Analyses of location differences using pesticide loadings did not change our findings

One study showed that this residential phaseout resulted in decreased air concentrations among low income households in New York City. However, these OPs are still used in agriculture and trends in residential contamination of these compounds have not been studied in agricultural communities, where pesticide drift and transport from fields on work clothing may impact indoor pesticide concentrations. It is also widely accepted that house dust is a reservoir for environmental contaminants with concentrations remaining fairly stable; however, to our knowledge, only one study has documented the temporal stability of pesticides in house dust focusing on the OP pesticide chlorpyrifos. Additionally, exposure data on other contemporary-use pesticides in low income households is limited. In this study, we characterized and compared house dust levels of agricultural and residential-use pesticides from low-income homes in an urban community and an agricultural community . We evaluated the correlation of several semi- and non-volatile pesticide concentrations in samples collected several days apart from the same general area in the home; and examined whether house dust concentrations of chlorpyrifos and diazinon declined in Salinas, CA after the U.S. Environmental Protection Agency’s voluntary residential phase-out of these compounds. Finally, we estimated resident children’s potential non-dietary ingestion exposures to these indoor contaminants to determine if exposures via this pathway exceeded current U.S. EPA recommended guidelines.Study participants included families with children between 3 and 6 years of age who were participating in a 16-day bio-monitoring exposure study conducted during July through September 2006. Through community health clinics and organizations serving low-income populations,movable racking system we recruited a convenience sample of 20 families living in Oakland, CA, and 20 families living in Salinas, CA .

Participating families were Mexican American or Mexican immigrants and all Salinas households included at least one household member who worked in agriculture. The University of California, Berkeley Committee for the Protection of Human Subjects approved all study procedures and we obtained written informed consent from parents upon enrollment.Using standard protocols, we collected dust samples from an area 1 to 2 m2 with a High Volume Small Surface Sampler which collects particles >5 μm. Most dust samples were collected from carpets where parents indicated children spent time playing, except for two homes with no carpets or rugs, for which we collected samples from upholstered furniture using an attachment on the HVS3. To assess the consistency of concentrations within homes, we collected up to two dust samples, 5-8 days apart, from the same general location in each home. Dust samples were then manually sieved to obtain the fine fraction , which is more likely to adhere to human skin. This fraction was stored at -80°C prior to shipment to Battelle Memorial Institute in Columbus, Ohio for laboratory analysis.Of the 40 homes sampled, 15 Salinas farm worker and 13 Oakland urban homes had sufficient sample mass for analysis after measurement of other analytes . We analyzed two dust samples per home except for one home in each location from which one sample was analyzed, yielding a total of 54 dust samples. For this study, a total of 25 analytes were measured in every sample. Analytes measured included the OP insecticides chlorpyrifos, diazinon, malathion, methidathion, methyl parathion, phorate, tetrachlorvinphos, and one oxidation product of diazinon, diazinon-oxon; the pyrethroid insecticides bifenthrin, allethrin , cypermethrin , cis- and trans-permethrin, deltamethrin, esfenvalerate, imiprothrin, and prallethrin; the pesticide synergist commonly added to pyrethroid formulations piperonyl butoxide; the herbicide chlorthal-dimethyl; and the fungicide iprodione.

We selected target analytes based on regional agricultural and non-agricultural use as reported in the California Department of Pesticide Regulation Pesticide Use Reporting database, active ingredients in pesticides used or stored indoors, detection in our prior studies, and laboratory feasibility. Select physico-chemical properties of the target analytes and information on county-level agricultural and non-agricultural pesticide use in both study locations are provided in the Additional files section . To measure analytes, we modified a previously published laboratory method. Briefly, 0.5 g dust aliquots were fortified with 250 ng of two surrogate recovery standards –fenchlorphos and 13C12- trans-permethrin–and extracted using ultrasonication in 1:1 hexane:acetone. We used solid phase extraction for sample cleanup, concentrated extracts to 1 mL and then fortified them with an internal standard, dibromobiphenyl. Concentrated extracts were analyzed with an electron impact gas chromatrography mass spectrometer in the multiple ion detection mode with temperatures programmed from 130-340°C at 6°C/min. For each sample analysis set, we analyzed seven calibration curve solutions ranging from 2 to 750 ng/mL and used a linear least squares regression and the internal method of quantification to prepare calibration curves. A solvent method blank, matrix spike sample , and duplicate study sample were included in each sample analysis set for quality assurance and quality control purposes. We also determined the relative percent difference of the duplicate samples for each analyte measured to ensure that the analytical precision was within acceptable limits. No analytes were detected in the four solvent method blanks, indicating no laboratory contamination. Analyte recoveries in four randomly-selected matrix spike samples averaged 117 ± 19% for OP analytes, 115 ± 16% for pyrethroid analytes, 82 ± 5% for chlorthal-dimethyl, 112 ± 14% for iprodione; and average SRS recoveries were 113 ± 6% and 128 ± 5% for fenchlorphos and 13C12- trans-permethrin, respectively. The average relative percent difference in concentration for the 12 analytes detected in duplicate samples was 14 ± 18% , indicating good analytical precision.We first summarized demographic characteristics and computed descriptive statistics for all analytes by location. For subsequent analyses, we focused on analytes frequently detected .

