Abstract B27: Neighborhood racial concentration and hormone receptor status among California women diagnosed with breast cancer

Abstract

Abstract Introduction: Recent evidence indicates that black women living in highly segregated metropolitan areas when diagnosed with breast cancer may receive their diagnosis at an earlier stage and have reduced cancer-specific and all-mortality rates if their neighborhood has a higher proportion of black residents. We conducted a cross-sectional, multilevel analysis of California Cancer Registry records merged with data from the California Neighborhoods Data System to determine whether the odds of having estrogen and progesterone negative (ER-/PR-) breast cancer are also reduced under similar neighborhood sociodemographic conditions. Methods: A total of 88,205 non-Hispanic white, non-Hispanic black, and Hispanic women, ages 18 to 108 at diagnosis with invasive breast cancer between 1996 and 2004 and lived in a census block group within a California metropolitan statistical area (MSA) were included in the analysis. Block groups were used as the neighborhood-level unit of analysis. Racial residential segregation was assessed at the MSA level and was operationalized using the multigroup entropy index, a measure of “evenness,” or the degree to which all racial groups present in an MSA are evenly distributed across its component parts (i.e., census tracts). Results: Controlling for block group socioeconomic status and individual sociodemographic and tumor characteristics, higher percent of black residents in a neighborhood significantly reduced the odds of having ER-/PR- breast cancer relative to having ER+/PR+ cancer among black women (-3.7% for every 10% increase in black neighborhood residents; p = 0.01). When the analysis was further stratified by multigroup entropy scores, black women residing in highly segregated MSA's had a similar reduction in ER-/PR- risk (-3.8% for every 10% increase in black neighborhood residents; p = 0.02), but the reduction was only marginally significant in less segregated MSA's (-4.1% for every 10% increase in black neighborhood residents; p = 0.10). There was a modest increase in the risk of ER-/PR- subtype among Hispanics not accounting for segregation(+4.6% for every 10% increase in black neighborhood residents; p = 0.053), and among Hispanics living in less segregated MSA's (+7.0% for every 10% increase in black neighborhood residents; p = 0.074), but not among Hispanics living in more segregated metropolitan areas. No associations between neighborhood black composition and ER-/PR- subtype were observed among whites. The neighborhood concentration of Hispanic residents generated a different pattern of ER-/PR- risk. Among Hispanic women, greater percentages of Hispanic neighborhood residents increased the odds of having ER-/PR- versus ER+/PR+ breast cancer (+2.7% for every 10% increase in Hispanic neighborhood residents, p < 0.01). Similar increases in risk were seen among non-Hispanic white women (+1.9% for every 10% increase in Hispanic neighborhood residents, p = 0.01), but there was no association among blacks. Stratifying on MSA-level racial segregation, we found similarly significant increases in ER-/PR- risk among whites and Hispanics within both more and less segregated metropolitan areas. No statistically significant relationship was noted among black women residing in low segregation MSA's, but there was a marginally significant decrease in ER-/PR- risk among black residents of more highly segregated MSA's (-4.1% for every 10% increase in Hispanic block group residents; p = 0.05). Conclusions: The racial/ethnic composition of neighborhoods is associated with the risk of ER-/PR- subtype among Californian women diagnosed with breast cancer. This relationship differs across racial/ethnic groups, and to a lesser extent, by the degree of race-based residential segregation across the broader metropolitan area. Citation Format: Erin Linnenbringer, Scarlett Lin Gomez. Neighborhood racial concentration and hormone receptor status among California women diagnosed with breast cancer. [abstract]. In: Proceedings of the Sixth AACR Conference: The Science of Cancer Health Disparities; Dec 6–9, 2013; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2014;23(11 Suppl):Abstract nr B27. doi:10.1158/1538-7755.DISP13-B27