Time-varying relationships between built environment factors, colon and rectum cancer prognosis, and survival

The burden of colorectal cancer (CRC) has been persistently unequal across racial/ethnic groups with mortality 30% higher among non-Hispanic Blacks (NHB) compared to non-Hispanic Whites (NHW). Moreover, the disparity remains after adjustment for important risk factors of CRC survival including tumor subsite, grade, stage at

diagnosis, health insurance, and treatment utilization. The persistence of racial/ethnic CRC disparities despite attempts to account for variation in healthcare and prognostic indicators has increased focus on the role of residential environmental factors. The residential social (e.g., residential racial/ethnic segregation, socioeconomic

deprivation) and built environments (e.g., residential walkability, physical disorder) are conceptualized as main drivers of cancer disparities by race. Several limitations, however, currently prevent robust and translatable results, including: overreliance on neighborhood socioeconomic deprivation as a marker of exposure; lack of individual

residential histories to estimate time-varying covariates, residential mobility patterns, and ‘health selection’ into neighborhoods; and underutilization of emerging technologies and cancer registry linkages that could lead to larger statistical power and novel translational targets. The purpose of this study is to utilize emerging methodologies –

residential history calculation and virtual neighborhood auditing – to investigate longitudinal relationships between modifiable, residential built environment factors and CRC disparities by race and geography. Residential built environment exposure histories will be assessed through the combination of separately developed methodologies:

residential history calculation of cancer cases within the New Jersey State Cancer Registry (NJCSR), and large- scale virtual neighborhood audits of >23,000 Google Street View (GSV) scenes across NJ. CRC case-specific built environment exposure histories will be constructed based on CRC cases’ residential histories and spatio-temporal

models of built environment assessments repeated at multiple dates between 2009-2023 per a GSV location. Specific aims are: 1) to construct built environment exposure histories (2009-2023) of residential walkability and physical disorder for those diagnosed with CRC between 2014-2019 and test relationships between built

environment factors and CRC prognostic factors (microsatellite instability testing, subsite, grade, stage at diagnosis); 2) to test relationships between time-varying built environment factors and CRC survival; and 3) to test whether relationships between race/ethnicity and CRC outcomes are mediated by time-varying built environment

exposure factors of walkability and physical disorder. An opportunity exists to build a more complete characterization of CRC disparities by race and geography through integration of built environment exposure histories into cancer surveillance. Findings from these novel and rigorous methods will motivate additional registry

linkages and more comprehensive epidemiologic studies, in turn, informing cancer surveillance systems, and public health interventions.