Document 4vjweaV5VEra5RE4VEkBD44Xe

+"':c"*1t8r;J' COt-yr' `1 be5rl 'e^uce:J w!fh p-r-nltslon ef ngotottjn it pf_ohib,to. A. J f /r Aj/J / ^ Pulmonary Function and Respiratory Symptoms in Polyvinylchloride Fabrication Workers1-3 i etca MICHAEL E. BASER, MELVYN S. TOCKMAN, and THOMAS P. KENNEDY M 271985 Introduction In 1978, a union representing employ ees of a company that fabricates poly vinylchloride (PVC) resins into vinyl sheeting and wallcovering submitted a request to the National Institute of Oc cupational Safety and Health (NIOSH) for an investigation of complaints concerning respiratory symptoms. The complaints were centered in the cal ender department, where PVC resins and other items were mixed under heat and pressure to form vinyl sheeting and wallcovering. Other exposures of con cern existed in the laminating, color, print and print service, and mainten ance departments (table 1). Production line areas were separated by brick walls that largely confined exposures within departments. On the basis of work place air monitoring by a NIOSH indus trial hygienist and interview of em ployees by a NIOSH physician, further investigation of the pulmonary com plaints was recommended. Pulmonary function abnormalities and dyspnea have been reported in vinyl chloride monomer (VCM) polymeriza tion workers (1-5), who are exposed to VCM and PVC dust. However, area monitoring of workplace air on the floor of the calender department indicated VCM concentrations of 1 ppm or less. Exposure to high PVC dust levels, such as occurs in manual bagging and load ing, is associated with radiographic ab normalities (6-11), dyspnea (9) and his topathologic changes (8, 12), although the evidence for pulmonary function im pairment is conflicting (9, 11). The PVC dust levels were not measured in the plant under study, but the only sources of PVC dust were fugitive emissions from over head pneumatic delivery systems, which do not produce dust levels comparable to manual operations. Because a haze was visible in the calender area in the plant under study, possible exposure to PVC thermal degra dation products was assessed. Polyvinyl chloride is thermally stable at tempera tures below 225 C. Between 225 C and 475 C (the ignition point), hydrochlo SUMMARY W* performed preshift and postahift spirometry and administered a standardized respi ratory symptoms questionnaire to 174 white males currently employed in polyvinylchloride (PVC) fabrication to examine the acute and chronic reapiratory eflecta of work exposure. Although there were no significant differences between the in-plant comparison group and any department with potential exposures, there was evidence for respiratory affects in the combined group of compari son and exposed workers. In the combined group, duration of employment was significantly as sociated with decrements in ad|usted cross-shift ratio of forced expiratory volume in one second to forced vital capacity (FEV,/FVC), preshift FEV,/FVC, and prevalence of chronic cough and chronic phlegm. In nonsmokers, the prevalences of chronic wheeze and chest tightness were high (36.0 and 50.5H, respectively). The age-ad|uated prevalence of chronic wheeze in nonsmokers was also elevated 3.54-fold when compared with that in a community study in the literature. We conclude that the cross-sectional design and in-plant comparison group may mask the effects of exposures on the entire plant population, and that employment in this plant is associated with patterns of obstructive air-flow limitation and respiratory symptoms consistent with exposure to pulmonary irritants. These results suggest that possible agents of pulmonary effects in PVC fabrication are not limited to vinyl chloride monomer, PVC dust, and PVC thermal