Public health concerns about pesticide exposure to young children have received increased attention following the publication of “Pesticides in the diets of infants and children” in 1993. In 1996, the U.S. Food Quality Protection Act required the U.S. Environmental Protection Agency to set food tolerances that account for dietary and non-dietary exposure and protect sensitive populations. Biomonitoring studies have confirmed that children are widely exposed to pesticides, including organophosphorus , pyrethroid, fungicide, and organochlorine pesticides. Diet is an important source of pesticide exposure in children. For example, Lu et al. reported that the median urinary concentrations of the specific metabolites for malathion and chlorpyrifos decreased to undetectable levels after the introduction of organic diets in school-aged children. Several studies have confirmed that children may also be exposed to pesticide contamination in home and daycare environments. Children living in agricultural areas may also be exposed to pesticides through drift during applications or volatilization from nearby fields and parental take-home exposures. Lu et al. found that children who live in agricultural communities had five times higher OP metabolite levels in their urine compared to children who resided in non-agricultural communities. These researchers also found higher residential OP pesticide contamination and/or elevated urinary metabolite levels in children living near orchards. Higher exposure to children living in agricultural areas has raised environmental justice concerns and has resulted in proposals to define farm worker children as a vulnerable population that need additional protections by the U.S. EPA. Identifying pesticide exposure determinants is needed to identify sources and pathways of pesticide exposure in children and contribute to policies aiming to reduce exposure. To date, no longitudinal studies have investigated factors associated with pesticide exposure in very young children. We hypothesize that exposure factors will vary over time given the changes in diet, behavior, and family practices that occur as children age. In this study,square flower bucket we report levels of OP pesticide metabolites in 6, 12, and 24 month old children participating in the CHAMACOS birth cohort study in the Salinas Valley of California, an agricultural area.
We examined potential determinants of exposure associated with OP urinary metabolite levels at each age point, including sex, child behavior, diet, home pesticide use, season, parental work status, and proximity of homes to fields. We focused on OPs because they are commonly used in the Salinas Valley and were the first pesticide class re-examined under the FQPA.Mothers were interviewed when the children were 6, 12, and 24 months old. Interviews were conducted in Spanish or English by bilingual interviewers. Information collected included demographics, household enumeration, occupational status, whether work clothes were worn into the home, home pesticide use, presence of pets, daily servings of child fruit and vegetable consumption based on a modified food frequency questionnaire, time spent in child care, location of child care relative to fields, and frequency of hand washing and how often child fingers, hands, or toes are placed in the mouth. The interview also included a Child Behavior Checklist which uses a standardized format to assess parent-reported behavioral characteristics of children. Based on the CBCL, we selected child temperament indicators that we hypothesized could be associated with behaviors that affect pesticide exposure: “Can’t sit still, restless, or hyperactive”, “Gets into everything”, “Quickly shifts from one activity to another”, and “Underactive, slow moving, or lacks energy.” Shortly after each interview, study staff conducted a home inspection. Recorded information included distance between the home and agricultural fields, carpeting, housekeeping quality, and a detailed inventory of home pesticides. Home visits were completed for 87%, 84%, and 87% of the enrolled children at 6-, 12-, and 24-months, respectively. Urine samples were analyzed by the Centers for Disease Control and Prevention in Atlanta, Georgia. We measured six non-specific DAP metabolites of OP pesticides metabolites: dimethylphosphate ; dimethyl-dithiophosphate ; dimethylthiophosphate ; and three diethyl alkylphosphate metabolites: diethylphosphate ; diethyldithiophosphate ; and diethyl-thiophosphate by isotope dilution gas chromatography-tandem mass spectrometry. We measured DAPs, rather than pesticide-specific metabolites, because there are no laboratory methods to measure specific metabolites of several OP pesticides used in the study area, such as oxydemeton-methyl. Approximately 80% of the OP pesticides used in the Salinas Valley devolve to a DAP metabolite . Creatinine concentrations were determined in urine using a commercially available diagnostic enzyme method . Laboratory quality control included repeat analysis of three in-house urine pools enriched with known amounts of pesticide residues whose target values and confidence limits were previously determined. The validity of each analytical run was determined using the Westgard rules for quality control. The limits of detection ranged from 0.08 g/L for DMDTP to 1.1 g/L for DMTP.
