Comparison of food consumption frequencies among NHANES and CPES children: Implications for dietary pesticide exposure and risk assessment

    loading  Checking for direct PDF access through Ovid


Characterizing food consumption patterns among children is critical to dietary pesticide exposure assessment. We have used public release data from the US National Health and Nutrition Examination Survey (NHANES) and the longitudinal Children's Pesticide Exposure Study (CPES) to illustrate the magnitude of potential error introduced by using national-scale, cross-sectional data to estimate the consumption frequencies for smaller cohorts. We focused on foods commonly consumed by children in the target CPES age and income group (3–11 years; annual household income > $75,000) and foods likely to contain organophosphorus or pyrethroid pesticide residues. We defined “percent eaters” as the percentage of study participants who reported eating a particular food in a 24-h period. We computed the weighted percent eaters and 95% confidence limits (CL) for the target age/income group using the NHANES 24-h dietary recall data and compared these with the CPES percent eaters by sampling day and season. For certain foods, particularly the seasonally available produce (for example, apples, peaches/nectarines, melon, grapes, pears, strawberries), soy milk, and peanut butter, the CPES percent eaters fell outside the NHANES 95% CLs on many sampling days. For other foods (for example, orange juice and cow's milk), differences were not readily apparent. Although the differences we observed for certain foods may be, in part, because of measurement error, they also likely reflect seasonal and geographic patterns among the CPES data that the public release NHANES data do not capture. Using NHANES data to estimate pesticide intakes from strawberries, for example, may underestimate the exposure of the CPES children, as significantly more CPES than NHANES children ate strawberries on many sampling days. For other sampling days or other foods, overestimation is also possible.

    loading  Loading Related Articles