Urinary 1-Hydroxypyrene as an Indicator for Assessing the Exposures

ACS2GO © 2018. ← → → ←. loading. To add this web app to the home screen open the browser option menu and tap on Add to homescreen...
0 downloads 0 Views 62KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 56-61

Urinary 1-Hydroxypyrene as an Indicator for Assessing the Exposures of Booth Attendants of a Highway Toll Station to Polycyclic Aromatic Hydrocarbons

surrogate indicator for assessing workers’ PAH exposures. Considering that the type of traffic designed for a given type of tollbooth is quite similar all over the world, the results obtained from this study, at least, could be served as a stepping-stone for providing a cheaper and convenient way for assessing traffic PAH exposures in the future.

P E R N G - J Y T S A I , * ,† T U N G - S H E N G S H I H , ‡ HSIAO-LUNG CHEN,† WEN-JHY LEE,§ CHING-HUANG LAI,| AND SAOU-HSING LIOU| Department of Environmental and Occupational Health, Medical College, National Cheng Kung University, 138 Sheng-Li Road, Tainan 704, Taiwan, Institute of Occupational Safety and Health, Council of Labor Affairs, Executive Yuan, 99 Lane 407, Heng-Ke Road, Shijr, Taipei, Taiwan, Department of Environmental Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan, and Department of Public Health, National Defense Medical Center, P.O. Box 90048-509, Nei-Hu, Taipei, Taiwan

Polycyclic aromatic hydrocarbons (PAHs) and their derivatives are the major culprits in urban areas that causing human lung cancer (1-3). Among various PAH emission sources, traffic source has been known to be the greatest contributor in urban areas (4-7). In principle, the technique associated with conducting personal PAH sampling in the field has created possibilities for directly assessing workers’ exposures (8). Because the above technique requires sophisticaed chemial analyses for quantitating a wide range of PAH compounds, the biological monitoring technique provides a potential alternative for individual-based exposure assessment. But it raises a question regarding how to select a suitable biomarker that is not only representative of the exposure of a specific PAH compound but is also representative of total PAH exposures. Among various biomarkers, metabolites of pyrene, such as 1-hydroxypyrene (1-OHP) and 1-hydroxypyrene glucuronide (1-OHPG), have been widely used in many studies to characterize workers’ exposures to PAHs (9-12). Nevertheless, conducting biological monitoring might become very costly and labor-intensive especially when many urine samples are needed to meet the research purpose. For traffic sources, the traffic densities of various traffic types have been successfully used as a surrogate indicator for assessing personal exposures in many epidemiological studies (13-16). In our previous study, we have successfully used the traffic densities of varioius traffic types to characterize tollbooth attendants’ PAH exposures (17). Considering the individual variations on motablizing pyrene, whether the above approach is adequate for assessing workers’ urinary 1-OHP concentrations required further investigating. In this study, the characteristics of PAH exposures to toll station workers is briefly described since it has been presented in detail in our earlier work (17). The objective of this study mainly focuses on evaluating (i) the feasibility of using urinary 1-OHP as a biomarker for assessing booth attendants’ PAH exposures and (ii) the possibility of using traffic densities of varioius traffic types for assessing workers’ urinary 1-OHP levels resulting from PAH exposures.

In this study, 32 booth attendants (the exposure group) and 21 in pre-job training to become booth attendants (the reference group) were randomly selected from a highway toll station. Personal PAH samplings were conducted on the exposure group on each day during the studied workweek. Pre-shift urinary 1-hydroxylpyrene levels (1-OHP) were measured on the first day of the workweek (BMpre) for both the exposure and reference groups, but the post-shift 1-OHP levels were measured on the last day of the workweek (BMpost) only for the exposure group. For the exposure group, we found that their mean total PAH exposure level (Ctotal PAHs) was 11 400 ng/m3 and that their mean BMpost was significantly higher than their mean BMpre () 3.02 and 0.910 µmol of 1-OHP/mol of creatinine, respectively). In addition, the mean BMpre for the exposure group were higher than that for the reference group () 0.410 µmol of 1-OHP/ mol of creatinine). The above results suggest that vehicle exhaust significantly affects the booth attendants’ 1-OHP levels. None of the three personal factors (age, work experience, and smoking habit), except for Ctotal PAHs, had a significant effect on predicting booth attendants’ BMinc levels () BMpost - BMpre) (R 2 ) 0.57). The above results suggest that urinary 1-OHP could be a suitable biomarker for characterizing workers’ PAH exposures. Similarly, we found that none of the three personal characteristics, except for the involved vehicle flow rates and vehicle types, had a significant effect on predicting booth attendants’ BMinc levels (R 2 ) 0.60). The above result suggests that the traffic densities of various traffic types could be a suitable * Correspondence author telephone: +886-6-2088390; fax: +8866-2752484; e-mail: [email protected]. † Department of Environmental and Occupational Health, National Cheng Kung University. ‡ Council of Labor Affairs. § Department of Environmental Engineering, National Cheng Kung University. | National Defense Medical Center. 56

