Document gRmBNjmopZom0jNv3w3zMqrG

* INSTITUTE OF OCCUPATIONAL MEDICI N S REPORT OF A SURVEY CARRIED OUT AT THE HILLHOUSE WORKS ------------- OF IMPERIAL CHEMICAL INDUSTRIES LIMITED Medical Branch, Institute of Occupational Medicine, Roxburgh Place, EDINBURGH SH8 9SU, (Tel. 031-667-5''31) October 1979 ucc 057606 4 (ii) CONTENTS Pare No, 1. 1.1 1.2 1.3 1.4 1.3 2. 3. 3-1 4*. 3^ 3.3 3.4 3.3 * 4. 4.1 4.2 4.3 5. 6. SOtMAXt..................................................................... (m) INTRODUCTION.............................................. ....... . Animal and in vitro Studies on Toxicity of PVC.......................................................... 1 Reports of Effects of PVC Dust in man ... Surveys of PVC Workers............................ 3 Hazards of PVC PyrolysisProducts .... The Present Study ...................................................... 1 2 4 3 THE FACTORY11 PLANTS mAND PROCESSES .... METHODS . ............................... Population to be Studied...... The Sample . . . . . ... Environmental Survey ....... Medical Survey................................................ 11 Methods of Analysis..................................13 6 8 8 9 10 RESULTS.......................................... 13 Environmental Survey ....... Medical Survey - Summary of Results ... Medical Survey - Detailed Results .... 13 13 19 DISCUSSIOH ...............................................................................ji* CONCLUSIONS AND RECOMMENDATIONS .... 38 ACKNOWLEDGMENTS.........................................39 TABLES (1-28)...................................................... 41-60 FIGURES (1-13)............................... ....... . REFERENCES........................................................ 75 . 61-73 APPENDIX I................................................... 79 ucc 057607 (iii) INSTITUTE OF OCCUPATIONAL MEDICINE AN EPIDEMICLCGICAL STUDY 0? RESPIRATORY DISEASE IN WORKERS EXPOSED TO POLYVINYLCHLORIDE DUST by C.A. Soutar, L.H. Copland, P.E. Thornley, J.7. Hurley, J. Ottery SUMMARY Polyvinylchloride PVC) is produced in industry in the form of a A ^ ^ 1,, - -- - . - i- . m . _ . a t*X tf* * . m* <r -* **wu d *. WiiWWW V J\# XdJ w* 4 V i J dUl ilWieUW4(/ 31(1044 to be inhaled into the smaller airways of the lungs (respirable dust), and recent reports in ".he medical literature have suggested that this dust may cause lung disease. ( A study of a sample of the present and past workforce of the Hillhouae Works of Imperial Chemical Industries Limited, which manufactures PVC, has assessed the prevalence of respiratory symptoms, chest radiographic and lung function abnormalities, and has related them to estimates of exposure to resnirable PVC dust based on measurement of current dust levels and detailed occupational histories. 1. Environmental Survey In an environmental survey of the plants manufacturing or further processing PVC, personal sampling of resnirable dust was carried out on 130 men enraged in occupations within the PVC plants. Selection of men to wear the dust samplers was based on subjective estimates of dustiness made during a preliminary visit to the Works and measurement of particle sizes of typical PVC dusts. The major sampling effort was directed towards the occupations thought to b dustiest. Results of the environmental survey indicated that respirable dust levels were highest for the drying and packing operations in an older plant making PVC by the emulsion method and the mixing process in a PVC. processing plant. The highest mean average respirable dust exposure for any occupation over a shift was P.83 mg/tn3 (SD - 1.3*0. Dust levels in other plants were generally lower. 2* Methods of Medical Survey On the basis of factory personnel records, a preliminary inspection of the factory and estimates of the particle sizes of typical PVC dusts, all current and many ex-employees were allocated to four broad categories corresponding to different probable exposures to PVC dust. The study sample was drawn in such a way that all those in the two higher exposure categories were included, and lower proportions of men in the other categories. Men .were seen by appointment: a detailed occupational history was taken; a questionnaire of resniratory symptoms administered; chest radiograph and lung function measurements tak n. The ucc 057608 (iv) appearances of the chest ra^io^rnrhs were read by thr*** medicallv qualified readers highly experienced in the intemretatior. of the radiographic appearances of occupational lung disease. 3* Results of Medical 5'irvey Results for Rl8 men were analysed, 663 nen currently employed at the factory, 9# pensioners and 57 wen who had left for reasons other than retirement. (a) Respiratory Systems Exposure to PVC dust was not associated with chronic cough or sputum production (chronic bronchitis), acute chest illnesses, dvwwy wry f XdOWUhlAe There was a relations tip between exposure to PVC dust and the complaint of breathlessness when hurrying on level ground or walking up a slight hill, but this relationship was found only in cigarette smokers and not in non- or ex-smokers. (b) Lung Function A measurement of the breathing capacity, the forced expired volume in one second (FSVi) was statistically significantly low r among men with higher PVC dust exposure. This effect was seen principally in current cigarette smokers, and not confirmed among non-smokers when considered separately. The pattern of results suggested that there wore real differences in the response to PVC dust related to smoking habit. The magnitude of this reduction in all men in relation to the mean dust exposure was approximately one-seventh of that caused hy ageing, and of a similar "cmnitudo re the loss caused by smoking 20 cigarettes a day. (The FEVi measures the rate at which air can be blown out of the lungs through the air passages, and reductions of the FEV1 are usually caused by narrowing of the air passages in the lung.) Another measure of breathing capacity is the forced vital - capacity (FVC), the volume of air expired in a full br ath, and this was also statistically significantly reduced at higher PVC dust exposure. The magnitude of this reduction was less than that of the FEVX. (c) Chest Radiographs One of the three expert readers who examined the chest radiographs detected small rounded opacities more commonly in the radiographs of men with higher exposures to PVC dust than in thos with lower or no exposure. The other two readers did not detect an effect of PVC dust, though they saw small rounded opacities in a few cases. The opacities seen by these readers were not more common in those with higher exposures to dust than in those with lower or no exnosure. Nevertheless, men with small rounded opacities also had reduced lung function, compared to those without. ucc t 051609 (v) (d) Autoclave Workers The observed effects could not be related to post exposure to vinylchloride monomer. Ker. working on the autoclaves may have been exposed to significant amounts of vinylchloride monomer in the past. However, the observed effects of PVC dust exposure could not be explained by an effect of working on the autoclaves, and thus were unlikely to be due to exposure to the monomer. (e) Factors other than PVC Dust Effects The associations with PVC dust exposure were detected against a background of other non-oceupntion?lly-induced <* * * w w * wuu^Uf w^u W ulii W V tC i i ^ wU; J <3 j l.i y v WM1<9 *919 strongly related to smoking habits. So also were abnormalities of lung function; both deterioration in lung function and small opacities in the chest radiograph wer strongly associated with age. The statistical analyses of the results have taken these factors into account. 4. Conclusion Exposure,to respirable PVC'dust is associated in a proportion of exposed workers with the presence of small rounded opacities on the chest radiograph and a decline in mean ventilatory capacity. This suggests that PVC dust causes a small but detectable effect on the respiratory health or the .workforce. / UCC -sm 057610 1. INTRODUCTION Polyvinylchloride is produced in industry by the polymerisation of vinylchloride monomer. The monomer is gaseous at normal temperature and pressure, and hazards of occupational exposure to it are now well known, notably aero-osteolysis (CORDIER et al., 1966), Raynaud's syndrome (SUCIU et al., 1967). angiosarcoma of the liver (CREECH and JOHNSON, 197*0 and hepatitis (SUCIU et al., 1967; LANGE et al.. 1973). Hazaris of exposure to the monomer are not the subject of this report, and will not be reviewed here. In the manufacture of PVC, pressurised liquid vinylchloride monoms.' in the form of an emulsion or suspension in water is polymerised by heating .under pressure. Particles of PVC are formed, and during the drying and bagging operations PVC dust may be released. Though the manufacturing process is not necessarily universall/ identical, the PVC dust produced has been reported by Italian workers t include a large proportion by number of small particles of respirable size (CASULA et al., 1977; MAFP et al.. 1978). The proportion of small particles is greater for the emulsion method than the suspension method (CASULA et al.. 1977). Such small particles are likely to be inhal d and denosited in the lung, and there have been suggestions in the medical literature that inhalation of PVC dust may be harmful. 1.1 Animal and In Vitro Studies on Toxicity of PVC In animal studies intmtrscheal adsinistratier* of PVC dust has been reported to cause peribronchitis and thickening of alveolar septa (BOITSOV, 1963) and in rats single large intra tracheal inoculations of PVC dust have caused vascular and inflammatory changes in the terminal and respiratory bronchioles and alveoli, and a proliferation of histiocytes and retlculin fibres (AGABWAL et al.. 1978). Multinucleated giant cells and granulomatous lesions developed, and serial measurements of enzymes from lung tissue showed increases in energy-linked and lysosomal enzymes. Small mammals k pt in the bagging area of a PVC plant develoned a histiocytic and giant multinucleate cell alveolar reaction (FRCNGIA et al., 197*+). PVC has also been rep rted to have a cytotoxic effect when ucc 057611 1.3 1.3.1 1.3.2 3 prevalence of respiratory disease have given conflicting results. Surveys of PVC Workers Chest Radiology A report by VEBTKIN et al. (1970) suggested diffuse pulmonary radiographic shadowing in of 96 PVC workers in the USSR, though no details of radiographic film reading methods are given. LILIS et al. (1975; 1976; 1977) reported a high prevalence of radiographic abnormality in workers at two of three PVC plants in North America. Prevalence of abnormality was related to years of work at the plants, but was common (15 - 1686) even among workers exposed to PVC dust for less than two y ars. No dust measurements were reported, and although the radi graphic appearances were recorded using the ILO U/C pneumoconiosis classification by five physicians experienced in this method, no information is given of reader agreement, nor are the criteria for 'abnormality' given. MAPP *t al. (1978) found a low prevalence of chest radiographic abnormality among' 258 PVC workers in a plant in Italy; only one radiograph showed ILO U/C small irregular opacities category 1/0, while 29 showed category 0/1. Lung Function Mims et al. (1975) and LILIS et al. (1976; 1977) reported a high prevalence of abnormal lung function, obstructive in type. In the two PVC plants in North America mentioned above. Th predicted normal values, based on regressions for age and height derived from other populations, and the criteria for abnormality may not have been appropriate to the regions, which were reported to have been highly industrialised. Adequate control groups were not included in the studies, and it was not possible to tak account of the effect of smoking. GAMBLE et al. (1976) and WAXWEILER et al. (1977) compared the lung function of workers in a PVC plant with vinylchloride monomer workers and rubber workers, and after carefully taking Into account age and smoking effects, concluded that there was. n evidence that PVC work rs had reduced lung function, even UCC 057612 .3.3 .4 k whi*n the same criteria for abnormality used by MILLER et al. (1975) and LILI3 et al. (1976; 1977) were applied. However, dust levels were not measured, and the control subjects were also exposed to other chemicals which may have influenced their lung function. MAPP et al. (1978) found that the prevalence of airways obstruction was related to duration of exposure to PVC dust, though it was not possible to allow for the effects of age and smoking. LANGE et al. (1973) reported a restrictive abnormality of lung function in 8 of 13 PVC workers, though numerical details are not given. VERTKIN et al. (1975) found lung function to b normal in the 96 workers mentioned above, the majority of whom had abnormal chest radiographs. Respiratory Symptoms LXLIS et al. (1976) and MAPP at al. (1978) reported relatively high prevalences of chronic cough and sputum among PVC workers, using standard questionnaires (2QK and 3^ respectively), but it was not clear to what extent this was due to smoking and local atmospheric pollution. VERTKIN et_ al. (1970) found a lower prevalence of cough (7#), but the method of questioning was not described. Hazards of PVC pyrolysis Products Respiratory symptoms have been reported in workers using hot wires to cut PVC film for food-wrapping, and this has been attributed to pyrolysis products of PVC (SCKCL et al., 1973)* Thesa symptoms appear more common in food wrappers who are cigarette smokers, and it has been suggested that the effects of PVC pyrolysis products are more pronounced in smokers (FALK and PORTNOY, 1976). This syndrome has been studied by other workers (ANDRASCH et al.. 1975; POLAKOFF et al., 1975; JOHNSON and ANDERSON, 1976; BROOKS and VANDERVORT, 1977) and will not be further reviewed here. In the PVC manufacturing process, the PVC is not heated sufficiently to release fumes, and smoking is not permitted within the plants, though PVC dust may be carried out of the ucc 057613 5. plants on skin and clothing. 