Document mpGE4B5JwLZvZG3Ezwo51d78Q

REPORT NO. TM/81/8 UDC 616.24-003.6 : 678.743 EPIDEMIOLOGICAL AND CLINICAL STUDIES * OF POLYVINYLCHLORIDE WORKERS CA Soutar S Gauld M Lloyd Iil Copland JF Hurley JULY 1981 VAB.0001192180 r INSTITUTE 07 OCCUPATIONAL MEDICINE EPIDEMIOLOGICAL AND CLINICAL STUDIES OF POLYVINYLCHLORIDE WORKERS by CA Soutar, S Gauld, M Lloyd, LH Copland, JF Hurley Medical Branch, Institute of Occupational Medicine Roxburgh Place, EDINBURGH EH8 9SU. (Tel* 031-667-513D * VAB.0001192181 <ii) CONTENTS Page No. SUMMARY................................................... - ..................................... (iii) 1. 2. 2*1 2*2 2*3 2.4 2.5 * INTRODUCTION . . . ....................................... * m* *m - STUDIES OF DUST-DISEASE RELATIONSHIPS ............................... Subjects . Methods Results for men at first factory Results for men at second factory .............................. Radiological results at first factory 1 3 * 3 3 4 5 5 both factories 8 3. 3*1 3.2 3*3 CASE/CONTROL STUDIES .................................................................. 9 Subjects . 9 Methods ................................................................. 10 Results ..*.............................................. 12 4. DISCUSSION...................................................................................................... 20 5. ACKNOWLEDGMENTS........................................................................................27 REFERENCES................................................................ 29-30 TABLES (1-28)..................................................................................................... 31 - 54 FIGURES (1-10) 55-59 APPENDICES (I - IV)...............................................................................................61-78 VAB.0001192182 REPORT NO. TM/8V8 INSTITUTE OF (iii) OCCUPATIONAL MEDICINE EPIDEMIOLOGICAL AND CLINICAL STUDIES OF POLYVINYLCHLORIDE WORKERS by -\ \ CA Soutar, S Gauld, M Lloyd, LH Copland, JF Hurley *\*J . SUMMARY We have reported previously an epidemiological survey of 8l8 present and past workers at a factory making polyvinylchloride (?VC) in which we showed that estimates of exposure to respirable PVC dust were inversely related to lung function after allowing for the effects of age and smoking, and related to prevalence of slight abnormalities of the chest radio graph; one of three experienced readers found that prevalence of small rounded opacities category 0/1 or greater was related to dust exposure after allowing for age. . The present studies extended the sample of the study population to include all ot^er members of the current workforce who had worked at any time in the dustier plants in the factory, in order to examine the prevalence of respiratory abnormality in these men and identify any individuals with clinically important chest illness, and :o examine in this extended sample and in the original study group the relationship between dust exposure and radiological appearances based on new readings of the chest radiographs. Also men selected from the first study on the basis of radiological and functional abnormality to which exposure to PVC dust may have made an important contribution have been examined in detailed case/control studies to attempt to identify clinical syndromes associated with PVC dust exposure, and to estimate the clinical severity of the effects of the dust on the lung. Lastly, men working at a second factory manufacturing PVC have been examined to determine the prevalence among them of respiratory symptoms and lung functional and radiological abnormality. Men were seen by appointment, and questionnaires of respiratory symptoms were administered. A detailed occupational history was recorded, lung function assessed by simple spirometry, and a chest radiograph taken. In the case/control studies, detailed lung function measurements were made, a detailed medical history taken, and a physical examination performed by a physician. An index of exposure to PVC dust was calculated from the detailed occupational history and current measurements of VAB.0001192183 t 4 (iv) airborne levels of respirable PVC dust made in an environmental survey the previous year. The chest radiographs for the original study population of 8l8 men were read again according to the ILO U/C Classification (ILO, 1972) by the same three experienced readers who read them previously*. The interval between readings was about a year, during which the results of the first reading had been released and discussed. All three readers recorded higher prevalences of small rounded opacities category O/l or greater on the second reading compared with the first. For two readers the changes in prevalence were from 2.2% to 3*8%, and 0.5% to 2.6% respectively; the third reader recorded a much higher prevalence on the second reading, a change from about 6% to about 6C%. The increase in prevalence for this reader was confined mostly to category O/l; the increase in category 1/0 or greater was from 10 to 17 radiographs only (1.2% to 2.1%). Prevalences of small rounded opacities category 0/1 or greater recorded by all three readers were related to In the previous readings, the two readers recording lower prevalences of small rounded opacities had not found a relationship with dust exposure. On these second readings one of them did find a relationship between small rounded opacities category 0/1 or greater and dust exposure after allowing for age, prevalence increasing from about 2.5% for men with the lowest PVC dust exposure to about 9% for men with the highest exposure. The reader who had previously detected the relationship with dust exposure (and whose second reading recorded a much higher prevalence cf these opacities than the first reading), again found a relationship with dust exposure after allowing for age, but on this reading it was with a higher category of profusion of small rounded opacities, category 1/0 or greater, Drevalence increasing from about 1.5% for men with the lowest dust exposure to about 7*3% for men with the highest dust exposure. He did not on this second reading find a relationship between the lower category of opacities and dust exposure. The chest radiographs for the original study population of 8l8 men were also read twice within a short time by a panel of five self-trained readers* who had not been informed of the results of other readings. Prevalences of small rounded opacities category 0/1 or greater were found by all readers to be related to age. After allowing for age, prevalences of these small opacities were found to be related to dust exposure by one reader on both readings arid another reader on one reading. Another reader found a relationship between small rounded opacities category 1/0 or greater and dust exposure after allowing for age. VAB.0001192184 (,) Ve conclude that our previous finding has been confirmed, that estimated. PVC dust exposure was related to abnormalities of the lungs demonstrated by the presence of small rounded opacities in the chest radiograph* The slight degree of the radiological abnormalities in most , and the difficulty in distinguishing them in an individual radiograph from the effects of age, combined with variations in readers' interpretations of these appearances, have caused some inconsistencies in the detection of the effects of PVC dust* An additional 229 men at this factory were examined in the present medical survey. There was no evidence of unusual prevalences among them of respiratory symptoms, abnormal lung function or chest radiographic abnormalities. When these groups were combined into one large group of 1,047 men, the relationships of estimates of dust exposure to lung function and radiological abnormality were similar to those found in the original study population. a One hundred and twenty-seven men seen at the second factory manufacturing PVC had a low prevalence of - respiratory symptoms and included few men with impaired lung function or clinically important chest radiographic abnormality. In four case/control studies, men were selected from the original study population on the basis of radiological or lung functional abnormality where PVC dust exposure might have made an important contribution. Men were chosen on the basis of the presence of small rounded opacities, presence of small irregular opacities, or (separately for non-smokers and smokers) low lung function with relatively high estimated PVC dust exposure. These cases, together with suitably matched controls, were given detailed clinical examinations. Twenty-eight men with small rounded opacities in the chest radiograph had slightly higher prevalences of chronic productive cough and crackles heard through the stethoscope in the mid and end parts of inspiration than controls matched for age, smoking habits and estimated dust exposure, confirming that these opacities were associated with minor functional abnormalities. The lung function of the cases was not significantly different from controls, but an exploratory analysis suggested that within the group of men with small rounded opacities, men with higher dust exposures had an average pattern of lung function abnormality intermediate between an obstructive and a restrictive type, in that the forced expiratory flow rates were reduced in parallel with the FEV^, but the gas transfer factor was well preserved. This, taken with the findings in the previous study, tends to suggest that the functional effects of PVC dust exposure are caused by disease of the peripheral airways, without airflow obstruction or much alveolar damage. Clinically important reduction of lung function was found in "f?* hUlT to ^ related to thVAB,0001192185 presence of small rounded opacities or to PVC dust exposure. (vi) Men with small irregular opacities (not thought to be related to PVC dust exposure) and reduced lung function had an obstructive pattern of lung function abnormality, with relative preservation of gas transfer factor. Tor non-smokers and smokers separately, "cases" with lower FEV^ than other men after allowing for age and height, and with relatively high estimated PVC dust exposure, were compared with controls who had similarly low , but had little exposure to PVC dust, so that the effect of dust could be studied* Non-smokers were matched for age, smokers for age and smoking habits. Among non-smokers, the lung function of the "cases" was within or close to the normal range and no differences were demonstrated between them and the controls. These men were probably in normal respiratory health. Among smokers, both "cases" and controls had a high prevalence of respiratory symptoms and abnormally low lung function of an obstructive pattern. The "cases" had a slightly higher prevalence of crackles heard through the stethoscope than controls, but there were no significant differences between the groups in lung function levels or pattern, and no differences in numbers of men with clinically important reduction of lung function. The respiratory disease evident among smoking "cases" and controls was probably the result of their smoking habit, and there was no evidence to suggest that PVC dust exposure had caused any serious additional damage. We conclude from the large epidemiological and small case/ control studies that exposure to PVC dust in the concentrations experienced in this factory caused a detectable reduction in lung function in the study population, and caused an increase in prevalence of slight abnormalities of the chest rariiogrranh. The clinical * features of the non-specific respiratory disease associated with dust exposure additionally included an increase in prevalence of productive cough in those with abnormalities of the chest radiograph, an increase in prevalence of crackles heard through the stethoscope during mid and late inspiration, and a pattern of lung functional abnormality characterised by a uniform reduction of ventilatory capacity, with relative preservat ion of gas transfer factor. These