Retinol-binding protein 4 versus albuminuria as predictors of estimated glomerular filtration rate decline in patients with type 2 diabetes
Abstract
Background: Since the increase in some tubular damage biomarkers can be observed at the early stage of diabetic nephropathy, even in the absence of albuminuria, we aimed to investigate if urinary albumin is superior than tubular damage marker, such as serum retinol?binding protein 4 (RBP4), in predicting renal function decline (defned as estimated glomerular fltration rate [eGFR] <60 mL/min/1.73 m2) in the cohort of patients with type 2 diabetes mellitus (T2D). Materials and Methods: A total of 106 sedentary T2D patients (mean [± standard deviation] age 64.9 [±6.6] years) were included in this cross?sectional study. Anthropometric and biochemical parameters (fasting glucose, glycated hemoglobin [HbA1c], lipid parameters, creatinine, RBP4, high sensitivity
C?reactive protein [hsCRP], urinary albumin excretion [UAE]), as well as blood pressure were obtained. Results: HsCRP (odds ratio [OR] =0.754, 95% confdence interval [CI] (0.603–0.942), P = 0.013) and RBP4 (OR = 0.873, 95% CI [0.824–0.926], P < 0.001) were independent predictors of eGFR decline. Moreover, although RBP4 and UAE as single diagnostic parameters of renal impairment showed excellent clinical accuracy (area under the curve [AUC] = 0.900 and AUC = 0.940, respectively), the Model which included body mass index, HbA1c, triglycerides, hsCRP, and RBP4 showed statistically same accuracy as UAE, when UAE was used as a single
parameter (AUC = 0.932 vs. AUC = 0.940, respectively; P for AUC di?erence = 0.759). As well, the Model had higher sensitivity and
specifcity (92% and 90%, respectively) than single predictors, RBP4, and UAE. Conclusion: Although serum RBP4 showed excellent clinical accuracy, just like UAE, a combination of markers of tubular damage, in?ammation, and traditional markers has the higher sensitivity and specifcity than UAE alone for prediction renal impairment in patients with T2D.
C?reactive protein [hsCRP], urinary albumin excretion [UAE]), as well as blood pressure were obtained. Results: HsCRP (odds ratio [OR] =0.754, 95% confdence interval [CI] (0.603–0.942), P = 0.013) and RBP4 (OR = 0.873, 95% CI [0.824–0.926], P < 0.001) were independent predictors of eGFR decline. Moreover, although RBP4 and UAE as single diagnostic parameters of renal impairment showed excellent clinical accuracy (area under the curve [AUC] = 0.900 and AUC = 0.940, respectively), the Model which included body mass index, HbA1c, triglycerides, hsCRP, and RBP4 showed statistically same accuracy as UAE, when UAE was used as a single
parameter (AUC = 0.932 vs. AUC = 0.940, respectively; P for AUC di?erence = 0.759). As well, the Model had higher sensitivity and
specifcity (92% and 90%, respectively) than single predictors, RBP4, and UAE. Conclusion: Although serum RBP4 showed excellent clinical accuracy, just like UAE, a combination of markers of tubular damage, in?ammation, and traditional markers has the higher sensitivity and specifcity than UAE alone for prediction renal impairment in patients with T2D.
Keywords
Albuminuria, diabetic nephropathy, in?ammation, retinol?binding protein 4