1.5 The Present Study In view of the published reports summarised abov , ICI Ltd. commissioned a study by the Institute of Occupational Medicine (IOM) of the health of workers in a FVC production and processing works. The study was planned to investigate the prevalence of respiratory symptoms, radiographic changes and pulmonary function abnormalities in a sample of the present and past workforce of the factory, and to relfte any abnormalities found to estimates of exposure to respirabLe PVC dust. Men were selected for medical studies from the present workforce of a factory manufacturing PVC, and from men who had retired from the factory, and men who had left for other reasons, in such a way as to include a high proportion of those workers likely to have been exposed to higher concentrations of respirable PVC dust for prolonged periods, together with samples of those men likely to hn/e been exposed to less or no dust. Aa environmental survey of the factory was carried out to measur current PVC dust levels to obtain an estimate of relative dust exposure between different jobs, and a detailed occupational history obtained from the men selected for study. On the basis of current respirable dust measurements and occupational history, an index of PVC dust exposure was derived, and compared with the measurements of respiratory health made in the medical survey. In the medical survey, information about respiratory symptoms was obtained by questionnaire, lung function measurements mad and chest radiographs obtained. The study was agreed beforehand by representatives of management and the unions after detailed discussions with the senior staff of the IOM. 'UH V L- J ucc 057614 6. 2. THE FACTORY PLANTS AND PROCESSES Plants 02 to 06 produce PVC from vinylchloride monomer. Plant Cl, now closed, was similarly employed. In these processes liquid vinylchloride monomer is admitted to autoclaves under pressure, and mixed with water to form either an emulsion or a suspension. The droplets of monomer and particles of PVC finally produced are smaller in the emulsion than the suspension method. Small quantities o' surface active agents and catalysts are used in both methods. The mixture is heated under pressure, the po1ymerisation reaction being complete in about eight hours. The remairing gaseous monomer is vented back to the gasholder. However, further monomer remains within the polymer, and this is stripped from the polymer by heating the slurry under reduced pressure. This is done within the autoclave in the lder plants, and in a separate lower chamber in the newer plants. When the slurry of water and polymer has been discharged the autoclave is cleaned with jets of water. After cleaning the autoclaves in the older plants are entered for inspection. Before the hazards of vinylchloride monomer were known, significant exposure to the monomer occurred during autoclave cleaning, but now the stripping procedure and flushing of the chamber with air ensure monomer concentrations well below strict regulatory control levels in the autoclave section. The PVC slurry passes through pipes to closed stock tanks over which air is circulated and vented to atmosphere to reduce further residual contamination with monomer. Op to this point in the process the PVC particles are wet, and little dust is apparent to the casual observer. The slurry then passes to the drying and packing areas, which are physically separate from the autoclave areas. In the plants producing emulsion polymers drying is in several large spray driers, each of which requires opening and cleaning every seven days, this being an obviously duslyoperation. In the plants producing other grades of PVC, the slurry is centrifuged and dried by standard solids drying operations, which are less dusty. The dry powder is conveyed by air to bulk storage silos or to seek packing machines. Packing can also be a dusty operation in the older plants, particularly in emulsion polymer plants. Information from the manufacturer indicated that the approximate size ranges of PVC particles from the emulsion process were 3 to 100 jjm and ucc 057615 7. from the suspension process 50 to 500 jum. The emulsion particles may be milled further to paste polymer, range 1 to 50 /im. Some simple spot checks by the IOM using the Coulter counter technique were consistent with this information* Part of the PVC produced in these operations is processed further in separate compounding plants (Plants W1, W2, D) in which the PVC powder is mixed with various additives, fillers, and pigments. These plants appear to the casual observer to be generally less dusty than the PVC polymerisation planta, although some dust is generated in the mixers and was*4 -- 41 . ___t w yv--x---a---wt x WU* Other plants in the Works operate in adjacent areas, to produce various chemicalj. These plants were not obviously dusty, and PVC dust levels within them were assumed to be negligible. uco 057616 8. 3. METHODS Population to be Studied Mon wars selected for study on the basis of information fro* factory personnel records. Records of place of work and occupation were reasonably complete except for a six-year peri d between 1966 and 1972. Records of men retiring from work and receiving pensions from the manufacturers were also available, and records of men.leaving the factory for reasons other than retirement were adequate for those leaving between 1966 until the present. The sample to be studied was therefore selected from 1,501 currently employed men, 339 pensioners who had retired since 1st January 1967, and 468 men leaving for other reasons since 1st January 1967. Mo women had been employed in dusty occupations, and none was included in the study. Tot the purpose of selection of men for medical studies, occupations and place of work were allotted to categories of dustiness derived from the preliminary inspection of th plant and measurements of the particle size ranges of typical PVC dusts. There were four such categories of dustiness, A to D, where A included the most dusty occupations and D negligibly dusty occupations* The process, maintenance and distribution work in plants Cl (an emulsion plant shut down in 19~4) and C4 was allocated to category A; process, maintenance and distri bution work in plants C2 and C3 (old suspension polymer plants), and process and maintenance work in plant V (an old compounding plant) ware allocated to category 3; work in plant C5, C6 and plant D, and all occupations in which exposure was slight or occasional, such- as laboratory, quality control and work study jobs were alloeated to category C. Occupations listed without sufficient detail for accurate categorisation were included in this group. In category D were all other jobs, mostly office work. This categorisation naturally involved some generalisations, and it is emphasised that these categories were for selection of men for medical studies, and were not used in the analysis. Detailed employs nt histories were obtained from the m n theme Ives during the study, and actual dust measurements were made subsequently. UCC 057617 9 On the basis of these categories of job, the men were allocated to dust/job groups 1 to V. Men who had worked 10 years or more in category A jobs or 15 years or more in category B jobs were defined as group I. Men not in group I who had worked five years or more, but less than ten, in category A jobs, or who had worked ten year3 or more, but less than 15, in category B jobs were defined as group II. Men not in groups I and II, who had worked one or more years, but less than five, in category A jobs or one or more, but less than ten, years in category 3 .lobs were defined as grouo III. Men no* in groups X, IX or XXX who had worked one or more years in category C jobs were defined as group IV. The remainder, men who had spent no time, or less than a year, in job categories A, B or C were defined as group V. The numbers of men in ach group are set out in TABLE 1.* The Sample * For sampling, the currently employed men, pensioners and other leaver*, were all considered separately. The sampl (TABLE 2) consisted of all those in groups I (332 men) and XI (196 men), randomly divided into two samples, a and b, and c and d, respectively; two random equal samples (e and f) from group III, the sum of the two groups totalling 96 of 501 men (2<3t); two random samples (h and i) from group XV, totalling 30 of 953 men (S&); two random samples (f and k) from group V, totalling 7^ of 326 men (31#). Additional random samples (g and 1) were drawn from groups . XXX' and V respectively, and used as reserves. If any man in groups a to f was unable or unwilling to take part, a reserve from group g was included; similarly group 1 acted as reserves for groups h to k. This intended total sample consisted of 750 men, the most heavily exposed men being highly represented in the sampl . A clerical error resulted in an additional 92 men being fully studied, and a further 158 had chest radiographs taken and spirometry performed. The results for this latter group ar recorded separately; th results for the additional men who w re fully studied are included in the analysis. These ucc 10.. additional men were in low dust/job groups. One-hundred and three reserves from the 'current' men, 31 pensioners and 22 leavers were drawn from sample g to replace men from groups i, 11 and III who did not attend the survey. In view of the 92 additional men studied as a result of the error mentioned above, no reserves were drawn to replace men from groups IV and V. Thus, finally, 841 men were fully studied. Single breath gas transfer factor for carbon monoxide Csee Methods) was measured in men in groups b, d, f, i, k. A error mentioned above (total 322 men). Environmental Survey Current personal exposures to respirable PVC dust were measured by occupation in each plant in which PVC was manufactured or processed. The sampling device used (HAPHIS and MAGUIBE, 1976) samples,a 'respirable1 fraction of total airborne dust- in accordance with the Johannesburg Convention (OHEKSTEIN, i960},, including 5& of all particles with an aerodynamic diameter of 5 pm and a density of 1.0 gm/cm3, and excluding 100# of the particles, above 7.1 pa. * Personal dust sampling was planned to include 180 men, with emphasis on those in the dustier occupations. Preliminary estimates of dustiness were based on subjective observation during a previous visit to the works and on particle size measurements of typical suspension and emulsion PVC dusts. Thus the major sampling effort was directed towards occupations