Document pxE6qz8kk0GkxaBOagpYdnBw

Lung Function Consequences of sure ibestos Cejnent Manufactunng Plants (a Hans Weill, MD; Morton M. Ziskind, MD; Carmel Waggenspack, MPH; Charles E. Rossiter, MA A comprehensive study ol health effects associated with the mixed dust exposure In this industry has Included the collec tion of clinical, radiographic, lung func tion, and dust exposure data on 859 work ers In two plants. Evidence Is presented supporting a dose-response relationship between Indexes of dust exposure and lung function, similar to the previously re ported relationship with extent of x-ray film changes using the ILO U/C classifi cation. Lung volumes and maximum expiratory flow rates decrease In relation to increas ing cumulative dust exposure while pul monary diffusing capacity (DJ is not dust-dose related. Workers who had crocldollte exposure had smaller lung vol ume*, lower expiratory flow rates, and re duced Dl when compared with those having only chrysotile exposure. When the study population Is divided Into exposure groups, data thua far ana lyzed suggest that the chest x-ray film will reveal small opacities as early as signifi cant functional changes can be detected, but Individuals may have functional re duction prior to the appearance of x-ray film changes. It has been the aim of investiga tors studying the inorganic dust present dust levels. The resulting dust dose-biologic response relation diseases to establish which methods ship will provide the basis for estab of biologic monitoring are most sensi lishment of safe dust levels in indus tive in detecting the earliest changes try and make it possible for those of diffuse pulmonary involvement and responsible for maintenance of the measuring the progression of the dis workers' health to use the most sensi ease over time. This objective can be tive methods of monitoring that will realized through data collected from provide the earliest indicators of an a carefully designed epidemiologic adverse effect on the lungs. study of workers exposed to a well- In a study of workers engaged in characterized environment, which the manufacturing of asbestos ce usually requires analysis of past and ment building products, we have shown previously a dear relationship between cumulative total dust expo Submitted for publication June 28, 1974; ac cepted Sept 18. From the Pulmonary Disease Section, Depart ment of Medicine, Tulane University School of Medicine, New Orleans (Dr. Weill, Dr. Ziskind, and Ms. Waggenspack), and the MRC Pneumoco niosis Research Unit, Penarth, South Wales (Mr. Rossiter). Reprint requests to the Department of Medi cine, Pulmonary Disease Section, Tulane Univer sity School of Medicine, 1700 Perdido St, New Orleans, LA 70112 (Dr. Weill). sure and prevalence and profusion of radiographic opacities in the lung fields.1 Since these workers are ex posed to both asbestos fiber and free silica dust, attention was directed to the separation of small rounded from small irregular opacities. Pulmonary physiologic pattern's emerged that were consistent with the hypothesis 88 Arch Environ Health/Vol 30, Feb 1975 - Lung Function/Weiii al HER 0001305 that the rounded opacities were pri marily associated with a silica effect C. and that irregular opacities indicated * the effect of inhaled asbestos. The ' generally narrow range of asbestossilica distribution of total dust expo sure is perhaps the primary reason that analysis of constituent dust ex posures could not confirm the sugges tion that small irregular opacities in dicated asbestosis, while the rounded shadows resulted from silicosis. It is the purpose of this report to examine the various indexes of pul monary function and relate reduction in mean values, as compared with a standard group, to dust exposure. In order to do this, the effects of age, race, smoking, and presence of radiographic changes have been accounted for, and the influence of fiber type has been examined. Finally, a threshold dust level has been established for this industry, below which significant reduction in mean values cannot be demonstrated. The relative value of simple and more complex pulmonary function tests in detecting a dust ef fect is evaluated. , METHODS AND MATERIALS l. Subjects for Study All workers who were employed in two asbestos cement manufacturing plants in the New Orleans area on Nov 3,1969, were recruited for this study. A major reduction in work force in one of the plants shortly after initiation of this investigation neces sitated special effort in recruiting into the study the 184 individuals laid off, with the result that 70% of them participated. The total study population, in whom complete studies of pulmonary function are avail able, numbers 839, or 91% of those who would have been available on the day the study began, before the reduction in num ber of workers employed. Pulmonary func tion data collection were completed in No vember 1971. Because of the small number of female employees in these plants (37), the study group consists only of men. The average age of the participants in the study was 43 years, ranging from 21 to 79 years. They had an average of 17 years of dust exposure, with a range of one month to 43 years. Analysis of ethnic dis tribution revealed that 396 (46%) were white and 463 (34%) were black. Fifty-one percent of the entire group were current smokers, 26% were exsmokers, and 23% had never smoked. The proportion of smokers decreased from 61% in the lowest