Performance of two general job-exposure matrices in a study of lung cancer morbidity in the Zutphen cohort

Our plain-English summary of Performance of two general job-exposure matrices in a study of lung cancer morbidity in the Zutphen cohort, by H. Kromhout, D. Heederik, L.M. Dalderup, et al (American Journal of Epidemiology, 1992).

The study

• Authors: H. Kromhout, D. Heederik, L.M. Dalderup, D. Kromhout
• Journal: American Journal of Epidemiology
• Published: 1992;136(6):698-711
• Cited by: 55 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Data from a general population cohort of 878 men from the town of Zutphen, The Netherlands, were used to evaluate the performance of two general job-exposure matrices. Exposures generated by the job-exposure matrices on the basis of job histories were compared. The validity of those exposures was measured against exposures reported by the participants in 1977/1978. The performance of the different exposure measures was assessed in proportional hazards analyses of lung cancer morbidity incidence. The two general job-exposure matrices generally disagreed with regard to exposure classification because of differences in exposure assessment and level of detail of the job axis. When compared with self-reported exposures, the sensitivity of both job-exposure matrices was low (on average, below 0.51), while the specificity was generally high (on average, above 0.90). Self-reported exposures to asbestos, pesticides, and welding fumes showed elevated risk ratios for lung cancer which were absent for exposures generated by the two job-exposure matrices. Thus, a population-specific job-exposure matrix is proposed as an alternative to general job-exposure matrices developed elsewhere. Such a matrix can be constructed from the results of in-depth interviews of a job-stratified sample of cohort members. Sound validation and documentation of exposure assessment methods used in job-exposure matrices are recommended.

In plain English

This study used data on 878 men from the town of Zutphen, the Netherlands, followed from 1960 to 1985. It tested two general job-exposure matrices, the Harvard matrix and the British Medical Research Council (MRC) matrix. These are tools that assign likely workplace exposures from job titles. The authors compared the two matrices with each other and with exposures the men themselves reported in 1977/1978, including asbestos. The two matrices mostly disagreed. Compared with the self-reports, the matrices had low sensitivity. When only high exposure counted, average sensitivity was 0.27 for the MRC matrix and 0.20 for the Harvard matrix. When any exposure counted, it was 0.51 and 0.33. Among 564 men followed from 1978 to 1985, the lung cancer hazard ratio for self-reported asbestos exposure was 2.14 (95 percent confidence interval 0.29-15.70; 8 men exposed). The four matrix-based asbestos estimates ranged from 1.00 (0.24-4.20) to 1.37 (0.33-5.77), and all these confidence intervals included 1. The authors noted wide confidence intervals because there were few lung cancer cases and follow-up was limited. They suggested that a population-specific matrix, built from interviews with cohort members, might be an alternative.

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: Cohort study evaluating two general job-exposure matrices (Harvard and Medical Research Council) against self-reported exposure, with proportional hazards analyses of lung cancer incidence adjusted for smoking and age
• Who was studied: Men from Zutphen, the Netherlands, born 1900-1919, drawn from a random sample of 1,088 invited men, of whom 878 were examined. They were followed 1960-1985. 603 men had 1977/1978 exposure questionnaires. 564 men (31 lung cancers) were in the 7-year analysis and 856 men (67 lung cancers) were in the 25-year analysis.
• Exposure: Self-reported exposure to 27 chemicals or groups of agents (including asbestos, pesticides, welding fumes and dust), compared with exposures assigned from job histories by the Harvard and MRC job-exposure matrices. Matrix exposures were classified leniently (any exposure) or strictly (high exposure only). Fiber types were not specified.
• Main results:
  • Agreement between the two matrices was poor for most agents (Cohen's kappa below 0.40). For asbestos, kappa was 0.24 (95% CI 0.18-0.30) with the lenient classification and 0.04 (-0.02-0.10) with the strict classification.
  • Compared with self-reports and counting only high exposure, average sensitivity was 0.27 for the MRC matrix and 0.20 for the Harvard matrix. Counting any exposure, average sensitivity rose to 0.51 and 0.33. Average specificity fell from 0.98 to 0.90 (MRC) and from 0.97 to 0.94 (Harvard).
  • For asbestos specifically, the Harvard matrix had a sensitivity of 0.33 and a specificity of 0.88 compared with self-reported exposure.
  • Among 564 men followed 1978-1985, the hazard ratio for lung cancer with self-reported asbestos exposure was 2.14 (95% CI 0.29-15.70; 8 men exposed).
  • Matrix-based asbestos hazard ratios in the same 7-year analysis were lower than the self-reported one: MRC lenient 1.36 (0.61-3.03; 115 men), MRC strict 1.37 (0.33-5.77; 25 men), Harvard lenient 1.21 (0.49-2.96; 92 men), Harvard strict 1.00 (0.24-4.20; 35 men). The authors wrote that the MRC matrix underestimated the hazard ratios for asbestos, pesticides and welding fumes compared with self-reports.
  • In the 7-year analysis, self-reported pesticide exposure had a lung cancer hazard ratio of 4.44 (95% CI 1.05-18.74; 10 men exposed).
  • In the 25-year analysis of 856 men, asbestos exposure assigned by the population-specific matrix had a hazard ratio of 1.40 (95% CI 0.44-4.48; 32 men). The comparable values were 1.00 (0.51-1.97) for MRC lenient and 0.75 (0.34-1.64) for Harvard lenient.
  • In the 25-year analysis, the population-specific matrix gave significantly elevated hazard ratios for dust 1.66 (1.00-2.75), soldering fumes 2.24 (1.17-4.29) and welding fumes 1.93 (1.05-3.55). These elevations were absent with the two general matrices.
• The authors' conclusion: The authors concluded that a population-specific job-exposure matrix, built from interviews with a sample of cohort members, might be an alternative to general matrices developed elsewhere for general population cohort studies.
• Limitations the authors state:
  • Hazard ratios had large confidence intervals because there were few lung cancer cases and follow-up was limited, which made interpretation difficult.
  • The validity of the self-reported exposures used as the 'gold standard' could not be assessed, because accurate quantitative exposure data were lacking.
  • The general matrices were not designed for use in the Netherlands.
  • The Harvard matrix's 'unknown level' codes were applied differently than in several other published studies that used it.
  • The 25-year analysis could use only each man's job held in 1960.
  • The performance of the population-specific matrix depends on the criterion used to assign exposure to a job (here, 10 percent of job holders reporting exposure).

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: Lung Cancer.

Related studies

• Performance of biomarkers SMRP, CA125, and CYFRA 21-1 as potential tumor markers for malignant mesothelioma and lung cancer in a cohort of workers formerly exposed to asbestos — M. Gube, D. Taeger, et al. (2011)
• Risk of stomach cancer associated with 12 workplace hazards: analysis of death certificates from 24 states of the United States with the aid of job exposure matrices — P. Cocco, M.H. Ward, et al. (1999)
• The asbestos-asbestosis exposure-response relationship: a cohort study of the general working population — I.B. Iversen, J.M. Vestergaard, et al. (2024)
• Retrospective assessment of asbestos exposure--I. Case-control analysis in a study of lung cancer: efficiency of job-specific questionnaires and job exposure matrices — W. Ahrens, K.H. Jöckel, et al. (1993)
• Comorbidity, postoperative morbidity and survival in patients undergoing radical surgery for malignant pleural mesothelioma — M. Infante, E. Morenghi, et al. (2016)
• Asbestos: toxicology and risk assessment for the general population in The Netherlands — G.K. Montizaan, A.G. Knaap, et al. (1989)

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