WORK A Journal of Prevention, Assessment & Rehabilitation
ELSEVIER
Work 11 (1998) 325···329
The effect of exposure to S02 on the respiratory system of power-station workers P. Frooma,*, G. Sackstein b , C. Cohen a , Y. Lerman a , Estella Kristal-Boneh a , J. Ribak a • Occupational Health and Rehabilitation Institute, Raanana, Israel "The Israel Electric Corporation, Tel Aviv, Israel
Abstract Sulfur dioxide (S02) is generally recognized as a respiratory irritant, but its effects if any at low levels of exposure are uncertain. We studied 38 power station technicians exposed to 0.8 ppm (parts per million) 8-h weighted levels of sulfur dioxide, and compared them to workers performing similar tasks without such exposure. Those exposed complained 5.8 times more frequently of cough (95% CI = 1.8-20.6, P < 0.001), and also had significantly more sputum production. There was also a trend for increasing prevalence of dyspnea. On the other hand there was no decrease in pulmonary function test values. In the eight exposed subjects who complained of dyspnea, there was a significant decrease in pulmonary flow values. We conclude that power station workers exposed to low levels of S02 have increased respiratory symptoms, and deserve compensation if their symptoms become chronic. The pulmonary function tests were not different from the control subjects, but there may be a small group who are prone to long-term morbidity. Additional studies are warranted to confirm our findings, and to define immediate and long-term morbidity due to low exposure to S02' © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. Keywords: Sulfur dioxide; Respiratory symptoms; Pulmonary function; Industry
1. Introduction Sulfur dioxide (S02) is produced in a number of industrial processes, such as refrigeration, pulp production for paper manufacture, copper smelt-
* Corresponding author, Department of Epidemiology, institute of Worker's Health and Rehabilitation, Raanana, PO Box 3, Israel. Fax: + 972 9 7712212; e-mail: paulf@ trendline.co.il
ing, and silicone-carbide production, as well as in corn refineries and electric power stations. It is generally recognized as a respiratory irritant, but the effect on the prevalence of symptoms and on pulmonary function is controversial (Kehoe et aI., 1932; Skalpe, 1964; Ferris et aI., 1967; Smith et aI., 1977; Archer and Gillam, 1978; Greaves et aI., 1984; Osterman et aI., 1989). Furthermore, the exposures were sometimes mixed (Archer and Gillam, 1978; Osterman et aI., 1989) and smoking
1051-9815/98/$ - see front matter © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. P/I S 1 051-9815(98)00049-7
326
P. Froom et al. / Work 11 (1998) 325-329
was not always taken into account (Kehoe et al., 1932), making it difficult to assess the isolated effect of S02 (Archer and Gillam, 1978; Osterman et al., 1989). Kehoe et al. (1932) reported an increased prevalence of symptoms indicative of upper respiratory tract irritation, such as cough, nasal irritation, expectoration, and prolongation of colds, in 100 refrigeration workers exposed to S02 as a refrigerant, compared to 100 unexposed control subjects. Furthermore, loss of taste and smell, increased shortness of breath and fatigability were significantly increased in the exposed group. The exposure, however, was relatively high in this study (7-70 ppm), and smoking was not taken into consideration. On the other hand, Ferris et al. (1967) studied 147 pulp-mill workers exposed to S02 levels of 0-33 ppm. They did not find a significant increase in chronic bronchitis as defined by respiratory symptoms. These workers, however, had a wide variation of exposure which was probably negligible for a significant number of them. Furthermore, there was voluntary transfer of workers from pulp-mill to paper-mill, and a demonstrable healthy worker effect in both mills, which may have reduced the apparent effect of exposure. Skalpe (1964), in a study of workers in pulp-mills exposed to similar levels of S02' did find a significantly higher incidence of cough, sputum, and dyspnea in younger exposed workers. Osterman et al. (1989) reported on 145 silicone-carbide workers who were exposed to low levels of S02 (0.2-1.5 ppm) over a long follow-up period. They found that the exposure to S02 was associated with chronic sputum, mild exertional dyspnea, and also wheezing in a dose-dependent relationship. The cough was found to be associated with smoking but not with S02 exposure, but the workers were not currently exposed to S02 and were also exposed to a significant amount of dust. Archer and Gillam (1978) reported on 953 smelter workers with intermediate exposure level (0.4-3 ppm) and also found an increase in morning cough, sputum, and shortness of breath compared to mine-shop worker control subjects. The authors, however, stated that they could not rule
