The Risk Factors Of Diabetic Nephropathy Progression And Establishment Of Prediction Equation

Background and objective: The incidence rate of diabetic nephropathy(DN)from diabetes mellitus has been raised considerably in recent decades.At present,diabetic nephropathy has become the main cause of end-stage renal disease(ESRD).Although there are a large number of patients with diabetes,not all diabetic patients will progress to DN or even to ESRD.The etiology and pathogenesis of DN has not been well understood until now.We found that the strict administration of blood glucose,blood pressure and the application of ACEI/ARB are not working to reduce the new incidence of ESRD.Previous studies have drawn different conclusions about the relationship between blood glucose,blood lipid and progress of renal function.Therefore,our study aims to characterize the risk factors of T2 DN progression to ESRD and to establish a preliminary clinical prediction model,so as to provide a reference for precise treatment of DN.Methods: The clinical data of 5452 patients with T2 DN without missing case information were retrospectively collected and analyzed.All patients were randomly divided into modeling data set and validation data set at a ratio of 7:3.In the modeling data set,the patients were divided into two groups according to whether they had progressed to ESRD or not,and the differences among the clinical indicators were compared.Spearman correlation analysis was used to explore the relationship between glomerular filtration rate and clinical indicators.Univariate and multivariate logistic stepwise regression analysis was established to analyze the independent risk factors affecting the progression of T2 DN to ESRD,and clinical prediction model were preliminarily constructed.P<0.05 showed statistical significance.The predictive model was used in the validation data set to calculate each patient’s risk of progressing to ESRD.Hosmer-Lemeshow goodness of fit test,calibration curve and Brier score were used to evaluate the calibration of the prediction model,and the area under the ROC curve(AUC)was used to evaluate the discrimination of the prediction model.Results: 1.The included clinical indicators showed no statistical difference between modeling data set and validation data set,indicating that the two groups of data had comparable value.2.In modeling data set,compared with non-ESRD group,the prevalence of stroke,hypertension,hypertension and carotid intimal plaque was higher in ESRD group.The utilization rate of RASB and the levels of DM duration,SBP,BUN,SCR,SUA,24 h UTP were significantly higher and Hb,Alb,HbA1c,TG,TC,and LDL were lower(P<0.05).3.Spearman correlation analysis showed that e GFR was negatively correlated with age,duration of diabetes,24-hour UTP,BUN,Scr,SUA and positively correlated with gender,BMI,Hb,Alb,TC,TG,LDL(P<0.05).4.Multiple Logistic regression showed that the independent risk factors for renal function progression in T2 DN patients were hypertension(OR=3.402,95%CI: 2.253~5.136,P<0.001),duration of DM(OR=1.064,95%CI: 1.045~1.088,P<0.001),carotid intima plaque(OR=1.520,95%CI: 0.944~2.447,P=0.045),SUA(OR=1.003,95%CI: 1.001~1.004,P<0.001)and TC(OR=1.272,95%CI: 1.030~1.571,P=0.026)were independent risk factors affecting the progression of ESRD in T2 DN patients.Hb(OR=0.931,95%CI: 0.925~0.937,P<0.001),Alb(OR=0.893,95%CI: 0.877~0.909,P<0.001)and Hb A1c(OR=0.920,95%CI: 0.850~0.996,P=0.039)were protective factors.5.The clinical prediction equation established by the history of hypertension,duration of DM,carotid intima plaque,SUA,HbA1c,Alb and TC as the predictors was as follows: P=ex/(1+ex),X=3.969+1.224×history of hypertension(0=no,1=yes)+0.062×duration of DM(years)+0.418×carotid intima plaque(0=no,1=yes)+0.003×SUA(μmol/L)+0.241×TC(mmol/L)-0.083×Hb A1c(%)-0.113×Alb(g/L)-0.071×Hb(g/L).6.Model evaluation: Hosmer-Lemeshow goodness of fit test was performed on the prediction equation in the modeling data set,chi-score value is 8.552(P= 0.381> 0.05)and the Brier score is 0.05.It shows that the calibration of the model is reasonable.The AUC value for the area under the ROC curve is 0.942(95%CI: 0.934~0.949,P<0.001),suggesting a high degree of differentiation of this model.Among them,Youden index is 0.763,sensitivity is 86.69%,specificity is 86.65%,the positive likelihood ratio is 6.72,the negative likelihood ratio is 0.12,the positive predictive value is 43.4%,and the negative predictive value is 98.7%.In the validation data set,the Brier score is 0.08 and the AUC value of the area under the ROC curve is 0.912(95%CI: 0.897~0.925,P<0.001),Youden index is 0.701,sensitivity is 83.23%,specificity is 86.92%,the positive likelihood ratio is 6.36,the negative likelihood ratio is 0.19,the positive predictive value is 40.6%,and the negative predictive value is 98.0%.Conclusions: Hypertension,duration of DM,carotid intimal plaque,SUA and TC are independent risk factors for ESRD in T2 DN patients,while HbA1c,Alb and Hb are protective factors.The clinical prediction equation established by the history of hypertension,the duration of DM,carotid intimal plaque,SUA,HbA1c,Alb,TC and Hb as the predictors has certain value in predicting whether T2 DN patients will progress to ESRD or not.Clinically,the monitoring and management of these factors are helpful to improve the prognosis of diabetic nephropathy patients.