Influence of exposure assessment and parameterization on exposure response. Aspects of epidemiologic cohort analysis using the Libby Amphibole asbestos worker cohort

Our plain-English summary of Influence of exposure assessment and parameterization on exposure response. Aspects of epidemiologic cohort analysis using the Libby Amphibole asbestos worker cohort, by T.F. Bateson, L. Kopylev (Journal of Exposure Science & Environmental Epidemiology, 2015).

The study

• Authors: T.F. Bateson, L. Kopylev
• Journal: Journal of Exposure Science & Environmental Epidemiology
• Published: 2015;25(1):12-7
• Cited by: 1 article, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Recent meta-analyses of occupational epidemiology studies identified two important exposure data quality factors in predicting summary effect measures for asbestos-associated lung cancer mortality risk: sufficiency of job history data and percent coverage of work history by measured exposures. The objective was to evaluate different exposure parameterizations suggested in the asbestos literature using the Libby, MT asbestos worker cohort and to evaluate influences of exposure measurement error caused by historically estimated exposure data on lung cancer risks. Focusing on workers hired after 1959, when job histories were well-known and occupational exposures were predominantly based on measured exposures (85% coverage), we found that cumulative exposure alone, and with allowance of exponential decay, fit lung cancer mortality data similarly. Residence-time-weighted metrics did not fit well. Compared with previous analyses based on the whole cohort of Libby workers hired after 1935, when job histories were less well-known and exposures less frequently measured (47% coverage), our analyses based on higher quality exposure data yielded an effect size as much as 3.6 times higher. Future occupational cohort studies should continue to refine retrospective exposure assessment methods, consider multiple exposure metrics, and explore new methods of maintaining statistical power while minimizing exposure measurement error.

In plain English

This study reanalyzed lung cancer deaths in the NIOSH cohort of workers at the vermiculite mine near Libby, Montana, whose ore contains Libby Amphibole asbestos, a mixture of amphibole fibers. The authors focused on 880 workers hired after 1959. For this group, job records were more complete, and the authors calculated that 85% of exposure time was covered by measured exposures. An earlier meta-analysis had attributed 47% coverage to the whole cohort. Followed through 2006, the subcohort had 32 lung cancer deaths. Using cumulative exposure lagged 10 years, the estimated lung cancer slope was 5.8 × 10^-3 per fiber/cc-year. This was higher than four earlier analyses of the larger cohort, which ranged from 1.61 to 4.2 × 10^-3. Nine cumulative-exposure measures, with or without allowance for decay over time, fit the data similarly and could not be clearly told apart. Residence-time-weighted measures did not fit. An indirect check for smoking did not indicate positive confounding. The authors concluded that excluding workers with poorer exposure data gave higher effect estimates, as expected with less measurement error. They noted that the subcohort had fewer deaths and less statistical power.

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: Reanalysis of an existing occupational cohort (the NIOSH Libby worker cohort). It used time-dependent Cox proportional hazards models in a subcohort of workers hired after 1959, compared several ways of expressing exposure, and compared results with four earlier analyses of the same cohort.
• Who was studied: The NIOSH cohort of 1,871 workers employed at operations of the vermiculite mine near Libby, Montana, USA. The analysis focused on the subcohort of 880 workers hired after 31 December 1959, with mortality follow-up through 31 December 2006. This subcohort had 32 lung cancer deaths.
• Exposure: Libby Amphibole asbestos, a mixture of amphibole fibers present in vermiculite ore, estimated from the NIOSH job exposure matrix. Exposure was expressed as cumulative exposure and as residence-time-weighted exposure, each also with exponential decay at half-lives of 5, 10, 15 or 20 years, and lagged 10, 15 or 20 years. In the subcohort, mean cumulative exposure (no lag) was 19.2 and median 3.4 fiber/cc-years.
• Main results:
  • Among the 991 workers hired before 1960, 811 had at least one job with an unknown assignment, and 706 had all department and job assignments before 1960 listed as unknown.
  • For the subcohort hired after 1959, the authors calculated that 85% of occupational exposure time was based on measured exposures. This compares with 47% attributed by Lenters et al. to the whole-cohort analysis.
  • In the subcohort hired after 1959, cumulative exposure lagged 10 years gave a lung cancer mortality slope of 5.8 × 10^-3 (0.0058) per fiber/cc-year. This was greater than the four earlier analyses of the larger cohort: Sullivan 4.2, Berman and Crump 3.96, Moolgavkar et al. 1.69, and Larson et al. 1.61 (all × 10^-3).
  • The effect size from the higher-quality exposure data was as much as 3.6 times higher than in previous whole-cohort analyses.
  • Of 30 exposure measures tested, nine cumulative-exposure-based models were statistically significant. Their fit (AIC) ranged from 358.4 to 361.5 and they could not be clearly differentiated. The residence-time-weighted models were not significant.
  • A 10-year lag of exposure best predicted lung cancer mortality compared with other lags, for all cumulative exposure measures.
  • In a sensitivity analysis, all estimated exposure intensities for 1960–1963 were reduced by 50%. This did not change the order of relative model fit.
  • In an indirect smoking check using chronic obstructive pulmonary disease deaths, the slope was negative (0.056 per fiber/cc-year below zero). The p-value was 0.102, which was not statistically significantly different from zero.
• The authors' conclusion: The authors concluded that excluding workers hired before 1960, whose exposure assessment was poorer, resulted in higher effect estimates, as would be expected when exposure data contain less measurement error. They concluded that their findings were consistent with Lenters et al. on job-history sufficiency and coverage by measured exposures.
• Limitations the authors state:
  • The subcohort had substantially fewer deaths because of its younger age distribution and shorter follow-up.
  • The range of exposure in the subcohort was substantially lower.
  • Improved accuracy came at the cost of reduced statistical power.
  • Slopes from different exposure measures cannot be directly compared.
  • Comparing model fit is empirical and does not necessarily reflect a biological cause.
  • The cohort lacked smoking data, so confounding by smoking was assessed only by an indirect method.
  • The coverage percentage could not be computed for the Larson et al. analysis.
  • It is unknown whether exposure quality has a similar impact on outcomes such as mesothelioma.

Mines named here

Asbestos mines this study is about. Each mine's page says who ran it, when it worked, and what it produced.

• Libby vermiculite mine, Montana

All asbestos mines

Topics

Medical terms in this study, explained in our Lexicon: Libby amphiboles, Epidemiology, Lung Cancer, Work History, Occupational Exposure, Cumulative Exposure.

Related studies

• Determining satisfaction with access and financial aspects of care for persons exposed to Libby amphibole asbestos: rural and national environmental policy implications — C.A. Winters, W. Hill, et al. (2011)
• Mortality in a cohort of vermiculite miners exposed to fibrous amphibole in Libby, Montana — J.C. McDonald, J. Harris, et al. (2004)
• HRCT/CT and associated spirometric effects of low Libby amphibole asbestos exposure — J.E. Lockey, K. Dunning, et al. (2015)
• A review of scientific literature examining the mining history, geology, mineralogy, and amphibole asbestos health effects of the Rainy Creek igneous complex, Libby, Montana, USA — B.R. Bandli, M.E. Gunter (2006)
• Separation and characterization of respirable amphibole fibers from Libby, Montana — J.S. Webber, D.J. Blake, et al. (2008)
• Trees as reservoirs for amphibole fibers in Libby, Montana — T.J. Ward, T. Spear, et al. (2006)

Browse more research in Journal of Exposure Science & Environmental Epidemiology, or return to medical research.