Serum metabolomic profiling by 1H NMR reveals distinct metabolic signatures in diabetic and non-diabetic chronic kidney disease.

Chronic kidney disease (CKD) and Type 2 diabetes mellitus (DM) are global health burdens, with diabetes being the leading cause of end-stage renal disease, both marked by systemic metabolic disruption. This study applied 1H NMR-based serum metabolomics to characterize metabolic alterations in CKD only (n = 31), DM (n = 33), and their comorbid state (CKD_DM; n = 45). A total of 45 polar metabolites and seven lipid signals were quantified, and statistical analysis was performed using MetaboAnalyst 6.0. The CKD_DM profile was not a linear combination of the CKD and DM profiles. Compared with DM, CKD_DM shows perturbation in pyruvate, glycerophospholipid, and butanoate metabolism, while tyrosine metabolism, TCA cycle, and cysteine-methionine were prominently altered in CKD only and CKD_DM group. CKD only and CKD_DM partially overlapped based on polar metabolites, Meanwhile lipid-associated signals clearly separated all three groups. mROC analyses identified a significant six-metabolite panel (creatinine, glucose, urea, myo-inositol, glycine, choline) that robustly distinguished CKD_DM from DM(AUC > 0.9). TMAO, isobutyrate, glycerol (AUC > 0.76), and a three-lipid-signal panel (AUC > 0.99) distinguished CKD only from CKD_DM, and the seven-metabolite panel separates CKD only from DM(AUC > 0.9). Overall, CKD_DM emerges as a distinct metabolic phenotype with integrated metabolite and lipid signatures, enhanced disease stratification, and improved diagnostics.