Application to Occupationally and Non-Occupationally Exposed Adult Populations

www.nature.com/jes

Environmental and biological monitoring of exposure to organophosphorus pesticides: Application to occupationally and non-occupationally exposed adult populations

GHISLAINE BOUVIERa, OLIVIER BLANCHARDb, ISABELLE MOMASa AND NATHALIE SETAa aEnvironmental and Public Health Laboratory, Faculty of Pharmaceutical and Biological Sciences, Rene´ Descartes University, Paris, France bHealth Risks Evaluation Unit, Chronic Health Effects Department, Institut National de l’Environnement Industriel et des Risques (INERIS), Verneuil en Halatte, France

The purpose of the study was to assess non-dietary exposure of workers and the general population in the Paris area to some organophosphorus (OP) pesticides. In total, 21 workers from different occupational places (two greenhouses, three florist shops and three veterinary departments) and 20 subjects assumed to be non-occupationally exposed were recruited. Indoor air, hand wipes, and three first morning urine samples were collected. Seven OPs were measured by GC/ECD and GC/TSD, and six urinary dialkylphosphate metabolites by GC/PFPD. All indoor air samples from the workplaces and only one-third of the samples from the residences contained at least one of the seven OPs. However, almost all participants were dermally exposed to OPs. Total OP indoor air and cutaneous levels were significantly higher for workers than for the general population (air median ¼ 185 pmol/m3 versus nondetectable, Po0.0001; hands median ¼ 1250 pmol/hands versus 475 pmol/hands, P ¼ 0.03). From the air, gardeners and florists were mainly exposed to methyl-OPs and veterinary staff to ethyl-OPs (mainly diazinon). From their hands, all subjects were exposed to methyl-OPs, with gardeners and florists exposed to somewhat but not significantly higher levels. Ethyl-OPs were more found frequently and at higher levels on the hands of veterinary workers. Total OP levels in indoor air and from hand wipes were significantly correlated (Spearman R ¼ 0.34, P ¼ 0.03). DAP detection frequencies and levels were not different between workers and the general population (workers median ¼ 168 nmol/g creat and general population median ¼ 241 nmol/g creat, P ¼ 0.31), and did not correlate with air or hand levels. Subjects not occupationally exposed showed significant residential exposure to OPs, more frequently from their hands than from the air. Different occupations were associated with different exposure profiles and levels. The lack of differences in DAP levels between the different groups of exposure suggests that dietary exposure to OP residues and exposure to other OPs are involved. Journal of Exposure Science and Environmental Epidemiology (2006) 16, 417–426. doi:10.1038/sj.jes.7500473; published online 25 January 2006

Keywords: organophosphorus insecticides, occupational exposure, environmental exposure, urinary metabolites, biological monitoring.

Introduction most acutely toxic pesticides for mammals, the others being highly (methyl-parathion, dichlorvos), moderately (chlorpyr- Organophosphorus (OP) insecticides are widely used for ifos, diazinon, fenthion), or slightly toxic (malathion) agricultural, residential and community health purposes. (ExToxNet, 2002). Regardless, lifetime exposure to low They progressively replaced organochlorine insecticides, doses of OP pesticides raises questions about potential known to be persistent in the environment; 25 different chronic health effects (Eskenazi et al., 1999; Ray and OPs are still used in France (EPHY, 2002). Six of the seven Richards, 2001; Farahat et al., 2003; Salvi et al., 2003). OPs analysed in this study (chlorpyrifos, diazinon, dichlor- Pesticides used or tracked-in indoors tend to have a longer vos, fenthion, malathion, and methyl-parathion) are regis- half-life than those in an outdoor environment, probably tered for use in the control of many insect pests on a wide because there is less water, ultra-violet and biological exposure variety of agricultural crops and in residential settings. Ethyl- and breakdown. This was suggested in studies finding non- parathion had very restricted uses at the time of our study, persistent pesticides, including OPs, in homes many years after and was banned at the end of 2002 in France. It is one of the their application (Lewis et al., 1994; Wright et al., 1994; Dingle et al., 1999). Studies comparing indoor and outdoor residential pesticide contamination showed greater indoor OP levels than outdoor ones (Whitmore et al., 1994; Simcox et al., 1. Address all correspondence to: G Bouvier, De´ partement Sante´ Publique 1995). From this, and considering the time spent indoors at et Environnement, Faculte´ des Sciences Pharmaceutiques et et Biologi- work or at home, we could assume non-dietary exposure to ques, Universite´ Paris 5, 4, avenue de l’Observatoire, 75270 Paris cedex 06, OP pesticides is largely from indoor contamination. France. Tel/Fax: 33 1 43 25 38 76. Besides agricultural workers, greenhouse workers are E-mail: [email protected] Received 31 August 2005; accepted 2 December 2005; published online 25 known to be exposed to OP pesticides they usually spray January 2006 (Aprea et al., 2001; Castro Cano et al., 2001). However, Bouvier et al. Organophosphorus pesticide exposure of adults in France passive exposure of florists and veterinary workers to OPs microenvironmental measurements (indoor air of residences previously used in floriculture and by pet owners has not been or workplaces), personal measurements (cutaneous measure- studied until now. These workers do not wear protection ments on hands), and biological monitoring (urinary DAP gloves for most of their activities. Thus, workers’ exposure metabolites analysis). The details about the nature of dialkyl can occur via contamination of the skin. The same exposure groups, DAP metabolites excreted, and agricultural and pathways could be expected when considering the general domestic uses in France of the seven OPs studied here are population, owing to contact with residues in garden, on recapitulated in Table 1. pets, and on home surfaces. Monitoring indoor air and hand contamination could be an interesting means of assessing non-dietary exposure of Methods subjects. In addition to this monitoring, the measurement of the actual pesticide amount that enters the body is useful, in Study Design order to accurately assess individual exposure. Biological This study took place from February to December 2002. monitoring, that is measuring a pesticide or one of its Indoor air, hand wipe sampling, and interview were metabolites in biological matrices, such as urine, provides a performed during a business day for workers and a day off quantitative measure of the internal dose. Thio or dithio OP for non-occupationally exposed participants. Three first compounds are first bioactivated by desulfuration to oxons in morning urine samples were collected by each subject during the liver. Then, OPs are rapidly hydrolysed and dearylated to a 1-week period. The first urine sample was taken the day form a specific metabolite and one or more dialkylphosphate after the air sampling. For workers, the second sample was metabolite(s). These metabolites are then excreted in urine, taken after another working day and the third after a day off. with biological half-lives from less than 30 min to more than For the general population, the two other samples were taken 24 h, depending on the OPs and the exposure pathways. The on two other days of the week. main studied urinary metabolites common to numerous OP pesticides are the six dialkylphosphates (DAPs) di- Study Population methylphosphate (DMP), dimethylthiophosphate (DMTP), The recruited subjects, aged more than 18 years, were living dimethyldithiophosphate (DMDTP), diethylphosphate (DEP), and working in the Paris area. Workers from two green- diethylthiophosphate (DETP), and diethyldithiophosphate houses of public botanical gardens, three flower shops and (DEDTP). three services of a veterinary school, and a group of non- Therefore, our purpose was to assess OP pesticide occupationally exposed subjects, recruited within the circle of exposure in some residential and occupational situations in acquaintances, were enrolled on a voluntary basis. Each a sample of adults, living in the Paris area. We used pesticide volunteer received oral and written information about the

