INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The list of causes of occupational asthma (1) continues to grow, with sensitization providing an explanation for the majority of cases (2). Severe respiratory insults, however, such as the inhalation of high concentrations of acid, can cause an asthmalike syndrome (independently of sensitization) known as the reactive airways dysfunction syndrome (RADS) (3). It is not known if chronic low-dose irritant exposure plays a role in the development of asthma or other chest diseases. Increased ambient levels of low molecular weight irritants increase respiratory symptoms (4, 5) and the incidence of acute bronchospasm in asthmatic patients (6, 7), but there is little evidence to support a role for irritants in the etiology of asthma. (8) Occupational exposure to low-dose irritant fumes offers an opportunity to compare symptoms and respiratory function in exposed and unexposed subjects.

Glass bottle factory workers in the United States have been reported to experience an excess of work-related respiratory symptoms (9). These symptoms of wheezing, chest pain, exertional dyspnea, and cough were thought to be due to the workers' exposure to hydrochloric acid fumes produced in the bottle-making process. We reported a cluster of cases of work-related symptoms in a local, similarly designed, glass bottle factory (10). These workers were concerned that they might have occupational asthma. We designed a matched pair comparison study to test the hypothesis that there was an excess of symptoms in the glass bottle factory. We also investigated the prevalence of bronchial hyperresponsiveness to inhaled methacholine and the cough response after inhalation of citric acid and capsaicin.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Glass bottles are manufactured on a "lehr line" consisting of hot and cold processing areas. Molten glass is shaped on moulds swabbed with sulphur, and then hardened with stannic chloride, which is sprayed on to the hot glass. To prevent sticking in the packing process, the bottles are sprayed with an organic lubricant. Employees work close to the machinery, often without respiratory protection. Thus, they are exposed to heat and low molecular weight irritant substances, including oxides of sulphur, hydrochloric acid (produced in the hydrolysis of stannic chloride by steam), and breakdown products of the organic lubricant. They are not exposed to any known sensitizing agent that could cause occupational asthma (1), and recorded levels of known irritant substances have been reported by glass industry surveys not to exceed occupational exposure limits.

Selection and Recruitment of Subjects

Full union membership lists were obtained for the factory manual work force. These lists included workers who had retained union membership but had left within the previous 5 yr. A total of 505 workers were listed, of whom 50 were selected using computer-generated random numbers. Of these, 41 (37 men and 4 women) agreed to participate in the study. Workers from the hospital nearest to the factory were recruited initially by advertisement on to a list of volunteers (87 men and 31 women). This list included details of age, sex, smoking habit, and history of allergy. When suitable matches were identified between randomly selected workers and listed volunteers, volunteers were contacted and invited to participate in the study; 45 hospital workers were invited to participate in the study.

Matching

Hospital workers were eligible only if presently or previously engaged in a task that would place them in the same socioeconomic group as the factory work force. All study participants lived in the same local area. Matching for risk factors for bronchial hyperresponsiveness (11, 12) was then carried out prior to analysis of symptoms, spirometry, serial PEFR, flow cytometry, and methacholine data. Cough data were analyzed after matching workers for sex, age, area of residence, and smoking.

Detection of Symptoms

Each worker was interviewed regarding respiratory symptoms in a structured manner by one of three investigators. Symptoms of upper and lower respiratory tract irritation or inflammation were recorded as being present or absent. Symptoms sought were ocular symptoms, excluding decreased visual acuity, and nasal or throat symptoms as well as lower respiratory symptoms of cough, sputum production, wheeze, and shortness of breath. Work-relatedness of symptoms was sought by asking for any relationship to work, improvement away from work at the weekends, and the effect of holidays. The interview then continued to include documentation of all the present and past employment of the workers.

Methacholine Challenge

Methacholine challenge testing was carried out using a Mefar MB3 dosimeter (Mefar, Brescia, Italy) and Vitalograph Spirotrac software version 2.02 (Vitalograph Ltd, Buckingham, UK). Nebulized methacholine was administered in doubling cumulative doses over the range 3.125 to 6,400 µg using a modified form of the method of Hendrick and colleagues (13). Airway responsiveness was expressed as the dose required to produce a 20% fall from the baseline FEV₁ (i.e., conventional PD₂₀). The reproducibility of this method in our hands has been assessed and found to be very similar to previously reported figures (14, 15).

