Urinary metabolites of polycyclic aromatic

1 downloads 0 Views 332KB Size Report
Feb 10, 2011 - mastic asphalt application and resulted in a median con- centration of 1.85 (1.26–3.02) g/m3 in bitumen-exposed workers and 0.21 (0.17–0.26) ...

Arch Toxicol DOI 10.1007/s00204-011-0680-7

T O X I CO K I N E T I S A N D M E T A BO L IS M

Urinary metabolites of polycyclic aromatic hydrocarbons in workers exposed to vapours and aerosols of bitumen Beate Pesch · Anne Spickenheuer · Benjamin Kendzia · Birgit Karin Schindler · Peter Welge · Boleslaw Marczynski · Hans-Peter Rihs · Monika Raulf-Heimsoth · Jürgen Angerer · Thomas Brüning

Received: 7 January 2011 / Accepted: 10 February 2011 © Springer-Verlag 2011

Abstract Urinary hydroxylated metabolites of polycyclic aromatic hydrocarbons (PAH) were investigated as potential biomarkers of bitumen exposure in a cross-shift study in 317 exposed and 117 non-exposed workers. Personal measurements of the airborne concentration of vapours and aerosols of bitumen during a working shift were weakly associated with post-shift concentrations of 1-hydroxypyrene (1-OHP) and 1-, 2+9-, 3- and 4-hydroxyphenanthrenes (further referred to their sum as OHPHE), but not 1- and 2-hydroxynaphthalene (OHNA). Smoking showed a strong inXuence on the metabolite concentrations, in particular on OHNA. Pre-shift concentrations of 1-OHP and OHPHE did not diVer between the study groups (P = 0.16 and P = 0.89, respectively). During shift, PAH metabolite concentrations increased in exposed workers and non-exposed smokers. Statistical modelling of postshift concentrations revealed a small increase in 1-OHP by a factor of 1.02 per 1 mg/m3 bitumen (P = 0.02) and 1.04 for OHPHE (P < 0.001). A group diVerence was observed Electronic supplementary material The online version of this article (doi:10.1007/s00204-011-0680-7) contains supplementary material, which is available to authorized users. B. Pesch (&) · A. Spickenheuer · B. Kendzia · B. K. Schindler · P. Welge · B. Marczynski · H.-P. Rihs · M. Raulf-Heimsoth · J. Angerer · T. Brüning Institute for Prevention and Occupational Medicine of the German Social Accident Insurance, Institute of the Ruhr University Bochum (IPA), Bürkle-de-la-Camp-Platz 1, 44789 Bochum, Germany e-mail: [email protected] B. K. Schindler · J. Angerer Institute and Outpatient Clinic of Occupational, Social- and Environmental Medicine, University of Erlangen-Nuremberg, Schillerstrasse 25/29, Erlangen, Germany

that was diminished in non-smokers. Exposed non-smokers had a median post-shift 1-OHP concentration of 0.42 g/l, and non-smoking referents 0.13 g/l. Although post-shift concentrations of 1-OHP and OHPHE were slightly higher than those in the general population, they were much lower than in coke-oven workers. The small content of PAHs in vapours and aerosols of bitumen, the increasing use of additives to asphalt mixtures, the strong impact of smoking and their weak association with airborne bitumen limit the use of PAH metabolites as speciWc biomarkers of bitumen exposure. Keywords Biomarker · Biomonitoring · Bitumen · Exposure · Polycyclic aromatic hydrocarbons

Introduction Bitumen, a complex mixture of agents derived from the distillation of crude petroleum oil, is widely used in road paving, rooWng and many other surface applications. In contrast, coal tar is a by-product from coke production or steel industry and contains aromatic compounds at much higher concentrations. During hot applications of bitumen, volatile compounds are released as complex mixtures, mainly consisting of aliphatic hydrocarbons. Progress is underway to reduce the emission of vapours and aerosols of bitumen (further referred to as fumes of bitumen) by lowering the temperature of asphalt laying and by improving ventilation. Investigation into potential health risks of bitumen fumes implies an appropriate assessment of the exposure to bitumen (Burstyn et al. 2000). However, measurements of bitumen compounds are challenging due to the complex nature of its vapours and aerosols (Herrick et al. 2007). Fumes of bitumen can be sampled with personal or stationary

