1476-072X-9-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Block G, Hartman AM, Dresser CM, Carroll MD, Ga Nnon J, Gardner L. A data-based approach to diet questionnaire design and testing. Am. J. Epidemiol. 1986;124:453–469. doi: 10.1093/oxfordjournals.aje.a114416. [DOI] [PubMed] [Google Scholar]
  5. Block G, Woods M, Potosky A, Clifford C. Validation of a self-administered diet history questionnaire using multiple diet records. J. Clin. Epidemiol. 1990;34:1327–1335. doi: 10.1016/0895-4356(90)90099-b. [DOI] [PubMed] [Google Scholar]
  6. Boeing H. Obesity and cancer-the update 2013. Best. Prac. Res. Clin. En-docrinol. Metab. 2013;27:219–227. doi: 10.1016/j.beem.2013.04.005. [DOI] [PubMed] [Google Scholar]
  7. Bradshaw PT, Ibrahim JG, Stevens J, Cleveland R, Abrahamson PE, Satia JA, Teitelbaum SL, Neugut AI, Gammon MD. Postdiagnosis change in bodyweight and survival after breast cancer diagnosis. Epidemiology. 2012;23:320–327. doi: 10.1097/EDE.0b013e31824596a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Caan BJ, Kwan ML, Hartzell G, Castillo A, Slattery ML, Sternfeld B, Weltzien E. Pre-diagnosis body mass index, post-diagnosis weight change, and prognosis among women with early stage breast cancer. Cancer Causes Control. 2008;19:1319–1328. doi: 10.1007/s10552-008-9203-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. California Cancer Registry. Cancer Reporting in California: Abstracting and Coding Procedures for Hospitals. California: Cancer Registry; 2009. p. 1. [Google Scholar]
  10. California Center for Public Health Advocacy, PolicyLink, and the UCLA Center for Health Policy Research. The Link Between Local Food Environments and Obesity and Diabetes. Designed for Disease; 2008. [Google Scholar]
  11. California Department of Food and Agriculture. California Certified Farmers’ Market Database. ed. California: Department of Food and Agriculture; 2010. [Google Scholar]
  12. California Department of Transportation. Highway Performance and Monitoring System. ed. California: Department of Transportation; 2004. [Google Scholar]
  13. Carmichael AR, Bates T. Obesity and breast cancer: a review of the literature. Breast. 2004;13:85–92. doi: 10.1016/j.breast.2003.03.001. [DOI] [PubMed] [Google Scholar]
  14. Chaix B, Ducimetiere P, Lang T, Haas B, Montaye M, Ruidavets JB, Arveiler D, Amouyel P, Ferrieres J, Bingham A, Chauvin P. Residential environment and blood pressure in the PRIME study: is the association mediated by body mass index and waist circumference? J. Hypertens. 2008;26:1078–1084. doi: 10.1097/HJH.0b013e3282fd991f. [DOI] [PubMed] [Google Scholar]
  15. Chen X, Lu W, Zheng W, Gu K, Chen Z, Zheng Y, Shu XO. Obesity and weight change in relation to breast cancer survival. Breast Cancer Res. Treat. 2010;122:823–833. doi: 10.1007/s10549-009-0708-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Conroy SM, Maskarinec G, Wilkens LR, White KK, Henderson BE, Kolonel LN. Obesity and breast cancer survival in ethnically diverse post-menopausal women: the Multiethnic Cohort Study. Breast Cancer Res. Treat. 2011;129:565–574. doi: 10.1007/s10549-011-1468-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dallal CM, Sullivan-Halley J, Ross RK, Wang Y, Deapen D, Horn-Ross PL, Reynolds P, Stram DO, Clarke CA, Anton-Culver H, Ziogas A, Peel D, West DW, Wright W, Bernstein L. Long-term recreational physical activity and risk of invasive and in situ breast cancer: the California teachers study. Arch. Intern. Med. 2007;167:408–415. doi: 10.1001/archinte.167.4.408. [DOI] [PubMed] [Google Scholar]
  18. Diez Roux AV, Mair C. Neighborhoods and health. Ann. NY Acad. Sci. 2010;1186:125–145. doi: 10.1111/j.1749-6632.2009.05333.x. [DOI] [PubMed] [Google Scholar]
  19. 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;16:175–190. doi: 10.1016/j.healthplace.2009.09.008. [DOI] [PubMed] [Google Scholar]
  20. Gomez SL, Glaser SL, Mcclure LA, Shema SJ, Kealey M, Keegan TH, Satar-iano WA. 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;22:631–647. doi: 10.1007/s10552-011-9736-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gomez SL, Shariff-Marco S, Derouen M, Keegan TH, Yen IH, Mujahid M, Satariano WA, Glaser SL. The impact of neighborhood social and built environment factors across the cancer continuum: current research, methodological considerations, and future directions. Cancer. 2015;121:2314–2330. doi: 10.1002/cncr.29345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gunier RB, Hertz A, Von Behren J, Reynolds P. Traffic density in California: socioeconomic and ethnic differences among potentially exposed children. J. Expo. Anal. Env. Epidemiol. 2003;13:240–246. doi: 10.1038/sj.jea.7500276. [DOI] [PubMed] [Google Scholar]
