Association Between Glycemic Traits and Primary Open-Angle Glaucoma: A Mendelian Randomization Study in the Japanese Population

Abstract

A meta-analysis suggests a relationship between abnormal glucose metabolism and primary open-angle glaucoma (POAG); however, the causal association between them remains controversial. We therefore conducted a Mendelian randomization (MR) study to assess the causal association between genetically predicted glycemic traits and the risk of POAG. Two-sample MR design. We examined the genetically predicted measures of fasting glucose, hemoglobin A1c (HbA1c), and fasting C-peptide, in relation to POAG. For the single nucleotide polymorphism (SNP)-exposure analyses, we meta-analyzed the study-level genome-wide associations of fasting glucose levels (n=17,289; n of SNPs=34), HbA1c (n=52,802; n of SNPs=43), and fasting C-peptide levels (n=1666; n of SNPs=17) from the Japanese Consortium of Genetic Epidemiology studies. We used summary statistics from the BioBank Japan projects (n=3980 POAG cases and 18,815 controls) for the SNP-outcome association. We observed no association of genetically predicted HbA1c and fasting C-peptide with POAG. The MR inverse-variance-weighted (IVW) odds ratios (ORs) were 1.44 (95% confidence interval [CI], 0.78-2.65; P=.25) for HbA1c (per 1% increment) and 0.92 (95% CI, 0.56-1.53; P=.76) for fasting C-peptide (per 2-fold increment). A significant association between fasting glucose (per 10 mg/dL-increment) and POAG was observed according to the MR IVW analysis (OR=1.48 [95% CI, 1.10-1.79, P=.009]); however, sensitivity analyses, including MR-Egger and weighted-median methods, did not support this association (P > .10). We did not observe strong evidence to support the association between genetically predicted glycemic traits and POAG in the Japanese population.