123

Arch Toxicol

devices and determined analytically according to a deWned standard (Burstyn et al. 2002). The measurement of vapours and aerosols of bitumen in the Human Bitumen Study is described by Breuer et al. (2011) in this issue. Biomonitoring allows for the assessment of internal exposure not only via the inhalative route of uptake but also by dermal uptake (Angerer et al. 2007; McClean et al. 2007). Hydroxylated metabolites of pyrene, phenanthrene and naphthalene are commonly used biomarkers of PAH exposure. 1-Hydroxypyrene (1-OHP) has successfully been used in biomonitoring to assess environmental or occupational exposure to PAH (Jongeneelen 2001). The isomeric hydroxyphenanthrenes became together with 1-OHP integral part of large population surveys (Becker et al. 2003; CDC 2005; Umweltbundesamt 1998). Also, asphalt workers have been subject of an extensive body of biomonitoring with PAH metabolites (for review, see Hansen et al. 2008). However, bitumen contains only small amounts of PAH. Additionally, non-occupational inXuences contribute to a wide variation of the metabolite concentrations in asphalt workers (NIOSH 2000). Only few studies have been conducted in asphalt workers with a more comprehensive statistical modelling of the association between external and internal exposure (McClean et al. 2004; Sobus et al. 2009). In order to investigate urinary hydroxylated metabolites of pyrene, phenanthrene and naphthalene as potential biomarkers of bitumen exposure and their associations with personal measurements of airborne bitumen, we present results from a nationwide cross-shift study in workers with and without occupational exposure to fumes of bitumen in Germany.

Materials and methods

tional hygienist documented exposure-related information of the construction site. All study subjects provided a written informed consent. The study was approved by the ethics committee of the Ruhr University Bochum and was conducted in accordance with the Helsinki Declaration. Exposure to vapours and aerosols of bitumen Details of the measurement of vapours and aerosols of bitumen are described by Breuer et al. in this issue. In brief, personal air sampling was performed during a shift in the worker’s breathing zone according to BGIA method No. 6305 (Breuer 2008). This GGP sampling system has been adapted for collecting agents in the gas phase as vapours and aerosols in the inhalable fraction of particulate matter. A 37-mm glass-Wbre Wlter is mounted in the Wlter holder for Wltration of particles. Vapours are adsorbed on 3 g puriWed Amberlite™ XAD-2 resin inserted into the GGP cartridge. The Xow rate was 3.5 l/min. The glass-Wbre Wlter and the XAD-2 cartridge were replaced after 2–4 h. The average shift concentration was calculated as time-weighted average of the concentrations derived from each sample. The samples were shipped to the Institute for Occupational Safety and Health of the German Social Accident Insurance for the extraction of bitumen-derived matter with hydrocarbon-free tetrachloroethylene as solvent and then analysed by Fourier transform infrared spectrometry (Thermo FTIR Avatar 370, C–H stretching vibration of alkanes at 2,800–3,000 cm¡1, 10-mm quartz glass cuvette, 16£ spectrum addition, resolution 4 cm¡1). The signals were calibrated against a mineral oil standard to derive the concentrations of vapours and aerosols of bitumen. The limit of quantitation (LOQ) was about 0.2 mg/m³. Urine collection and determination of PAH metabolites, cotinine and creatinine

Study population A cross-sectional cross-shift study was conducted at 59 German construction sites with and without hot bitumen applications between 2001 and 2008. Details of the study design and conduct are described by Raulf-Heimsoth et al. (2011a, b) in this issue. In brief, the study population of this biomarker analysis comprised 317 male bitumen-exposed workers and 117 roadside construction workers as referents. In order to measure airborne exposure to fumes of bitumen, all bitumen-exposed workers and 68 workers of the reference group were equipped with a personal air sampler during the working shift. Before and after shift, the workers provided spot urine for the determination of urinary PAH metabolites. A structured questionnaire was applied in a face-to-face interview to assess demographic characteristics, smoking habits and other data. An occupa-

