Background: There is increasing clinical interest in accurate measurement of urine albumin as an early indicator of kidney disease. However, urine is a highly variable sample matrix that may exert matrix effects on assays for urine albumin.
Methods: Variation in urine composition was assessed for components routinely measured in the clinical laboratory over 1 year. Solutions of representing different concentrations of urine components were prepared and spiked with a constant amount of albumin. Fourteen urine specimens of variable composition were ultrafiltered and spiked with a constant amount of albumin. Concentrations of albumin in solutions were determined with an immunoturbidimetric assay on the Beckman LX20 analyzer and with a competitive immunoassay on the Siemens Immulite analyzer.
Results: Variation in concentrations of most constituents altered measured urine by <10%. High NaCl concentration increased assay values by 10% or more. Addition of 0.5% Triton X-100 increased values in the LX20 assay by about 20%. Most ultrafiltered urine specimens had <10% effect on measured albumin at a concentration near 20 mg/l, but one specimen was noted to decrease measured albumin by approximately 20% in the LX20 assay.
Conclusions: Urine matrix components exert small but potentially significant effects on assays for urine albumin. Urine matrix effects should be considered in the design and evaluation of assays for albumin.