dust exposure group to 41% in the highest dust level group. Pulmonary Function Expiratory flows and volumes were de termined with a 13.5 liter water-sealed spirometer, with calculations of vital ca pacity (VC), forced vital capacity (FVC), forced expiratory volume, one second (FEVt), FEV./VC, and forced expiratory flow, 25% to 75% (FEFm-m). Functional residual capacity (FRC) was determined by the closed-circuit helium dilution method and led to calculation- of residual volume (RV) by subtraction of the expira tory reserve volume obtained during the equilibration procedure and of total lung capacity (TLC) by addition of VC to RV. We determined pulmonary diffusing ca pacity (transfer factor) (Dl) by the breath-holding carbon monoxide tech nique, using an automated apparatus for timing of events and valve sequencing, by which inspired, alveolar, and washout vol umes as well as breath-holding time are preset3 Diffusion constant (K.) was calcu lated as a ratio of to alveolar volume (VA), obtained by addition of inspired vol ume to the separately determined residual volume, ie, Dt/VA. For purposes of com parison VA was also determined by dilution of helium during the single breath. Two levels of exercise on a chair ergometer were chosen to approximate 1 and L5 liters of oxygen uptake (V) while measure ments of total ventilation and oxygen and carbon dioxide concentrations of the ex pired air allowed calculation of actual V. Ventilation was then corrected to the two standardized levels of Radiographic An EPA and two oblique chest x-ray films were obtained on each study partici pant Only the posteroanterior films were read using the ILO U/C International Classification for the Pneumoconioses by two members of the committee that de vised the classification. Analysis of the oblique projections will form the basis of a future report Films of 69 unexposed con trols who had never worked in this indus try were randomly interspersed among the films of the workers and were all read as 0/0 (66) or 0/1 (3), for small opacities. The heading "any x-ray change'' in this report is defined as category 1 or more for small opacities or grade 1 or more for pleural thickening. "No x-ray change" refers to those subjects who did not exhibit these changes on the x-ray film. A radiographic classification was assigned to an x-ray film if that classification was read by either of the expert readers. Agreement on the readings between the two readers has been analyzed and is considered very good. Using the stated definitions, 584 (68%) of the 859 total population had no x-ray change while 275 (32%) had any x-ray change; 130 (15%) had showed pleural changes only, 71 (8%) had small opacities only, and 74 (9%) had opacities and pleural changes. Dust Exposure Assessments Plant A produced only shingles, while plant B initially produced shingles and roofing, later adding pipe and, ultimately, flooring was added to the operations. In both plants, each work area was assigned a code number. In many instances several code numbers had several identical job ti tles. Detailed job histories that contained the work area code number, job title, and dates of starting and finishing each job were transcribed from personnel records. Because code numbers and job titles had changed frequently over the years, re peated consultation with personnel offi cers and older employees was required to describe jobs property. For plant A, dust exposure assessments were based on dust samples collected at various dates between 1952 and 1969. The primary instrument used was the midget impinger. These dust samples were col lected by federal and state agencies and industry, and the results were provided to us by the companies. For periods prior to 1952, estimates of exposure were made by interviewing employees of long service in an effort to compare recent dust levels with conditions of earlier dust exposure. Results and presampling estimates were recorded in million particles per cubic feet of air (mppcf). Plant B provided an exposure coding for each job identified, ranging from 5 mppcf to over 50 mppcf, also based on midget impinger sampling data. The primary type of asbestos used in plant A was chrysotfle. A small amount of amosite was added to all corrugated prod ucts in 1957 and minimal amounts of croddolite to corrugated bulkhead alone in 1962. Before 1957 no silica was added, but the cement at that time contained a high concentration of silica. Analyses of the dust samples suggested that the propor tion of silica in the airborne dust was only slightly less than that in the product Plant B had a wider range of jobs and exposure than plant A. Chrysotile was the primary type of fiber used; a small amount of crocidolite was added to one product; no amosite was used. Silica was added to three of the four products, and mica and talc were used in two of the four. Both Arch Environ Health/Vol 30, Feb 1975 Lung Function/Weill et al 89 / HER 0001306 plants provided information regarding the percent of each type of fiber and of silica used in the products manufactured. Comparison of Dust Exposure Measures In the Two Plants The aim was to grade equal dust expo sures in the two plants equally. However, because dust sampling was carried out by various industrial hygienists, agencies, and apparatus, comparability is not pre cise. These considerations also imply that the dust levels in this study cannot neces sarily be taken to be the same as in any other study. An exception to this is in a mortality study of retired asbestos work ers where the same industrial hygiene team and methods were used as in plant B.