123

Spot urine was collected from each subject before and after shift. The analysis of the monohydroxylated metabolites of pyrene and phenanthrene was performed as described elsewhere with slight modiWcations (Lintelmann and Angerer 1999; Marczynski et al. 2002). In brief, samples were collected in polypropylene tubes, shipped on dry ice to the Institute and Outpatient Clinic of Occupational, Social and Environmental Medicine at University of ErlangenNuremberg and stored at ¡20°C until preparation. The determination of 1-OHP and 1-, 2+9-, 3- and 4-hydroxyphenanthrenes (further referred to their sum as OHPHE) was carried out by high-performance liquid chromatography (HPLC) and Xuorescence detection. After enzymatic hydrolysis, the metabolites were enriched on a pre-column, consisting of a copper phthalocyanine-modiWed silica gel, separated on a RP-C18 column and quantiWed by Xuorescence detection.

Arch Toxicol

LOQs of the injected volume of 3 ml ranged between 12 and 48 ng/l. The metabolites of naphthalene and 1- and 2-hydroxynaphthalene (OHNA) were determined by 3D liquid chromatography and Xuorescence detection as described elsewhere (Preuss and Angerer 2004). In brief, after enzymatic hydrolysis, the metabolites were trapped on a restricted access material (RAM RP-8) pre-column for a Wrst online-clean-up step. The analytes were transferred to a silica-based cyanophase column in back-Xush mode to cut oV further matrix compounds and pre-separate the analytes. Final separation of 1- and 2-hydroxynaphthalene was achieved on a C12 bonded reversed phase column with trimethylsilyl endcapping. LOQs were 4.5 g/l (1-hydroxynaphthalene) and 1.5 g/l (2-hydroxynaphthalene) for the injected volume of 350 l. Urinary cotinine was determined by HPLC–UV at 500 nm. After centrifugation, the urine sample is injected into the HPLC system. The König reaction forms a red-coloured complex after addition of potassium cyanide, chloroamino-T and 1,3-diethylthio barbituric acid. Separation from matrix is achieved by liquid chromatography and detection by UV/Vis spectrometry. Aqueous standard solutions are used for calibration and processed in a corresponding manner as the urine samples. Limit of detection was 5 g/l. A cut-oV of 100 g/l was applied to deWne smoking status. Urinary creatinine (crn) was determined photometrically according to the JaVé method (Taussky 1954). Statistical analysis All calculations were performed with SAS/STAT version 9.2 (SAS Institute Inc., Cary, NC, USA). Metabolite concentrations below LOQ were substituted by 2/3 LOQ. Smoking status was based on self-assessed information from the questionnaire. Non-smokers with cotinine concentrations over 100 g/l were reclassiWed as smokers. The distributions of metabolite concentrations were presented with median and interquartile range (IQR) stratiWed by exposure group (bitumen-exposed workers, reference group) and smoking status (current smokers, non-smokers) in pre- and post-shift samples. DiVerences in the distributions were assessed with Wilcoxon–Mann–Whitney test. Crude associations between the various urinary variables and with airborne bitumen exposure were estimated with Spearman rank correlation coeYcients (rs) with 95% conWdence limits (CI). Statistical modelling of the associations between the external and internal dose was carried out with three linear regression models for the potential predictors of the logtransformed pre- or post-shift concentrations of 1-OHP, OHPHE and OHNA in all workers and in non-smokers

only. Exposure to fumes of bitumen was assessed with the personal measurements of vapours and aerosols of bitumen. Missing measurements (N = 49, 42%) of fumes of bitumen in the group of reference workers were imputed by the median among non-exposed workers. In addition, a binary group variable (bitumen-exposed workers, reference group) was implemented into all models in order to assess potential residual eVects, e.g. exposure. The concentrations of fumes of bitumen were implemented into the post-shift model in order to estimate the confounder-adjusted regression coeYcient for a dose–response relation with PAH metabolites. The estimates of the regression coeYcients were back-transformed in order to calculate the relative change of metabolite concentrations per unit of airborne bitumen exposure. In addition, to separate pre-shift and post-shift models, a mixed model with both measurements was applied to calculate the shift eVect for the metabolites. All models were adjusted for urinary creatinine, smoking (current smokers, non-smokers), age (per 10 years) and nationality (non-German, German). An interaction term for exposure and smoking was included into the models by multiplying the exposure-group variable with the smokingstatus variable. In the mixed model with repeated measurements applied to estimate the shift increase, interaction terms between time of measurement and exposure group and between time of measurement and smoking were included. Subject was implemented as random factor, whereas all other factors were Wxed in the mixed model. The restricted maximum likelihood (REML) was used to estimate the parameters of the mixed model, and the assumed structure was compound symmetry.