degradation products. AM REV RESPIR OIS 1985; 131:203-208 ric acid (HC1) is first released, then, at increasing temperatures, other com pounds such as carbon monoxide, car bon dioxide, benzene, and VCM are formed (13). At temperatures in excess of 475 C, phosgene and chlorine are formed. Commercial meat wrappers ex posed to fumes from PVC film cut with a hot wire (14, 15) report an increase in respiratory symptoms, although pulmo nary function impairment in these work ers has not been unequivocally demon strated (16). Exposure of guinea pigs to sublethal doses of PVC thermal decom position products produces a depression of the normal ventilatory response to C02 (17) and changes in pulmonary func tion, lung mixed-function oxidase activ ity, and lung organ to body weight ratio (18). In an industrial accident involving overheating of PVC in a PVC fabrica tion plant, workers experienced acute up per and lower respiratory irritation; symptoms and abnormal pulmonary function persisted in some workers for several months after the accident (19). However, in the plant under study, the operating temperature of the mixing pro cess in the calender department was 200 C, and airborne concentrations of HC1 were below detectable limits. Although historical incidents of PVC overheating were reported by one worker, the fre quency and duration of these incidents were not known. Polyvinylchloride fabrication is a widespread process in industry. Because previously reported agents associated with respiratory effects were not present at this plant, we designed this study to determine if work exposure in PVC fabri cation under normal operating condi tions was associated with acute or chronic pulmonary function impairment or ex cessive respiratory symptomatology, rel ative to a comparison group of workers in production line departments subjec tively free of fume, dust, and odor. Methods The 5 departments with potential exposures and a comparison group of 3 departments subjectively free from fume, dust, and odor were studied (table 1). The comparison group consisted of workers from the shipping, in spection, and receiving and materials depart- (Received in original form April II, 1984 and in revised form August 6, 1984) ' From the Departments of Environmental Health Sciences and Medicine, The Johns Hop kins Medical Institutions, Baltimore, Maryland. 1 Supported by Grant No. 210-79-00-88 from the National Institute of Occupational Safety and Health. 1 Requests for reprints should be addressed to Michael E. Baser, Department of Environmental Health Sciences, The Johns Hopkins School of Hygiene and Public Health, 615 N. Wolfe Street, Room 7032, Baltimore, MD 21205. 203 BFG08327 20941001 204 BASER. TOCKMAN. ANO KENNEDY TABLE 1 PROCESSES AND PRINCIPAL CHEMICALS IN STUDY DEPARTMENTS Study Departments Process Principal Chemicals Departments with potential exposures Calender Laminating Print and print service Color Maintenance Mixing of PVC resins and additional items under heat and pressure to form a soft vinyl dough, which is milled into thin sheets Lamination of layers of vinyl sheeting or backing under heat and pressure Mixing of printing inks and decorative printing and embossing of vinyl sheeting Mixing of solvent-based inks Plant clean-up and equipment repair PVC resin powders Liquid plasticizers Stabilizers Fire retardants Color dispersions Vinyl adhesives 1.1,1 trichloroethane Ink pigment concentrates Solvents* Solvents' Lead chromate pigment powders PVC and copolymer resins Dependent on location within plant Comparison departments Shipping Inspection Receiving & materials Handling of feedstocks or finished product None ' Solvents include methyl ethyl ketone, methyl isobutyl ketone, toluene, xylene, and cyclohexanone. ments, which