  23. Hauner D, Janni W, Rack B, Hauner H. The effect of overweight and nutrition on prognosis in breast cancer. Dtsch. Arztebl. Int. 2011;108:795–801. doi: 10.3238/arztebl.2011.0795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hu C, Yang X, Cormier JN, Chang GJ. Assessing the utility of cancer-registry-processed cause of death in calculating cancer-specific survival. Cancer. 2013;119:1900–1907. doi: 10.1002/cncr.27968. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. John EM, Hopper JL, Beck JC, Knight JA, Neuhausen SL, Senie RT, Ziogas A, Andrulis IL, Anton-Culver H, Boyd N, Buys SS, Daly MB, O’malley FP, Santella RM, Southey MC, Venne VL, Venter DJ, West DW, Whittemore AS, Seminara D. The Breast Cancer Family Registry: an infrastructure for cooperative multinational, interdisciplinary and translational studies of the genetic epidemiology of breast cancer. Breast Cancer Res. 2004;6:R375–R389. doi: 10.1186/bcr801. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. John EM, Horn-Ross PL, Koo J. Lifetime physical activity and breast cancer risk in a multiethnic population: the San Francisco Bay area breast cancer study. Cancer Epidemiol. Biomark. Prev. 2003;12:1143–1152. [PubMed] [Google Scholar]
  27. John EM, Miron A, Gong G, Phipps AI, Felberg A, Li FP, West DW, Whit-temore AS. Prevalence of pathogenic BRCA1 mutation carriers in 5 US racial/ethnic groups. JAMA. 2007;298:2869–2876. doi: 10.1001/jama.298.24.2869. [DOI] [PubMed] [Google Scholar]
  28. John EM, Phipps AI, Davis A, Koo J. Migration history, acculturation, and breast cancer risk in Hispanic women. Cancer Epidemiol. Biomark. Prev. 2005;14:2905–2913. doi: 10.1158/1055-9965.EPI-05-0483. [DOI] [PubMed] [Google Scholar]
  29. John EM, Phipps AI, Sangaramoorthy M. Body size, modifying factors, and postmenopausal breast cancer risk in a multiethnic population: the San Francisco Bay Area Breast Cancer Study. Springerplus. 2013;2:239. doi: 10.1186/2193-1801-2-239. http://dx.doi.org/10.1186/2193-1801-2-239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. John EM, Sangaramoorthy M, Phipps AI, Koo J, Horn-Ross PL. Adult body size, hormone receptor status, and premenopausal breast cancer risk in a multiethnic population: the San Francisco Bay Area breast cancer study. Am. J. Epidemiol. 2011;173:531–542. doi: 10.1093/aje/kwq345. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Keegan TH, John EM, Fish KM, Alfaro-Velcamp T, Clarke CA, Gomez SL. Breast cancer incidence patterns among California Hispanic women: differences by nativity and residence in an enclave. Cancer Epidemiol. Biomark. Prev. 2010;19:1208–1218. doi: 10.1158/1055-9965.EPI-10-0021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Keegan TH, Shariff-Marco S, Sangaramoorthy M, Koo J, Hertz A, Schupp CW, Yang J, John EM, Gomez SL. Neighborhood influences on recreational physical activity and survival after breast cancer. Cancer Causes Control. 2014;25:1295–1308. doi: 10.1007/s10552-014-0431-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Keller C, Todd M, Ainsworth B, Records K, Vega-Lopez S, Permana P, Coonrod D, Nagle Williams A. Overweight, obesity, and neighborhood characteristics among postpartum Latinas. J. Obes. 2013;2013:916468. doi: 10.1155/2013/916468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Krieger N. Theories for social epidemiology in the 21st century: an ecosocial perspective. Int. J. Epidemiol. 2001;30:668–677. doi: 10.1093/ije/30.4.668. [DOI] [PubMed] [Google Scholar]
  35. Kwan ML, Chen WY, Kroenke CH, Weltzien EK, Beasley JM, Nechuta SJ, Poole EM, Lu W, Holmes MD, Quesenberry CP, Jr, Pierce JP, Shu XO, Caan BJ. Pre-diagnosis body mass index and survival after breast cancer in the After Breast Cancer Pooling Project. Breast CancerRes. Treat. 2012;132:729–739. doi: 10.1007/s10549-011-1914-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kwan ML, John EM, Caan BJ, Lee VS, Bernstein L, Cheng I, Gomez SL, Henderson BE, Keegan TH, Kurian AW, Lu Y, Monroe KR, Roh JM, Shariff-Marco S, Sposto R, Vigen C, Wu AH. Obesity and mortality after breast cancer by race/ethnicity: the California Breast Cancer Survivorship Consortium. Am. J. Epidemiol. 2014;179:95–111. doi: 10.1093/aje/kwt233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lin DY, Wei LJ. The robust inference for the Cox Proportional Hazards Model. J. Am. Stat. Assoc. 1989;84:1074–1078. [Google Scholar]
