The aim of this study was to investigate the socioeconomic outcomes of subjects who experienced workrelated asthma symptoms in the absence of demonstrable occupational asthma (OA) and to compare these outcomes with those found in subjects with documented OA.
Subjects (n=157) who were being investigated for workrelated asthma, were surveyed. Of these 86 had OA, ascertained by a positive specific inhalation challenge (SIC), and 71 subjects had a negative SIC response. After a median interval of 43 months (range 12–85 months), the subjects were interviewed to collect information on employment status, income changes, and asthmarelated work disability.
Rates of work disruption and income loss at followup were similar in subjects with negative SIC (46% and 59%, respectively) and in those with OA (38% and 62%). The median loss as a percentage of initial income was 23% in subjects with negative SIC and 22% in subjects with OA. Asthmarelated work disability, defined as any job change or work loss due to asthma, was slightly more common in subjects with OA (72%) than in those with negative SIC (54%).
This study shows that, even in the absence of demonstrable occupational asthma, workrelated asthma symptoms are associated with considerable socioeconomic consequences.
This study was supported by the Services Fédéraux des Affaires Scientifiques, Techniques et Culturelles, Programme d'appui scientifique à la protection des travailleurs grant SSTC PS/10/44.
Epidemiological surveys have highlighted the importance of workplace exposure in initiating or triggering asthma symptoms 1, 2. Workrelated asthma symptoms can result either from immunological occupational asthma (OA) with a latency period or from concurrent or preexisting asthma worsened by irritants or physical stimuli in the workplace 3. A substantial proportion of subjects experiencing worsening of asthma symptoms at work have failed to demonstrate objective evidence of workrelated changes in physiological indices of asthma 2, 4. However, distinguishing between subjects with OA and those who are symptomatically worse at work is important for both medical and medicolegal reasons 5. Immunological OA is characterized by the development of specific bronchial hyperreactivity to occupational agents, although the precise mechanisms involved in OA caused by low molecular weight chemicals have not yet been identified precisely 3. In subjects with OA, repeated exposure to the causative agent is associated with an increase in airway inflammation and in nonspecific bronchial hyperresponsiveness 6, 7, which can lead to progressive worsening of asthma. Therefore, the treatment for OA is to remove the affected worker from exposure to the causative agent. Several studies have explored the economic consequences of OA 8–14. By contrast, little information is available on the outcome of subjects who are symptomatically worse at work 9.
The aim of this study was to characterize the socioeconomic outcome of subjects with workrelated asthma symptoms in the absence of demonstrable OA. Therefore, a group of subjects who experienced workaggravated asthma symptoms, but failed to develop a positive response to specific inhalation challenge (SIC) with occupational agents, were investigated. The socioeconomic outcomes of these subjects were compared with those in subjects with OA.
Subjects (n=186) who were referred to the authors' centre, between the years 1994–1999, for investigation of asthma symptoms temporally related to workplace exposure were surveyed. Based on the results of SIC, the subjects were categorized into two groups: 1) those who demonstrated positive SIC were regarded as having OA and 2) those who failed to develop a bronchial response during SIC were categorized as experiencing workaggravated asthma symptoms without objective evidence of OA.
A total of 201 subjects had been investigated for workrelated asthma symptoms during this period. Six subjects were excluded from the cohort because their clinical history was consistent with irritantinduced asthma. In these subjects, the onset of asthma symptoms clearly occurred within hours after unusually high level exposure to irritants at work. Three of these six subjects underwent SIC in order to exclude sensitization to the causative agent and these SICs were negative. Nine other subjects were not enrolled in the followup study because the results of SIC could not be interpreted accurately: three subjects could not be challenged with the agent or process that was suspected of provoking asthma symptoms at work, and six subjects showed nonreproducible changes in forced expiratory volume in one second (FEV1) during SIC.
Initial assessment at the time of the diagnostic investigation included a detailed occupational and medical questionnaire, skinprick testing with a battery of common inhalant allergens and with occupational agents when relevant extracts were available, measurement of spirometry and nonspecific bronchial hyperresponsiveness to histamine and SIC. The subjects' medical records were analysed by a research assistant in order to retrieve information on sociodemographical, occupational, and medical characteristics. Information on asthma symptoms and use of medications were collected either during the last month before SIC, if the subject was still exposed to the offending agent(s), or during the last month at work, if the subject had been removed from exposure.