Results Table 1 presents the characteristics of the exposure groups. Median age of 317 bitumen-exposed workers was 41 (range 17–63) years and 41 (range 18–64) years among 117 non-exposed workers. The reference group comprised 52.1% current smokers, while 62.1% of the exposed workers were current smokers. There was a higher fraction of German workers among the non-exposed workers (82.1% vs. 68.8%). Both groups did not diVer by anthropometric measures. Personal air monitoring yielded a median concentration of vapours and aerosols of bitumen of 3.45 (IQR 1.80–5.90) mg/m3 in exposed workers and 0.20 (0.07–0.32) mg/m3 in the reference group as background exposure. US EPA PAH compounds were determined with area sampling at 27 construction sites with mastic asphalt application and resulted in a median concentration of 1.85 (1.26–3.02) g/m3 in bitumen-exposed workers and 0.21 (0.17–0.26) g/m3 at 10 worksites of the reference group.

123

Arch Toxicol Table 1 Characteristics of bitumen-exposed workers and a group of roadside construction workers without exposure to vapours and aerosols of bitumen enrolled in the Human Bitumen Study Variable

Reference group N = 117

Bitumen-exposed workers N = 317

Age (years)

Median (range)

41 (18–64)

41 (17–63)

Current smokers

N (%)

61 (52.1)

197 (62.1)

German nationality

N (%)

96 (82.1)

218 (68.8)

Body mass index (kg/m2)

Median (interquartile range)

26.3 (24.3–29.4)

26.9 (24.3–30.4)

Median (interquartile range)

0.20 (0.07–0.32)a

Bitumen condensate standard (conversion)

Median (interquartile range)

0.29 (0.10–0.47)

a

Polycyclic aromatic hydrocarbons (ng/m3)b

Median (interquartile range)

210 (172–258)

Vapours and aerosols of bitumen (mg/m3) Mineral oil standard

a

3.45 (1.80–5.90) 5.07 (2.64–8.67) 1,847 (1,256–3,023)

Personal measurements to vapours and aerosols of bitumen (n = 68)

b

Stationary measurements of 16 US EPA priority PAH compounds and benzo[j]Xuoranthene at construction sites with (n = 27) and without (n = 10) exposure to vapours and aerosols of bitumen

Table 2 depicts the distributions of cotinine, creatinine and hydroxylated PAH metabolites by exposure group and smoking status with median and IQR in pre- and post-shift urine samples. Exposed and non-exposed smokers had similar cotinine concentrations (1,556 and 1,549 g/l, respectively). During working shift, creatinine concentrations increased, especially in exposed workers (P < 0.001). A corresponding table with creatinine-adjusted concentrations is provided as supplemental material (Table A1). In non-smoking referents, 1-OHP levels did not increase during shift (median 158 ng/l pre- vs. 133 ng/l post-shift). The corresponding pre- and postshift concentrations were 615 vs. 664 ng/l for OHPHE and 5.1 vs. 6.9 g/l for OHNA. In non-smoking bitumen-exposed workers, all metabolite levels increased during shift (median 1-OHP: 193 vs. 419 ng/l; OHPHE: 618 vs. 1,414 ng/l; OHNA: 6.8 vs. 11.5 g/l). Smokers had higher PAH metabolite concentrations than non-smokers both before and after shift. In order to explore the associations between the study variables, Spearman rank correlation coeYcients were calculated among all and non-smoking bitumen-exposed workers with concentrations in the post-shift urines. 1-OHP and OHPHE but not OHNA correlated with bitumen fume in non-smokers (1-OHP: rs = 0.23, 95% CI 0.05-0.39; OHPHE: rs = 0.28, 95% CI 0.10–0.43; OHNA: rs = 0.03, 95% CI -0.15–0.21). Figure 1 shows the associations between vapours and aerosols of bitumen and post-shift metabolite concentrations among exposed workers by smoking status. There were no correlations of individual PAH metabolites with the area measurements of PAH compounds (data not shown). These associations did not improve when PAH exposure was estimated as potential personal exposure level based on the correlation between personal and stationary bitumen measurements. A comprehensive statistical modelling of the association between bitumen exposure and PAH metabolites was