out enhancement of S02 toxicity by dust or trace metals. On the other hand, factors in their study design and population characteristics may have actually reduced the difference between the exposed and control subjects. Smith et al. (1977), in a similar study of 113 smelter workers with exposure to 0.2-2.4 ppm of S02 found an increase in chronic cough and sputum in workers at higher exposure (> 1 ppm) who also smoked, compared to those at lower levels of exposure. In corn-refinery workers, Greaves et al. (1984) found a dose-dependent increase in cough, sputum, and wheezing with exposures under 5 ppm. In the following study we compared the prevalence of respiratory symptoms, and the values of pulmonary function tests in two groups of workers performing similar tasks but with only one exposed to S02' The exposure resulted from leaks from a positive-pressure burning chamber in the power unit, whereas the control group workers in proximity to negative-pressure burning chambers had no exposure to S02' 2. Methods Thirty-eight male operators at the power station who worked in proximity to the burning chambers for at least 2 years were studied. They were exposed to levels of S02 which ranged from o to 15 ppm, with an estimated TWA of 0.7 ppm over the 8-h shift. They were compared to 34 workers who worked in proximity to a negativepressure burning chamber and were not exposed to S02' Their work tasks were identical to those of the exposed operators. All workers were examined. The workers completed a self-administered questionnaire based on that of the American Thoracic Society (ATS). Height and weight were recorded as was pulse, and blood pressure measured with patients in the sitting position. Pulmonary function testing was performed with the Autospiro-400 instrument as recommended by the ATS. Calibration was done at monthly intervals. Age and height were used to determine predicted values. The following parameters were recorded; PVC (functional vital capacity), FEVI
327
P. Froom et al. / Work 11 (1998) 325-329
(forced expiratory volume - 1 s), FEV1/FVC, FEF (forced expiratory flow maximum) 25-75%, FEF75, FEF50 and FEF25. Finally all episodes of absenteeism from work over the last 2 years were obtained from the personnel department.
2.1. Statistical analysis The difference between groups for continuous variables was tested for statistical significance by the Student's t-test for normally distributed variables, and by the Kruskal-Wallis one-way AOV for data not distributed normally. For non-continuous variables the x2-test with Yates correction was used, and where small numbers were present, Fisher's exact test was substituted. A P-vahie of < 0.05 was considered significant. 3. Results All workers in the study group were examined and none refused to fill out the questionnaire or undergo pulmonary function testing. Over the preceding 2 years, 25 workers had retired, two workers changed jobs, and four left for medical reasons not related to respiratory illness. There were no significant differences in age, years of employment, smoking history, or physiological parameters (Table 1). Smoking history was nearly identical between the exposed and the control subjects (smokers and ex-smokers; 32%,
Table 1 Exposed power station workers compared to control subjects Variable
Exposed (n = 38)
Controls (n = 34)
P-value
Age (years) Employment (years) BM1b (kg/m2) SBpb (mmHg) DBPb (mmHg) Pulse (no/min)
37 ± 9" 11 ± 7 24.7 ± 3.4 119 ± 16 77 ± 9 67 ± 9
35 ± 7 10 ± 6 25.3 ± 2.8 120 ± 13 78 ± 10 67 ± 9
0.21 0.47 0.46 0.77 0.58 0.95
± S.D. body mass index; SBP, systolic blood pressure; DBP, dia~tolic blood pressure.
"Mean b BM !,
13% and compared to 26%, and 15%, respectively). Cough and sputum production were significantly more common in the exposed group, and there was also a trend of an increasing prevalence of dyspnea (Table 2). There appeared to be a synergistic effect of exposure and smoking on the prevalence of cough, non-exposed/non-smokers 16% (4/25), non-exposed/smokers 22% (2/9), exposed/non-smokers 50% (13/26), and exposed/smokers 67% (8/12) (P = 0.0077). There were no differences in pulmonary function tests in the exposed compared to the control subjects (Table 3). In the exposed workers, however, with complaints of dyspnea there were significant decreases in flow values (Table 4). This was not found for those who complained of a cough.
