Metabolomic Analysis Of Chronic Kidney Disease

Chronic kidney disease (CKD) is a progressive condition characterized by the gradual loss of kidney function, significantly impacting metabolic homeostasis. However, the differential abundance of metabolites in different stages of CKD are unknown. This study aimed to identify differentially abundant (DA) metabolites in individuals without CKD and across four stages of CKD, and in doing so, uncover potential biomarkers and understand disease-related metabolic alterations. I also tried to classify individuals CKD into unaffected or into stage I–IV using DA metabolites or clinical features. I identified 15 DA metabolites between controls and early-stage CKD cases, 40 between early-stage CKD and late-stage CKD cases, and 79 between controls and late-stage CKD cases. There was enrichment of DA metabolites in 24 pathways in controls vs and early-stage CKD cases, 19 in controls vs late-stage CKD cases, and 40 in early-stage vs late-stage CKD cases. Machine learning models that attempted to classify individuals based on their DA metabolites had greater than 80% accuracy when classifying between controls and early-stage cases or early-stage cases and late-stage cases and greater than 88% accuracy when classifying between controls and late-stage cases. However, multi-outcome predictions only performed modestly well. The results also indicate that classification using metabolites performs better than classification with clinical features. This study identified stage-specific metabolic alterations, underscores the promise of precision medicine in CKD diagnosis, which can help the early treatment of CKD. However, we did not identify a solid biomarker, which requires the need for further research. The insights in DA metabolites of CKD can help improve the quality of care and reduce global healthcare disparities by improving screening programs, especially for at-risk populations, and supporting the design of cost-effective interventions to slow the progression of chronic renal failure.