Ta bl e 1 . Characteristics of the analysed organophosphorus (OP) pesticides and their uses in France.

Type of alkyl Common name of Dialkyl-phosphate Domestic uses in France Agricultural uses in France groups the compound metabolitesa

Methyl-OPs Dichlorvos DMP Insecticidal sprays Aerial and soil treatments for Diffusive devices against mosquitoes numerous crops Insect control of premises Fenthion DMP, DMTP Pet spot-ons against ﬂeas and ticks Aerial treatments for numerous crops Withdrawn since December 2004 Malathion DMP, DMTP, DMDTP Capillar formulations against lice Aerial treatments for numerous crops Plant protection products against cochineals Insect control of premises Methyl parathion DMP, DMTP None Aerial treatments for numerous crops Withdrawn since May 2003 Ethyl-OPs Chlorpyrifos DEP, DETP Ant and cockroach bait traps Aerial and soil treatments for Insecticidal sprays numerous crops Insect control of premises Diazinon DEP, DETP Pet collars against ﬂeas and ticks Soil treatments for numerous crops Ant powders Insecticidal soil treatments Ethyl parathion DEP, DETP None Aerial treatments for numerous crops Withdrawn since October 2002 aDMP ¼ dimethylphosphate; DMTP ¼ dimethylthiophosphate; DMDTP ¼ dimethyldithiophosphate; DEP ¼ diethylphosphate; DETP ¼ diethylthiophosphate; DEDTP ¼ diethyldithiophosphate.

418 Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) Organophosphorus pesticide exposure of adults in France Bouvier et al. study and signed a statement of formal consent before 100, 50, 40, 30, 20, 10, 5, and 1 ng/ml for GC/ECD were participating. prepared by diluting the working solutions in acetone. Samples (PUF þ filter and swabs) were soxhlet extracted Questionnaires with dichloromethane (Fluka, Sigma-Aldrich Chimie) during Each participant was interviewed to identify demographic 16 h, then concentrated to 100 ml in a TurboVap II rotary characteristics and information like smoking habits and type evaporator (Zymark, Roissy, France), diluted in 2 ml of habitation (i.e., apartment or single family home). acetone, and kept frozen (À201C) until analysis. Gardeners detailed recent OP pesticide spraying in green- The GC/TSD was a CP-3800 gas chromatograph, houses and the observed re-entry interval. Nonoccupation- equipped with CP-8200 autosampler, 1079 programmable ally exposed subjects were asked about pesticides used at vaporizing temperature split/splitless injector and thermionic- home or in the garden during the past year. Information specific detector (TSD), all purchased from Varian (Les Ulis, gathered was kept anonymous. France). The EC-5 capillary column 5% phenyl-95% methylpolysiloxane (30 m Â 0.32 mm i.d., 0.25 mmfilm Sampling Methods thickness) was from Alltech (Templemars, France). Helium The indoor air sampling method was based upon the ASTM was used as carrier gas (flow rate: 1 ml minÀ1) and as make- D 4861-00 Standard (ASTM, 2000). We used a MiniPartisol up gas. Injector and detector were held at 3001C; GC column air sampler 2100 (Rupprecht & Patashnik, East Greenbush, was held at 401C for 10 min, then programmed to 2701Cat NY, USA), equipped with a precleaned glass cartridge 41C/min and finally held at 2701Cfor5min.Internal containing a polyurethane foam (PUF) (ref. 226-92, SKC, standard azobenzene was added before injection; splitless Blandford Forum, UK) and a quartz fibre filter (ref. QM-A mode injection was performed for 1 min. 1851, Whatman, Maidstone, UK). The pump (24 h; flow The GC/ECD was a CP-3800 gas chromatograph, rate ¼ 5 l/min) was placed at a height of approximately 1.6 m, equipped with CP-8400 autosampler, two CP-1177 isother- in the main working room or in the living room of the mal split/splitless injectors, two columns and a double residences. electron capture detector (ECD), all purchased from Varian. Both hands were wiped with disposable sterile swabs The capillary columns were CPSiL 8CB (50 m Â 0.25 mm (Mesoft, Mo¨ lnlycke Healthcare, Wambrechies, France) i.d., 0.25 mm film thickness) and CPSiL 24CB wetted with 100% isopropanol (Carlo Erba, Val de Reuil, (60 m Â 0.25 mm i.d., 0.25 mm film thickness) from Chrom- France), according to a standardized protocol. The wipes Pack (Varian). Helium was used as the carrier gas (flow rate: were then placed in a precleaned glass flask (Fisher Scientific 1mlminÀ1) and nitrogen as the make-up gas. Injector and Labosi, Elancourt, France). detector were held at 3001C; GC columns were held at 801C Urine samples were collected and stored by the partici- for 3 min, then programmed to 1501Cat251C/min, then to pants at þ 41C until collection by the investigator. Total 2801Cat31C/min, and finally held at 2801C for 30 min. urine volume collected and pH were measured before Internal standard PCB101 was added before injection; freezing, and urinary creatinine was analysed using the Jaffe splitless mode injection was performed for 1 min. method. Foreachseriesofsamples,analiquotofcontrol,consisting All samples were shipped on ice, blinded, and rapidly of an acetone solution with known concentrations of each OP frozen (À201C) until analysis. and kept frozen at À201C, was analysed. Laboratory blanks wereanalysedwitheachanalysisbatchandshowedno Analytical Methods contamination. Field blanks (15% of total samples) were blinded and analysed as the other samples. When necessary, Indoor air and hand wipe sample analysis OP pesticides sample values have been corrected by the corresponding field were analysed by Gas Chromatography (GC) with electron blank value from the same batch. capture detection (ECD) for chlorpyrifos, or thermionic Peaks were integrated using the Varian Star workstation specific detection (TSD) for diazinon, dichlorvos, fenthion, analysis software. Identifications and quantifications were malathion, ethyl parathion, and methyl parathion, based based on the retention time and the area under curve relative upon the ASTM D 4861-00 Standard (ASTM, 2000). to those of the internal standards azobenzene (GC/TSD) or Analytical standards of the seven OP compounds of PCB101 (GC/ECD). 97.4–99.9% purity were obtained from Riedel-de Hae¨ n Limits of detection (LODs) and limits of quantification (Sigma-Aldrich Chimie, Lyons, France), internal standard (LOQs) were calculated on the basis of a signal three times azobenzene 99% from Sigma-Aldrich Chimie, and Poly- and ten times the background noise, respectively. The LOQs chlorobiphenyle 101 (PCB101) 99.8% from LGC Promo- were 2.2, 5.5, 5.0, 4.5, 4.5, 9.0, and 10.3 ng/ml for chem (Molsheim, France). Stock solutions and working chlorpyrifos, diazinon, dichlorvos, fenthion, malathion, ethyl solutions were prepared in acetone (Carlo Erba). Calibration parathion, and methyl parathion, respectively. The detector standards 1000, 500, 100, 50, and 10 ng/ml for GC/TSD and response was linear (Pearson’s correlation coefficient

Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) 419 Bouvier et al. Organophosphorus pesticide exposure of adults in France

R40.99) for the concentration ranges LOD – 1000 ng/ml for added to the filtrate, which was then evaporated to dryness GC/TSD and LOD – 100 ng/ml for GC/ECD. Average under a nitrogen stream at ambient temperature. The extract recoveries from acetone solutions ranged from 91% to 96% was solubilized in desiccated ACN and evaporated to dryness for all analytes (min.: 83%, max.: 108%). The method was again. K2CO3, desiccated ACN and PFBBr solution were repeatable, the relative standard deviations (RSDs), calcu- then added to the dry extract for the derivatization reaction lated from the replicate analysis of standard solution (n ¼ 6 at 401C overnight. The derivatization mixture was then injections), ranging between 2.6% and 12.7% and repro- brought to ambient temperature and mixed with hexa- ducible, the RSDs from between-day analysis ranging from ne þ osmosed water (1:1). The organic phase was then 5.3% to 17.5%, depending on the compound. filtered with Na2SO4. The aqueous phase was extracted To assess overall recovery of the method, a series of six again, and both organic phases were combined and PUF þ filter cartridges spiked with 10 mg of each compound evaporated to a final volume of 0.5 ml and analysed. in acetone, and two series of seven isopropanol-wetted swabs A Varian Gas Chromatograph CP-3800, equipped with spiked either with 0.1 or 0.8 mg of each pesticide in acetone CP-8400 autosampler, CP-1177 split/splitless injector and were tested. Spiked samples were frozen at À201Candthen Pulsed Flame Photometric Detector (PFPD) was used. The handled as described above. Average recoveries always capillary column SPB-20 (ref. 24088, 30 m Â 0.32 mm i.d. comprised between 73% and 93%, except for dichlorvos Â 0.25 mm film thickness) was purchased from Supelco (55% on PUF þ filter and 26% on swabs). Coefficients of (Sigma-Aldrich Chimie). The detector was held at 2501C, variation assessed during this experiment were between 3.9% and injector at 2251C. The oven was held at 801Cfor1min, and 9.5% for all analytes, except for dichlorvos (25% on then raised to 1701Cat151C/min, held at 1701C for 10 min, PUF þ filter and 20% on swabs). then finally raised to 2701Cat301C/min and held at this Results were expressed as ng/m3 for indoor air samples temperature for 5 min. Helium was used as carrier gas at a and as ng/hands for hand wipes (including contamination of flowrate of 2 ml min À1, and flame gas was a mixture of air both hands). andhydrogen.Injectionwasperformedinthesplitlessmode during 1 min, and a split ratio of 20 after 1 min was applied Urine Analysis The analytical method has been adapted (volume injected: 1 ml). from those of Aprea et al. (1996a) and Hardt and Angerer An aliquot of a control (blank urine spiked with known (2000). concentrations of each DAP) was analysed with each sample Analytical standards (495% purity) DMP, DEP, DMTP, series. DMDTP, and DETP as sodium salts were purchased from Peaks were integrated using the Varian Star software. Applichem (Moleluka Fine Chemicals, La Tour du Pin, DAPs were identified and quantified by using internal France), DEDTP as ammonium salt and internal standard standards DBP for DMP and DEP, and TBP for DMTP, dibutyl-phosphate (DBP) from Accros Organics (Noisy Le DMDTP, DETP, and DEDTP. LOQ calculated from spiked Grand, France), and internal standard tributyl-phosphate blank urine was 10 ng/ml for each DAP. (TBP) from Chem Service (Greyhound Chromatography, The detector response was linear (Pearson’s RZ0.985) for UK). Acetonitrile, diethyl-ether, hexane, and methanol were the concentration ranges LOD–200 ng/ml for the six DAPs. from Carlo Erba. Derivatization reagent 2,3,4,5,6-Penta- Average recoveries from spiked urine samples ranged fluorobenzyl bromide (PFBBr) was obtained from Supelco between 85.7% and 98.7% for all analytes (min.: 66%;