features suggest that the functional defect is predominantly in peripheral airways in the lung. Important clinical illness could not be shown to be related to PVC dust exposure in spite of careful selection of'cases" and controls, and we conclude that it is unlikely that exposure to PVC dust at these levels has caused serious respiratory disability, though the possibility of a rare idiosyncratic response to the dust cannot be excluded. VAB.0001192186 1. INTRODUCTION In a previous epidemiological, study (SOUTAR et al., 1980) of 8l8 men sampled from the workforce of a factory manufacturing polyvinylchloride (PVC), an index of individual cumulative exposure to respirable PVC dust was found to be related to the dyspnoea, reduction of forced, vital 4 forced vital capacity (FVC), and to complaint of mild exertional p capacity in one second (FEV^) prevalence of small rounded and opacities in the chest radiographs. Three experienced medical radio- graphic readers recorded different prevalences of small rounded opacities (category 0/1 or more on the ILC U/C Classification (ILO, 1972)), and the reader recording the greatest prevalence of these opacities found a relationship with dust exposure. The conclusions reached from this study were that exposure to respirable PVC dust was related to a non-specific disease of the lungs manifested by complaints of slight breathlessness, a functional defect probably of a mixed obstructive and restrictive type, and the presence of slight abnormalities in the chest radiograph. The present studies were designed to examine the prevalence of respiratory abnormality in all other members of the current workforce who had worked at any time in the dustier plants in the factory and who had not been included in the first survey, to identify any individuals with clinically important chest illness, and to examine in this population and in the original study group the relationship between dust exposure and radiological appearances based on new readings of the chest radiographs. Men were also selected from the first study on the basis of radiological and functional abnormality to which exposure to PVC dust may have made an important contribution, and have been examined in detailed case/control studies to attempt to identify clinical syndromes associated with PVC dust exposure, and to estimate the clinical severity of the effects of dust on the lung. In these studies, men with small rounded opacities in the chest radiograph were examined and compared with men without opacities matched for age, smoking and dust exposure to study the clinical features associated with these opacities. Men with small irregular opacities (not thought to be related to PVC dust exposure) were also studied in this way. Additionally, for smokers and non-smokers separately, the men with VAB.0001192187 2 I I 1 1 IA the lowest TEVl after allowing for age* height, weight and dust exposure among those with higher dust exposures were compared with men similarly selected from those with lower dust exposures. The purpose was to identify clinical features in these men attributable to PVC dust exposure, on the assumption that the higher exposure groups would be likely to include men whose lung function was substantially * impaired by PVC dust exposure (if such men existed), while all groups would be likely to include men with impaired lung function from other causes* Lastly, men working at a second factory manufacturing PVC have been examined to determine the prevalence among them of respiratory ** h symptoms and lung functional abnormality, and identify any men with clinical respiratory illness* In this report the methods and results of each of these studies are reported separately, and a single discussion covers all three* + + VAB.0001192188 2* STUDIES OF DUST-DISEASE RELATIONSHIPS . 2*1 Subjects .Men working at the first factory who had worked at any time in those plants where higher FVC dust levels had been found during the environmental survey (plants Cl, C4, W1 and W2) but who had not been studied were invited for examination, together with some men seen in the previous survey but who had not been included in the analysis because of incomplete data* s + f 4* t ah men currently working at the second factory were invited for examination* This factory was situated in Cheshire, and had been in operation since 1971* .i A a' ; 2*2 Methods * p A medical survey team visited each factory and men were seen by appoint- * meht* At the first factory, medical and occupational data were collected in the same way as in the first survey; that is a detailed occupational history was taken, and a respiratory symptoms questionnaire 4 (KSC, 1976) administered, by trained staff* Simple spirometry was a performed using a Gaensler spirometer and a full size chest radiograph * was taken* Details of the methods have been described elsewhere (SOUTAR et al*, 1980). aw * * P At the second factory, a shorter respiratory symptoms questionnaire was a administered (RAE et al*, 1971), simple spirometry was performed and a , chest radiograph taken* m % f* Chest radiographs were read for clinical purposes by a chest physician, and those from the first factory were re-examined under controlled a 4 conditions by the three experienced readers who read them previously, and m who classified them according to the ILO U/C Classification (ILO, 1972)* aM 4 On this occasion the earlier 818 radiographs were re-read together with r the radiographs from the new survey* The interval between first and second readings of these 818 radiographs was approximately one year, p during which the results of the first reading were published. A panel of five self-trained readers also read these radiographs* This panel * consisted of four non-medically qualified persons, and one medically VAB.0001192189 * 4 1 qualified person none of whom had undergone special training in radiology or chest disease; all five had trained themselves independently to read radiographs according to the HO U/C Classifi cation (COPLAND et al., in press) They hat! no prior information about the origin of the radiographs, and no knowledge*of the results of the first reading at the time of the second. The interval between readings was approximately one month. An index of dust exposure had been calculated for some of the men at to the first factory on the basis of their occupational histories and on measurements of airborne respirable dust made in a detailed environmental survey of the factory carried out the previous year a (SOTJTAR et al. 1980). (These estimates do not represent actual dust exposure, since only present dust levels were known, but represented an index of relative dust exposure for the purposes of comparison with medical responses.) Relationships between lung function measurements and dust exposure were studied by multiple regression methods, taking account of age, height, weight and smoking habits. Prevalences of radiographic abnormalities were also examined using linear logistic analysis. No dust exposure data are available for men surveyed at the second factory. Lung, function.results were compared with predicted values based on the 155 non-smoking currently employed men and leavers seen in the previous study at the first factory. 2.3 Results for men at first facto 2.3.1 Svmntoms and lunx function Two hundred and twenty-nine men attended for examination out of a r possible 263. 'Their mean age, years worked at the factory, mean index of dust exposure and prevalence of chronic cough and sputum were less than those of the 8l8 men studied in the first survey (Table 1). The forced expired volume in one second (FEVj) in relation to * predicted values derived from the non-smokers in the first study population are set out in Table 2. Twenty men (936) had FEV'sx below VAB.0001192190 f 5 A 8g of predicted, and all but two of these were or had been smokers 4 These features indicated a lower frequency of respiratory disease *% than in the original study population*- The lung function results for these 229 sen were included with * men originally studied, total 1,047 and FEV, and forced vital capacity (FVC) compared with dust exposure after allowing for age, height, weight, smoking habits, and employment status* The inverse relationships between FE^ and FVC and dust exposure were confirmed for this larger population (see Appendix I). The magnitude * FEV, was of a similar order to that found previously* 2.4 Results for men at second factory 2*4*1 toms and lung function / One hundred and twenty-seven of the male workforce of 134 men were seen Age distribution and smoking habit are set out in Table 3* Kean age was 39-5 years* Prevalence of lifelong non-smokers was 3556- Prevalences of respiratory symptoms were low* Prevalence of chronic cough and/or sputum was 1526, and of complaints walking up hill or hurrying on level ground %. The distribution of FEVj values expressed as percentage of predicted values are set out in Table 4* Six 7&) had FEVj values below 8006, and eight (6*356) above 12C& 4 Table 5* Clinical details of the six men below 8056 are FEV^/FVC ratios suggest that three airflow obstruction, while three had lung function which was probably normal These results indicate a low prevalence of respiratory disease* tm 2.5 Radiological results at first facto 2.5.1 nal study copulation The chest radiographs of 8l8 men examined in 1979 were read twice independently and randomly by three experienced readers* The interval between readings was approximately one year, during which the results .of the first reading were released and discussed* were also read by a panel of self-trained readers These radiographs VAB.0001192191 i * 6 # I 4 * 1 I Readings by experienced medically qualified readers All readers recorded higher prevalences of small rounded opacities r category 0/1 or more on the second reading than on the first (Table 6). One reader recorded much higher prevalences than the other two on both occasions, and the prevalence on his second reading was strikingly * higher than on the first* The increase was found principally among classifications by this reader of category O/l. The increase in prevalence of category 1/0 or more was from ten to seventeen radiographs only* Tables of consistency are set out in Appendix II* Prevalences of small irregular opacities found by three readers were similar in both readings (Table ?) * The relationships of the results of the first readings with age and index of dust exposure have already been reported (SOTTAR et al*, 1980), and these relationships for the second readings are presented here. Prevalences of small rounded opacities category 0/1 or greater recorded by all three readers were related to age ( Figure 1). After allowing for age, only reader 03 found a relationship between this category of small opacities and index of dust exposure (P < 0.03) (Figure 2). After allowing for dust exposure, these opacities recorded by reader 03 were related to age at the 105 significance level* However, prevalence of small rounded opacities category 1/0 or greater recorded by reader 17 was related both to age and to index of dust exposure (P < 0*03 and P < 0*001 respectively, each after allowing for the other) (Figures 3 and 4)* The other two readers found that this category of small opacities was related to age but not to dust exposure* Prevalence of small irregular opacities was found by all three readers to be related to age (Figure 3), but not to index of dust exposure. Readings trained nanel A self-trained panel of five readers read 816 radiographs from the same series (two were unavailable at the time of reading) on two occasions separated by an interval of several weeks, without knowledge of the results of their own first readings or the medical panel's readings* Prevalences of small opacities recorded by these readers are set out in and Differences in prevalences VAB.0001192192 * 7 on other series of radiographs have been described previously (COPLAND et al., in press). + 44 *1 Prevalence of small rounded