Cough Challenges

Cough challenges to citric acid and capsaicin were performed as previously described (16). Briefly, logarithmic incremental inhaled concentrations of citric acid over the range 10 to 1,000 mM and capsaicin over the range 1 to 100 µM were administered using a Mefar dosimeter. Four inhalations, each lasting 1 s, were given at each concentration. The number of coughs in the 10-s period after each inhalation was counted, and a D2 (concentration at which two coughs occurred after each inhalation) calculated from the intercept of a plot of concentration against mean number of coughs after inhalation.

RAST and Flow Cytometry

A measure of atopic status was obtained by RAST for IgE to common environmental allergens (house dust mite, cat fur, and mixed pollens) using ¹²⁵I-labeled anti-human IgE (Pharmacia, Uppsala, Sweden). A RAST percent binding greater than 1% was considered positive. Three-color flow cytometry was used to measure both phenotypic and inducible cell surface markers on lymphocytes (CD3, CD8, CD25, CD4), and 100 µl aliquots of whole blood were stained with appropriate fluorescently labeled monoclonal antibodies and lysed after a 15-min incubation using the Immunoprep system (Coulter Electronics, Hamptom, Herts, UK). Samples were analyzed on a calibrated Epics-XL four-channel flow cytometer: 10,000 lymphocytes were analyzed in each test sample.

Statistics and Power Calculation

The sample was calculated to detect a 20% difference in the proportion of workers in each group recording a quantifiable (i.e., less than 6,400 µg) methacholine challenge result with 80% power and 95% confidence intervals. The frequency of symptoms was compared using McNemar's test (17), and confidence intervals were calculated for the difference in proportion between paired samples. In the analysis of the methacholine data, Wilcoxon's paired rank sum test was used due to the skewed nature of the data and the need for a right censored correction. The ² test for trend was used to compare the methacholine data obtained from both glass bottle workers and control subjects with that obtained in a large community survey (18). Normal probability plots (SPSS for Windows) were used to confirm that log-transformed cough data obeyed a normal distribution. These data were then compared using student's paired t test.

Ethics

Permission for the study was given by both the South Sheffield Research Ethics Committee and the local Research Ethics Committee. Written informed consent was obtained from all participants.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population and Matching

The matching process resulted in 37 pairs of bottle workers and control subjects. The results of this matching are shown in Table 1. The pairs were found to be well matched for baseline and predicted FEV₁. Thirty glass bottle workers and 30 control subjects completed cough challenge. These volunteers were therefore rematched.

Symptoms and Work-relatedness

There was an excess of symptoms among the bottle workers (Table 2). Wheeze was the only symptom for which there was not a significant difference between bottle workers and control subjects (p = 0.15). Approximately three quarters of bottle workers with upper respiratory symptoms described these as being worse at work, and 64% of those reporting cough described this as being worse at work; 29% of bottle workers reported work-related exacerbations of wheeze and 24% reported shortness of breath being worse at work. Eight bottle workers (22%) and three control subjects (8%) met a clinical case definition of chronic bronchitis.

Methacholine Challenge

Methacholine challenge results (Figure 1a) did not differ significantly between the matched groups (p > 0.2). Because of the close similarity of methacholine challenge methods, and the known equivalence of reproducibility with the technique, (14, 15), a comparison with a large randomly selected community survey (18) was carried out. Comparison with this survey showed that the results were not significantly different from that in the survey in both the bottle factory (² = 4.29, df = 2, p > 0.2) and control (² = 1.24, df = 2, p > 0.2) groups.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Inhalation challenge results in groups: (a) methacholine challenge testing; (b) cough test with citric acid; (c) cough test with capsaicin.

Flow Cytometry

Bottle workers and control subjects did not show changes consistent with asthma in either phenotypic or inducible cell markers on peripheral blood lymphocytes. Changes sought were increases in CD25 expression on both CD4 and CD8 populations and changes in the absolute and relative numbers of CD3, CD4, and CD8 lymphocytes.

