Expos Trabajadores Asfalto

7/29/2019 Expos Trabajadores Asfalto

1/14

Ann. occup. Hyg., Vol. 44, No. 1, pp. 43-56, 2000 2000 British Occupational Hygiene SocietyPublished by Elsevier Science Ltd. All rights reservedP I I : S0003-4878(99)00101-5 P r i n t ^ in Great Britain0003^878/00/520.00

Statistical Modelling of the Determinants ofHistorical Exposure to Bitumen and PolycyclicAromatic Hydrocarbons Among Paving Workers1IGOR BURSTYNf, HAN S KRO MH OUTf*, TIMO KAUPPINENJ,PIRJO HEIKKILAJ and PAOLO BOFFETTAtEnvironmental and Occupational Health Group, Wageningen University, PO Box 238, 6700 AEWageningen, The Netherlands; ^Institute of Occupational Health, Topeliuksenkatu 4laA, 00250,Helsinki, Finland; Unit of Environmental Cancer Epidemiology, International Agency forResearch on Cancer, 150 cours Albert-Thomas, 69372 Lyon Cedex 08, France

Introduction: An industrial hygiene database has been constructed for the exposure assessmentin a study of cancer risk among asphalt workers.

Aim: To create models of bitumen and polycyclic aromatic hydrocarbons (PAH) exposureintensity among paving workers.Methods: Individual exposure measurements from pavers (N = 1581) were collected from 8countries. Correlation patterns between exposure measures were examined and factors affecting

exposure were identified using statistical modelling.Results: Inhalable dust appeared to be a good proxy of bitumen fume exposure. Bitumen

fume and vapour levels were not correlated. Benzo(a)pyrene level appeared to be a goodindicator of PAH exposure. All exposures steadily declined over the last 20 years. Masticlaying, re-paving, surface dressing, oil gravel paving and asphalt temperature were significantdeterminants of bitumen exposure. Coal tar use dictated PAH exposure levels.Discussion: Bitumen fume, vapour and PAH have different determinants of exposure. Forpaving workers, exposure intensity can be assessed on the basis of time period and productioncharacteristics. 2000 British Occupational Hygiene Society. Published by Elsevier ScienceLtd. All rights reserved.

Keywords: determinants of exposure; road construction; asphalt; mastic; coal tar; retrospective exposureassessment; database; multicentric study; bitumen

INTRODUCTIONThe International Agency for Research on CancerReceived 30 June 1999; infinalform 28 September 1999. /TAr>/^\ *i J L 4 j c* A . U * U A u IJ v A A A ( I ARC ) is currently engaged in a cohort study or*Author to whom correspondence should be addressed. v ' . J & & JFax- ( + 31) 317 485 278; cancer risk am on g Eu rop ean asph alt wo rkers. The

E-mail: Hans.Kromhout@ staff.eoh.wau.nl ob jec tive of the exposure assessment in the study is'The authors write on behalf of the asphalt study group. to estimate exposures of road paving workers toInstitute of Occupational Health: (Finland); Timo bit um en and coal tar. An investigation of the po-Partanen; National Hospital, University of Oslo (Norway): t e n t i a l c a n c e r r i s k d u e t o bitumen exposure is theSverre Langard; Bremen Institute for Prevention Research p r i m a r y g o & 1 o f t h g g t u d y H o w e v e r j c o a l t a r ex_and Social Medicine (Germany): Rainer Frentzel-Beyme; . . . c , .. . tT^T^^T^A. ,,- x r i ,1 r. i TT TT posure is an important conrounder, since it is cau-INSERM (France): Isabelle Stucker; Umea University . , . , , , u -t i /c A \ D * T u 1 r u w i sally associated with cancer in humans ( IARC,Hospital (Sweden): Bengt Jarvholm; Danish Working J v 'Environment Service: Ole Svane; Wageningen University 1 9 8 5 ) a n d h a s b e e n u s e d i n P a v m Processes in the(The Netherlands): Dick Heederik; Occupational Health Pa st (P ar ta ne n et al, 1994). Because polycyclic aro-and Rehabilitation Institute (Israel): Judith Shaham and matic hydrocarbons (PAH) are a 'major corn-

byguestonNovember14,2012

http://annhyg.oxfordj

ournals.org/

D

ownloadedfrom
http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/http://annhyg.oxfordjournals.org/


2/14

44 I. Burstyn et al.priori that PAH exposure is an appropriate measureof exposure to coal tar.

Previously published inhalable dust, bitumen andPAH exposure data for the asphalt industry havebeen reviewed (Claydon et al, 1984; Brandt et al,1985; Darby et al, 1986; Monarca et al, 1987;Ekstro m, 1991; Hanse n, 1991; Brand t andCordingley, 1992; CR OW , 1992; W ong et al, 1992;Lien, 1993; Machado et al, 1993; Brandt andDe Gr oot, 1994; Levin et al, 1994; Greenspan et al,1995; Hicks, 1995; Kitto et al, 1997). They revealedthat in modern road paving workers are typicallyexposed to 0.1-2 mg m ^ 3 of bitumen fume whichincludes 10-200 ng m ~ 3 of benzo(a)pyrene. It wasfurther concluded that these reports were of limitedvalue in assessing historical exposure levels in theindustry since they lack comprehensive evaluationof the determinants of exposure in road construc-tion (Burstyn et al, 2000a). However, the publishedreports could be used to create the following crudegrouping of road construction workers into groupswith similar exposure (in order of decreasing ex-posure level to both bitumen and PAHs from leftto right):

mastic laying ^ re-paving > surface dressing >hot mix paving > cold applications

Substantial non-differential misclassification of ex-posure can be expected to be present in such agrouping, and the degree of this misclassificationwould be impossible to estimate. Therefore, giventha t published re ports a ppea r to be insufficient foran accurate exposure assessment in the study, adatabase of individual exposure measurements inthe asphalt industry of participating countries(AWE database) was assembled (Burstyn et al,2000b). The analysis of the paving workers' ex-posures collected in the database is presented in thisreport .The specific goals of this investigation were: (a)to ascertain appropriate measures of exposure levelsfor use in the epidemiological study; (b) to identifyfactors influencing exposure intensity; and (c) toconstruct statistical models that can be used to inferaverages of long-term exposure in the past for agroup of road paving workers.