Table 2 The prevalence of respiratory symptoms Symptom
Chest pain Dyspnea on effort Dyspnea at work Dyspnea (any) Cough (any) Cough (at work) Sputum production
Exposed (n = 38) n (%) 2 (5)
5 (13) 5 (13) 8 21 18 8
(21) (55) (47) (21)
cn
Controls (n = 34) n (%)
ORa (95%
1 (3) 2 (6) 1 (3) 2 (6) 6 (18) 3 (9) 0(0)
1.8 (0.1-111.4) 2.4 (0.4-26.7) 5.0 (0.5-243.5) 4.3 (0.8-43.6)
5.8 (1.8-20.6) 9.3 (2.2-53.9)
--------------------------------------------"Odds ratio with 95% confidence intervals. b Fisher's exact test.
P-value
0.54b 0.26b
O.l3 b 0.06 b
< O.OOl < 0.01 0.01
P. Froom et al. / Work 11 (J998) 325-329
328
find an increased prevalence of cough. These studies, however, as mentioned above had varying exposures and serious methodological flaws. We found no decrease in pulmonary function tests. In only a few studies has pulmonary function testing been performed in workers exposed to SOz. Ferris et al. (1967) did not find any decrease in pulmonary function tests in pulp-mill workers compared to paper-mill workers. Those studying higher exposures have found reductions in both lung volume (Archer and Gillam, 1978) and flow values (Smith et al., 1977; Archer and Gillam, 1978). However, Greaves et al. (1984) reported results similar to ours in corn-refinery workers exposed to 0-5 ppm SOz. It is likely that the decrease in pulmonary function is dose-dependent and occurs at levels higher than those needed to cause respiratory symptoms. This assumption is supported by the findings of Smith et al. (1977) who reported that reduction in flow was found only in workers exposed to levels above 1 ppm. As SOz is highly water-soluble, the nasopharynx and upper respiratory mucosa are the most vulnerable to its effects. At the relatively low concentrations in the industrial setting, 99% of the gas is deposited in the nasopharynx and hypopharynx. It would therefore be expected that exposure to SOz would cause increased respiratory symptoms due to irritation of the upper airway, while pulmonary flow patterns, reflecting lower airway function, remain essentially intact. There may, however, be a subset of workers whose small airways are affected also by very low levels of SOz. For example a deficiency of sulfite oxi-
Table 3 Pulmonary function tests (percent of predicted) Tests
Exposed = 38)
Controls = 34)
P-value
(n
90.2 ± 92.4 ± 90.3 ± 88.3 ± 89.3 ± 97.9 ±
92.1 ± 91.4 ± 87.6 ± 91.7 ± 86.7 ± 91.3 ±
0.57 0.53 0.60 0.55 0.79 0.45
(n
FVC FEV1 FEF25-75 FEF75 FEF50 FEF25
11.6 13.2 23.2 21.7 23.7 35.1
12.2 10.0 22.0 16.5 20.6 35.7
Absence from work due to illness was no different between the two groups, 19/36 (52.8%) in the exposed compared to 23/34 (67.6%) in the control subjects . 4. Discussion The main finding of our study of a small but very homogenous group of power station workers was an effect of exposure to low levels of SOz on the prevalence of cough, sputum production and a trend for increasing complaints of dyspnea. Furthermore, there appeared to be a synergistic effect of smoking and exposure on the prevalence of cough. This increase in symptoms has been reported consistently previously (Kehoe et al., 1932; Skalpe, 1964; Ferris et al., 1967; Smith et al., 1977; Archer and Gillam, 1978; Greaves et al., 1984; Osterman et al., 1989), but usually at much higher exposure levels than were present in our study. Ferris et al. (1967) did not find an increase in symptoms, and Osterman et al. (1989) did not
Table 4 The exposed power station workers with either cough or dyspnea Test
Dyspnea = 8)
(n
..