handled only finished products (shipping and inspection) or packaged raw materials (receiving and materials). The 201 white males currently employed in these departments were eligible for study. Demo graphic groups other than white males were not studied because of inadequate represen tation in the departments under study. Eligi ble workers were excluded from study if they were unavailable for study (n = 15) or if they declined to participate (n = 12). The 174 workers studied, representing 86% of the eligi ble population, were stratified by department and by work shift and were randomly allo cated to be tested over the 5 days of the study. On the day of his test, each worker was ad ministered the American Thoracic Society (ATS) respiratory symptoms questionnaire (20) in a NIOSH trailer adjacent to the plant by interviewers previously trained in standard technique (21). Trained pulmonary function technicians administered preshift and post shift spirometry within 90 min of beginning and ending the work shift. Each worker was tested before and after his shift by the same technician. Spirometry was performed in a climate-controlled room in a clean area of the plant and in accordance with the ATS Snow bird conference criteria (22), using Stead-Wells water-filled spirometers (Warren E. Collins, Braintree, MA). Spirometers were calibrated with a stopwatch and a 3-L syringe before test ing each shift. Forced vital capacity (FVC) and forced expiratory volumes in one, three, and six seconds (FEV,, FEVu and FEVS) were corrected to btps and analyzed. (The ratio of FEVj to FEV6 is an index of flow towards the end of the forced expiration and may re flect small airways changes (23)). Work his tories indicating the calendar dates and du ration of employment in each department were obtained from company records. The Abbey-Tonascia modification of the Feldstein multiple linear regression technique was used to simultaneously adjust pulmon ary function and symptom prevalence for age, height, smoking, shift, day of week, du ration of employment, and current depart ment (24). Within each independent varia ble, comparisons were made to the adjusted function or rate of a reference class. The values for these reference classes were analogous to predicted values, in that they indicate the value that would be expected after adjustment for all independent variables other than the one being tested. Comparisons were made to de termine if the adjusted outcome changes over the range of the independent variable under study. For example, the adjusted function of persons in the age classes 30 to 39, 40 to 49, and 50 or more yr was compared with the ad justed function of persons younger than 30 yr of age, the reference class, to determine if age was associated with function. The chisquare test with Yates' continuity correction was used for discrete variables, such as respir atory symptoms, and the two-tailed t test was used for continuous variables, such as pul monary function (25). Differences that returned p values of less than 0.05 were con sidered significant. Results Age and duration of employment were not significantly different between the 174 eligible workers studied and the 27 eligible workers not studied. Of the 174 eligible workers studied, 170 (98%) had acceptable quality preshift spirometry, 165 (95%) had acceptable quality preshift and postshift spirometry, and 163 (94%) had acceptable quality spirometry plus available work histories. Demographic characteristics of the 174 workers studied are presented in table 2. 