METHODS AND MATERIALSDescription o f the A WE database

A detailed description of the Asphalt WorkerExposure (AWE) database is given in the compa-nion paper (Burstyn et al, 2000b). However, keyfeatures of the database will also be summarizedhere. The exposure data collected into the databasecomprised measurements of exposure levels for avariety of agents among asphalt workers, plus sup-

plementary information. The supplementary infor-mation was analogous to collected data from acompany questionnaire on production character-istics in companies enrolled in the study. Thisensures that AWE data can be linked directly toother data that will be used in the exposure assess-ment. All available exposure data were collected ineight participating countries (N = 2007). The majorcontributors (70% of samples) were fourScandinavian countries with 35% of samples orig-inating from Norway. Germany contributed 20%.The remaining 10% of samples were approximatelyequally distributed among France and TheNetherlands. A small portion of the data camefrom Israel. The majority of measurements werepersonal samples (92%) with the exception ofsamples collected from German asphalt plants,which were mostly stationary (273/283). Stationarysamples were employed to determined vehicleexhaust emissions (CO and NO 2 monitors placedon the machines employed in road paving) as wellas to measure bitumen fume/vapour emissions atdifferent positions in asphalt plants and on roadpaving equipment.

The earliest collected samples originated from thelate 1960s, while the majority of samples were col-lected in the late 1970s and between 1985 and 1997.The data set was sufficiently large to permit statisti-cal modelling of the intensity of exposure to inhal-able dust, organic matter in the dust (bitumenfume), organic matter content of gaseous emissions(bitumen vapour) and PAHs, since each agent wasmeasured in 500-1000 samples. The majority ofmeasurements were made among road pavingworkers (N = 1581), and only a few came from as -phalt plants (351) and waterproofing or roofing op-erations (62). The overwhelming majority ofmeasurements have not yet been described in peer-reviewed literature. Measurements were originallycollected primarily to test compliance, and forindustrial hygiene research. Half of the studiesclaimed to have used a representative samplingstrategy.

In nine data sets included in the AWE database,up to six repeated measurements within individualswere identified. These studies were conductedbetween 1977 and 1996 in France, The Netherlands,Sweden and Norway. Exposures to dust, bitumenfume, vapour, and PAH in a variety of road pavingand asphalt mixing operations were measuredrepeatedly for individual workers. This was typi-cally achieved by monitoring exposures of a singlecrew over a number of consecutive days.

Correlation between exposure measuresPrior to statistical modelling, the correlationbetween different exposure measures has beenexplored in order to establish which exposuremeasure would be the best available proxy of agents



ournals.org/

D

ownloadedfrom


3/14

Bitumen and PA H exposure in paving w orkers 45Table 1. Definition of com pan y question naire -base d variab les used in statistical mode lling

Variable type Definition CommentsDummyTar use

Hot mixSurface dressingOil gravelEmulsionOther pavingMastic layingRecyclingoperationRecycle bitumenRecycling tarModified bitumenCategoricalCountryContinuousApplicationtemperatureYear

T if a positive answer was given to thefollowing question, '0' otherwise'Recycling tar containing asphalt ' or'Application of coal tar solution' or'Application of coal tar pitch products' or'Application of coal tar containing bitumensolution' or 'Application of coal taremulsion''Paving hot mixes''Surface dressing''Oil gravel paving''Application of emulsion''Other paving''Mast ic paving'Recycling operation; not part of companyquest ionnairea'Recycling bitumen binders''Recycling tar containing asphalt 'Polymer, scrap rubber, fly ash or othermodified bitumenCountry where survey was performedApplication temperature in C bDifference between 1997 and vear of the

In AWE data, five indicators of coal tar use werecorrelated with primary cause of coal tar use being'Application of coal tar solution'

start date of survey Maximum duration of a survey was one year

aConsidered to be a potential predictor that can be incorporated into the exposure assessment protocol.bMissing values were not uniformly distributed among job titles, therefore they were substituted with job title-specificaverages.considered. Correlation coefficients between inhal-able dust, bitumen fume and bitumen vapour ex-posure levels were computed. A total of 27 differentPAHs were determined in some samples in theAWE database. These PAHs were often determinedin both fume and vapour, resulting in up to 54different possible PAH levels for a sample.However, in some samples only benzo(a)pyrenelevels were determined. Since we were primarilyinterested in modelling a proxy for exposure to allPAHs (such as benzo(a)pyrene), the sum of ben-zo(a)pyrene levels in fume and vapour was used asa measure of benzo(a)pyrene exposure when bothwere available for a sample. We investigated thecorrelation between benzo(a)pyrene levels and bitu-men fume/vapour levels. In addressing the questionof whether benzo(a)pyrene levels are representativeof all other PAH levels, we have resorted to princi-pal component analysis. Principal component analy-sis identified groups of variables that share acommon feature. Only those groups of variableswhich contributed substantially to explaining mul-tiple correlation in the data were examined in detail.Each variable comprising a principal componentwas judged by the absolute magnitude of its eigen-vector (relative to other variables) in order to seehow strongly it was associated with the factor com-mon to all variables in a given principal com-ponent.

Building predictive models of exposureWe examined the frequency distributions of ex-

posure levels in order to determine if transform-ations prior to statistical modelling were warranted.Skew in the distributions signalled that logarithmictransformation was needed to satisfy the assump-tions of inferential statistical analysis.Variables used to represent determinants of ex-

posure were defined on the basis of the companyquestionnaires. All company questionnaire ques-tions had analogues in the AWE database. Theirdefinitions can be found in Table 1. In addition tothese variables, other variables were constructedrepresenting workers, sampling under conditionswhen highest exposures were anticipated, stationarysamples, type of sampling head used, and sorbenttype used to collect gas-phase emissions.

Multiple linear models for bitumen fume andbitumen vapour were constructed in the sequentialprocedure described below. At each step of themodel building process, all variables that were notstatistically significant (0.10 level) were removedfrom the models. Newly added variables were keptin the models if they were statistically significant,and if alterations in regression coefficients for othervariables appea red to be interpre table. A t step 1 allvariables listed in Table 1, except for time- andcountry-related variables, were forced into eachmodel. Patterns of correlation between variables in



ournals.org/

D

ownloadedfrom


4/14

46 I. Burstyn et al.

Table 1 were explored by computing correlationcoefficients, and if a correlation existed(r > 0.70), only one of the correlates was used infurther modelling. The choice of the correlatewas driven by considerations of ease in interpret-ation, should it be included in the final form ofthe model. At step 2 a variable representing timetrend was added to the models. At step 3 acountry variable was added to the models. Atstep 4 variables representing sampling strategy(worst case versus representative) and samplepositioning (area versus personal) were added tothe models. At step 5 variables representingsampling methods were introduced into themodels. At step 6 country-related variables wereremoved from the models in order to study theireffect on model fit. Negligible change in modelfit signalled that country-related variables couldbe removed without harming the predictivepower of the models. The rationale for this wasthat country-related effects are difficult to inter-pret and it was considered that small losses inmodel fit could be compensated for by anincrease in generalization of the model. Finally,at step 7 a random worker effect was introducedinto the models (Proc Mixed in SAS 6.12,restricted maximum likelihood algorithm, com-pound symmetry covariance matrix). All effectsincluded at step 6 were treated as fixed effects,and worker identity was introduced as a randomvariable. This consequently permitted the within-and between-worker components of exposurevariability to be quantified. These values areneeded for calculation of the maximum likelihoodestimates of the average of long-term exposurefor a group of individuals.