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 1, 2004

Introduction

Materials and Methods Selected Toll Station. The selected toll station contained 20 tollbooths. Among them, 4 booths were designed for collecting both cash and prepaid tickets from buses and trucks (i.e., bus/truck tollbooths), 12 booths were designed for collecting prepaid tickets from cars and vans (i.e., car/ticket tollbooths), and 4 booths were designed for collecting cash from cars and vans (i.e., car/cash tollbooth). All tollbooths had the same dimension (L × W × H ) 1.5 m × 1.0 m × 2.1 m). Each of them had one door (W × H ) 75 cm × 190 cm) opened completely toward the vehicle lane for collecting tolls from vehicles and one window installed on the opposite side (L × W ) 65 cm × 30 cm). Toll collections were conducted by 61 female booth attendants, and no male booth attendant worked at the studied toll station. The whole study was conducted in winter, and we found that the only window was closed and that there was no ventilation in each tollbooth. 10.1021/es030588k CCC: $27.50

 2004 American Chemical Society Published on Web 11/22/2003

Studied Exposure and Reference Groups. Thirty-two female booth attendants were randomly selected from the studied toll station. All booth attendants followed a weekly schedule and, as a rule, worked continuously for 4 days and then took 1 day off. Each booth attendant performed her task (i.e., collecting tolls from vehicles) at 1-3 tollbooths during one workshift. We recruited 21 female workers, who were in pre-job training to become booth attendants, as the reference group. All participants’ background information (including age, work experience, and smoking habit) was carefully registered. All participants were requested to fill in a consent form before field samplings were conducted. This study was approved by the institutional review boards at the Tri-Service General Hospital, Taipei, Taiwan. Environmental Monitoring and Sample Analysis. For environmental sampling, we collected personal samples from the breathing zone of each booth attendant of the exposure group for approximately one full workshift (i.e., sampling time ) ∼8 h) on each day during the studied workweek. The sampling method was modified from Method 5515 suggested by National Institute for Occupational Safety and Health (NIOSH) in 1994 and was adopted by our research group for conducting personal samplings on carbon black workers (9, 18). The sampling train consisted of a filter cassette (IOM personal inhalable aerosol sampler; SKC Inc., Catalog No. 225-70) and followed by a sorbent tube (washed XAD-2, 3.5 g/0.5 g) with the sampling flow rate specified at 2.0 and 0.2 L/min for collecting particulate PAHs of the inhalable fraction and gaseous PAHs, respectively. Before the sampling, all filters and sorbent tubes were cleaned and extracted with a solvent solution (mixture of n-hexane and dichloromethane, v:v ) 1:1) for 24 h in a Soxhlet extractor to ensure that it was free from contamination. After the samplings were conducted, both filter and XAD-2 resin were placed again in the above-mentioned solvent solution in a Soxhlet extractor for 24 h extraction. The extract was then concentrated, cleaned up, and reconcentrated to exactly 1.0 or 0.5 mL. PAH contents were determined by using a gas chromatograph (GC) (Hewlett-Packard 5890A) with a mass selective detector (MSD) (Hewlett-Packard 5972) and computer workstation. The GC/MS equipped with a Hewlett-Packard capillary column (HP Ultra 2; 50 m × 0.32 mm × 0.17 µm) and an HP-7673A automatic sampler. Operating conditions specified for the GC/MS included injection volume: 1 µL; splitless injection: 310 °C; ion sources temperature: 310 °C, oven from 50 °C to 100 °C at 20 °C/ min, 100 °C to 290 °C at 3 °C/min, and hold at 290 °C for 40 min. The masses of the primary and secondary PAH ions were determined using the scan mode for pure PAH standards. Qualification of PAHs was performed using the selected ion monitoring (SIM) mode as that used in our previous studies (9, 18-21). The concentrations of 22 PAH compounds for both filter and XAD-2 resin were determined, including naphthalene (Nap), acenaphthalene (AcPy), acenaphthene (Acp), fluorene (Flu), phenanthrene (PA), anthracene (Ant), fluoranthene (FL), pyrene (Pyr), cyclopenta[c,d]pyrene (CYC), benz[a]anthracene (BaA), chrysene (CHR), benzo[b]fluoranthene (BbF), benzo[k]fluoranthene (BkF), benzo[e]pyrene (BeP), benzo[a]pyrene (BaP), perylene (PER), indeno[1,2,3,-cd]pyrene (IND), dibenz[a,h]anthracene (DBA), benzo[b]chrycene (BbC), benzo[ghi]perylene (BghiP), coronene (COR), and dibenzo[a,e]pyrene (DBP). The total PAH content was defined as the sum of the contents of the 22 PAH compounds. Detailed QA/QC procedures were described in our previous studies (9, 18) and hence were not repeated here. This study yielded recovery efficiencies (corrected by the five internal standards of NaP-d8, Acp-d10, PA-d10, CHR-d12, and PER-d12) and the limit of quantification (LOQ) of the 22 PAH compounds fell to the ranges 78.7-102.3% and 40.6-639

