The objective was to evaluate all-cancer and cancer-specific mortality in relation to residential racial segregation in all individuals and specifically for black individuals. The Mortality Disparities in American Communities (MDAC) study is a record linkage of the nationally representative American Community Survey (ACS) with the National Death Index (NDI). Census county-level segregation estimates were calculated for exposure (isolation and interaction) and evenness (dissimilarity) for black race as minority and non-Hispanic white as majority populations. Age-adjusted mortality rates (per 100,000) were calculated for quintiles, and p-trend was calculated using quintile as continuous covariate in parametric survival regression models. Due to small numbers (black deaths<400) in other cancers, only colon, rectum, and anus (ICD C18-C21); pancreas (ICD C25); respiratory (trachea, lung, and bronchus) (ICD C33-C34); breast (ICD C50); and prostate (ICD C61) were evaluated. There were >4,300,000 individuals, representing 31,400,000 days at-risk. In all individuals, there were >69,500 deaths from all malignant neoplasms (ICD C00-C97) (age-adjusted mortality rate in all: 159.6; in black individuals: 181.9). In all individuals, mortality from all malignant neoplasms was significantly associated with interaction (Q1: 168.5 to Q5: 154.1, p<0.01) and isolation (Q1: 137.9 to Q5: 172.7, p<0.001). In black individuals, similar associations were observed for interaction (Q1: 199.0 to Q5: 142.1, p<0.05) and isolation (Q1: 127.8 to Q5: 199.5, p<0.001), and dissimilarity was also significantly associated (Q1: 174.3 to Q5: 205.5, p<0.01). In cancer-specific analyses, dissimilarity was positively associated with pancreatic cancer mortality in all individuals (p<0.05) and respiratory cancer mortality in black individuals (p<0.01). Interaction was inversely associated with mortality from pancreatic cancer (p<0.01), respiratory (p<0.001), and breast cancer (p<0.05) in all individuals, but not associated with specific cancers in black individuals. Isolation was positively associated with mortality from pancreatic cancer (p<0.001); respiratory (p<0.05); and, breast cancer (p<0.001) in all individuals, and positively associated with respiratory (p<0.05) in black individuals. Small numbers of cancer-specific deaths and not having the ability to adjust for key health behaviors were limitations. However, a strength was the ability to look at the role of individual-level race and geographic-level segregation, and results suggest that residential segregation measures are associated with cancer mortality, and impact on black individuals shares some similarity to the overall population associations, but dissimilarity may have more adverse associations for black individuals. Additional work is needed to evaluate other dimensions of residential segregation, evaluate other races, and understand the cancer-related characteristics of these geographic areas to help inform mechanisms.

Citation Format: Cara L. Frankenfeld, Jahn K. Hakes, Timothy F. Leslie. All-cancer and cancer-specific mortality is associated with black race segregation in the United States: Mortality Disparities in American Communities data [abstract]. In: Proceedings of the Eleventh AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2018 Nov 2-5; New Orleans, LA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(6 Suppl):Abstract nr C069.

©2020 American Association for Cancer Research.
2020
American Association for Cancer Research.