Bovine serum albumin (BSA), fetal bovine serum, goat anti-rabbit immunoglobulin G (whole molecule) peroxidase conjugate (GAR-HRP), H₂O₂, Tween 20, dimethyl sulfoxide, 3,3′,5,5′-tetramethylbenzidine (TMB), N,N-dimethyl-formamide (DMF), N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) were obtained from Sigma-Aldrich (St. Louis, MO, USA). All other chemicals were reagent grade or better from Fisher Scientific (Pittsburgh, PA, USA) or Merck (Darmstadt, Germany). The samples were extracted using Strata Screen-A mixed mode solid phase extraction columns (8B-S019-EAK, Phenomenex, Torrance, CA, USA). The ELISA method was performed on Nunc Maxisorp Immunoplates (96F, 442404, Roskilde, Denmark) and read spectrophotometrically with a microplate reader (Tecan Sunrise, Männedorf, Switzerland). A creatinine kit (Sysmex Corporation, Kobe, Japan) was used to measure the concentration of creatinine in urine according to the manufacturer’s instructions.

The present study was a cross-sectional study designed to assess pyrethroid exposure among consumers and farmers. Farmer volunteers were assumed to be occupationally exposed. The Fang district of Chiang Mai province, one of the most intensive agricultural areas in northern Thailand, is composed of eight sub-districts. The four sub-districts with the most intensive agriculture were selected as study sites. These were Mae Kha, Mae Ngon, Wiang, and Mae Ka sub-districts. These four sub-districts contained 56% (63,879) of the total Fang district population (112,402).^[28] The subjects included 100 consumers (50 females, 50 males) and 100 farmers (50 females, 50 males). Farmers and consumers in these selected sub-districts were sampled using a random number of their household numbers in order to provide the same probability of being selected into the study. Numbers of farmers and consumers participated in the study described by sub-district is shown in Table 1. This study was approved by the Human Experimentation Committee, Research Institute for Health Sciences (RIHES), Chiang Mai University (No. 32/2006).

A consent form was introduced and signed by individual participants prior to collection of personal data, blood, and urine samples. Blood and urine samples were taken in the morning of the same day. Approximately, 10 mL of venous blood samples in vacutainer heparinized tubes and 50 mL of midstream morning void urine samples in 250 mL polypropylene bottles were collected and placed in the ice chest. The samples were then transported to Toxicology Laboratory, Research Institute for Health Sciences, Chiang Mai University (about 150 km away). On the next day, blood samples were centrifuged at 2500 rpm for 15 min at 4°C to collect plasma. Plasma and urine samples were aliqouted into 1 mL and 10 mL, respectively, and then frozen (−20°C) until analysis. On the day of analysis, samples were thawed at room temperature, gently mixed to homogeneity and then sampled for analysis. The analyses of all samples were finished within 1 year.

The plasma 3-PBA ^[27] and urinary 3-PBA analyses ^[15,29] were modified from previous studies. To prepare plasma samples, briefly, 100 μL sodium hydroxide (6 N) was added to the plasma (0.5 mL) and placed in a heating block at 100°C for 1 hr to hydrolyze the 3-PBA adduct. After cooling, 1 mL of 0.2 M sodium acetate buffer, pH 4.5 was added to adjust the pH to around 12. For further cleanup, liquid-liquid extraction (LLE) was employed by adding 2 mL of ethyl acetate and the sample shaken vigorously for 10 min. The ethyl acetate was discarded. The ethyl acetate removes organic contaminants while the 3-PBA remains in the aqueous phase. To the remaining aqueous phase was added 120 μL of hydrochloric acid (6 N) to adjust the pH to around 3. This was extracted twice with 2 mL of ethyl acetate. The organic phases were evaporated to dryness under a gentle stream of nitrogen. The residue was dissolved in 200 μL of methanol. Two milliliters of 0.2 M sodium acetate buffer was added to adjust the pH to around 5 and mixed thoroughly. The sample was then ready for solid phase extraction (SPE). The SPE cartridge used was the Strata Screen-A (55 μm, 70 Å), 100 mg/mL containing a mixed mode silica-based sorbent consisting of mixed-mode SPE hydrophobic (C8) and ionic (anion exchange) phases. The SPE method was described previously.^[15]

For measurement of 3-PBA in urine, we used a common acid hydrolysis method to release any 3-PBA that may have been in the conjugated form such as glucuronides, sulfates or amino acids. A 0.5 mL aliquot of urine was added into a glass vial. The urine sample was treated with 100 μL of 6 N hydrochloric acid and heated at 100°C for 1 hr. The hydrolyzed urine was neutralized by adding 1 mL of 0.2 M sodium acetate buffer (pH 4.5) followed by 100 μL of sodium hydroxide (6 N) and mixing thoroughly to adjust the pH to around 4.5. To reduce the matrix effect of the hydrolyzed urine, the same SPE method as for the plasma 3-PBA was used.

