Abstract B034: Association of Neighborhood-Level Food Environment with Prostate Cancer-Specific Mortality Among Men in Georgia

Abstract Introduction and Objectives Food swamps (FS) –areas with a high density of ultra-processed food outlets relative to grocery stores and supermarkets –have emerged as contextual determinants of cancer risk and survival. We hypothesized that greater FS exposure would be associated with higher prostate cancer-specific mortality (PCSM) and that this association would differ by race in Georgia, a state marked by persistent prostate cancer (PCa) disparities. Methods We conducted a retrospective analysis of men diagnosed with PCa in Georgia (2000–2022) using Georgia Cancer Registry data. Census tract–level Federal Information Processing Series (FIPS) codes linked neighborhood-level food data from the National Neighborhood Data Archive (NaNDA; 1990–2021) and the Modified Retail Food Environment Index (RFEI) with patient-level cancer registry data. The RFEI, defined as the ratio of ultra-processed food outlets to grocery stores, farmers markets, and supermarkets, was categorized into low, moderate, and high FS exposure. Chi-square tests and ANOVA assessed group differences, and Cox proportional hazards models estimated associations between FS exposure and PCSM, stratified by race. Fully adjusted models controlled for age, race, insurance status, Gleason grade, PSA, stage at diagnosis, and primary treatment. Results Among 154,480 men (95,517 non-Hispanic [NH] White; 55,288 NH Black), 70% (N=108,146) lived in moderate/high FS tracts. Men in high FS areas (N = 33,330) were more likely to have public insurance, localized disease, and undergo surgery (p<0.05). Median follow up time was 8 years (IQR 4-13). There were 14,656 (9.5%) PCa-deaths. After age adjustment, each 1-unit increase in FS score corresponded to a 1% higher risk of PCSM (aHR 1.01 [95% CI 1.00–1.02], p = 0.005). The FS score ranged from 0-17 (median 1.16, mean 1.77; IQR 0.50–2.33). The association was attenuated after full adjustment (aHR 1.01 [95% CI 0.99-1.03], p=0.23). NHB men in moderate (aHR 1.14 [95% CI 1.01 – 1.28], p=0.04) or high (aHR 1.18 [95% CI 1.01 – 1.37], p=0.04) FS tracts had a higher risk of PCSM versus NHW men in low FS census tracts. No difference was observed among NHW men in high (aHR 1.09 [95% CI 0.96 – 1.24], p=0.20) FS census tracts. Conclusions The neighborhood food environment represents a potentially modifiable, contextual determinant of cancer outcomes and potential target for intervention. The higher mortality risk among NHB men across FS levels underscores the enduring effects of systemic inequities on PCa outcomes. Source of Funding: None Citation Format: Divya Lagisetti, Samuel Kennedy, Katherine J. Kim, R. Seth. Rozelle, Phillipe Gaillard, Avirup Guha, Zachary Klaassen, Martha K. Terris, Justin X. Moore, Malcolm S. Bevel, Ashanda R. Esdaille. Association of Neighborhood-Level Food Environment with Prostate Cancer-Specific Mortality Among Men in Georgia [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Innovations in Prostate Cancer Research and Treatment; 2026 Jan 20-22; Philadelphia PA. Philadelphia (PA): AACR; Cancer Res 2026;86(2_Suppl):Abstract nr B034.