Structural disparities in early-onset colorectal cancer mortality: A national county-level analysis.

10579 Background: Early-onset colorectal cancer (CRC) incidence and mortality are rising, with significant disparities by geography and socioeconomic status. We examined county-level disparities in early-onset CRC mortality (ages < 55) across the United States to identify modifiable factors associated with mortality rates. Methods: We analyzed age-adjusted CRC mortality rates from CDC WONDER. Predictors included rurality (Rural–Urban Continuum Codes 2023: non-metro vs metro), socioeconomic factors (median income, % uninsured, % poverty from County Health Rankings 2025), demographic composition (% non-Hispanic Black, % Hispanic), health behaviors (current smoking, obesity, routine checkup from PLACES 2025), and healthcare access (gastroenterologist, primary care physician, and hospital counts from AHRF 2024–2025). Multivariable linear regression assessed associations between predictors and mortality, with continuous variables standardized to z-scores. Counties with missing or unreliable mortality data were excluded (p < 0.05 for significance). Results: The final sample included 672 counties. Mean mortality was 4.26 per 100,000 (range 1.7–11.9). Non-metro counties comprised 12.2% (n = 82). The model explained 54.8% of variance in mortality (F = 66.5, p < 0.001). Current smoking was the strongest predictor (β = 0.54 per SD, 95% CI: 0.38–0.70, p < 0.001), indicating a one-SD increase in smoking prevalence corresponded to a 0.54 per 100,000 higher mortality rate. Non-metro counties had 0.80 more deaths per 100,000 than metro counties (95% CI: 0.56–1.05), a 19% increase relative to the mean. Poverty (β = 0.21 per SD, p = 0.011), uninsured rates (β = 0.20 per SD, p < 0.001), and % Black (β = 0.21 per SD, p < 0.001) were also significant. Higher Hispanic population was associated with lower mortality (β = −0.20 per SD, p = 0.001). Healthcare provider variables (gastroenterologist count, primary care count, hospital count) were not significantly associated (p≥0.051). Conclusions: Geographic, socioeconomic, and health behavior factors accounted for 54.8% of variation in early-onset CRC mortality, with smoking showing the strongest association. Non-metro counties and areas with higher poverty, uninsured rates, and smoking prevalence had higher mortality, though these relationships are not causal. Remaining unexplained variance suggests roles for screening uptake, stage at diagnosis, treatment quality, and genetic susceptibility. Findings identify high-risk counties and highlight the need for future causal analyses to clarify pathways underlying disparities and support targeted interventions.