pg/m3, respectively. Analyses of field blanks, including the glass fiber filter and XAD-2 sorbent tube, found no significant contamination (i.e., GC/MS integrated area < detection limit). Biological Monitoring and Analysis of Urine Sample. We collected pre-shift urine samples on the first day of the workweek (i.e., BMpre) from both the exposure and reference groups. But we simply collected the post-shift urine samples on the last day of the workweek (i.e., BMpost) from the exposure group. We used a PE bottle, first pretreated with 10% of nitric acid and then rinsed with distilled water, to collect urine sample from each participant. All participants were requested to wash their hands to ensure the contamination-free collection of urine specimens. Immediately after each urine sample was collected, 5 mL of the urine sample was sent to an accredited laboratory to determine its creatinine contents, and the rest was stored at -80 °C until analysis. All urine samples were analyzed to determine their 1-OHP contents by the HPLC method. This method was first developed by Jongeneelen and his colleagues in 1987 (12) and was adopted in our previous study (9). For each collected urine sample, 10 mL of the urine specimen was first adjusted to pH 5.0 by adding a buffer solution (the mixture of 1 N hydrochloric acid and 0.1 M acetate) to a final volume of 30 mL. Then, the sample was incubated for 24 h with 15 µL of glucoronidase arylsulfatase (134 600 unit/mL, Sigma, Lot 97H3386) at 37 ( 0.5 °C in an electronically controlled rotary shaking bath (Hotech, shaker bath model 903). A sample purification and enrichment cartridge, packed with C18 reversed-phase liquid chromatograph material (Waters, 500 mg/3 mL) was used to extract the metabolites. The flow rate of the treated urine passing through the cartridge was specified at ∼3 mL/mim. The cartridge was washed with 10 mL of distilled water and 3 mL of 50% methanol. Final elution of 1-OHP was performed with 10 mL of methanol. The solution was evaporated to dryness and reconstituted with 2 mL of methanol. The HPLC system consisted of a highperformance liquid chromatograph with an auto-injector (Hewlett-Packard 1100) and a fluorescence detector (Hewlett HP-1046A). The extracts (20 µL) were injected onto a 150 × 4.0 mm column (Supelco RP-18) with the column temperature and flow rate specified at 40 °C and 1.0 mL/min, respectively. The excitation wavelength (λex) and the emission wavelength (λem) of the fluorescence detector were specified at 241 and 395 nm, respectively. The detection limit of the method was ∼5.43 ng according to the data obtained from seven repeated analyses at a concentration 15.0 ng/dL. The reproducibility of the method, determined by repeated analysis of urine samples at concentrations from 2.85 to 9.99 µg/dL, was found with variation coefficients ranging from 1.87% to 8.40%. To minimize the effects arising from the hydration states of workers, the urinary 1-OHP concentrations were calibrated by their corresponding creatinine concentrations and were presented by using the unit of micromoles of 1-OHP per moles of creatinine. Measuring the Involved Traffic Densities for Various Traffic Types. As mentioned earlier, each booth attendant was found to collect tolls from vehicles at 1-3 tollbooths during one workshift. We carefully registered the durations for each selected booth attendant staying at each individual tollbooth during the workshift. In addition, the traffic for each tollbooth during one workshift was counted by using a pneumatic tube, which was laid across the vehicle lane and was connected to an automatic data logger. The above measurements allowed us to estimate the vehicle flow rates of various traffic types experienced by each selected booth attendant during one workshift.

Results Studied Population. Personal background information for both the exposure and the reference groups are shown in VOL. 38, NO. 1, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

57

TABLE 1. Personal Background Information of the Exposure and Reference Groups statistics type of workers

mean

SD

Exposure Group (n ) 32) nonsmokers (n ) 30) age (yr) 26.1 5.8 work experience (yr) 2.5 2.0 smokers (n ) 2) age (yr) 21.9 3.1 work experience (yr) 0.6 0.2 total (n ) 32) age (yr) 25.9 5.7 work experience (yr) 2.4 1.9 Reference Group (n ) 21) nonsmokers (n ) 14) age (yr) 22.5 2.4 work experience (yr) 0.25 0 smokers (n ) 7) age (yr) 23.3 2.7 work experience (yr) 0.25 0 total (n ) 21) age (yr) 23.3 2.7 work experience (yr) 0.25 0