(Sigma-Aldrich Chimie), potassium carbonate (K2CO3)and max.: 124%). The repeatability (within-day RSDs from 13% anhydrous sodium sulphate (Na2SO4)of499% purity were to 17%) and reproducibility (between-day RSDs from 8% to from Acros Organics, and chlorhydric acid (HCl) and 19%) of spiked urine analysis were satisfactory, except for sodium chloride (NaCl) from ProLabo (Briare, France). DMP, whose between-day RSD was 28%. The extraction solvent was a mixture of diethyl Results were expressed as nmol DAP/g creatinine to avoid ether þ desiccated acetonitrile 1:1 (V/V); the derivatization urine dilution bias (Aitio and Kallio, 1999). Statistical solution was made by diluting PFBBr in desiccated analysis was performed with and without samples with acetonitrile to a proportion of 1:2 (V/V). creatinine content o0.3 g/l or 43.0 g/l (n ¼ 3 out of 122 Stock solutions and working solutions were made in urine samples), and results were unchanged. methanol. Calibration standards were made with blank urine spiked with working solutions at a final concentration of 25, Statistical Analysis 50, 100, and 200 ng/ml of each DAP. Owing to the relatively small sample size, we performed non- DBP and TBP were added to 5 ml centrifuged urine, then parametric tests, and the level of statistical significance was mixed with 5 ml extraction solvent þ NaCl þ HCl and ao0.05. centrifuged. The aqueous phase was extracted again; both For indoor air and hand wipes, the molar concentrations organic phases were combined, desiccated by the addition of of methyl-OPs (dichlorvos, fenthion, malathion and methyl- anhydrous Na2SO4 and filtered on a paper filter. K2CO3 was parathion), ethyl-OPs (chlorpyrifos, diazinon, and ethyl-

420 Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) Organophosphorus pesticide exposure of adults in France Bouvier et al. parathion) and total OPs (sum of the seven OPs molar However, only seven of the products used contained diazinon concentrations) were combined. When the value of a given (pet collars or ant powder), one dichlorvos, and another one indoor air or dermal pesticide level was below the LOQ, we malathion. This was not related to the presence of these assigned it a value of half the LOQ. compounds in the air or on the subject’s hands. For each individual, the molar concentrations of urinary Indoor air and hand wipe samples were correctly collected metabolites were combined to generate three value categories: for the 41 subjects; a total of 30 indoor air samples (n ¼ 10 in SMP, sum of dimethylphosphates (DMP þ DMTP þ workplaces, because several participants worked in the same DMDTP), SEP, sum of diethylphosphates (DEP þ DETP þ workplace) and 41 hand wipe samples were taken. The mean DEDTP), and SDAP, sum of the six dialkylphosphates. air volume was 7.170.4 m3/24 h per sample. A total of 122 A value of a given metabolite belowthe LOQ wasassigned urine samples were collected. Indeed, three urine samples a value of zero. Concordance between urinary DAP levels were collected for 40 subjects, but the volume of the third from three urine samples of the same subject was assessed urine sample was insufficient for one florist. using Friedman coefficient of concordance. In the case of Table 2 details mean methyl, ethyl, and total DAP levels in a lack of significant differences, the mean of the three urine samples 1–3, and the means of the three sample values. concentrations for SMP, SEP, and SDAP for each The Friedman test for concordance showed that there were individual were calculated, and were subsequently used for no significant differences between levels of the three samples. statistical analyses. Consequently, in Tables 3 and 4, means from the three urine Correlations between results from the three different media samples for each participant are presented. (air, hand wipes, and urine) were tested using Spearman Table 3 indicates detection frequencies and levels of total correlation coefficient. OPs in indoor air samples and hand wipes, and total DAPs in Frequencies of detection and levels in each media were urine. Only one-third of residential indoor air samples were compared between groups of subjects by using w2 test and contaminated by at least one OP, whereas all workplaces Kruskal–Wallis or Mann–Whitney tests, respectively. When showed airborne contamination. Almost all participants regarding total OP pesticide exposure, subjects were divided showed a dermal exposure to OPs, whether they were into two groups, the group of workers and that of the general occupationally exposed or not. However, workers experi- population. Indeed, there were no significant differences enced significantly higher respiratory and cutaneous exposure between gardeners and florists, on one hand, nor between the to the seven measured OPs than the general population did. plant-related workers and the pet-related workers on the Plant workers were selectively exposed to methyl-OPs, and other hand, for total OP exposure in air, on hands or in urine pet workers to ethyl-OPs in indoor air (see Table 4). Indeed, (Kruskal–Wallis test, data not shown). When comparing the most frequently found OPs in air were malathion and methyl-OP and ethyl-OP exposure, subjects were divided into methyl-parathion, followed by fenthion and dichlorvos, in three groups: plant-related workers, pet-related workers and plant-related places, compared to diazinon, and less fre- the general population. To assess differences between the quently, dichlorvos, in pet-related places. Fenthion and groups, we paired each group and performed multiple dichlorvos were the most frequently found in the air of comparisons after the Kruskal–Wallis test. residences. Median methyl-OP levels on hands were higher for greenhouse workers and florists than for veterinary workers and the general population, but not significantly so. Results Malathion was the most frequently detected methyl-OP on the hands, followed by fenthion on the hands of florists and Five gardeners, seven florists, and nine veterinarians or general population, methyl-parathion on the hands of veterinary workers were recruited. The twenty non-occupa- gardeners, and dichlorvos on the hands of veterinary tionally exposed people recruited were employees from the workers. Dermal exposure of veterinary workers to ethyl- civil service or from business services. They were 63% OPs (mainly diazinon) was significantly more frequent and females, with a mean age of 3479 years. Most of the higher than that of plant workers and the general population. participants (76%) were living in apartments. There were Detectable DAP levels were observed for 93% of urine only six smokers in our subject sample. samples, and methyl metabolites were much more frequently Greenhouse workers reported that OP insecticides sprayed found than ethyl metabolites. DMTP was the most in the previous 3 days were methidathion and methyl frequently detected (85% of urine samples), followed by parathion in the first greenhouse, and malathion in the DMP (83%) and DEP (46%). Medians were below the second. The first greenhouse workers also reported the LOQ, except for DMTP (115 nmol/g creat) and DMP possible use of 16 other OPs in the past year, among which (72 nmol/g creat). As seen in Table 3, no significant were chlorpyrifos, dichlorvos, and ethyl parathion. Regard- differences were observed for urinary SDAP levels between ing general population, 85% reported recent use of at least workers and the general population, although median level one insecticide at home and less often in the garden. was higher in the general population than in the workers’

Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) 421 Bouvier et al. Organophosphorus pesticide exposure of adults in France

Ta bl e 2 . Urinary dialkylphosphate metabolites: comparison of con- Table 3 . Total indoor air and cutaneous organophosphate levels and centrations (nmol/g creatinine) from three ﬁrst morning voids collected total urinary dialkylphosphate metabolite concentrations in workers on a 1-week period. and the general population.