FVC FEV1 FEF25-75 FEF75 FEF50 FEF25
85 ± 11 82 ± 12 75 ± 19 87 ± 28 77 ± 21 73 ± 22
None = 30)
P
92 95 94 89 93 105
0.15 0.02 0.03 0.72 0.09 0.02
(n
Cough = 21)
None = 17)
P
(n
(n
90 ± 12 94 ± 13 95 ± 22 88 ± 26 90± 23 108 ± 37
90 ± 11 90 ± 14 85 ± 24 89 ± 26 88 ± 26 85 ± 28
0.87 0.30 0.18 0.83 0.69 0.06
"-.--.-.----~-
± 12 ± 12 ± 22 ± 20 ± 24 ± 35
P. Froom et al. / Work 11 (1998) 325-329
dase in some individuals could make them more susceptible to cellular protein damage by sulfite (Archer and Gillam, 1978). In our study group, those with dyspnea had significantly reduced lung flow values, and we cannot rule out a small negative effect of SOz on lung flow values because of the lack of power of our study due to the small number of subjects. In conclusion our study supports the hypothesis that upper respiratory symptoms are caused by exposure to low levels of S02. We did not find evidence for a decrease in pulmonary function tests, and the long-term effects on workers who have retired is uncertain. Further studies are warranted to more clearly define immediate and long-term morbidity due to low exposures to S02. References Archer YE, Gillam JD. Chronic sulfur dioxide exposure in a smelter. J Occup Med 1978;20:88-95.
329
Ferris BG, Burgess WA, Worcester J. Prevalence of chronic respiratory disease in a pulp mill in the United States. Br J lnd Med 1967;24:26-37. Greaves lA, Ferris BG, Essex D. Respiratory effects of sulfur dioxide (S02) among corn refinery workers. Am Rev Respir Dis 1984;129:AI57. Kehoe RA, Machle WF, Kitzmiller K, leBlanc TJ. On the effects of prolonged exposure to sulfur dioxide. J lnd Hyg 1932;14:159-173. Osterman JW, Greaves lA, Smith TJ, Hammond SK, Robins JM, Theriault G. Respiratory symptoms associated with low level sulfur dioxide exposure in silicon carbide production workers. Br J Ind Med 1989;46:629-635. Skalpe 10. Long-term effects of sulfur dioxide exposure in pulp mills. Br J Ind Med 1964;21 :69-73. Smith TJ, Peters JM, Reading JC, Castle CH. Pulmonary impairment from chronic exposure to sulfur dioxide in a smelter. Am Rev Respir Dis 1977;116:31-39.
ELSEVIER
Work 11 (1998) 325···329
The effect of exposure to S02 on the respiratory system of power-station workers P. Frooma,*, G. Sackstein b , C. Cohen a , Y. Lerman a , Estella Kristal-Boneh a , J. Ribak a • Occupational Health and Rehabilitation Institute, Raanana, Israel "The Israel Electric Corporation, Tel Aviv, Israel
Abstract Sulfur dioxide (S02) is generally recognized as a respiratory irritant, but its effects if any at low levels of exposure are uncertain. We studied 38 power station technicians exposed to 0.8 ppm (parts per million) 8-h weighted levels of sulfur dioxide, and compared them to workers performing similar tasks without such exposure. Those exposed complained 5.8 times more frequently of cough (95% CI = 1.8-20.6, P < 0.001), and also had significantly more sputum production. There was also a trend for increasing prevalence of dyspnea. On the other hand there was no decrease in pulmonary function test values. In the eight exposed subjects who complained of dyspnea, there was a significant decrease in pulmonary flow values. We conclude that power station workers exposed to low levels of S02 have increased respiratory symptoms, and deserve compensation if their symptoms become chronic. The pulmonary function tests were not different from the control subjects, but there may be a small group who are prone to long-term morbidity. Additional studies are warranted to confirm our findings, and to define immediate and long-term morbidity due to low exposure to S02' © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. Keywords: Sulfur dioxide; Respiratory symptoms; Pulmonary function; Industry
1. Introduction Sulfur dioxide (S02) is produced in a number of industrial processes, such as refrigeration, pulp production for paper manufacture, copper smelt-
* Corresponding author, Department of Epidemiology, institute of Worker's Health and Rehabilitation, Raanana, PO Box 3, Israel. Fax: + 972 9 7712212; e-mail: paulf@ trendline.co.il
ing, and silicone-carbide production, as well as in corn refineries and electric power stations. It is generally recognized as a respiratory irritant, but the effect on the prevalence of symptoms and on pulmonary function is controversial (Kehoe et aI., 1932; Skalpe, 1964; Ferris et aI., 1967; Smith et aI., 1977; Archer and Gillam, 1978; Greaves et aI., 1984; Osterman et aI., 1989). Furthermore, the exposures were sometimes mixed (Archer and Gillam, 1978; Osterman et aI., 1989) and smoking