PAH compound

max

min

43.5 8.9

19.2 0.3

24.1 0.8

19.7 0.5

43.5 8.9

19.2 0.3

28 0.25

20 0

29 0.25

21 0

29 0.25

21 0

Table 1. The mean ages for the exposure and reference group were 25.9 and 23.3 yr, respectively. The mean work experience for the exposure group (2.4 yr) was longer than that for the reference group (0.25 yr; all newly recruited on pre-job training status). The fraction of current smokers for the exposure group (6.25%) was lower than that for the reference group (33.3%). PAH Exposure Profiles. Table 2 shows the PAH exposure profile of the exposure group. The mean total PAH exposure level was 11 400 ng/m3, and the mean fractions for the particulate phase and gaseous phase PAHs were 11.7% and 88.3%, respectively. To assess the PAH homologue distribution of the exposure profile, PAH compounds with low molecular weights (LM-PAHs, containing 2-3-ringed PAHs), middle molecular weights (MM-PAHs, containing 4-ringed PAHs), and high molecular weights (HM-PAHs, containing 5-7-ringed PAHs) were also determined. This study yielded the mean fractions 86.1%, 2.29%, and 11.6% for LM-PAHs, MM-PAHs, and HM-PAHs, respectively. Urinary 1-OHP Concentrations. Table 3 shows biological monitoring results for both exposure and reference groups. Among booth attendants of the exposure group, the mean BMpre for smokers was higher than that for nonsmokers () 1.34 and 0.882 µmol of 1-OHP/mol of creatinine, respectively), but the above difference was statistically insignificant (p ) 0.26; Mann-Whitney test). Among booth attendants of the reference group, we found that the mean BMpre for smokers was significantly higher than that for nonsmokers () 0.686 and 0.272 µmol of 1-OHP/mol of creatinine, respectively) (p ) 0.01; Mann-Whitney test). The mean BMpre for nonsmokers of the exposure group was significantly higher than that for nonsmokers of the reference group (p < 0.001; Mann-Whitney U test). On the other hand, although the mean BMpre for smokers of the exposure group was higher than that for smokers of the reference group, the above difference was statistically insignificant (p ) 0.24; Mann-Whitney test). Nevertheless by combining both smokers and nonsmokers, the mean BMpre for the exposure group was significantly higher than that for the reference group () 0.910 and 0.410 µmol of 1-OHP/mol of creatinine, respectively) (p < 0.001; Mann-Whitney U test). For nonsmokers of the exposure group, we found that their BMpost values were significantly higher than their BMpre () 3.01 and 0.882 µmol of 1-OHP/mol of creatinine, respectively) 58

9

TABLE 2. PAH Exposure Profiles for the Exposure Group (n ) 32)a

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 1, 2004

Nap AcPy Acp Flu PA Ant FL Pyr CYC BaA CHR BbF BkF BeP BaP PER IND DBA BbC BghiP COR DBP total PAHs

mean 8850 267 142 186 264 95.9 103 105 14.1 22.7 31.5 91.6 215 185 126 96.1 17.4 23.6 218 40.4 267 39.8 11400

fractions of PAH phases (%) particle phase gas phase fractions of PAH homologues (%) LM-PAHs MM-PAHs HM-PAHs a

SD 1774 107 28.8 42.5 82.5 19.7 59.1 31.3 13.7 19.0 21.7 84.1 220 146 48.7 67.2 22.0 30.8 108 24.7 186 65.5 2210

max

min

12300 518 212 283 422 133 253 165 52.4 59.4 72.5 300 731 474 247 325 80.4 112 481 79.4 699 250 15100

5040 131 86.8 100 111 51.5 41.1 51.2 0.00 0.00 0.00 0.00 15.8 26.2 44.5 22.9 0.00 0.00 58.11 0.00 30.0 0.00 6310

11.7 88.3

8.62 5.62

39.9 92.4

7.6 60.1

86.1 2.29 11.6

6.10 0.798 5.46

94.8 4.10 23.9

72.8 1.08 3.97

In units of ng/m3.

TABLE 3. Urinary 1-OHP Concentrations of BMpre, BMpost, and BMinc for Exposure and Reference Groupsa workers nonsmokers (n ) 30) BMpre BMpost BMinc smokers (n ) 2) BMpre BMpost BMinc total (n ) 32) BMpre BMpost BMinc nonsmokers (n ) 14) BMpre smokers (n ) 7) BMpre total (n ) 21) BMpre a

mean

SD

max

min

Exposure Group 0.882 3.01 2.13

0.388 0.662 0.593

1.88 4.14 3.16

0.420 1.41 0.424

1.34 3.20 1.92

0.792 1.80 1.24

1.94 4.68 2.70

0.601 1.48 0.779

0.910 3.02 2.12

0.417 0.717 0.612

1.94 4.68 3.16

0.420 1.41 0.424

Reference Group 0.272

0.148

0.460

0.014

0.686

0.313

1.13

0.250

0.410

0.289

1.13

0.014

In units of µmol of 1-OHP/mol of creatinine.