* Calculation of Dust Exposure for Each Person There were 146 and 241 different job ti tles identified at plants A and B, respec tively. There were ten mutually exclusive dust exposure categories at plant A and 57 such categories at plant B. Using the de tailed occupational history that had been abstracted for each person, total dust ex posure for an individual was calculated by multiplying the dust exposure level of each job by the time spent in the job. The re sults were totaled for all jobs and ex pressed in mppcf-yr. Ethnic Differences and Standardization of Lung Function Preliminary analyses showed that there were considerable differences in pulmo nary function between the black and white subjects included in this study. A detailed analysis of these differences suggested that they were related to size differences in chest volume for given height in blacks and whites.* Calculations based on anthro pometric data on African and Europeans from Rhodesia' gave a 13.2% difference in chest volume, which agrees almost com pletely with differences in the major lung volume indexes between blacks and whites in a subgroup (244) of this population who had had little exposure to asbestos dust and who had no major respiratory symp toms or radiographic changes. This sub group will be termed the standard group in this report As an example, in this stan dard group after increasing the total lung capacity of the blacks by 13.2% there was no difference in the relations to age and height in the two races. Standardizing to age 40 years and height 175 cm (5 feet 9 inches) the whites and blacks had average total lung capacities of 6.32 and 6.31 liters, respectively. Remarkable agreement with these ethnic differences in lung function has also been reported recently.* The factor, 13.2% has been used to scale total lung capacity, VC, FVC, FEV, FEF, and the alveolar volumes. A smaller in crease of 8% was proposed for functional residual capacity and residual volume, again based on anthropometric data. For diffusing capacity a factor of 8% was also proposed, with a corresponding factor for the diffusion constant of --4.6%. Application of these scaling factors to the lung function indexes for the blacks eliminated any racial difference. A com mon regression equation proved to be ap plicable for each index, and details of these common regression equations calculated from the standard low-exposure group are given in Table 5 of a previous publication.* Whenever standardized values are pre sented in this report, they are based on these regression relations after scaling the indexes for the blacks by the factors given above. For example, a man aged 40 years with a height of 175 cm (5 feet 9 inches) would have a standard total lung capacity of 6.31 liters. If his actual total lung capac ity were 6.0 liters, then his standardized value would be 95.1%. This standard group will always have average standardized values of 100%, and the expected value for any other group will also be 100%. For tests of significance, the observed standardized value has always been compared to the expected value of 100%. Effect of Smoking ' In order to determine whether smoking affected the relationship of pulmonary function and dust exposure, this relation ship has been separately analyzed for smokers, exsmokers, and nonsmokers in the total study population. When indexes of pulmonary function in each of the smok ing groups were compared with those in the standard group, in which smoking habits were mixed, it initially appeared that synergism between smoking and ex posure existed, since function in relation to the standard group was lower in each ex posure category in the smokers than in the nonsmokers; the difference between the two groups generally increased with expo sure and age. As described in the previous section, the standardization or prediction equations tised in this study were based on a standard population, consisting of 142 current smokers, 45 exsmokers, and 57 nonsmokers of cigarettes. A detailed anal ysis of FEFm-75%, in the standard group, now also divided into subgroups according to smoking habits, showed that not only did the smokers have reduced function, but also that the FEFm-m, declined with age much faster in the smokers than in nonsmokers, with exsmokers intermediate (Fig 1). A similar pattern was found for other indexes of ventilatory function. When these differing regressions for each smoking group are used to "predict" or "standardize" lung function (ie, correcting for smoking habits), the exposure-function relations do not differ significantly be tween smoking groups. It was also found that the proportion of nonsmokers is al most constant in the five exposure groups (23%), while the proportion of current smokers actually declines from 60% to 40% as exposure increases. For these reasons it was concluded that the effect of exposure to asbestos.is not influenced appreciably by smoking habits and that there is no synergism between smoking and asbestos dust as regards an effect on pulmonary function in this population. It was, there fore, thought reasonable that in the follow ing analyses, smoking need not be consid ered and the overall regression relations published previously* are used to provide the standardized values of pulmonary function against which the effects of expo