Genetic evidence for causal relationships between vegetarian biomarkers and esophageal cancer risk

Esophageal cancer (EC) poses a significant global health burden with complex pathogenic mechanisms. While biomarkers specific to vegetarian populations have been implicated in cancer development, their causal relationship with EC remains unclear.

We leverage genetic approaches to investigate the bidirectional causal relationships between vegetarian-specific biomarkers and EC risk. Using genome-wide association study (GWAS) data from European populations, we performed Mendelian randomization (MR) analyses on vegetarian biomarkers as exposures and EC as the outcome, with validation in an independent EC cohort. Our comprehensive analytical framework incorporated inverse-variance weighted (IVW) method, supplemented by MR-Egger and weighted median approaches, along with extensive sensitivity analyses.

After false discovery rate (FDR) correction, we identified significant positive genetic correlations between EC risk and genetically predicted levels of direct LDL-cholesterol interactions with vegetarianism, both with and without BMI adjustment (LDL-C × Veg and LDL-Cadj × Veg). Reverse MR analyses revealed negative associations between EC and total bilirubin interactions with vegetarianism (TBIL × Veg and TBILadj × Veg), while positive associations were found with six biomarkers, including albumin interactions (ALB × Veg and ALBadj × Veg). Notably, colocalization analyses provided strong evidence for shared genetic variants between EC and albumin interactions, suggesting common genetic determinants.

Our findings provide novel genetic evidence for bidirectional causal relationships between vegetarian-specific biomarkers and EC risk. These results enhance our understanding of EC pathogenesis and may inform future therapeutic strategies, particularly for vegetarian populations.