Creatinine adjustment of biological monitoring results

Background Biological monitoring (BM) aids exposure assessment but where this is based on incomplete collections of single urine voiding measurement of creatinine is often used to adjust analyte concentrations for the effects of fluid balance. Aims To provide reference data on creatinine concentrations in urine samples from a population of UK workers. Methods Urine samples sent to the Health and Safety Laboratory were analysed for creatinine by an automated kinetic Jaffe technique using alkaline picric acid and the results stored in a database. Statistical analysis of the data used linear mixed effects models on the natural log-transformed data. Results Between 1996 and 2007, the laboratory analysed 49 506 urine samples from 20 433 UK adult workers. In the 42 817 samples where gender was known, 93% were from men and 7% were from women. The overall mean and median creatinine concentrations were both 12 mmol/l corresponding to 1.36 g/l. The mean (13 mmol/l) and median (12 mmol/l) creatinine concentrations for men were higher than those (9 and 10 mmol/l, respectively) for women. Conclusions Gender differences in creatinine concentrations and the range of 0.3-3.0 g/l (2.653 and 26.53 mmol/l) traditionally used for confirming acceptability of urine samples mean that 2.5% of samples from male and 9% from female workers were flagged as 'low creatinine' and required a repeat sample. In addition, care should be taken interpreting any apparent gender differences in BM results to ensure that they are due to exposure and not an artefact of creatinine adjustment.