* Become familiar with the characteristics and health impact of exposure to Libby vermiculite and Libby amphibole asbestos (LAA), and the previous findings in a cohort of LAA-exposed workers.
* Summarize the follow-up findings in LAA-exposed workers and the association between estimated cumulative fiber exposure (CFE) and the findings on chest imaging studies and spirometry.
* Discuss the implications for understanding the health impact of commercial and noncommercial asbestos exposure.
Vermiculite ore mined near Libby, Montana, in the United States supplied nearly 80% of the vermiculite used worldwide from the 1920s to 1990.1 The US Environmental Protection Agency estimated that in the United States alone, 13 million people were exposed living near vermiculite exfoliation (expanding) facilities and 106 million people were exposed to consumer products containing vermiculite, including home insulation.2 This widespread use of Libby vermiculite (LV) makes it important to understand the potential effect on human health of low-level exposure to the amphibole fibers within the ore.
The ore from Libby contains amphibole fibers including winchite, richterite, and tremolite and is identified as Libby amphibole asbestos (LAA).3 The potential health impact of exposure first appeared in the medical literature based on our 1980 study of workers employed at a plant expanding LV in Marysville, Ohio.4 A 2004 follow-up study demonstrated an increased prevalence of pleural changes on chest radiographs from 2.0% in 1980 to 28.7%.5 This included 15% prevalence of pleural changes in workers with a lifetime cumulative fiber exposure (CFE) of 2.2 or less fiber-year/cm3. A subsequent mortality study identified three workers with malignant mesothelioma.6 This morbidity study examines the 1980 cohort 30 years after the last production use of LV at the plant. It includes subjects with high-resolution or conventional chest computed tomography (HRCT/CT) and/or chest radiography and spirometry. The primary study objectives were to evaluate the relationship between low working lifetime CFE and HRCT/CT findings and assess the impact of pleural and parenchymal changes on lung function using a more comprehensive job exposure matrix (JEM). A secondary sensitivity analysis pooled the radiographic findings from all sources including chest radiographs.
Living members (n = 431) of the original 1980 cohort (n = 513) were eligible to participate. Participants signed an informed consent approved by the University of Cincinnati Institutional Review Board. An occupational and pulmonary questionnaire was administered by telephone. Chest HRCTs and posteroanterior chest radiographs were obtained at facilities convenient for each participant. High-resolution computed tomographies were performed in the prone position to permit differentiation of mild posterior basilar fibrosis from dependent atelectasis, and 1-mm slices were taken at 15-mm increments from the costophrenic angle toward the apices, not exceeding 15 axial slices.7 For those not participating in the 2010 study, attempts were made to collect their previous HRCTs/CTs and/or chest radiographs from 2003 to 2009.
Images were masked of identifiers and interpreted independently by experienced chest radiologists (R.S., C.M., C.P.) qualified as National Institute for Occupational Safety and Health B-readers. Agreement by two or more readers (the median reading) was used for all normal/abnormal determinations.
For chest HRCT/CT scans, localized pleural thickening (LPT) was defined as well-circumscribed region(s) of pleural thickening with mesa-like appearance not involving the apices or costophrenic angle. Diffuse pleural thickening (DPT) was defined as homogeneous pleural thickening with tapering margins crossing multiple rib interspaces and included involvement of the costophrenic angle. Parenchymal changes (interstitial fibrosis) were characterized as subpleural irregular opacities and irregular interface of the lung with the pleural surfaces with or without traction bronchiectasis or honeycomb lung.7 Subpleural curvilinear lines and parenchymal bands in isolation were not considered findings of parenchymal changes.8
Chest radiographs were evaluated using the 2000 International Labor Office International Classification of Radiographs of Pneumoconiosis.9 Pleural changes included both LPT and DPT with or without calcification. Localized pleural thickening was defined as unilateral or bilateral focal pleural and/or diaphragmatic density (plaque) not associated with costophrenic angle blunting. Diffuse pleural thickening was defined as unilateral or bilateral diffuse pleural soft tissue density including costophrenic angle involvement greater than or equal to the 1/1 t/t International Labor Office standard radiograph. Parenchymal changes were defined as irregular opacities, profusion category of 1/0 or greater.