0941002 TABLE 2 DEMOGRAPHIC CHARACTERISTICS OF STUDY POPULATION, BY CURRENT DEPARTMENT' Current Department Characteristic Calender Laminating Print and Print Service Color Maintenance Comparison Group n Height, inches Duration of total employment, yr Age, yr Smoking status, n Nonsmoker Ex-smoker Current, < 1 pack/day Current, > 1 pack/day 41 70 2 3.3 10.6 5.2t 37.7 9.9t 18 4 4 15 23 70.4 2.9 15.0 3.8 47.3 9.0 10 4 3 6 38 69.9 i 2.5 11.5 i 5.63 39.8 9 7 13 5 4 16 6 69.1 0.9 8.2 5.8t 43,7 18.7 2 2 1 1 25 70.6 2.6 10.8 5.8t 42.2 7.5 11 3 1 10 41 69.6 2.1 14.2 3.7 44.0 10.7 18 7 2 14 Total 174 70.1 2.7 12.2 5.2 41.8 10.1 75 25 15 65 * Values are mean SD. t Two-tailed r test, p < 0.05 as compared with that in comparison group. BFG08328 M PULMONARY FUNCTION IN POLYVINYLCHLORIDE WORKERS 205 The mean age of calender workers was significantly less than the comparison group (37.7 yr versus 44.0 yr, respectively, p < 0.05). The mean duration of employ ment of workers in all departments ex cept laminating was significantly less than that of workers in the comparison group. There were no significant differ ences in mean height or in the distribu tion of smoking status by department. Mean adjusted cross-shift changes in pulmonary function are presented in ta ble 3. Cross-shift changes were seldom greater than approximately 1% of preshift values. However, the slight de cline (-0.19%) in adjusted FEV./FVC over Shift 2 was significantly less (p < 0. 05) than the 1.30% increase over Shift 1. The 0.24% increase in adjusted cross shift change in FEV,/FVC in persons employed for 6 months or more was sig nificantly less (p < 0.01) than the 3.22% increase seen in persons employed for less than 6 months. Aside from the expected changes in ad justed preshift pulmonary function with increasing age and smoking habit, there were no significant differences by shift, day of week, duration of employment, or current department (table 4). However, the duration of employment-related change in the acute response suggested to us that persons whose acute response was greatest might migrate to nonexposed departments, thereby attenuating differ ences between exposed and comparison departments seen on cross-sectional view. If this were true, total duration of em ployment in departments, rather than current department membership, would be the most appropriate work history measure to assess employment-associated chronic effects on pulmonary function. When we regressed preshift FEV,/FVC against total duration of employment in each department (table 5), the slope for each department except maintenance, the smallest group, was significantly less than zero, even after adjustment for age, height, smoking, shift, day of week, and duration of employment in other depart ments. Several patterns are apparent in table 6, which presents the mean adjusted prevalences of lower respiratory symp toms. The adjusted prevalences of chronic cough, chronic phlegm, and chronic wheeze were significantly lower (p < 0.05) in persons tested on Wednes day than in those tested on Monday. The prevalence of wheeze plus shortness of breath increased daily, from 16.4 to 34.9% over the work week. The adjusted prevalences of chronic cough and chronic TABLE 3 MEAN ADJUSTED CROSS-SHIFT CHANGE- IN PULMONARY FUNCTION BY CHARACTERISTIC Mean Cross-shift Change in Adjusted Pulmonary Function (%) Characteristic Class FEV, FEV./FVC FEV,/FEV, Age Smoking status Shift Day of week Duration of employment, yr Current department < 30 30-39 40-49 50 + Nonsmoker Ex-smoker Current smoker 1 2 3 Mon Tue Wed Thu Fri < 0.5 S 0.5 Comparison group Calender Laminating Print and print service Color Maintenance 0.52 0.38 0.32 0.62 0.62 0.79 -0.15 0.64 -0.52 1.33 0.01 0.62 -0.14 -0.47 3 44 2.41 0.17 1.30 0.56 -0.15 1.16 0.99 -2.09 0.56 0.76 0.53 0 53 1.06 0.51 0.18 1.30 -0 -19t -0.07 1.05 0.19 1.21 0.29 0.10 3 22 0 24* 0.69 0.89 0.75 1.01 -0.73 -0.50 -0.50 0.08 0.36 -0.02 0.29 -0.36 -0.03 000 -0.01 0.30 0.08 0.43 0.38 -0.53 -0 20 0.04 0.06 -0.14 0.00 -0.20 1.30 1.17 0.07 ' {Pre-Post)/Pre X 100. t Two-tailed t test, p < 0 05 * Two-tailed t test, p < 0.01 TABLE 4 MEAN ADJUSTED PRESHIFT PULMONARY FUNCTION BY CHARACTERISTIC Mean Adjusted Preshift Pulmonary Function Characteristic Age, yr Smoking status Shift Day of week Duration of employment, yr Current department Class < 30 30-39 40-49 50 + Nonsmoker Ex-smoker Current smoker 1 2 3 Mon Tue Wed Thu Fri < 0.5 > 0.5 Comparison group Calender Laminating Print and print service Color Maintenance FEV, (L) 4.38 3.96* 3 62* 3 04* 3.79 3.84 3.50- 3.73 3.63 3.56 381 3.68 3.54 3.73 3.54 3.85 365 3.64 3.58 3 66 3.76 3.37 3.79 FEV./FVC (%) 82 0 77.676.1* 70.8* 77 5 79.1 73.4* 75.7 75 6 77.0 76.1 74.7 76.1 772 75.2 76.0 75.9 75.4 75.3 74.7 77 6 71.8 77.3 FEV,/FEV, (%) 97.3 94 9* 94 1* 92.2* 94.5 95.7 93 6 946 93 8 94 6 94 0 93.5 93.8 95 2 94 0 94.9 94.2 94.0 94 1 94 4 94.8 92.0 94.4 * Two-tailed t test, p < 0 05. t Two-tailed t test, p < 0 01 * Two-tatied t test, p < 0 001 BFG08329 20941003 t 206 BASER, TOCKMAN. ANO KENNEDY TABLE 5 REGRESSION OF ADJUSTED* PRESHIFT FEV,/FVC ON TOTAL DURATION OF EMPLOYMENT IN DEPARTMENTS Department* Calender Laminating Print and print service Color Maintenance Comparison group Explained Correlation Variance Intercept a Slope b nr r1 (%) (% yr) too -0.168 0.020 78.13 -0.16 95 -0.426 0.181 78.46 -0.35 88 -0.368 0.135 78.75 -0.29 41 -0.312 0 098 78.43 -0.50 29 -0.209 0.044 76.26 -0.17 92 -0.287 0.082 77.87 -0.29 P 0.047 < 0.001 < 0.001 0.023 0.139 0.003 * Adjusted for age, height, smoking, shift, day of week, and duration of employment in other departments, t The sum of number of individuals in each department does not equal 174 because a person may have been employed in more than one department over his work history. As a result, the departmental groups are not mutually exclusive, and comparison of slopes between departments is not appropriate TABLE 6 MEAN ADJUSTED PREVALENCE OF LOWER RESPIRATORY SYMPTOMS BY CHARACTERISTIC Characteristic Age, yr Smoking status Shift Day of week Duration of employment, yr Current department Class < 40 > 40 Nonsmoker Ex-smoker Current smoker 1 2 3 Mon Tue Wed Thu Fri < 0.5 > 0.5 Comparison group Calender Laminating Print and print service Color Maintenance Adjusted prevalence (%) Chronic Cough Chronic Phlegm Chronic Wheeze Wheeze and Short* ness of Chest Breath Tightness 21.0 18.2 13.9 6.9 27 6' 19.3 17.8 19.7 29.3 10.5 10.2* 27.5 16.6 0.0 21.8* 10.8 21.2 26.8 15.9 41.3 20.7 25.9 20.8 17.4 16.1 30.5 22.8 21.0 26.5 29 0 18.4 8.0* 35.1 25.0 2.1 25.8* 13.6 28.4 33.6 23.0 47.9 13.6 38.4 38.6 36.1 12.3* 49.3 44 0 30.0 36.4 48.9 49.8 20.7* 43.1 25.4 31.9 39.4 34.1 42.2 46.8 37.9 46.8 30.8 23.7 24.4 26.7 9.0 26.7 28.0 11.4 26.5 16.2 20.0 23.8 29.2 34.9 11.2 25.9 13.6 30.6 31.3 28.6 18.3 18.6 42.4 49.7 44 8 38.0 51.0 50.5 45.4 37.4 41.5 62.2 40.6 45.2 40.2 40.4 474 39.0 42.6 60.4 55.1 458 40.0 * Two-tailed f test, p < 0.05. phlegm were higher (p < 0.05) in persons employed for 6 months or more than in persons employed for less than 6 months. Last, the adjusted prevalences of chronic wheeze and chest tightness in all non smoking PVC workers (36.0 and 50.5 %, respectively), although not significantly different between characteristic catego ries, were higher than expected when compared with community studies in the literature (26-28). To explore the latter finding, we com pared the age- and smoking-specific prevalences of lower respiratory symp toms in PVC fabrication workers with the prevalences expected if the PVC workers experienced the same symptom preva lences as a community population, ob tained from a published study that used the MRC respiratory symptoms question naire (29). The MRC and ATS