in the drying and packing areas in plants C2 to C6 and mixing areas in plants W1 and W2 (plant Cl is now closed). Plants making emulsion and paste polymer received more study than those making suspension polymer. In the event only 130 reliable personal samples w re obtained; the dust tended to cause failure of the sampling pumps, which required frequent servicing. The numbers of men sampled in each occupation and plant are set out in TABLES 3 to 12 in the Results section. Numbers of men in each shift and numbers of samples taken are also indicated. Office staff w re not included in the survey. Plant A, which proc seed polyethylene, was the only non-PVC plant included in the surv y. 1mu -W, f w1 'ii'.'am 1 1 1-- 1 UCC 057619 ucc 057620 11 Three shifts were worked daily at each plant, one six-hour shift and two nine-hour shifts. Samples were taken on the afternoon shift (1 pm to 10 pm) except for one occasion when a morning shift (7 am to 1 pm) was monitored. Plants C2 to C6 and W1 and W2 were monitored on more than one day to reduce errors due to daily variations in dust concentrations. All personal samples were taken for full shif: periods on Sartorius cellulose ester membrane filters (type 11501), diameter 57 mm, pore size 8,0 pm. Each filter was ins rted into z Cursl1 i c^ (MAGJI3E ct nl., T70) which was fixed to the workman's lapel in a vertical pos: tion. A Caaella personal sampling pump (an 'intrinsically safe' model in appropriate factory areas) was used to draw air through the filters at a flow rate of 1.9 l/m. Sampling rates and revolution counter readings w re recorded at the beginning and end of the shift, and the volume of air sampled calculated. A check was made during the shift to ensure that the instruments were functioning correctly. A small number of static samples were taken in plants C2 to 06 and Wl and V2. The sampling instruments were placed-in selected work areas to measure dust concentrations in the general atmosphere. Filters were weighed twice before and after sampling. Prior to each weighing, the filters were allowed to reach equilibrium with the balance room atmosphere for 24 hours so that their weights were stable. Every tenth filter was retained as a control and not used for sampling, to calculate correction factors for changes in filter weights arising from changes in atmospheric humidity. The airborne dust concentrations for each sample were calculated from the dust weight and the volume of air sampled. 3.4 Medical Survey A medical survey team visited the factory and the a n were seen by appointment. A full occupational history was recorded by a train d clerk who was familiar with the topography and operation of the factory and types of job within each plant. Details of type of work and plant within the plastics factories were recorded with dates. Dates and type of work outside the 12 The Medical Research Council Questionnaire of Respiratory Symptoms (MRC, 1976) was administered by a trained clerk. Th cleric was instructed not to prompt; if a man had difficulty answering yes or no, encouragement was given in the form of a standard sentence and the question was repeated. If the answer was again equivocal, 'NO* was recorded. Forced expiratory volume in one second (FEVt) and forced vital capacity (FVC) were measured using a modified Gaensler spirometer (GAENSLER, 1951; McKERRQW, i960). After a practice VApAAwaWAW. ..itf aMt,U0W vVwUUAW WU4J *WIWU..............J ...A^AA-a.Vi*AWUO were recorded. The largest values of FEV1 and FVC, not necessarily in the same breath, were used for the analysis, and the FEVl/FVC ratio derived from these figures. Ucc 057821 The men were weighed fully clothed, and their standint heights measured. A full-size postero-anterior chest radiograph was ta! a standard technique (95 - 120 kV). The chest radiographs were read independently using ILO U/C International Classification of Radiographs of Pneumoconiosis (International Labour Office, 1972) by tt medically qualified National Coal Board readers highly experienced in this work. The readers were chosen to represent a range of interpretations of the presence of small radiographic opacities, on the basis of systematic inter reader differences recorded in previous work. The radiographs were also read by a chest physician for clinical purposes. When an abnormality was found the man and his general practitioner were informed by letter. If a man's lung function tests were abnormal he was informed of this and advised not to smoke. W .t - _s a . Data were checked, coded, punched on cards, recorded on magnetic tape, re-cheeked and edited. Every occupation within the plasties divisionsite at Hillhouse Works was allotted a PVC dust concentration on the basis of the measured current dust levels, the`results of which are described in the next section. Sixty-six such job/dust categories were derived from the 130 measurements made and from knowledge of the factory conditions. . Details of this allocation are described in Appendix 1. A 'dust index* was derived from the current dust measurements and the occupational history such that the current dust levels in each of a man's occupations in the plastics factory was multinlied by the years he had spent in that occupation, and this figure added to the figures derived from his other occupations in the factory, i.e. n 'Dust index' * Z (number of years in occupation, x mean dust i*1 concentration in occupation J1 where n is the number of occupations an individual has worked in the plastics factory. This index does not represent the man's actual dust exposure, for only present dust levels have been measured, and subjective impressions of airborne dust over the years suggest that dust levels have been higher in the past than those* presently experienced. This index can be considered an approximate estimate of relative dust exposure. It is expressed in years, weighted by present dust levels. For instance a man who had worked for 30 years in an office in the factory had a dust index of zero; a man who had worked over 30 years in the dustiest occupations had a dust index of 90* 14 Criteria for respiratory symptoms, based on the questionnaire were as.follows:Chronic cough; yes to questions 1 or 2 plus yes to question 3 Sputum; yes to question 4 or 5 plus yes to question 6 Period of increased cough and phlegm lasting 3 weeks or more in the last 3 years; yes to question 7a Chest illness during last 3 years which has kept subject from normal activities for as much as a week; yes to . question 11a Dyspnoea; shortness of breath when hurrying on level ground or walking up a slight hill; yes to question 8a Asthma; attacks of shortness of breath with wheezing; yes to question 10a and/or history of asthma; yes to question 12g. , . Men were grouped by smoking category; lifelong non-smokers, currant cigarette smokers (including men smoking cigars or a pipe as well as cigarettes), ex-smokers and 'other' smokers, including present pipe and/or cigar smokers, excluding cigarette smokers. Lifetime cigarette consumption in packs (mean daily cigarette consumption x years smoked x 4^; ) was calculated for cigarette smokers and ex-smokers on the basis of the answers to the detailed smoking history included in the questionnaire. If a man smoked cigars or a pipe, this was not considered in the figure for lifetime cigarette consumption. Data were examined by standard methods including multiple linesr regression and subsequent examination of residuals, and logistic analysis. ucc 057623 15. 4. RESULTS 4.1 Environmental Survey The mean airborne respirable dust concentrations to which 130 workers in various occupational groups were exposed during the survey are shown, by plant, in TABLES 3-12. The occupations are listed in TABLE *3 in order of dust concentration. Respirable dust levels were generally low except for the C4 drying and packing operations and mixing process in plant Wi. The highest mean average respir-.ble dust exposure over any occupation over a shift was 2.83 mg/m3. The distribution of calculated dtat indices (years x mg/m3) for the 818 men studied in the medical survey are shown in FIGURE 1. The concentrations measured by the static samplers w re low compared with those from the personal samplers (TABLE 14), and have not been used in the analysis. 4.2 Medical Survey - Summary of Results Measures of respiratory symptoms, lung function and chest radiographs are typically affected by many factors at once. The effect attributable to any one of these factors (age, dust index, smoking status, ...) depends to some extent on what other factors are also considered to contribute to the response. . Correspondingly, in deriving the levels of significance quoted here for any one variable allowance was made at the same time for the effect of other variables of interest. 4.2.1 PVC Dust Effects 4.2.1.1 Respiratory Symptoms There was an association of complaints of mild breathlessness on exertion with PVC dust exposure index among cigarette smokers (P < 0.02). Chronic cough or sputum, recent acute cheat illness or history of asthma were not related to PVC dust exposure index. 4.2.1.2 Lunr Function Index of PVC dust exnosure was significantly related to reduction of forced expired volume in one second (FEVX) (P < 0.025). ucc 057624 4.3 4.3.1 4.3.2 19. Medical Survey - Detailed Results Features of Population Studied Eight hundred and forty-one men were seen, of whom 23 men were excluded from the analysis because of incomplete records of age, occupational history or spirometry. The results for the remaining 818 men have therefore been analysed. Six hundred and sixty-three men were currently employed at the factory ('current men'); <?8 were pensioners, and 57 men had left for reasons other than retirement ('leavers') (TABLE 15). Current men and leavers were of similar mean age (44.2 and 43.2 years respectively); pensioners were older (63.4 years)A Mean dust indices were similar in curr nt men and pensioners (13.4 and 14.2 units * respectively) and less in leavers (4.9 units ). The current men and pensioners had sp nt mean periods of 14.3 and 16.6 years respectively air the factory 1 the leavers only 5-5 years. The FEVt, FVC and FEV1/FVC ratio were similar in the current men and leavers, and lower in the (older) pensioners (TABLE 15). Numbers of current men, pensioners and leavers are set out by age group and smoking category in TABLE 16. In TABLE 17, all men are shown by these categories, with mean dust indices for the groups. One hundred and forty-eight (18$) of the whole population were lifelong non-smokers; 378 (46&) were current cigarette smokers or smoked cigarettes as well as cigars or a pipe; 58 (1%) smoked cigars or a pipe without cigarettes (other smokers); 234 ( 293S) were ex-smokers. Preliminary Analysis Multiple linear regression was used to screen possibl explanatory variables in relation to response. There was no evidence that years worked at the PVC plant or work in an utside industry in which noxious materials ware used were related to prevalence of symptoms, lung function or radiographic abnormality. * Dust index units are years x mg/m3, but this does not represent actual dust exposure, as only current dust levels are known. ucc 057625 20 after age, index of dust exposure, smoking habits, and (in the case of lung function) height and weight were taken into account. *.3.3 Respiratory Symptoms Analysis of respiratory symptoms was basad on the answers to the questionnaire. The only respiratory symptom showing a relationship with index of ?VC dust exposure was dyspnoea. *.3.3.1 Chronic Cough or Sputum One hundred and fifty-two men (18.6Gb) reported chronic cough for as much as three months in each year. One hundred and thirty-four men (16.