Abstract
Given that homocysteine (Hey) may be a viable target for intervention and that there is uncertainty regarding the causal relationship between plasma Hcy levels and colorectal cancer (CRC), this study used Mendelian randomization (MR) to investigate the relationship between Hey and CRC. We summarized the data in this work using genome-wide association studies, identified single nucleotide polymorphisms that were strongly correlated with plasma Hcy levels as instrumental variables, and ran MR analysis on 2 separate sets of outcome data. To make sure the results were stable, a meta-analysis was carried out. MR-Egger, weighted median MR analysis, and the inverse variance method were among the specific analysis techniques used. The leave-one-out method, MR-Egger intercept, MR-PRESSO, and Cochran Q test were also used to assess the stability and dependability of the MR analysis results. In 2 separate European population-based datasets (UK Biobank: OR = 0.9992, 95% CI = 0.9963-1.0021, P = .5951, FinnGen: OR = 0.9771, 95% CI = 0.8370-1.1408, P = .7698), inverse variance method analysis did not reveal any significant causal connection between plasma Hcy levels and CRC. The MR-Egger and weighted median analyses yielded nonsignificant relationships. Both the Cochran Q test and the MR-Egger intercept indicated the absence of considerable heterogeneity and horizontal pleiotropy. The conclusions were further corroborated by the results of the MR-PRESSO analysis and leave-one-out. The pooled results of the Meta analysis also failed to demonstrate significant causality (OR = 0.9992, 95% CI = 0.9963-1.0021, P = .5914), therefore providing additional confirmation of the findings from the individual studies. There was no discernible causal relationship between plasma Hcy levels and CRC risk, according to MR analysis. The findings of this study indicate that while Hcy is a possible target for intervention, there may not be a direct causal relationship between it and the risk of CRC. This finding requires further validation in larger sample sizes and in different populations.