opacities category 0/1 or greater was found by all five readers oh both readings to be related to age (all relationships were statistically significant at the 50ft level except for one which approached this level of significance). Observed data for the second reading are illustrated in Figure 6 After allowing for age, prevalence of this category of small opacity was found to be related to index of dust exposure by reader L5 on both readings and by reader L3 on the second reading. These relationships with dust exposure were statistically significant (P < 0.01 and P < 0.03 * respectively) but in both cases were based on low prevalences of * opacities. Some other readings suggested a relationship with dust exposure, but these did not reach statistical significance at the % level (Figure 7) Prevalence of small rounded opacities category 1/0 or greater recorded by reader L2 on the first reading was related to age and dust exposure (P < 0.001 and P < 0.05 respectively, each after allowing for the other). itude of effects on chest radiogranhic anoearances The prevalences of small rounded opacities category 0/1 or greater found by individual medical readers on second reading in relation to age ranged from zero for the youngest age group (< 35 yrs) to about 115ft for the oldest age group (^ 60) for reader 15, about 25ft and about 65ft respectively for reader 03, and about 53& and about 6/5 for reader 17* The prevalences of these opacities in relation to index of dust exposure ranged from about 2.50ft for the lowest two dust categories together (less than 4.5 dust index units) to about for the highest dust category O 30 units) for reader 03 (the other two readers did not find exposure for this category of opac The prevalences of small rounded opacities category 1/0 or greater found reader 17 ranged from zero for the youngest age group to about 65ft fo: the oldest group, and from about 0.5# for the lowest two dust categories to about 7.^ for the highest dust category. VAB.0001192193 Relationships between snail rounded opacities category 0/1 or greater and dust exposure were found by two self-trained readers recording low prevalences of opacities. On the second reading reader L3 recorded a prevalence of about 1% for the lowest two dust categories and about for the highest dust category, while for reader L5 the prevalences * were CfZ and 2.5# respectively on the second reading. 2.3*2 Additional men studied in 1980 The chest radiographs of 229 additional nen studied in 1980 were read once by the nedically qualified panel of readers, at the same time as the second reading of the 818 radiographs for the original study population. Prevalences of small rounded and small irregular opacities found by each reader are set out in Tables 10 and 11. Prevalences of these opacities tended to be similar to or lower than the prevalences found in the original study population on the second reading. Within this group, prevalence of small rounded opacities was found by all three readers to be related to age, but not related to index of dust * When the results from the additional men and the original population were considered together (total 1,047 men), the significances of the relationships of small rounded opacities with dust and age were increased slightly, but the estimates* of the size of the effect were not' substantially altered. 2.6 Clinical readings of chest radiographs at both factories Three radiographs at the second factory and eight at the first factory were thought on clinical readings to show clinical abnormalities (pleural thickening or calcification in five, cardiomegaly in three, apical fibrosis in two, and discrete opacities requiring investigation in one). These men, and those with abnormal lung function, were referred to their h general practitioners. VAB.0001192194 9 3. CASE/CONTBOL STUDIES 3*1 Sublets 4 * For the clinical studies, eases and controls were selected from the 818 men examined at the Hillhouse factory in February 1979 on the basis of chest radiographic abnormality or low FEV^ after allowing for age and height* 3*1*1 Small rounded opacities In the previous study, one reader had recorded the presence of small rounded opacities category 0/1 or greater on the ILO U/C scale in the chest radiographs of fifty men (a higher prevalence than the other two readers recorded)* Ten of these were of category 1/0 or greater, and all of these were invited to attend for further study, together with a further 20 men randomly selected from the remainder* Reserves were selected in case a man did not attend for examination, and were invited for examination when selected men did not attend* The controls were men in whose chest radiograph all three readers agreed small opacities were not present, group-matched by three ranges of index * of PVC dust exposure, three ranges of age, by smoking category and by employment status (current employment at the factory, pensioner, and those who left for reasons other than retirement). The index of dust exposure was not actual lifetime dust exposure, but was calculated from current personal dust exposure levels and a detailed occupational history for each man (SOUTAR et al., 1980)* 3*1*2 Small irregular opacities In the previous study, 90 men were thought by one or more readers to have Y+ small irregular opacities category 0/1 or more* Twenty men, thought by 4 one or more readers to have these opacities and with no small round opacities by any reader, were randomly selected. Twenty controls * thought by all three readers not to have opacities of either type, were matched by the same dust and smoking criteria as in the study of men with small rounded opacities, but matched by only two age categories. VAB.0001192195 3.1.3 Men with low FEV Men with high dost exposure and low FEV1 after allowing for age, height, weight, employment status and dust exposure were also selected for further study. By "low FEV^1 we mean that the observed FEV* was less than that predicted by a multiple regression analysis taking these variables into account (negative residual) The largest would be associated with the observed FEV. which was comuarison with Among non-smokers the sixteen men with largest negative residual FEV* among those with an index of dust exposure greater than 10 units were selected together with sixteen k controls among those with an index of dust exposure less than 3.5 units matched by residual FEVt and age. Twenty cigarette smoking cases and twenty controls were selected similarly, but the selection criteria for dust exposure were greater than 13.3 or less than 8.5 units respectively (non-smokers tended to have lower dust exposures). 3.2 Methods A medical team visited the factory and men were seen by appointment. None of the medical team knew to which study group any of the subjects belonged. Same of the cases had received medical reports the previous year informing them of abnormal lung function results, though not of abnormal radiological results (since the presence of low categories of small opacities in the chest radiograph seen in the clinical context had not justified a report of abnormality). One of us (CAS) carried out a clinical examination of the chest, recording physical signs on a questionnaire. Subsequently he recorded a medical history of the subject, reading the questions in standardised format, and recording the answers as given. In ease of prolonged hesitation or expressed doubt, individual guidance was given. The questionnaire is appended (Appendix III). The fora of the questions on cough, sputum, recent exacerbations of cough and sputum, recent chest illness and smoking history was almost identical to that in the Medical Research Council Questionnaire of Respiratory Symptoms (MRC, 1976). The questionnaire additionally asked about personal history of asthma, hay fever, eczema, perennial rhinitis, pneumonia, acute bronchitis, tuberculosis; m childhood history of bronchitis, unduly frequent colds, tonsillitis and ear infections; first degree relative history of chronic bronchitis. VAB.0001192196 4 I - ' m\ + 4i .* I ; i * * t * I *4 a* 4 i *. * * * t 4 11 * frequent chest colds, asthma, hay fever and eczema, and parental smoking habit*. Skin prick tests to mixed grass pollens, Dermatophagoides pterosyssinus, house dust mix and Aspergillus fumigatus (Bencard) were performed and read- at 15 minutes* Weal diameters of 4 am or more were taken as ap positive* Men for whom the control solution showed a response of 2 am or more were excluded from the analysis of skin test results* Forced expiratory flow/volume curves were recorded on a* rolling seal spirometer and X-Y recorder using air and following three vital capacity breaths of 80# helium/2G# oxygen Three technically satisfactory flow/ volume curves were recorded on air and two on the helium mixture* The curve selected for.analysis. was that with, the greatest sum of forced vital capacity (FVC) and forced expired volume, in one second (FEV^* Maximum flows at 5C# and 25# of vital capacity (i*e* when 25# of the FVC %^ has still to be expired) on air (V max__ and V max__) and helium mixture 50 75 (V maxCrt ,, and V max.. ,, ) and forced expired time were measured on thi curve peak expiratory flow rate (PEFR) was also though the slow response time of the X-Y plotter (relative to the rate ' changes in the flow rate measured) could have caused a systematic under estimation of the high flow rates* Single breath carbon monoxide transfer factor (TLCe) was measured i: duplicate using an automated spirometer system (PK* Morgan Ltd.)* Effective alveolar volume (V _^,) and Tuco were calculated for each manoeuvre and the means of the readings were analysed* 3*2*1 Methods of analysis m. * i For the four case/control studies the purpose was to compare men in the two groups (cases and controls) on the many measurements taken during clinical examinations* The emphasis was on an exploratory approach to identify suggestive patterns in the data* The scope of the statistical work carried out, and the methods used, are described in Appendix IV. The main results are presented descriptively in -this report in terms of average values and variability in each group* Observed average differences in the clinical measurements between cases and controls were tested for statistical significance'one at a time, using conventional^ 0001192197 12 Ii I j methods (chi-squared tests for binary variables, independent two-sample lung function measurements) In the context of this study these tests are not exact or optimal v but they do give reliable pointers to where real differences exist between the groups studied* Additionally, patterns of abnormality of lung function after taking account of the level of FEV^ were studied by examining the ratios of the r values of other lung function tests to the value of the FEVj (for example or ' Qa the assumption that the FEVt expresses an overall FEV1 L of lung function (of a non-specific type) and the pattern of the lung function ratios reflects type of abnormality, though did not necessarily express adequately its clinical severity* Predicted values for FEVX were based on regressions for age, height and weight in 148 non-smokers, excluding pensioners, studied- in the previous epidemiological survey. / ** The accuracy to which the sample of men seen represented the population under study was assessed by examination of their FEV1 recorded at the previous survey, in 1979* 3*3 Results * 3*3.1 Men with small rounded opacities Twenty-eight men in whose chest radiograph one reader reported small rounded opacities were seen, and 29 controls in whose radiograph all readers agreed no small opacities were present* Characteristics of these * groups are set out in Table 12* The majority were smokers* * * Sample bias * * Possible sample bias was examined by analysis of the FEV^ values of the year before* Differences of predicted from actual FEVj (FEV1 residuals) were calculated from regressions in the non-smokers studied in the previous survey* The distribution of FEV1 residuals