sure are assessed. This matter is discussed further in a separate publication.' RESULTS Characteristics of Subjects in Regard to Dust Exposure The study population of 859 men is classified by dust exposure group, race, smoking history, and x-ray re sults changes in Table 1, and by total dust exposure group, age, and con stituent exposure pattern in Table 2. The whites tended to have had a higher total of exposures than the blacks, 51% of the former and 43% of the latter being in the two highest ex posure groups. There were also a higher proportion of smokers in the lowest exposure group, and an in crease from 18% to 36% in the propor tion of exsmokers, as exposure in creased. The proportion with small opacities of category 1 or more as recorded by either of two readers rose markedly with exposure from 4% in the lowest group to 30% in the high est. There was also a rise in the pro portion with pleural changes from 11% to 30% with increasing exposure. In Table 2, where average age is given for each dust exposure group, it is apparent that while the lowest ex- 90 Arch Environ Health/Vol 30, Feb 1975 Lung Function/Weill et al HER 0001307 Age, yr Fig 1.--Relation between FEVmvtsx and age for the smokers, exsmokers, and nonsmokers in the standard group (height taken as 175 cm [5 feet 9 inches]). Table 1.--No. of Subjects by Exposure Group, Race, Current Cigarette Smoking Habits, and Radiographic Change Category All men Whites Blacks Smokers Exsmokers Nonsmokers No x-ray changes Small ooacities only Pleural changes only Both small opacities and pleural change ' <S0 233 9B 135 139 43 51 202 6 22 3 Total Dust Exposure, mppcf-yr* 50-100 92 43 49 45 28 19 73 3 14 100-200 130 52 78 68 30 32 85 8 28 200-400 245 126 119 119 67 59 138 29 40 2 9 38 * Million particles per cubic feet-years. 4004- ' 159 77 82 65 57 37 86 25 26 22 Totals 859 396 463 436 225 198 584 71 130 74 Table 2.--Exposure Characteristics of Subjects in Each Total Dust Exposure Group Total Dust Exposure, mppcf-yr* Index Number Average age Years in industry Total dust, mppcf-yr Chrysotile index Crocidolite index Amosite index Silica index ' <50 238 35.3 5.9 19.2 3.1 0.1 0.1 3.8 50-100 89 46.4 17.8 71.4 11.2 0.3 0.3 13.6 100-200 129 48.2 20.7 146.0 22.4 0.8 0.3 322 200-400 244 48.8 21.9 303.0 51.3 4.2 0.3 88.0 * Million particles per cubic feet-years. 400+ ' 159 52.2 24.9 565.4 104.4 2.8 0.8 139.7 Totals 859 45.3 17.4 225.3 39.3 1.9 0.3 58.2 posure group is much younger, aver age age does not differ much in the remaining four groups. The average exposure indexes, including simply years in the industry, rise steadily. The exception is the crocidolite index that is very low in the first three ex posure groups, rises strikingly in the fourth and drops slightly in the fifth group. This pattern may be explained by the date of opening of the pipe manufacturing section in plant B. Lung Function Changes With Exposure In a previous section, the need to scale the values for the pulmonary function indexes for the blacks to ac count for ethnic differences was dis cussed. The effect of this scaling is to eliminate the problem that would oth erwise be caused by the differing pro portions of white and black subjects in the five exposure groups. Also, smoking habits have been discussed but for the reasons indicated previ ously, no correction for smoking has been made in the following analyses. Table 3 shows how each pulmonary function index varied with exposure. For example, total lung capacity de clined from 6.25 liters on average in the lowest exposure group to 5.85 li ters in the highest The changes in average values in some of the other indexes are more marked, for exam ple, for the indexes of flow. Also shown, in parentheses for each index, is the average standardized value, taking into account age and height As the standardizing relations were determined from people within this study population, the expected value for each group is 100% and so each standardized value has been com pared with 100%. For total lung capacity, the stan dardized value drops from 98.2% to 94.0% with increasing exposure. While it is reduced significantly in the lowest exposure group, the larg est and most significant changes oc cur in the two highest exposure groups. Figure 2 shows that the re duction in the lowest exposure group occurs entirely in those with x-ray changes. Those with x-ray changes have much lower total lung capacity, but even in those with no x-ray Arch Environ Health/Vol 30. Feb 1975 Lung Function/Weill et al 91 HER 0001308 Table 3.--Pulmonary Function Indexes Dust Exposure, mppcf-yrt Index Number Total lung capacity, liters Vital capacity, liters FVC. liters FEV,, liters FEV,/VC 1%) FEFtf,.;,v liters/sec Functional residual capacity, liters Residual volume, liters RV/TLC(%) Diffusing capacity, ml/mln/mm Hg Diffusion constant, Di/V* Alveolar volume (multibreath), liters Alveolar volume (single-breath), liters Number Ventilation at oxygen uptake of 1.0 liters/mln Ventilation at oxygen uptake of 1.5 llters/min <50 233 6.39 (98.2) 45 4.72 (98.4) S 4.63 (98.2)5 3.80 (98.0)5 80.4 (99.5) 4.21 (97.7) 2.96 (98.6) 1.62 (98.5) 25.3 (100.0) 34.1 (97.5) 6.04 (99.1) 5.66 (98.0)5 5.22 (99.3) 216 25.1 (100.2) 37.8 (100.4) 50-100 92 6.39 (99.8) 4.41 (98.1) 4.32 (98.3) 3.48 (99.1) 78.8 (101.3) 3.81 (98.2) 3.13 (98.6) 1.93 (101.2) 30.0 (100.9) 32.9 (103.6) 5.78 (193.9) 5.77 (101.5) 5.07 (98.2) 70 25.7 (102.3) 39.2 (104.2) 100-200 130 6.18 (98.0) 4.42 (97.4)5 4.16 (97.8) 3.28 (97.4)5 77.4 (100.2) 3.50 (94.5) 3.04 (96.4) 1.88 (96.8) 30.0 (99.0) 31.1 (100.8) 5.68 (103.4) 5.52 (98.5) 4.94 (97.8) 97 25.6 (102.6) 37.8 (100.6) 200-400 245 5.99 (95.2)11 4.04 (93.1)11 354 (93.1)11 3.11 . (93.0)11 77.0 (99.9) 3.32 (89.9)11 2.99 (94.5)11 1.89 (96.6) 31.5 (102.3) 29.9 (97.7) 5.66 (103.5)1 5.30 (94.9)11 4.66 (92.7)11 170 26.0 (103.6)1 39.2 (104.3)11 * Average values and values standardized for age, height, and race for the five total dust-exposure groups. * Million particles per cubic feet-years. t Number In parentheses below average value refers to average % standardized. Significance of difference between standardized value and 100%: $ P <.05; II P <.001; 1P <.01. 