Urine (N)a SMPb SEPc SDAPd Workers group General population P-valuea (n ¼ 21) (n ¼ 20) Urine 1 (41) % 4LOQ 78 29 78 Total OPs b Median 139 NDe 139 Indoor air (pmol/m3) Range ND – 1855 ND – 271 ND – 1922 % 4LOQ 100 35 Median 185 ND o0.0001 Urine 2 (41) Range 13 – 602 ND – 71 % 4LOQ 80 19.5 80.5 Median 185 ND 193 Hand wipes (pmol/2 hands) Range ND – 1266 ND – 347 ND – 1266 % 4LOQ 95 95 Median 1250 475 0.034 Urine 3 (40) Range ND – 10 662 ND – 7361 % 4 LOQ 66 19.5 66 Median 108 ND 109 Total DAPs c Range ND – 1645 ND – 342 ND – 1680 Urine (nmol/g creat) % 4LOQ 90 95 Friedman statistics Median 168 241 0.309 Coefficient of concordance 1.51 0.82 1.32 Range ND – 788 ND – 1525 P-value 0.47 0.66 0.52 aMann–Whitney U-test. Kendall W coefficient 0.018 0.010 0.016 bTotal OPs ¼ sum of the molar concentrations of the seven organophosphate pesticides dichlorvos, fenthion, malathion, methyl-parathion, Mean levelsf chlorpyrifos, diazinon, and ethyl-parathion. % 4 LOQ 93 49 93 c Total DAPs ¼ mean of the three urine sample values for the sum of the six Median 216 ND 226 dialkylphosphates DMP, DMTP, DMDTP, DEP, DETP, and DEDTP Range ND – 1525 ND – 123 ND – 1525 molar concentrations. aN ¼ number of valid samples. bSMP ¼ sum of DMP, DMTP and DMDTP molar concentrations. cSEP ¼ sum of DEP, DETP and DEDTP molar concentrations. morning void urine samples for the same individual. This dSDAP ¼ sum of the six dialkylphosphate molar concentrations. suggests that OP exposure of workers as well as non- eND ¼ non detectable. f occupationally exposed subjects was quite stable over this Mean of molar concentrations from urine samples 1–3 for each subject. period. Meeker et al. (2005) recently demonstrated that a single measure of urinary trichloropyridinol and 1-naphtol group. Frequencies of detection for methyl and ethyl DAPs adequately represented average exposure of men over a three were similar between the three groups; this is also true when month period. In addition, Kissel et al. (2005) showed that comparing SMP and SEP levels (Table 4). However, the first morning void samples were the best predictors of highest levels of SMP and SEP (1525 nmol/g creat and estimated 24 h DETP and DMTP metabolite concentrations 123 nmol/g creat, respectively) were observed in the general in the urine of preschool children. Thus, this sample timing population. seems to be the most reliable for the evaluation of continuous A significant correlation was observed between ethyl-OP exposure to OPs through urinary DAP metabolite measure- levels in air and ethyl-OP levels on hand wipes ments. (RSpearman ¼ 0.50, Po0.0001), and total OP levels in air In the Non-Occupational Pesticide Exposure Study and total OP levels on hand wipes (RSpearman ¼ 0.34, (NOPES), Whitmore et al. (1994) showed a strong correla- P ¼ 0.03) for the 41 subjects. There was no significant tion between US residential indoor air and personal air levels correlation between methyl-OP levels in air and on hand for chlorpyrifos, diazinon, and dichlorvos. They also wipes, nor between urinary methyl, ethyl, and total DAP reported outdoor air levels of these OPs that were much concentrations, and air or cutaneous methyl, ethyl, and total lower than indoor air levels. Similarly, residential indoor air OP levels. OP pesticide levels were higher than outdoor air levels in the We found no significant differences in urinary DAP levels study from Mukerjee et al. (1997) in Texas. Thus, we can according to gender, age, smoking habits, and housing type. assume that indoor air sampling is representative of the respiratory exposure of our subjects to OP pesticides. Discussion Except for dichlorvos, the semivolatile OP pesticides studied here are more likely to deposit on surfaces and soils, We showed that, over a 1-week period, there were no leading to a greater potential for dermal than respiratory significant differences between DAP levels in three first exposure. However, the extent of percutaneous absorption of

422 Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) Organophosphorus pesticide exposure of adults in France Bouvier et al.