of cases and controls is set out in Table 13* The mean residual FEVj of 28 men with small rounded opacities attending for examination was - 0*29 litres, whereas that for 3 men who were invited but did not attend was - 1.14 litres, suggesting some self-selection of iller men out of the sample * (R < 0*01)* The mean residual FEV^ of all men invited, whether seen or not, was lower than, but not statistically significantly different from, VAB.0001192198 H those not invited (- 0*42 litres and -- 0*72 litres respectively)* Thus the bias (mean FSVX residual of men seen was - 0*29 litres, and for sen not seen was - 0*8l litres, P < 0*01) was partly attributable to * selfseleetion for five iller men out of the sample* Amongst the controls, there was no significant difference in mean FEVj residual * between men seen and men not seen, but the distribution of FEVX % residuals of men invited but not seen again suggested some self-selection of ill men out of the study* Examination of the indices of PVC dust exposure available for these men from the previous study (Table 14) shows that men invited but not attending tended to have dust exposures in the lower ranges, suggesting * that the absence of these men was unlikely to have seriously affected the results of this study of dust effects on the lung* The mean residual FEV* in 1979 of all men with small rounded opacities invited for examination in 19&0, whether seen or not, was - 0*26 litres lower than that for all controls without opacities, a difference which was not statistically significant in these small samples, but which was of the some order as the difference found in the previous analysis of the much larger population* Comparison of men with small rounded opacities and controls Ken with small rounded opacities in the chest radiograph more often complained of chronic sputum production (9 men, 32#) than the controls (1 man, A first degree relative history of asthma, hay fever or eczema was reported by fewer cases (5 men, 21#) than controls (12 men, 46#), though this was not accompanied by similar differences in prevalences of personal history of allergic symptoms or positive skin tests* No other statistically significant differences in personal or family medical history were demonstrated (Table 15). * * Physical signs of' gross respiratory dysfunction were rare, and were associated with very low FEVt in both cases and controls. Crackles heard through the stethoscope in the middle or late parts of inspiration (heard separately or in combination) were more frequent in cases than controls (Table 15) Wheezes were not audible in these subjects. VAB.0001192199 1'+ 1 Differences between cases and controls in the mean observed results of the lung function measurements were small, and none reached A statistical significance at the % level (Table 16). Multiple regression analyses, taking account of age, height, weight and dust exposure confirmed the lack of any substantial or statistically significant differences between cases and controls (Appendix IV), nor were any mean differences in pattern of lung function between cases and controls identified. A The FEV} of seventeen cases and controls was less than. 8C0 of the * predicted value for age and height. Some had substantial impairment of lung function likely to be associated with symptoms, but these men were not significantly over-represented among those with small rounded opacities. The patterns of lung function abnormality in these men, in conventional clinical terms, were obstructive in six men v/FrsEVsVCsr, ratio less than 0.65), restrictive in seven men (ratio greater than 0.75) and intermediate in four men (ratio 0.65 - 0.75)* More detailed * examination of lung function patterns in these men did not demonstrate + any obvious relationships with rounded opacities. Thus although there was in men with small rounded opacities further evidence of abnormalities of the lung, namely increased bronchial mucus secretion and abnormal mechanical functioning of alveolar units (crackles), this lung damage was insufficiently gross to be associated with clinically important loss of lung function in excess of that related to age and smoking habit, and was not associated with a pattern of lung function abnormality which could be distinguished from that found in men without small rounded opacities. Comparisons of men with low or high dust exposures among nwa/* A 4* 4 me The men with small rounded opacities were divided into approximately equal groups of those with dust exposures less or greater than 15 dust index units, thus giving a group of men likely to include those whose small rounded opacities were related to dust exposure, and a group in whom they were not likely to be dust--related. The-distribution of dust exposures in these men is described in Table 14. VAB.0001192200 15 There was no suggestive evidence of differences in prevalences of symptoms and physical signs between those with low or high dust exposure after allowing for age differences* The means for the lung -function tests were -very- similar -between the two groups . * Patterns of lung function were- analysed by examining the test results higher in the high dust group (high dust group Tico ratio 10,28, low dust group 8,56 (Table 17, Figure 8))# After allowing for age and height differences, which were not themselves statistically significantly dust men remained statistically significant only at the 1C& level* When r 9- the cases with low dust exposure were combined with controls with low dust exposure and compared with eases and controls with high dust exposure, the difference was confirmed* A similar comparison of the effect of dust exposure within controls alone showed similar but slighter trends of those ratios, which could easily have arisen by chance* The forced expiration ratios were similar in all subgroups of men, including that group with no opacities and low dust exposure* Substantial impairment of lung function (FEVX less' than 80ft> of predicted) was not significantly more common in high or low dust groups* * We conclude that there was suggestive evidence that small rounded opacities 9 related to PVC dust exposure were associated, with a pattern of lung function abnormality characterised by relatively better preservation of gas transfer factor for the level of FEVj than men whose opacities were not related to dust exposure, but no differences in pattern of airflow during forced expiration* Furthermore there were no differences in pattern of airflow between these men. and controls without opacities, after allowing for the level of FEVX. This suggests that PVC dustrelated small rounded opacities were associated with less alveolar damage relative to the level of ventilation than rounded opacities not related to PVC dust, and that reduction of airflow rates during forced expiration tended to be uniform (concentric) These VAB.0001192201 features taken together suggest a physiological defect intermediate between an obstructive and a restrictive defect, probably the result of disease of very peripheral airways*. However, the lung damage * indicated by this defect appeared to be slight, for clinically important reductions of lung function were not found more commonly ' among these men with higher `dust exposure than those with lower dust exposure* 3*3*2 Ken with small irregular opacities e Eighteen men in whose chest radiograph one or more readers reported small irregular opacities were seen, and 20 controls in whose radio graphs all readers agreed no small opacities were present* Characteristics of these groups are set out in Table 18. Sacmle bias Possible sample bias was examined by analysis of the FEV*, values of the year before, as described for men with small rounded opacities. The mean residual FEVl of 18 men with small irregular opacities attending for examination was - 0*30 litres, whereas that for two men who were Invited but did not attend was - 0*53 litres* There was no evidence of self-selection of iller men out of the sample of cases, nor among the controls* Medical histories and lung function The prevalences of positive answers to selected questions on respiratory symptoms and personal and family medical history are summarised in Table 19* There were no differences in prevalences of chronic respiratory symptoms between cases and controls, but men with irregular opacities more often complained of recent chest illnesses causing time off work or recent increases in cough and phlegm* There were no statistically a- significant differences in history of allergic symptoms in men seen nor in results of the skin tests. A family history of asthma, hay fever or ecsema was more common among the cases, and family history of recurrent bronchitis or chest colds was marginally more common among them than among the controls* There were no important differences in * auscultatory signs between the groups; the slightly greater prtTalence of aid--inspiratory crackles wopg the controls seems an isolated finding and is difficult to explain* VAB.0001192202 1 I* J *. i 4 4 ri * 1 1 4 3 3- m Y*I V' .- * J 1*1 i * : /- +' v > r -t * * .. I t * ** 'I i * # # * 4 I * 1 .t * 17 % Differences in seen lung function test`results between cases and controls were small and did not approach statistical significance (Table 20)* Multiple regression analyses for each lung function # measurement, taking account of age, height, -weight, and dust index r confirmed the lack of any substantial or statistically significant differences between cases and controls (Appendix IV)* Examination of lung function patterns as described for the men with small rounded opacities shoved some suggestive but not statistically significant differences between men with irregular opacities and controls The pattern suggested more airflow obstruction for the level of FEV1 among cases than controls in that the FVC tsrtt ratio appeared to be slightly elevated and the pgy >0 and Vpgmya"x25 ratios slightly lower in the cases than controls (Table 21 )* The patterns were therefore examined among men whose FEV^ was less than of predicted, and who would therefore be * regarded as probably having abnormal lung function* The results indicated that the pattern of lung function associated with small irregular opacities and low FEVX was associated with an excessive reduction of relative expiratory flow rates at 5C& and 22& of vital capacity and an increase in FVC ratio when compared with men with low FEV1 but no opacities (Table 22, Figure 9) We conclude that the presence of small irregular opacities in the chest radiograph was associated with history of recent chest illness, and when these opacities were accompanied by clinically abnormal lung function, the pattern of functional abnormality was of obstructive type# These men were likely to have been suffering from chronic n.on--specific airflow obstruction, not related to PVC dust exposure* 3.3.3 smokers with low FEV Thirteen non-smokers with large negative residual FEV* and high index dust, exposure were seen, and twelve non-smoking controls with large negative residual FEV, and low index of dust Characteristics are Matching was imperfect rols were younger, taller and heavier than .the cases* Two cases and three controls did not attend for examination their FEVj res previous year's measurement, and dust exposures VAB.0001192203 U* p Prevalences of symptoms were low, and there were no significant differences between eases and controls (Table 24). The medical histories did not suggest that these men were suffering from symptoms of any respiratory disease* * * > * There appeared to be substantial differences between cases and controls in some of the lung function results, but this could be attributed to the age and height differences* After allowing for age and height differences, no statistically significant differences in lung function remained (Table 25 and Appendix TV) . The pattern of lung function after allowing for level of FEVj was also examined, and no distinctive differences in pattern were seen