*#) reported sputum for as much as thre months in the year. Ninety-four of these men (11.59) r ported both cough and sputum. Chronic cough was positively associated with sputum (P 0.001), and prevalence of chronic cough (irrespective of sputum) and prevalence of sputum (irrespective of cough) each showed highly significant differences b .'tween smoking categories (TABLE 18). The association of cough and sputum was similar in the different smoking categories when the different prevalences were allowed for. Chronic cough alone, sputum alone and cough and sputum have been considered together in the subsequent analysis (called 'cough or sputum'). One hundred and ninety-two men (23.5&) reported chronic cough or sputum. There were marked differences in prevalence of cough or sputum between smoking categories (X3 test, P < .001), and smoking category influenced the apparent relationship of cough or sputum with other variables including age, lifetime cigarette consumption and dust index. For this reason men in the four smoking categories were considered separately by logistic analysis. Non-smokers Only five of 148 men (3*$) reported oough or sputum, and there were no evident relationships with dust index or age. Cigarette Smokers There was a higher prevalence of cough or sputum among cigarette smokers (3**.**) than among non-smokers, and this ucc 057626 .21 effect was related to lifetime cigarette consumption. Logistic analysis showed a strong positive relationship between cough or sputum and cigarette consumption (P < .001). Using the log of lifetime cigarette consumption in the logistic function gave a better fit with cough or sputum, and, for these do4,a, a simpler mathematical expression for the relationship*. Using this form of relationship between cough or sputum and cigarette consumption, the age effect was no longer appar nt and there was still no effect of index of dust exposure. Qther_Smokers In pipe and cigar smokers there was a greater prevalence of cough or sputum O'#) than in non-smokers, similar to that ,in*cigarette smokers. No relationship with years smoked, dust, or age could be demonstrated. Ex-Smokers Logistic modelling taking account of age, lifetime cigarette consumption and dust simultaneously, showed a relationship with age (P < 0.05), a relationship with cigarette consumption which did not reach significance (P 3 0.11), and no dust effect. (Lifetime cigarette consumption was taken as zero in cigar and pipe smokers.) *i.3 32 Asthma No relationship was demonstrated between a history of asthma and-index of FVC dust exposure. Ninety-two men (11.2S) reported suffering from asthma or Fitting the logistic function: p(x) 3 ;p-U+,*> 1 + exp (a + b log x) gives b = 1.008, very close to 1. Taking b as 1, the expression becomes P(x) c, * " 1 . C x X>0 C = ea C is estimated as 0.06. Thi* fit is illustrated in FIGURE 2. Where P(x) is the probability of cough or sputum as x varies x = lifetime cigarette consumption (units = 1,000 packs), a, b being unknown constants estimated by the curve-fitting procedure (using maximum likelihood methods), UCC 057627 .22 attacks of shortness of breath with wheezing. More men in the smoking groups (cigarette smokers (12.70, other smokers (IJ.QU), ex-smokers (12.8^6) ) reported these symptoms than in the non smoking group (*#;). The differences between the four categories are significant (Xs test, P < 0.04). Logistic likelihood analysis confirmed the differences between smoking groups at a sharper level of significance (P < 0.01), but no effect of age or dust was apparent. There was no discernible interaction between smoking and PVC dust. **3.3.3 Dyspnoea There was a significant slight association of breathlessness on exertion with index of dust exposure. Two hundred* and fifteen men (26.3%) admitted to dyspnoea on exertion grade I (* shortness of breath when hurrying on level ground or walking up a slight hill'). Prevalence was approximately three times higher in the smoking groups (32$ in cigarette smokers) than in the non-smokers (1CK). Linear logistic analysis confirmed this (P < 0.0001), and shoved a strong age effect (P < 0.00001). A dust effect approached significance (P a 0.05**) after adjustment for age and smoking, and this effect was pronounced in current cigarette smokers (P < 0.02) after allowing for age and lifetime cigarette consumption, but not present in the other smokine categories. Prevalence of more severe grades of dyspnoea showed a strong relationship with age and a weak relationship with smoking, but no dust effect. **.3.3.** Acute Chest Illnesses No relationships were demonstrated between acute chest illness and index of dust exposure. A history of recent periods of increased cough and phlegm was mors common in cigarette smokers and other smokers than non-smokers or ex-smokers, though this difference only approached significance (X* test, P = 0.063; likelihood test, P < 0.03), and no effect of index of dust exposure or age was apparent after smoking differ nces had been taken into account. ucc 057628 4.3.4 4.3.4.1 23 Prevalence of a history of recent chest illness was not significantly related to smoking, age or index of dust exposure. Combination of these two features of chest illness showed no significant relationships with smoking, age or index of dust exposure, nor did the inclusion of positive histories of pneumonia, pleurisy or (acute) bronchitis. Lung Function FEVt and FVC were significantly inversely related to the index of PVC dust exposure. Forced Expired Volume in One Second (F5V,) Preliminary analysis indicated the anticipated relationship between FEVl and age and smoking category (FIGURE 3)* cigarette smokers having the greatest loss of FEV^ with age, non-smokers the least. Dust effect was studied after standardisation for age, height and weight by multiple regression. Pensioners wer found to have significantly lower mean FEV* after age, height, weight and dust index had been taken into account than current men or leavers (FIGURE 4), though the slope of the regression for dust index was similar in all three groups. For subsequent analysis the slopes of the regression for FEVj with dust index were constrained to be the same for current men, pensioners and leavers: the intercepts for current men and leavers were constrained to be the same, but the intercept for pensioners was allowed to vary independently. Introducing different.intercepts for smoking categories and additional regression for lifetime cigarette consumption into the model did not seriously alter the overall regression with dust index, though there was a significant inverse relationship of FEVX with cigarette consumption (P < 0.0001) (TABLE 19). There was a significant negative effect on FEV1 of increasing dust index (P < 0.025) after adjustment for age, height, weight and smoking. Further expansion of this model by allowing different slopes of regression of FEV1 with dust index for the different smoking categories (TABLE 20) showed that the dust effect was most marked in cigarette smokers, and not apparent in non-smokers, though , ucc 057629 24 the magnitude of difference in slopes between these two groups is just significant at the 1CE6 level. The intermediate values of the slopes for ex-smokers and other smokers provide slight additional evidence that the dust effect varied according to smoking status. The effect was significant among smokers (P < 0.00$), and was not demonstrated in non-smokers, but this may partly be the result of their lower mean dust exposures (TABLE 1?), and small number of men (148). More detailed modelling allowing different regression of FEVt with age for each smoking grown confirmed a dust effect in cigarette smokers, and also indicated a slight but statistically non-significant effect of dust in the non-smokers. In conclusion an overall effect of PVC dust was demonstrated, and suggestive evidence of differences between smoking categories was identified. Addition of chronic cough or sputum, and recent chest * illness to the model (with a single regression for dust effect) indicated that each of these was associated with a lower level of FEVt (P < .001, P < 0.002 respectively). Association of _ radiological abnormality with FEVt is reported in the section on chest radiographs* Magnitude of FEV! Loss After taking into account the effects of age, height and weight, the overall regression of PEV1 with dust index was -0.0040 litres per dust index unit (years x mg/m3). This was equivalent to a loss of 52 ml of FEV^ for the mean dust index. As an illustration this would be equivalent to a loss for all men of 5.2 ml of FEV1 per year during exposure to the mean dust index over a period of ten years. In cigarette smokers considered separately this loss was 8.9 ml of FEVj per year for similar exposure, additional to the loss caused by smoking and age* It is emphasised that dust index does not represent actual dust exposure for it is based on current dust levels only. This may be compared with the mean-losses due to age (35 ml per year) or smoking one pack of cigarettes a day (6.2 ml per year). ucc -XT*** 057630 25. *+.3.4.2 Forced Vital Capacity (FVC) FVC was significantly inversely related to index of PVC dust exposure among currently employed men. Preliminary analysis indicated the anticipated adv rse effects of age and smoking category on FVC (FIGURE 5). Multiple regression techniques, allowing different intercepts and different slopes of regression with dust index for current men, pensioners and leavers showed a negative association between dust index and FVC among the current men (P < 0.05) (TABLE 21). Pensioners had a lower mean level of FVC after adjustment for age, height and weight for any dust index, and no dust effect was apparent. Leavers had a similar mean level of FVC to the current men but no dust effect (FIGUBS 6). The pattern of dust effects among smoking groups was similar to that for FEVl, though the differences did not approach significance at conventional levels. 4.3.4,3 Satio of Forced Expired Volume in One Second to Forced Vital Capacity . tFSVyFVC ratio) The FEV1/FVC ratio was not significantly related to index of PVC dust exposure. Preliminary analysis of FEV1/FVC ratios demonstrated age and smoking effects (FIGUHE 7). Multiple regression for log(FEVl/FVC) in the same manner as for FEY. and FVC confirmed age, height and weight effects, but dust effects did not reach significance. Pensioners did not have significantly lower FEV^/FVC ratios. The introduction of chronic cough or sputum and recent chest illness into the model indicated a significant association of chronic cough or sputum with lower FEVl/FVC ratio (P < 0.001) (TABLE 22). 4.3*4.4 Gam Transfer for Carbon Monoxide (TL^q) TLgQ, Kgg and V^eff were not significantly related to ind x of PVC dust exposure. Multiple regression for TL^ in 322 men showed significant age, weight and smoking effects but no dust effect. Pensioners had lower TL^^ adjusted for age, height, weight, and smoking than other groups bux no dust effect (TABLE 23). ucc 057631 26 Diffusion constant, K^, showed significant inverse relationships with age and height for all men, significant smoking effects but no dust effect. Alveolar volume, V^eff, showed age and height effects but no dust effect. Pensioners had lower V^eff adjusted for age, height and smoking. 4.3.5 Chest Radi^yranhs A significant relationship of the prevalence of small rounded opacities with index of PVC dust exposure was found by one reader only. Results for the three expert readers are described separately. The anticipated differences between readers in interpretation of small opacities were apparent (TABLE 24). ^ 4.3.5.1 Prevalences cf Small Opacities e The prevalences of small rounded and small irregular opacities of 0/C category 0/1 or greater, and 1/0 or greater, are set out in TABLE 25. Reader 1? found the greatest pr valence of small rounded opacities; 30 radiographs <6.IK) with category O/l or more, ten (1.2#) with category 1/0 or more. (These ten consisted of six radiographs with category 1/0 and four with category 1/1.). Reader 15 found more small irregular opacities than the other readers; 53 radiographs (6.5#) with category c/l or more, most of whom were classified as 1/0 or more (46, 5.6# of the entire sample) (TABLE 24). Of these 46 radiographs, 16 were 1/0, 20 were l/l, 1 was 1/2, 6 were 2/1 and 3 were 2/2. For two readers (03 and 1?) there was s slight tendency for films classified as showing ant type of small opacity to be classified as showing the other type as well. Agreement between readers in classifying type of small opaeitias is sat out in TABLES 25 and 26. 