Models of PAH exposure were constructed byevaluating the statistical significance of significantpredictors of total organic matter (that is, bitumenfume and vapour) plus the variable representingtar use. The rationale for this approach was basedon the expectation that PAHs originate from bothbitumen and coal tar.The fit of the predictive models was evaluatedgraphically by plotting observed versus predictedexposure levels. Residuals were analysed by plot-ting predicted values against residual errors.Statistical analysis was performed using SAS 6.12(SAS Institute Inc.), data management tasks wereperformed using Microsoft Access 2.0 (MicrosoftCorp.) and graphs were prepared using SigmaPlotversion 4.01 (SPSS Inc.).

RESULTSExposure levels and correlation between exposuremeasuresArithmetic and geometric summary statistics forbitumen fume, bitumen vapour and benzo(a)pyr-

a 5

X! co Z-! i

R

noy

1

SD

SI Q0

0

Ra

o

1pno

>bAM

0

0

;ae

s? aoZ

002

0231

ON NO

m


5/14

Bitumen and PAH exposure in paving workers 47ene exposures are summarized in Table 2. Itis apparent that the exposures cover a wide rangeof values and tend to be higher during masticlaying.

There was a high degree of correlation betweenfume and inhalable dust levels (Fig. 1). This allowedus to construct the following regression model:(mg m~3 bitumen fume) = 0.93 x (mg m~ 3 inhalabledust), which accounted for 97% of variability inbitumen fume levels. The intercept of the abovemodel was not different from zero, and was there-fore excluded from the final form of the model.Thus, we have concluded that inhalable dustemitted in road paving consists primarily of organicparticulate matter, which we refer to as 'bitumenfume'. In all subsequent analyses, when an inhalabledust level was reported for a sample but the organicmatter content of that dust was not determined, weassumed that bitumen fume constituted 93% ofinhalable dust and used that value in statisticalmodelling.

No consistent correlation between bitumen fumeand bitumen vapour levels appeared to exist. Itsmagnitude and statistical significance varied greatlybetween countries (Table 3). Thus, we concludedthat we were justified in creating separate predictivemodels for bitumen fume and vapour.

Not all PAHs were determined in the 414samples for which series of PAHs were quantified,so missing values were replaced with averages ofreported levels for each PAH. Furthermore, foreach PAH a sum of its levels in particle andvapour phase was used as an indicator of 'total' ex-posure. The first two principal componentsaccounted for 50% of the multiple correlationbetween PAH levels in fume and vapour. The firstprincipal component appeared to be dominant (i texplained 38% of the multiple correlation). Noneof the PAHs contributed disproportionately to thefirst two principal components. In effect, wheneigenvectors indicated that a set of PAHs was con-tributing to a principal comp one nt ( > 0.20), alleigenvectors for the set of 'significant' PAHs wereapproximately equal (0.20-0.44). The set of PAHscomprising the first principal component consistedof 13 out of 27 measured PAHs. The first two prin-cipal components included 18 out of 27 monitoredPAHs. Benzo(a)pyrene was an important contribu-tor only to the second principal component andhad a high degree of correlation with it (r = 0.79,P < 0.0001, N = 414). Thus, we have concludedthat total benzo(a)pyrene is a good proxy of one ofthe principal sources of PAH exposure. In sub-sequent statistical modelling we have consideredboth the first principal component and total ben-zo(a)pyrene as proxies of PAH exposure. The ex-pression used to calculate the value of the firstprincipal component is given below (all PAH levelswere in ng m~ 3):

0.28 x (acenaphthene) + 0.27 x (acenaphthylene)+ 0.27 x (anthanthrene) + 0.26x (benzo(a)anthracene) + 0.22 x (benzo(a)fluorene)+ 0.29 x (benzo(k)fluorathene) + 0.28x (fluoranthene) + 0.30 x (fluorene) + 0.30x (phenanthrene) + 0.26 x (pyrene) + 0.23x (2-methylnaphthalene) + 0.29x (1-methylnaphthalene) + 0.30 x (biphenyl)

Benzo(a)pyrene levels had a statistically significantcorrelation with bitumen fume (r = 0.42,P < 0.0001, N = 419), but bitumen fume exposuresexplained only 18% (0.42 2) of variability in ben-zo(a)pyrene levels. This moderate correlation wasdriven by the high correlation between benzo(a)pyr-ene and bitumen fume levels during mastic laying(mastic laying: r = 0.96, P < 0.0 00 1, TV = 12;other paving: r = 0.32, P < 0.001, N = 407). Ingeneral, there was no correlation between benzo(a)-pyrene and bitumen vapour exposures (r = 0.01,P < 0.87, TV = 364). How ever, th e c orrelatio nbetween mastic vapour and benzo(a)pyrene wasvery strong (r = 0.99, P < 0.0001, N = 10). Thus,it appeared that factors other than bitumen fumeand vapour levels have an important bearing onbenzo(a)pyrene exposure levels. The first principalcomponent of PAHs was not correlated with eitherbitumen fume or vapour.

Thus, on the basis of the correlation patterns pre-sented above, we selected bitumen fume (N =1193), bitumen vapour (N = 510), total benzo(a)-pyrene (N = 487) and the weighted sum of PAHlevels comprising the first principal (N = 414) asdependent variables in subsequent statisticalmodels.Statistical distribution of exposure levels and treat-ment of non-detectable values

All four measures of exposure levels followedskewed distributions, which were normalized by log-arithmic transformations (base e) . Thus, naturallogarithms of exposure levels were used in the sub-sequent statistical modelling in order to stabilizevariance. Values below the limit of detection wereset to be equal to that limit. This will result in anupward bias in the absolute levels of bitumen fumeand vapour. Nevertheless, this should have negli-gible effect on the identification of determinants ofexposure, since non-detectable values representedonly a small proportion of bitumen fume (184/1193) and vapour (5/510) measurements. A largeproportion (321/487) of benzo(a)pyrene valuesappeared to be non-detectable, even if, in order toqualify as a non-detectable value, both fume and



ournals.org/

D

ownloadedfrom


6/14


o


7/14

Bitumen and PAH exposure in paving workers 49Table 4. Statistical models of log e (m g m b itu m en f um e), log,, (m g mbenzo(a)pyrene) bitumen vapour) and log e (ng m

Variable

Mastic layingMastic laying x worst case (indoors)bRecyclingRecycling x worst casebSurface dressingOil gravelTar useYears before 1997Application temperature in non-mastic paving(C)Area sample (yes/no)Sorbent type eSilicaCharcoalSampling head837 mm closed faceG G PInterceptPercentage variance explained by fixed effectsNumber of observationsB W S ^ ( B R O . 9 5 )wwSy (wRo.95)

