A meta-analysis of the relation between cumulative exposure to asbestos and relative risk of lung cancer

Our plain-English summary of A meta-analysis of the relation between cumulative exposure to asbestos and relative risk of lung cancer, by T.L. Lash, E.A. Crouch, L.C. Green (Occupational and Environmental Medicine, 1997).

The study

• Authors: T.L. Lash, E.A. Crouch, L.C. Green
• Journal: Occupational and Environmental Medicine
• Published: 1997;54(4):254-63
• Type of article: Meta-analysis
• Cited by: 35 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

Objectives: To obtain summary measures of the relation between cumulative exposure to asbestos and relative risk of lung cancer from published studies of exposed cohorts, and to explore the sources of heterogeneity in the dose-response coefficient with data available in these publications.

Methods: 15 cohorts in which the dose-response relation between cumulative exposure to asbestos and relative risk of lung cancer has been reported were identified. Linear dose-response models were applied, with intercepts either specific to the cohort or constrained by a random effects model; and with slopes specific to the cohort, constrained to be identical between cohorts (fixed effect), or constrained by a random effects model. Maximum likelihood techniques were used for the fitting procedures and to investigate sources of heterogeneity in the cohort specific dose-response relations.

Results: Estimates of the study specific dose-response coefficient (kappa 1.i) ranged from zero to 42 x 10(-3) ml/fibre-year (ml/f-y). Under the fixed effect model, a maximum likelihood estimate of the summary measure of the coefficient (k1) equal to 0.42 x 10(-3) (95% confidence interval (95% CI) 0.22 to 0.69 x 10(-3)) ml/f-y was obtained. Under the random effects model, implemented because there was substantial heterogeneity in the estimates of kappa 1.i and the zero dose intercepts (Ai), a maximum likelihood estimate of k1 equal to 2.6 x 10(-3) (95% CI 0.65 to 7.4 x 10(-3)) ml/f-y, and a maximum likelihood estimate of A equal to 1.36 (95% CI 1.05 to 1.76) were found. Industry category, dose measurements, tobacco habits, and standardisation procedures were identified as sources of heterogeneity.

Conclusions: The appropriate summary measure of the relation between cumulative exposure to asbestos and relative risk of lung cancer depends on the context in which the measure will be applied and the prior beliefs of those applying the measure. In most situations, the summary measure of effect obtained under the random effects model is recommended. Under this model, potency, k1, is fourfold lower than that calculated by the United States Occupational Safety and Health Administration.

In plain English

This was a meta-analysis that pooled published studies; no new data were collected. The authors found 15 cohorts of asbestos-exposed workers, reported in 22 publications, that gave lung cancer risk by level of cumulative asbestos exposure. The cohorts came from asbestos cement, textile and other manufacturing plants and from mines (chrysotile, crocidolite, and a vermiculite mine whose ore contained tremolite). The authors estimated how much the relative risk of lung cancer rose per fibre-year per millilitre of cumulative exposure. Single studies gave values from zero to 42 x 10^-3. Pooled, the value was 0.42 x 10^-3 (95% CI 0.22 to 0.69 x 10^-3) under a fixed effect model. Under a random effects model it was 2.6 x 10^-3 (95% CI 0.65 to 7.4 x 10^-3); this model allows for differences between studies, and the authors used it because those differences were substantial. Industry type explained significant variation only under the fixed effect model. Fibre type added no significant variation once industry was accounted for. The authors recommended the random effects estimate for most situations and said it was fourfold lower than the figure calculated by the US Occupational Safety and Health Administration (OSHA).

