Diagnostic prediction formula for diabetic nephropathy based on serum TFF3, SFRP5, Gal-3 and NES-1 / 国际生物医学工程杂志

ABSTRACT

Objective:To investigate the expression and correlation of serum trefoil factor 3 (TFF3), serum secreted frizzled-related protein 5 (SFRP5), galectin-3 (Gal-3), and nesfatin-1 (NES-1) in patients with type 2 diabetes(T2DM), diabetic nephropathy(DN), chronic kidney disease (CKD), and healthy controls. To explore the relationship between the above factors and the diagnosis of DN and to establish a diagnostic formula for the diagnosis of DN combined with the above four factors.Methods:In each group 36 patients hospitalized in Tianjin Third Central Hospital from April 2017 to June 2019 were enrolled. 36 healthy volunteers were also chosen as the healthy control group. After 8 to 10 hours of fasting, the venous blood of the subjects in each group was centrifuged, the serum was collected for detection, the serum levels of TFF3, SFRP5, Gal-3, and NES-1 were compared, and the Pearson correlation analysis was performed. According to whether the diagnosis of DN was repeated, the subjects were divided into the DN group and the non-DN group (including a healthy control group, T2DM group, and CKD group). The four datasets were analyzed by binary logistic regression, and the diagnostic prediction model of DN was established, which was further verified by receiver operating characteristic (ROC).Results:The serum levels of TFF3, Gal-3 and NES-1 in DN groups were significantly higher than those in healthy control group, T2DM group and CKD group (all P<0.05), but the serum level of SFRP5 in DN group was significantly lower than that in healthy control group, T2DM group and CKD group (all P<0.05). The differences between the four groups in the four aforementioned indicators were all statistically significant (all P<0.001). The Pearson correlation analysis showed that there was a significant correlation between the above four indicators (all P<0.01). The area under the ROC curve of TFF3, SFRP5, Gal-3, and NES-1 was 0.849, 0.807, 0.882, and 0.841 respectively. The area under the curve diagnosed by the combination of four indicators (0.986) was significantly higher than that of a single indicator, and the difference was statistically significant ( Z=3.75, 4.08, 3.63, 4.06, all P<0.05). Conclusions:The joint prediction model based on serum TFF3, SFRP5, Gal-3, and NES-1 can effectively improve the diagnostic accuracy of DN and provide an important basis for clinical diagnosis of DN.