1051-9815/98/$ - see front matter © 1998 Published by Elsevier Science Ireland Ltd. All rights reserved. P/I S 1 051-9815(98)00049-7
326
P. Froom et al. / Work 11 (1998) 325-329
was not always taken into account (Kehoe et al., 1932), making it difficult to assess the isolated effect of S02 (Archer and Gillam, 1978; Osterman et al., 1989). Kehoe et al. (1932) reported an increased prevalence of symptoms indicative of upper respiratory tract irritation, such as cough, nasal irritation, expectoration, and prolongation of colds, in 100 refrigeration workers exposed to S02 as a refrigerant, compared to 100 unexposed control subjects. Furthermore, loss of taste and smell, increased shortness of breath and fatigability were significantly increased in the exposed group. The exposure, however, was relatively high in this study (7-70 ppm), and smoking was not taken into consideration. On the other hand, Ferris et al. (1967) studied 147 pulp-mill workers exposed to S02 levels of 0-33 ppm. They did not find a significant increase in chronic bronchitis as defined by respiratory symptoms. These workers, however, had a wide variation of exposure which was probably negligible for a significant number of them. Furthermore, there was voluntary transfer of workers from pulp-mill to paper-mill, and a demonstrable healthy worker effect in both mills, which may have reduced the apparent effect of exposure. Skalpe (1964), in a study of workers in pulp-mills exposed to similar levels of S02' did find a significantly higher incidence of cough, sputum, and dyspnea in younger exposed workers. Osterman et al. (1989) reported on 145 silicone-carbide workers who were exposed to low levels of S02 (0.2-1.5 ppm) over a long follow-up period. They found that the exposure to S02 was associated with chronic sputum, mild exertional dyspnea, and also wheezing in a dose-dependent relationship. The cough was found to be associated with smoking but not with S02 exposure, but the workers were not currently exposed to S02 and were also exposed to a significant amount of dust. Archer and Gillam (1978) reported on 953 smelter workers with intermediate exposure level (0.4-3 ppm) and also found an increase in morning cough, sputum, and shortness of breath compared to mine-shop worker control subjects. The authors, however, stated that they could not rule
out enhancement of S02 toxicity by dust or trace metals. On the other hand, factors in their study design and population characteristics may have actually reduced the difference between the exposed and control subjects. Smith et al. (1977), in a similar study of 113 smelter workers with exposure to 0.2-2.4 ppm of S02 found an increase in chronic cough and sputum in workers at higher exposure (> 1 ppm) who also smoked, compared to those at lower levels of exposure. In corn-refinery workers, Greaves et al. (1984) found a dose-dependent increase in cough, sputum, and wheezing with exposures under 5 ppm. In the following study we compared the prevalence of respiratory symptoms, and the values of pulmonary function tests in two groups of workers performing similar tasks but with only one exposed to S02' The exposure resulted from leaks from a positive-pressure burning chamber in the power unit, whereas the control group workers in proximity to negative-pressure burning chambers had no exposure to S02' 2. Methods Thirty-eight male operators at the power station who worked in proximity to the burning chambers for at least 2 years were studied. They were exposed to levels of S02 which ranged from o to 15 ppm, with an estimated TWA of 0.7 ppm over the 8-h shift. They were compared to 34 workers who worked in proximity to a negativepressure burning chamber and were not exposed to S02' Their work tasks were identical to those of the exposed operators. All workers were examined. The workers completed a self-administered questionnaire based on that of the American Thoracic Society (ATS). Height and weight were recorded as was pulse, and blood pressure measured with patients in the sitting position. Pulmonary function testing was performed with the Autospiro-400 instrument as recommended by the ATS. Calibration was done at monthly intervals. Age and height were used to determine predicted values. The following parameters were recorded; PVC (functional vital capacity), FEVI
327
P. Froom et al. / Work 11 (1998) 325-329
(forced expiratory volume - 1 s), FEV1/FVC, FEF (forced expiratory flow maximum) 25-75%, FEF75, FEF50 and FEF25. Finally all episodes of absenteeism from work over the last 2 years were obtained from the personnel department.
