Assessment of sublingual microcirculation for the screening of diabetic nephropathy

Objective To investigate the potential of employing sublingual microcirculation as an early noninvasive screening technique for diabetic nephropathy (DN). Research design and methods We recruited 89 patients with type 2 diabetes mellitus (T2DM) and 41 healthy subjects in this cross-sectional observational study. All participants underwent fluorescein fundus angiography, vibration perception testing, 10 g (Semmes–Weinstein) monofilament examination, nerve conduction velocity, and 24-h urine microalbumin determination. HbA1c, fasting plasma glucose, blood lipid, and estimated glomerular filtration rate(eGFR) were measured. Sublingual microcirculatory images were captured using side-stream dark-field (SDF) microcirculation microscopy, and total and perfused vascular density (TVD and PVD) were calculated. Results The sublingual microcirculatory parameters denoting microvascular density and perfusion were negatively correlated with both fasting plasma glucose (TVD, r = − 0.316, P < 0.001; PVD, r = − 0.350, P < 0.001; PPV, r = − 0.279, P = 0.001) and HbA1c (TVD, r = − 0.367, P < 0.001; PVD, r = − 0.423, P < 0.001; PPV, r = − 0.399, P < 0.001). Diabetes patients already had a reduction in sublingual microcirculation compared with healthy control, and more severe reductions in TVD (7.07 ± 1.64 vs. 9.67 ± 1.94 mm/mm 2 , P < 0.001) and PVD (5.88 ± 1.82 vs. 8.64 ± 2.46 mm/mm 2 , P < 0.001) were found in those diabetes patients developed microvascular complications. Sublingual microcirculation impairment was accompanied with higher urinary albumin creatinine ratio (UACR). Receiver operating characteristic (ROC) analysis showed that TVD (area under the curve, AUC = 0.890 [0.836 0.944], P < 0.001) and PVD (AUC = 0.883 [0.826, 0.940], P < 0.001) could be indicators for DN screening. We derived a combined predictor index (CPI) considering both TVD and PVD for screening DN, and both the AUC (0.892, [0.838 0.945], P < 0.001) and cutoff point of 11.30 mm/mm 2 showed great improvement (sensitivity: 95.5%, specificity: 67.4%). Conclusions Diabetes patients experienced impaired sublingual microcirculation, which was closely correlated with UACR. Sublingual microcirculation monitoring could be used for the noninvasive early detection of DN.