4.J.5.2 Small Rounded Opacities - Category 0/1 or Greater Reader 17 found a significant relationship betwwen prevalence of small rounded opacities category 0/1 or greater and index of PVC dust exposure. ucc 057632 27. Preliminary analysis of the data for readers 03 and 1? showed a relationship between the prevalence of small rounded opacities and age (FIGURE 8). The results for reader 03 were dominated by a relatively high prevalence (7-9) in the oldest age group. The results for reader 17 showed a progressive increase in prevalence of small rounded opacities with increasing age, the highest prevalence being 11.4# in the old .st age group. Results for reader 03 showed no relationship with dust index (FIGURE 9), but reader 17 found an additional relationship of . w****.^^ _ 4* ******** 4* ******** <*** " * ** 1 VS age was taken into account (FIGURE 10). The 2ero prevalence of small rounded opacities category 0/1 or more in the lowest dust index category (TABLE 27) does not necessarily indicate that dust exposure is required for the development of these opacities. The subsequent logistic analysis indicated a strong age effect independent of dust effect (for two readers). Reader 15 classified only four radiographs as rounded opacities 0/1 or more. These showed no striking relationships with dust or age, and have not been analysed further. Results for readers 03 and 17 were examined by logistic analysis. Reader 03 found a clear relationship of the prevalence of small rounded opacities category O/l or more with age (P < 0.002), and no relationship with dust when age was taken into account (P s 0.S3). The fitted model indicated a prevalence of 1.52 for men aged 45 years, subsequently increasing approximately twofold for each additional ten years of age. Smoking category did not influence these results. Reader 17 found highly significant relationships of small rounded opacities category O/l or more with both age and dust index. Age and dust index were independently related to prevalence of opacities. The relationship with dust index after allowing for age was significant at P < 0.001 (FIGURE 10). The relationship with age after allowing for dust index was significant at P < 0.001. The fitted model estimates a prevalence of small rounded opacities category 0/1 or more of among 45-year-old men with Len dust index units. This estimate increases by a factor of approximntel 5/3 for each additional ten years of age, and (separately) by a factor ucc --------------------- .. _ -.................. --- - --------- 057633 .28 of approximately 4/3 for each additional ten dust index units. An alternative representation of the fitted model is illustrat in FIGURE 11. Predicted prevalence of rounded opacities category 0/1 or more is plotted against dust index for differ* age groups. Reader 17's actual readings for all ages are al; shown. This model predicts a ?& prevalence of small round' opacities category 0/1 or more for 65-year-old r.en never expo to PVC dust. Twenty PVC dust index units increases this prevalence at age 65 to 13.^6 - Age and dust effects for reader 17 ware nn . the same fm smoking categories. The dust effect was seen in th ex-smo! and 'other* smokers, among whom no age effect was seen* Am non-smokers, who were younger,'and of whom only three of 143 had small rounded opacities category 0/1 or more, an age effect was indicated but notconfirmed, and the dust effect v. doubtful. Amongst cigarette smokers there was a clear relationship with age, and no dust effect. The reasons fo: these differences were not apparent, and the results presen are those for all smoking categories combined. Age for age the pensioners had a'lower prevalence of s rounded opacities when age and dust had been taken into ace Relationship of Respiratory Symptoms to Prevalenc of Small Rounded Opacities " There was no avtaenca chat tne presence of chronic co sputum, or recent acute chest illness was related to the p of small rounded opacities when age and dust had b en take account. 4.3.5*3 Small Rounded Opacities - Category 1/0 or Greater Reader 17 found ten radiographs showing small roundec opacities category 1/0 or greater, and in spite of this sr number, a significant relationship with both age (P < 0.0 dust index (P < 0.02) (after adjusting for the other) was The estimated prevalence of small rounded opacities l/O or more for men aged 45 with ten dust index units was the prevalence increasing by a factor of approximately each ten years of life and approximately if for each fur- dust index units. UCC 057634 29 4.3.5.** Small Irrerular Qnacities - Caterory 0/1 or Grocer No significant rel?'*"1 .---ipo were found between prevalence of aprsV j-regular opacities and index of FVC dust exposure. Preliminary analysis showed a clear relationship betw en prevalence of small irregular opacities category 0/1 or greater and age for all three readers (FIGURE 12). There appeared to be no or little relationship with dust index (FIGURE 13). Logistic analysis confirmed the age effect for all three reader* (reader f>3: P < (1.001; reader 15; P < O.OOOOI r reader 17, P < 0.001). In the case of reader 15, for example, the estimated prevalence among men aged 4$ years was 2.&S. This increased by a factor of three (approximately) for each additional ten years of age* There was no relationship of prevalence of small irregular opacities with dust index for any of the three readers. For reader 15 the relationship of small irregular opacities with age could in small part be explained by a positive history of chronic cough or sputum (P < 0.03), and by history of chest illness. The criteria of chest illn ss, in this instance only, included a history of pneumonia, pleurisy or (acute) bronchitis, or a period of increased cough and phlegm lasting three weeks or more in the last three years, or a ch st 11literal during the lest three y**rs which h*d kept the subject from his normal activities for as much as a week. Pr valence - of small opacities was related to cheat illness by these criteria (P < 0.03). Men with chronic cough or sputum, or men with a history of chest illness ware each approximately twice as likely to have small irregular opacities as those without these features. Chronic cough or sputum and history of chest illneas were to some extent alternatives to each other: the inclusion of each in the model reduced the significance of the other, while a positive history of both waa more strongly associated with small irregular opacities. Reader 17 also found a relationship between small irregular opacities and chronic cough or sputum (P a 0.061), but no relationsh with history of chest illness. For both readers, the age effect remained extremely strong even when these symptoms were taken into account. ucc 051635 30. Pesder 03 found differences between smoking categories. Prevalence oi '-=11 irregular opacities was lowest in non-smokers (0.7%) and highest am-^.g w>... cri (12.1%). Ex-smoker= (3.8%) were similar to cigarette smokers v. ' Thc-j differences were formally significant (P = 0.021) after allowing for differences in age distribution between the smoking groups. There was no dust effect. Neither of the other two readers identified smoking differences. 50 omaii irregular opacities - Category 1/0 or Greater Analysis of readings of small irregular opacities eat gory 1/0 or more (TABLE Z1* shows prevalence) confirmed the age effects and did not show any dust effect. Besults for reader 15 confirmed the relationship of small irregular opacities with history of chronic cough and sputum and/or acuta chast illness (P < 0.02). Reader 17*s results indicated a relationship with chronic cough and sputum (P * 0.054). 4*3.5*6 Relationship of Lung Function to the Preaanca of Small Rounded Opacities Men with small rounded opacities had significant reductions of FEVl, FVC and V eff. Reader 17 found small rounded opacities category 0/1 or greater in 50 radiographs. The lung function of th*** men was significantly reduced. The FEV1 was reduced by a mean of 219 ml (P < 0.01) after age, height, weight, smoking, cigarette consumption and dust index had been taken into account. The FVC was reduced by a mean of 264 ml (P < 0.01), using the same model, and FEVj/FVC ratio not significantly different. TL^ and Kqq were measured in 21 of these men, and ware not found to be significantly reduced* V^eff was reducad in these 21 men by a mean of 328 ml (P < 0.05)* The same reader found small rounded opacities category 1/0 or more in ten radiographs. In these ten subjects the mean FEVt was reduced by 293 ml but this difference failed to reach significance (P < 0.2) because of the small numbers of men. Similarly the FVC was 323 ml low r but this difference also failed to reach significance (P < 0.2). The TLg0 was measured in only one of these men* ^ , 051636 31- These results suggest that the men with small rounded opacities had an associated impairment of lung function, predominantly of obstructive pattern, 4.3.5-7 Relationship of Lung Function to the Presence of Small Irrerular Opacities Hen with small irregular opacities in the cheat radiograph had significantly reduced FEVt, FVC and FEV1/FVC ratio. Radiographs of 90 men were thought by one or more readers to show small irregular opacities category O/l or greater. The lung function of these men was significantly reduced. Their mean FS.^ was reduced by 210 ml after adjusting f r age, height, weight, smoking category, lifetime cigarette consumption and dust index (P < 0.01). The mean FVC was reduced by 147 ml, using a similar model (P < 0.05). The log of the ratio FEV1/FVC .was also significantly reduced (P < 0.01). TL^ was measured in 34 of these men, and was reduced by a mean of 1.59 ml/min. am Hg (P < 0.05). and ware not significantly different. Radiographs of 52 men ware thought by one or more readers to show small irregular opacities category 1/0 or greater. The mean FSVj of these men was reduced by 325 mi (P < 0.001) using a similar modal to that used above. The mean FVC was reduced by 276 ml (P < 0.01). The log(FEV1/FVC) ratio also was significantly reduced (P < 0.01). TL^q was measured in 25 of these men, and was reduced by a mean of 1.97 al/min. am Hg (P<0.05). These findings suggested that the presence of small irregular opacities was associated with an obstructive lung defect, and that greater profusion of opacities was associated with more severe airflow obstruction. 4.3.6 Autoclave Workers Autoclave workers may have been exposed to significant concentrations of vlnylchloride monomer in the past. The effect of a history of being an autoclave worker waa examined in the respiratory symptoms, lung function and chest radiographic abn rmality models. With one exception, a history of having been an autoclave worker had n influence on symptoms, lung _- ........ UCC 057637 Lxn *i 32. function or chest radiographic abnormalities. The exception was in the presence of small rounded opacities category 0/1 or more found by reader 03; men who had worked at some time on the autoclaves had a significantly lower prevalence of small rounded opacities category 0/1 or more, after age and PVC dust effects (P < 0.04). This was not the result of differences in smoking habi. Thus apparent PVC dust effects were not the result of exposure to vinylchloride monomer during autoclave working. The readings of the c',,eat radiographs of l8l men n t included in the analysis are given in TABLE 28. These comprise the radiographs of 23 men excluded from the analysis because of inadequate data, together with 1?8 mep whoee chest radiographs were taken but who were not intended to be part of the selected study population. The prevalences of small opacities were similar to those of the study population. 4.3.8 Clinical Readings of Chest Radiographs As a result of the readings of chest radiographs by two chest physicians for. clinical purposes, 84 men were referred to their general practitioners. Most of these referrals were for minor abnormalities of the chest radiographs, particularly pleural shadows consistent with post infections, apical shadows consistent with inaetiv or healed tuberculosis, probable nipple shadows, and slight diffuse shadowing in the lungs. Cardiomegaly was also observed. In six men the appearances were thought to require clinical investigation to exclude or confirm malignant disease. At the present time, serious disease has been excluded in four m n, one has refused further investigation and one is still undergoing investigation. The chest radiograph of one man showed advanced diffuse pulmonary fibrosis. This man had a chest radiograph taken though he was not in thq,intended study sample. He had worked in plants Cl and C4 for 32 years, though not in employment clos to the process. ' He was asymptomatic, though on examination he ucc 051638 33. had clubbing of the fingers and many crepitations were audible over the lower parts of his lungs, He was referred to his general practitioner for investigation, and it is hoped that further information 'will be available later. ucc 057639 3^. 