2.1. Statistical analysis The difference between groups for continuous variables was tested for statistical significance by the Student's t-test for normally distributed variables, and by the Kruskal-Wallis one-way AOV for data not distributed normally. For non-continuous variables the x2-test with Yates correction was used, and where small numbers were present, Fisher's exact test was substituted. A P-vahie of < 0.05 was considered significant. 3. Results All workers in the study group were examined and none refused to fill out the questionnaire or undergo pulmonary function testing. Over the preceding 2 years, 25 workers had retired, two workers changed jobs, and four left for medical reasons not related to respiratory illness. There were no significant differences in age, years of employment, smoking history, or physiological parameters (Table 1). Smoking history was nearly identical between the exposed and the control subjects (smokers and ex-smokers; 32%,
Table 1 Exposed power station workers compared to control subjects Variable
Exposed (n = 38)
Controls (n = 34)
P-value
Age (years) Employment (years) BM1b (kg/m2) SBpb (mmHg) DBPb (mmHg) Pulse (no/min)
37 ± 9" 11 ± 7 24.7 ± 3.4 119 ± 16 77 ± 9 67 ± 9
35 ± 7 10 ± 6 25.3 ± 2.8 120 ± 13 78 ± 10 67 ± 9
0.21 0.47 0.46 0.77 0.58 0.95
± S.D. body mass index; SBP, systolic blood pressure; DBP, dia~tolic blood pressure.
"Mean b BM !,
13% and compared to 26%, and 15%, respectively). Cough and sputum production were significantly more common in the exposed group, and there was also a trend of an increasing prevalence of dyspnea (Table 2). There appeared to be a synergistic effect of exposure and smoking on the prevalence of cough, non-exposed/non-smokers 16% (4/25), non-exposed/smokers 22% (2/9), exposed/non-smokers 50% (13/26), and exposed/smokers 67% (8/12) (P = 0.0077). There were no differences in pulmonary function tests in the exposed compared to the control subjects (Table 3). In the exposed workers, however, with complaints of dyspnea there were significant decreases in flow values (Table 4). This was not found for those who complained of a cough.
Table 2 The prevalence of respiratory symptoms Symptom
Chest pain Dyspnea on effort Dyspnea at work Dyspnea (any) Cough (any) Cough (at work) Sputum production
Exposed (n = 38) n (%) 2 (5)
5 (13) 5 (13) 8 21 18 8
(21) (55) (47) (21)
cn
Controls (n = 34) n (%)
ORa (95%
1 (3) 2 (6) 1 (3) 2 (6) 6 (18) 3 (9) 0(0)
1.8 (0.1-111.4) 2.4 (0.4-26.7) 5.0 (0.5-243.5) 4.3 (0.8-43.6)
5.8 (1.8-20.6) 9.3 (2.2-53.9)
--------------------------------------------"Odds ratio with 95% confidence intervals. b Fisher's exact test.
P-value
0.54b 0.26b
O.l3 b 0.06 b
< O.OOl < 0.01 0.01
P. Froom et al. / Work 11 (J998) 325-329
328
find an increased prevalence of cough. These studies, however, as mentioned above had varying exposures and serious methodological flaws. We found no decrease in pulmonary function tests. In only a few studies has pulmonary function testing been performed in workers exposed to SOz. Ferris et al. (1967) did not find any decrease in pulmonary function tests in pulp-mill workers compared to paper-mill workers. Those studying higher exposures have found reductions in both lung volume (Archer and Gillam, 1978) and flow values (Smith et al., 1977; Archer and Gillam, 1978). However, Greaves et al. (1984) reported results similar to ours in corn-refinery workers exposed to 0-5 ppm SOz. It is likely that the decrease in pulmonary function is dose-dependent and occurs at levels higher than those needed to cause respiratory symptoms. This assumption is supported by the findings of Smith et al. (1977) who reported that reduction in flow was found only in workers exposed to levels above 1 ppm. As SOz is highly water-soluble, the nasopharynx and upper respiratory mucosa are the most vulnerable to its effects. At the relatively low concentrations in the industrial setting, 99% of the gas is deposited in the nasopharynx and hypopharynx. It would therefore be expected that exposure to SOz would cause increased respiratory symptoms due to irritation of the upper airway, while pulmonary flow patterns, reflecting lower airway function, remain essentially intact. There may, however, be a subset of workers whose small airways are affected also by very low levels of SOz. For example a deficiency of sulfite oxi-