Ta bl e 4 . Methyl and ethyl organophosphate levels in indoor air and frequency between veterinary staff (after each pet ausculta- on hands, and urinary methyl and ethyl dialkylphosphate metabolite tion), gardeners, and florists (only before lunch and at the concentrations: comparisons between the different exposure groups. work shift), and the general population may have also influenced the dose effectively absorbed. Some studies a Plant working Pet working General P-values showed that for pesticide sprayers, legs and forearms were group group population (n ¼ 12 (n ¼ 9 (n ¼ 20 the most exposed parts of the body (Castro Cano et al., 2001; workers) workers) subjects) Frenich et al., 2002; Martinez Vidal et al., 2002), but in our sample, evaluation of hand contamination was more relevant 3 Indoor air (pmol/m ) because, for gardeners, we took measurements after re-entry Methyl-OPsb % 4LOQ 100 33 33 0.0001 and not during spraying tasks. No consensus has been Median 170 ND ND reached on howto measure hand contamination. Wipes or Range 9 – 598 ND – 137 ND – 68 washes with ethanol, 2-propanol, mixtures of water and 2- propanol have been used, but the reliability and the c Ethyl-OPs reproducibility of the different methods have been assessed % 4LOQ 0 100 5 o0.0001 Median ND 106 ND for some but not all the existing OPs (Fenske et al., 1991, Range ND 51 – 176 ND – 43 1999; Fenske and Lu, 1994; Geno et al., 1996). Thus, we chose a convenient and less-solvent consuming method, with Hand wipes (pmol/2 hands) only one investigator using the same wiping procedure for all Methyl-OPs participants. % 4LOQ 100 78 95 0.11 Median 920 398 357 In a study assessing fenitrothion exposure of greenhouse Range 157 – 10 476 ND – 3675 ND – 6889 workers, levels ranged between 800 and 17 400 pmol on the hands and 9700–81 200 pmol/m3 in personal air (Aprea Ethyl-OPs et al., 1999). In another study, assessing exposure to three % 4LOQ 42 89 30 0.0042 OPs, they found levels in personal air of 400–17 400 pmol/m3, Median ND 475 ND Range ND – 301 ND – 6954 ND – 3084 depending on the OP (Aprea et al., 2001). These levels were higher than in our study, but this could be explained by Urinary DAPsd (nmol/g creat) different re-entry intervals, differences in doses applied, and SMPe spraying methods employed. % 4LOQ 100 78 95 0.53 We also explored exposure of workers not directly using Median 204 153 221 Range 23 – 582 ND – 766 ND –1525 pesticides, such as florists and veterinary workers. Florists were mainly exposed to methyl-OPs widely used in France SEPf for plant and flower treatments against various pests, and % 4LOQ 58 33 50 0.55 veterinary workers to diazinon, widely used in pet collars in Median 15 ND 7 France. We did not find any published study assessing Range ND – 107 ND – 116 ND – 123 exposure of these worker categories to OP pesticides; indeed, a Kruskal–Wallis one-way rank analysis of variance. owing to the great number of sources, this exposure is bMethyl-OPs ¼ sum of the molar concentrations of dichlorvos, fenthion, difficult to predict. malathion, and methyl-parathion. cEthyl-OPs ¼ sum of the molar concentrations of chlorpyrifos, diazinon, Regarding general population exposure, one study evalu- and ethyl-parathion. ating exposure of children living in an agricultural commu- d DAPs ¼ dialkylphosphates. nity expressed their results as sums of OP molar levels on the eSMP ¼ sum of the molar concentrations of dimethylphosphate, dimethylthiophosphate, and dimethyldithiophosphate (mean of the three hands as in our study (Shalat et al., 2003). Despite the fact urine sample values). that the population studied was different, levels were fSEP ¼ sum of the molar concentrations of diethylphosphate, diethylthio- comparable to ours in urban adults, with a mean7SD ¼ phosphate, and diethyldithiophosphate (mean of the three urine sample 143072860 pmol/hands and median ¼ 300 pmol/hands values). (molar sum of 12 OP levels). We found that nearly all subjects were dermally exposed to OP insecticides, whereas chemicals such as OPs is still under debate because of the lack only half of the children’s hands presented OP contamination of experimental data (Griffin et al., 1999; Sartorelli, 2002; in the study of Shalat et al. (2003). Semple, 2004). In addition to physico-chemical character- Chlorpyrifos was detected very rarely in residences from istics, the formulation, the dose applied, the site and the our study, compared to its ubiquity in US studies (Roinestad duration of application, the film thickness on the skin surface, et al., 1993; Lewis et al., 1994; Whitmore et al., 1994; and other characteristics such as perspiration, temperature, Gordon et al., 1999; Buck et al., 2001; Moschandreas et al., and type of activity can play a role on the absorbed dose 2001; Whyatt et al., 2003). Also, given the wide differences in (Semple, 2004). In our study, differences of hand washing agricultural crops, pesticide licences and population habits

Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) 423 Bouvier et al. Organophosphorus pesticide exposure of adults in France between countries, it seems hazardous to generalize results contribution of environmental DAP exposure to the urinary from residential exposure assessment studies performed in excretion of DAPs remains unclear. one country to other countries. It would be preferable to Several limitations of our study should be mentioned. perform pesticide exposure assessment studies in each First, for dichlorvos, the lowreliability of the analytical country, to assess frequency of detection and levels of method prevented us from adequately assessing exposure to exposure, and to determine which compounds cause concern this OP, which was probably underestimated. Second, we and deserve further investigation, as is the case for diazinon, used flame photometric detection to analyse urinary meta- dichlorvos, fenthion, and malathion in our general popula- bolites, which is less specific compared to mass spectrometry, tion sample. used in the NHANES study for instance, and preventing us DAP metabolites are biomarkers of interest because they from accurately comparing our results with those of studies allowresearchers to assess global exposure to a great number applying this detection method. Third, the non-dietary of OP pesticides. However, as other urinary metabolites, they exposure to OPs of the participants is not exhaustive, reflect aggregate exposure by multiple ways (i.e., oral, because of the great number of other OP compounds on respiratory, and dermal), and from multiple sources. Our the market. Last, our small sample size does not allowus to study, the first performed in France, found DAP concentra- generalize our results to the general population and workers tions for general population comparable to those reported by living in France. Cocker et al. (2002) in UK (mean ¼ 194 nmol/g creat in 463 In conclusion, we show that both the general population adults), and Aprea et al. (1996b) in Italy (DMTP þ DMP, and workers are exposed to OP pesticides, through geometric mean ¼ 126.9 nmol/g creat in 124 adults). Detec- respiratory and dermal pathways. Indoor air exposure levels tion frequency was greater for methyl-DAPs and lower for to OP pesticides was generally not elevated in workers and in ethyl-DAPs than those found in the CDC Second National the general population, whereas dermal exposure was Report (CDC, 2003; Barr et al., 2004). Median DMTP and frequently found and at higher levels for both groups. DEP levels were much higher in our study, but our sample Despite this difference, levels of contamination from the air size was small and not representative of the country and skin were of the same order of magnitude. Air and hand population, compared to the CDC sample. Creatinine exposure assessment provided complementary results, and adjustment has been discussed by numerous authors; lack of correlation for methyl-OPs suggested that exposure however, for adults, it is widely used, to take dilution due to contact with treated pets, plants, fruits, and vegetables into account when spot samples are collected (Mage could result in hand but not indoor air contamination, owing et al., 2004). to the lowvolatility of numerous OPs. Similar urinary DAP We found that, despite greater respiratory and dermal levels between workers and the general population could be exposure, the workers we studied did not excrete more explained by exposure to other OP pesticides and dietary urinary DAPs than the non-occupationally exposed subjects. exposure. Relatively lowexposure of the workersrecruited, Similarly, Aprea et al. (1999, 2001) did not find any owing to low doses used and sufficient re-entry interval for significant differences in DAP levels between greenhouse gardeners and frequent aeration and hand washing for all workers and the general population in two studies. The lack workers, may also explain the similarity in DAP excretion of correlation we observed between urinary DAP and relative to the general population. environmental OP levels could also indicate that dietary exposure plays an important role in the complete human exposure to OP insecticides, although results from previous Acknowledgements studies were inconclusive (MacIntosh et al., 2001; Kieszak et al., 2002; Clayton et al., 2003). Moreover, a given human We thank all study participants for their cooperation, and urinary pesticide metabolite can also be an environmental Robin Aujay, Laure Malleret, Jerome Beaumont, Sebastien degradate. Indeed, recent studies found DAPs in fruits and Fable, Didier Granier, Halina Koszdras, Fabrice Marlie` re, juices (Lu et al., 2005), and specific metabolite of chlorpyr- and Marie-Pierre Strub (INERIS) for sampling advice and ifos in environmental samples (Wilson et al., 2003). laboratory support. However, these polar metabolites would probably not be Ghislaine Bouvier received a doctoral grant from the well absorbed in the gastro-intestinal tract and could also be Agence de l’Environnement et de la Maıˆ trise de l’Energie metabolized to compounds such as mono-alkylphosphates (ADEME) and the Institut National de l’Environnement and organic and inorganic phosphates. To date, only one Industriel et des Risques (INERIS). Japanese study reported that after oral administration of DAPs to rats, a small percent of the dose was excreted unchanged in urine, whereas inorganic and organic phos- References phates were excreted at amounts of 30% to more than 60% Aitio A. and Kallio A. Exposure and effect monitoring: a critical appraisal of their of the DAP dose (Imaizumi et al., 1993). Thus, the practical application. Toxicol Lett 1999: 108: 137–147.