5. dtscudsion A survey of the literature has shown that there is some evidence that inhalation of PVC dust might be associated with the development of respiratory disease, expressed as radiographic changes, respiratory symptoms or functional abnormalities (VERTKIN et al,, 1970; MILLER et ail.. 1975; LILIS et al., 1976; ARNAUDet al., 1978; MAPPet al., 1978). However, one euidemiologieal study failed to demonstrate any functional abnormality related to PVC dust (GAMBLE et al.. 1976). The results of this and the previous studies were equivocal, and this has generally been Lccaua* eucli sympluins, dysfunction and radiological changes have multiple potential causes, of which inhalation of PVC dust is only one. It is necessary therefore in studying the effect of inhalation of PVC dust to take into account other factors known to cause respiratory symptoms, radiological changes and reduced lung function. Moreover, in studying any working population it is desirable to include those workers who have left the labour force either to move to other jobs or on retirement, as such men may include disproportionate numbers of less healthy men (FOX and COLLIER, 1976). This study was planned to sample the present workforce and those who had previously worked in the factory. It was not possible to study the entire population and so a random sample was selected, weighted to include a relatively high proportion of those who had the highest lifetim exposure to PVC dust. The initial selection was based on rough estimates of past exposure, but analysis of the effects of exposure to PVC dust was based on more accurate estimates calculated from a careful occupational w history and on measurements of current exposure to respirable PVC dust. Nevertheless, these estimates remain inexact, as it was not possible to obtain measurements of oast dust exposures in the factory. Insofar as it is probable that earlier levels were higher than those obtaining at present, our exposure estimates art probably rather lower than the actual historical values especially for those men with the longest working history, and this should be borne in mind when drawing conclusions about dust/disease relation ships. The assumption was also made, and is likely to have been valid, that the ratios of exposure between difference jobs had remained approximately constant. In the selected sample of the population, the study was planned to measure the prevalence of respiratory symptoms, the levels of lung function ucc ______________________________ ____________________ __________ h057640 35. (both of the airways and of the peripheral, gas-exchanging parts of the lung) and the amount of radiological abnormality, and to relate these findings to the estimates of exposure to repairable PVC dust. In so doing it was necessary to allow for the eifects of other factors that were known to affect, or were suspected of affecting, these measurements. In the case of symptoms and respiratory function, age and cigarette Broking (and size with regard to lung function) are known to have important effects, while in the case of radiological changes it was thought probable that, observer differences and effects due to age and snrking might be observed. The analysis was therefore planned to take tbiao factors into account as well as the possible effects of PVC dust exposure. In the ease of this epidemiological study, a clinical reading of the chest radiograph was carried out, as in any such study a proporti n of msn are found with serious and usually*coincidental disease that requir s medical attention. Among those found to have abnormalities was one with pulmonary fibrosis. There is no way of knowing at present whether this man's disease was related to his work, but he is being investigated further clinically and it is hcped to acquire further information lat r. No excesa of other disease, such as lung cancer, was apparent in this study. We found a relationship between estimates of exposure to PVC dust and symptoms, lung function and radiological abnormalities. In the cas of symptoms, tha relationship was between exposure Lo dual la cigarette smokers end grade I (or slight) dyspnoea. This relationship was not present in non- or ex-smokers or in other smokers. The effects of age and cigarette smoking on this symptom were, as anticipated, much stronger and it seems unlikely that the effect of PVC dust on symptoms alone is of clinical significance. However, the relationship between PVC exposure estimates and respiratory function showed that inhalation of the dust has an ffect on the FSVt that is comparable to that of smoking, though not as pronounced as that of age. The effect of PVC dust exposure on the FEVt was seen strongly in cigarette smokers, and not confirmed among aon-smokers considered separately. ucc 057641 36. and the pattern of the regressions for dust effect in the different smoking categories suggested that there were real differences in FVC dust effect according to smoking habit. This suggested interaction between FVC dust and smoking might be caused by a combined effect of dust and smoking acting separately on the airways, or possibly to PVC pyrolysis products released from cigarettes contaminated with PVC dust. Smoking was not permitted within the factory plants, but the dust could be carried out on skin and clothing. Respiratory symptoms are known to occur in workers cutting PVC film with hot wire (SOKOL et al., 1973), and it has been suggested that these effects are more pronounced in smokers (FALK and PORTNOY, 1976). While this suggested interaction between smoking and effect of PVC dust on lung function could be caused by either of these mechanisms, the effect of PVC dust og the chest radiographs was not clearly related to smoking, and therefore presumably not related to PVC pyrolysis products. * . We only measured respirable dust, but an effect on the airways could also have been due to dust of larger particle size, also present in the air, and which would be deposited less peripherally in the lung than the respirable fraction* Measures to reduce respirable dust will of eours reduce dust of larger particle size as well. This effect on FEV1 is important. Although the mean decline with exposure is small,within the range there are likely to be workers whose decline, by analogy with the known response to cigarette smoking, is of clinical significance. The effect of dust exposure on FEV1 was mirrored by a similar small effect on FVC. However, no effect on transfer factor or alveolar volume was found. This suggests that dust exposure has a predominant effeot on airways and that there is no evidence of an effect, such as lung f. m + fibrosis, on the gas-exchanging parts of the lung from this study. This is in keeping with what has so far been reported from human and animal pathological studies, where pulmonary fibrosis has not been a prominant finding. The effect of dust exposure on airways may be a non-specific effect, related to the mass and number of dust particles, or it may be a specific effect due to a particular property of the dust. Which of these is correct must await further pathological and experimental studies, but some evidence may be adduced from the radiological findings. Interpretation of chest radiographs is a largely subjective exercise, large variations occurring between different readers. We have chosen to ucc 057642 37 usa those readers trained specifically in the assessment' of radiographs according to the ILQ V/C classification (International Labour Office, 1976), and known to 3how little intra-reader variability. They did, however, show some variability between themselves, which had been apparent in previous studies and was seen in this work. All readers detected small shadows, both rounded and irregular, in a small proportion of films. Almost all were of a low category and there was no evidence of advanced pulmonary fibrosis in the survey population. The small irregular opacities were clearly related to age by all three readers, and two readers found them to be related to cough and sputum. This, taken with the finding of a reduced FEV, in these men, suggests that age and bronchitis are '.raportant factors in the development of these opacities. Small rounded opacities were found to be related to age by two readers, but one of these also found a relationship to dust exposure. Again, men with small rounded opacities 3howed a reduction in their FEVl and FVC and V^eff. The transfer factor was not significantly reduced. This suggests that the reduced ventilatory capacity was due primarily to obstructive (bronchial) lesions, but with an additional restrictive component (small airways or alveoli). UCC 057643 38 6. CONCLUSIONS AMD PUCCI CATIONS We have thus found the clearest evidence yet obtained that inhalation of PVC dust is associated with respiratory dysfunction and radiographic changes. We have not found evidence of a serious progressive form of pulmonary fibrosis, but we have shown that radiographic abnormalities occur in a proportion of exposed men and that they are associated with measurable abnormalities of airway function. This suggests that the effect is a specific one on airways associated with structural changes. We have also shown that the effect of PVC dust inhalation on lung function i *P ^ -mv wu** **t4iw* u a *4* - ___i i v wm w^wmityeia Miu ap^A, UAiiUauolJf </ f (Ji Uiittl# of ageing. While these results are not unduly alarming, they do indicate that further measures to reduce individual dust exposure would be prud nt. Moreover, consideration should be given to monitoring the ventilatory function and radiographs of those men in the highest exposure categories. * Clinical follow-up of these men wjlth radiographic abnormalities is desirable in order to determine whether ths condition is progressive or indeed if it regresses. ucc 057644 39. acknov.i^dg:`~:t3 Thi3 3tudy wa3 carried out with the full co-operation of the management and workforce of ICI Ltd,, to whom we are grateful for financial support. We are grateful to Dr. J.A. Dick, Dr. J.CS. Bennett and Dr. D.J. Thomas for the epidemiological radiographic readings and to Dr. A. Seaton, Dr. M. Jacobsen and Mr. J. Dodgson for much advice and help. ucc 057645 Group I; * 10 years or sore in category A jobs or 15 " '* " '* B " Currently employed Pens:oners Leavers Total men 12*i '5 156 29 2fi0 Hr 6 145 2 18? 8 332 Group II: Not in group I 5 years or more, but lees than 10 years in category A jobs or 10 years or m:.re, but less than 15 years in category B jobs 44 9 94 25 138 :*4 18 71 6 125 24 196 Group III: Not in groups I or II 1 year or more, but less than 5 years in category A jobs 95 2 53 150 or 1 year or more, but less than 10 years in category jobs 18? 51 113 351 e 282 53 166 501 Group IV: Not in groups I, II or III 1 or more years in category C jobs - 601 1*8 204 953 Group V: No time in A, B or C (= remainder) 200 eo 66 326 Totals (population) 1,501 359 *t68 2,308 ucc 057646 TABLE 1 Sampling Frame: groups of .n n by dust/job category. Group I (all): Sample a "h Group II (all): Sanole c "d Group III (3& sample); Sample p "f Reserves g were drawn from this group. Group IV (5# sample): Sample h "* Group V (33# sample): Sample j "k Reserves 1 were drawn from this group. Total (10 samples, excluding reserv samples) Currently employed Pens:oners Leavers Total men 168 26 112 '8 280 *4 5 3 8 332 84 20 14 5*t '4 10 138 ;4 24 196 33 6 20 22 4 13 55 '0 33 98 19 5 6 13 3 4 32 8 10 50 2? 8 19 .5 9 6 46 13 551 1C9 15 74 90 750 ucc 057647 TABLE 2 Sampling from total population of 2,308 men. 4^ Occupational Group Autoclaves Driers Packers Recovery operator Fork truck driver Drier forerun Cleaner Maintenance Numher of I'en per Shift Number of Ten Sampled Mean n'esnirahle Dust Concentration Standard Deviation (mc/m3) 13 4 67 9 14 11 22 11 11 93 0.46 - 0.18 2.59 - 1.56 2.38 1 1.84 0.65 0.89 0.94 1.19 0.84 t o.48 TABLE 3 Dust concentrations in occupational groups in Plant C4. Occupational Group Number of Men per Shift Number of Men Sampled Mean Respirable Dust Concentration - Standard Deviation (mrc/ra3) Autoclaves Driers Packers Fork truck driver 82 35 57 11 T *v / ___________ ________________________ 0.1? 