Insulinemic and Inflammatory Dietary Patterns and Colorectal Cancer Risk: A Dietary Data Harmonization Study of One Million Participants in the COMETS Consortium
Menée à partir de données internationales portant sur 501 892 femmes et 407 390 hommes (durée médiane de suivi : 14,9 ans), cette étude analyse l'association entre des régimes alimentaires faiblement insulinémiques et anti-inflammatoires et le risque de cancer colorectal (16 525 cas)
Background: Inflammatory and insulinemic dietary patterns have been associated with colorectal cancer (CRC) risk, but generalizability across diverse populations with heterogeneous food supplies and dietary behaviors has not been established.
Objective: We harmonized disparate dietary and covariate data on a large scale to compute the reverse Empirical Dietary Index for Hyperinsulinemia-rEDIH, reverse Empirical Dietary Inflammatory Pattern-rEDIP, and Healthy Eating Index (HEI)-2015 scores, and tested their associations with CRC risk.
Methods: We leveraged data among 501,892 women and 407,390 men from six cohorts across the U.S. (NIH-AARP, MESA, PLCO, SCCS) and Europe (EPIC, ATBC) with varying sociodemographic characteristics, participating in the Consortium of Metabolomics Studies. We harmonized nomenclature and nutritional information for more than 800 unique food items across cohorts. We used multivariable-adjusted Cox regression, adjusting for demographic, clinical, and lifestyle factors, to calculate hazard ratios (HR) and 95% confidence intervals (95% CI) for the associations between the dietary indices and CRC risk per cohort, then meta-analyzed the estimates.
Results: During a median follow-up of 14.9 years, 16,525 incident CRC cases were diagnosed. Participants in the highest quintile of rEDIH (low-insulinemic diet) had an 18% reduced risk of CRC (HR 0.82; 95% CI: 0.78, 0.86) compared to those in the lowest quintile. For the same comparison, similar risk reductions were observed for rEDIP (anti-inflammatory diet) (HR 0.84; 95% CI: 0.80, 0.89) and HEI-2015 (overall dietary quality) (HR 0.80; 95% CI: 0.76, 0.85). Heterogeneity between cohorts in the meta-analyzed estimates was low for rEDIH (I2=22.3%) compared to rEDIP (I2=62.5%) and HEI-2015 (I2=83.9%).
Conclusion: Using carefully harmonized data from nearly one million individuals in the U.S. and Europe, we observed significant CRC risk reduction with habitual intake of low-insulinemic and anti-inflammatory dietary patterns, comparable to higher overall dietary quality. Study findings underscore the utility of these dietary patterns for global cancer prevention efforts.
The American Journal of Clinical Nutrition , article en libre accès, 2025