Biological reference materials for assaying human albumin in urine

Summary To assess the interlaboratory variation in the results of albumin measurements, we prepared albumin solutions in human urine at various concentrations within the normal range. Since some investigators have reported that albumin is unstable in some human urine samples stored at –20°C, we screened urine samples from 21 persons to identify samples that were stable under these conditions and that had low native albumin content. The urine of two donors met these criteria, and they provided urine, which we prefiltered, sterile-filtered, and spiked with commercially available human serum albumin. The albumin was characterized as pure by a Lowry assay of protein content with National Institute of Standards and Technology bovine serum albumin (standard reference material 926) as the standard and by the appearance of one band on agarose gel electrophoresis. To evaluate the necessity for additional stabilization when urine samples are stored at –20°C, a surfactant was included in one set of materials and not included in another. The materials with surfactant have been evaluated for 10.5 months and those without surfactant for 5 months. The preserved materials showed no significant loss of activity during this period. The unpreserved materials remained stable for 2 months, and then the two higher level materials appeared to loose activity. The negative slope of the highest level of unpreserved material was statistically significant (p=0.01) during this period. In our laboratory, the albumin recovered by enzyme immunoassay was 106.7% and 115.9% in two preserved normal-range materials and 102.2% and 106.3% in similar unpreserved materials.