Potential role of renal artery resistance index in differential diagnosis of diabetic nephropathy and construction of a predictive model

Objective To analyze the potential role of renal artery resistance index (RI) for the differential diagnosis of diabetic nephropathy (DN), and construct a quantitative predictive model to guide clinical application.

Methods A total of 312 patients with type 2 diabetes were retrospectively included, including DN group(187 patients with DN) and non-diabetic nephropathy (NDN) group(125 patients) according to renalpuncture and pathology results. The clinical data between the two groups were compared. Multivariate Logistic regression analysis was used to screen the risk factors of DN and the predictive model was established. Receiver operator characteristic (ROC) curve was drawn, and the areas under the curve (AUCs) of the model and RI for DN diagnosis were compared.

Results Compared with patients in the NDN group, systolicblood pressure, fasting blood glucose, serum creatinine, urea nitrogen and RI of patients in the DN group were significantly higher, the course of disease ≥ 60 months, glycosylated hemoglobin (HbA1c) ≥ 7.0% and ratio of patients complicating diabetes retinopathy (DR) were significantly increased, while hemoglobin and estimated glomerular filtration rate (EGFR) were significantly reduced (P < 0.05). Multivariate Logistic regression analysis showed that the course of disease ≥ 60 months (OR=3.526; 95%CI, 2.425 to 5.023; P < 0.001), DR (OR=5.528; 95%CI, 4.426 to 6.325; P < 0.001), HbA1c ≥ 7.0% (OR=1.958; 95%CI, 1.235 to 3.526; P < 0.001) and RI ≥ 0.65 (OR=4.025; 95%CI, 3.265 to 5.524; P < 0.001) were the independent risk factors to DN (P < 0.05). ROC curve showed that AUC of nomograph model containing RI for DN diagnosis was significantly higher than RI alone and the nomograph without RI(0.701, 0.799; P < 0.001). Spearman correlation analysis showed that RI was positively correlated with systolic blood pressure, course of disease, blood creatinine and urea nitrogen, but negatively correlated with eGFR (P < 0.05).

Conclusion Noninvasive quantitative detection of RI by ultrasound has important potential for the differential diagnosis of DN. The nomograph model combined with disease course, DR, HbA1c and RI has high efficiency in the diagnosis of DN.