Causal relationship between type II diabetes mellitus, metformin, insulin, gliclazide, and esophageal cancer-insights from two-sample Mendelian randomization study and meta-analysis.

Abstract

Background: Over recent decades, findings on the potential correlation between type II diabetes mellitus (T2DM) and the risk of esophageal cancer (EC) have displayed considerable heterogeneity. Furthermore, metformin has emerged as a potentially protective agent against certain site-specific malignancies. This study aims to explore the causal relationship between T2DM, medication treatments (metformin, insulin, gliclazide), and EC risk while addressing the notable variability in previous research findings.

Methods: To elucidate the causal associations between T2DM, medication treatments, and EC, we employed a synergistic methodology that integrates the two-sample Mendelian randomization (MR) approach with meta-analysis. The genome-wide association studies (GWAS) pertaining to each exposure and EC were acquired from a publicly accessible database.

Results: For MR analyses, three out of seven GWAS datasets within the T2DM cohort exhibited statistical significance. Conversely, all MR analyses yielded non-significant results in the medication cohort. Meta-analyses suggested that a genetic predisposition to T2DM correlated with a reduced risk of EC [odds ratio (OR), 0.999612; 95% confidence interval (CI): 0.999468-0.999756; P=0.01; I²=0%]. Moreover, metformin intake was causally linked to a decreased prevalence of EC (OR, 0.988954; 95% CI: 0.979044-0.998963; P=0.03; I²=0%), whereas neither insulin nor gliclazide manifests statistical significance.

Conclusions: Our findings indicate T2DM and metformin are causally associated with diminished risk of EC, while no causal associations exist between insulin, gliclazide, and EC.