(p < 0.05; Mann-Whitney test). But for smokers of the exposure group, although their mean BMpost was higher than their mean BMpre () 3.20 and 1.34 µmol of 1-OHP/mol of creatinine, respectively), the above difference was not statistically significant (p ) 0.101; Mann-Whitney test). By combining both smokers and nonsmokers together, we found that their mean BMpost was significantly higher than their mean BMpre () 3.02 and 0.910 µmol of 1-OHP/mol of creatinine, respectively) (p < 0.05; Mann-Whitney test).

TABLE 4. Relationship between BMInc and Their Corresponding CTotal-PAHs and the Three Personal Factors of Age, Work Experience, and Smoking Habit

TABLE 6. Relationship between BMInc and the Involved Vehicle Flow Rates of Qcar/cash, Qcar/ticket, and Qbus/truck and the Personal Factors of Age, Work Experience, and Smoking Habit

BMInc (µmol of 1-OHP/mol of creatinine)

BMinc (µmol of 1-OHP/mol of creatinine)

regression parameters

coeff

SE

Ctotal PAHs (ng/m3) 0.113 0.023 age (yr) -0.012 0.010 work experience (yr) 0.054 0.032 smokinga 0.046 0.041 intercept 1.04 0.350 a

regression

p value 0.00 0.28 0.11 0.27 0.01

R2

F p ratio value

0.57 8.77

0.00

Yes ) 1, no ) 0.

parameters

type of vehicle flow rate

mean

SD

range

Qcar/cash Qcar/ticket Qbus/truck

865 2779 686

989 2730 1007

0-3883 0-7271 0-3158

a

In units of vehicles/shift.

Using Urinary 1-OHP Concentrations To Characterize PAH Exposures. Considering the complexity of PAHs in their compositions, we need to examine the followings before the worker’s urinary 1-OHP can be regarded as a suitable indicator for assessing worker’s PAH exposure: (i) the correlation between workers’ pyrene exposure levels (CPyr) and their corresponding total PAH exposure levels (Ctotal PAHs) and (ii) the relationship between workers’urinary 1-OHP concentrations and their corresponding Ctotal PAHs levels. We conducted Pearson correlation analysis to examine the relationship between Ctotal PAHs and CPry (22). This study yielded a correlation coefficient as high as 0.87. As mentioned earlier that BMpre and BMpost were defined respectively as workers’ urinary 1-OHP concentrations of the pre-shift on the first workday and concentrations of the post-shift on the last day of the workweek, we hence defined BMinc as the increase of the urinary 1-OHP concentration during the whole workweek (i.e., BMinc ) BMpost - BMpre). It is known that worker’s personal factors (including age, work experience, and smoking habit) might play an important role on her BMinc. We conducted multivariate regression analysis to relate workers’ BMinc levels to their corresponding Ctotal PAHs levels and three personal factors. Results show that none of the three personal factors, except for Ctotal PAHs, was statistically significant (Table 4). This study yielded R 2 ) 0.57, indicating that the increase of urinary 1-OHP concentrations during the workweek could be fairly explained by PAH exposures arising from vehicle exhausts. Using the Involved Traffic Densities and Vehicle Types To Relate Workers’ Urinary 1-OHP Concentrations. In this study, the involved three types of vehicle flow rate (denoted as Qcar/cash, Qcar/ticket, and Qbus/truck for the traffics measured at the car/cash tollbooth, car/ticket tollbooths, and bus/truck tollbooth, respectively) for each booth attendant were summarized in Table 5. Again, because worker’s personal factors (i.e., age, work experience, and smoking habit) might play an important role on BMinc, the above factors were taken into account while conducting multivariate regression analysis. Results show that none of the three personal factors, except for the three involved vehicle flow rates (i.e., Qcar/cash, Qcar/ticket, and Qbus/truck) were statistically significant (Table 6). This study yielded R 2 ) 0.60, indicating that the increase of urinary 1-OHP concentrations during the workweek could

SE

p value

R2

F p ratio value

Qcar/cash (102 vehicle/shift) 0.022 0.008 0.01 0.60 6.13 0.00 Qcar/ticket (102 vehicle/shift) 0.018 0.004 0.00 Qbus/truck (102 vehicle/shift) 0.049 0.010 0.00 age (yr) -0.008 0.011 0.47 work experience (yr) 0.065 0.033 0.06 smokinga 0.066 0.041 0.12 intercept 1.01 0.037 0.01 a

TABLE 5. Involved Three Types of Vehicle Flow Rate of Qcar/cash, Qcar/ticket, and Qbus/truck for the Exposure Groupa

coeff

Yes ) 1, no ) 0.

be fairly explained by the involved vehicle flow rates and vehicle types.

