Risk prediction of type 2 diabetic kidney disease based on a nomogram model： a retrospective cohort study

ObjectiveThis study aimed to investigate the risk factors associated with progression from type 2 diabetes mellitus （T2DM） to diabetic kidney disease （DKD）， and to develop and validate a DKD risk nomogram prediction model to provide evidence-based support for clinical intervention.MethodsThis retrospective cohort study included 344 hospitalized patients with T2DM who underwent renal biopsy between January 1， 2019 and September 30， 2024 in Center for Kidney Diseases， the First Affiliated Hospital of Xinjiang Medical University （165 confirmed DKD cases and 179 nondiabetic kidney disease ［NDKD］ cases）. Through univariate screening and multivariable regression modeling， a nomogram was constructed using the rms package in R software. Receiver operating characteristic （ROC） curve analysis was used to evaluate discrimination， while calibration curve analysis and the Hosmer-Lemeshow goodness-of-fit test were used to assess calibration. Decision curve analysis was performed to determine the clinical utility of the model.ResultsMultivariable regression analysis showed that diabetes duration ≥7 years （OR=16.36， 95%CI： 9.18-29.17）， 24 h urinary microalbumin level （OR=1.00， 95%CI： 1.000-1.001）， and glycated hemoglobin variability （OR=7.48， 95%CI： 2.28-24.51） were independent predictors of DKD occurrence（P<0.05）. The nomogram model constructed based on these parameters showed good predictive performance， with an area under the ROC curve of 0.85 （95%CI： 0.81-0.89）， a sensitivity of 74%， and a specificity of 87%. Model calibration was confirmed by the Hosmer-Lemeshow test （P=0.668>0.05）， and decision curve analysis further demonstrated its clinical applicability.ConclusionThe dynamic prediction tool developed in this study facilitates early identification of patients at high risk for DKD and provides a data-driven， individualized support tool for clinical practice.