Cough Challenge

Both citric acid (p = 0.031) and capsaicin (p = 0.036) cough D2 were significantly lower in bottle workers (Figure 1b and c). The mean differences and 95% confidence intervals for the differences for citric acid and capsaicin were 1.96 mM (1.07 to 3.61 mM) and 2.0 µmol (1.67 to 2.39 µmol), respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, randomly selected glass bottle workers exposed to low-dose irritant fumes were shown to exhibit a significant excess of respiratory symptoms compared with matched hospital artisans. This excess of symptoms was predominantly work-related, but it did not include wheeze, and neither was there an excess of bronchial hyperresponsiveness during methacholine testing. Peak flow data collected were of too poor a quality to analyze. There was a significantly lower cough threshold to irritant stimuli in glass bottle workers. There was therefore evidence of airway irritation but not asthma.

Is it consistent with this finding, however, that all the asthmatic workers had left the factory because of the adverse conditionsthe so-called "healthy worker effect?" This seems unlikely given that the population sampled was a union membership list including workers who had left the workplace as much as 5 yr previously. Workers with medical complaints were more likely to retain union membership in order to seek medicolegal help. Second, community surveys (11, 12, 19) have shown that the major determinants of bronchial reactivity in the general population are age, atopy, and airway caliber. In our study, we have attempted to match bottle workers and control subjects for these factors. The distribution of bronchial reactivity in our study showed a strong similarity to that of a much larger community survey (18). It seems likely that the randomization and matching were therefore successful and that there was no large local effect on the prevalence of bronchial hyperresponsiveness in the factory.

This failure of correlation between irritant exposure and asthma, despite a symptom excess, has been reported in other industries. A study of 668 synthetic fiber plant workers (20) failed to show a correlation between irritant exposure, symptoms, and airway responsiveness. In a large study of shipyard workers, no correlation was found between questionnaire reported symptoms and methacholine challenge results (21, 22). In a study of 22 symptomatic workers exposed to potassium aluminum tetrafluoride flux (23), only two subjects had a decrease of 20% in the FEV₁ after 0.1% methacholine provocation. Our inability to demonstrate bronchial hyperreactivity in irritant-exposed workers is therefore consistent with previously published findings. Previous studies have observed a similar excess of symptoms (9, 20, 24), but in our study there was a striking absence of reported wheeze. We believe that this is strong evidence that chronic exposure to low molecular weight irritants such as that seen in our study population does not lead directly to chronic bronchial inflammation or to the syndrome of occupational asthma.

Flow cytometric studies have demonstrated specific changes in the peripheral blood lymphocytes in atopic asthma (25) and isocyanate-induced asthma (26). It is in keeping with the respiratory findings above that these flow cytometric changes were not reproduced in this group.

Cough challenging produced a significantly lower threshold for both tussive stimuli despite random selection from a large work force. There are two important implications inherent in these results. First, chronic exposure of workers to irritant fumes may have respiratory sequelae other than that currently described as occupational asthma or RADS. Second, exposure to irritants in the community may cause respiratory sequelae other than that detected by conventional estimation of bronchial hyperreactivity. In both of these contexts, the long-term consequences of cough, which may indeed be favorable in terms of lung protection, are unknown. This can only be adequately tested in longitudinal studies. In the meantime, cough remains an unwelcome symptom.

The tussive agents used in this study are thought to stimulate sensory nerves, acting on rapidly adapting receptors and nonmyelinated C-fibers (27). Low-dose irritant exposure may alter the local milieu surrounding these afferent sensory receptors. Neurogenic inflammation could then, in extreme cases, provide the link with the reactive airways dysfunction syndrome (30). It has been hypothesized in RADS that a large insult leads to inflammation and subsequent atypical repair with increased bronchial reactivity (31). Our study supports the view that chronic low-dose irritant exposure does not cause inflammation sufficient to cause airway hyperresponsiveness, but it may lead to a syndrome of cough and airway irritancy (CAIS).

In conclusion, concentration on the prevention and early detection of occupational asthma may obscure the true diagnosis in irritant-exposed workers. In the general population, irritant exposure may have important consequences on respiratory morbidity that are undetected by community surveys of bronchial hyperreactivity.