424 Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) Organophosphorus pesticide exposure of adults in France Bouvier et al.

Aprea C., Sciarra G. and Lunghini L. Analytical method for the determination of ments in the national human exposure assessment survey (NHEXAS) pilot urinary alkylphosphates in subjects occupationally exposed to organopho- study in Arizona: preliminary results for pesticides and VOCs. J Expo Anal sphorus pesticides and in the general population. J Anal Toxicol 1996a: 20: Environ Epidemiol 1999: 9: 456–470. 559–563. Griffin P., Mason H., Heywood K. and Cocker J. Oral and dermal absorption of Aprea C., Sciarra G., Lunghini L., Centi L. and Ceccarelli F. Evaluation of chlorpyrifos: a human volunteer study. Occup Environ Med 1999: 56: 10–13. respiratory and cutaneous doses and urinary excretion of alkylphosphates by Hardt J., Angerer J. Determination of dialkyl phosphates in human urine using gas workers in greenhouses treated with omethoate, fenitrothion, and tolclofos- chromatography-mass spectrometry. J Anal Toxicol 2000: 24: 678–684. methyl. Am Ind Hyg Assoc J 2001: 62: 87–95. Imaizumi H., Nagamatsu K., Hasegawa A., Ohno Y. and Takanaka A. Aprea C., Sciarra G., Orsi D., Boccalon P., Sartorelli P. and Sartorelli E. Urinary Metabolism of toxicity of acid phosphate esters, metabolites of organopho- excretion of alkylphosphates in the general population (Italy). Sci Total sphorus insecticides, in rat [in Japanese]. Jpn J Toxicol Environ Health 1993: Environ 1996b: 177: 37–41. 39: 566–571. Aprea C., Sciarra G., Sartorelli P., Ceccarelli F. and Centi L. Multiroute exposure Kieszak S.M., Naeher L.P., Rubin C.S., Needham L.L., Backer L., Barr D. and assessment and excretion of urinary metabolites of fenitrothion during manual McGeehin M. Investigation of the relation between self-reported food operations on treated ornamental plants in greenhouses. Arch Environ Contam consumption and household chemical exposures with urinary levels of selected Toxicol 1999: 36: 490–497. nonpersistent pesticides. J Expo Anal Environ Epidemiol 2002: 12: 404–408. ASTM. D 4861-00. Standard Practice for Sampling and Selection of Analytical Kissel J.C., Curl C.L., Kedan G., Lu C., Griffith W., Barr D.B., Needham L.L. Techniques for Pesticides and Polychlorinated Biphenyls in Air. http:// and Fenske R.A. Comparison of organophosphorus pesticide metabolite levels www.astm.org, 2000. in single and multiple daily urine samples collected from preschool children in Barr D.B., Bravo R., Weerasekera G., Caltabiano L.M., Whitehead Jr R.D., Washington State. J Expo Anal Environ Epidemiol 2005: 15: 164–171. Olsson A.O., Caudill S.P., Schober S.E., Pirkle J.L., Sampson E.J. and Lewis R.G., Fortmann R.C. and Camann D.E. Evaluation of methods for Jackson R.J., Needham L.L. Concentrations of dialkyl phosphate metabolites monitoring the potential exposure of small children to pesticides in the of organophosphorus pesticides in the US population. Environ Health Perspect residential environment. Arch Environ Contam Toxicol 1994: 26: 37–46. 2004: 112: 186–200. Lu C., Bravo R., Caltabiano L.M., Irish R.M., Weerasekera G., Barr D.B. The Buck R.J., Ozkaynak H., Xue J., Zartarian V.G. and Hammerstrom K. Modeled presence of dialkylphosphates in fresh fruit juices: implication for organopho- estimates of chlorpyrifos exposure and dose for the Minnesota and Arizona sphorus pesticide exposure and risk assessments. J Toxicol Environ Health A NHEXAS populations. J Expo Anal Environ Epidemiol 2001: 11: 253–268. 2005: 68: 209–227. Castro Cano M.L., Martinez Vidal J.L., Egea Gonzalez F.J., Martinez Galera M. MacIntosh D.L., Kabiru C.W. and Ryan P.B. Longitudinal investigation of Gas chromatographic method for assessing the dermal exposure of greenhouse dietary exposure to selected pesticides. Environ Health Perspect 2001: 109: applicators to dimethoate and malathion. J Chromatogr Sci 2001: 39: 145–150. 345–350. Mage D.T., Allen R.H., Gondy G., Smith W., Barr D.B. and Needham L.L. CDC. Second national report on human exposure to environmental chemicals, Estimating pesticide dose from urinary pesticide concentration data by NCEH Publication No 02-0716. National Center for Environmentral creatinine correction in the Third National Health and Nutrition Exami- Health, Centers for Disease Control and Prevention. http://www.cdc.gov/ nation Survey (NHANES-III). J Expo Anal Environ Epidemiol 2004: 14: exposurereport/3rd/pdf/thirdreport.pdf, 2003. 457–465. Clayton C.A., Pellizzari E.D., Whitmore R.W., Quackenboss J.J., Adgate J. and Martinez Vidal J.L., Egea Gonzalez F.J. and Garrido Frenich A., Martinez Sexton K. Distributions, associations, and partial aggregate exposure of Galera M., Aguilera P.A., Lopez Carrique E. Assessment of relevant factors pesticides and polynuclear aromatic hydrocarbons in the Minnesota Children’s and relationships concerning human dermal exposure to pesticides in green- Pesticide Exposure Study (MNCPES). J Expo Anal Environ Epidemiol 2003: house applications. Pest Manag Sci 2002: 58: 784–790. 