123

performed. Table 3 depicts the pre- and post-shift models of 1-OHP and OHPHE, while OHNA was dominated by smoking (data not shown). The pre-shift models identiWed current smoking but not exposure at group level as an inXuencing factor with a weaker eVect for OHPHE than for 1-OHP. Current smokers had about twofold higher 1-OHP concentrations and about 1.5-fold higher OHPHE concentrations than non-smokers before shift. In the postshift models, the bitumen concentration was additionally implemented. An increase in the airborne bitumen concentration by 1 mg/m3 resulted in an increase in 1-OHP by a factor of 1.02 (95% CI 1.00–1.04) and of OHPHE by 1.04 (95% CI 1.02–1.05). Furthermore, there was an additional group diVerence with higher concentrations in exposed workers compared with non-exposed workers. The smoking eVect was slightly stronger post- than preshift. Age and nationality were not consistently associated with the levels of PAH metabolites. A clear interaction between smoking and exposure on the metabolite concentrations was not identiWed. Creatinine was a signiWcant predictor of the metabolite concentrations and implemented in all models. Due to the strong smoking eVect, Table 4 presents the corresponding models for non-smokers only. Here, we also present the results for OHNA with higher pre-shift concentrations in mastic asphalt workers. The post-shift model of OHNA did not reveal an eVect of fumes of bitumen, but a diVerence between the exposure groups (factor 1.51, 95% CI 1.13–2.03). Comparing the models of 1-OHP and OHPHE in non-smokers with data from all workers, the associations with the airborne bitumen concentrations were similar, whereas the group diVerence was slightly smaller in non-smokers. Figure 2 depicts the shift increase for 1-OHP, OHPHE and OHNA by exposure and smoking status estimated with

Arch Toxicol Table 2 Urinary metabolites pre- and post-shift by exposure to fumes of bitumen and smoking presented with median and interquartile range Time

N (N < LOQa) Reference group

Bitumen-exposed workers

Non-smokers (N = 56)

Smokers (N = 61)

Non-smokers (N = 120)

Smokers (N = 197)

Cotinine (g/l)

Post-shift

411

2.50 (2.50–6.80)

1,549 (893–2,247)

2.50 (2.50–11.0)

1,556 (741–2,320)

Creatinine (g/l)

Pre-shift

432

1.20 (0.78–1.55)

0.94 (0.63–1.38)

1.17 (0.81–1.61)

1.32 (0.93–1.77)

434

1.12 (0.79–1.53)

1.18 (0.95–1.54)

1.49 (1.08–2.08)

1.53 (1.15–2.15)

1-Hydroxypyrene (ng/l)

Pre-shift

430 (7)

158 (70–261)

271 (144–458)

193 (94–385)

407 (202–743)

Post-shift

432 (4)

133 (86–296)

453 (237–736)

419 (216–678)

793 (432–1,519)

431

242 (115–451)

312 (169–460)

241 (154–407)

474 (268–681)

433 (1)

263 (128–533)

450 (257–734)

494 (257–748)

697 (438–1,218)

Post-shift

1-Hydroxyphenanthrene (ng/l)

Pre-shift Post-shift

2+9-Hydroxyphenanthrene (ng/l) Pre-shift 3-Hydroxyphenanthrene (ng/l) 4-Hydroxyphenanthrene (ng/l)

431

160 (74–289)

188 (111–313)

152 (98–220)

315 (183–497)

Post-shift

433

140 (85–308)

305 (211–485)

353 (193–585)

559 (344–912)

Pre-shift

431

143 (81–285)

275 (147–414)

186 (111–288)

476 (254–742)

Post-shift

433

171 (99–350)

496 (222–673)

450 (275–880)

954 (526–1,547)

28 (10–62)