Spirometry was conducted at a single hospital and only for those living within 50 miles of the plant. Testing was conducted by experienced respiratory therapists using a Viasys Vmax test system (Yorba Linda, CA). Before testing, a study team member (R.M.) conducted an on-site visit to evaluate testing equipment/procedures to comply with 2005 ATS-ERS Guidelines.10 Shortly after data collection began, a quality assurance review was conducted to evaluate compliance with study-specific procedures.
Estimates of exposure by job at the facility from first vermiculite usage in 1957 through 1980 were calculated from a recently updated JEM.11 Briefly, the update utilized 899 airborne fiber measurements and improved information on overtime and the history of vermiculite ore sources. Initially, ore was received from South Carolina, then Libby, Montana, first beginning in 1959, Palabora, South Africa, in 1970, and Louisa County, Virginia, in 1979. Workplace exposure samples were analyzed using phase contrast microscopy, with the result reported as fiber/cm3. Particles greater than 5 μ in length and smaller than 3 μ in diameter with an aspect ratio of 3:1 or more were counted, without identification of chemical composition or crystalline structure. The JEM included estimates by job activity and time, adjusted for ore source and overtime. Minor changes to the exposure assessment were made subsequent to the Environmental Protection Agency Toxicological Review of Libby Amphibole Asbestos.12
Detailed job histories from the original cohort study4 were used to calculate CFE from 1957 through 1980 when the use of LV was discontinued. Duration of work in a job was multiplied by the time-specific exposure estimate in the JEM and summed across all jobs. For those participants still employed at the plant after 1980, job histories were updated using the occupational history collected in 2004 or 2010 to 2011. For workers hired prior to 1957, employment duration calculations began on January 1, 1957, to coincide with first vermiculite usage in the facility. Individual CFE was calculated from date of hire to 1980 for the primary CFE analysis and for comparison to the 1980 and 2004 studies. Vermiculite ore from Africa, Virginia, and South Carolina contains less than 1% asbestiform minerals compared with up to 26% for LV ore.12–14 To account for this, a second CFE value was calculated from date of hire to last date of employment or May 1, 2001, when vermiculite was no longer used at the facility, whichever date was earlier.15
Generalized estimating equations evaluated the relationships among CFE, duration of employment, and risks of pleural and parenchymal changes. Primary models were constructed for HRCT/CT participants using CFE. Generalized estimating equations models analyzed the relative risk ratio (RR) since outcome events were not rare (greater than 10%). Parameter values comparing each CFE or duration category to the reference were estimated from the regression model.
Using backward elimination for the pleural models, age, body mass index (BMI), and smoking were not significant and removal did not affect the relationship between CFE and pleural change, so these were excluded from final models. Age and smoking were included in the parenchymal analyses because these covariates have historically been associated with parenchymal changes. The HRCT/CT pleural and parenchymal log-binomial models failed to converge when including those with commercial asbestos exposure (yes/no). Penalized logistic models determined that potential commercial asbestos exposure was not significantly associated with HRCT/CT changes. Nonetheless, repeat analyses were conducted for participants reporting no commercial asbestos exposure. Because industrial hygiene measurements began in 1972, pleural change models were repeated for workers hired on or after 1972 (n = 99) but not for parenchymal models because there were only two cases (CFE 0.26 and 0.67 fiber-year/cm3). A sensitivity analysis was performed for the pleural model by adding subjects with only chest radiographs to the HRCT/CT group to determine the robustness of the findings.
General linear regression models investigated the relationship between HRCT/CT findings and percent predicted (already adjusted for sex, race, height, and age) forced vital capacity (FVC), forced expiratory volume in the first second of expiration (FEV1), and FEV1/FVC.16 Covariates included pack years (log transformed) and BMI (less than 30; 30 or more). Cumulative fiber exposure was not included because it was a predictor of pleural changes. Analyses were conducted for the following three subgroups: LPT only; LPT and/or DPT without parenchymal changes; and DPT and/or parenchymal changes. Participants with surgical changes of lobectomy or coronary artery bypass graft surgery on HRCT/CT or acute cardiopulmonary findings or significant neuromuscular morbidities were excluded.17 Those participants who had HRCT/CT findings of rounded atelectasis were included in the DPT category.18 These analyses were repeated adjusting for HRCT/CT evidence of emphysema and/or bronchiectasis (yes/no). SAS (Version 9.3, Cary, NC) was used for all analyses.