between cases and controls. None of the cases had an FEVj of less than 8C& of predicted value. Of the controls three, were below 8(#, none below 75&. We conclude that among non-smokers selected for large negative residual 4 FEVj, no detectable differences in symptoms, physical signs or pattern of lung function could be found between men with high estimated dust exposure and those with low exposure* normal respiratory health. These men were probably all in * 3*3*4 Smokers with low FEV, Eighteen current cigarette smokers with large negative residual TEV1 and high, index of dust exposure were seen, and 18 smoking controls with large negative residual FEV* and low index of dust exposure. Five cases and eight controls were invited but not seen* Their FEVX residuals calculated from measurements made the previous year were similar to those of the men who attended* Characteristics of these groups are set out in Table 26* controls tended to be taller* Age-matching was close, though * personal history of hay fever or perennial rhinitis was more common Bong the controls than the cases past history pneumonia or pleurisy and late inspiratorv crackles heard auscultation were slightly commoner among the cases than controls, but these differences did not reach statistical significance at the 556 (Table 27)* Respiratory symptoms were more common in both cases m and controls than the non-smokers* VAB.0001192204 19 No substantial or statistically significant differences of lung function between cases and controls were apparent (Table 28)* A Multiple regression analysis for each lung function test allowing for age, height, weight and FEV* confirmed that differences in lung function between the groups were small and not statistically significant (Appendix IV) The pattern of lung function after allowing for level of FEVX was also examined, and no distinctive differences in pattern were seen between cases and controls* Substantial differences in these patterns between smokers and non- smokers were seen (Figure 10), but these have not been analysed in detail* Seventeen of the smoking cases had an FEVX less than 8C& of predicted, and six were under 6C$. Fifteen of the smoking controls were b under 8C6& and seven under 6($ Thus the selection procedure had f selected men with clinically important impairment of lung function equally among both cases and controls* We conclude that among smokers selected for large negative residual FEV* no detectable differences in pattern or level of lung function would be detected between men with high estimated dust exposure and those with exposure history of asthma among the controls probably indicated avoidance of dusty conditions by these men* view of this the slightly higher prevalences of history of acute bronchitis, pneumonia or pleurisy and of crackles heard through the stethoscope among cases than controls could be related t the differences in prevalence of asthma in these groups rather than an effect of PVC dust exposure, although the timing of the crackles late inspiration suggests that they could have been the result of du exposure, rather than the result of smoking, which tends to be relat inspiratory crackles however, there were no other important differences between those men with high or with low dust exposure, and we conclude that these men did not have recognisably different clinical syndromes from those with low exposures* VAB.0001192205 4. discussion 4 Our previous study of 8l8 present end past workers at a factory making polyvinylchloride (PVC) had shown that estimates of exposure to respirable PVC dust were inversely related to lung function after allowing for the effects of age and smoking* and related to prevalence of slight abnormalities of the chest radiograph; one of three experienced readers found that prevalence of small rounded opacities category C/l or greater was related to dust exposure after allowing for The present studies extended the sample of the study population to include all other members of the current workforce who had worked at any time in the dustier plants in the factory* to examine the ' prevalence of respiratory abnormality in these men* and to examine in this extended population and in the original study group the relation ship between dust exposure and radiological appearances. Also men selected from the first study on the basis of radiological and functional abnormality have been examined in detailed case/control studies to attempt to identify clinical syndromes associated with PVC dust exposure* and estimate if possible the clinical severity of the effects of the dust on the lung. These an extended population at the same factory an inverse relationship of lung function with estimates of exposure to PVC dust* and confirm the relationship between dust exposure and the presen small rounded opacities in the chest radiograph The radiographs from the original study population were re-read in the present work by the three experienced readers who read them previously* 4 and were also read twice by a self-trained panel of readers. The majority of readers found that the prevalence of small rounded opacities was related to age* and in some readings relationships were found between prevalence of small rounded opacities and estimates of PVC dust exposure, after allowing for age. Not only did readers differ from each other in the detection of the relationship with dust exposure* but also some readers detected this effect on one reading but not on the other. The # three experienced readers recorded higher prevalences of small rounded opacities on the second reading than the first* and this is likely to h have been the result of a greater awareness of this type of opacity * VAB.0001192206 4I 4 a 21. & induced by knowledge and discussion of the results of the first reading. The reader who had first detected the relationship with dust exposure recorded a much greater prevalence of small rounded opacities on the *p second reading, and in spite of his change in interpretation of small rounded opacities, the relationship of these opacities with dust exposure was again found for his readings,, this time between a higher category of profusion and dust exposure. The reasons for these various differences in detection of these effects are likely to include the slight degree and non-specific nature of the radiographic appearances related to exposure to PVC dust, which are indistinguishable from those related to age, as well as variability in readers' interpretations of the appearances. This helps to explain the differing results reported by other workers (VERTKIN et al., 1970; LIUS et al.. 1976; HAPP et al., 197^), and emphasises that in studies of the relationship between non-specific radiological appearances and dust exposure, it is important to take account of the confusing effects of other age-related factors. Xn the case/control studies we examined men most likely to have been affected by PVC dust, by selecting men whose radiographs were thought to show small rounded opacities, and also men whose estimated PVC dust exposure was in the higher range and whose FEVX was at the lower end of the range after allowing for age, height and weight, together with suitably matched controls. The men with small rounded opacities had slightly increased prevalences of sputum production and of crackles heard through the stethoscope, indicating disturbances of the secretory function of the lung and of the mechanical function of alveolar units, thus confirming that these radiographic abnormalities were associated with minor functional abnormalities. The audible crackles tended to be of the end-inspiratory type, usually associated with diseases in which diffuse fibrosis of the lung occurs. (ipRGACS, 1978). In spite of this we were unable to show that important clinical illness was related to exposure or to the presence of small rounded opacities in the chest radiograph. The previous epidemiological study of 818 men had shown that the presence of small rounded opacities in the chest radiograph was associated with slightly lower lung function than would be expected for age, height, smoking habit and dust exposure. The 1192207 case/control study has not confirmed this difference, probably because of the smaller numbers of men in this study, and because some iller men did not attend for examination* .These missing men, however, * tended to have low estimates of FVC dust exposure, so it is unlikely that their omission has caused us to underestimate the seriousness of the effects of FVC dust since their radiological appearances were likely to be related to age rather than PVC dust exposure. We conclude that the presence of small rounded opacities of low categories of profusion in the chest radiograph, while associated with a slight reduction of lung function detectable when these men were studied as part of a large population, does not in an individual worker give cause for concern that exposure to FVC .dust has seriously affected his respiratory health. 4* The ease/eontrol studies of men whose lung function was at the lower en of the range did not show any characteristic differences in symptoms or pattern of lung function in men with higher dust exposures compared wit men with low dust exposures and similarly low values of lung function. While the non-snokers selected in this way were probably all in normal respiratory health, the substantial amount of clinical illness found among these selected smokers was probably the result of their smoking habit, and showed no characteristic features which would suggest that their illness was the result of exposure to PVC dust. The one exception to this was a slightly greater prevalence of crackles heard through the stethoscope of late inspiratory type. While this could partly he the result of a greater prevalence of men with allergic conditions in the control group (possibly self-selected by their avoidance of dusty conditions), the timing of the crackles in the respiratory cycle suggests that these could have been the result of FVC dust exposure (since smoking tends to cause early and mid-inspiratory crackles)., a conclusion consistent with the findings in the men with small rounded opacities, However, there were no other important differences between these ten with high or low dust exposure, and conclude that these men with high exposures did not have recognisably different clinical syndromes from those with low exposures. i These ease/control studies, while showing no obvious physiological differences between men with small rounded opacities in the chest VAB.0001192208 I +- * * ft-#* 4 radiograph and controls of similar age and smoking habit but normal radiographs, did suggest among the men with small rounded opacities a difference between men with lower and men with dust exposure the men with higher dust exposures had a pattern * of lung function abnormality in which the. flow rates during forced expiration were normal after adjusting for the level of FEV1 f but the gas transfer factor appeared to be relatively better preserved than the FEVt. This form of analysis was an exploratory one, and was based on a convention that the level of FE^ represented the overall level of lung function, reasonably sensitive to all types of lung function defe and that examination of tyoes of lung functional abnormality among oth lung function tests should first take account of the level of FEV1 A simple if crude method of taking account of the level of FEV1 when considering the value of a second lung function test is to divide the second test result by the value of the FEVt to* obtain a testsFEVj^ ratio In this exploratory analysis the ratios of various lung function tests to the FEV^ have been compared In the ease of the men with small rounded opacities and high dust exposure the values of the ratios of flow rates during forced expiration were very similar to those for men with low dust exposure, whether or not they had small opacities, but there was suggestive evidence that the ratios of gas transfer factor