Eligible subjects were sent an invitation to participate, detailed information on the study protocol, and an informed consent form. The invitation explained that information obtained through the study would not be used for medicolegal purposes. The protocol of the study was approved by the Ethics Committee of MontGodinne University Hospital. The subjects were then contacted by telephone and given a structured questionnaire to obtain the following information: job changes related to asthma symptoms, current working status, current exposure to agent(s) causing asthma symptoms, perceived magnitude of income changes, actual income during the last year at work prior to initial assessment, income at the time of followup and compensation status.
Lung function assessment
At the time of diagnostic evaluation, FEV1 and forced vital capacity (FVC) were measured according to the recommendations of the American Thoracic Society 15 using a pneumotachograph instrument (Medisoft, Dinant, Belgium). Predicted values were those of the European Coal and Steel Community 16. Nonspecific bronchial hyperresponsiveness to histamine was assessed using the method described by Cockcroft et al. 17. The level of nonspecific bronchial hyperresponsiveness was expressed as the provcocation concentration of histamine causing a 20% fall in FEV1 (PC20). PC20 values <16 mg·mL−1 were considered as reflecting significant bronchial hyperresponsiveness.
Specific inhalation challenges
SICs were performed as previously described 18. On the first day, the subjects were exposed to a control substance to ensure that there was no significant fluctuation of the FEV1. On subsequent days, the subjects were challenged with occupational agent(s) suspected of causing workrelated asthma based on clinical history and inspection of the workplace by hygienists from the Workers' Compensation Board (WCB). Exposure to occupational agents was produced in a realistic way, by reproducing the physical features (e.g. temperature, dust, aerosol, vapour or fume) encountered at the workplace. The duration of exposure was gradually increased to a total of 2 h on the first challenge day. In the absence of significant change in FEV1, challenge exposure was repeated for 2–3 h on at least one subsequent day. Baseline histamine PC20 value was determined at the end of the control day and reassessed 6–8 h after the end of each challenge exposure in the absence of a significant change in airway calibre. A SIC result was considered positive if there was: 1) a sustained fall in FEV1 of ≥20% or 2) a significant (>3‐fold) decrease in postchallenge histamine PC20 as compared with the baseline value 18.
Analysis of results
Asthmarelated work disability was defined as any selfreported job change or work loss due to asthma 19. The severity of asthma was graded using a score proposed by Blanc et al. 20 for assessing work disability among adult asthmatics. This score (ranging 0–28) is based on hospital admissions, frequency of asthma symptoms, and use of asthma medications. Atopy was defined by the presence of a positive skin test to at least one common allergen.
Data are presented as the median with 25th and 75th percentiles. Comparison between subjects with OA and those with a negative SIC response was made using the Chisquared test, Fisher exact test or Wilcoxon ranksum test as appropriate. The Spearman coefficient was used to assess correlations between variables. Multivariate logistic regression analysis was carried out to explore the determinants of unemployment at followup. This analysis included the following independent variables: positive or negative SIC, asthmaseverity score, the logarithm of histamine PC20 value at initial examination, time elapsed since onset of workrelated symptoms, age, level of education and size of the company. A p‐value <0.05 was considered significant.
At the time of the followup assessment, five subjects (including one subject with OA) had died from nonrespiratory disease. Of 181 subjects, 157 (87%) agreed to participate, while 24 subjects either could not be traced or declined to participate in the study (n=7). Nonparticipants included a higher proportion of current smokers (10 of 24 versus 21 of 157, p=0.003) and of subjects referred by the WCB (22 of 24 versus 97 of 157, p=0.004) than participants. The asthmaseverity score at initial assessment was slightly lower in nonparticipants (median score 6, 25–75th percentile 4–8) than in participants (median score 7, 25–75th percentile 5–9, p=0.037).