From the 1980 cohort (n = 513), 82 (16.0%) were deceased by October 15, 2004, and 125 (24.4%) did not participate in either the 2004 or 2010 follow-up studies. Of the remaining 306 participants (59.6%), 175 provided HRCTs, 16 conventional CTs, and 115 chest radiographs dating from 2003 to 2011. Of the 16 conventional CT scans made available from participants and evaluated in the HRCT/CT analysis, seven were normal (four of seven with spirometry), three demonstrated DPT (two of three with spirometry), two demonstrated parenchymal change with LPT and/or DPT (two of two with spirometry), and four demonstrated LPT only changes (one of four with spirometry). The HRCT and conventional CT scan results were combined for the analyses.
Pleural and Parenchymal Changes by HRCT/CT
Demographic and exposure characteristics of the 191 participants with HRCT/CT are shown in Table 1. Mean CFE and employment were substantially greater in participants with HRCT/CT findings. Of these 191 participants, 101 (52.9%) and 25 (13.1%) had pleural and parenchymal changes, respectively. Of the 101 with pleural changes, there were 71 (70.3%) with LPT only, 10 (9.9%) with LPT and DPT, 15 (14.8%) with LPT and parenchymal changes, and 5 (5.0%) with LPT, DPT, and parenchymal changes (data not shown). Of the 25 with parenchymal changes, 5 (20%) had parenchymal changes only, and pleural changes in the remaining 20 were predominantly bilateral (90%). Thus, a total of 106 (55.5%) had HRCT/CT abnormalities. For those with pleural changes, 79% (80/101) were bilateral, and 84% (21/25) with parenchymal changes (all involved lower lung fields) were bilateral.
Table 2 shows an increasing prevalence of pleural changes with increasing CFE and employment through September 1, 1980. Cumulative fiber exposure categories of 0.15 fiber-year/cm3 or more had higher proportions of participants with pleural changes across all exposure categories, with significant RR ranging from 5.0 to 10.4. Results were similar for participants reporting no commercial asbestos exposure. Again, when CFE was calculated to May 1, 2001, findings were similar for all participants as well as for those reporting no commercial asbestos exposure (data not shown).
For the 99 participants hired on or after 1972, there was an increased prevalence of pleural changes with increased CFE and employment through September 1, 1980. Cumulative fiber exposure categories of 0.15 fiber-year/cm3 or more were associated with significant RR ranging from 4.1 to 10.2 (Table 3). For participants reporting no commercial asbestos exposure, RR were 3.4 (95% confidence interval [CI], 1.2 to 9.9), 2.9 (95% CI, 0.9 to 10.0), and 8.4 (95% CI, 3.2 to 22.0) for the sequential nonbaseline exposure categories. For CFE computed to May 1, 2001, RRs were significantly increased for all exposure categories compared with the reference category for all participants and those reporting no commercial asbestos exposure (data not shown).
Parenchymal changes for participants hired between January 1, 1957 to September 1, 1980 were also significantly associated with increasing CFE and employment through September 1, 1980 with adjusted (adj)RR ranging from 3.5 to 5.8 (Table 4). Results were similar for participants reporting no commercial asbestos exposure. The exposure response relationships were similar for CFE computed to May 1, 2001 (data not shown).