0.54 - 0.31 0.60 t 0.53 0.70 m a ^-- Va* 4l^^' TABLE 4 Dust concentrations in occupational groups in Plant C2 and C3. UCC 057648 Occupational Group Autoclaves Driers Packers (including polymer recovery) Fork truck driver Foreman Number of Men per Shift Number of Men Sampled Mean Resnirable Dust Concentration - Standard Deviation (mo/m3) 19 0 6 4 0.50 i 0.28 9 7 0.38 - C.12 11 0.29 11 0.35 TABLE 5 Dust concentrations in occupational grouns in Plant C5 hk. Occupational Group Preraix team Mixing unit operators Foreman Fork truck driver Cleaner Number of Men oer Shift Number of Men Sampled Mean hesnirable Dust Concentration - Standard Deviation (m<Vnr ) 74 9 17 l1 21 11 1.41 * 0.74 1.86 1 1.27 0.49 0.36 0.72 TABLE 6 Dust concentrations in occupational groups in Plant Wl. Occupational Group Premix team Dry blend units Unit controller Ground floor team Fork truck driver Cleaners Supervisor Number of Men per Shift Number of Men Sampled Mean Resnirable Dust Concentration - Standard Deviation (mm/rr ) 6 4 1.04 t 0.48 6 5 0.86 t 0.34 22 0.57 4 6 0.37 - 0.20 10 - 22 11 0.23 0.61 TABLE 7 Dust concentrations in occupational groups in Plant W2. Occupational Group Assembly Presa operator Polishing Fork truck driver Trimmer Mill cutter Premix operator Number of Number of Mean Respirable Men per Men Dust Concentration Shift Samnled (mtr/m3) 41 21 11 11 11 11 11 0.13 0.38 0.49 0.22 0.14 0.37 0.10 TABLE ft Dust concentrations in occupational groups in Plant D. ucc 057649 **5- Occupational Group Number of "en per Shift Number of Ken Sampled Kean Hesnimble Dust Concentration - Standard Deviation (mcr/nr'') Testers .. 6 3 0.34 - 0.11 TABLE 9 Dust concentrations in occupational groups in lrmoratory (Plant Wl). Occupational Group Chemists Testers Number of Men per Shift Number of Ken Sampled Kean Respirable Dust Concentrr.tion - Standard Deviation (nut/m3) 32 0.19 3 3 0.2? - 0.17 TABLE IQ Dust concentrations in oecunational grouns in laboratories (Plants C2/C3, C5/C6). Occupational Group Number of Men per Shift Number of Men Sampled Kean Respirable Dust Concentration - Standard Deviation (mff/m3) instrument technicians 12 3 0.52 - 0.30 TABLE 11 Dust concentrations in the occupational group at the central workshops. Occupational Group Mixer operators Packers Number of Men per Shift Number of Men Sampled Mean Resoirable Dust Concentration - Standard Deviation (mst/a3) 3 3 0.35 - 0.05 21 0.47 TABLE 12 Dust concentrations in occunational groups in Plant A (not PVC) ucc 057650 Occupational Group Mean Dust Concentration (mg/m3) C4 packers 04 driers W1 mixing operators W1 premix team C4 cleaner W2 premix team 04 drier foreman C4 fork truck driver W2 dry blend units C4 maintenance C2/C3 fork truck driver W1 foreman, cleaner C4 recovery ^operator W2 sunervisor C2/C3 packers C2/C3 driers Central workshops C4 autoclaves C5/C6 driers Plant A (non-PVC) packers C5/C6 packers ^rvuUU HUOt Plant A mixer operators (non-PVC) Semi-technical laboratory Plant D (all groups) W2 cleaner C5/C6 fork truck driver All C plants shift laboratory (testers) All C plants shift laboratory (chemists) C2/C3 autoclaves 2.88 2.59 1.86 1.4i 1.19 1.04 0.94 0.94 0.86 0.84 0.70 0.69 O.65 0.61 0.60 0.54 0.52 0.46 0.50 0.47 0.38 0.37 0.35 0.34 0.27 0.28 0.29 0.23 0.19 0.23 TABLE 13 Occupational groups in order of dust concentrations. ucc 057651 4? Plant C4 C2/C3 C5/C6 . W1 Location Driers Packing Packing Packing Packing Packing Packing Premix Unit Mixer Unit Dust Concentration (mr/rrr5) 0.19 0.22 0.15 0.15 0.11 0.13 0.16 0.29 0.06 TABLE *|4 Mean dust concentrations measured uointr static aampl rs in selected areas in Plants Cl to 6 and W1. ucc 057652 48 Number Mean age (years) Mean he.ght (cm) Mean weipht (Kg) Kean dfcrt index* Years at the plant Mean FEVl (litres) Mean FVC (litres) MeanFEV1/rvc * Current Men Pensioners Leavers 663 98 57 44.2 63.4 43.2 173.0 170.2 171.2 77.7 75.6 77.4 13.39 14.23 4.86 14.34 16.63 5.46 3.54 4.62 2.43 3.41 3.53 4.54 0.76 0.70 0.78 TABLE 15 Features of 8l8 men studied, according to currently employed, pensioner or other leaver status. *Duat index is exnressed in years x mg/m3, but does not represent actual du3t exposure, since only current dust levels are tenown. ucc 057653 49 Age Groups (vr c) < 35 35-44 45-49 50-54 55-59 60 + Total Current men Non-smokers 48 35 11 14 5 6 119 Cigarette smokers 59 88 54 55 44 12 312 'Other* smokers . 3 10 12 7 11 4 47 Ex-smokers 32 4a 34 37 25 15 185 -1 4 -it 9c *rj CCt t WWy Pensioners Non-smokers 0 0 0 0 0 13 13 Cigarette smokers 1 0 2 3 7 28 4l 'Other* smokers 0 0 0 1 0 7 8 Ex-smokers 0 1 0 3 5 27 36 Total 1 1 2 7 12 75 98 Leavers Non-smokers 2 6 5 2 0 1 16 Cigarette smokers 5 9 4 2 4 1 25 'Other' smokers 0 1 0 1 1 0 3 Ex-smokers 5 5 1 1 1 0 13 Total 12 21 10 6 6 2 57 TABLE 16 Numbers of current men, pensioners and leavers by age group and smoking category. ucc 057654 Non-smokers < 35 ;>5-44 Age groups (yrs) 45-49 50-54 55-59 60 + All 50 (*1.07) 4l (6.71) 16 (8.19) 16 (12.3) 5 (15-i-) 20 (15.2) 148 (8.02) Cigarette smokers 65 (5.23) 97 (12.0) 60 (16.4) 60 (16.6) 55 (18.U 41 (16.9) 378 (13.7) Other smokers Ex-smokers 3 (3.05) 11 (14.7) 12 (13.8) 9 (19.4) 12 (19.') 11 (11.7) 58 (15.0) * 37 (6.63) 48 (12.9) 35 (14.5) 4l (16.5) 31 (16.3) 42 (18.1) 234 (14.2) All 155 (5.15) 197 (11.3) 123 (14.5) 126 (16.2) 103 (17.';) 114 (16.5) 818 (12.9) TABLE 17 Numbers of men and (mean dust indices) of 8l8 men grouped by age anc smoking category. (Units of dust indices aie years x rag/m3). ucc 057655 1 No. of men All men 818 Nonsmokers 148 Cigarette Smokers 37S Other Smokers 58 Ex-smokers 3 d.f. P value* 23i Chronic cough (disregarding sputum) 152 (l8.6> 5 (3.4*) 111 (29.4*) 36 (27.6*) 20 (8.3*) 57.2? .001 Sputum (disregarding cough) 134 06.40 7 <4.#) 88 (23.3*) 15 (25.9*) 2't (10.2?*) 31.80 .001 Chronic cough and/or sputum 192 (23.5SO 10 (6.8*) 134 (J4.40* 18 (31.0*) 30 (12.1,*) 53.46 .001 TAtll.E ifi Number of nen (prevalence) with chronic cough and/or sputum among smoking category. * Significance of difference between smoking categories. men grouped by ucc 057656 ucc 057657 Variables Units Age Height Weight Dust index Lifetime cigarette consumption Years cm kf! Years x mg/nf 1,000 packs Constant for current men and leevers (pon-smokers) Constant for pensioners (non-smokers) Mean 46.4 172.5 77.4 12.9 7.42 Regression Coefficient - 0.0350 0.0424 - 0.0022 - 0.0040 - 0.0169 Constants (Intercepts) - 1.8845 - 2.1527 t - 15.31 11.04 - 0.99 - 2.2? - 5.28 P < 0.001 < 0.001 N.S. < 0.05 < 0.001 * Effect on intercept of: Cigarette smoking Other smoking Ex-smoking - 0.0757 - 0.0949 0.0975 - 1.16 - 1.02 1.44 N.S. N.S. N.S. TABLE 19 Multiple regression of FEV1 with age, height, weight, dust index and lifetime cigarette consumption. Different intercepts alloved for smoking categories and pensioners. vtoji All smoking categories Regression with dust index - O.CO*iO Cigarette smokers Other smokers Ex-smokers Non-smokers - 0.0069 - 0.C033 - 0.0016 O.OG25 t value - 2.2? - 2.82 - 0.62 - 0.53 0.48 P < 0.025 < 0.005 NS NS NS TABLE 20 Further expansion of the model for regression of FEVj shown in Table 19. Different slopes for the regression with dust index have been allowed in the model, which still includes age, height, weight end lifetime cigarette consumption, and different intercepts for pensioners and smoking categories. ucc 057658 * Variables Units Age Height ''eight Dust index (current men) Dust index (pensioners) Dust index (leavers) lifetime cigarette consumption Years cm kg years x mg/o 1 M II II 1,OCO packs Constant for current men ( non-smokers) Constant for pensioners (non-smokers) Constant for leavers (nen-smokers) Effect on intercept of: Cigarette smoking Other smoking Ex-smoking Regression Coefficient t p - 0.032*1 0.0634 - 0.0089 - o.oo*? 0.0011 O.OO85 - 0.0136 - 13.14 15.35 - 3.70 - 2.27 0.20 0.68 - 3.97 < 0.001 < 0.001 < 0.001 < 0.05 N.S. h'.S. < 0.001 Constants (Intercepts) -4.08l8 - 4.5541 - 4.2108 - 0.0017 0.0176 0.1036 - 0.02 0.18 1.43 N.S. H.S. N.S. ucc 057659 TABLE 21 Multiple regression af FVC allowing for different intercepts ani slopes for employment categoric 3. ucc 057660 Variables Units -- ---- ---jW" Mean Regression Coefficient Age Height . t'eirht bust index Lifetiae cigarette consumption Years cm Rb Years x ag/n3 1,000 packs 46 .4 172.5 77.4 12.9 7.42 - 0.0032 - 0.0022 0.0019 - 0.0003 - 0.0017 Constants (Intercepts) Constant for current men and leavers (non-smokers) Constant for pensioners (non-smokers) * 0.1323 0.1220 t - 7.68 - 3.C7 4.62 - 0.83 - 2.83 Effect on intercept of: Cigarette smoking Other smoking Ex-smoking Chronic cough or sputum Recent exacerbation of cough and phlegm Recent chest illness - 0.0121 - 0.0277 - 0.0001 - 0.060? - 0.0030 - 0.0169 - 1.03 - 1.61 - 0.03 - 6.03 - 0.29 - 1.85 TABLE 22 Regression of log|j^1lwith explanatory and associated variables. P < 0.001 < 0.01 < 0.001 N.S. < 0.01 N.S. H.S. H.S. < 0.001 N.S. < 0.1 oon Variables Units Age Height Weight bust index Lifetime cigarette consumption Years cm kg Years x mg/:a* 1,000 packs Kean *t6.* 172.5 77.*t 12.9 7.*f2 Constant for current men and leavers (non-smokers) Constant for pensioners (non-smokers) Effect on intercept of: Cigarette smoking Other smoking Ex-smoking Regression Coefficient - 0.21*16 0.0831 0.1*t85 0.0121 - 0.0717 Constants (Intercepts) 15. **156 12.5996 t - 8.t)0 1.8*4 5.35 0.61 - 2.81 - 3.7`*1*5 - 2.9082 - 1.1565 - *t.{(7 - 2.58 - I.* 6 P < 0.001 < 0.1 < 0.001 N.S. < 0.05 < 0.001 < 0.01 N.S. TABIE 23 Regression for TLc^(ml/min* nm Hg) allowing different intercepts for employment and smoking categories* uai\ Reader Small rounded opacities Small irregular opaciti.-* I Both rounded and irregular er.ncit.ies Category 0/1 or more Category 1/0 or more Category 0/1 or more Category 1/0 or more Category 0/1 or more 03 18 (2.2%) 2 (0.2%) 32 (3.91%) 7 (0.(6%) b (0.^9%) 15 b (0.b&) 2 (0.2%) 53 (6.**aS) *t6 (5.(2%) o (o.oq;) 17 50 (6.11$) 10 (1.22$) 29 (3.5%) b (0.V&) /--s & O oo ucc 057662 TABLE 2b Prevalences of categories of email opacities found by three readers in 8l8 chest radiographs. (Radiographs in which both rounded and irregular opacities were found are also represented here in the rounded and irregular categories.) Reader 15 Absent Present Total Absent Render 03 freserrt 796. 18 4 800 0 18 Vc ini 3l4 4 818 Reeder 17 Absent present Total Reader 03 Absent Present 759 9 41 9 13 Total 768 50 8'*8 Reader 1? Absent Present Total Bunder Absent '5 Present 766 2 48 814 24 Total 768 50 Si8 TABLE 25 Agreement between readers for presence of small rounded opacities categ:ry 0/1 cr core. Reeder 15 Absent Present Total Bender Absent 03 Present 748 17 38 786 15 32 Total 765 53 818 Reeder 1? Absent Fresent Total Boeder 03 Absent Present 763 ' 26 23 786 6 32 Total 7<)-9 29 8l3 Reader 1? Absent Present Total B juder Absent 15 Present 745 44 20 765 9 53 Tc tel 789 29 Sic TABLE 26 Agreement between readers for presence of small irregular opacities category 0/' or more. VCJOl 991$q ooC *99t$o 000 0 1 DUST INDEX RANGE'S (years 3: nr/"5) b.o - 1.5 1.5 - **-5 4.5 - 7.5 7-5 - 12.0 12.0 - 18.C iS.O - 30.0 Ar:e (yrei % with Q/U < 1*1* (nilliber) Total 0.0C> (0) 72. 1.1*5 (1) 69 2.94 (2) 68 5.66 .(;> 53 0.00 (0) 4l 2.56 (1) 39 -- 30.0 + ni * -454* 20.GO (2) IO 2.56 (9) 352 1*5 - 1*9 % with C/W {number) Toti.l 0.00 (0) 12 7-69 (1) 13 0.00 (0) ia 10. J4 (3) 29 0.00 (0) 16 12.50 (2) 16 15-79 (3) n9 7.32 (9) 123 S with 0/It (nuaher) Total 0.00 (0) 4 55 - 59 with C/1+ (number) Total 0.00 (O) 6 7.14 (1) 1<* 10.00 (1) 10 7.41 (2) 27 18.1a (2) 11 9.09 (2) 22 10.53 (2) 19 0.00 (0) 15 7.14 (1) 14 4.1? (1) 24 10.34 (3) 29 10.00 (2) 20 14.29 (2) 14 6.35 (8) 126 10.68 (11) 103 % with 0/1+ 0.0c 9.C9 12.50 10.53 10.?' ii.il 23-53 11.40 60* (nanher) (0) (1) (1) (2) (M (2) (4) (13) Total 1) 11 3 19 28 18 1? 114 All Hen with 0/1+ (nu.T.ber) Total 0.00 10) 10? 4.27 (5) 11? 5.30 (?) 152 8.45 (12) 142 3-51 (4) 114 7.14 (9) 126 16.25 (13) 80 6.11 (50) BlE TABLE 2? Kinder 17; presence of anull rounded opacities category C/l or mere by age and dust index rarM;ei*. vDVJ1 Header Small rounded opacities Category Q/l or more Category 1/0 or -mere Small irregular opacities Category 0/1 or more Category 1/0 or more Both rounded and irr^pi'.nr edacities Category 0/1 or "arc 03 6 (3.3SO 2 <1.U!) 1-T (6.1?i) 4 2 2( . -;) 2 (1.1;0 15 2 (1.US) 1 (0.60) 10 (5.5fO 8 (4.4;;) 1 (0.6O 17 9 (5.06) 2 (1.T5> 5 (2.f&) 3 (1.70 0 (o.c?0 TABLo 28 Prevalence of categories of small opacities found by 3 readers in chest radiographs of l8l men not included in the analysis. ucc 057665 I j \ i oc acr> gO CaDn riGUBE 1 Frt:'|i.ci-y ..r 'licit in'licetj. 