Bitumen fumepa

0.881.710.891.67 N S d- 1 . 5 1 N S0.062 N SN SN A fN A

- 1 . 3 21.20- 2 . 0 9 4111930.99 (49)1.08 (59)

SE a0.220.290.250.370.28

0.008

0.150.160.11

0.20s0.221

Bitumen vapourP

0.781.70 N SNPC1.880.48 N S0.1350.0092.26- 2 . 7 4- 0 . 9 5

N AN A- 1 . 1 9 365101.16(68)1.26(81)

SE0.470.70

0.240.470.0340.0030.370.240.37

0.42

0.3110.27 j

Benzo(a)pyreneP

1.273.071.51 N S0.38- 0 . 6 51.680.107 N SN SN SN SN SN P0.91 434870.43 (13)1.71 (168)

SE0.470.840.280.200.380.280.022

0.16

0.2810.32*ap-. regression coefficient; SE: standard error of regression coefficient.bSymbol ' x ' denotes a multiplicative interaction of variables.cNot possible to estimate.dVariable is not statistically significant and therefore is not included in the model or did not improve model fit uponinclusion in the model.eReference category is XAD2 sorbent.fNot applicable.8Reference category includes: 37 mm open face cassette, 25 mm c losed face cassette (Millipore) a nd P AS 6.hBw5'y: variance of the distribution of logarithmic means of an individual 's exposures (between-worke r); w w ^ : var-iance of the distribution of logarithmic means of exposure from day to day for an individual (within-worker); 8-^0.95:ratio of 97.5th to 2.5th percentile of the betw een-w orker distrib ution of exposure s; w-Ro.95: ratio of 97.5th to 2 .5th p er-centile of the distribution of exposures within a worker.'Standard deviation of variance components' estimates generated in procedure whereby a model was evaluated 300times on random subsets of 50% of the data.

vapour models, except with respect to coal tar use.Indeed, when bitumen fume levels and coal tar usewere considered as the only predictors of benzo(a)-pyrene exposure, the following model resulted,accounting for 32% of variability in benzo(a)pyreneexposure:loge (ng irT 3 benzo(a)pyrene) =0 .1 0 x (mg m~3 total

organic fume) + 2.71 x (tar use) +1.50When tar use is 'yes', the value used is 1; when taruse is 'no' the value used is 0. In the model, tar useper se accounted for 20% of variability in benzo(a)-pyrene exposure.

Examination of the residuals of the three modelsrevealed no substantial deviations from the assump-tion of homoscedasticity.

Table 5 presents the estimates of variance com-ponents with and without correcting for the signifi-cant determinants of exposure (fixed effects). Theresults presented in Table 5 indicate that predictors

of exposure identified in our models are predomi-nantly associated with between-worker variancecomponents. It would appear that fixed effectsexplain a greater proportion of between-worker var-iance in the benzo(a)pyrene model ({2.00 0.43}/2.00 = 79% ) tha n they do in the bitumenfume ({2.27 - 0.99 J/2.27 = 56% ) and bitum envapour models ({2.51 - 1.16J/2.51 = 54%) .

DISCUSSIONComparison with the literature and interpretation ofthe observed pattern of determinants of expo sure

Organic matter content of inhalable dust emittedduring paving. Our finding of a high degree of corre-lation between inhalable dust and its organic mattercontent implies that little information is gainedabout bitumen fume exposure by analysing organicmatter content of samples compared to gravimetricanalysis. Organic matter content of asphalt dust hasbeen reported to vary considerably, especially underlow inhalable dust exposure conditions (Monarca et



ournals.org/

D

ownloadedfrom


8/14

50 I. Burstyn et al.1000

100

j?u,O


9/14

Bitumen and PAH exposure in paving workers 51Table 5. Influence of fixed effects and random effects on variance components in models of bitumen fume, bitumenvapour and benzo(a)pyrene exposure among paving workers

Model3

Worker and fixed effectsWorkerN6ke

lo ge (m ge2 bBW^y0.992.27

m " 3

1193904

bitumen fume)W W < S y

1.081.18

Dependent variablelog,, (mg m~ 3 bitumen vapour)c2 c2BWJy WWOy1.16 1.262.51 1.30510346

lo g c (ng mBwSy0.432.00

3 benzo(a)pyrene)

1.711.7448 729 2aAll models include intercept.bEstimate of logarithmic between-worker variance.cEstimate of logarithmic within-worker variance.dNumber of exposure measurements .eNumber of workers .

al , 1987; Brandt and Cordingley, 1992; CROW,1992). It is evident from Fig. 1 that the same pat-tern holds for our data: there is more variability inorganic matter content of particulate emissionswhen their overall level is low. Removing exposurelevels that were above 30 mg m ~3 from Fig. 1 hadthe predictable effect of reducing the range of vari-ables and increasing variability in the studied re-lationship: R2 decreased to 0.64. The slope of therelationship also decreased from 0.93 to 0.74. Thus,our results are in agreement with previously pub-lished reports. However, it is quite possible that theapparent change in organic matter content of inhal-able dust emitted in road paving is an artefact ofincreased uncertainty in quantifying relatively lowparticulate and organic matter levels. Even if thereis a 20% decrease in bitumen fume content fromhigh to low dust operations, the effect is likely to beunnoticeable in a statistical model of bitumen fume,since much stronger predictors of exposure are pre-sent in road paving (factors of 2-5). In fact, apply-ing an alternative relationship for translatinginhalable dust into bitumen fume levels, assumingincreasing content of organic matter with increasingparticulate matter, hardly alters the parameters ofthe bitumen fume exposure model (data notshown). The coefficient most affected was that fortime trend, but it was not significantly differentfrom the one presented in this study (95% confi-dence, two-tailed Mest). Thus, statistical modelsappear to be robust with respect to the aboveassumptions that could have been made about therelationship between inhalable dust and bitumenfume exposure. Nevertheless, since it has beensuggested that organic matter content of inhalabledust is dependent on application temperature andasphalt type (Brandt and Cordingley, 1992), thisissue should be investigated further. The efficiencyof different solvents and quantification methodsmay further confound the issue, but it was notpossible to explore it using our data.