Baseline characteristics of participants
Of the 157 participants, 86 (55%) subjects showed a positive SIC response and were diagnosed as having OA, whereas 71 subjects demonstrated a negative SIC. Low molecular weight agents were more frequently involved in subjects with negative SIC (62%) than in those with OA (44%, p=0.026) (table 1⇓). Of the 71 (52%) subjects with negative SIC 37 underwent separate challenges with more than one occupational agent before excluding the diagnosis of OA. Subjects with OA were slightly younger, had a higher level of education and professional qualification, and included a lower proportion of smokers than subjects with negative SIC (table 1⇓). Interestingly, a history of asthma before employment was less prevalent among subjects with workaggravated asthma symptoms (five of 71, 7%) than among those with OA (17 of 86, 20%, p=0.022). The prevalence of preexisting asthma was similar in subjects referred by the WCB for medicolegal purposes (13 of 97, 13%) and in those who were referred by their attending physician (nine of 60, 15%). Also, a history of asthma before employment was not affected by the nature of the causative agent. The agent suspected of causing workrelated asthma symptoms was a high molecular weight compound in 72 of 135 (53%) subjects with OA and in 10 of 22 (46%) subjects with negative SIC.
At initial assessment, there were no significant differences in FEV1 and global asthmaseverity score between subjects with OA and those with negative SIC (table 2⇓), although histamine PC20 values were lower in subjects with OA (0.72 mg·mL−1 (0.09–2.57)) than in those with negative SIC (1.46 mg·mL−1 (0.23–5.27)).
The impact on employment and income was only slightly different in subjects with OA than in those with negative SIC (table 3⇓). The proportion of those currently working was 54% in subjects with negative SIC and 62% in those with OA. Retirement was more frequent in subjects with negative SIC (11 of 71, 15%) than in those with OA (two of 86, 2%, p=0.005). Asthma was the selfreported cause of work disruption (i.e. unemployment, chronic sick leave, or retirement) in 27 of 33 (82%) of subjects with OA and in 20 of 33 (61%, p=0.057) of those with negative SIC. Thirtyeight subjects with negative SIC (54%) reported asthmarelated work disability, as defined by being unemployed or having changed jobs because of asthma symptoms, as compared with 62 subjects (72%) in the OA group. Only 5% of subjects in both groups benefited from a professional rehabilitation programme. A similar proportion (∼30%) of subjects in both groups remained exposed to the agent(s) causing asthma symptoms. However, a higher proportion of subjects with negative SIC (20%) than with OA (7%) were exposed to unchanged levels of the causal agent(s), whereas the level of exposure was reduced in 26% of those with OA and in only 8% of those with negative SIC. The results of the logistic regression analysis exploring the factors that determined employment status at followup are presented in table 4⇓. The risk of being unemployed was significantly associated with an older age and a lower level of education.
A similar proportion of subjects with OA (62%) and with negative SIC (59%) reported that they suffered a reduction in income (table 3⇑). There was a good correlation between perceived and actual loss of earnings (rho=0.87, p<0.001). The median (25–75th percentile) actual loss as a percentage of initial income was 22% (0–44) in the OA group and 23% (0–41) in the group with negative SIC. The magnitude of perceived loss of earnings correlated inversely with the level of exposure to the causal agent at followup (rho=−0.48, p<0.001). A loss of income was reported by four of 20 (20%) subjects with persistent exposure, nine of 28 (32%) subjects with reduced exposure, and 83 of 107 (78%) subjects who were no longer exposed. Of 82 subjects with OA who had filed a claim for compensation to the WBC, 75 (91%) had been compensated at the time of followup. A lower proportion of subjects with negative SIC (78%) had attempted to claim compensation and of these, only two received compensation. Among subjects with OA, the loss of earnings was offset by the WCB permanent disability indemnity in only 10 of 45 (22%) subjects who reported an income loss for whom this information was available.
Limitations of the study
Potential limitations of this study should be carefully considered, since the validity of the inferences that can be drawn from the findings depends to a great extent on the method used to identify the disease of interest and on the populations being studied. Much controversy arises from the absence of objective criteria for defining workrelated asthma in the absence of immunological OA or reactive airwaysdysfunction syndrome 3. Terms such as exacerbation, aggravation, and worsening of asthma at work refer to selfreported symptoms but not to physiological indices of asthma 2, 3, 5. A group of subjects who demonstrated a negative SIC response were investigated as a proxy for those who experienced worsening of asthma symptoms at work without objective evidence of workrelated worsening of asthma disease. Subjects with a clinical history consistent with reactive airwaysdysfunction syndrome or acute irritantinduced asthma were excluded from this study. Accordingly, the group with negative SIC included subjects with preexisting or coincidental asthma whose symptoms were temporally related to workplace exposure, although the underlying pathophysiological mechanisms were unknown.
