ObjectiveTo investigate the potential causal relationships between blood metabolites and diabetic kidney disease (DKD) in East Asian populations and to identify ethnicity-specific metabolic risk factors. MethodsWe utilized summary-level data from genome-wide association studies (GWAS) comprising 112 blood metabolites in a Chinese population and DKD in an East Asian population. Genetic variants significantly associated with metabolites were selected as instrumental variables. We performed bidirectional two-sample Mendelian randomization (MR) analyses, with primary causal estimates derived using the inverse variance weighted (IVW) method. Sensitivity analyses included MR-Egger, weighted median, and Bayesian-weighted MR (BWMR). Metabolites showing suggestive evidence of association were further investigated using multivariable MR and colocalization analysis. Robustness of findings was assessed via Cochran’s Q test for heterogeneity, the MR-Egger intercept test for horizontal pleiotropy, and leave-one-out analysis. ResultsForward MR analysis identified suggestive associations between elevated levels of tyrosine, glutamine, leucine, and the albumin-to-globulin ratio with DKD risk. In multivariable MR analysis, leucine (OR=0.084, P=0.026) and the albumin-to-globulin ratio (OR=0.161, P=0.048) retained statistically significant independent protective effects against DKD, whereas the associations for tyrosine and glutamine were attenuated. Colocalization analysis did not support the presence of shared underlying causal genetic variants between the identified metabolites and DKD. Sensitivity analyses revealed no significant evidence of heterogeneity or horizontal pleiotropy. Reverse MR analysis did not suggest the presence of reverse causation. ConclusionOur findings suggest that leucine and the albumin-to-globulin ratio may act as independent protective factors against DKD in East Asian populations. These results provide new genetic insights into the ethnicity-specific metabolic mechanisms contributing to DKD.