UNLABELLED: There is an increasing need for new biomarkers to improve prediction of chronic kidney disease (CKD) in individuals with type 2 diabetes (T2D). We aimed to identify blood-based epigenetic biomarkers associated with incident CKD and develop a methylation risk score (MRS) predicting CKD in individuals with newly diagnosed T2D. DNA methylation was analyzed epigenome wide in blood from 487 individuals with newly diagnosed T2D, of whom 88 developed CKD during an 11.5-year follow-up. Weighted Cox regression was used to associate methylation with incident CKD. Weighted logistic models and cross-validation (k = 5) were performed to test whether the MRS could predict CKD. Methylation at 37 sites was associated with CKD development based on a false discovery rate of <
5% and absolute methylation differences of ≥5% between individuals with incident CKD and those free of CKD during follow-up. Notably, 15 genes annotated to these sites, e.g., TGFBI, SHISA3, and SLC43A2 (encoding LAT4), have been linked to CKD or related risk factors, including blood pressure, BMI, and estimated glomerular filtration rate. Using an MRS including 37 sites and cross-validation for prediction of CKD, we generated receiver operating characteristic (ROC) curves with an area under the curve (AUC) of 0.82 for the MRS and AUC of 0.87 for the combination of MRS and clinical factors. Importantly, ROC curves including the MRS had significantly better AUCs versus the one only including clinical factors (AUC = 0.72). The combined epigenetic biomarker had high accuracy in identifying individuals free of future CKD (negative predictive value of 94.6%). We discovered a high-performance epigenetic biomarker for predicting CKD, encouraging its potential role in precision medicine, risk stratification, and targeted prevention in T2D. ARTICLE HIGHLIGHTS: There is an increasing need for new biomarkers to improve the prediction and prevention of chronic kidney disease (CKD) in individuals with type 2 diabetes (T2D), a leading cause of morbidity and mortality in this population. We investigated whether new blood-based epigenetic biomarkers predict incident CKD in individuals with newly diagnosed T2D. We discovered a novel blood-based epigenetic biomarker, composed of a combination of a methylation risk score and clinical factors, capable of predicting CKD during an 11.5-year follow-up (area under the curve of 0.87, negative predictive value of 94.6%) in individuals with newly diagnosed T2D. The epigenetic biomarker could provide a valuable tool for early risk stratification and prevention of CKD in individuals with newly diagnosed T2D, supporting its future use for precision medicine.