Lung function of sheet metal workers exposed to fiber glass

Lung function of sheet metal workers exposed to fiber glass. Scand j work environ health 9 (1983) 9-14. Out of 532 registered and contacted sheet metal workers, 251 responded, but only seven pairs were acceptable for the present study due to the following requirement: no history of smoking, pleural plaques or asbestos exposure. Seven of these workers were exposed daily to fiber glass, and seven were almost never exposed. In a second step nine additional exposed workers were included. Ordinary spirometry, lung volumes, closing volumes and the slope of the alveolar plateau, the maximum expiratory flow in air and after helium-oxygen breathing, and the elastic recoil pressures were measured. No evidence of small airway dysfunction or restrictive or obstructive ventilatory impairment was found, but the elastic recoil pressures of the exposed group were slightly increased. Fiber glass can conceivably cause a corresponding faint and probably harmless fibrous reaction in the lung parenchyma.

Fiber glass has been commercially available for more than 40 years. The increased awareness of the harmful effects of asbestosie, fibrosis of t h e lung parenchyma and, in recent years, also its cancerogenicityhas focused interest on t h e possible injurious effects of fiber glass. Occasional cases have been reported in which this material has been pointed out as t h e possible harmful agent (2,11,14). The causal association has been uncertain however.
Several epidemiologic studies based on the radiographic examination of the lungs of workers heavily exposed to fiber glass have failed to show an increased preva-lence of pathological findings (15,24). Studies of lung function have also failed to reveal any chronic effects of fiber glass inhalation (3, 10, 22).
The development of new lung function tests, more sensitive than e.xlier methods in revealing subtle changes in the small peripheral airways, made further studies on workers exposed to fiber glass of interest, particularly since, according t o some of these tests, apparently healthy asbestosexposed subjects (9) and welders (18) have been shown to b e affected. The object of the present study was to determine whether these new tests could reveal changes in subjects exposed to fiber glass. posed. The study was confined to life-long nonsmokers. By means of a questionnairedistributed through the trade unionwe obtained 65 nonsmoking subjects (table l), 15 of whom had never been exposed to fiber glass and could therefore be used as referents for comparison with 15 exposed subjects. The exposed subjects and referents were pair-matched for age (+ 4 a) and height (+ 4 cm).
Six of the 15 subjects in the exposed group had been exposed to asbestos for more than three months. Two of them had pleural plaques, and one had pleural plaques and parenchymal changes consistent with asbestosis. In the reference group, pleural plaques were found in 5 out of 15 subjects, but none had been exposed to asbestos. Consequently, because of exposure to asbestos for more than three months and/or pleural plaques, we excluded seven pairs of subjects from further analysis. Yet another pair was excluded because the referent (not exposed to asbestos) had a history of repeated pleuritic illness. The final analysis thus comprised seven exposed subjects and seven matched referents. The mean values (range) of the age, height, and accumulated time of fiber glass exposure of the groups are given in table 2.

Methods
A rolling-seal servospirometer with electric output for volume and flow was used to obtain an ordinary spirogram [vital ca-VC inhalations of a helium-oxygen mixture (80120) (4). Closing volume, the slope of the alveolar plateau, and the amplitude of the cardiogenic oscillations were assessed by the nitrogen technique (16). Total lung capacity (TLC) was measured with a body plethysmograph (5). The static elastic recoil pressure of the lung at different lung volumes (%TLC) was assessed by the esophageal balloon technique (12). The volume and pressure signals were calibrated at each measurement. The group mean values of the lung function variables have been expressed as the percentage of the predicted level.
The exposure to fiber glass has been given as accumulated exposure (eg, 3 monthslyear for 10 years = 30 months). It was not possible to estimate the fiber content in the air.
The paired t-test has been applied to test the difference between the group means.

Results
In the exposed group, two subjects claimed cough in the morning and sputum production during the day in the wintertime. In connection with exposure to fiber glass, three subjects noticed cough and phlegm, two cough only, and one subject cough and dyspnea. In the reference group two subjects claimed to wheeze when having a cold.
The group mean values for TLC, VC, and FEV, are given in table 3. No individual showed values below 80 % of the predicted level. The group differences were small and nonsignificant. Table 4 summarizes the results of the nitrogen test and the flow-volume curves. The group means for the closing volume and the slope of the alveolar plateau were not significantly different. The mean amplitudes of the cardiogenic oscillation were similar for the groups. Furthermore, no single subjectwhether exposed or not   (1) and total lung capacity according to Grimby & Soderholm (6). NS = p > 0.05 (paired t-test). Table 4. Results of the single-breath nitrogen test, maximal expiratory flows, and the change in the exposed subjects (all nonsmokers) and in the referents (all nonsmokers) following helium-oxygen breathing   exhibited an abnormally high closing volume or slope of the alveolar plateau. The group mean value of the MEF~o was slightly lower for the referents, but the difference was not significant (p > 0.1, table 4). The response to helium-oxygen breathing did not separate the exposed group from the referents. One subject in the reference group had a AMEF,, of less than 20 %, ie, there was evidence of density independence (4). The elastic recoil pressure was consistently higher in the exposed group in comparison to the reference group. The difference was significant at lung volumes between 70 and 90 % of the TLC (table 5).