to FEVX were higher, implying that the gas transfer factor was relatively better preserved than the FEV1 This pattern of lung functional abnormality could therefore be considered intermediate between an obstructive type * or restrictive type, in that the lung volumes and expiratory flow rates were reduced in a concentric way, a feature of restrictive defects, but the gas transfer factor was not reduced proportionately, even though this expected in a classical restrictive The results large epidemiological survey had also suggested an intermediate or mixed type of defect, for the FEV* and the forced vital capacity were both Inversely related to dust exposure, but the ratio between them was not* It is probable that this defect was the result of disease of the airways accompanying alveolar damage This is in keeping with what is so far known about the histological appearances of the lungs in exposed animals and man, where fibrosis has not been a w prominent feature (FRONGIA et al*, 197*M ARNAUD et al, 1978) and in seme animal studies an inflammatory reaction has been observed round respiratory terminal bronchioles (AGARWAL VAB.0001192209 Men with small irregular opacities in the chest radiograph (not thought to be related to PVC dust exposure) and reduced lung function had an obstructive type of functional defect, with relative preservation of the gas transfer factor, demonstrated clearly by the analysis of lung + function test:FEV4 ratios* We conclude from the large epidemiological and small case/eontrol studies that exposure to PVC dust in the concentrations experienced in this factory caused a small but detectable reduction in lung function in the study population, and caused an increase in prevalence of slight abnormalities of the chest radiograph. The clinical features of the non specific respiratory disease associated with dust exposure additionally included an increase in prevalence of productive cough in those with abnormalities of the chest radiograph, an increase in prevalence of crackles heard through the stethoscope during mid and late inspiration, and an average pattern of lung functional abnormality characterised by a uniform reduction of ventilatory capacity, with relative preservation of gas transfer factor. Important clinical illness or serious reduction of lung function has not been shown to be related to PVC dust exposure at the levels occurring in this factory, and in view of these case/control studies, we conclude that PVC dust exposure at these levels is unlikely to have caused serious disability* While differences in PVC formulations or levels of exposure could have accounted for some of the conflicting results reported in the literature, it is probable that differences in study design have been the main cause 4 of these inconsistencies. The high prevalence of abnormalities of lung function occurring as a result of factors other than PVC dust exposure indicates the importance of making comparisons with careful estimates of # dust exposure or carefully selected control groups. Where this is not done, the effect of PVC dust exposure may not be detected, especially when the study population has been exposed to other potentially noxious agents (GAMBLE et al more properly attributed agents including smoking or general atmospheric pollution may be thought to be related to PVC dust exposure (LXLIS et al view degree of the reduction of lung function we found PVC dust exposure, studies in which defined "abnormal" levels of lung function are compared with exposure would be unlikely to demonstrate - of PVC dust on the lungs* VAB.0001192210 The possibility of a rare idiosyncratic disabling response to PVC dust cannot be entirely excluded, and it should be borne in mind that a case of diffuse lung disease in a PVC worker has been reported (AENAUD et al , 1978) 1' and in our first epidemiological'Study,, one man was found to have clinically obvious lung fibrosis. The relationship of this man's illness to his PVC dust exposure was not clear, though further details of his case will be reported. It is conceivable that different formulations of PVC or exposure to other agents may be related to apparently idiosyncratic reactions in PVC workers. In our studies it has not been possible to take account of the effects of differences in -t- formulations of the PVC, nor of the effects of exposure to vinylchloride monomer, except in a limited examination in the previous study of the effect of probable exposure to monomer on the relationships between PVC dust exposure and disease. * The prevalence of respiratory disease at a second factory manufacturing PVC was low, possibly related to the younger age of the workforce, and a 1 greater proportion of non-smokers among them, and appears to give no cause for concern about an industrial health hazard there* VAB.0001192211 i i- <* 1 * * 27 * 4 4* * 4 I * *1i # I* a t 41 4J to I J ; f 4 m * #4 m f : * i * 4 h 4+ * ACKNOWLEDGKEWTS < This study was carried out with the full co-operation of the management and workforce of ICX Ltd., to whom we are grateful for financial support* We are grateful to Dr. J.A. Dick, Dr. J.G. Bennett and Dr. D.J. Thomas for the epidemiological radiographic readings and to Dr. A. Seaton and Dr. M. Jacobsen for much advice and help. VAB.0001192212 REFERENCES 9 # AGARWAL DK, KAW JL, SRIVASTAVA SP, SETH PK (1978) Some biochemical and histopathological changes induced by polyvinyl chloride dust in rat lung. Environmental Research; 16: 333 * 341. ARNAUD A, FOMMIER de SANTI P, GARBS Lt PATAN H, CHARPIN J (1978) Polyvinyl chloride pneumoconiosis* Thorax; 33s 19 - 25# * COPLAND- L, BURNS J, JACOBSEN M (in press) Classification of chest radiographs for epidemiological purposes by persons not experienced in the radiology of pneymoconiosis. FORGACS P (1978) Lung Sounds. London: Bailliere Tyndall. FRONGIA N, SPINAZZOLA A, BUCARELLI A (1974) Lesioni polmonari sperimentali da inalazione prolungata di PYC in ambiente di lavoro. Medicina del Lavoro; 65* 321 342. * GAMBLE J, LIU St MeMICHAEL AJ, WAXWEILER RJ (1976) Effect of occupational and nonoccupational factors on the respiratory system of vinyl chloride and other workers Journal of Occupational Medicine; 18: 65<9 - 670. W INTERNATIONAL LABOUR OFFICE (1972) U/C International classification of radiographs of # the pneumoconioses, 1971. Geneva: ILO, 1972. (Occupational Safety and Health Series No. 22). 29. VAB.0001192213 LILIS R, ANDERSON K, MILLER A, SELIKOFF IJ (1976) Pulmonary changes among vinyl chloride polymerization vorkers * Chest; 69 (Suppl. 2, Feb.): 299 - 303. KAPP C, FABBRI L, ROSSI A, KORO G, CERVI G (1978) Alterazioni funzionali respiratorie da esposizione eronica a cloruro di vinile monomero e polimero. Medicina del Lavoro; 69* 151 - 162* MEDICAL RESEARCH COUNCIL (1976) Report of the Working Party on research into chronic bronchitis. London: Medical Research Council. RAE S, WALKER DD, ATTFIELD KD (1971) * Chronic bronchitis and dust exposure in British coalminers. In: V/alton WH, ed. Inhaled Particles III. Cld Woking, Surrey: Unwin Bros: fiP-3 - 2-94 SCUTAR CA, COPLAND LH, THOENLLTr PE, BURLEY JF, OTTERY J, ADAMS WGF, BENNETT B (1980) Epidemiological study of respiratory disease in workers exposed to polyvinylchloride dust. Thorax; 35: 644 - 652. VER7KIK YI, MAMONTOV YR (1970) The state of the bronchi and lungs in workers employed in the manufacture of nolvvinvl chloride articles. Gigiena Truda i Professional ' nye Zabolevaniya; 14: 29 - 32. 51 ft '* 4 *1 9 #* -> , i* ft / * 515' men seen in 229 further men first survey seen in second 4 survey i ft Mean (3D) Mean (SD) Age * 46.4 (12.1) 36.5 (13.7) Years worked at the factory *r ^* r Index of dust exposure a * * * l4.0( 8.6) m s 12.9 (12.6) * * ** " . *, s 9.6 ( 9.7) * 6.1 ( 8.3) ** ft Non-smokers 148 ) * Cigarette smokers and other current smokers Ex-smokers Prevalence of chronic cough and/or sputum <36 *}11 234 *" + H 192 (53* ) (28* ) * (2# > 69 (3C% ) 117 (51* ) 43 (19* ) 34 (IS* ) TABLE 1 .i 3 * ^* * .J "i t*# ft # f populations studied in first and second surreys factory* ** VAB.0001192215 _ fe + * * 01 >% T3 9 t *oco oo * c 4> <o\J* 1 <M + o H4>49} *9 O H *C9i CL +O c gO *0H -01 i-l H9 9(X * a cH o m9 W 00 Pm 9 O9 eOH -<tHo4 + O 9 U <8 CL > > *4T93 >s to *C9 eoB o4O> oC<VTjI* CO W -*-> e <0 h QH LOh * TABLE 3 Ages and smoking h a b its o f 127 men a t the second fa c to ry VAB.0001192216 i 33 * VAB.OOOl192217 TABLE 5 D e ta ils o f the s ix men w ith FBVj below o f the predicted value Category of profusion of small rounded opacities 1i f * *! 'H * T 15 1st reading 2nd reading 814 797 4 (0.56) 21 (2.656) 2 (0.256) 18 (2.256) 17 1st reading 2nd reading 768 326 50 (6*156) 492 (60.156) 10 (1.256) 17 (2.196) TABLE 6 Prevalences of small rounded opacities recorded by three experienced medically qualified readers on first and second readings of 818 chest radiographs. Reader ' Category of profusion of smallL irregular opacities 0/0 0/1 or greater I/O or great ez; 03 1st reading 2nd reading 786 788 32 (3.954) 30 (3.754) 7 (0.90 12 (1.554) 15 1st reading 2nd reading * 765 781 53 (6.54) 37 (4.50 46 (5.6*)' 34 U.z') 17 1st reading 2nd reading 1 789 780 29 (3.554) 38 (4.696) 6 (1 C56) 7 (0.954) TABLE 7 Prevalences of small irreeular opacities recorded by three experienced medically qualified readers on first and second readings of 818 chest radiographs. 4 4 VAB.0001192218 I t 35 U 1st reading 2nd reading 782 789 34 (4.#) 27 (3.390 L5 1st reading 2nd reading 813 811 3 (0.496) 5 (0.696) TABLE 8 Prevalences of small rounded trained readers on first and radiographs. 1 34 (4*30 27 (3.396) 2 (0.296) 2 (0.296) ri VAB.0001192219 p Category of profusion of small Irregular opacities 0/0 0/1 1/0 or greater LI 1st reading 2nd reading 778 791 38 (4.796) 25 (3.190 37 (4.90 25 (3.196) L2 1st reading 2nd reading 813 812 3 (0.496) 4 (0.9<) 3 (0.496) 2 (0.296) L3 1st reading 2nd reading 14 1st reading 2nd reading X.5 1st reading 799 17 ( 2.196) 812 4 (0.96) ( . )809 i 7 0 96 810 6 (0.796) 813 3 (0.456) 814 2 (0.30 10 (1.296) 4 (0.96) 7 (0.96) 6 (0.796) 2 (0.296) 2 (0.296) TABLE'9 Prevalences of small irregular opacities recorded by five selftrained readers on first and second readings of 816 chest radiographs* VAB.0001192220 4 t j i _# 4* 37 Category1 of profusion of small rounded opacities Header 15 17 TABLE 10 a 0/0 O/l or greater 1/0 or greater * - - .. 223 t t 6 (2.6#) m 2 (0.906) 228 1 (0.46) 1 (0.490 114 115 (50.36) 4 (1.796) * Prevalences of small rounded opacities recorded by three experienced medically qualified readers in the chest radiographs of 220 additional men. Category of profusion of small irregular opacities 1/0 or greater 03 224 5 (2.2*0 1 (0.496) 15 224 5 (2.36) 5 (2.36) 17 227 2 (0.996) 0 (0.06) TABLE 11 Prevalences of small irregular opacities recorded by three experienced medically qualified readers in the cheat radiographs of 229 additional man. # VAB.0001192221 % Number of men Age (yr) Height (cm) Weight (kg) Dust exposure (yr x mgm Non-smokers Current smokers Ex-smokers Currently employed (at time of original surrey) Pensioners Mean (SD) 28 Controls Mean (SD) 1i 29 52.4 ( 9.1) 51.3 ( 9.1) 172.7 ( 6.2) 174.0 ( 6.0) 4 80.8 (10.A) 78.1 (12.7) 2*t.3 (20.2) 18.6 (14.5) TABLE 12 Study of men with radiographic small rounded opacities characteristics of groups studied. VAB.0001192222 39. A 1979 FEVj re s id u a ls ( l i t r e s ) * C ts *f&e i I ion* V CmM O 4f34t p4uiafc*>sO 1 I ! | I icn^ o V mo a*e ae oa >c o0wo} OC>> 4oe4 f9t ' a<a1 raOHs 44 uco sn SP o a tehfjt iHhO 9S Co o *9 a a> 9 *ur 8 S. raaha-(9HOd4> **i99Oo9* H 9e ON + o ocH f9t aaos *oM bC O C K\ W4 6-* *I VAB.0001192223 Index of dust exposure Cyr x mgm"*) 0 - 14.9 15 - 29-9 30 - V*. 