This study group could also include some subjects with falsenegative SICs, although this is unlikely to have occurred. A long interval of time between the investigation and last work exposure could lead to falsenegative results 3. Although approximately half of the subjects with positive and negative SIC had been removed from their jobs at the time of the tests, the duration of removal did not differ between the two groups. The subjects were challenged with the suspected agent(s) for prolonged periods (at least 4 h) before excluding OA. In addition, subjects who demonstrated a significant and reproducible decrease in postchallenge histamine PC20 value were considered as having OA, even in the absence of an asthmatic reaction 18, 21. Some subjects may not have been challenged with the agent that actually caused asthma at work. However, SICs were carried out in a realistic way, aimed at reproducing the workplace exposure. Approximately half of the subjects were challenged with multiple agents to which they were exposed at work. Also, subjects with indeterminate results of SIC and those for whom the SIC in the laboratory could not adequately reproduce the mode of exposure at work were excluded from the study.
This study assessed subjects who had been referred to a specialized centre by their attending physicians (38%), including general practitioners, chest physicians and occupational physicians, or by the WCB (62%). The cohort in this study was not formed from a selection bias towards more complex cases. Frenchspeaking workers who filed a claim for compensation have been systematically referred to the authors' centre for SIC since 1993, although SICs with flour were requested only when there was discordance between the results of the other diagnostic procedures. Therefore, OA caused by flour may be underrepresented in the study population. Nevertheless, it is likely that the subjects represented only a subset of those who experience workrelated asthma, since their symptoms were severe enough for them to seek specialized medical advice or claim compensation. Conversley, severe asthmatics were probably underrepresented in this study population, as SIC are contraindicated in subjects with severe airway obstruction. For instance, during the study period (1994–1999), SIC could not be carried out in 10 subjects referred to the authors' centre because their asthma was unstable despite appropriate treatment.
The rates of unemployment and income loss in subjects with negative SIC (46% and 59%, respectively) did not differ significantly from those observed in subjects with OA in this study (38% and 62%) or in other European studies 8, 9, 11–14. However, determining the specific effects of workrelated asthma requires a control group of subjects experiencing asthma of similar severity without symptoms at work, since there is some suggestion that asthma per se is associated with a negative impact on employment and job effectiveness 19, 22, 23. Axon et al. 24 found that subjects with occupational asthma were more likely to be unemployed than subjects with nonoccupational asthma (3%), although the subjects were not matched for the severity of asthma. The findings in this study in subjects with negative SIC were similar to those reported in the only study that assessed the outcome of workexacerbated asthma 11. In the study by Cannon et al. 11, subjects investigated for possible OA were considered as having workexacerbated asthma, OA, or asthma unrelated to work based on the clinician's interpretation of a combination of diagnostic procedures, including the clinical history, measurement of specific antibodies, peakflow recordings and SIC, although the criteria used for diagnostic categorization were not precisely defined. The rate of unemployment was not different in workexacerbated asthma (31%), OA (39%), or asthma unrelated to work (32%), although a reduction of income was more frequently reported by subjects with workexacerbated asthma (65%) and OA (62%) than by those with asthma unrelated to work (38%). The effects of asthma on work disability defined by breathingrelated job changes have been investigated in a populationbased study of Swedish adults 19. Respiratory work disability was found in 2.3% of subjects without asthma. 13% of subjects with selfreported asthma, and 22% of asthmatics with documented airway hyperresponsiveness. Using the same definition, 54% of subjects with workaggravated asthma symptoms and 72% of subjects with OA in this study should be considered as having asthmarelated work disability.
In this study population, logistic regression analysis showed that old age and low education level were the most significant predictors of unemployment. Previous studies of subjects with OA identified various sociodemographic factors that adversely affected the employment status, including manual socioeconomic groups 9, low level of education 13, old age 10, young age 13 and small family with a low number of dependants , and smallsized company 13, while the severity of asthma was not an important determinant of the working status.
To conclude, there is controversy as to whether a distinction should be made between nonspecific exacerbation of asthma symptoms at the workplace and immunologicallymediated occupational asthma for prevention, management, and medicolegal purposes 3, 5, 25. This study shows that, even in the absence of demonstrable occupational asthma, workrelated asthma symptoms are associated with a considerable socioeconomic impact. The environmental and host factors that determine worsening of asthma. symptoms at work should be further investigated in order to improve the medical management of this common condition and to avoid unwarranted professional and financial consequences.
- Received August 14, 2001.
- Accepted February 2, 2002.
- © ERS Journals Ltd