The specific antibody and hapten for the target analyte (3-PBA) and the preparation of coating antigen, the buffers and the procedure for the indirect competitive ELISA was previously described.^[29] Briefly, 96-well plates were coated overnight at 4°C with 100 μL of coating antigen, 3-PBA-BSA, (0.5 μg/mL). The following day, plates were blocked for 30 min with 200 μL 0.5% BSA in PBS (phosphate buffered saline) at room temperature. After 30 min the blocking solution was removed. Then 50 μL of extracted samples (plasma or urine) and 50 μL of antiserum (1:7000) were added to each well and mixed for 1 hr at room temperature. The plates were washed five times. The goat anti-rabbit IgG-horseradish peroxidase conjugate diluted 1/10000 (100 μL) was added for 1 hr. Following a wash, the TMB substrate (100 μL) was added for 15 min then the reaction stopped and absorbance measured at 450 nm. The concentrations of unknown samples were calculated from the calibration curve using a 4-parameter fit equation and expressed as 3-PBA equivalents. Urinary 3-PBA concentrations were normalized to creatinine levels to correct for dilution of urine and is often used to normalize when spot samples are taken rather than 24 hr samples.^[30] Although plasma 3-PBA may be referred to as the 3-PBA-adduct, the concentrations reported here are a measure of the total of bound and free 3-PBA in the plasma.

Previous work from this laboratory described suitable conditions for the ELISA analysis of plasma 3-PBA.^[27] An IC₅₀ value of 26.7 ng/mL and reproducibility (%CV < 15%) was obtained. The IC₅₀ value is expressed as the sensitivity of the immunoassay and approximates the concentration of analyte giving 50% inhibition. The recovery of spiked free 3-PBA in plasma was 85.9–99.4%. More than 80% 3-PBA was recovered from the adducted form (3-PBA-BSA) when subjected to hydrolysis. The limit of quantitation (LOQ) was 5 ng/mL, 30- to 40-fold more sensitive than a previous study (150–200 ng/mL) that used HPLC analysis.

For laboratory quality assurance and control, precision measurements were conducted prior to the analysis of real samples. Plasma and urine controls (10 tubes) were analyzed for day-to-day variation of samples to obtain an optimum condition variance (OCV) range (mean ± 2SD). Each plasma analysis batch consisted of 15 samples, 1 sample blank and 2 controls using pooled plasma that were placed in the first and last tubes. For urine, 25 samples were analyzed in each batch with 3 pooled urine controls placed in the first, middle and last tubes. The controls in each batch were analyzed for intra-batch variation of samples to obtain within-day CV (%). Inter-batch variation of samples was analyzed from controls in all batches analyzed (n = 14 and 9, respectively) to obtain between-day CV (%).

Data from laboratory results were analyzed by SPSS Statistical Software Package (SPSS, Thailand) version 17 as follows; (1) the concentrations of 3-PBA were tested for normal distribution (parametric or nonparametric), (2) % detection, median and range of 3-PBA concentrations were computed, (3) the plasma 3-PBA and urinary 3-PBA concentrations were compared between consumers and farmers using the paired sample t-test, (4) the correlation between plasma 3-PBA and urinary 3-PBA concentrations was analyzed using the Spearman test.

The optimal conditions for the ELISA analysis of plasma 3-PBA was previously developed in this laboratory and described above.^[27] The assay for 3-PBA in urine was also previously reported. The IC₅₀ value in that study was reported at 1.65 ± 0.7 ng/mL.^[29] In our hands, the optimized assay using antiserum 294 (diluted 1:7000, final dilution in well) and a coating antigen concentration of 0.5 μg/mL gave an IC₅₀ value of 15.3 ± 0.77 ng/mL (Fig. 1). Although the IC₅₀ was different, it was highly reproducible, so no further optimization was done. It is not unusual for there to be systematic differences in IC₅₀ between laboratories. Minor differences between operations, preparation of calibrators, or reagent degradation are among some of the causes of such variations.^[31]

Most immunoassays exhibit some cross reactivity to structurally related compounds. The assay used here does not cross react with a variety of parent pyrethroids (permethrin, cypermethrin, esfenvalerate, deltamethrin and cyfluthrin) that contain the phenoxybenzyl moiety. However, 4-fluoro-3-phenoxybenzoic acid, a metabolite of cyfluthrin does cross react at 72% ^[29] while 4-hydroxy-3-phenoxybenzoic acid and 3-phenoxybenzaldehyde cross react at 103 and 75%, respectively. Using the same methods as reported here, urine samples collected from forestry workers were analyzed by both LC/MS/MS and immunoassay for 3-PBA. The methods correlated well with slope, intercept and r² of 1.14, 0.24 and 0.964, respectively.^[15] Thus, these other cross-reacting metabolites likely did not greatly impact the determination of urinary 3-PBA for that study population and were not considered significant in the study reported here.