question naires elicit similar prevalences of chronic cough, chronic phlegm, and chronic wheeze, but the ATS questionnaire elicits a higher prevalence of any wheeze than does the MRC questionnaire (28); we therefore excluded wheeze plus shortness of breath from this comparison. As in dicated in table 7, compared with ex pected frequencies, the nonsmoking PVC fabrication workers had a 3.54-fold in crease in chronic wheeze (p < 0.001). The PVC workers who were moderate smokers had a 1.64-fold excess of chronic wheeze (p < 0.001). Other lower respira tory symptoms, however, were lower than expected in PVC workers who smoked (chronic phlegm in moderate smokers, 0.58-fold, p < 0.05; chronic cough in heavy smokers, 0.40-fold, p < 0.005). Discussion The results of this cross-sectional study indicate that acute and chronic obstruc tive ventilatory changes and an increased prevalence of lower respiratory symp toms, consistent with the presence of pul monary irritants, are associated with em ployment in this PVC fabrication plant. The 3.22% cross-shift increase in FEV,/FVC in persons employed for less than 6 months may be due to a training effect, which is significantly attenuated in persons employed for more than 6 months by cumulative effects of work place exposures. The decreased cross shift change in FEV,/FVC during Shift 2 relative to Shift 1 may be related to nat ural circadian rhythm (30). Although ad justed preshift pulmonary function is not significantly different between workers currently employed in departments with potential exposures and the comparison group, adjusted preshift FEV,/FVC declines significantly with total duration of employment in all but one department, suggesting an employment-associated chronic obstructive effect. The daily increase over the work week in the prevalence of wheeze plus short ness of breath suggests cumulative ex posure to a lower respiratory irritant. The adjusted prevalence of chronic cough and chronic phlegm increases with increas ing duration of employment, and the seemingly high overall prevalence of chronic wheeze, as supported by compar isons with a community study, also sug gest that employment is associated with an increase in lower respiratory symp toms. The interpretation of the observed variation in prevalence of chronic symp toms by day of week is unclear. The fact that none of the foregoing conclusions are evident from interdepart mental comparisons is notable, and un derscores the limitations of the cross-sec tional design and in-plant comparison group. A healthy working population may result from differential initial em ployment, selective migration within areas of the plant, and employment ter mination (31). In this study, for exam ple, although persons employed for 6 months or more had a significantly IroOTfrGOZ BFG08330 - PULMONARY FUNCTION IN POLYVINYLCHLORIDE WORKERS 207 TABLE 7 OBSERVED AND EXPECTED* LOWER RESPIRATORY SYMPTOMS BY AGE AND SMOKING STATUS Observed and Expected Symptom Frequencies Smoking Status Never Previous Light to moderate Heavy Age (yr) 18-44 45 + Total O/E 18-44 45 + Total O/E 18-44 45 + Total O/E 18-44 45 + Total O/E Chronic Cough Chronic Phlegm Chronic Wheeze n Obs. Exp. Obs. Exp. Obs. Exp. 43 10 6 8 6 27 9 32 12 7 6 7 19 4 75 22 13 14 13 46 13 1.69 1.08 3.54* 13 6 3 3 3 4 5 12 2 5 2 5 5 5 25 8 8 5 8 9 10 1.00 0.62 0.90 43 18 14 10 14 38 25 17 3 10 4 10 13 6 60 21 24 14 24 51 31 0.88 0.58* 1.64* 10 4 7 6 7 9 8 10 2 8 3 8 9 15 20 6 15 9 15 18 13 0.40* 0.60 1.38 * Expected frequencies derived from reference 30. t Chi-square test, p < 0 05 * Chi-square test, p < 0 005 Chi-square test, p < 0 001 smaller cross-shift change in FEV,/FVC than did persons employed for less than 6 months, in both groups FEV,/FVC in creased over