Sources of P AHs. The PAHs that comprised thefirst principal component do not vary across differ-ent production conditions in road paving. They rep-resent an overall PAH exposure common to allroad paving workers. The identity of the underlyingsource of these PAHs is unclear. The PAHs in thesecond principal component and benzo(a)pyreneboth appear to represent tar use and PAHs in bitu-men fume. These two variables are very good pre-dictors of benzo(a)pyrene exposure levels. We wereunable to test whether the second principal com-ponent is related to changes in bitumen vapour ex-posure levels, since no bitumen vapourmeasurements were collected during the use of tar.However, a comparison of the patterns of re-gression coefficients in bitumen vapour and ben-zo(a)pyrene models (see the discussion below)indicates that changes in total benzo(a)pyrene ex-posure may mimic, to some extent, those in bitu-men vapour.Mastic laying. Bitumen fume and vapour ex-posures arising during mastic laying were 2-13times higher than those observed in associationwith hot mix paving. This corroborates findings ofother researchers, who have indicated that masticlaying results in much higher bitumen fume ex-posure levels than other paving activities (Brandt etal , 1985; Hansen, 1991; CROW, 1992). In masticlaying elevated bitumen fume and vapour exposurescan be expected, since the source of exposure (thatis, the application mixture) is closer to the exposedperson's breathing zone owing to the fact that mas-tic is habitually applied by hand. It has beenreported that up to a twofold decrease in inhalabledust exposure levels during paving can be explainedby an increase in distance between the exposed per-son's breathing zone and the source of exposure(Claydon et al, 1984). Furthermore, higher tem-peratures of mastic mixes can lead to more fuming.In fact, it was not possible to separate the effects ofapplication temperature from those of mastic layingin the models. Differences in composition andphysical properties of bitumen used in mastic pav-



ournals.org/

D

ownloadedfrom


10/14


ing may also account for this effect. 'Worst case'mastic laying corresponded to applying mastic as-phalt in indoor environments (for example a cov-ered parking lot). The apparent fivefold to sixfolddifference between indoor and outdoor exposuresobserved in our models is the same order of magni-tude as those reported by Brandt et al. (1985). Itwould also appear that the industrial hygienist whoconducted the surveys entered into the AWE data-base was able to consistently identify indoor masticlaying as a 'worst case' scenario.The association of mastic laying with a fourfoldto eightyfold increase in PAH exposure relative tohot mix paving is rather surprising, since coal tar isnot supposed to be used in the formulation of mas-tic mixes. The explanation for the observed effectmay lie with the fact that even though PAH contentof mastic emissions may be low, the overall amountof these emissions is so high that it results in PAHexposure comparable to that incurred during theuse of coal-tar-containing materials. This hypothesisis supported by: (a) the high correlation betweenlevels of mastic fume or vapour and benzo(a)pyreneexposures reported earlier; and (b) patterns of coef-ficients in statistical models. However, owing to thecontroversy about cancer risk in mastic laying andthe uncertainty about the contribution of the under-layer with coal tar to PAH exposures during masticlaying (Wong et al., 1992), this issue deservedfurther attention. If the source of PAHs in masticlaying is the under-layer with coal tar (Wong et al.,1992), then owing to the fact that coal tar is nolonger used in paving, we can assume that the highPAH exposures resulting from mastic laying are athing of the past. However, if the mechanism wehave suggested is correct, then high PAH exposures(and the associated health risks) due to mastic lay-ing may still be present, especially in indoor en-vironments.

Re-paving or in situ recycling. Re-paving or in siturecycling operations have been associated with elev-ated bitumen fume and benzo(a)pyrene exposure.This may be due to the fact that coarse particulateis generated during recycling operations in thecourse of grinding the old asphalt layer. These par-ticles may absorb bitumen fume and vapour gener-ated in heating old asphalt, resulting in elevatedbitumen fume rather than vapour exposures. It isnot clear as to what defined worst case scenarios inrecycling operations. However, there are some indi-cations that the re-paving technique in which oldasphalt is heated with a propane burner (hot in siturecycling) produced these high exposure situations(CROW, 1992). An alternative to hot in situ recy-cling is milling and recycling at ambient tempera-tures. Since bitumen vapour was not monitoredduring these operations, the effect of milling andrecycling on bitumen vapour exposure was notpossible to estimate. As in the case of mastic laying,

it would appear that industrial hygienists conduct-ing the surveys of recycling operations were able toidentify a fivefo ld increase in exposure as 'worstcase'. The association between PAH levels and recy-cling operations may be due to the correlationbetween fume and PAH concentrations. It is alsopossible that the same mechanism that leads to elev-ated fume levels during recycling operations is alsoresponsible for preferential absorption of heavierPAH, such as benzo(a)pyrene, by particles compris-ing fume. This would lead to elevated PAH contentof fume during recycling/re-paving operations.Surface dressing and oil gravel paving. On thebasis of limited data, CROW (1992) has concludedthat surface dressing is associated with elevatedfume and PAH exposure relative to hot mix paving.Our results indicate that there is no difference inbitumen fume exposure levels between surface dres-sing and hot mix paving. It is possible that smalldifferences between these two types of applicationsdo exist, but their magnitude is likely to be small(less than a factor of two), resulting in negligiblecontrasts in average fume exposure levels. Ourobserved association between surface dressing andbitumen vapour can be attributed to differences inmodes of applying bitumen to road surfaces. In sur-face dressing, a liquid bitumen solution (cut-backor emulsion) is sprayed onto the road surface, whilein other applications (with the exception of oilgravel paving) bitumen is deposited as a part ofsolid asphalt (a mixture of bitumen and mineralaggregate). It is possible that the spraying of liquidbitumen solution is more conducive to the for-mation of small droplets and vapour, which arepreferentially absorbed by sorbent tubes. It is alsolikely that bitumen used in surface dressing has ahigher proportion of volatile hydrocarbons, orig-inating either from bitumen itself or from the cut-backs which are customarily added to bitumen sol-utions used in surface dressing. The process ofvapour emission during surface dressing is governedby application temperature and vapour pressure ofthe constituents of the mixture. Given that we can-not expect all surface dressing mixtures to have thesame composition, we can expect that applicationtemperature per se will not account for the amountof vapour originating in surface dressing. The influ-ence of bitumen composition on the relationshipbetween the amount of bitumen fume emitted andtemperature has been reported (Machado et al.,1993).

Oil gravel resulted in bitumen fume exposures,which were lower than those in the reference cat-egory (which includes hot mix paving). This wasprobably the case because fume formation requiresbitumen to be heated (as in surface dressing).However, during oil gravel paving, binding materialis applied at ambient temperatures, precluding theoccurrence of evaporation and condensation pro-



ournals.org/

D

ownloadedfrom


11/14

Bitumen and PAH exposure in paving workers 53cesses that lead to fume formation. The weak posi-tive association between oil gravel paving and or-ganic vapour levels is probably a result of theapplication method, similar to surface dressing, inwhich vapour and small aerosol droplets are gener-ated during spraying of a liquid. It is also likelythat volatile organic materials not present in hotmix asphalt or surface dressing mixtures evaporateduring oil gravel paving, resulting in elevatedvapour exposure.The observed pattern of relationships betweenbenzo(a)pyrene exposure and these two applicationmethods suggests that it mirrors patterns of organicmatter levels in fume and vapour.Effects of application temperature. The depen-dence of bitumen fume, vapour and PAH exposureon application temperature has been reported byother investigators (Ekstrom, 1991; Brandt andCordingley, 1992; Lien, 1993; Machado et ai, 1993;Brandt and DeGroot, 1994; Kitto et al, 1997). Therationale for temperature dependence of the levelsof asphalt emissions is that the emissions areformed via evaporation of the organic componentsof asphalt. This evaporation process is governed, atleast in part, by the temperature of a given asphaltmix. In our data, application temperatures weremost commonly extracted from production records,but in one survey they were measured at the lo-cation of a worker. Separate analysis by the methodof assessing application temperature did not yielddiverging results. We only observed the effect of ap-plication temperature, independent of mastic laying,on bitumen vapour. Despite some authors' claimsthat application temperature is the primary predic-tor of bitumen fume and vapour exposure levels(Ekstrom, 1991; Brandt and Cordingley, 1992;Brandt and DeGroot, 1994), in the AWE data theapplication temperature alone accounts for 7-14%of variability in bitumen fume and vapour exposurelevels (Fig. 2). However, the observed slope of therelationship is of the same order of magnitude asreported in the literature.

