Diet and nondiet predictors of urinary 3-phenoxybenzoic acid in NHANES 1999-2002

BACKGROUND: 3-Phenoxybenzoic acid (3PBA), a pyrethroid metabolite, was detected in 75% of urine samples analyzed for pesticides in the U.S. National Health and Nutrition Examination Survey (NHANES) 1999-2002. NHANES also includes 24-hr diet data and information on household pesticide use, activities, occupation, demographics, and other exposure factors. OBJECTIVES: The objective of our study was to explore the relative importance of diet versus nondiet predictors in explaining variability in urinary 3PBA. A secondary objective was to explore whether the NHANES data could be used to identify particular foods driving 3PBA levels. METHODS: We divided subjects into child (6-10 years of age), teen (11-18 years), and adult (>= 19 years) age groups and restricted our analyses to subjects in the morning sampling session who fasted for >= 8 hr beforehand. Regression modeling consisted of several model-building steps and a final Tobit regression on the left-censored log 3PBA measurements. We also conducted bootstrap analyses to evaluate the stability of the regression parameters. RESULTS: Reported household pesticide use was not significantly associated with urinary 3PBA in any age group. Diet was significant for all three groups, and certain foods appeared to contribute more than others. Among adults, tobacco use was positively associated with 3PBA (p = 0.0326), and positive associations were suggested with the number of cytochrome p450-inhibiting medications taken (p = 0.0652) and minutes spent gardening (p = 0.0613) in the past month. CONCLUSIONS: Although exploratory, our findings underline the importance of collecting accurate data on household pesticide use and dietary intake when evaluating pyrethroid exposure-biomarker relationships.