50 (27–84)

Pre-shift

431 (51)

Post-shift

433 (31)

33 (7–80)

65 (40–123)

47 (22–99)

41 (18–90)

74 (39–129)

1-, 2+9-, 3-and 4-Hydroxyphenanthrene (ng/l)

Pre-shift

1-Hydroxynaphthalene (g/l)

Pre-shift

409 (66)

1.8 (1.0–5.8)

9.5 (5.6–16.7)

2.3 (1.0–4.4)

Post-shift

410 (43)

2.7 (1.0–4.8)

17.1 (6.2–27.5)

3.6 (2.1–6.9)

22.5 (12.0–36.4)

Pre-shift

411 (29)

2.5 (0.3–5.0)

13.0 (6.2–18.1)

3.3 (1.6–8.7)

23.2 (11.2–35.9)

Post-shift

412 (5)

3.3 (1.8–6.5)

19.6 (7.4–25.7)

7.0 (3.9–13.4)

34.7 (19.5–56.3)

409

5.1 (1.4–9.4)

22.9 (12.8–34.2)

6.8 (3.5–12.9)

42.9 (20.4–63.6)

410

6.9 (3.2–11.0)

39.5 (12.5–56.2)

11.5 (7.3–24.3)

58.3 (31.8–93.5)

Post-shift

2-Hydroxynaphthalene (g/l)

(1+2)-Hydroxynaphthalene (g/l) Pre-shift Post-shift

859 (488–1,243)

18 (8–39)

431

615 (314–1,018)

618 (393–1,041) 1,381 (732–1,973)

433

664 (320–1,323) 1,283 (712–2,156) 1,414 (788–2,403) 2,256 (1,367–3,788) 15.7 (8.4–25.1)

LOQ limit of quantitation

a mixed linear model that included both pre- and post-shift concentrations. Table A2 presents the adjusted eVect estimates from this model as supplemental material. During shift, the concentrations of 1-OHP increased about 1.6-fold in non-smoking and twofold in smoking mastic asphalt workers. No shift increase was observed in non-exposed non-smokers after adjustment for creatinine and other factors. Similar eVects were observed for OHPHE, although smoking showed a smaller inXuence. OHNA concentrations were again mainly predicted by smoking.

Discussion Hydroxylated metabolites of the non-carcinogenic parent compounds pyrene, phenanthrene and naphthalene are commonly applied urinary markers of internal exposure to PAH in both environmental and occupational settings, including hot bitumen applications as reviewed by NIOSH (2000). In the present cross-shift study, urinary pre- and post-shift concentrations of 1-OHP, OHPHE and OHNA were determined in 434 workers with and without exposure to fumes of bitumen at various outdoor and indoor

construction sites. PAH metabolites integrate exposure from various sources and routes. All metabolites were strongly inXuenced by smoking. OHNA concentrations were even dominated by smoking. Therefore, we focused on 1-OHP and OHPHE and analysed bitumen eVects also in non-smokers only. Before shift, 1-OHP and OHPHE concentrations were similar in exposed and non-exposed workers indicating clearance over night. Smokers had already higher pre-shift levels than non-smokers. Both 1-OHP and OHPHE concentrations increased during the shift in exposed workers. Statistical modelling revealed weak associations between the shift concentrations of bitumen fume and the post-shift concentrations of these PAH metabolites. Exposure to bitumen is associated with a wide range of compounds, and as a result of this chemical complexity, no biomarker has been developed that is speciWc for bitumen exposure. Although PAH metabolites are not speciWc enough to serve as biomarker for bitumen exposure, they can indicate accidental tar exposure. Another health concern in bitumen applications is its possible modiWcation with polymers or other additives (Vaananen et al. 2003). The mastic asphalt mixtures at the construction sites

123

Arch Toxicol 100000

Smokers Non-smokers

1-OHP [ng/L]

10000

1000

100

10

1 0.1

1

10

100

Fumes of bitumen [mg/m³]

OHPHE [ng/L]

1000

100

10

1 0.1

1

10

100

Fumes of bitumen [mg/m³]

OHNA [µg/L]

