Abstract C091: Deprivation/segregation indices and survival in the AACES ovarian cancer cohort: A mediation analysis

Abstract

Abstract Deprivation indices (DIs), constructed from area-level socio-economic indicators, attempt to allow for the acknowledged relationship between socio-economic status and health outcomes. Segregation indices (SIs) address the area-level divisions in ethnicity and socio-economic status. In this study we evaluated the effects of deprivation and segregation on ovarian cancer survival among African-American women in a Bayesian hierarchical joint mediation framework. Deprivation and segregation effects considered were the most commonly used: 4 dimensions of Kolak social determinants of health, Yost index, Singh index, the percentage of people below the poverty line, ICE income and race segregation measures, and the concentrated disadvantage index. We analyzed a cohort (n=559) of African-American ovarian cancer patients from the states of Alabama, Georgia, Illinois, Louisiana, New Jersey, North Carolina, Ohio, South Carolina, Tennessee, and Texas and the metropolitan area of Detroit, as part of the African American Cancer Epidemiology Study (AACES). We evaluated the effects of mediation via stage at diagnosis, histology and delayed diagnosis and examined confounding variables: age, current BMI, income, smoking status, and physical activity level, using Bayesian Gibbs variable selection to select the confounders for the final models. Our main result, with the stage as the mediator, showed that increased urbanization, mobility, socio economic status of Kolak dimensions and decreased deprivation measured by Yost index have positive effects on overall survival. 1 unit decrease of deprivation measured by Yost index will lead to 26% decrease of the hazard for ovarian cancer (CI: 0.56,0.91). Similarly, increased urbanization gives positive effect for the overall survival. 1 unit increase of urbanization lead to 27% decrease of hazard (CI: 0.57,0.95). Similar results are found for histology and diagnosis delay: when histology is the mediator, increase of urbanization and socio-economic status, and decreased deprivation measured by Singh, Yost and concentrated disadvantage indices have positive association with overall survival. When delayed diagnosis was our mediator, increased urbanization, socio-economic status and mobility, and decreased deprivation measured by Yost, ICE income and concentrated disadvantage indices have the positive association with overall survival. The results of mediation analyses revealed that Kolak dimensions and decreased deprivation was associated with increased survival from ovarian cancer . In summary, these results indicate that several measures of area-level deprivation, when mediated by intermediate pathway variables, are associated with survival among African-American women diagnosed with ovarian cancer. Their particular direction and level of impact varies with DI or SI chosen. Citation Format: Joanne Kim, Andrew B. Lawson, Joellen Schildkraut, Edward S. Peters, Courtney E. Johnson. Deprivation/segregation indices and survival in the AACES ovarian cancer cohort: A mediation analysis [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr C091.