Discussion
The inclusion of smokers may be expected to reduce the possibility of revealing the discrete effects of occupational exposure on lung function (18). We therefore decided to use only nonsmoking subjects, who constituted approximately 25 % of the total male population available for study. In order to obtain a sufficiently large number of subjects for study, all sheet metal workers in the Goteborg area who were members of the trade union were contacted (N = 522). As is obvious from table 1, the response rate was very low (47 %), in spite of our cooperation with the trade union and two additional contacts. We can only speculate about the factors guiding the individual in his decision of whether or not to participate and about the extent to which this large primary dropout influences the general applicability of the findings. One of the consequences of the low response rate and our decision to study only nonsmokers was that the number of presumptive referents was limited to 15. The group available for analysis was further reduced as a number of subjects had to be excluded because of exposure to asbestoswhich has a potential effect on lung function that almost certainly outweighs the potential effect of fiber glass (3). These considerations highlight two problems inherent in every study on occupational hazards, ie, the need for a very large basic population and the difficulty in ascertaining pure exposure.
It seems that symptoms involving the respiratory tractunrelated to occupa-tional exposurewere equally common in both our groups. On the other hand, irritation of the upper airways during heavy exposure has been reported (131, and in our exposed group six out of seven subjects claimed cough when exposed. This result may indicate that, periodically, there is a high number of airborne fibers with an aerodynamic diameter associated with deposition in the larger airways. The net time of exposure may seem rather short (table 2), but it corresponds to occupational handling of fiber glass for 6 to 10 years.
Our study gives no spirometric evidence of the development of restrictive ventilatory impairment, a result which is in line with that of previous reports (3, 10). In earlier studies the nitrogen test has revealed tobacco-induced changes in smokers without symptoms (18) and changes presumably located in the small peripheral airways or the lung parenchyma in nonsmoking welders (17). In the present study, however, the nitrogen test did not reveal significant differences between subjects exposed to fiber glass and referents, and we were thus unable to demonstrate any effect of fiber glass on the small peripheral airways. The flow-volume curves gave results to the same effect. If anything, it seems that the reference group deviated more from the predicted values than the exposed group did. Furthermore, the similarity in the AMEF5, and AMEFZ5 of our groups indicates that the distribution of the central and peripheral airway resistance was equal, which also appears to contradict the possibility of adverse effects of fiber glass.
While neither the nitrogen test nor the flow-volume curves separated the subjects exposed to fiber glass from the referents, we did find a significant difference in the elastic properties of the lung, with slightly but consistently higher recoil pressure in the exposed group. No single subject exhibited an abnormally high recoil pressure however. The interpretation of this small difference is uncertain. Fibrosis of the parenchyma would tend to increase elastic recoilas is the case in asbestosis and silicosiswhereas obstructive disease may instead be associated with lowered elastic recoil. A fundamental question is then whether the difference between the pressure-volume relationships should be taken to indicate increased elastic recoil in the exposed group or a corresponding decrease in the referents.
In comparison to our laboratory standard, it certainly appears as if the exposed group deviates towards high elastic recoil (table 5). The same impression prevails when the exposed group is compared with generally accepted reference material (211, although such a comparison is inherently uncertain due to possible differences in the populations and methodology.
When, in an attempt to verify our findings, we extended our study by another nine exposed subjects (nonsmokers, mean age 44.2 years, mean net exposure 3.2 years) who were not investigated initially because there were no appropriate referents, the exposed group did not deviate significantly from the reference group in respect to elastic recoil pressure at any lung volume. Including all exposed subjects, we did, however, still find a shift to the right of the pressure-volume curve (fig 11, which is significant at a lung volume of 60, 70 and 80 % of the TLC. Is this slightly elevated recoil pressure indicative of a fibrotic process in the lung parenchyma of subjects exposed to fibrous glass? Alveolar deposition is favored when the fiber diameter is less than 3 p and the length is shorter than 200 E(. (8,20). In addition, the fibrinogenic effect is reported to be increased when the fiber length is above 10 p (19). Indoor measurements of airborne fibers at places of work indicate that the majority of glass fibers are less than 3 p in diameter and between 10 and 200 p in length. Therefore the possibility cannot be excluded that fiber glass can elicit a fibrous reaction in the lung parenchyma. Asbestos fibers, which are known to cause fibrosis of the lung, fulfill the size requirements for alveolar deposition, but the concentration of respirable asbestos fibers is probably much higher (7) and therefore more hazardous.
On the other hand, it must be realized that selection mechanisms are already in play in the choice of occupation, and it may be totally unwarranted to infer that small deviations of subgroups (eg, occupational categories) from a reference mean are caused by pathological processes.
Even if the increase in elastic recoil found in the exposed group is in fact caused by fiber glass exposure, the prog- nosis nevertheless appears favorable since epidemiologic studies have failed to reveal an increased mortality or incidence of lung disease among subjects exposed to fiber glass (22,24). In summary, in a small number of subjects exposed to fiber glass, we found slightly increased elastic recoil but no evidence of small airway dysfunction, restrictive ventilatory impairment, or airway obstruction. It is not possible from this study to generalize as to the hazard of handling fiber glass, but it may serve to indicate a need for further studies on the effect of fiber glass inhalation on lung function.