1000

100

10

1 0.1

1

10

100

Fumes of bitumen [mg/m³] Fig. 1 Associations between vapours and aerosols of bitumen and post-shift concentrations of 1-hydroxypyrene (1-OHP), the sum of 1-, 2+9-, 3- and 4-hydroxyphenanthrenes (OHPHE), and 1- and 2-hydroxynaphthalene (OHNA) in exposed workers by smoking status

enrolled in our study did not contain polymers or waste products. However, additives are increasingly used for the reduction in processing temperatures. It has been discussed whether risk estimates derived from historical cohort studies of asphalt workers might be confounded by former tar exposure (BoVetta et al. 2003; Partanen and BoVetta 1994). Coal tar and bitumen diVer by source (coal or petroleum) and consequently chemical composition. Fumes of bitumen contain much less PAH

123

compounds than coal tar emissions (Jongeneelen 2001). Although the use of coal tar in road paving or other applications was discontinued in Germany since the 1980s, tarcontaining layers can still occur. We detected accidental tar exposure by means of biomonitoring at a contaminated mastic asphalt worksite where the average 1-OHP postshift concentration was 6.61 g/g crn in comparison to a median of 0.44 g/g crn in all other bitumen-exposed workers (Raulf-Heimsoth et al. 2008). These seven tar-exposed workers were excluded from the present analysis. Depending on the various sources of exposure, PAH metabolites show a wide range of concentrations (Brandt and Watson 2003). In our study, mastic asphalt workers had a median 1-OHP post-shift concentration of 0.44 g/g crn, ranging between 0.03 and 8.24 g/g crn. Jongeneelen proposed 4.44 g/g crn (2.3 mol/mol crn) as exposure limit for 1-OHP in coke-oven workers (Jongeneelen 2001). This value is much higher than the concentrations found in mastic asphalt workers, whereas workers from settings with high PAH exposure can exceed this benchmark (Marczynski et al. 2009; Pesch et al. 2007; Rihs et al. 2005). Only three out of 317 bitumen-exposed workers were above that threshold, suggesting the possibility of accidental tar exposure. The workers in this study handled mastic asphalt for special surface applications at temperatures that are higher than in rolling asphalt for road construction. Furthermore, the construction sites of the present study included also settings with limited ventilation (e.g. basement garages) as presented by Spickenheuer et al. in this issue. This might contribute to higher concentrations of fumes of bitumen than in road paving. DiVerences in PAH emissions from rolled and mastic asphalt are less well explored. RaulfHeimsoth et al. (2011a, b)report in this issue on a paving team that worked at two consecutive worksites, Wrst with rolled asphalt and a few days later with mastic asphalt. Although personal measurements revealed much higher concentrations of bitumen fume when applying mastic asphalt, concentrations of airborne PAHs and urinary PAH metabolites did not obviously diVer by type of asphalt. This also supported by post-shift 1-OHP concentrations in nonsmoking asphalt workers from Italy (Buratti et al. 2007). Their median (0.44 g/l) was similar to the average in nonsmoking mastic asphalt workers from our study (0.42 g/l). It is noteworthy that the 1-OHP concentrations in mastic asphalt workers were on average even lower than in US pavers (McClean et al. 2004). Although hot bitumen laying results in much lower airborne PAH concentrations than coke production or tar applications, asphalt workers were exposed to higher PAH levels than the reference group or general population as shown by Breuer et al. (2011) in this issue. The metabolite distributions in mastic asphalt workers diVer from the

Arch Toxicol Table 3 Estimation of the eVects of exposure to vapours and aerosols of bitumen, current smoking and other factors on the log-transformed urinary concentrations of 1-hydroxypyrene and the sum of 1-, 2+9-, 3-, and 4-hydroxyphenanthrene Pre-shift model

1-Hydroxypyrene (ng/l) (N = 430) Exp ()

Intercept

1-, 2+9-, 3-, 4-Hydroxyphenanthrene (ng/l) (N = 431)

95% CI for Exp ()

37.74

P value

Exp ()

23.63–60.28

95% CI for Exp ()

192.40

P value

140.07–264.30

Bitumen-exposed workers

1.23

0.92–1.63

.162

1.01

0.84–1.23

.886

Creatinine (g/l)

2.54

2.21–2.91