HRCT/CT Pleural and Parenchymal Changes and Spirometry Results
Of the 191 participants with HRCT/CTs, 168 had a pulmonary function test. Two participants with lobectomies, nine with radiographic findings of previous coronary artery bypass graft surgery, and three with acute cardiopulmonary or significant neuromuscular morbidities were excluded, leaving 154. Of these 154 participants, spirometry was obtained within 24 hours of HRCT/CT for 84.4% (n = 137). Of the remaining 24, 9.1% (n = 14) were obtained within 3 years of the HRCT/CT and 6.5% (n = 10) were obtained 6 to 7 years prior to the HRCT/CT scan. Adjusting for pack years and BMI, participants with HRCT/CT findings had lower estimated percent predicted pulmonary function test values compared with those with normal HRCT/CTs (Table 5). Localized pleural thickening alone was associated with a −6.1 and −4.3 loss in percent predicted FVC (P ≤ 0.05) and FEV1 (P = 0.14), respectively. Any pleural change (LPT and/or DPT) and DPT and/or parenchymal changes were significantly associated with greater loss in percent predicted FVC and FEV1 (Table 5). Excluding participants with potential commercial asbestos exposure or adding a variable to adjust for emphysema and/or bronchiectasis to the models did not appreciably change the findings (data not shown). Percent predicted FEV1/FVC ratio was not statistically significant in any model in Table 5 (data not shown). Parenchymal changes with LPT only (n = 14 with spirometry) were associated with a −11.5 and −5.7 loss in percent predicted FVC (P = 0.01) and FEV1 (P = 0.23), respectively, a finding which persisted when emphysema and/or bronchiectasis was added to the model (data not shown).
Pleural Changes With Combined HRCT/CT or Chest Radiograph Results
As a sensitivity analysis and to assess potential systemic error in overestimating any affect because of participation bias, particularly within the lower CFE categories, additional analyses were conducted. Pleural changes by HRCT/CT or, when available, chest radiograph results (n = 115) were evaluated. In addition, the 125 living nonparticipants were included, assuming they had normal chest imaging studies. Table 6 shows a significant exposure–response relationship of all exposure categories utilizing the recently updated JEM,11 including the 0.15 to less than 0.45 CFE category with significant RR ranging from 6.4 to 13.4 (P < 0.001). This analysis was repeated after excluding nonparticipants (n = 125). Significant RR ranged from 5.3 to 10.4 (P < 0.001). The results were similar when CFE for participants and nonparticipants was estimated through May 1, 2001; the RR ranged from 6.2 to 13.1 (P < 0.001) (data not shown).
When evaluated by HRCT/CT, 52.9% (101/191) of participants demonstrated pleural changes significantly related to increasing CFE and employment duration. Of those participants with no reported commercial asbestos exposure within the 0.15 to less than 0.45 and 0.45 to less than 1.35 CFE categories, 43.2% and 53.1% had pleural changes, respectively. These working lifetime CFE levels are approximately 3 to 10 times less than the 45-year working lifetime US Occupational Safety and Health Administration standard of 4.5 fiber-year/cm3 for commercial asbestos exposure.19 The results of the HRCT/CT analyses for participants hired on or after 1972 when airborne fiber levels were measured continued to demonstrate a significant effect with pleural changes with increasing CFE and employment duration. These findings related to low lifetime CFE are relevant not only for consumers who used LV as home insulation but potentially for community exposures to amphibole asbestos fibers in general through man-made environmental contamination or naturally occurring outcroppings of amphibole asbestos or other asbestiform minerals.
The prevalence rate for parenchymal changes by HRCT/CT was 13.1% (25/191) and was significantly related to increasing CFE and employment duration. It is noteworthy that in the 0.45 to less than 1.35 CFE group, 15.8% (6/38) had parenchymal changes (five of six with correlative bilateral LPT). It has not been previously documented that exposure to LAA is associated with parenchymal changes at these low lifetime CFE levels. Early stage pulmonary fibrosis based on HRCT results and commercial asbestos exposure has been associated with a CFE of 25 or less fiber-years/cm3.20,21 A toxicokinetic explanation for the apparent increased propensity of LAA to cause pleural and parenchymal changes at lower CFE levels compared with that commonly associated with commercial asbestos exposure requires further investigation.
