Associations between food frequency checklist responses and internal biomarkers of exposure to pesticides and heavy metals among U.S. pregnant women.

A self-administered food consumption frequency checklist was completed over a 3-day period by 640 pregnant women to assess its utility for identifying associations between exposure to chemical contaminants and food consumption. Responses were segregated by food categories. Demographic information was self-reported. Chemical measurements included cadmium, lead, and methylmercury in blood samples and urinary (creatinine corrected) total arsenic, six organophosphate metabolites (dialkylphosphates), and summed polyaromatic hydrocarbon (PAH) metabolites. Frequencies for each food group were compared to demographics and geometric mean concentrations of each contaminant to determine any associations using a one-way analysis of variance. Age, race/ethnicity, education, and household income showed significant associations with several food groups (p < 0.05). Methylmercury concentrations were significantly associated with consumption of fish. Lead concentrations were significantly associated with consumption of milk, dairy, grains, fruits, and vegetables (p < 0.05). Urinary metabolite concentrations indicated significant associations between total arsenic and fish; dimethyldithiophosphate and water, protein, and fish; dimethylthiophosphate and fish; diethylthiophosphate and fruit and fish; and summed PAHs and water and grains (p < 0.05). Demographic characteristics play a role in potential exposures to these participants. The usefulness of the food frequency checklist to predict chemical exposures through biomarkers in blood and urine was mixed across chemicals and food types in that sometimes the prediction matched expected results (e.g., methylmercury and total arsenic concentrations associated with fish consumption) and sometimes it did not (e.g., no association between summed PAH concentrations and consumption of protein).