Short-term temporal variability of urinary biomarkers of organophosphate flame retardants and plasticizers.

ABSTRACT

BACKGROUND: Exposure to organophosphate flame retardants and plasticizers (PFRs) is commonly estimated by measuring biomarker concentrations in spot urine samples. However, their concentrations in urine can vary greatly over time due to short biological half-lives and variable exposure, potentially leading to exposure misclassification. In this study, we examined the within- and between-individual and within- and between-day variability of PFR metabolites in spot and 24-hour pooled urine samples during five consecutive days. METHODS: We collected all spot urine samples from 10 healthy adults for 5 days. On one additional day, we collected 24-hour pooled urine samples. Samples were analyzed by solid-phase extraction coupled to high-performance liquid chromatography tandem mass spectrometry. We calculated intraclass correlation coefficients (ICCs) to assess the reproducibility of metabolite concentrations in morning void and spot samples. RESULTS: Fair-to-good reproducibility was observed for serial measurements of bis(1,3-dichloro-2-propyl) phosphate (BDCIPP), 1-hydroxy-2-propyl bis(1-chloro-2-propyl) phosphate (BCIPHIPP), 2-hydroxyethyl bis(2-butoxyethyl) phosphate (BBOEHEP) and 2-ethyl-5-hydroxyhexyl diphenyl phosphate (5-HO-EHDPHP) (ICC 0.396 - 0.599), whereas concentrations of diphenyl phosphate (DPHP) and 2-ethylhexyl phenyl phosphate (EHPHP) were more variable in time (ICC 0.303 and 0.234). Reproducibility improved significantly when only morning void samples were considered and when concentrations were adjusted for urinary dilution. Collecting 24-hour pooled urine could be a reliable alternative for PFR biomarkers with poor short-term temporal variability. CONCLUSIONS: The between-day variability was minor compared to variability observed within the same day, which suggests that collecting multiple samples could reduce exposure missclassification. Differences in the observed between- and within-individual variance were compound specific and related to both the nature of the exposure (e.g., diet vs other exposure routes, multiple sources) and the individual toxicokinetic properties of the investigated PFRs.