057667 KJflllKfci 2 Observed and fitted prevalence of chronic cough or sputum among related to lifetine cigarette consumption. cigarette smokers, 899Z.S0 oon FSiV (Litres) rio. of Ken Are ranges 50 65 3 37 : 55 <*1 97 11 48 "1 35 - 44 16 16 5 6o 60 55 1?! 9 12 35 41 31 1 45 - 491 50 ~ 5* 155 - 59 1 20 Non-smokers 41 Cigarette smokers 11 Other snokers 42 Ex-smokers 60t- 1 ucc 057669 FTflURL *t iJuc*t index (yenro x tr-p/n?) kci're:>H j on of :KVl will du.it index (KfcV^ adjusted for a^e t height and vch-U). ucc 057670 r ranges 6.0 --> 50 65 3 37 : 35 41 97 ,11 48 1 35 - 44 16 60 12 35 re - m 16 5 60 55 9 12 41 31 50 - 541 55 - 59 1 5.0 - 20 Non-smokera 4l Cigarette smokers 11 Other smokers 42 Ex-smokers 6<>r 1 * % *A Non-smokers Cigarette smokers Other smokers Ex-smokers vOn'! 4.0 -- 5.0 -- 25 i----------------------------r 35 45 TT1 55 t>5 75 Mean Age tyrs) MOIIkK 5 Relationship of FVC with age - crude data # FVC (1 i t res) ( Adjusted for nge, height, weipht, stroking) ucc 057671 FKilIHE 6 Uust index (years x mr/m*) degression of FVC with dust Index (FVC nd.lusted for age, height mtc weight). i i No. of Men A?e ranges 50 65 3 371 < 35 41 97 11 48 1 55 - **4 16 16 60 60 12 9 35 41 | 45 - 49 | 50 - 5 55 12 31 155 - $9 |----------- 20 Non-smokers 4l Cigarette smokers 11 Other smokers 42 Ex-smokers 6C+ ....... 1 i 0.80 -- Non-smokers A- Cigarette smokers Other smokers Ex-smokers 0.75 -- 'CTJi DM/ lA3i 0.70 057672 o.6s --; I 35 '5 FEV, 65 75 Mean 're (yrs) KEY-READER CODE ------- 03 * ------x 15 ucc 057673 FIGURE ft Prevalence of snail rounded opacities category (l/l or more, found by three readers, related to age group. Humbera of Men nre given on the graphs. Prevalence of rrall rounded or- ic i t ies. KEY-READER CODE --------- -- fc------ x 05 15 f--------+ 17 vcOr\ ucc 057674 KlOtJRE 9 > Prevalence of email roun4e'l opacities category 0/1 or more, found by ;hrce readers^ related to duet in-lex. Prevalence {.%] oT Email rounded opacities. ucc FIGURE 10 Prevalence of small rounded opacities category C/1 or more, >y ape proup, reader 17, related to PVC dust index. (Gee Table 26 for nunbers of men.) KEY AGE *-------- *-------- * *----------* ---------e 35 ^5 55 65 j-'lHUWE 11 Reader 17: observed prevalence of small rounded opacities, category O/' or more; and, for each of four age groups, predicted prevalence; related to lpdex of dus . exposure.  ucc 057677 t 75. HErERENCES AGARWAL, D.K., KAV/, J.L., SRIVASTAVA, S.P. and SETH, P.K. (1978) Some biochemical and histonathological changes induced by polyvinylchloride duet in rat lung. Environ. Research 16. 333 - 3^1. ANDRASCH, R.H., KOSTER, F., LA'/ZSCN, k.H. and BARDANA, E.J. (1975) Meat wrappers asthma - an appraisal of a new occupational syndrome. J. Allergy Clin. Immunol. 13c. ARNAIJD, A., POMMIER DE SANTI, P., GASBE, L., PAYAN, H. and CHARPIN, J. (1978) Polyvinyl chloride pneumoconiosis. Thorax 33, 19 - 25. 4ftArmrcMf 4 at / WA AWV * <* \ I r* uuxeiiia u* iuuuoli ial tijfcieutf in. new plant** for the manufacture of synthetic resins and plastics. Gig. Trud. 11, 20 - 27. BROOKS, S.M. and VANDERVdW, R. (1977) Polyvinylchloride film thermal decomposition products as an occupational illness. J. Occup. Med. 19. 192 - 196. CORDIER, J.M., F1EVEZ, C., LEFEVRE, M.J. and SEVRI", A. (1966) Acroooteolyse et lesions cutar.fes associees ches deux ouvriers affeetees au nettoyage d'autoclaves. Cah. Med. trav. <4, 14 - 19. CREECH, J.L. and JOHNSON, K.N. (1974) Angiosarcoma of liver in the manufacture of polyvinyl chloride. J. Occup. Med. 16. 150 - 151. CASULA, D., CHERCHI, P., SPIGA, G. and SPINAZZCLA, A. (1977) Environmental dust in a plant for the production of polyvinylchloride. Annali Instituto Superiore di Sanita 13. 189 - 195. DARKE, C.S. (1976.) Vinvl Chloride and the nrnHont.i on of PVC (TK . Proc. Roy. Soe. Med. 6, 280. DE HAAN, R.L. (1971) Toxicity of tissue culture media exposed to polyvinylchloride plastic. Nature New Biology, 231. 85 - 86. FALK, H. and PORTNOY, 3. (1976) Respiratory illness in meat wrappers. J. Am. Med. Ass. 23, 915 - 917. FOX, A.J. and COLLIER, P.F. (1976) Low mortality rates in industrial cohort studies due to selection for work and survival in the industry. Brit. J. Prev. Soe. Med. 30, 225 - 230. FRONGIA, N., SPINAZZQLA, A. and BUCARELLI, A. (197*0 Lesioni polmonari sperimentali da inalazione prolungata di PVC in ambienta di lavoro. Medicina Lavoro 6j5, 321 - 342. GAENSLER, E.A. (1951) Analysis of the ventilatory defect by timed capacity measurements. Am. Rev. Tuberculosis 64, 256 _ 278. ucc 057679 76, GAMBLE, J., LIU, 3., MeMICHAEL, A.J. anti WAXWEILER, R.J. O976) , Effect of occupational anti nonoccupational factors on the respiratory system of vinyl chloride and other workers. J. occup. Med. 1ft. 659 - 670. GRASSO, P., GAYDGN, J. and HENDY, R.J. (1973) The safety testing of medical plasties II. An assessment of lysosomal changes as an index of toxicity in cell cultures. Fd. Cas. Toxicol. 1^, 255 - 263. HARRIS, G.W. and MAGUIRE, 3.A. (1968) A gravimetric dust sampling instrument (SIMPEDS) : preliminary results. Ann. occup. Hyg. 195 - 201. INTERNATIONAL LABOUR OFFICE (1972) ILO U/C International classification of radiographs of pneumoconioses 1971. Occupational Safety and Health 22. Geneva. JOHNSON, C.J. and ANDERSON, J.W. (1976) Meat-wrappers asthma : a case study. J. occup. Med. 18, 102 - 10^. LANGE, C.E., JUNE, S., STEIN? G. and VELTMAN, G. (1973) Die sogenannte vinylchlorid-krankheit-cine berufsbedingte syatemsklerose? Int. Arch. Arheitmed. 32, 1 - 32. LILIS, R., ANDERSON, H., NICHOLSON, W.J., DAUK, S., FISCH3EIN, A.S. and SEUKOFF, I.J. (1975) Prevalence of disease among vinylchloride and polyvinyl chloride workers. Ann. N.Y. Acad. Sei. 2<*6. 22 - 4l. LILIS, R., ANDERSON, H., MILLER, A. and SELIfCOFF, I.J. (1976) Pulmonary changes among vinyl chloride polymerisation workers. Chest 69, Supplement (2 February), 299 - 303. LILIS, R., ANDERSON, ll., MILLER, A. and 3ELIK0FF, I.J. (1977) Modifications pulmonaires et exposition au chlorure et polyehlorure de vinyle. Med. Hyg. 3, 15^2 - 15^5. MAGUIRE, B.A., BARKER, D. and BADEL, D.A. (1970) SIMPEDS 70 : an improved version of the SIMPEDS personal gravimetric dust sampling instrument. In: Inhaled Particles III, Vol. II (Ed. W.H. Walton), 1053 - 1057, Unwin Bros., Old Woking, Surrey, U.K. MAPP, C., FABBRI, L., ROSSI, A., MORO, G. and CERVI, G. (1978) Alterazioni funzionali respiratorie da esposizione cronica a cloruro di vinile monomero e polimero. Med. Lavoro 69, 151 - 162. MASTRCMATTEO, S., FISHER, A.M., CHRISTIE, H. and DANZIGER, H. (i960) Acute inhalations toxicity of vinylchloride to laboratory animals. Amer. Ind. Hyg. Ass. J. 21, 39** - 398. ucc 057680 77 MEADE, F., SAUNDERS, M.J., HY3TT, F., REYNOLDS, J.A., PEARL, N. and COTES, J.E. (1965) Automatic measurement of lung function. Lancet 2, 573 - 575- MILLER, A., TEIRST2IN, A.S., CHUANG, M. and 3EUK02T, I.J. (1975) Changes in pulmonary function in workers exposed to vinyl chloride and polyvinyl chloride. Ann. N.Y. Acad. Sci. 246, 42 - 52. McKERROV, C.B., HeDERMOTT, M. and GILSON, J.C. (i960) A spirometer f r measuring the forced expiratory volume with a simple calibrating device. Lancet 149 - 151. MRC (1976) Medical Research Council Working Party on Research into Chronic Bronchitis. Publications Group, Medical Research Council, 20 Park Crescent, London WIN 4AL. OREHSTEIN, A.J. (i960) ProcPneumoconiosis Conference, Johannesburg 1959* London, J.A. Churchill, pp. 619* FELLING, D., SHARRATI, M. and HARDT, J. (1973) The safety testing f medical plastics : I, An assessment of.methode. Fd. Cos. Toxicol. 11, 69 - 83. PIGGOTT, G.H. and ISHMAEL, J.' (1979* A comparison between in vitro toxicity of PVC powders And their tissue reaction in vitro. Ann. occ. Hyg. In press. POLAKOFF, P.L., LAPP, N.L. and REGER, R. 0975) Polyvinylchloride pyrolysis products : a potential cause for respiratory impairment. Arch, envir. Health_30, 269 - 271. RAE, S., WALKER, D.D. and ATTFIELD, M.D. (1971) Chronic bronchitis and dust exposure in British coalminers. in: Inhaled Particles III, Vol. II (Ed. W.H. Walton), 897 - 900. Unwin Broa., Old Woking, Surrey, U.K. RICHARDS, R.J., DESAX, R., HEXT, P.M. and ROSE, F.A. (1975) Biological reactivity of PVC dust. Nature 256. 664 - 665 RICHARDS, R.J., DESAI, R. and ROSE, F.A. (1976) A surface-active agent involved in PVC-induced haemolysis. Nature 260, 53 - 54. SALTHOUSZ, T.N., MATLAGA, B.F. and O'LEARY, R.K. (1973) Microspectro photometry of macrophage lysosomal enzyme activity : A measure of polymer implant tissue toxicity. Toxicology and Applied Pharmacology 2, 201 - 211. SOKOL, W.N., AELONY, Y. and BEALL, G.N. (1973) Meat wrappers asthma : a new syndrome. J. Am. Med. Ass. 226(b). 639 - 64l. SUCIU, I., DREJMAN, I. and VELASKAI, M. (1967) Study of disease caused by vinyl chloride. Med. Lavoro j>8, 26l - 271. jjqq 057681 78 SUCIU, I., PRODAH, L., ILEA, E., PODURARU, A. and PASCU, L. (1975,) Clinical manifestations in vinylehloride poisoning. Ann. N.Y. Acad. Sci. 246, 53 - 69. SZEUDE, B., LAPIS, K., NEMES, A. and PINTER, A. (1970) Pneumoconiosis caused by the inhalation of polyvinylchloride dust. Med. Lavoro 6^, 433 - 436. VERTKIN, T.I. and MAMONTOV TU, R. (1970) The state of the bronchi and lungs in workers employed in the manufacture of polyvinyl chloride articles. Gig. Trud. 14. 29 - 32. WAXWEILER, R.J., FALK, H., McMICHAEL, A., MALLOV, J.S., GRIVAS, A.S. and STRINGER, W,T. (1977) A cross-sectioral epidemiologic survey of vinylehloride workers. O.S. Dept, of HEW., Public Health Service Centre for Disease Control. NIOSH, Cincinatti, Ohio. UCC 057682 APPWDTX 1 79- PVC dust concentrations allotted to occupations on the basis of measured dust concentrations for 130 men. Plant Cl/4, Cl, C4 C2/3, C2, C3 C5/6, C5, C6 'ci/4, Cl, C4 C2/3, C2, C3 C5/6, C5, C6 C2/3, C2, C3 (J unspecified Cl/4, Cl, C4 C5/6, C5, C6, C C unspecified Cl/4, Cl, C4 .Cl/1*, Cl, C4 ClA, Cl, C4 C1/4, Cl, C4 C1/4, Cl, C4 CT/4, Cl, C4 ClA, Cl, C4 C1/4, Cl, C4 C1/4, ci, C4 ClA, Cl, C4 C A TAflU C2/3, C2, C3 C unspecified C2/3, C2, C3 . C2/3, C2, C3 C2/3, C2, C3 C2/3, C2, C3 C2/3, C2, C3 C2/3, C2, C3 C2/3, C2, C3 C unspecified Cl/4, Cl, C4 Cl/4, ci, C4 ClA, Cl, C4 Occupation Concentration Laboratory l It Work Study/Quality Control It It Tester ft It Instrument Technician Process - Work Study/Quality Control General Process - Mixing e Process - Milling Investigation Estimator Foreman Superintendent Chargehand Fork Lift, Driver, Distribution Maintenance/Engineers T'Ltaixttrv t InvettligtiLi'ja Fork Lift, Driver, Loader Fork Lift, Driver, Distribution Process - Packing Foreman Superintendent Chargehand General Process - Washing Process - Drying Foreman Process - Autoclaving Process - Recovery Process - Reaction 0.21 tl It n It It It It ft It 1.41 0.98 It It tl 19 0*94 It It 0.89 0.84 0.82 0.70 0.69 0.60 0.58 0.58 0.58 0.58 0.58 0.54 O.51 0.50 0.50 0.50 ucc 057683 82 Plant Occupation Concentration All Plants All Plants (PVC) All Workers It Distribution Research It t Yard II Stores II Construction South Support Group If It Alkathene Process - Packer, Bagger Alkathene Process - Mixer, Gardner Operator Alkathene Engineer, Inspector, Investigation, Distribution Alkathene Alkathene Alkatheno Alkathene Process - Milling, Machine Operator Process - Filming, Tubing, Extrusion, Pigments Process - Filtering, Steeping Charcoal Process - Lead Weighing, Washing Alkathene General, Work Study, Management/ Office Alkathene vn/2, w W1/2, w Wl/2 w Fluon Fork Lift, Foreman, Chargehand, Supervisor Process - Washing, Lead Weighing, Others Process - General, Relief Operator, Press Tester, Investigation, 'Work Study/ Quality Control Investigation, Storekeeper, Superintendent All Nylon It Terylene It Fluon Lab It Squireseate Welvic 3 PES II II Polythene 0.24 II It tt It II It IV 0.47 0.35 0.38 0.38 0.38 0.38 0.38 0.24 0.24 0.59 0.59 0.59 0.59 0.0 It It It II II II II IJCC 057684 Plant Occupation Ferspex LPT CL7 PI (D P2 Office Staff (outwith nlants or at 0,0 nlants)Restaurant Training All It II II II ft 11 H Concentration 0.0 11 fl It It It n it ucc 057685