Objective To elucidate the causal relationship between high-density lipoprotein cholesterol (HDL-C) and colorectal cancer (CRC) through Mendelian randomization.

Methods Mendelian randomization analysis was conducted using genetic instrumental variables selected from a genome-wide association study dataset. The main methods included inverse variance weighted, MR-Egger, weighted median, simple mode, and weighted mode method; among which, inverse variance weighted method served as the primary analytical approach. Sensitivity analyses were performed to verify the robustness of results.

Results A total of 41 genetic instrumental variables associated with HDL-C were identified. Inverse variance weighted method (OR=0.84, 95%CI: 0.73-0.96, P=0.01) and weighted median method (OR=0.82, 95%CI: 0.67-0.99, P=0.04) indicated a negative correlation between genetically-determined HDL-C and CRC risk. Sensitivity analyses confirmed the absence of heterogeneity and horizontal pleiotropy (P>0.05).

Conclusion A causal relationship exists between HDL-C and CRC risk, with rs1077834 as a potential key determinant in the influence of HDL-C on CRC risk.