There are several possible explanations as to whywe did not observe a very strong influence of appli-cation temperature on exposures in road paving.First, laboratory studies (Machado et al., 1993;Brandt and DeGroot, 1994; Kitto et al., 1997) donot take into account a multitude of factors thatmay and probably do influence personal exposurein the workplace. Thus, a variable can be expectedto have lower predictive power in the 'real world'than it does under controlled laboratory conditions.Second, the model developed by Brandt andDeGroot (1994) in a set of laboratory experimentshas been validated on a very small number of fieldobservations, and thus its generalization remainsuncertain. Third, two reports claiming that appli-cation temperature has a strong effect on bitumenfume and vapour exposure (Ekstrom, 1991; Lien,

1993) have not been published in peer-reviewedscientific literature, thus making it difficult to ascer-tain how the authors arrived at such a conclusion.Incidentally, the data from the last two reports areincluded in the AWE database and form part of thedata set used to build the statistical models pre-sented in this report.From the point of view of exposure assessment inan epidemiological study, it is quite fortunate thatwe do not have to rely very heavily on applicationtemperature in making inferences about exposurelevels. Our experience has indicated that collectingaccurate information about typical application tem-peratures in the past is difficult, with uncertainty onthe order of 10-20C. Thus, for example, if we hadto assign a twofold change in exposure level forevery 15C, as suggested by Ekstrom (1991), a sig-nificant misclassification of exposure would haveresulted.Even though application temperature may notaffect the amount of bitumen fume emissions, itmay influence their composition (Kitto et al., 1997).Norseth et al. (1991) found that health effects ofbitumen fume exposure were associated with appli-cation temperature. This may indicate that the toxicproperties of asphalt emission change with changein application temperature. Thus, one of the princi-pal limitations of the AWE database and our pre-dictive models becomes apparent: they do not takeinto account either bitumen composition or qualityand quantity of non-bitumen components of thebinder. Questions about bitumen composition areat present difficult to answer (CROW, 1992; Kittoet al., 1997) since bitumen is defined on the basis ofits physical properties rather than its chemical com-position. Therefore, the above-mentioned limitationis not an inherent fault of our data, but is merely areflection of the overall poor current understandingof bitumen chemistry.

Coal tar in paving mixes. The strong relationshipbetween coal tar use and PAH levels observed inour study supports our a priori assumption that anassessment of exposure to coal tar can be based onPAH measurements. Our results reveal that in thepresence of coal tar, PAH exposures in road pavingare elevated by a factor of five (that is, exp(1.68)).However, on the basis of the results of laboratoryexperiments, PAH content of tar can be assessed tobe 100 or even 1000 times higher than that of bitu-men (Lindstedt and Sollenberg, 1982; Brandt et al,1985; Darby et al., 1986; Machado et al, 1993).This apparent difference in the observed effects ofcoal tar on PAH exposure levels in road paving canbe explained by the observation that most benzo(a)-pyrene levels were non-detectable, unless coal-tar-containing materials were used in paving. Thus, theabsolute magnitude of the observed coal tar effect isdependent on the treatment of non-detectable ben-zo(a)pyrene levels in statistical analysis. On the



ournals.org/

D

ownloadedfrom


12/14


other hand, some doubts have been raised abouthow representative laboratory-generated emissionsare of those encountered in workplace (Niemeier etal., 1988).Time trend. According to the time trend observedin the AWE data, bitumen fume, bitumen vapourand benzo(a)pyrene exposures have decreased by afactor of 2-3 each decade. The possibility that thetime trend in benzo(a)pyrene levels was due to themanner in which we treated non-detectable valueswas rejected since there was no time trend in thevalues we used to replace the non-detectable ben-zo(a)pyrene exposure measurements. We have notechnical data that would explain these patterns.The role of other factors that may affect the timetrends, such as those related to social and economicconditions, is difficult to ascertain. The use of timetrend for which we have no explanation will lead tobiased estimates of exposure only if the time trendsthemselves are estimated with a bias. There is noreason to believe that our time trend estimates forthe asphalt industry are biased. For unbiased ex-posure assessment for epidemiological purposes it iscritical to explain ho w exposure levels have chan-ged, not why they have done so.

A recent investigation of world-wide time trendsin exposure data (Symanski et al., 1998a, b) hasindicated that when occupational exposuresdecrease over time, they generally do so at a rate of6-17% per year (statistically significant downwardtrends, P < 0.05). Our results are in agreement withthese trends. The limitations of the meta-analysisconducted by Symanski et al. (1998a) were the in-ability (a) to obtain individual measurements, and(b) to adjust time trends for the effects of samplingstrategy (notably worst case versus representative)and sampling methods. Our estimates of time trendsdo not surfer from these limitations, since they arebased exclusively on individual measurements (cor-rected for the effect of area sampling) and areadjusted for sampling strategy, analytical methodsand production factors. Thus, our results provideadditional evidence of the validity of the time trendsobtained by Symanski et al. (1998a, b).

