Abstract:Objective To investigate the correlation between serum pancreatic regenerating (Reg) protein Ⅰα and chronic kidney disease (CKD) in diabetic patients, to develop a risk prediction model for diabetes combined with CKD. Methods The clinical data of 500 diabetic patients admitted to the Zhongda Hospital Southeast University from September 2021 to September 2023 were collected for a retrospective study, including general data, medical history, and laboratory indexes［blood routine tests, fasting plasma glucose (FPG), blood urea nitrogen (BUN), serum creatinine (Scr), serum uric acid (UA), and urinary albumin/creatinine ratio (UACR)］. Additionally, the estimated glomerular〖LM〗 filtration rate (eGFR) was calculated. Serum Reg Ⅰα level was measured by enzyme-linked immunosorbent assay, and patients were classified into groups R1 (Reg Ⅰα≤38.60 ng/mL,n=98), R2 (38.60＜Reg Ⅰα≤82.17ng/mL,n=97), R3 (82.17＜Reg Ⅰα≤196.15 ng/mL,n=97) and R4 (Reg Ⅰα＞196.15 ng/mL,n=97), according to quartile levels. Patients were randomly divided into a training set (234 cases) and a validation set (155 cases) using a random number table method in a ratio of 6∶4. Multivariate logistic stepwise regression was utilized to identify model variables and create a nomogram model for predicting the risk of diabetes combined with CKD, and evaluate its efficiency. Results (1) A total of 389 diabetic patients with complete data were included, and 210 (53.98%) patients with CKD. Patients in the R4 group exhibited significantly higher levels of FPG, BUN, Scr, and UA, increased rates of hypertension, coronary heart disease, and CKD, as well as lower levels of red blood cell count, hemoglobin(HGB), and eGFR compared to those in the R1 group (P＜0.05). (2) The indicators in the training and validation sets were found to be well-balanced (P＞0.05). In the training set, the CKD group showed higher levels of Reg Ⅰα, white blood cell count, BUN, Scr, and UA, higher prevalences of hypertension and coronary artery disease, and lower levels of erythrocyte count, HGB, and eGFR compared to the NCKD group (P＜0.05). Binary logistic regression analysis was performed on the basis of collinearity diagnosis,and the results showed that history of hypertension (OR=2.901), Reg Ⅰα＞82.17 ng/mL, decreased HGB (OR=0.965) and increased UA (OR=1.005) were the risk factors of diabetes complicated with CKD (P＜0.05). (3) The areas under the ROC of the training set and validation set were 0.846(95%CI: 0.796-0.896) and 0.920(95%CI： 0.875-0.965), respectively. The calibration curves and the Hosmer-Lemeshow test indicated good agreement of the predicted and observed outcomes (P＞0.05). Conclusion Hypertension, elevated Reg Ⅰα levels, decreased HGB and elevated UA are risk factors for CKD in diabetic patients, and the nomogram model constructed in this study has good predictive power and clinical value.