the work shift. We conclude that the cross-sectional design and inplant control group may mask the effects of exposures on the PVC workers as a whole. Other explanations exist for these results. The lack of interdepartmental differences may simply indicate that em ployment in PVC fabrication is not associated with effects on pulmonary function or symptoms. Flowever, the con sistent associations of duration of em ployment with cross-shift change in FEV./FVC, preshift FEV./FVC, and prevalence of lower respiratory symptoms argue for work-associated effects. In this cross-sectional study design, we cannot differentiate between cohort effects, in which time-dependent effects would be due to high past exposure levels (or to introduction of a toxic agent), and ef fects caused by longer exposure to a simi lar dose level. Concern about a possible cohort effect is raised by the anecdotal reports of PVC overheating in the plant under study. Because the calendering area is enclosed by a brick wall, it is likely that exposure to PVC thermal degradation products would be limited to calender department employees. However, the sig nificant declines in adjusted preshift FEV./FVC in relation to total duration of employment were not limited to the calender department, but existed in all but one department (table 5). Although this study cannot identify specific agents associated with the ob served ventilatory effects, the industrial hygiene results rule out VCM. HC1. and PVC thermal degradation products. The PVC additives that partially volatilize at temperatures of 200 C or less, and chem icals used in noncalendering areas of the plant, might be contributing to the de velopment of respiratory symptoms and functional abnormalities in the work force. For example, the lead chromate in pigments used in the color department is a carcinogen, and chromate dust, when hydrated in the respiratory tract to form chromic acid, would be a pulmonary ir ritant (32). The large number of chemi cals used in this plant alone complicates the search for an agent or agents. How ever, PVC fabrication is a widespread process in industry, and the consistent evidence from this study of employmentassociated pulmonary effects should motivate further research. Acknowledgment We thank Drs. Edward Emmett and Harold Menkes for guidance in the initial stages of the study, and Peter Terry for manuscript re view; Mr. Stan Salisbury for industrial hy giene monitoring; J. Baetz, D. Ordin, J. Fer rara, R. Marnales, M. Kaczorowski, P. Pyzik, J. Jacobi, and P. Mason for technical assistance; M. Donathan and P. Wilkinson for computer programming; S. Graber and C. Sudbrook for secretarial assistance. References 1. Key MM, Henschel AF, Butler J, et at, eds. Oc cupational diseases: a guide to their recognition. USDHEW. Washington, D.C., U.S. Government Printing Office, 1977. 2. Litis R, Anderson H, Miller A, Selikoff 1. Pul monary change among vinyl chloride polymeriza tion workers. Chest 1976; 2(Suppt:299-305). 3. Miller A, Teirstein AS, Chuang M, Selikoff I. Changes in pulmonary function in workers exposed to vinyl chloride and polyvinyl chloride. Ann NY Acad Sci 1975; 246:42-52. 4. Gamble J, Liu S, McMichae! AJ, Waxweiler R. Effect of occupational and non-occupational factors on the respiratory system of vinyl chloride and other workers. J Occup Med 1975; 18:659-70. 5. Cordasco EM, Demeter SL, Kerkay J, et at. Pul monary manifestations of vinyl and polyvinyl chlo ride (interstitial lung disease). Chest 1980; 78: 828-34. 6. Mastrangelo CT, Manno M, Marcer G, et at. Polyvinyl chloride pneumoconiosis: epidemiolo gical study of exposed workers. J Occup Med 1979; 21:540-2. 7. Arnand A, Pommier de Sanki P. Garbo L, Payan H, Charpin J. Polyvinyl chloride pneu moconiosis. Thorax 1978; 33:19-25. 