Table 3 Pulmonary function tests (percent of predicted) Tests
Exposed = 38)
Controls = 34)
P-value
(n
90.2 ± 92.4 ± 90.3 ± 88.3 ± 89.3 ± 97.9 ±
92.1 ± 91.4 ± 87.6 ± 91.7 ± 86.7 ± 91.3 ±
0.57 0.53 0.60 0.55 0.79 0.45
(n
FVC FEV1 FEF25-75 FEF75 FEF50 FEF25
11.6 13.2 23.2 21.7 23.7 35.1
12.2 10.0 22.0 16.5 20.6 35.7
Absence from work due to illness was no different between the two groups, 19/36 (52.8%) in the exposed compared to 23/34 (67.6%) in the control subjects . 4. Discussion The main finding of our study of a small but very homogenous group of power station workers was an effect of exposure to low levels of SOz on the prevalence of cough, sputum production and a trend for increasing complaints of dyspnea. Furthermore, there appeared to be a synergistic effect of smoking and exposure on the prevalence of cough. This increase in symptoms has been reported consistently previously (Kehoe et al., 1932; Skalpe, 1964; Ferris et al., 1967; Smith et al., 1977; Archer and Gillam, 1978; Greaves et al., 1984; Osterman et al., 1989), but usually at much higher exposure levels than were present in our study. Ferris et al. (1967) did not find an increase in symptoms, and Osterman et al. (1989) did not
Table 4 The exposed power station workers with either cough or dyspnea Test
Dyspnea = 8)
(n
..
FVC FEV1 FEF25-75 FEF75 FEF50 FEF25
85 ± 11 82 ± 12 75 ± 19 87 ± 28 77 ± 21 73 ± 22
None = 30)
P
92 95 94 89 93 105
0.15 0.02 0.03 0.72 0.09 0.02
(n
Cough = 21)
None = 17)
P
(n
(n
90 ± 12 94 ± 13 95 ± 22 88 ± 26 90± 23 108 ± 37
90 ± 11 90 ± 14 85 ± 24 89 ± 26 88 ± 26 85 ± 28
0.87 0.30 0.18 0.83 0.69 0.06
"-.--.-.----~-
± 12 ± 12 ± 22 ± 20 ± 24 ± 35
P. Froom et al. / Work 11 (1998) 325-329
dase in some individuals could make them more susceptible to cellular protein damage by sulfite (Archer and Gillam, 1978). In our study group, those with dyspnea had significantly reduced lung flow values, and we cannot rule out a small negative effect of SOz on lung flow values because of the lack of power of our study due to the small number of subjects. In conclusion our study supports the hypothesis that upper respiratory symptoms are caused by exposure to low levels of S02. We did not find evidence for a decrease in pulmonary function tests, and the long-term effects on workers who have retired is uncertain. Further studies are warranted to more clearly define immediate and long-term morbidity due to low exposures to S02. References Archer YE, Gillam JD. Chronic sulfur dioxide exposure in a smelter. J Occup Med 1978;20:88-95.
329
Ferris BG, Burgess WA, Worcester J. Prevalence of chronic respiratory disease in a pulp mill in the United States. Br J lnd Med 1967;24:26-37. Greaves lA, Ferris BG, Essex D. Respiratory effects of sulfur dioxide (S02) among corn refinery workers. Am Rev Respir Dis 1984;129:AI57. Kehoe RA, Machle WF, Kitzmiller K, leBlanc TJ. On the effects of prolonged exposure to sulfur dioxide. J lnd Hyg 1932;14:159-173. Osterman JW, Greaves lA, Smith TJ, Hammond SK, Robins JM, Theriault G. Respiratory symptoms associated with low level sulfur dioxide exposure in silicon carbide production workers. Br J Ind Med 1989;46:629-635. Skalpe 10. Long-term effects of sulfur dioxide exposure in pulp mills. Br J Ind Med 1964;21 :69-73. Smith TJ, Peters JM, Reading JC, Castle CH. Pulmonary impairment from chronic exposure to sulfur dioxide in a smelter. Am Rev Respir Dis 1977;116:31-39.