Pleural changes on the basis of HRCT/CT were shown to have an impact on pulmonary function. There was an increasing loss in percent predicted FVC for LPT only, LPT and/or DPT, and DPT and/or parenchymal changes across all participants and among those with no reported commercial asbestos exposure. Similar results were demonstrated for percent predicted FEV1. These decrements in percent predicted FVC and FEV1 are not likely related to other potential structural or physiologic abnormalities. Those with chest surgery were excluded, and adjusting for HRCT/CT findings of emphysema and/or bronchiectasis did not change the findings. The cause of the restrictive spirometry pattern related to LPT is unclear. Potential mechanisms include impaired diaphragm excursion, limited chest wall expansion, or undetected early interstitial fibrosis.22–24 The latter is unlikely because of the high predictive value of HRCT/CT scans to detect early interstitial fibrosis.7 Previous studies comparing chest radiographs or HRCT findings and pulmonary function have demonstrated a similar restrictive pattern in relationship to DPT and/or parenchymal findings.7,25–27 Localized pleural thickening only studies have demonstrated more varied findings.28–30 Review articles have estimated the reduction in FVC attributable to LPT (plaques) from commercial asbestos exposure ranging from 4% to 10%.7,31 Clin et al32 utilized HRCT to evaluate LPT alone, excluding other radiographic findings that could impact pulmonary function and found a 3.8 and 4.0 loss in percent predicted FVC and FEV1, respectively, compared with 6.1 and 4.3, respectively, reported in this study.
Studies of LV miners and millers have demonstrated both pleural and parenchymal chest radiograph changes associated with increased CFE.33,34 Lower exposure has also been associated with pleural changes in a more recent study, but there were indicators of potential exposure misclassification.35 A study of urban residents living near a plant using LV also demonstrated an increased risk of pleural changes at lower estimated CFE.36,37
A study of spirometry results from a community screening program of Libby residents demonstrated increased odds ratios between parenchymal changes and DPT as well as LPT only based on chest radiographs and restrictive spirometry (FEV1/FVC greater than lower limits of normal and FVC less than lower limits of normal).38 Weill and colleagues39 found that male never smoker Libby residents had a 4.3% and 23.8% loss in percent predicted FVC for LPT and DPT and/or costophrenic angle blunting, respectively. Libby miners with LPT only based on HRCT demonstrated losses in percent predicted for FVC and FEV1 of 5.3 and 5.7, respectively, similar to this study results.40
Participation bias was a potential study limitation because participants had higher mean CFE and longer employment duration than living nonparticipants. Nevertheless, these differences are not likely to impact the lowest nonreference exposure category (0.15 to less than 0.45), where 44% of the HRCT/CT participants had pleural changes.
Although 899 fiber measurements were available from 1972 to 1994, a second limitation is potential exposure misclassification, especially for the 1957 to 1971 period when no measurements were available. Nevertheless, the findings of similar results for participants hired on or after 1972 despite smaller sample sizes support the findings of radiographic changes with low lifetime CFE.
The use of the 175 HRCT results may have resulted in minor underestimation of true prevalence of LPT; that is, noncontiguous evaluation of the pleural surfaces may have omitted plaque that occurred in the 14-mm increments between the 1-mm slice thickness. In turn, the use of the 16 conventional CT scans in the analysis could result in minor underestimation of parenchymal changes in comparison with the more sensitive HRCT technique. In view of the changes identified in the chest CT scans previously outlined and the combined HRCT/CT analysis, any misclassification of pleural or parenchymal changes would be both minor and an underestimation, making the data even more compelling.
The potential for a bias related to the time interval between spirometry and HRCT/CT is minimal because 84.4% were obtained within 24 hours of the HRCT/CT and 93.5% within 3 years. The remaining 6.5% were obtained 6 to 7 years prior to HRCT/CT, which could result in underestimating the spirometric impact associated with the HRCT/CT findings.
This study demonstrates that exposure to low lifetime levels of LAA (CFE 1.35 or less fiber-years/cm3) is significantly associated with LPT (plaque) and parenchymal changes. These findings are important because these changes occur at substantially lower CFE levels than that commonly associated with commercial asbestos. The occurrence of LPT alone and DPT and/or parenchymal changes was associated with a mean 6.1 and 18.0 loss in percent predicted FVC, respectively. These decrements can be particularly relevant when potentially combined with other respiratory comorbidities that can occur over a person's lifetime that can impact lung function. In summary, fiber exposure that includes LAA is associated with progressive radiographic changes at low CFE levels, and these radiographic changes, including isolated LPT based on HRCT/CT, are associated with declines in spirometric values.