9 * 13 6 e 4 5 Cases invited but not seen 41 00 Controls seen 16 Controls invited but not seen 2 6 4 52 11 TABLE 14 Distribution of dust indices among men with small rounded opacities and controls. VAB.0001192224 History of chronic cough History of chronic sputum f all grades Dyspnoea i grade I 1 grades IX or III History of asthma History of recent chest illness History of acute bronchitis, pneumonia or pleurisy History of hay fever or perennial rhinitis Childhood history of recurrent bronchitis, asthma, frequent chest colds or eczema Family history of recurrent bronchitis or frequent chest colds Family history of asthma, hay fever or eczema ParentCs) smoked Auscultatory crackles: Inspiratory: early mid late Expiratory All Wheezes present Skin prick test(s) positive A Cases Controls 9 (28) 9 ( ") 13 ( ") 11 ( ") 2 ("> 2 ( ") 15 ( ") 13 ( ") 5 (29) 1 ( ") 11 ( ") 7 ( ") 4 ( ") 1 ( ') 10 ( ") 10 ( ") 4 ( ") 7 ( ") 5 ( ") 4 ( ") 13 (25) 12 (28) p 1 5 (24) 12 (26V 22 (24) 25 (28) 2 (27) 8 ( ") 7 ( ") 0 ( ") 9 ( ") 0 ( ") 5 ( ") 3 (29) 2 ( M) 2 ( ") 1 ( ") 5 ( ") 0 ( ") 3 ( ") TABLE 15 Men with small rounded opacities : Summary of selected results from medical history questionnaire, physical examination and skin prick tests. Figures in parentheses are numbers of men for whom the result was recorded* (Statistical significance of group differences - 0*1 > P > 0.05; ** 0*5 > P > 0.01; P > 0.01) VAB.0001192225 TABLE 16 28 men with small rounded opacities and 29 controls : mean observed values of lung function data. Differences between cases and controls slight and not statistically significant* VAB.0001192226 Number Age (yr) Height (cm) Dust exposure (yr x mgm") Difference of FEV* from predicted (litres) FVC/FEV1 PEFR/FEV1 m V.ax_0^FEV1 tfmax^VFEV^ Tlco/FEVj Kco /FEVt Va/FEVj Lower dust exposure Mean (SD) Higher dust exposure Mean (SD) 13 15 49 55 173 7.5 - 0.30 (0.4l) 172 39 - 0.52 (0.61) 1.45 (0.06) 2.40 (0.24) 1.10 e 0.29 (0.19) (0.06) 8.56 0.63) 1.52 (0.56) 1.95 (0.25) 1.37 (0.13) 2.40 (0.23) 1.30 (0.28) 0.26 (0.09) 10.28 (2.51)** 1.85 (0.68) 2.19 (0.43) TABLE 17 Men with small rounded opacities ; lung function results divided by FEV1# (The Tlco and Kco results for the low dust exposure group are based on 12 technically satisfactory readings.) group significance VAB.0001192227 r * * *+1 VAB.0001192228 I 45 A Controls nistory- of chronic cough History of chronic sputum f all grades Dyspnoea < grade I l grades II or III History of asthma History of recent chest illness History of acute bronchitis, pneumonia or pleurisy History of hay fever or perennial rhinitis Childhood history of recurrent bronchitis, frequent chest colds, asthma or eczema Family history of recurrent bronchitis or frequent chest colds Family history of asthma, hay fever or eczema Parent(s) smoked Auscultatory crackles: Inspiratory: early mid late Expiratory AH Wheezes present - Skin prick test(s) positive 08) 5 ( "> 8 ( "> 4 ( ") 4 ( ") 0 ( ") 9 ( ") 11 ( ") 7 ( ") 1 ( ") 15 ( ") 12 (17) 17 08) 5 08) 2( ) 5( ) 0( ) 10 ( ) 2() 2( ) 5 (20) 5 ( ") 8 ( ") 3 ( "> 3 ( ") 2 ( ') 4 ( ") 12 ( ") 6 ( ") 3 ( ") 11 (19) 5 06) 17 09) 2 (20) 7 ( ") 9 ( ") 0 ( ") 11 ( ") 0 ( ") 4 ( ") TABLE 19 Men with small irregular opacities : Summary of selected results from medical history questionnaire, physical examination and skin prick tests* Figures in parentheses are numbers of men for whom the result was recorded* (Statistical significance of group differences -- * 0.1 > P > 0.05; 0.05 > P > 0.01) VAB.0001192229 Controls Kean (SD) FEV, (litres) 2.6 (0.9) 2.6 (0.6) Difference from predicted FEV, (litres) - 0.61 (0.64) - 0.48 (0.65) FVC (litres) (17 cases) (19 controls) 3.9 (1.1) 3.5 (0.8) PEFR (litres/sec) (19 controls) tfinaxL-. (litres/sec) (17 cases) (19 controls) 6.4 (1.9) 2.5 (1.4) 6.5 (1.4) 2.7 (1.0) VmaXgc (litres/sec) (17 cases) (19 controls) *""50 Helium (17 cases) (16 controls) 0.7 (0.4) 0.8 (0.3) 3.5 (2.2) 3.8 (2.0) * Vmax... li-mi, (lit**es/sec) (17 cases) (16 controls) 0.9 (0.5) 0.9 (0.5) Tice ml/rain/cim Hg (17 cases) (18 controls) V* ff (litres) (17 cases) (18 controls) 24.6 (6.2) 25.6 (5.4) 3.8 (1.0) 5.7 (1.0) * TABLE 20 18 men with small irregular opacities and 20 controls: mean observed lung function data. No statistically significant differences between groups VAB.0001192230 k? A Cases Mean (SD) Controls Mean (SD) FVC/FEVj PEFR/FEVl V maxoc/FEV1 O Tlcc/FEVx Kco/FEVx Va/FEVx * 1.58 (0.42) I.Vl (0.20) 2.57 (0.43) 2.56 (0.37) 0.89 (0.34) 1.02 (0.28) 0.25 (0.07) 0.28 (0.09) 10.8 (4.6 ) 10.14 (3.4 ) 2.0 (1.2 ) 1.8 (0.8 ) 2.6 (1.3 ) 2.2 {0.5 ) TABLE 21 Patterns of lung function of all men with irregular opacities and controls* Differences not statistically significant* Numbers of men consistent with those in Table 20. VAB.0001192231 i Number 4 Age (yr) Height (cm) Dust exposure (yr x mgm~9) Difference of FEV* from predicted (litres) FVC/FEVj FEFK/FEVj. ^ max-g/FEVj fr raaotjj/FEVj Ttco /FEVi * Kco/FEVj P Va/FEV1 Cases Mean (SD) 6 Controls Mean (SD) 1 ~L fT . _~[ . 8 1 IIII 1 68.8 61.1 167.7 173.6 db 15.0 13.7 - 1.35 (0.46) 1.99 (0.43) - 1.04 (0.54) .. 1.35 (0.39) * 2.93 (0.44) 2.72 (0.39) 0.57 (0.27) 1.03 (0.25) * 0*20 (0.05) 0.26 (0.05) V CVI tf\ M* UN 11.5 (5.2 ) 3.2 (1.3 ) 2.3 (1.0 ) 3.9 (1.5 ) 2.3 (0.7 ) TABLE 22 Patterns of lung function among men with and without small irregular opacities and FEV1 less than 80# of predicted* (Statistical significance of differences between group averages - 0*1 > P > 0*05; * P < 0*05) VAB.0001192232 49 Number of men * Age (yr) Height (cm) Weight (kg) Dust exposure (yr x mgra"3) Currently employed "Leavers" Pensioners Cases { Controls Mean (SD) 1 Mean (SD) 13 12 46.1 (12.8) 34.3 (8.2) 1- 172.2 ( 6.9) 177.1 (8.1) ?6.4 (10.6) 81.8 (8.0) 17.2 ( 8.8) 1.4 (1.2) 11 10 02 20 TABLE 23 Characteristics of non-smokers with large negative residual FEV1, cases with high dust exposure, controls with low dust exposure* VAB.0001192233 Controls Number of men History of chronic cough History of chronic sputum Dyspnoea (none gre grade I) 4- History of asthma than History of recent chest illness History of acute bronchitis, pneumonia or pleurisy History of hay fever or perennial rhinitis Childhood history of recurrent bronchitis, frequent chest colds, asthma or eczema Family history of recurrent bronchitis or frequent chest colds Family history of asthma, hay fever, eczema Parent(s) smoked Auscultatory crackles: Inspiratory (late only) Expiratory All Wheezes Skin prick test(s) positive TABLE 24 Non-smokers with large negative residual FEVX, cases with high dust exposure, controls with low exposure: Summary of selected results from medical history questionnaire, physical examination and skin prick tests. None of the differences reached statistical significance at the 10# level. VAB.0001192234 51 9 -3 * P<PS 9 **9ri bHO P9S 9 9 *4 -o W-OSi *C9 -49* 3 to 9 M OU o *9m <to *A9 od o CcQ 9 9 SC 5 s A V rp VA CM CMAV oAV A 5? a Va ao O# o oooooooO i11 1 <49 (90 -P TJ bC CO 9h Q PS CQ *9 w T9 4) 3 9 *9t<O **(9Q0 OC 9A 9 cm T3 p PoooSf h jS A (0 bO Q. 9 9 TJ PS *9 <o** 9h TJ 0 Sg tt.fg <8h M'S 9 40 9 (4 c o hC r*O- H 9 En 9A > ,, hO <2 O H +> o ^ C0 +J LA fl bC-^ O Sco ,H2 BN. g 9- 2 CO TJ * jC9 C 9 9 U9 <2 9 9 !S O C K\ 9O 9 *9 9 9 iou 99 9 (4 9> 4J 9 bO 9C 9bC J. p9-C _ *? auto ao a o> *(ao0 t0ao) >9 to d aa to a o & s 44>) * *9 9 9P .Q U 0 OH TJ 9 s% 9 MPu 9(4 coi CAm 9 Cm Q*H s t a t is t ic a l s ig n ific a n c e a t the 1C&S le v e l I ir\ cm HpS 63-t VAB.0001192235 h r # T s * CO CO CO A 00 00 & ' 1 * ' t 4 1 r 1 k ,i ' i a No. of men b> Age (yr) * Height (cm) Weight (kg) Cases 18 a Controls i 18 b P 171.9 (5.1) 173.2 ( 9-1) ?4.9 (8.8) ?4.3 (21.6) Dust exposure (yr x mgm"s) mm Currently employed * nLeavers" Pensioners 25.9 (7.7) 4.6 ( 2.6) 17 + 1 b 0 16 1 1 TABLE 26 Characteristics of smokers with large negative residual FEVj, cases with high, controls with' low index of dust exposure* VAB.0001192236 53 ft Cases 18 10 9 10 6 4 0 11 14 2 5 Controls ~u 18 11 8 9 6 3 2 10 9* 8** 5 TABLE 11 12 87 16 18 34 13 7 0 3 2* 0 10m 5* 25 14 Smokers with large negative residual FEV1, cases with high dust exposure** controls with low exposure atistical significance of group 0.1 > P > 0.05; ** 0.05 > P > * VAB.0001192237 r 9**1 FEV, (litres) FVC (litres) FEFR (litres/sec) VmaXeQ (litres/sec) )Vm ( 11 " Helium (1? cases) Vm Helium (litres/sec) (17 cases) Tecs ml/min/am Hg (litres) Cases Mean (SD) Controls Mean (SD) 2.2 (0.6) 2.3 (0.9) 3.5 (0.8) 3.6 (0.8) 5.1 (1.3) 5.7 (1.6) 1.6 (0.8) 1.9 (1.3) 0.5 (0.2) 0.6 (0.4) 2.1 (1.1) 2.5 (1.7) 0.3 (0.2) 0.7 (0.6) 24.2 (5.9) 24.6 (7.5) 5.7 (0.8) 5.6 (0.9) TABLE 28 iction results for smokers FEVj f 18 cases with high dust exposure Is with low dust exposure* No significant cases and controls VAB.0001192238 55 * * Figure 1 Age (years) # Prevalences of small rounded opacities category C/1 or greater recorded by three medical readers on second reading, compared with age* Figures on graph are numbers of men. 1.5 7.5 12 18 SO Figure 2 Index of dust exposure (years x mg. m j Prevalences of small rounded onacities category 0/1 or greater recorded by three medical readers on second reading, compared with index of dust exposure. Figures on graph are numbers of men. VAB.0001192239 56. / j. Figure 3 Age (years) Prevalences of email rounded opacities category 1/0 or greater recorded by three medical readers on second reading, compared with age* Figures on graph are numbers of men* * Figure k 1.5 7*5 12 18 30 Index of dust exposure (years x mg. m3 ) Prevalences of small rounded opacities category 1/0 or greater recorded by three medical readers on second reading, compared with Index of dust exposure* Figures on graph are numbers of men* t VAB.0001192240 # p Figure 5 Age (years) Prevalences of small irregular opacities category 1/0 or greater recorded by three medical renders on second, reading, comnared with ge. Figures on graph are numbers of men. * Figure 6 Age (years) * Prevalences of small rounded opacities category C/1 or greater recorded by five lay nanel readers on second reading, compared with age. Figures on graph are numbers of men. * VAB.0001192241 J * * 58 l -f 1.5 7.5 12 18 30 Index of dust exposure (years x mg. m -3 ) Figure 7 Prevalences of small rounded opacities category 0/1 or greater recorded by five lay panel readers on second reading, compared with index of dust exposure. Figures on graph arc numbers of men. PEFB VMaXg0 VMaXjg TLCO KCO VA Figure 8 Lung function values relative to FEV. in men with rounded opacities (* P < 0.05) VAB.0001192242- p * t tp p '* ' >*t -J * * 'I -TM -*** % '* * " *. M "- ' * - *** * * t > i>t ,v* -'* ;t i-V-l :r ^ p' ^ ** * ":-`**'iv/'uv*'*^ii -:i t* - i , 4 *" *1 -# -* * I V" : -\ 3lf *i i *4 *\ tF * - : L j -, i, ` , . ' .* 59. A 2 Ratio 1 0 Figure 9 VMaXgQ VMax2- VC TLCO KCO VA Lung function values relative to FEYj in men with irregular opacities and FEV less than 80% predicted. VAB.0001192243 .1 t APPENDIX I REGRESSIONS FOR LUNG FUNCTION FOR 1,047 MEN Height Weight Age r Non-Smokers Ex-Smoke rs ] Other Smokers ^ Cigarette Smokers Lifetime Cigarette Consumption (19000 packs) Dust Index (yr x mg/m*) 0.0426 - - 0.0014 - 0.0285 - O.O37O - 0.0483 - 0.0397 - 0.0143 - 0.0033 < 0.0001 n.s. * < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.0001 < 0.05 TABLE 1 Regression of FEVt - different intercepts for employment and smoking categories allowed but not shown. A Status Current Workers Number Cigarette Smoker Smoking Category Other Ex- Smokers Smokers NonSmokers - 0.0041 - 0.0037 - 0.0022 - 0.0050 418 < 0.1 58 n.s. n.s. 188 n.s. Leavers Number - 0.0266 25 0.1 0.0466 3 n.s. 0.0227 13 n.s. 0.0086 16 n.s. Pensioners - 0.0018 - 0.0159 - 0.0011 0.0182 Number 41 8 36 13 n.s n.s n.s. n.s. VAB.0001192244 TABLE 2 Regression coefficients for dust index against FEV, L. 1 62 r APPENDIX I Contd. A Variable Height Weight 4 Age r- Non-Smokers ! Ex-smokers 1 Other Smokers ^ Cigarette Smokers Lifetime Cigarette Consumption (1,000 packs) Dust Index (yr x mg/m9) Regression Coefficient P-value 0.06*5 - 0.0067 < 0.0001 * < 0.002 - 0.0236 < 0.0001 - 0.0307 < 0.0001 - 0.0414 < 0.0001 - 0.0319 < 0.0001 - 0.0136 < 0.0001 - 0.0035 < 0.05 TABLE 3 Regression of FVC - different intercepts for employment and smoking categories allowed but not shown. Status Current workers Number Cigarette Smoker Smoking Category Other Smokers ExSmokers NonSmokers - 0.0049 - 0.0029 - 0.0024 - 0.0084 4i8 P < 0.1 58 n.s 228 188 n.s n.s Number - 0.0197 25 n.s. 0.0359 3 0.0402 13 P<0 0.0080 16 Pensioners Number 0.0002 41 n.s. 0.0029 8 n.s. 0.0015 36 n.s 0.0148 13 n.s. TABLE 4 Regression coefficients for dust index against FVC. Expansion of model of regression for FVC illustrated in Table 3* i. .1 t K h t VAB.0001192245 b +1T * APPENDIX II AGREEMENT BETWEEN FIRST AND SECOND READINGS OVER PRESENCE OF SMALL ROUNDED OPACITIES CATEGORY O/l OH MORE FOR THREE MEDICALLY QUALIFIED READERS 63. - f '* i ** r 4 *I * 4 i 2nd reading Small rounded opacities Absent Present Total I Absent 1st reading Present 1* Total 778 9 787 22 800 9 18 31 8l8 2nd reading Reader 17 Absent 2nd reading 6*4 t t APPENDIX II Contd. AGREEMENT 6ETV/EEN FIRST AND SECOND READINGS OVER PRESENCE OF SMALL ROUNDED OPACITIES CATEGORY 0/1, AND CATEGORY 1/0 OR MORE FOR READER 17 2nd Reading VAB.0001192247 4 APPENDIX III CONFIDENTIAL INSTITUTE OF OCCUPATI 0 N A L MEDICINE 65 PERSONAL DETAILS Surname: Initials: Date of Birth: Reference No.: PLACE OF EXAMINATION: DATE OF EXAMINATION: INSTRUCTIONS FOR COMPLETING QUESTIONNAIRE: For "Y : N" answers: Mark answer which does apply. For "numeric" answers: Fill in appropriate numbers right justified, i.e. using the right-most boxes. VAB.0001192248 APPENDIX III Contd. QUESTIONNAIRE Reference No COUGH 1 Do you usually cough first thing in the morning in the winter? 