Estimating exposures in the asphalt industry for an international epidemiological cohort study of cancer risk

Our plain-English summary of Estimating exposures in the asphalt industry for an international epidemiological cohort study of cancer risk, by I. Burstyn, P. Boffetta, T. Kauppinen, et al (American Journal of Industrial Medicine, 2003).

The study

• Authors: I. Burstyn, P. Boffetta, T. Kauppinen, P. Heikkilä, O. Svane, T. Partanen, I. Stücker, R. Frentzel-Beyme, W. Ahrens, H. Merzenich, D. Heederik, M. Hooiveld, S. Langård, B.G. Randem, B. Järvholm, I. Bergdahl, J. Shaham, J. Ribak, H. Kromhout
• Journal: American Journal of Industrial Medicine
• Published: 2003;43(1):3-17
• Cited by: 41 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Background: An exposure matrix (EM) for known and suspected carcinogens was required for a multicenter international cohort study of cancer risk and bitumen among asphalt workers.

Methods: Production characteristics in companies enrolled in the study were ascertained through use of a company questionnaire (CQ). Exposures to coal tar, bitumen fume, organic vapor, polycyclic aromatic hydrocarbons, diesel fume, silica, and asbestos were assessed semi-quantitatively using information from CQs, expert judgment, and statistical models. Exposures of road paving workers to bitumen fume, organic vapor, and benzo(a)pyrene were estimated quantitatively by applying regression models, based on monitoring data, to exposure scenarios identified by the CQs.

Results: Exposures estimates were derived for 217 companies enrolled in the cohort, plus the Swedish asphalt paving industry in general. Most companies were engaged in road paving and asphalt mixing, but some also participated in general construction and roofing. Coal tar use was most common in Denmark and The Netherlands, but the practice is now obsolete. Quantitative estimates of exposure to bitumen fume, organic vapor, and benzo(a)pyrene for pavers, and semi-quantitative estimates of exposure to these agents among all subjects were strongly correlated. Semi-quantitative estimates of exposure to bitumen fume and coal tar exposures were only moderately correlated. EM assessed non-monotonic historical decrease in exposures to all agents assessed except silica and diesel exhaust.

Conclusions: We produced a data-driven EM using methodology that can be adapted for other multicenter studies.

Copyright 2003 Wiley-Liss, Inc.

In plain English

This article described how the authors built an exposure matrix for an International Agency for Research on Cancer cohort study of cancer among asphalt workers from eight countries. It did not report cancer results. Using company questionnaires, a database of 2,007 exposure measurements, statistical models and expert judgment, the authors estimated exposures for 217 companies plus the Swedish asphalt paving industry in general. Asbestos, with no fiber type named, was one of seven agents assessed, and only semi-quantitatively. Each asbestos score began as zero or one for reported absence or presence of contact with source material, then was multiplied by the fraction of work time the material was used and adjusted for job class and shift length. The matrix showed declines over time for most agents. Coal tar exposure fell by a factor of 200 from before 1960 to 1990–1996, while silica and diesel exhaust stayed virtually unchanged. The rank correlation between asbestos and coal tar scores was 0.30. The authors concluded that their method could be adapted for other multicenter studies. They noted that questionnaire answers could not be standardized across companies and countries.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Exposure assessment methods study: development of a job-exposure matrix (country, company, job class, time period, agent) for a multicenter international cohort study; no health outcomes reported
• Who was studied: 217 companies enrolled in the IARC asphalt worker cohort (Denmark, Finland, France, Germany, The Netherlands, Norway, Sweden, Israel), plus the Swedish asphalt paving industry in general; exposure database of 2,007 measurements collected as of February 1999, mostly from the late 1970s and 1985 to 1997
• Exposure: Coal tar, bitumen fume, organic vapor, polycyclic aromatic hydrocarbons, diesel exhaust, silica and asbestos (inhalable asbestos fibers, no fiber type named). Asbestos was assessed semi-quantitatively only. An initial intensity of zero or one was set from questionnaire-reported absence or presence of contact with source materials. It was then multiplied by the fraction of work time the material was used and adjusted between job classes and for work-shift duration. Building construction workers were generally assigned asbestos intensities twice those in ground construction. Quantitative estimates were made only for bitumen fume, organic vapor and benzo(a)pyrene among road pavers.
• Main results:
  • Among firms enrolled in the cohort, 51% paved and 94% manufactured asphalt; some were also in ground construction (27%), building construction (6%) and waterproofing/roofing (6%)
  • Waterproofing and roofing were assessed on average to have the highest bitumen fume, organic vapor and PAH exposures, exceeding those in paving by a factor of 1.5 to 3
  • Across the matrix, the steepest decline over time was in tar exposure: a factor of 200 from before 1960 to 1990–1996
  • Assessed exposures to silica and diesel exhaust remained virtually unchanged over time
  • Among pavers, quantitative estimates were strongly correlated with one another, with rank correlations from 0.78 (bitumen fume and organic vapor) to 0.93 (bitumen fume and benzo(a)pyrene)
  • The rank correlation between semi-quantitative asbestos scores and coal tar scores was 0.30 (4,390 pairs of matrix cells)
  • Replacement values for missing data were used for 2.7% of exposed person-years for asbestos (also 2.7% for bitumen fume and for organic vapor)
• The authors' conclusion: The authors concluded that they produced a data-driven exposure matrix using a methodology that can be adapted for other multicenter studies. They also concluded that health risks of bitumen fume, organic vapor and PAH could not be distinguished from one another, while adjustment for coal tar should be possible.
• Limitations the authors state:
  • Questionnaires were filled out by individuals of varying experience and knowledge, so responses could not be standardized among companies and countries
  • Sparse monitoring data meant some semi-quantitative assessments rested exclusively on professional judgment of occupational hygienists
  • Quantitative assessment was possible only for paving workers
  • Benzo(a)pyrene models underestimated exposures overall by 51%
  • Questionnaire information could not be checked against production records

Topics

Medical terms in this study, explained in our Lexicon: Cohort Study.

Related studies

• A case-control study of lung cancer nested in a cohort of European asphalt workers — A. Olsson, H. Kromhout, et al. (2010)
• Simulation tests to assess occupational exposure to airborne asbestos from artificially weathered asphalt-based roofing products — P. Sheehan, F. Mowat, et al. (2010)
• A case-control study of asphalt and tar exposure and lung cancer in minorities — M.D. McClean, K.T. Kelsey, et al. (2011)
• A review of changes in composition of hot mix asphalt in the United States — D.J. Mundt, K.M. Marano, et al. (2009)
• Retrospective exposure assessment for carcinogenic agents in bitumen waterproofing industry in Finland and denmark — P. Anttila, P. Heikkilä, et al. (2009)
• Carcinoma of the pharynx and tonsils in an occupational cohort of asphalt workers — F. Zanardi, R. Salvarani, et al. (2013)

Browse more research in American Journal of Industrial Medicine, or return to medical research.