Circulating metabolomic profiles associated with incident age-related eye diseases and improved risk prediction

To evaluate the associations between circulating metabolites and the risk of multiple age-related eye diseases, and to develop risk prediction models integrating metabolite information to improve early risk stratification. This study was based on data from the UK Biobank and included analyses of age-related macular degeneration (AMD), diabetic retinopathy (DR), primary open-angle glaucoma (POAG), and cataract. A total of 168 circulating metabolites were examined as exposures. Cox proportional hazards regression models were applied to assess the associations between metabolites and the incident risk of each disease. Subsequently, metabolite-based predictive models integrated with traditional clinical risk factors were developed using the XGBoost algorithm. Model performance was evaluated using the area under the curve, concordance index, net reclassification improvement, integrated discrimination improvement, and decision curve analysis. In addition, two-sample Mendelian randomization analyses were conducted to explore causal relationships between metabolites and disease outcomes. A total of 442,275 participants for AMD, 69,197 for DR, 103,577 for POAG, and 385,967 for cataract were included, with a median follow-up of approximately 13 years. Multivariable Cox regression analyses identified 146, 139, 119 and 111 metabolites significantly associated with the risk of DR, AMD, POAG and cataract, respectively. Compared with models including only traditional clinical risk factors, integrated models improved predictive performance. Mendelian randomization analyses suggested a positive causal association of citrate with DR and of 3-hydroxybutyrate with AMD. Circulating metabolites are associated with the risk of multiple age-related eye diseases. Incorporating metabolite information into predictive models improves risk prediction and may support the feasibility and clinical translation of metabolite-based testing for targeted risk assessment and early identification of age-related eye diseases.