PREPARATION OF URINE SAMPLES FOR TRACE-METAL DETERMINATION - A STUDY WITH ALUMINUM ANALYSIS BY INDUCTIVELY-COUPLED PLASMA OPTICAL-EMISSION SPECTROMETRY

Urinary analysis of trace metals forms a significant role in clinical chemistry, but the optimal preparation and analysis of urine samples h as not been investigated. Human urine is generally supersaturated with dissolved solids. Therefore, samples often precipitate following coll ection. X-ray microanalysis showed that this precipitate was predomina ntly rich in calcium and phosphorus but could include some trace metal s from urine, potentially lowering their concentrations in solution. H ence, the precipitate must be fully redissolved for accurate analysis of trace metals in urine. Methods are emphasized for the best collecti on and preparation of urine samples for subsequent trace metal analysi s; in this work inductively coupled plasma optical emission spectromet ry (ICPOES) was used for the analysis of aluminium. For optimal accura cy, peak profiles were collected over 396.147 nm-396.157 nm. Urinary a luminium levels were investigated from 10 healthy volunteers and conce ntrations were obtained using either aqueous, pooled or individual uri ne-based standard curves. Since urine has a highly variable matrix, in dividual sample-based standards, which are unique to that particular s ample, gave the most accurate results. However, where sample size is s mall or sample numbers are unfeasibly large, pooled sample-based stand ards give good approximations to within 15% and, with appropriate vali dation, other elements as internal standards could also be used for ap proximations. Aqueous standards should be avoided. Spike-recovery expe riments confirmed these data since individual sample based standards s howed optimal recovery [99.3 (4.4)%], while pooled sample-based standa rds were a close proxy [101.6 (9.2)%] but aqueous standards were inapp ropriate [137.4 (12.8)%]. Postprandial urinary aluminium levels of the 10 volunteers were [7.2 (3.7)mu g/L] after analysis using individual sample-based standard curves.