8. Szende B, Lapis K, Nemes A. Pinter A. Pneu moconiosis caused by the inhalation of polyvinyl chloride dust. Med Lav 1970; 61:433-6. 9. Souter CA, Copland LA. Thornley PE, et at. Epidemiological study of respiratory disease in workers exposed to polyvinylchloride dust. Tho rax 1980; 35:644-52. 10. Mastrangelo G, Saia B, Mercer G, Piazza CT. Epidemiological study of pneumoconiosis in the Italianpoly(vinylchloride)industry. Environ Health Perspect 1981; 41:153-7. 11. Chivers CP, Lawrence-Jones C, Paddle GM. Lung function in workers exposed to polys inyl chlo ride dust. Br J Ind Med 1978; 37:147-51. 12. Agarwal DK, Kaw JL, Srivasta SP. Seth PK. Some biochemical and histopathological changes induced by polyvinyl chloride dust in rat lung. En viron Res 1978; 16:333-41. 13. Dyer RF, Esch WH. Polyvinyl chloride toxic ity in fires. JAMA 1976; 235:393-7. 14. Smith TJ. Caferella JJ, Chelton C, Crowley S. Evaluation of emissions from simulated com mercial meat wrapping operations using PVC w rap. Am Ind Hyg Assoc J 1983; 44:176-83. 15. Pauli G, Bessot JC, Kopferschmitt MG, et at. Meat wrapper's asthma: identification of the causal agent. Clin Allergy 1980; 10:263-9. 16. Krumpe PE, Finley TN, Martinez N. The search for expiratory obstruction in meat wrappers studied on the job. Am Rev Respir Dis 1979; 119: 611-8. 17. Wong KL, Stock MF, Alarie YC. Evaluation of the pulmonary toxicity of plasticized polyvinyl chloride thermal decomposition products in guinea pigs by repeated CO, challenges. Toxicol Appl Phar macol 1983; 70:236-48. 18. Jaeger RJ, Sknornik WA. Heimann R. Ther mal decomposition products of PVC plastics: ef fects on guinea pig lung mechanics and pulmonary mixed function oxidase activity. Am Ind Hyg As soc J 1982; 43:900-7. 19. Froneberg B, Johnson PL, Landrigan PH. Res piratory illness caused by overheating of polyvinyl 20941005 BFG08331 208 BASER, TOCKMAN, AND KENNEDY chloride. Br J [nd Med 1982; 39:239-43. 20. Ferris BG. Epidemiology standardization proj ect. Am Rev Respir Dis 1978; 118(Suppl:7-55). 21. Survey Research Center, University of Michi gan Institute for Social Research. Interviewers man ual. Rev ed. Ann Arbor: University of Michigan. 1966. 22. American Thoracic Society. Snowbird Con ference on Standardization of Spirometry. Am Rev Respir Dis 1979; 119:831-8. 23. Permutt S, Menkes H. Spirometry: analysis of forced expiration within the time domain. In: Macklem PT, Permutt S, eds. The lung in the tran sition between health and disease. New York: Mar cel Dekker, 1979:113-79. 24. Snedecor GW, Cochran EG. Statistical methods. 6th ed. Ames: Iowa State University Press, 1978. 25. Feldstein M. A binary variable multiple regres sion method of analyzing factors affecting peri natal mortality and other outcomes of pregnancy. J R Stat Soc (A) 1966; 219:61-73. 26. Ferris BG, Higgings ITT, Higgins MW, Peters JM, Van Ganse WF, Goldman MD. Chronic non specific respiratory disease, Berlin, New Hampshire, 1961-1967: a cross-sectional study. Am Rev Respir Dis 1971; 104:232-44. 27. Lebowitz MD, Knudson RJ, Burrows G. Tuc son epidemiologic study of obstructive lung dis eases. 1. Methodology and prevalence of disease. Am J Epidemiol 1975; 102:139-52. 28. Helsing KL, Comstock GW, Speizer FE, ef aL Comparison of three standardized questionnaires on respiratory symptoms. Am Rev Respir Dis 1979; 120:1221-31. 29. Lebowitz MD. Occupational exposures in a community population. Environ Res 1977; 14: 59-67. 30. Guberan E, Williams MK, Walford J. Circa dian variation of FEV in shift workers. Br J Ind Med 1969; 26:121-5. 31. Monson RR. Occupational epidemiology. Boca Raton: CRC Press, 1981. 32. Parkes WR. Occupational lung disorders. 2nd ed. London: Butterworths, 1982. 20941006 BFG08332