Causal Association of Galectins and Diabetic Retinopathy: A Two-Sample Mendelian Randomization Study.

Purpose: Diabetic retinopathy (DR) is a common microvascular complication of diabetes. Previous observation studies identified the association between Galectins (Gal) and the risk of DR, but the causal relationship between Gal and the risk of DR is still unclear. This study aimed to explore the causal associations between circulating galectins (Gal) family levels and the risk of diabetic retinopathy (DR) and its phenotypes including proliferative DR (PDR), background DR (BDR), and severe non-proliferative background DR (SNPDR) using two-sample Mendelian randomization (MR) analysis.

Methods: The single nucleotide polymorphism (SNP) as instrumental variables (IVs) were identified from previous genome-wide association studies (GWAS). Qualified SNPs associated with Gals were selected via strict screening process. The causal association between Gal family levels and the risk of DR was analyzed using MR analysis. The main MR method was Inverse variance weighted (IVW). Causal estimates before and after outlier removal were presented as odds ratio (OR) with 95% confidence interval (CI).

Results: The results of IVW indicated that Gal-4 levels were causally related to elevated risk of DR (OR = 1.068, 95%CI: 1.006-1.133), and PDR (OR = 1.084, 95%CI: 1.004-1.169). Gal-3 had causal effect on the risk of SNPDR (OR = 1.455, 95%CI: 1.103-1.919). Gal-3 binding protein was causally associated with higher risk of SNPDR (OR = 1.295, 95%CI: 1.014-1.654). In addition, Gal-9 had causal association with decreased risk of SNPDR (OR = 0.669, 95%CI: 0.491-0.911). Leave-one-out analysis depicted that no individual SNP could affect the causal associations of Gal-4 with DR and PDR, Gal-3, Gal-3 binding protein and Gal-9 with SNPDR.

Conclusion: Gal-4 levels were causally related to elevated risk of DR, and PDR. Gal-3 and Gal-3 binding protein had causal effect on increased risk of SNPDR while Gal-9 had causal effect on decreased risk of SNPDR.