Concentrations below the limit of detection were assigned a value of LOD/√2 [30] and results were considered significant at p < 0.05. We used Fisher’s Exact tests to determine if analyte detection frequencies differed between locations. To assess differences in concentrations between study locations, we used linear regression models with a generalized estimating equations approach in order to report robust inference that accounts for the non-independence of repeated samples within households. Given the limited number of homes sampled and the homogeneity of the study population, we excluded demographic characteristics as covariates in GEE models. We also examined location differences using analyte loadings, ng/m2. We calculated loadings by multiplying analyte concentrations by the sieved fine mass and dividing by the area sampled. To determine the correlation of analyte concentrations between the first and second collections,vertical farming products we computed Spearman rank-order correlations. To examine temporal trends of chlorpyrifos and diazinon concentrations in farm worker homes after the residential phase-out, we used Wilcoxon Mann-Whitney tests to compare dust concentrations in the 15 Salinas farm worker households sampled in 2006 from our present study with dust concentrations from a subset of 82 Salinas farm worker homes of participants in the CHAMACOS study sampled between 2000 and 2002 , and 20 similar households sampled by Bradman et al. in 2002. The same laboratory and collection methods were used in all studies. In addition, we restricted comparisons to those study homes located in the same zip codes as the homes in the present study. If multiple dust samples were available from any of the study homes in the same year, including the present study, the mean analyte dust concentration was used in our analyses. There were no demographic or household differences between our previous studies and the present study; e.g., all households had at least one farm worker residing in the home and study participants generally represented the farm worker population in Salinas Valley: primarily Mexican or of Mexican descent; Spanish-speaking; low literacy; low income; and frequently reported pesticide applications in the home and wearing work clothes and shoes indoors. Homes were also located >200 feet from the nearest agricultural field. Using the California Department of Pesticide Regulation Pesticide Use Reporting database, we also computed county-level agricultural and non-agricultural usage of these OP pesticides during 1999-2007 to determine whether temporal changes in residential dust concentrations were concurrent with regional use patterns. Non-agricultural uses included applications for landscape maintenance, public health, commodity fumigation, rights-of-way, and structural pest control applications by licensed applicators which are reported to the state. Finally, to determine if exposures via the non-dietary ingestion pathway exceed U.S. EPA guidelines for the children in the present study, we calculated hazard quotients for the majority of the detected analytes.

We focused on the children given their unique vulnerabilities to environmental toxicants. We calculated the HQ as the ratio of the child’s potential daily toxicant intake at home via non-dietary ingestion to the specific toxicant chronic reference dose, RfD, .Except for farm worker status, demographic characteristics were similar in both study locations . Participating households were within 200% of the poverty line and approximately 50% or more of the homes had at least six household members. Although not statistically significant, pest sightings were more commonly reported in Oakland urban homes compared to Salinas farm worker homes. Most participants reported using pesticides indoors in the three months preceding the study and the most common location of use was the kitchen. Hand-held pyrethroid sprays were the most common formulation and application method in both locations; applications were mostly targeted at ants and cockroaches. Participants from three homes reported applying pyrethroid insecticides between the two sampling dates. No products with OP insecticides were stored or reported applied in the homes, at the workplace or on pets. Most participants from Salinas households reported that farm workers residing in the home wore their work clothing indoors and about half of them also wore their work shoes indoors. Approximately 27% of Salinas farm worker homes were located <1/4 mile from the nearest agricultural field or orchard.We detected 21 of the 25 analytes measured . The majority of homes had at least three analytes detected in dust; 79% of the homes had at least six analytes detected and <1% of Salinas farm worker homes had up to 14 analytes detected in one sample. Cis- and trans-permethrin were the only insecticides detected in every home. Commonly detected OP pesticides included diazinon and chlorpyrifos. Diazinon was detected in 79% and 52% of the samples collected from Salinas farm worker and Oakland urban homes, respectively. Chlorpyrifos was detected in 55% and 36% of the samples collected from Salinas farm worker homes and Oakland urban homes, respectively. Other commonly detected analytes in samples collected from both locations included: allethrin , cypermethrin , and piperonyl butoxide . Detection frequencies were only significantly different between locations for chorthal-dimethyl, which was detected solely in Salinas farm worker homes. Median concentrations of diazinon, chlorpyrifos, permethrins, allethrin, and chlorthal-dimethyl were higher in Salinas farm worker homes compared to Oakland urban homes; however, only chlorthal-dimethyl concentrations were significantly different between locations.Dust concentrations from furniture samples in two farm worker homes were comparable to those collected from carpets in other farm worker homes for frequently detected OPs , piperonyl butoxide, and chlorthal-dimethyl, while for frequently detected pyrethroids, concentrations were generally at the upper end of the distribution. We observed the same general pattern when using loadings. Maximum permethrin concentrations in farm worker homes were observed in furniture samples; however, the highest permethrin concentrations were observed in carpet samples from urban homes. The highest loading observed for cypermethrin was collected from a furniture sample; however, higher loadings were observed in carpet samples from urban homes. No location differences in pesticide concentrations or loadings were observed when we excluded furniture samples from our analysis. Some of the less frequently detected analytes were detected with greater frequency in Salinas farm worker homes and at higher maximum concentrations than in Oakland urban homes. Conversely, the 95th percentile and maximum concentrations for malathion and deltamethrin were higher among Oakland urban homes . Although not statistically significant, we generally observed higher dust concentrations in homes that reported recent pesticide use when pesticide containers were available to confirm the active ingredients. For example, in one home where bifenthrin had been applied less than a week before the first sample collection, concentrations were up to 200 times higher than the median concentration observed in other homes. Cypermethin was applied in one farm worker home, while imiprothrin was applied in two urban homes between the two sampling dates. For the farm worker home, cypermethrin dust concentrations were at the upper end of the distribution among other farm worker homes .