  38. Liu H. Cochran-Armitage Trend Test using SAS. Rahway, NJ: Merck Research Labs, Merck & Co., Inc.; 2007. [Google Scholar]
  39. Meijer M, Rohl J, Bloomfied K, Grittner U. Do neighborhoods affect individual mortality? A systematic review and meta-analysis of multilevel studies. Soc. Sci. Med. 2012;74:1204–1212. doi: 10.1016/j.socscimed.2011.11.034. [DOI] [PubMed] [Google Scholar]
  40. Molarius A, Seidell JC. Selection of anthropometric indicators for classification of abdominal fatness - a critical review. Int. J. Obes. Relat. Metab. Disord. 1998;22:719–727. doi: 10.1038/sj.ijo.0800660. [DOI] [PubMed] [Google Scholar]
  41. Northridge ME, Sclar ED, Biswas P. Sorting out the connections between the built environment and health: a conceptual framework for navigating pathways and planning healthy cities. J. Urban Health. 2003;80:556–568. doi: 10.1093/jurban/jtg064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Protani M, Coory M, Martin JH. Effect of obesity on survival of women with breast cancer: systematic review and meta-analysis. Breast Cancer Res. Treat. 2010;123:627–635. doi: 10.1007/s10549-010-0990-0. [DOI] [PubMed] [Google Scholar]
  43. Reynolds P, Hurley SE, Quach AT, Rosen H, Von Behren J, Hertz A, Smith D. Regional variations in breast cancer incidence among California women, 1988–1997. Cancer Causes Control. 2005;16:139–150. doi: 10.1007/s10552-004-2616-5. [DOI] [PubMed] [Google Scholar]
  44. Shariff-Marco S, Yang J, John EM, Sangaramoorthy M, Hertz A, Koo J, Nelson DO, Schupp CW, Shema SJ, Cockburn M, Satariano WA, Yen IH, Ponce NA, Winkleby M, Keegan TH, Gomez SL. Impact of neighborhood and individual socioeconomic status on survival after breast cancer varies by race/ ethnicity: the Neighborhood and Breast Cancer Study. Cancer Epidemiol. Bio-mark. Prev. 2014;23:793–811. doi: 10.1158/1055-9965.EPI-13-0924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Thornton LE, Pearce JR, Kavanagh AM. Using Geographic Information Systems (GIS) to assess the role of the built environment in influencing obesity: a glossary. Int. J. Behav. Nutr. Phys. Act. 2011;8:71. doi: 10.1186/1479-5868-8-71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. U.S. Census Bureau. Census 2000 Summary File 3 Technical Documentation. 2002. [Google Scholar]
  47. Vance V, Mourtzakis M, Mccargar L, Hanning R. Weight gain in breast cancer survivors: prevalence, pattern and health consequences. Obes. Rev. 2011;12:282–294. doi: 10.1111/j.1467-789X.2010.00805.x. [DOI] [PubMed] [Google Scholar]
  48. Walls Associates. National Establishment Time-Series (NETS) Database, 2009 ed. Oakland, CA: Walls & Associates; 2008. [Google Scholar]
  49. West-Wright CN, Henderson KD, Sullivan-Halley J, Ursin G, Deapen D, Neu-hausen S, Reynolds P, Chang E, Ma H, Bernstein L. Long-term and recent recreational physical activity and survival after breast cancer: the California Teachers Study. Cancer Epidemiol. Biomark. Prev. 2009;18:2851–2859. doi: 10.1158/1055-9965.EPI-09-0538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. World Health Organization. Obesity: Preventing and Managing the Global Epidemic: Report of a WHO Consultation. Geneva, Switzerland: 2000. [PubMed] [Google Scholar]
  51. Xu H, Short SE, Liu T. Dynamic relations between fast-food restaurant and body weight status: a longitudinal and multilevel analysis of Chinese adults. J. Epidemiol. Commun. Health. 2013;67:271–279. doi: 10.1136/jech-2012-201157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Yang D, Bernstein L, Wu AH. Physical activity and breast cancer risk among Asian-American women in Los Angeles: a case-control study. Cancer. 2003;97:2565–2575. doi: 10.1002/cncr.11364. [DOI] [PubMed] [Google Scholar]
  53. Yen IH, Michael YL, Perdue L. Neighborhood environment in studies of health of older adults: a systematic review. Am. J. Prev. Med. 2009;37:455–463. doi: 10.1016/j.amepre.2009.06.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. 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;12:703–711. doi: 10.1023/a:1011240019516. [DOI] [PubMed] [Google Scholar]

RESOURCES