Influence of biological and analytical variation on urine measurements for monitoring exposure to cadmium.

Abstract

OBJECTIVE: To define the mean intraindividual biological variation in urinary concentrations of cadmium and retinol binding protein (RBP) in untimed, random urine samples and the influence of creatinine or specific gravity correction on reducing this variation. The relation between biological variation and analytical variation in defining uncertainty in a single measurement and significant differences between successive measurements was explored. METHODS: Repeat measurement study in subjects with either high historical exposure to cadmium but without current exposure, or unexposed volunteers. Standard statistical tools used in clinical laboratory medicine were applied to define intraindividual biological and analytical variation. RESULTS AND CONCLUSION: Both creatinine and specific gravity correction of urinary cadmium measurements in random urine samples seem to reduce the intraindividual variability compared with uncorrected values. With a standard definition, acceptable long term analytical precision for measurements of cadmium and RBP combined with creatinine analyses should be < 9% and < 15% respectively. The mean intraindividual biological variation of cadmium and RBP, expressed as creatinine corrected, was 18% and 40% respectively in the subjects exposed to cadmium. With the analytical precision used, significant differences (p < 0.05) between consecutive measurements for creatinine corrected urinary cadmium and RBP would need to show changes of > 54% and > 110% respectively. The relation between significant differences in consecutive results and differences in the analytical precision of the method used to measure the samples is described.