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: meta-analysis of published occupational cohort studies
• Who was studied: 15 occupational cohorts reported in 22 publications. They covered workers in asbestos cement, textile, friction products, insulation and other asbestos products plants, and miners. The cohorts were in the United States, Canada, England, Sweden, Austria, Italy and Western Australia. Studies were identified from an existing overview, a Medline search from 1966 to December 1995, and citations in the studies found.
• Exposure: Cumulative asbestos exposure in fibre-years per millilitre (other units converted with conversion factors). Fibre types across cohorts included chrysotile, crocidolite, amosite and tremolite. Cohorts were grouped by industry (mining and milling; cement and cement products; manufacturing and textile products) and by fibre type.
• Main results:
  • Study-specific estimates of the dose-response coefficient ranged from zero to 42 x 10^-3 ml/f-y.
  • Fixed effect model, all 15 cohorts: summary coefficient 0.42 x 10^-3 (95% CI 0.22 to 0.69 x 10^-3) ml/f-y.
  • Random effects model: summary coefficient 2.6 x 10^-3 (95% CI 0.65 to 7.4 x 10^-3) ml/f-y. The intercept A was 1.36 (95% CI 1.05 to 1.76); A is the fitted relative risk of lung cancer at zero exposure compared with the standard population.
  • Under the fixed effect model, industry category was a significant source of heterogeneity (P less than 0.001).
  • Fixed effect coefficient for the mining and milling cohorts: 0.3 x 10^-3 (95% CI 0.01 to 0.5 x 10^-3) ml/f-y.
  • Fixed effect coefficient for the cement products cohorts: 3.4 x 10^-3 (95% CI 0.1 to 8.8 x 10^-3) ml/f-y.
  • Fixed effect coefficient for the manufacturing and textile cohorts: 7.7 x 10^-3 (95% CI 4.7 to 12 x 10^-3) ml/f-y.
  • The authors concluded that industry type is a source of heterogeneity, but that its contribution is overstated under the fixed effect model. Under the random effects model they found insufficient evidence that industry category was a significant source (P = 0.58).
  • Under the fixed effect model, after accounting for industry type, fibre type (predominantly chrysotile or not) added no significant heterogeneity (P = 0.58).
  • Under the fixed effect model, when industry was ignored, the multiplier applied to the coefficient for cohorts exposed to predominantly chrysotile was 0.05 (95% CI 0.02 to 0.14). Under the random effects model the corresponding multiplier was 0.19 (95% CI 0.02 to 1.6), and fibre type did not contribute significantly to heterogeneity.
  • Under the random effects model, fibre type did not contribute significantly to heterogeneity, even independent of industry type.
  • The summary estimates were 24-fold lower than OSHA's figure under the fixed effect model and fourfold lower under the random effects model. OSHA's figure was 10 x 10^-3 ml/f-y, adopted in its 1986 rules from an earlier semiquantitative review. Both 95% CIs excluded 10 x 10^-3 ml/f-y.
  • In four cohorts studied at several stages of follow-up, the coefficient was lowest at the latest follow-up.
  • In one cohort with data stratified by smoking, the authors found weak evidence that the dose-response coefficient was larger among non-smokers than among smokers, and that the opposite was true for the intercept.
• The authors' conclusion: The authors concluded that the appropriate summary measure depends on the context and the prior beliefs of those applying it, and that no unique correct answer exists. In most situations they recommended the summary measure from the random effects model.
• Limitations the authors state:
  • Unpublished studies, or published studies unknown to the authors, may exist that could alter the findings, possibly toward the null.
  • Only cohorts showing a positive relation may have been broken down into exposure categories, which would bias the results away from the null. Investigators with null results may also have broken down their data to look for effects. The authors could not assess how these possibilities balance.
  • For most cohorts, the midpoint of each exposure range had to be used instead of the mean, which can artificially inflate or deflate the coefficient.
  • Equivalent cumulative exposure values in different cohorts likely reflect very different actual doses, because measurement and exposure-assignment methods differed between studies.

Topics

The National Library of Medicine indexes this study under subjects our site covers: Lung cancer.

Medical terms in this study, explained in our Lexicon: Meta-Analysis, Cumulative Exposure, Relative Risk, Lung Cancer, Dose-response relationship, confidence interval, Habit.

Related studies

• Lung cancer risk at low cumulative asbestos exposure: meta-regression of the exposure-response relationship — S. van der Bij, H. Koffijberg, et al. (2013)
• Environmental exposure to asbestos and risk of pleural mesothelioma: review and meta-analysis — V. Bourdès, P. Boffetta, et al. (2000)
• Modeling community asbestos exposure near a vermiculite processing facility: Impact of human activities on cumulative exposure — J.L. Adgate, S.J. Cho, et al. (2011)
• Non-occupational exposure to asbestos and risk of pleural mesothelioma: review and meta-analysis — G.M. Marsh, A.S. Riordan, et al. (2017)
• Relative risk of mesothelioma among railroad machinists exposed to chrysotile — T.F. Mancuso (1988)
• A simple approach for fitting linear relative rate models in SAS — D.B. Richardson (2008)

Browse more research in Occupational and Environmental Medicine, or return to medical research.