The value of retinal microvascular OCTA parameters for early detection of diabetic kidney disease in patients with non-proliferative diabetic retinopathy

INTRODUCTION: This study used optical coherence tomography angiography (OCTA) to investigate the early diagnostic value of OCTAparameters for diabetic kidney disease (DKD) in patients with non-proliferative diabetic retinopathy (NPDR). MATERIAL AND METHODS: This prospective study enrolled patients with type 2 diabetes mellitus (T2DM) into three groups: those with T2DMwithout retinopathy (group A), those with non-proliferative diabetic retinopathy (NPDR) without DKD (group B), and those with NPDR andDKD (group C). All participants underwent OCTA imaging to measure vessel density (VD) in the superficial capillary plexus (SCP)and deep capillary plexus (DCP), as well as the foveal avascular zone (FAZ) area, perimeter (PERIM), and acircularity index (AI). Pearsonor Spearman correlation analysis was used to explore the relationships between OCTA parameters and renal function indicators, including24-hour urinary microalbumin, urinary albumin-to-creatinine ratio (UACR), and blood urea nitrogen (BUN). Multivariate logistic regressionwas performed to identify independent factors associated with NPDR with DKD. Receiver operating characteristic (ROC) curveswere generated to assess the diagnostic performance of individual and combined OCTA parameters. RESULTS: Significant differences in OCTA parameters and renal function indicators were observed among the three groups. Correlation analysis revealed that SCP-VD and DCP-VD were negatively correlated with 24-hour urinary microalbumin, UACR, and BUN, whereas FAZ area, PERIM, and AI were positively correlated with these renal markers. Multivariate logistic regression identified SCP-VD [oddsratio (OR) = 0.67], DCP-VD (OR = 0.68), PERIM (OR = 1.30), and AI (OR = 1.04) as independent factors associated with NPDR and DKD.Analysis of diagnostic performance showed that among the single parameters, SCP-VD had the best diagnostic value [area under the curve(AUC) = 0.80]. Among the different combination models, the combination of all four parameters demonstrated the highest diagnosticperformance (AUC = 0.94). CONCLUSIONS: OCTA parameters (SCP-VD, DCP-VD, PERIM, AI) are closely associated with NPDR and DKD. A combined multiparametermodel demonstrated excellent diagnostic efficacy for early detection. Retinal OCTA parameters may reflect early microvascular alterationsassociated with diabetic kidney disease.