2 Do you usually cough during in the winter? - or at night - If YES to 1 or 2 Do you cough like this three months each year? much PHLEGM Do you usually bring up any phlegm from first thing in the morning- in winter? Do you usually bring up any phlegm from during the day -- or at night - in the v: Y :N to 4 or 5 Do you bring up phlegm like this on most days for as much as three months each year? PERIODS OP COUGH AND PHLEGM In the past three years have you had a period of (increased) cough and phlegm lasting for three weeks or more? If YES more than one such period? CHEST ILLNESSES During the past three years have you had any chest illness which has kept you from your usual activities for as much as a week? If YES Did you bring up more phlegm than usual in any of these illnesses? If YES Have you had more than one illness like this in the past three years? VAB.0001192249 p * k * f * ^ p `i 4 '' * * l *f * % `* ** 4m APPENDIX III Contd BREATHLESSNESS If the subject is disabled from walking by any condition other than heart or lung disease, omit question 9 and enter 1 here 9a Are you troubled by shortness of breath when hurrying on level ground or walking up a slight hill? If YES 9b Do you get short of breath walking with other people of your own age on level ground? If YES 9c Do you have to stop for breath when walking at your own pace on level ground? ASTHMA 10a Have you ever had asthma? If YES 10b Have you had attack of asthma in the past 5 If YES 10c When you have an attack, noise in your chest? you aware of a whistling CHILDHOOD ILLNESSES When you were a child (before 15 years old) did you suffer from: 11a Catarrh? If YES 11b Bose? lie Chest? lid Discharging ears? lie Ilf 11h Hi Bronchitis more than other children? * Head colds more than other children? Chest colds more than other children? or Asthma? Whooping cough? Eczema of the skin when you were an infant? 67 YsN Y:N Y sN Y :N Y:N Y:N Y:N YN YN YN YN YN YN YN YN YN VAB.0001192250 APPENDIX III Contd PAST ILLNESSES Have you ever had (in adult or childhood life): 12a An injury or operation affecting your chest? 12b Heart trouble? 12c Bronchitis? 12d Bo colds usually go to your chest? 12e Pneumonia? 12f Pleurisy? 12g 12h 121 Tuberculosis (T.B.)? * Other chest trouble? * m Eczema of the skin in front of your elbows or behind your knees? 12j Hay fever (itching, sneezing or a running or stuffy nose, in the summer months)? 12k Apart from colds, itching of your nose, sneezing, or a running or stuffy nose, at other times of the year? 121 Sinusitis? If TBS 4 12m One attack? or 12n Recurrent attacks? FAMILY HISTORY Is your father alive? * Does/Did your father suffer from: 13a Frequent colds on the chest/or recurrent chestiness? 13b Chronic bronchitis? 13c Asthma? 13d Eczema of the skin? n 13 Hay fever? Y: N Yi N Y: N Y: N Y: N Y: H Y: N Y: N Y: N Y:N Y: N Y: N Y: N Y:N Y: N w Y:N Y: N Y:V Y:K Y: N VAB.0001192251: APPENDIX III Contd * Is your mother alive? Does/Did your mother suffer from; 13f Frequent colds on the chest/or recurrent chestiness? 13f Chronic bronchitis? 13h Asthma? 131 Eczema of the skin? 13d Hay fever? If you have brothers them suffered from: sisters. any one of Frequent colds on the chest/or recurrent chestiness? Chronic bronchitis? Asthma? 13n Eczema of the skin? Hay fever? If you have children, has any one of them suffered from: Frequent colds in the chest/or recurrent chestiness? Bronchitis? Asthma? 13s Eczema of the skin in front of the elbows or behind the knees? If YES As an infant? or Hay fever? life? Y: N Y:N Y: N Y:N Y: N Y: N Y: N Y;N YsN Y: N Y: N Y: N Y: N i Ys N Y:N ' Y: N Y: N 69 VAB.0001192252 APPENDIX III Contd. ** Y: N For present smokers: V 22a Have you been cutting down your smoking over the past year? If YES How many cigarettes did you usually smoke per day before you cut down? Y: N Month Y: H YN YN YN Y: N YN YH YN p J VAB.0001192253 75 4 APPENDIX IV Case/control studies : statistical approach Each case/control study provided data for two groups of man* Multiple measurements were available for each man. Most were measurements of lung function or presence/absence of symptoms; the studies were designed to compare men in the two groups in these respects. The remainder (age, height, weight, smoking status) were characteristics which, if ignored, might have obscured real comparisons between the groups The men studied were selected to broadly comparable in terms of these characteristics had n hoped that pair-matched studies would be possible. This however, would have led to unmatched cases and so a less stringent objective of group comparability was set This was achieved in three of the studies, but not in the fourth - non-resoonse was an added \ complication* In all studies there were some differences between cases and controls in how these "nuisance" characteristics were in related. The studies were not designed to test specific hypotheses about * particular lung function values or about particular clinical signs. Rather they were carried out to identify suggestive patterns of * differences, if any, between the groups over a wide range of clinical measurements. The data were considered from several viewpoints to draw out what patterns existed. In this way the emphasis was on exploratory analysis rather than on formal statistical testing. Such tests were carried out, but interpreted as a check on subjective assessments of differences rather than as the main focus of the analysis. The exploratory nature of the work also influenced the other main emphasis of the analysis, i.e. a.study of variables one-at-a-time based on clinical significance, and subsequent teasing out of patterns, rather than a formal multivariate exercise without full regard for the meaning of the measurements. In testing for statistical significance, the binary variables (i.e. responses to the medical questionnaire) were examined using (Pearson, not log--likelihood) chi--squared tests, without continuity correction. The continuous variables (lung function measurements) were tested using VAB.00D1192254 4 I 4 independent two sample t-tests assuming equality of variances. The Tables show that in general this assumption was realistic. Significance levels reported were not adjusted in any way to take account of the multiplicity of tests carried out on each group, or of the inter-relationship of the many response variables. Additionally, covariance analyses (using multiple linear regression programs) were carried out for the lung function measurements. These analyses were designed to eliminate bias, insofar as the matching had been imperfect, and also to reduce unexplained within-group variability and to obtain a better tests of group differences. The detailed results are shown in the accompanying Tables. Plots of residuals were examined to see if the summary statistics (mean, residual s.d.) used in tests of group comparisons represented the data adequately. The identification of outliers (i.e. men with unusually low values) was indeed of some intrinsic interest also, though again this was not considered the main focus of the work - it was considered a priori likely that some ill men P would be found among both cases and controls, and this in fact happened. Finally, the pattern of lung function measurements was * studied by standardising for FEVj level, as outlined in the report. VAB.0001192255 ft ` *4 & ION ANALYSIS FOR UJNC 1 UNOTION Vi APPENDIX IV : Table I 75 US tn 2 K m us US * US * s* {; 1 NO CM 3 % in a 5 US * CuMs I o o d d o o oft 1-s! 4* p & 4 8 g St & s a tt *H ft O ot i g fu"nisIToS O01 CoMi a0t ^o SS ddS II us oi o a 8 CM US doS Cl o "55 oi ot S o *s 3 )0 ?7 <VJ S3 ddg 00 o mo n* 1 M O O SB II os s iia cm 89 co VO 01 oi 2 4> dI dI g m dog ii *- VO d cl * VO o* o cn ?7* us o* oCM* was *3 e4 3 45 O Vft t * lila s s <M 4 >s 0 1 2 -aft > X IooT^UooS* fot- Oo /v V V b<| Pft Oft Pft flu PS II P 1 as oa s& *8 og M o oi oi 0a1s o o cn o o;g O2T S^O o o* aCO m\o ?8 oda *5 CM 4 d^i oS d ^ ft CO 0t 3* 1 8 ^5 ddt mOS d'^s oO oos aws O SO o o a03 82 o o CSO \00 5 O so u O O SB * O r-C*O US vp 00 O CM 6AZ s IS moot*CrMt-* *** us^ CD O CM 6At it 2Os S8O OI CMI * CM S O* CM** ?T* U-VCM S'S! OI CMI o sO gC4 oAt tt mo\* * tft Ol mI * ?8 di iAt mo s> 5 O r! g II s as os in b9 OS tmICfSS OO s CO CM vo Os 09 os CM o It o o CM NO 01 ft I 9 flu 0*0 0 X4 4 JO o o o 44 0 4 0 40 0 4 jo O oo o c o 8J33 a 008 4I* C-4fl 4^0 1 VAB.0001192256 yI 76. 2 & P O 2 >4 3 i 2o 1 Ch g p 32 Pd P 3 cn 03 SwR OIH CO * c2s |V3 2 S 03 5 hO APPENDIX IV : Tabl 2 i4 e VO NO NO ncn m m UN i stfS & tt On ON i o no in In* O in * on CM NO vo NO o oo O M CO g. 0 O 1S1 ^J^m^-Ca0.gM01 iOOn*N oCM*\13 0I 0I 2 vo no CM CM CM C 0 0*023 CM 00 to 0I 0I 2 Oin cinm di i S no m nv m * 2 002 II nnoo mmv*2 0I 0I 2 icnv CO i ?* * in 0 0* 2 II ^CM COON *2 002 % f * 83 it 1s I 4> p odS ti ooS ii s0^s0 w 0I 0I 2 CinM o~s lI 002 II On *o i CO ddg It m On O N- 2 II 8 CmM oi d $ 3 5S S p > e 5O 6 & $3 *O5 * *8 o* w 00 2 O i&n O O;s #n ov CoM OmN 002 8$ o o ac $ in j O2 CM CM 88 ddS s, $ $ s 4 C O o CM Gc j I US K 6&I c m C>xH iOn iOn*o- O OOOo Av v v frt 2 2 A* ;?; ityw n o di iZ O CM * II 5? O* m* * II #v-namo ** on* II CM & o OI CMI * CnrO-- ori+n% t *V O II o *** II s O CM 0I ^I 2 VO 8z m NO NO On o m I 00I m 4 <o o * o o * *0 O CiP o 4 niOnn CO GiOOnONn NO CM T 4 -O O 4 #0 o 4 O 4 o o oc o Pk Vt oo o c8 o a o 4o & & | *H PS 4 CO jD O VAB.0001192257 L 1 f i # * '-\4 -i t A ft i . -k-.l *" i * - 1 - .*1 f ** 1 r # * -- , Ti . . *; - -- ,< .- ' * a 1 * - - V*f * *I 4 h* -1 ;v *;-f* * * . 4* . * j H4 -i * 1 4 4 It Ii r 2 t < '4 i* d i -p rt >p i p * l * *I I -P i i A APWHDEC IV : Tabl* 3 77 P. S UCMN fit S8i4* o o C 4 ** ;ri| Ss llsii ?7a QUN\UCNM O CM f 5 > "S 4) ^ BO r* * CcO0M1 GUUoOINN sc e* sM2 4* ft OCNl UONN O ^fw* ?3 ss ddg UCMN UCMN CM CM CM UCMN CM 1 UN NUOONN NCOM 00 o o &UN o yo o s o fN m ve$ O^z ir\ O f0*1 no n Xs ooz tUN 1ft o iis 88 ooz NO 8 CO <?7* S UNflO f *- CO O CM* * 50 ON 2 8 nni * *^NO O52f* 3 ooS sO sNO O f 0) OO8c5nf*01 01 oI *01 UN UN OO mf* *on O 8 On dI oI * o \o 01 oI *OT IO? fO?Z0I 5$ US ^OOIC MCMI*f*g OCMN 01 CCOMM ffUN oI oI* 0*1 N&O NmO O CM * 8 o_ o c*n CM 5ft oI oI z rsf 8o5*- 2z CO o01 oI C*fl UN S.n di oi * co CM OI* OI 0*1 oO 8OoN*0*1 ^m in *: i c c3 8 1WMp 4Ci <-*hh*8 i| 6 +8*^&1 >?<**H 6 $SS8 o oIot*o a M /*v v v P< At * COUMNN I NCCOMM n N<Oy CM CO CMI NcOo UN * 00I 15 <3 :2 i m* o ** 0*0 * o P* 4*0 &oo O a 4a> s 'H P o& i m40 8888 8m o04 s. t* >oM fi < cP 8 0 s P #H *4 &P | *&H VAB.0001192258 1 % ?8. APPBfOIX IV : Tabl 4 2.:! 1 IS t aO eg 6 1,0 H <m >> 90 ^ H w 1 So -* >a? 4o4 o olfMT\ o O oooo /* V V V M CU 0* Pm t *3 *oo VOi c o V 1*4 H fetfl Sis 4.0 0 VAB.0001192259