Question and Answer

Purpose/Background: Rising mortality rates due to early onset colorectal cancer (EOCRC), defined as diagnosis before age 50, are poorly understood. We investigated the relationship between local food environment, social vulnerability, and overall survival in patients with EOCRC across a statewide health system.

Methods/Interventions: We conducted a retrospective study to evaluate non-clinical drivers of overall survival among adults aged 20-49 years diagnosed with EOCRC between January 2013 and March 2025 within a statewide academic health system. Residential addresses were geocoded to U.S. census tracts and categorized into quartiles of 1) an adapted Unhealthy Food Environment Index (UFEI), derived from the CDC's modified Retail Food Environment Index (mRFEI), and 2) social vulnerability, using the Centers for Disease Control and Prevention's (CDC) Social Vulnerability Index (SVI). The UFEI was defined as the proportion of unhealthy retailers relative to the total number of unhealthy and healthy retailers within a half-mile radius of each census tract (range: 0-100). Food establishment data were obtained from Data Axle, a marketing firm with a comprehensive business database. We generated Kaplan-Meier survival curves and performed log-rank tests to compare overall survival (OS) across the highest and lowest quartiles of UFEI ("most unhealthy" and "least unhealthy") and SVI ("most socially vulnerable" and "least socially vulnerable", respectively).

Results/Outcomes: A total of 642 patients were diagnosed with EOCRC between 2013 and 2025. Ten-year overall survival was 64.7% for patients in the least unhealthy census tracts (n=152, UFEI≤61.67) and 60.12% for patients in the most unhealthy tracts (n=237, UFEI=100). Kaplan-Meier curves suggested that patients in the most unhealthy tracts had lower survival probability than those in the least unhealthy tracts (Figure 1a). However, the log-rank test (χ2=2.73, p=.09) did not reach statistical significance. Ten-year overall survival was 62.9% for patients in the least socially vulnerable census tracts (n=152, SVI≤0.26) and 49.7% in the most socially vulnerable (n=153, SVI≥0.75). Kaplan-Meier curves showed that patients in the most socially vulnerable tracts had lower survival probability than those in the least socially vulnerable (Figure 1b). The log-rank test confirmed a significant difference in survival (χ2=3.58, p=0.05).

Conclusion/Discussion: Social vulnerability was associated with worse overall survival among patients with EOCRC, while the relationship between unhealthy food environment and survival showed a similar but non-significant trend. Accounting for area-level clustering in future analyses may clarify this relationship. Overall, these findings highlight the importance of population-level strategies in strengthening the built environment and reducing survival disparities among patients with EOCRC.

Figure 1a: Kaplan-Meier Curves for Overall Survival in Patients with EOCRC by Unhealthy Food Environment Index Quartiles
Curves represent the lowest and highest quartiles for each index. Abbreviations: EOCRC, early-onset colorectal cancer; UFEI, Unhealthy Food Environment Index; SVI, Social Vulnerability Index
Figure 1b: Kaplan-Meier Curves for Overall Survival in Patients with EOCRC by Social Vulnerability Index Quartiles
Curves represent the lowest and highest quartiles for each index. Abbreviations: EOCRC, early-onset colorectal cancer; UFEI, Unhealthy Food Environment Index; SVI, Social Vulnerability Index