Development of an airborne fiber size specific job-exposure matrix (JEM) based on transmission electron microscopy (TEM) data, Charleston, South Carolina asbestos textile cohort.
The National Institute for Occupational Safety and Health (NIOSH) is currently extending the follow-up period for a cohort of chrysotile asbestos workers previously employed in an asbestos textile plant located in Charleston, South Carolina. This plant has many unique features that make further study and analyses important in order to provide new information concerning exposure-response relationships for asbestos related pulmonary fibrosis and lung cancer. Unique features of this plant and worker cohort include: 1) use of almost entirely chrysotile for a period of approximately 50 years, 2) availability of historical exposure measurements from the 1930's until asbestos textile production ceased in 1977, 3) availability of archived membrane filter samples for more detailed transmission electron microscopy (TEM) analyses of airborne fiber size, and 4) a cohort of sufficient size and with sufficient follow-up to allow meaningful analyses of exposure-response. Previous studies of mortality among the Charleston asbestos textile cohort have demonstrated significant excess mortality due to asbestos related fibrotic lung disease and a steep exposure response relationship for lung cancer [Dement, 1980; Dement et al., 1981; Dement et al, 1983a; Dement et al, 1983b]. Both cohort and nested case-control study designs were used to investigate exposure-relationship relationships and to control for potential confounders. Chrysotile exposure levels, expressed as fibers longer than 5 micrometer per cubic centimeter of air by phase contrast microscopy (PCM), by plant areas (department and operations), specific textile jobs, and calendar years were estimated based on historic data and were used in connection with detailed worker job histories to calculated cumulative lifetime exposures [Dement et al 1981; Dement et al, 1983]. Individual lifetime cumulative exposures were calculated and used for estimating exposure response relationships. The overall lung cancer SMR for this cohort was 1.97 (95% CI=1.69-2.28) while the risk for pneumoconiosis was 3.11 (95% CI=2.52-3.80). The risk of lung cancer was higher for white males (SMR=2.30; CI=1.88 2.79) and white females (SMR=2.75; CI=2.06 3.61). Using data for the most recent follow-up of this cohort through 1990, Stayner et al.  evaluated alternative exposure response models for lung cancer and asbestosis with Poisson regression. These analyses evaluated evidence of a threshold response. A highly significant exposure response relation was found for both lung cancer and asbestosis. The exposure response relation for lung cancer seemed to be linear on a multiplicative scale; however, the exposure response relation for asbestosis seemed to be nonlinear on a multiplicative scale. There was no significant evidence for a threshold in models for either lung cancer or asbestosis. The excess lifetime risk for white men exposed for 45 years at the recently revised OSHA standard of 0.1 fiber/ml was predicted to be about 5/1000 for lung cancer, and 2/1000 for asbestosis. A non-linear relationship between pulmonary fibrosis and chrysotile exposure in this cohort also has been suggested by pathological analyses of parenchymal lung tissues [Green et al., 1997]. The slope of the exposure response relationship observed in Charleston is among the highest seen in asbestos exposed cohorts, irrespective of fiber type or industry. Independent study of these same chrysotile asbestos textile workers has corroborated the results obtained by Dement et al. and Stayner et al. [McDonald et al, 1983b]. Subsequent nested case control analyses of this population were conducted to examine possible confounding effects from the use of mineral oil in the process of textile manufacturing [Dement, 1991; Dement et al, 1994]. These analyses concluded that mineral oil exposure was not a plausible explanation for the observed risk and confirmed the role of chrysotile exposure as the most likely cause for the increase in lung cancer and the strong exposure response pattern observed. Similar lung cancer exposure response patterns have been observed in other asbestos textiles [McDonald et al., 1983b; Peto et al, 1985]. In contrast to the risks observed for chrysotile workers in textiles, the lung cancer exposure response pattern observed among chrysotile miners and millers has been less dramatic [McDonald et al 1980], suggesting possible differences in airborne fiber exposures characteristics between mines and mills versus production plants. Prior transmission electron microscopic (TEM) analyses of airborne fibers collected in plants using chrysotile have shown reasonably large differences in the distribution of fibers by both length and diameter with airborne fibers in textile plants tending to be longer and thinner [Dement and Wallingford, 1990]. Airborne fiber size measurement data for chrysotile mining and milling demonstrate a significantly greater proportion of fibers (95 98%) less than five micrometers in length compared to data from textile plants using chrysotile [Gibbs and Hwang, 1980; Dement et al., 1994], providing some support for increased pathogenicity for fibers produced in textile operations. In addition to differences in airborne particle size by type of plant, TEM data show considerable variation in fiber size distributions among different operations within the same plant [Dement and Harris, 1979; Dement, 1980; Dement and Wallingford, 1990; Dement et al.,1994]. For these analyses, airborne fibers collected on membrane filters were analyzed by TEM using direct filter preparation methods in order to determine the bivariate distributions of airborne fibers. Due to the high magnification required to resolve the smallest diameter fibers, the maximum length of fiber that could be measured was 10 microm. While limited, these data suggest differences in particle size by plant operation that might be of biological significance. Further support for differences in the risk of lung cancer by plant operation is provided in the nested case-control analyses of Dement et al.  where spinning operations were found to have a higher risk after controlling for confounders and cumulative lifetime asbestos exposures by PCM. Limitations of these data include: 1) no analyses by jobs within plant operations, and 2) no information on the bivariate distribution of airborne fiber longer than 10 microm. The current research expands upon these studies by: 1) analyzing archived filters by TEM in order to further determine the airborne fiber sizes (length and diameter) with lengths greater than 10 microm, and 2) using these airborne fiber data to develop a TEM airborne fiber size specific job-exposure matrix (JEM). The size-specific JEM will be used in updated mortality analyses of the Charleston cohort extended with additional years of follow-up.
Textile-workers; Textiles-industry; Asbestos-fibers; Asbestos-industry; Asbestos-workers; Exposure-assessment; Respiratory-system-disorders; Pulmonary-function; Pulmonary-system-disorders; Laboratory-techniques; Lung-cancer; Mortality-rates; Risk-analysis; Pneumoconiosis; Region-4; Fibrous-dusts; Analytical-models; Statistical-analysis; Workplace-studies