OBJECTIVE - Assessment and follow-up of early renal dysfunction is important in diabetic nephropathy. Plasma creatinine is insensitive for a glomerular filtration rate (GFR) >50 ml/min and creatinine clearance is unwieldy and subject to collection inaccuracies. We aimed to assess the reproducibility, reliability, and accuracy of plasma cystatin C as a measure of GFR ranging from normal to moderate impairment due to type 1 diabetes in the presence of a normal plasma creatinine concentration. RESEARCH. DESIGN AND METHODS - A sensitive immunoturbidimetric cystatin C assay was examined in 29 subjects with type 1 diabetes and 11 nondiabetic subjects. Duplicate measurements of the following were collected from each subject, 2 weeks apart: cystatin C, enzymatic plasma creatinine, 24-h creatinine clearance, GFR estimated from plasma creatinine by the Cockcroft-Gault equation, and iohexol clearance as a gold standard. RESULTS - Iohexol clearance ranged from 35 to 132 ml.min(-1).1.73 m(-2). Plasma cystatin C compared well with the other clinically used tests. The reliability of cystatin C, as assessed by the discriminant ratio, was superior to creatinine clearance (3.4 vs. 1.5, P G 0.001) and the correlation of cystatin C with iohexol clearance (Rs -0.80) was similar to that of creatinine clearance (Rs -0.74) and superior to that of plasma creatinine and the Cockcroft-Gault estimate (Rs -0.54 and 0.66, respectively). Duplicate estimations were used to provide an unbiased equation to convert plasma cystatin C to GFR. CONCLUSIONS - Based on this study, cystatin C is a more reliable measure of GFR than creatinine clearance, is more highly correlated with iohexol clearance than plasma creatinine, and is worthy of further investigation as a clinical measure of GFR in type 1 diabetes.
