Possible Causal Association Between Thyroid-Related Traits and Diabetic Retinopathy Risk: Evidence From 23 Medication-Taking Traits

Purpose To evaluate whether thyroid-related medication-taking traits show evidence consistent with a possible causal relationship for diabetic retinopathy (DR).Methods We utilized genome-wide association study (GWAS) summary statistics from UK Biobank, which included 23 medication-taking traits in a sample of approximately 320,000 individuals and DR data from 1,652 cases and 60,577 controls. We employed Mendelian randomization (MR) approach to investigate the possible causal relationships between medication-taking characteristics and DR. Possible causal effects were screened with TwoSampleMR inverse-variance weighting, confirmed by univariable MR after pleiotropic SNPs were removed via MR-PRESSO; reverse MR excluded reverse causality, and public GWAS data verified the thyroid-disease associations.Results We found drugs used in diabetes and thyroid preparations showed a significant association with DR (p < 0.05/22). After removing pleiotropic SNPs, the UVMR analysis suggested a potential relationship between thyroid preparations and DR (IVW: OR = 1.18, p = 1.91 × 10−3). Furthermore, the MR analysis indicated that both hypothyroidism and hyperthyroidism were associated with a significantly higher risk of DR (OR = 1.31, p = 1.75 × 10−6; OR = 1.32, p = 1.05 × 10−4). The findings from different single-nucleotide analytic methods were consistent, suggesting the robustness of the possible causal associations.Conclusions Our findings suggest thyroid dysfunction appears to causally elevate DR risk, informing screening and prevention in diabetes. The precise mechanisms may involve thyroid-related oxidative stress and inflammation.