The authors thank the men and women who were participants in this study and their families who provided information. The authors also appreciate the cooperation of the company personnel and thank Connie Thrasher, from the University of Cincinnati, for her administrative support.
2. US Environmental Protection Agency. Health Assessment Document for Vermiculite. Washington, DC: US Environmental Protection Agency; 1991. Publication No. EPA/600/8–91/037.
3. Meeker GP, Bern AM, Brownfield IK, et al. The composition and morphology of amphiboles from the Rainy Creek complex, near Libby, Montana. Am Mineral. 2003;88:1955–1969.
4. Lockey JE, Brooks SM, Jarabek AM, et al. Pulmonary changes after exposure to vermiculite contaminated with fibrous tremolite. Am Rev Respir Dis. 1984;129:952–958.
5. Rohs AM, Lockey JE, Dunning KK, et al. Low-level fiber-induced radiographic changes caused by Libby vermiculite: a 25-year follow-up study. Am J Respir Crit Care Med. 2008;177:630–637.
6. Dunning KK, Adjei S, Levin L, et al. Mesothelioma associated with commercial use of vermiculite containing Libby amphibole. J Occup Environ Med. 2012;54:1359–1363.
7. American Thoracic Society. Medical section of the American Lung Association: diagnosis and initial management of nonmalignant diseases related to asbestos. Am J Respir Crit Care Med. 2004;170:691–715.
8. Webb WR, Muller NL, Naidich DP. Diseases characterized primarily by linear and reticular opacities. In: High-Resolution CT of the Lung. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001; Chapter 4:236–251.
9. International Labor Office. Guidelines for the Use of the ILO International Classification of Radiographs of Pneumoconioses, Revised 2000. Ed. Geneva, Switzerland: International Labor Office; 2002.
10. Miller MR, Hankinson J, Brusasco V, et al. Standardisation of spirometry. Eur Respir J. 2005;26:319–338.
11. Borton EK, Lemasters GK, Hilbert TJ, Lockey JE, Dunning KK, Rice CH. Exposure estimates for workers in a facility expanding Libby vermiculite: updated values and comparison with original 1980 values. J Occup Environ Med. 2012;54:1350–1358.
13. Atkinson GR, Rose D, Thomas K, Jones D, Chatfield EJ, Going JE. Collection, Analysis and Characterization of Vermiculite Samples for Fiber Content and Asbestos Contamination. Kansas City, MO: Midwest Research Institute EPA Office of Pesticides and Toxic Substances; 1982. Publication No. EPA 68–01–5915.
14. Moatamed F, Lockey JE, Parry WT. Fiber contamination of vermiculites: a potential occupational and environmental health hazard. Environ Res. 1986;41:207–218.
15. NIOSH Site Visit Report 2002–0437. The Scotts Company, Marysville, Ohio. October 6, 2004. Respiratory Disease Hazard Evaluations and Technical Assistance Program. Field Studies Branch. Division of Respiratory Disease Studies. NIOSH.
16. Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S. population. Am J Respir Crit Care Med. 1999;159:179–187.
17. Berrizbeitia LD, Tessler S, Jacobowitz IJ, Kaplan P, Cunningham JN. Effect of sternotomy and coronary bypass surgery on postoperative pulmonary mechanics. Comparison of internal mammary and saphenous vein bypass grafts. Chest. 1989;96:873–876.
18. Kee ST, Gamsu G, Blanc P. Causes of pulmonary impairment in asbestos-exposed individuals with diffuse pleural thickening. Am J Respir Crit Care Med. 1996;154(pt 1):789–793.
20. Paris C, Benichou J, Raffaelli C, et al. Factors associated with early-stage pulmonary fibrosis as determined by high-resolution computed tomography among persons occupationally exposed to asbestos. Scand J Work Environ Health. 2004;30:206–214.
21. Paris C, Martin A, Letourneux M, Wild P. Modelling prevalence and incidence of fibrosis and pleural plaques in asbestos-exposed populations for screening and follow-up: a cross-sectional study. Environ Health. 2008;20:7:30.