13: 100–111. Meeker J.D., Barr D.B., Ryan L., Herrick R.F., Bennett D.H., Bravo R. and Cocker J., Mason H.J., Garfitt S.J., Jones K. Biological monitoring of exposure to Hauser R. Temporal variability of urinary levels of nonpersistent insecticides in organophosphate pesticides. Toxicol Lett 2002: 134: 97–103. adult men. J Expo Anal Environ Epidemiol 2005: 15: 271–281. Dingle P., Williams D., Runciman N., Tapsell P. Pesticides in homes in Western Moschandreas D.J., Karuchit S., Kim Y., Ari H., Lebowitz MD MK O.R., Australia. Bull Environ Contam Toxicol 1999: 62: 309–314. Gordon S. and Robertson G. On predicting multi-route and multimedia EPHY. Catalogue officiel français des produits phytopharmaceutiques et de leurs residential exposure to chlorpyrifos and diazinon. J Expo Anal Environ usages. http://e-phy.agriculture.gouv.fr/, 2002. Epidemiol 2001: 11: 56–65. Eskenazi B., Bradman A., Castorina R. Exposures of children to organophosphate Mukerjee S., Ellenson W.D., Lewis R.G., Stevens R.K., Somerville M.C. and pesticides and their potential adverse health effects. Environ Health Perspect Shadwick D.S. An environmental scoping study in the lower Rio Grande 1999: 107(Suppl 3): 409–419. Valley of Texas – III. Residential microenvironmental monitoring for air, ExToxNet. Pesticide Information Profiles. ExToxNet database http://ace.orst.edu/ house dust, and soil. Environ Int 1997: 23: 657–673. info/extoxnet, 2002. Ray D.E., Richards P.G. The potential for toxic effects of chronic, low-dose Farahat T.M., Abdelrasoul G.M., Amr M.M., Shebl M.M., Farahat F.M. and exposure to organophosphates. Toxicol Lett 2001: 120: 343–351. Anger W.K. Neurobehavioural effects among workers occupationally exposed Roinestad K.S., Louis J.B. and Rosen J.D. Determination of pesticides in indoor to organophosphorous pesticides. Occup Environ Med 2003: 60: 279–286. air and dust. J AOAC Int 1993: 76: 1121–1126. Fenske R.A., Curry P.B., Wandelmaier F. and Ritter L. Development of dermal Salvi R.M., Lara D.R., Ghisolfi E.S., Portela L.V., Dias R.D. and Souza D.O. and respiratory sampling procedures for human exposure to pesticides in Neuropsychiatric evaluation in subjects chronically exposed to organopho- indoor environments. J Expo Anal Environ Epidemiol 1991: 1: 11–30. sphate pesticides. Toxicol Sci 2003: 72: 267–271. Fenske R.A., Lu C. Determination of handwash removal efficiency: incomplete Sartorelli P. Dermal exposure assessment in occupational medicine. Occup Med removal of the pesticide chlorpyrifos from skin by standard handwash (London) 2002: 52: 151–156. techniques. Am Ind Hyg Assoc J 1994: 55: 425–432. Semple S. Dermal exposure to chemicals in the workplace: just how important is Fenske R.A., Simcox N.J., Camp J.E. and Hines C.J. Comparison of three skin absorption? Occup Environ Med 2004: 61: 376–382. methods for assessment of hand exposure to azinphos-methyl (guthion) during Shalat S.L., Donnelly K.C., Freeman N.C., Calvin J.A., Ramesh S., Jimenez M., apple thinning. Appl Occup Environ Hyg 1999: 14: 618–623. Black K., Coutinho C., Needham L.L., Barr D.B. and Ramirez J. Nondietary Frenich A.G., Aguilera P.A., Gonzalez F.E., Castro Cano M.L., Martinez Galera ingestionofpesticidesbychildreninanagriculturalcommunityontheUS/ M., Martinez Vidal J.L. and Soler M. Dermal exposure to pesticides in Mexico border: preliminary results. J Expo Anal Environ Epidemiol 2003: 13: greenhouses workers: discrimination and selection of variables for the design of 42–50. monitoring programs. Environ Monit Assess 2002: 80: 51–63. Simcox N.J., Fenske R.A., Wolz S.A., Lee I.C. and Kalman D.A. Pesticides in Geno P.W., Camann D.E., Harding H.J., Villalobos K. and Lewis R.G. household dust and soil: exposure pathways for children of agricultural Handwipe sampling and analysis procedure for the measurement of dermal families. Environ Health Perspect 1995: 103: 1126–1134. contact with pesticides. Arch Environ Contam Toxicol 1996: 30: 132–138. Whitmore R.W., Immerman F.W., Camann D.E., Bond A.E., Lewis R.G. and Gordon S.M., Callahan P.J., Nishioka M.G., Brinkman M.C., O’Rourke M.K. Schaum J.L. Non-occupational exposures to pesticides for residents of two US and Lebowitz M.D., Moschandreas D.J. Residential environmental measure- cities. Arch Environ Contam Toxicol 1994: 26: 47–59.

Journal of Exposure Science and Environmental Epidemiology (2006) 16(5) 425 Bouvier et al. Organophosphorus pesticide exposure of adults in France

Whyatt R.M., Barr D.B., Camann D.E., Kinney P.L., Barr J.R., Andrews H.F., Wilson N.K., Chuang J.C., Lyu C., Menton R. and Morgan M.K. Aggregate Hoepner L.A., Garﬁnkel R., Hazi Y., Reyes A., Ramirez J., Cosme Y. exposures of nine preschool children to persistent organic pollutants at day care and Perera F.P. Contemporary-use pesticides in personal air and at home. J Expo Anal Environ Epidemiol 2003: 13: 187–202. samples during pregnancy and blood samples at delivery among urban Wright C.G., Leidy R.B. and Dupree Jr H.E. Chlorpyrifos in the air and soil of minority mothers and newborns. Environ Health Perspect 2003: 111: houses eight years after its application for termite control. Bull Environ Contam 749–756. Toxicol 1994: 52: 131–134.

426 Journal of Exposure Science and Environmental Epidemiology (2006) 16(5)