Discussion We found that total PAH exposure levels for booth attendants (11 400 ng/m3) were mainly contributed by the gaseous phase PAHs rather than particulate phase PAHs () 88.3% and 11.7%, respectively) (Table 2). The above result can be confirmed by examining the homologue distributions of the collected PAHs. Also shown in Table 2, we can see that the collected PAHs were mainly contributed by both LM-PAHs and MMPAHs with high volatility (in total accounted for ∼88.4% of total PAHs). On the other hand, low volatility PAHs (i.e., HMPAHs) accounted for only ∼11.6% of total PAHs. Based on these, it is not so surprising to see that the collected PAHs were mainly contributed by gaseous PAHs rather than particulate PAHs. Because all selected booth attendants were requested to collect tolls from vehicles, it is expected that their PAH exposures should be mainly contributed by traffic emissions. But still the above statement can be further confirmed by comparing the results obtained from this study with the results obtained from our previous laboratory studies on investigating PAH emissions from gasoline-powered and heavy-duty-diesel (HDD) vehicle engines (20, 23). For gasoline-powered engines (under idling condition), we found that LM-PAHs, MM-PAHs, and HM-PAHs accounted for 82.2-92.8%, 1.50-9.47%, and 5.66-7.37% of total PAHs in the engine exhaust, respectively (20). Similar results (96.9%, 1.04%, and 2.05%, respectively) were found in HDD engine exhaust (under idling condition) (23). In principle, the above two PAH homologue distributions shared the same trend with the results that were found in this study. Nevertheless, there still is some difference in PAH homologue distributions between that found in the laboratory studies (20, 23) and that found in this field study. This could be partly because the driving condition tested in our previous laboratory studies (i.e., idling) was not able to fully reflect the real driving condition at tollbooth (i.e., from low speed, idling, to low speed). In this study, we found that the mean BMpre for nonsmokers of the exposure group was significantly higher than that for nonsmokers of the reference group (Table 3). In principle, the above results were consistent with that found by Merlo et al. (24). In their study, they compared BMpre levels among 94 traffic police officers and 52 reference subjects in Genoa, Italy. The mean BMpre for nonsmokers in the exposure group (0.102 µmol of 1-OHP/mol of creatinine) was higher than that in the reference group (0.067 µmol of 1-OHP/mol of creatinine). In both booth attendants of the exposure and reference groups, we found that the mean BMpre for smokers was higher than that for nonsmokers (Table 3). VOL. 38, NO. 1, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

59

The above results clearly indicate that the smoking habit did play an important role on BMpre. To eliminate the effect of the smoking habit on worker’s urinary 1-OHP concentrations, we hence used BMinc levels (BMpost - BMpre) to assess workers’ PAH exposures associated with traffic emissions. We conducted Pearson correlation analysis to examine the relationship between Ctotal PAHs and CPry. This study yielded a correlation coefficient as high as 0.87. The above result is not so surprising since all selected booth attendants were exposed to similar emission sources (i.e., traffic sources). On the other hand, the above result also suggests that the individual compound pyrene could be regarded as a suitable indicator for total PAH exposures for the exposure group. On this basis, it is expected that BMinc of the exposure group might be able to be explained by their total PAH exposures (i.e., Ctotal PAHs). But considering that a worker’s personal factors (i.e., age, work experience, and smoking habit) might play an important role on her BMinc, we conducted multivariate regression analysis to examine the relationship between workers’ BMinc levels and their corresponding Ctotal PAHs levels and their three personal factors. Interestingly we found that none of the three personal factors, except for Ctotal PAHs, was statistically significant (Table 4). The former two factors (i.e., age and work experience) were found with insignificant effect on BMinc could be due to their small variations in the exposure group (Table 1), as compared with the variations of Ctotal PAHs (Table 2). Regardless, the smoking habit was also found to have an insignificant effect on BMinc, although the above result was consistent with that found in other studies (25-27) and still worth further discussion. In principle, it is known that the contribution of the “smoking habit” to urinary 1-OHP was not negligible for the nonoccupational exposure group. For example, Viau et al. has suggested that the contribution of the smoking habit to the increase of urinary 1-OHP was as high as ∼0.105 µmol of 1-OHP/mol of creatinine (27). But here it should be noted that the above value was not so significant as compared with BMinc that was found in this study for booth attendants (i.e., 2.12 µmol of 1-OHP/mol of creatinine; see Table 3). Based on the above comparison, it is suggested that the results obtained from this study could be theoretically plausible. For the same reason, we found that none of the three personal factors, except for the three involved vehicle flow rates (i.e., Qcar/cash, Qcar/ticket, and Qbus/truck), had a significant effect on workers’ BMinc levels (Table 6). Here, it should be noted that the magnitudes of the regression coefficients for the three involved vehicle flow rates were not the same and hence warrant the need for further discussion. This study yielded the regression coefficients 0.022, 0.018, and 0.049 for Qcar/cash, Qcar/ticket, and Qbus/truck, respectively (Table 6). The above result suggests that Qbus/truck had the highest contribution to workers’ BMinc among the three involved vehicle flow rates. In our previous studies, we have found that total PAHs in the heavy-duty diesel engine exhaust (1500 µg/m3, under idling conditions) (23) was higher than that in the gasolinepowered engine exhaust (337 µg/m3, using grade 95 leadfree gasoline under idling conditions) (20). In addition, in this study we found that the time for a vehicle spent at the bus/truck tollbooth, car/cash tollbooth, and car/ticket tollbooth (for paying the toll) was ∼7.6, 6.3, and 3.8 s, respectively. Based on these, it is not surprising to see that Qbus/truck (because most involved vehicles with diesel engines and the longest toll collection time) had the highest contribution to workers’ BMinc levels than Qcar/cash and Qcar/ticket (because most involved vehicles with gasoline powered engines and shorter toll collection time). Considering both Qcar/cash and Qcar/ticket shared the same type of traffic, we found that the regression coefficient for Qcar/cash was greater than that for Qcar/ticket, which might be due to their intrinsic differences in the toll collection times (i.e., 6.3 s > 3.8 s). 60