400+159 5.83 (94.0)11 3.84 (91.8)11 3.73 (91.6)11 2.94 (92.1)11 76.2 (100.1) 3.02 (85.6) il 2.89 (91.4) 1.93 (95.7) 32.8 (102.5) 30.0 (101.8) 5.79 (108.8)11 5.23 (94.9)11 4.61 (93.3)11 103 26.5 (105.6)11 39.4 (104.6)1 Totals 859 6.14 (96.7)11 4.26 (95.5)11 4.16 (95.5)11 353 (95.5)11 78.0 (100.0) 3.59 (92.8)11 2.99 (95.8)11 1.83 (97.5) S 29.7 (101.0) 31.6 (99 5) 5.80 (103.3)11 5.47 (97.0)11 4.89 (96.0)11 656 25.7 (102.4)11 38.6 (102.5)11 changes there are significant reduc tions in the two highest exposure groups. Vital capacity (Fig 2) and FEV, (Fig 3) show the same pattern of marked reductions in function in those with x-ray changes, but also significant changes in the highest ex posure groups for the total group and in those with x-ray changes. Forced vital capacity is not shown in the Fig ures because it is almost indistin guishable from VC. The FEV,/VC ra tio shows no effect of exposure confirming that FEV, and VC show the same relation to exposure. The changes in the FEFut-ts* (Fig 3) are larger than for any of the other indexes, but it is also the most vari able being about 2.4 times as variable as the FEV,. Functional residual capacity (Fig 2) is also a highly variable index. The pattern of change is similar to that for FVC and VC, showing reduced values with increased exposure, but in the no x-ray change group it is only in the highest exposure group that the average standardized value differs significantly from 100%. Residual vol ume shows a tendency to fall with in creasing exposure, but not signifi cantly so. However, its ratio to total lung capacity shows very little evi dence of an exposure effect, indicat ing that the two indexes move in the same direction. ' Pulmonary diffusing capacity (Fig 4) shows only an irregular pattern with exposure, although it does show a paradoxical tendency to rise with increasing exposure in those with out roentgenographic changes and to fall in those with roentgenographic changes. Since alveolar volume de creases with increasing exposure, there must be compensatory changes in the diffusion constant. Figure 5 shows that the diffusion constant rises with increasing exposure and that it is apparently unaffected by the disorder that causes the roentgen ographic changes. The two standard indexes of venti lation during exercise show similar results. The ventilation rate at an oxygen uptake of 1 liter/min (Fig 4) rises with exposure in the total group and in those with x-ray changes. The pattern in those with no roentgen ographic change is less regular, so that virtually all the rise with expo sure may be attributed to the increas- 92 Arch Environ Health/Vol 30. Feb 1975 Lung Function/Weill et al HER 0001309 Total Lung Capacity % Standardized Vital Capacity % Standardized Change Change Change Fig 2.--Relation between lung volumes and dust exposure. Change Table 4.--Characteristics of Workers Who Had More Than 75% of Their Employment or Had Never Been In the Pipe Area Number Total years of exposure Total exposure, mppcf-yr* Crocidolite exposure Chrysotile exposure Silica exposure Age Proportion of smokers. % ' Million particles per cubic feet-years. 75% or More In Pipe Area 108 22.8 324 6 48 120 49 51 Never Worked In Pipe Area 100 23.0 290 0.6 52 78 52 51 ing proportion of people with roentgenographic changes. Other Analyses As well as the averages for each ex posure group, the proportions in the tails of the distributions of standard ized values were considered. These proportions followed the patterns ex pected from the pattern of the aver ages, and suggest that the effect of exposure was to reduce function over all rather than in a few susceptible individuals. Figure 5 shows that for FEV, the distribution of standardized values in the highest exposure group is a little skewed compared to that for the lowest exposure group, but that the major difference between the dis tributions is one of position. This con firms for this index the general im pression of a general reduction in function as the main response to dust exposure. The comparison of average stan dardized values in the five exposure groups makes no assumptions about the linearity of the changes in func tion with respect to dust exposure. Linear regression of the standardized values of lung function on total expo sure confirms the findings based on the main analysis. All regression co efficients of function on exposure are highly significant (P<001) except those for residual volume (P <.01) and for diffusing capacity, FEV,/VC ratio and RV/TLC ratio that are not sig nificant. Effect of Crocidolite In one of the two plants studied, crocidolite has been used regularly in the pipe making area only, although a few other people have been slightly exposed in maintenance work. In an attempt to separate the effects of cro