Metabolite levels below the LOD were randomly imputed based on a log-normal probability distribution. Because individual OP pesticides can devolve to more than one DAP metabolite, we summed the DAPs on a molar basis to reflect total DMAP or DEAP metabolites. Frozen field blanks, prepared earlier by CDC, were defrosted, re-packaged in the field in a manner identical to collection procedures for actual samples, and then shipped blinded to CDC. The mean levels of individual DAPmetabolites in 57 blank field samples were <2 g/L. The median values of the DAP metabolites in the field blanks were all below the detection limit. All data analyses were performed with Stata Version 10 . We first computed descriptive statistics and percentiles for individual and total DMAP and DEAP metabolites at each sampling time point. We used Pearson correlations and ANOVA to assess bivariate associations between the metabolite levels and potential exposure determinants selected a priori, including sex, age, produce intake, breastfeeding, season, distance to agricultural fields, occupation of household members, wearing work clothes or shoes into the home, home pesticide use, presence of carpets, presence of pets, and housekeeping quality. We examined post facto additional determinants which may be related to drift of pesticides from fields, including daily rainfall, behaviors which may modify exposures , time spent in child care, and proximity of child care to agricultural fields. We then constructed generalized linear mixed models with log10-transformed DMAP or DEAP metabolite levels as the dependent variables and potential exposure determinants found to have significant bivariate relationships. The models included a random effects term to adjust for the lack of independence of repeated measures on the same subject. Because children’s development, diet, and behavior differ at different age points, we also examined whether age modified any associations, with 12-month olds and 24-month olds compared to 6-month olds as the reference. All interaction terms were included in the final DMAP and DEAP models. Based on the final models, we used linear combination equations to compute the percent differences in log DMAP and DEAP metabolites for the predictor variables to determine the effect of these predictors on metabolite levels among the 6-, 12- and 24-month old children. To assess bias due to loss to follow up, we ran the models with weights equal to the inverse probability of inclusion in the final sample at each time-point. We then performed the analyses without the weights for comparison. For statistical analyses, we present results that are not adjusted for creatinine.We also included urinary creatinine as an independent variable in the final multi-variable mixed DMAP and DEAP models for comparison with models without the urinary creatinine variable.We investigated the relationship between potential exposure determinants and urinary pesticide metabolite levels in ~400 children followed through infancy and toddlerhood living in an agricultural community. All children had detectable levels of OP metabolites in their urine. Consistent with previous studies,black flower bucket the DMAP metabolite levels were higher than the DEAP metabolite levels. We observed three-fold higher DMAP levels in 24-month olds and two-fold higher levels in 12-month olds relative to 6 month olds; however DEAPs declined between 12 and 24 months. Nearby agricultural use of dimethyl and diethyl OP pesticides was generally stable over the study period, however, most residential uses of chlorpyrifos and diazinon, two diethyl OP pesticides, were cancelled.
CHAMACOS children turned 12 months during the first year of the residential ban, which was phased in gradually. Thus, the decrease in DEAP metabolite levels among 24-month olds may be related to reduced indoor contamination of chlorpyrifos and diazinon , due to the residential use ban. This hypothesis is supported by our finding in a separate study that chlorpyrifos and diazinon house dust levels declined in Salinas Valley homes between 2000 and 2006. However, the ontogenetic increase in DMAP levels cannot be explained by changes in dimethyl pesticide use which did not change substantially during this time. The increase in DMAP levels may be due to increasing exposure-related behaviors and changes in diet as the children age in an environment where dimethyl OP pesticide use was relatively constant. Associations between the two classes of DAP metabolites and exposure determinants were not consistent at different age points. Possible reasons include differences in usage patterns, physical-chemical properties of the pesticides, field degradation, environmental transport, and metabolism of the dimethyl versus the diethyl OP pesticides. For example, malathion, which devolves to a DMAP metabolite, has a relatively high vapor pressure compared to other OP pesticides, and, thus, may result in greater exposures via inhalation. The spring/summer season, when malathion use is higher, was associated with higher DMAP levels in six-month olds, who are not yet crawling, suggesting an inhalation exposure pathway. We also found that recent rainfall was associated with lower DMAP levels in the younger children, a finding consistent with our previous study that showed rainfall was associated with lower OP levels in air. Together, these findings support the hypothesis that inhalation may be an important pesticide exposure route for very young children. Overall, our findings suggest that agriculture-related determinants of pesticide exposure may be associated with measured exposure at some ages, but we did not observe consistent associations across age points, or between DMAP and DEAP metabolites. The high variability in pesticide application frequency and the nature of transient, non-persistent exposures in young children may create too much variability to statistically model the association of these variables and child exposures. In contrast, intake of fruits and vegetables was consistently and positively associated with both classes of urinary metabolites in children at all ages, and was statistically significant for DMAP metabolites in 6- and 24-month old children, suggesting that diet is an important pesticide exposure pathway. This finding is consistent with recent studies that indicate diet is an important source of pesticide exposure to children.Few studies report levels of pesticide metabolites in children 6- to 24-months old. Median total DAP metabolite levels in the CHAMACOS children at 6, 12, and 24 months of age were lower than levels in 10 crawling infants and 10 toddlers sampled in the Salinas Valley in 2002. These twenty children were from farm worker homes and sampled in the summer, when levels might have been higher; direct comparisons, however, are limited by the small sample size. Median total DMAP and DEAP metabolite levels in the CHAMACOS 6- to 24-month olds were lower by ~30–70% than levels in children 24- to 72-months old living in Washington state agricultural or suburban areas; however, the Washington children were older than the CHAMACOS participants and the samples were collected between 1997 and 1999, before restrictions on residential use of chlorpyrifos and diazinon were implemented. Thus, these populations may not be directly comparable. Creatinine-adjusted levels were similar to adjusted concentrations reported in 41 5- to 73-month old farm worker children living on the US/Mexico border. Due to age differences, it was not possible to compare DAP levels in these CHAMACOS children with levels in older children studied by the National Health and Nutrition Examination Survey. Representative pesticide-exposure studies of national and state-wide populations are needed to compare to regional or local studies in impacted communities. Our study has several limitations. In a setting where multiple OP pesticides are used, measurement of the non-specific DAP metabolites does not provide information on exposure to the specific parent OP compound. As noted above, the many OP pesticides used in the Salinas Valley have widely varying usage, environmental persistence, and physical-chemical properties, adding variability to bio-monitoring measurements and possibly biasing statistical models toward null results.