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 1, 2004

It is known that, besides traffic emissions, personal diet (28-30) and background exposures to ambient environments and indoor pollutant sources, such as burning of incense (31) and mosquito coils (32), cooking (33, 34), and use of fireplace (35), might significantly affect the worker’s urinary 1-OHP level. But in this study the effect of above factors might be negligible since all booth attendants were known living in the same dormitory and eating the same diet. This study yielded R 2 ) 0.60 (Table 6), which does suggest that tollbooth attendants’ BMinc levels could be fairly explained by their involved vehicle flow rates and vehicle types. In addition, it should be noted that all urine specimens were collected from a group of female-only operators. Therefore, whether the results obtained from this study are applicable to a corresponding male population might worth further discussion. In our previous study conducted on carbon black industries (9), we found that the effect of gender on the increase of worker’s urinary 1-OHP level was statistically insignificant. The above result suggests that the methodology used in this study could be also suitable for assessing workers’ PAH exposures for a corresponding male population. Finally, according to the statistic data provided by the Department of Transportation in Taiwan, we found that the average fleet ages for cars (including vans with less than 12 seats), buses (including vans with 12 seats and above), and trucks are 7.47, 7.74, and 9.58 years, respectively. It is known that the vehicle condition might have a significant effect on PAH emissions from vehicle engines. Therefore, the results obtained from this study should be used with caution in other countries with different mean fleet ages. Yet, it could be true that whether the results obtained from this study are applicable to other toll stations with different surroundings and topographical features warrant the need for further investigations. Nevertheless, considering that the type of traffic designed for a given type of tollbooth is quite similar all over the world, the results obtained from this study, at least, could be served a steppingstone for providing a cheaper and convenient way for assessing tollbooth attendants’ PAH exposures in the future.

Acknowledgments The authors thank the Institute of Occupational Safety and Health (IOSH) of the Council of Labor Affairs in Taiwan for funding this research project.

Literature Cited (1) Doll, R.; Peto, R. Natl. Cancer Inst. J. 1981, 66, 1191. (2) Speizer, F. E. Environ. Health. Perspect. 1986, 70, 9. (3) Westerholm, R.; Alsberg, T. E.; Frommelin, A. B.; Strandell, M. E.; Rannug, U.; Winquist, L.; Grigoriadis, V.; Egeba¨ck, K. Environ. Sci. Technol. 1988, 22, 925. (4) Henderson, T. R.; Sun, J. D.; Li, A. P.; Hanson, R. L.; Bechtold, W. E.; Harvery, T. M.; Shabanowitz, J.; Hunt, D. F. Environ. Sci. Technol. 1984, 18, 428. (5) Pyysalo, H.; Tuominen, J.; Wickstrom, K.; Skytta, E.; Tikkanen, L.; Salomaa, S.; Morsa, M.; Nurmela, T.; Mattla, T.; Pohjola, V. Atmos. Environ. 1987, 21, 1167. (6) Tuominen, J.; Salomaa, S.; Pyysalo, H.; Skytta, E.; Tikkancen, L.; Nurmela, T.; Sorsa, M.; Pohjola, V.; Sauri, M.; Himberg, K. Environ. Sci. Technol. 1988, 22, 1228. (7) Baek, S. O.; Field, R. A.; Goldstone, M. E.; Kirk, P. W.; Lester, J. N.; Perry, R. A. Water Air Soil Pollut. 1991, 60, 273. (8) Sexton, K.; Ryan, P. B. Assessment of human exposure to air pollution: methods, measurements, and models. In Air Pollution, the Automobile, and Public Health; Watson, A. Y., Bates, R. R., Kennedy, D., Eds.; National Academy Press: Washington, DC, 1988; p 207. (9) Tsai, P.-J.; Shieh, H.-Y.; Lee, W.-J.; Chen, H.-L.; Shih, T.-S. Ann. Occup. Hyg. 2002, 46, 229. (10) Strickland, P.; Kang, D. Toxicol. Lett. 1999, 108, 191. (11) Jongeneelen, F. J.; Bos, R. P.; Anzion, R. B. M. Scand. J. Work Environ. Health 1986 12, 137. (12) Jongeneelen, F. J.; Anzion, R. B. M.; Henderson, P. T. J. Chromatogr. 1987, 413, 227.