cidolite from those of chrysotile, two groups of workers were defined who had worked for between 20 and 30 years in the industry. One group had worked for at least 75% of their time in the pipe area and the other group had never worked in that area. Table 4 lists some of the character istics of these groups. The crocidolite group had exposure to considerably more crocidolite, slightly more silica, and somewhat less chrysotile. Mean age was slightly younger and there was an identical proportion of smokers. Table 5 shows the average pulmo nary function values, expressed as the proportion of the standardized values. The crocidolite group had smaller major lung volumes, lower FEV, and reduced diffusing capacity. On additional analysis, it was found that in this group of workers these differences in lung function were not Arch Environ Health/Vol 30. Feb 1975 Luna Function/Weill et al 93 HER 0001310 accounted for by the differing mean values for silica exposure, and, in fact, when the study population was divided into high and low silica expo sure groups, average lung function values were higher in those with high silica exposure. COMMENT The role of pulmonary function in the detection and quantitation of ad verse effects related to asbestos dust exposure has received considerable attention during the last 15 years. While initial reports were concerned primarily with the characterization of the physiologic pattern in well ad vanced asbestos13 associated with ra diographic abnormalities, more re cent investigation has been directed toward the use of pulmonary function to detect the earliest adverse biolog ic response before the appearance of radiographic changes. Attempts have also been made to resolve the unset tled question of the relative sensi tivity of differing measurements of pulmonary function in detecting this early dust effect The task has been complicated by recent evidence indi cating that the biologic effects asso ciated with asbestos fiber inhalation probably depends, to considerable ex tent, on the type of occupational ex posure, eg, mining and milling, man ufacturing, and application, as well as fiber type. Nonfibrous dust exposures in asbestos-exposed workers, partic ularly dust of the fibrogenic variety, proride further complicating features that require analysis. While the ulti mate objective to use lung function as an indicator of a departure from respiratory health is to develop doseresponse relationships and threshold limits of exposure below which the production of disease is prevented, the difficult task of reconstructing past dust exposures has limited markedly the available information pertaining to this important objec tive. Most studies of pulmonary function in asbestosis directed attention to the well-established disease, in which pre dictably the major abnormalities in cluded reduced pulmonary volumes with "stiff" lungs and disturbance of alveolar gas transfer.*-11 In general, it FEV,/VC % Standardized All Subjects <r 105 100 951- All Subjects FEF 25%-75% No X-Ray Change Any X-Ray Change ---- 1771 No X-Ray Change Any X-Ray Change %Standard!zed Fig 3.--Relation between expiratory flow and dust exposure. Table 5.--Lung Function (Standardized Values)* t 75% or More In Pipe Area Never Worked In Pipe Area Vital capacity 92.7 97.9 Functional residual capacity 95.2 95.1 Residual volume 97.2 94.5 Total lung capacity 94.6 97.6 RV/TLC ratio Forced vital capacity 103.1 92.1 98.4 98.5 FEV, 91.2 98.2 FEFjj*.,,% 84.5 88X1 FEV/VC ratio 98.4 100.2 Diffusing capacity 95.5 103.2 Diffusion constant 102.7 105.5 Alveolar volume (single breath) 91.1 98.4 Alveolar volume (multibreath) 93.7 97.8 Ventilation at oxygen uptake of 1.5 llters/mln 105.2 106.0 P <01 ... <1 ... <01 <01 ... <02 <01 <1 * Values In workers who had spent more than 75% of their employment or had never been in the pipe area. Men, 20 to 30 years In the plant. 94 Arch Environ Health/Vol 30. Feb 1975 f'l ' Lung Function/Weill et al HER 0001311 Diffusing Capacity % Standardized Diffusion Constant % Standardized Ventilation at Oxygen Uptake of 1 Liter/min % Standardized Fig 4.--Relation between indexes of gas exchange and dust exposure. was thought that a significant effect on the airways does not result from asbestos dust exposure. In several re cent studies, however, data have been presented that imply that the early adverse effect of asbestos dust expo sure may indeed be on airways func tion with particular reference being made to the "small airways." This lat ter effect can be expected to precede radiographic evidence of parenchy mal disease.14-11 All of the studies reviewed above are importantly limited by inade quate data in regard to individual asbestos exposure of the workers un der investigation. In most cases, when the degree of exposure was esti mated, it could be expressed only in terms of total years on the job. It has, therefore, not been possible to estab lish dose-response relationships in any meaningful way. Results .from a study of over 1,000 Quebec chrysotile mining workers who had dust expo sure estimated from their work his tory and periodic dust levels deter mined by sampling, indicated that simple spirometric measurements in cluding inspiratory capacity, FEV, and FVC are most sensitive for the detection of early dust effects, while the diffusing capacity at rest is a poor indicator of level of exposure.1* The dose-response relationship was better in nonsmokers than in smokers, and the effect of dust