To assess the performance of urinary total protein measurements in timed 24-h urine collection (24-h UP) and in a diurnal random urine specimen (RUS) for the screening and diagnosis of overt diabetic nephropathy.
A total of 167 diabetic patients (20 type 1 and 147 type 2 diabetic patients; 78 women and 89 men), aged 20–84 years, collected 217 timed 24-h urine specimens. Albumin was measured by immunoturbidimetry, total protein by sulfosalicylic acid technique, and creatinine by Jaffe's method. According to the timed 24-h urinary albumin excretion rate (UAER), samples were divided into three groups: normoalbuminuric (NORMO) (UAER <20 μg/min; n = 84), microalbuminuric (MICRO) (UAER 20–200 μg/min; n = 78), and macroalbuminuric (MACRO) (UAER ≥200 μg=min; n = 55). Eighty-six patients also collected 105 RUSs (NORMO, n = 47; MICRO, n = 37; MACRO, n = 21), and urinary protein concentration (UPC) and urinary protein-to-creatinine ratio (UPCR) were measured. The receiver operating characteristics (ROC) curve approach was used to analyze the performance of the diagnostic tests.
Spearman's coefficient of correlation of 24-h UAER versus 24-h UP was 0.95 (P < 0.001), and of 24-h UAER versus UPC and UPCR were 0.77 and 0.72, respectively (P < 0.001). The calculated areas (±SEM) under the ROC curve for the diagnosis of overt diabetic nephropathy were 0.9987 ± 0.001 for 24-h UP, 0.9926 ± 0.006 for UPC, and 0.9751 ± 0.014 for UPCR. In the ROC curves, the first points with 100% sensitivity were 541 mg (95.7% specificity) for 24-h UP, 431 mg/1 (92.9% specificity) for UPC, and 0.2 (76.2% specificity) for UPCR.
Measurements of proteinuria presented almost perfect accuracy for the screening and diagnosis of overt diabetic nephropathy. Protein measurement in spot urine is a reliable and simple method for the screening and diagnosis of overt diabetic nephropathy.
