Investigating the relationship between blood metabolites and diabetic retinopathy using two-sample mendelian randomization and in vivo validation.

We addressed fundamental questions about the influence of metabolites on the development of Diabetic retinopathy (DR), and explored the related pathological mechanism. Genome-wide association study (GWAS) database data for metabolites and DR were used to perform Mendelian randomization (MR) studies. The inverse variance weighting (IVW) was chosen as the primary analysis method. Sensitivity analysis was conducted using MR-PRESSO, leave-one-out and Cochran's Q test. Confounding factors were eliminated to ensure robustness. We also conducted metabolic pathway analysis. In vivo experimental validation was conducted using Sprague Dawley rats. The serum metabolites of the DR group rats and normal group rats were examined to evaluate the MR results. The screen identified eighteen metabolites associated with DR risk, twelve of which were known components. Seven metabolites were positively correlated with DR risk, while five could reduce it. Eight metabolites associated with proliferative DR (PDR) risk were identified, four of which are known components. Three of these were positively associated with PDR risk and one metabolite reduced PDR risk. Additionally, two possible metabolic pathways involved in the biological mechanism of DR were identified. The ELISA results showed that the serum levels of isoleucine and 4-HPA were significantly increased in DR rats, while the level of inosine was decreased. This study offers novel insights into the biological mechanisms underlying DR. Metabolites that are causally linked to DR may serve as promising biomarkers and therapeutic targets.