22. Fridriksson HV, Hedenstrom H, Hillerdal G, Malmberg P. Increased lung stiffness in persons with pleural plaques. Eur J Dis. 1981;62:412–424.
23. Singh B, Eastwood PR, Finucane KE, Panizza JA, Musk AW. Effect of asbestos-related pleural fibrosis on excursion of the lower chest wall and diaphragm. Am J Respir Crit Care Med. 1999;160:1507–1515.
24. Broderick A, Fuortes LJ, Merchant JA, Galvin JR, Schwartz DA. Pleural determinants of restrictive lung function and respiratory symptoms in an asbestos-exposed population. Chest. 1992;101:684–691.
25. Wilken D, Velasco GM, Manuwald U, Baur X. Lung function in asbestos-exposed workers, a systematic review and meta-analysis. J Occup Med Toxicol. 2011;26:6–21.
26. Schwartz DA, Galvin JR, Yagla SJ, Speakman SB, Merchant JA, Hunninghake GW. Restrictive lung function and asbestos-induced pleural fibrosis. A quantitative approach. J Clin Invest. 1993;91:2685–2692.
27. Becklake MR. Asbestos and other fiber-related diseases of the lungs and pleura. Distribution and determinants in exposed populations. Chest. 1991;200:248–254.
28. Staples CA, Gamsu G, Ray CS, Webb WR. High-resolution computed tomography and lung function in asbestos-exposed workers with normal chest radiographs. Am Rev Respir Dis. 1989;139:1502–1508.
29. Copley SJ, Wells AU, Rubens MB, et al. Functional consequences of pleural disease evaluated with chest radiography and CT. Radiology. 2001;220:237–243.
30. Borbeau J, Ernst P, Chrome J, Armstrong B, Becklake MR. The relationship between respiratory impairment and asbestos-related pleural abnormality in an active work force. Am Rev Respir Dis. 1990;142:837–842.
31. Rockoff SD, Chu J, Rubin LJ. Special Report: asbestos-induced pleural plaques-A disease process associated with ventilatory impairment and respiratory symptoms. Clin Pulm Med. 2002;9:113–124.
32. Clin B, Paris C, Ameille J, et al. Do asbestos related pleural plaques on HRCT scans cause restrictive impairment in the absence of pulmonary fibrosis? Thorax. 2011;66:985–991.
33. McDonald JC, Sebastien P, Armstrong B. Radiological survey of past and present vermiculite miners exposed to tremolite. Br J Ind Med. 1986;43:445–449.
34. Amandus HE, Althouse R, Morgan WK, Sargent EN, Jones R. The morbidity and mortality of vermiculite miners and millers exposed to tremolite-actinolite: part III. Radiographic findings. Am J Ind Med. 1987;11:27–37.
35. Larson TC, Antao VC, Bove FJ, Cusack C. Association between cumulative fiber exposure and respiratory outcomes among Libby vermiculite workers. J Occup Environ Med. 2012;54:56–63.
36. Alexander BH, Raleigh KK, Johnson J, et al. Radiographic evidence of nonoccupational asbestos exposure from processing Libby vermiculite in Minneapolis, Minnesota. Environ Health Perspect. 2012;120:44–49.
37. Christensen KY, Bateson TF, Kopylev L. Low levels of exposure to Libby amphibole asbestos and localized pleural thickening. JOEM. 2013;55:1350–1355.
38. Larson TC, Lewin M, Gottschall EB, Antao VC, Kapil V, Rose CS. Associations between radiographic findings and spirometry in a community exposed to Libby amphibole. Occup Environ Med. 2012;69:361–366.
39. Weill D, Dhillon G, Freyder L, Lefante J, Glindmeyer H. Lung function, radiological changes and exposure: analysis of ATSDR data from Libby, MT, USA. Eur Respir J. 2011;38:376–383.
40. Clark KA, Flynn J, Goodman JE, Zu K, Karmaus WJJ, Mohr LC. Pleural plaques and their effect on lung function in Libby vermiculite miners. Chest. 2014;146:786–794.