Variability of bitumen and PAH exposures in pav-ing operations. Estimating an arithmetic mean ex-posure for a group of workers using multiple linearmodels presented in this report requires one to havean estimate of within-worker (day-to-day) variabil-ity of exposure (Tornero-Velez et al., 1997). Ourestimates of within-worker variance presumablyunderestimate true day-to-day variability in bitumenand PAH exposures. This arises from samplingstrategies employed in collecting repeated exposuremeasurements, in which exposures were monitoredover the duration of one project on several consecu-tive days. Thus, the collected measurements cannotrepresent a variety of exposure circumstances (thatis, projects) that paving workers experience in a

given year. Nevertheless, the within-worker logar-ithmic variance for outdoor work, such as roadpaving, was of similar magnitude (1.40) in otherdata (Kromhout et al., 1993). The reason for this ispossibly related to the fact that outdoor work isoften not very different from one day to another.Consequently, exposure levels can be expected to bedetermined to a large extent by environmental fac-tors such as wind and rain, which were notaccounted for in either data set.Our results suggest that the predictors of ex-posure present in our models are related to differ-ences in exposure between workers. Thesepredictors explain at least half of the variability inexposure concentrations between-workers. It hasbeen suggested that process-related variables influ-ence between-worker variance components(Kromhout et al., 1993). However, the grouping ofworkers solely on the basis of identified predictorvariables will not result in uniformly exposedgroups, since mean exposures for each worker inthese groups can be expected to vary over a 13-70-fold range (as assessed by B^O.95 values).Methods of measuring bitumen exposure. Areasamples of bitumen vapour were 10 times higherthan personal ones. A possible explanation for theobserved effect is that the stationary samples wereplaced in areas of high exposure (close to thesource, such as the screed of a paving machine)which are avoided by workers. Even though areasamples have been generally reported to underesti-mate personal exposures (Scheeper et al., 1995), thebias in area samples is probably highly dependenton their position. Thus, in estimating an indivi-dual's exposure level, stationary sampling should beavoided.Different sorbents have demonstrated varyingaffinities to bitumen vapour. It is not clear as towhy this is the case. Given that the effect of sorbentchoice on the measured bitumen vapour level canbe substantial (a factor of 0.06 in the case of silicagel, as compared to XAD2), this issue should bethe subject of further research.The type of sampling head used to collect bitu-men fume appears to have had a significant impacton the observed exposure levels. We have chosenthe sampling head employed to collect the majorityof samples (37 mm open face Millipore sampler) asa reference category, even though we are aware thatits collection efficiency deviates significantly fromthe toxicologically relevant inhalable dust fraction(Kenny et al., 1997). However, these deviations forsmall particles, such as those that can be expectedto form bitumen fume (< 40 p.m in aerodynam ic di-ameter), are not dramatic (Kenny et al., 1997).Nevertheless, the GGP sampler, a modified versionof the GSP sampler whose collection efficiency bet-ter adheres to the inhalable convention (Kenny etal., 1997), would probably have formed a more



ournals.org/

D

ownloadedfrom


13/14

Bitumen and PAH exposure in paving workers 55appropriate reference category. Unfortunately, theGGP sampler was used to collect only a limitednumber of samples, all of them from Germany,making the generalization of the effect of the GGPsampler somewhat questionable. Furthermore, theGGP sampler is expected to outperform the 37 mmopen-face sampler mostly at low wind speeds (lessthan 4 m s"1), but at high wind speeds both sam-plers deviate significantly from collecting inhalabledust (Kenny et al., 1997). According to the AWEdatabase, wind speeds during road paving rangedbetween 0.5 and 33 m s"1 (N = 303, AM = 3.1 ms"1, GM = 2.1m s"1, GSD = 2.3). Under suchvariable wind conditions it is very difficult to pre-dict how the collection efficiencies of samplerswould be affected. Unfortunately, wind speed wasnot recorded during sampling with the GGP, mak-ing it impossible to correct the observed 'samplereffect' for wind speed. Thus, absolute levels of bitu-men fume predicted by our model can underesti-mate levels of fume available for inhalation by afactor of approximately three. This has importantimplication for standard setting and risk assess-ment, indicating that relative performance of differ-ent samplers in outdoor environments should be thesubject of further research.

In general, characterization of exposure to bitu-men has suffered from a lack of standard methods,which have been demonstrated to collect physiologi-cally and toxicologically relevant fractions of bitu-men em issions.

our understanding of dermal exposure in the as-phalt industry.

CONCLUSIONChanges in bitumen fume, bitumen vapour andPAH exposures are each associated with somewhatdifferent factors in road paving. In assessing theintensity of exposure among road paving workers,adjustments for time period and the type of workthey have performed must be made. Time trend anda small set of production parameters can be used toexplain a substantial portion of variability in ex-posures among paving workers. Our models, despitetheir limitations, will help to make the assessmentof exposures among road paving workers in a mul-ticentric cohort study a data-driven, reproducible

process. The models have also revealed a muchmore complex pattern of determinants of exposurethan could have been ascertained, with any degreeof certainty, from the review of published reportsalone.AcknowledgementsAn IARC Special Training Awardsupported Igor Burstyn. The study was partially supportedby a grant of the European Commission, DG-XII,Biomed-2 programme (Contract No. BMH4-CT95-1100),European industrial associations, EAPA, Eurobitume andCONCAWE. Pamela Cruise edited the manuscript. WimBraun created SAS programs used in investigating thestability of models. Chava Perez provided valuable com-ments on the final manuscript.

Limitations of the predictive modelsThe AWE data for paving workers are unba-lanced. We attempted to compensate for this bychoosing the appropriate statistical model(restricted maximum likelihood algorithm) (Searle,1988). Nevertheless, we cannot alter the nature ofthe data. Somewhat different subsets of the AWEdata were used to construct predictive models forthe three agents raising, the possibility that factorswhich operated in some subsets may not have beenpresent in others. Consequently, the classical pro-blem of unbalanced data is manifested by the factthat some effects cannot be estimated. Thus, thechallenge of obtaining a data set which would allowone to conduct a comprehensive evaluation of thedeterminants of bitumen (fume and vapour) andPAH exposure remains.

There is a noticeable lack of dermal exposuremeasurements in the AWE database. This reflectsthe virtual absence of data on dermal exposure inthe European asphalt industry, with the exceptionof one study which measured dermal deposition ofPAH during road construction (Jongeneelen et al.,1988). Since the dermal route of bitumen exposurecan potentially play a role in increasing the risk ofadverse health effects among bitumen-exposed indi-viduals, more research is clearly needed to increase

REFERENCESBrandt, H. C. A. and Cordingley, N. (1992) In PersonalExposures to Bitumen Fumes and PAC During RoadPaving and in L aboratory Simulations, 5th EurasphaltCongress, pp. 166-179. EAPA, The Hague, TheNetherlands.Brandt, H. C. A. and DeGroot, P. C. (1994) A laboratoryrig for studying health aspects related to bitumen fume

exposure. Rassegna de l bitume 23, 9-15.Brandt, H. C. A., DeGroot, P. C , Molyneux, M. K. B.and Tindle, P. E. (1985) Sampling and analysis of bitu-men fumes. Annals of Occupational Hygiene 29, 27-80.Burstyn, I., Kromhout, H. and Boffetta, P. (2000a) A lit-erature review of levels and determinants of exposure topotential carcinogens and other agents in road construc-tion industry. American Industrial Hygiene AssociationJournal, In press.Burstyn, I., Kromhout, H., Cruise, P. J., Brennan, P. et al.(2000b) Designing an international industrial hygienedatabase of exposures among workers in the asphaltindustry. Annals of Occupational Hygiene 44, 57-66.Claydon, M. F., Eyres, A. R., Guelfo, G., Lentge, H.,Mitchell, J. G., Molyneux, M. K. B., Moore, J. P.,Rousseaux, G., Sanderson, J. T. and Webb, C. L. F.(1984) Review of Bitumen Fume Exposures andGuidance on Measurement. In Report no. 6/84.CONCAWE, The Hague, The Netherlands.CROW (1992) Exposure to PAC in Bituminous RoadConstruction. In Record 8. Centrum voor Regelgevingen Onderzoek in de Grond-, Water-, en Wegenbouw ende Verkeerstechniek, Ede, The Netherlands, (in Dutch).Darby, F. W., Willis, A. F. and Winchester, R. V. (1986)



ournals.org/

D

ownloadedfrom


14/14

56 I. Burstyn et al.Occupational health hazards from road constructionand sealing work. Annals of Occupational Hygiene 30 ,4 4 5 ^ 5 4 .