exposure and smok ing on lung function did not appear to be additive. We have reported previously the ra diographic patterns associated with the mixed fibrogenic dust exposure in the same workers as in this investiga tion who have been exposed to both asbestos and silica.1 The prevalence of both small rounded and small irregu lar lung opacities has been correlated with increasing levels of total dust exposure, findings similar to those re ported by the Quebec group. These re lationships have been established in spite of the inherent limitations and uncertainties of grading past dust ex posures. Confidence in these dust as sessments has been strengthened by the results of this investigation of the effect of exposure on pulmonary func tion. Progressive reduction in a num ber of lung function measurements is clearly associated with increasing level of total dust exposure in this population. Of particular interest is the demonstration of the dose-re sponse relationship in those workers without radiographic evidence of a dust disease, an indication that indi viduals may have reduction in pul monary function prior to the time when radiographic abnormalities ap pear. The relationship between dust exposure and pulmonary function is maintained after the effect of age, height, ethnic group, smoking history, and presence of radiographic changes have been considered. Previous inves tigators have not taken these impor tant variables fully into account. In this study, simple spirometric measurements of pulmonary volumes and maximum expiratory flow rates have provided the benefit of full bio logic monitoring in assessment of an adverse functional effect on respira tory health. Vital capacity, FEV,, and FEFz-ro, are indicators of an early detrimental effect as sensitive as or more sensitive than measurements such as total lung capacity, pulmo nary diffusing capacity, and exercise tests. They suggest that an early ef fect on the airways may, indeed, de velop as a consequence of asbestos ex posure. These conclusions are based on cross-sectional and prevalence data that have been analyzed up to this point and future examination of longitudinal data now being gener ated could result in modifications of the conclusions concerning the rela tive usefulness of the various lung function tests. Arch Environ Health/Vol 30. Feb 1975 Lung Function/Weill et al 95 HER 0001312 Biologic monitoring of workers ex posed to a hazardous dust must be directed to the detection of early unfavorable changes in the hope that removal from exposure will prevent the progression of disease. Since chest radiography and pulmonary function are most widely used to monitor the respiratory health of as bestos-exposed workers, it is appro priate to review the relative sensi tivity of these two methods in detecting an adverse biologic re sponse to dust exposure in this indus try. At the top of Fig 6, selected simple spirometric measurements in average percent of the standardized value are plotted by exposure group. It can be seen that after a cumulative exposure of 100 mppcf-yr there is a gradual reduction in the mean values as dust exposure increases. At the lower half of the figure, the propor tion of workers with irregular or rounded small opacities at the one or two category levels, is plotted for each exposure group. It is apparent that the major increase in workers at each level of radiographic change also begins to appear after a cumulative exposure of 100 mppcf-yr, which again reveals a dose-response rela tionship. One must conclude from these data that in considering groups of workers with varying levels of dust exposure, radiographic evidence of small opacities within the lung fields and pulmonary functional change are equally sensitive in detecting diffuse parenchymal abnormalities associ ated with exposure to these fibrogenic dusts. However, it must be re membered that even in the absence of x-ray film changes, individuals may show decline in pulmonary function unexplained by other factors and pre sumably related to dust exposure. The reverse is also true, ie, radiograph ic abnormalities may appear before changes in pulmonary function. It is suggested that both methods of bio logic monitoring be used in order to detect the earliest abnormalities in the largest number of exposed work ers. In contrast to studies of mortality from malignant diseases in asbestosexposed workers, there seems to be little interaction between cigarette Standardized Value, % Fig S.--FEV,: Distribution of standardized values. smoking and the diffuse fibrogenic ef fect of dust exposure in this indus try. As indicated previously, this is demonstrated by the equally good ex posure-pulmonary function relation ship in nonsmokers after allowance for personal smoking habits. This is not to say that cigarette smoking has not reduced pulmonary function in workers who smoke; that it has is shown by a lower position of pulmo nary function measurements, while the additional rate of decline result ing from dust exposure is similar for the smoking groups. The diffuse parenchymal and pos sible airways damage associated with asbestos dust exposure seem to be rel atively unaffected by host factors, as evidenced by the distribution of pul monary function in the lowest and highest exposure groups. In the FEV, for example, we demonstrated pri marily a shift in the group mean ob served values (downward in the highexposure group), rather than a signif icant "skewing" or "tailing" effect Available evidence suggests that crocidolite asbestos has a greater car cinogenic effect than chrysotile.1 This suspicion has led previously to a dif ferent standard for these two fiber types in the United Kingdom. It has not been demonstrated previously that crocidolite has a more marked ef fect on the diffuse pulmonary paren chymal process. Although crocidolite exposures in the pipe-making area of one of the two plants studied were considered reasonably low, there ap pears to have been a significantly greater adverse effect on pulmonary function in workers exposed to cro cidolite than on a matched group in 96 Arch Environ Health/Vol 30. Feb 1975 Lung Functlon/WeiH et al HER 0001313 ployed in these two plants will deter mine if there is an excess risk of the development of neoplasms in workers employed in this industry. Fig 6.--Selected spirometric indexes of lung function and prevalence of small opac ities on the chest x-ray films in relation to dust exposure. the same plant for whom no crocidolite exposure was documented. The crocidolite group has smaller lung volumes, lower FEV,, and reduced DL. This subject will be separately consid ered in the future report. Finally, threshold limit values must obviously take into consideration all demonstrated adverse health effects. This report considers only the diffuse fibrogenic effect of asbestos exposure m the asbestos cement products manufacturing industry. However, the data from this study demonstrate not only a dose-response relationship but suggest that a threshold dust level exists in this industry below which diffuse pulmonary fibrosis does not occur. While past dust levels were-expressed in total particulate units, con version to fiber levels is desirable for evaluation of asbestos exposure. A conservative calculation using data obtained previously from simultane ous sampling with the impinger and fiber counting method, leads to a con version ratio of 2 fibers/ml for 1 mppcf." The cumulative total dust exposure level of 100 mppcf-yr would therefore be equivalent to 200 fiber/yr or an exposure to 5 fibers/ml for a working lifetime of 40 years. While it is hoped that this level of ex posure is "safe" in regard to the de velopment of diffuse lung changes, any such analysis must be cautiously interpreted, since follow-up of exposed workers continuing well beyond the time of their active employment may increase the incidence of radiographic and functional abnormalities. These important additional data should re suit from the longitudinal investiga tion that involves a cohort of this population.' A mortality study of all workers who have ever been em Supported by grant from the Institute of Oc cupational and Environmental Health of the Quebec Asbestos Mining Association and a Spe cialized Center of Research (SCOR) grant. Divi sion of Lung Diseases, NHLI (HL-15092). References L Weill H, Waggenspack C, Bailey W, et al: Radiographic and physiologic patterns among workers engaged in manufacture of asbestos ce ment products. J Occup Med 15:243-252,1973. 2. Meade F, Saunders MJ, Hyett F, et al: Au tomatic measurement of lung function. Lancet 2373-575,1965. 3. Enterline PE, Henderson V: Type of as bestos and respiratory cancer in the asbestos in dustry. Arch Environ Health 27312-317, 1973. 4. Rossiter CE, Weill H: Ethnic differences in lung function: Evidence for proportional differ ences. Int J Epidemiol 355-61, 1974. 5. Paul R, Fletcher GH, Addison G: A com parative study between Europeans and Africans in the mining industry of northern Rhodesia. Med Proe 6:69-74,1960. 6. Lapp NL, Amandus HE, Hall R, et al: Lung volumes and flow rates in Mack and white sub jects. Thorax 29:185-188,1974. 7. Rossiter CE, Weill H: Synergism between dust exposure and smoking: An artefact in the statistical analysis of lung function? Bull Physicepathcl Retpir, to be published. 8. Williams R, Hugh-Jones P: The signifi cance of lung function changes in asbeatosis. Thorax 15:109-119,1960. 9. Leathart GL: Clinical, bronchographic, ra diological and physiological observations in ten cases of asbestosis. Br J Ind Med 17213-227, 1960. 10. Bader ME, Bader RA, Tierstein AS, et al: Pulmonary function and radiographic changes in 598 workers with varying duration of exposure to asbestos. J ML Sinai Hoep 37:492-500, 1970. 1L Regan GM, Tagg B, Walford J, et al: The relative importance of clinical, radiological and pulmonary function variables in evaluating as bestosis and chronic obstructive airway disease in asbestos workers. Clin Sri 41569-582,1971. 12. Murphy RLH, Gaensler EA, Redding RA. et al: Low exposure to asbestos: Gas exchange in ship pipe coverers and controls. Arch Environ Health 25253-264,1972. 13. Gaensler EA, Carrington CB, Coutu RE, et al: Pathological, physiological, and radiological correlations in the pneumoconioses. Ann NY Acad Sri 200574-607,1972. 14. Muldoon BC, Tumer-Warwick M: Lung function studies in asbestos workers. Br J Die Chat 66:121-132,1972. 15. Jodoin G, Gibbs GW, Macklem PT. et al: Early effects of asbestos exposure on lung func tion. Am Rev Retpir Die 104525-535, 1971. 16. Becklake MR, Foumier-Massey G. Rossiter CE, et al: Lung function in rhrysotile as bestos mine and mill workers of Quebec. Arch Environ Health 24:401-409,1972. 17. Lpich JR, Ayer HE, Johnson DL The in terrelationships of selected asbestos exposure in dices. Am Ind Hyg Assoc J 3159'WO*. 1970. Arch Environ Health/Vol 30. Feb 1975 Lung Function/WeiII et al .97 HER 0001314