The 3-PBA immunoassay has been assessed for matrix effects from plasma. Recovery of a laboratory-generated 3-PBA-protein adduct (the 3-PBA was adducted to bovine serum albumin) spiked into fetal bovine serum were greater than 80%. In addition, we found that an additional LLE cleanup was necessary before SPE extraction to eliminate the matrix effect of plasma samples. This resulted in high recovery values (85.9–99.4%) with a limit of quantitation (LOQ) at 5 ng/mL.^[27] In this study, for urinary 3-PBA, SPE extraction was necessary to determine 3-PBA metabolites. The SPE cleanup technique gave high recovery values (87.3–98.0%) with limit of quantitation at 2.5 ng/mL (Table 2), comparable to values obtained earlier.^[15]

The limit of blank, limit of detection and limit of quantitation as defined by Armbruster and Pry^[32] are shown in Table 3. The urine spiked with levels lower than the limit of quantitation were not detected (ND, Table 2) as expected. Analysis of 3-PBA spiked into fetal bovine serum at 4 ng/mL gave values nearly identical to 5 ng/mL.^[27] This higher recovery in serum may be due to a matrix effect hence the LOQ was set to 5 ng/mL. Samples with concentrations lower than 5 and 2.5 ng/mL, in serum and urine respectively, were not quantitated. The methods reported here are selective and sensitive enough to assess samples from both acute and chronic exposures to pyrethroids. For example the RfD for pyrethroids is 0.25 mg/kg/day.^[9] Assuming an average person weighs 70 kg, has a blood volume of 5 L, and consumes, in a single day, 0.25 mg of pyrethroid and all was absorbed and circulating in the blood, then the expected concentration in the blood would be about 3.5 μg/mL, well above the LOQs reported here.

For quality control, pooled samples of plasma or urine were used. Two or three replicates of the pooled samples were run with each batch of samples. The coefficient of variation (% CV) was less than 5% for both plasma and urine. Moreover, all control concentrations were within the OCV range (mean ± 2SD; 5.79–6.83 and 14.8–17.8 ng/mL, respectively) as shown in Table 4.

From the questionnaire, data showed that the volunteers in this study used pesticides in the home and in agriculture practices. Of the total study population (consumers and farmers; n = 200), 56% used pesticides indoors. The concentration of plasma 3-PBA in the volunteers who used indoor pesticides (n = 106) was significantly higher (P < 0.005) than who did not use them (n = 94). Twenty-five percent of the farmers (n = 100) applied pesticides agriculturally. The farmers (n = 100) used a variety of pesticide types: 83% insecticides including synthetic pyrethroids, 75% herbicides, 72% chemical fertilizers, 56% organic fertilizers and 41% plant growth substances.

The % detection of plasma 3-PBA was less than 50% overall while the detection of urinary 3-PBA was greater than 50% for both groups (Table 5). Although the rates of detection in plasma or urine were similar between the groups, the range of urinary concentrations was generally higher for the farmer group and the median concentrations were significantly different a P < 0.01 likely reflecting recent acute exposures in the farmer group. Longer-term exposure seems more similar between the groups since the median and range of plasma 3-PBA concentration was similar between groups.

In Table 6, the % detection of plasma 3-PBA and urinary 3-PBA for consumer and farmer groups are broken out by concentration ranges. Plasma concentrations for all subjects were less than 10 ng/mL although the rate of detection was nearly double in farmers compared to consumers in the 5.1–10 ng/mL range. Similarly more farmers than consumers had urinary concentrations of 3-PBA greater than 10 μg/g creatinine and the upper range of concentrations were much higher for farmers. Additionally, the concentrations of urinary 3-PBA in farmers in Chiang Mai at selected percentile values in the present study were higher than found by previous researchers (Fig. 2).^[33,34] This may be due to different agricultural practices.

Thus, the farmer group appears to have greater acute exposure most likely due to their agricultural activities.^[33] Consumers that lived in communities where crops were produced also likely had more potential for exposure to insecticides.^[24] However, there are other pathways and routes by which both farmers and consumers may be exposed to these insecticides. Consumption of pesticide-contaminated vegetables and fruits was thought to be the major route by which volunteers were exposed to pesticides. The other routes were dermal contact and the respiratory route from insecticide drift in the air from vector control programs ^[35] and from usage in the household for controlling insects.

Interestingly, there was no correlation between plasma 3-PBA and urinary 3-PBA concentrations in the study population. However, the farmer population likely had more recent acute exposures that would lead to transient high urinary 3-PBA concentrations. Also, metabolism and urinary elimination of pyrethroids varies by individual.^[36] Further studies are necessary to determine the precise reasons for the difference.