Ekstrom, L. G. (1991) Bitumen fumes and exposure to sol-vents in connection with the laying of asphalt: Reportfrom exposure measurements carried out in 1990,Johanneshov, Sweden. In BitTech 625. Nynas BitumenAB.Greenspan, C. A. , Moure-Eraso, R. , Wegman, D. H. andOliver, L. C. (1995) Occupational hygiene characteriz-ation of a highway construction project: a pilot study.Applied Occupational and Environmental Hygiene 10, 50-58.Hansen, E. S. (1991) Mortality of mastic asphalt workers.Scandinavian Journal of W ork and Environmental Health17, 20-24.Hicks, J. B. (1995) Asphalt industry cross-sectional ex-posure assessment study. Applied Occupational andEnvironmental Hygiene 10, 840-848.IARC (1985) Polynuclear Aromatic Compounds, Part 4,Bitumens, Coal-tars and Derived Products, Shale-oilsand Soots. In IARC Monographs on the Evaluation ofCarcinogenic Risks to Humans, Volume 35. Internat ionalAgency for Research on Cancer, Lyon.Jongeneelen, F. J., Scheepers, P. T. J., Groenendijk, A.,Van Aerts, L. A. G. J. M., Anzion, R. B. M., Bos, R.P. and Veenstra, S. J. (1988) Airborne concentrations;skin contamination; and urinary metabolite excretion ofpolycyclic aromatic hydrocarbons among pavingworkers exposed to coal tar derived road tars. AmericanJournal of the Hygiene Association 49, 600-607.

Kenny, L. C , A itken, R., Chalmers, C , Fabries , J. F. ,Gonzalez-Fernandez, E. , Kromhout , H. , Liden, G. ,Mark, D., Riediger, G. and Prodi, V. (1997) A colla-borative European study of personal inhalable aerosolsampler performance. Annals of Occupational Hygiene41, 135-153.Kit to, A. M., Pirbazari , M., Badriyha, B. N. , Ravindran,V., Tyner, R. and Synolakis, C. E. (1997) Emissions ofvolatile and semi-volatile organic compounds and par-ticulate matter from hot asphalts. EnvironmentalTechnology 18, 121-138.Kromhout , H. , Symanski , E. and Rappaport , S. M. (1993)A comprehensive evaluation of within- and between-worker components of occupational exposure to chemi-cal agents. Annals of Occupational Hygiene 37, 253-270.Levin, J. O., Andersson, K. and Hallgren, C. (1994)Exposure to low molecular polyamines during road pav-ing. Annals of Occupational Hygiene 38, 257-264.Lien, J. (1993) The Asphalt Project 1991-92: Survey ofOccupational Exposures of Asphalt Workers (Englishsummary). Statens vegvesen, Vegdirektoratet, Oslo.Lindstedt, G. and Sollenberg, J. (1982) Polycyclic aromatichydrocarbons in the occupational environment: with

special reference to benzo(a)pyrene measurements inSwedish industry. Scandinavian Journal of Work andEnvironmental Health 8, 1-19.Ma cha do, M. L., Beatty, P. W., Fetzer, J. C , Glickman,A. H. and McGinnis, E. L. (1993) Evaluation of the re-lationship between PAH content and mutagenic activityof fumes from roofing and paving asphalts and coal tarpitch. Fundamentals of Applied Toxicology 21, 492-499.Monarca, S., Pasquini, R., Scassellati, S. G., Savino, A.,Bauleo, F. A. and Angeli, G. (1987) Environmentalmonitoring of mutagenic/carcinogenic hazards duringroad paving operations with bitumens. InternationalArchives of Occupational and Environmental Health 59 ,393-402.Niemeier, R. W., Thayer, P. S., Menzies, K. T., vonThuna, P., Moss, C. E. and Burg, J. (1988) A compari-son of skin carcinogenicity of condensed roofing as-phalts and coal tar pitch fumes. In PolynuclearAromatic Hydrocarbons: A Decade of Progress, eds M.Cooke and A. J. Dennis, Tenth InternationalSymposium, pp. 609-647. Battelle Press, Columbus, OH.

Norseth, T., Waage, J. and Dale, I. (1991) Acute effectsand exposure to organic compounds in road mainten-ance workers exposed to asphalt. American Journal ofIndustrial Medicine 20, 737-744.Partanen, T., Heikkila, P. and Boffetta, P. (1994)Feasibility Phase of a Study on Cancer Risk amongWorkers in the European Asphalt Industry. In Internalreport 94/005. International Agency for Research onCancer, Lyon.Scheeper, P., Kromhout, H. and Boleij, J. S. M. (1995)Wood-dust exposure during wood-working processes.Annals of Occupational Hygiene 39, 141-154.Searle, S. R. (1988) Mixed models and unbalanced data:wherefrom, whereat and whereto? Communications inStatistical-Theory Methods 17, 935-968.Symanski, E., Kupper, L. L., Hertz-Picciotto, I. andRappaport, S. M. (1998a) Comprehensive evaluation oflong term trends in occupational exposure: part 2.Predictive models for declining exposures. Occupationaland Environmental Medicine 55, 310-316.Symanski, E., Kupper, L. L. and Rappaport, S. M.(1998b) Comprehensive evaluation of long term trendsin occupational exposure: part 1. Description of thedatabase. Occupational and Environmental Medicine 55,300-309.Tornero-Velez, R., Symanski, E., Kromhout, H., Yu, R.C. and Rappaport, S. M. (1997) Compliance versus riskin assessing occupational exposures. Risk Analysis 17,279-292.Wong, O., Bailey, W. J. and Amsel, J. (1992) Cancer mor-tality and incidence in mastic asphalt workers: letter tothe editor and author's reply. Scandinavian Journal ofWork and Environmental Health 18 , 133-141.



ournals.org/

D

ownloadedfrom

Expos Trabajadores Asfalto

Documents

Transcript of Expos Trabajadores Asfalto