Evaluation of method performance for oxidative stress biomarkers in urine and biological variations in urine of patients with type 2 diabetes mellitus and diabetic nephropathy

Oxidative stress biomarkers such as superoxide dismutase (CuZnSOD), catalase (CAT) and malondialdehyde (MDA) play an important role in the pathogenesis or progression of numerous diseases. Data regarding the biological variation and analytical quality specifications (imprecision, bias and total error) for judging the acceptability of method performance for oxidative stress biomarkers in urine are conspicuously lacking in the literature. Such data are important in setting analytical quality specifications, assessing the utility of population reference intervals (index of individuality) and assessing the significance of changes in serial results from an individual (reference change value; RCV). 20 patients with type 2 diabetes mellitus (T2DM), 20 patients with diabetic nephropathy (DN) and 14 healthy individuals as control were involved in this study. Timed first morning urine samples were taken from patients and healthy groups on the zero, 1st, 3rd, 5th, 7th, 15th and 30th days. Index of individuality and reference change value were calculated from within-subject and between-subject variations. Methods of oxidative stress biomarkers in human blood were adopted in human urine and markers were measured as spectrophotometrically. Also, analytical quality specifications for evaluation of the method performance were established for oxidative stress biomarkers in urine. Within-subject variations of oxidative stress biomarkers were significantly higher in patients with DN and T2DM compared to healthy subjects. MDA showed low individuality, and within-subject variances of MDA were larger than between-subject variances in all groups. However, CAT and CuZnSOD showed strong individuality, but within-subject variances of them were smaller than between-subject variances in all groups. RCVs of all analytes in diabetic patients were relatively higher, because of high within-subject variation, resulting in a higher RCV. Also, the described methodology achieves these goals, with analytical CVs of