(13) Savitz, D. A.; Feingold, L. Scand. J. Work, Environ. Health 1989, 15, 360. (14) Weiland, S. K.; Mundt, K. A.; Ruckmamm, A.; Keil, U. Ann. Epidemiol. 1994, 4, 243. (15) Feychting, M.; Svensson, D.; Ahlbom, A. Scand. J. Work, Environ. Health 1998, 24, 8. (16) Raaschou-Nielsen, O.; Hertel, O.; Thomsen, B. L.; Olsen, J. H. Am. J. Epidemiol. 2001, 153, 433. (17) Tsai, P.-J.; Shih, T.-S.; Chen, H.-L.; Lee, W.-J.; Lai, C.-H.; Liou, S.-H. Atmos. Environ. 2004, 38, 333. (18) Tsai, P.-J.; Shieh, H.-Y.; Lee, W.-J.; Lai, S.-O. J. Occup. Health 2001, 43, 118. (19) Tsai, P.-J.; Shieh, H.-Y.; Hsieh, L.-T.; Lee, W.-J. Atmos. Environ. 2001, 35, 3495. (20) Mi, H.-H.; Lee, W.-J.; Tsai, P.-J.; Chen, C.-B. Environ. Health Perspect. 2001, 109, 1285. (21) Lee, W.-J.; Liow, M.-C.; Tsai, P.-J.; Hsieh, L.-T. Atmos. Environ. 2002, 36, 781. (22) Farant, J. P.; Garie´py, M. Am. Ind. Hyg. Assoc. J. 1998, 59, 758. (23) Mi, H.-H.; Lee, W.-J.; Chen, S.-J.; Lin, T.-C.; Wu, T.-L.; Hu, J.-C. Chemosphere 1998, 36, 2031. (24) Merlo, F.; Andreassen, A.; Weston, A. Cancer Epidemiol. Biomarkers Prev. 1998, 7, 147. (25) Buchet, J. P.; Gennart, J. P.; Mercado-Calderon, F.; Delavignette, J. P.; Cupers, L.; Lauwery, R. Br. J. Ind. Med. 1992, 49, 761.

(26) Ny, E. T.; Heederik, H.; Kromhout, H.; Jongeneelen, F. Am. Ind. Hyg. Assoc. J. 1993, 54, 277. (27) Viau, C.; Vyskocil, A.; Martel, L. Sci. Total Environ. 1995, 163, 191. (28) Buckley, T. J.; Lioy, P. J. Br. J. Ind. Med. 1992, 49, 113. (29) Kang, D. H.; Rothman, N.; Poirier, M. C.; Greenberg, A.; Hsu, C. H.; Schwartz, B. S.; Baser, M. E.; Groopman, J. D.; Weston, A.; Strickland, P. T. Carcinogenesis 1995, 16, 1079. (30) Van Rooij, J. G..; Veeger, M. M.; Bodelier-Bade, M. M.; Scheepers, P. T.; Jongeneelen, F. J. Int. Arch. Occup. Environ. Health 1994, 66, 55. (31) Wallace, L. Appl. Occup. Environ. Hyg. 2000, 15, 39. (32) Lin, J. M.; Lee, J. K. Bull. Environ. Contam. Toxicol. 1997, 59, 868. (33) Levy, J. I.; Dumyahn, T.; Spengler, J. D. J. Exposure Anal. Environ. Epidemiol. 2002, 12, 104. (34) Van Winkle, M. R.; Scheff, P. A. Indoor Air 2001, 11, 49. (35) McCrillis, R. C.; Burnet, P. G. Toxicol. Ind. Health 1990, 6, 95.

Received for review August 8, 2003. Revised manuscript received October 8, 2003. Accepted October 9, 2003. ES030588K

VOL. 38, NO. 1, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

61