P503 - Dietary Index for Gut Microbiota and Cancer Mortality: A Prospective Comparison with HEI and DII in NHANES 2007–2018

Purpose/Background: Emerging evidence suggests that dietary patterns influencing gut microbiota composition may affect carcinogenesis and cancer survival. The Dietary Index for Gut Microbiota (DI-GM) captures the overall quality of a microbiota-supportive diet. However, its association with cancer and colorectal cancer (CRC) mortality has not been evaluated in a nationally representative cohort. We aimed to investigate both cross-sectional and prospective relationships of DI-GM with CRC prevalence and cancer-related mortality, and to compare its predictive ability with established dietary indices-the Healthy Eating Index (HEI-2015) and Dietary Inflammatory Index (DII).

Methods/Interventions: Data were obtained from adults in NHANES 2007-2018 linked with mortality follow-up through 2019. DI-GM, HEI, and DII were derived from two 24-hour dietary recalls. Cross-sectional associations between DI-GM and prevalent CRC were estimated using survey-weighted logistic regression adjusting for demographic, socioeconomic, lifestyle, and metabolic factors. Prospective associations with cancer and CRC mortality were assessed using weighted Cox proportional-hazards models, sequentially adjusted for the same covariates. Early deaths within 24 months were excluded in sensitivity analyses. Non-linear associations were examined with restricted cubic splines (RCS).

Results/Outcomes: Participants with higher DI-GM scores were older, had higher education and income, and demonstrated lower BMI and waist circumference with healthier lifestyles (p < 0.001). Cross-sectionally, DI-GM was not significantly associated with CRC prevalence after multivariable adjustment (p-trend > 0.05). Prospectively, higher DI-GM predicted lower cancer mortality (HR per unit = 0.94; 95% CI 0.89-0.99; p = 0.04) and CRC-specific mortality (HR = 0.90; 95% CI 0.81-1.00; p = 0.05). RCS analyses showed an approximately linear inverse relationship. When compared across dietary indices, higher HEI also related to lower cancer mortality (HR ≈ 0.95 per 10-point increase), whereas higher DII (pro-inflammatory diet) was positively associated (HR ≈ 1.08 per SD). The magnitude of the DI-GM association was comparable or slightly stronger than HEI and opposite to DII. Sensitivity analyses excluding early deaths or stratifying by BMI and age yielded consistent findings.

Conclusion/Discussion: A microbiota-supportive diet, reflected by a higher DI-GM score, was prospectively associated with reduced risks of overall and CRC-specific mortality in U.S. adults. These findings highlight the potential value of DI-GM as a gut-microbiota-relevant dietary quality metric that complements HEI and DII and underscores the importance of promoting microbial-friendly dietary patterns for cancer prevention and survivorship.

Kaplan-Meier Curves for Cancer Mortality by DI-GM Quartiles
Survival probability for overall cancer mortality across quartiles of the Dietary Index for Gut Microbiota (DI-GM) in NHANES 2007-2018, adjusted for demographic, socioeconomic, lifestyle, and metabolic factors (Lag 24 months excluded).
Restricted Cubic Spline for DI-GM and Cancer Mortality
Adjusted hazard ratios and 95% confidence intervals for cancer mortality across continuous DI-GM values using restricted cubic spline models, showing an approximately linear inverse association. Models weighted and adjusted for full covariates.