Temporal variability in urinary pesticide concentrations in repeated-spot and first-morning-void samples and its association with oxidative stress in healthy individuals

Exposure of humans to pesticides is widespread. Measurement of urinary levels of pesticides and their metabolites is often used in the assessment of body burdens and exposure doses to these chemicals. An understanding of temporal variability in urinary levels of pesticides within individuals is critical for accurate exposure assessment. We examined within- and between-individual variability in concentrations of nine organophosphate and pyrethroid insecticides as well as two phenoxy herbicides in urine collected consecutively for up to 44 days from 19 individuals. Seven oxidative stress biomarkers also were measured in urine samples to elucidate their relationship with pesticide exposure. Intraclass correlation coefficients (ICCs) were calculated to assess reproducibility in urinary pesticide concentrations from repeated measures. Sensitivity and specificity analyses were performed to evaluate the suitability of spot urine to characterize average exposures. Data analysis was further limited to seven pesticides and their metabolites, which had a detection frequency of > 60%. Poor reproducibility was found for the seven pesticides and their metabolites in both spot (ICCs <= 0.24) and first-morning-void (FMV) samples (ICCs < 0.38) collected during the 44-day study period. Use of single-spot or FMV sample to classify high (top 33%) concentrations showed high specificities (0.73-0.85) but low sensitivities (0.45-0.70). The minimum number of samples (k) required per individual to estimate participant-specific mean value for pesticides (within 20% of the true values) were 28-140 and 18-119 for spot and FMV samples, respectively. Repeated longitudinal measurements of these pesticides and their metabolites in urine showed considerable within-individual variability in both spot and FMV samples. Urinary concentrations of seven pesticides and their metabolites were significantly correlated with oxidative damage to lipids, proteins, and DNA.