A community-based case-control study of asthma and chronic bronchitis in relation to occupation. Final report on HSE Contract No. 3149/R58053)

Background: A community-based epidemiological study was conducted in the West Lothian and Central regions of Scotland in 1994 by postal questionnaire to seek information on the prevalences of occupationally related asthma and chronic bronchitis (Cowie et al, 1997). A total of 16990 valid questionnaires were returned, which represented a 50% response rate. The prevalence of asthma of all kinds was estimated to be 8% and the prevalence of chronic bronchitis was estimated to be 15%. The prevalence of asthma, and of chronic bronchitis, related to work was estimated to be 0.5% and 2.2% respectively. Clinical assessments of a sample of respondents confirmed the reliability of diagnoses of asthma using a self-administered questionnaire, but suggested that the work-relatedness of asthma was less reliably diagnosed.Each respondent was asked to complete a full occupational history including all jobs held since leaving school. The purpose of the current study was to investigate the associations between occupations previously held and prevalent symptoms of asthma and chronic bronchitis. A further aim was to use the occupational histories to investigate the effect of the onset of asthma symptoms on career progression.Methods: A case-control design was used, with cases and controls selected from respondents to the earlier prevalence study. A total of 424 asthma cases (224 work-related, 200 non-work-related) were individually matched by age to 424 controls, to within one year of age. In separate analyses, 408 chronic bronchitis cases (208 work-related, 200 non-work-related) were compared, unmatched, to the same set of controls. Asthma cases were diagnosed using responses to standard questions on wheezing and whistling in the previous twelve months. Chronic bronchitis cases were diagnosed from responses to standard MRC questions on regular coughing and phlegm production. In both analyses, all work-related cases identified in the prevalence study were included, together with a random sample of non-work-related cases. Controls were randomly sampled from those respondents reporting no history of respiratory symptoms. Occupations and industries in the occupational histories were coded and categorised into a smaller number of occupational groups which have commonly been used to report occupational respiratory disease rates.Results: Smoking status was an important predictor of respiratory symptoms in the analysis of asthma and, in particular, chronic bronchitis. Those reporting symptoms of asthma or chronic bronchitis were more likely to be smokers than non-smokers or ex-smokers. Those reporting symptoms of chronic bronchitis were more likely to be male than female, but the proportions were similar among cases of asthma.After adjustment for smoking, significantly raised odds of asthma symptoms were reported among those who had previously been employed in food processing, particularly bakers, and as caterers, cleaners, textiles processors and general labourers. In part due to the profile of local industries, those employed as food processors were commonly involved in the processing of meat products, in particular poultry meat, while textiles processors were most often employed in the manufacture of clothing. Also associated with raised odds of asthma symptoms were print workers, metal treatment workers and spray painters, but estimates were imprecise due to the small numbers involved.After adjustment for smoking, raised odds of chronic bronchitis symptoms were associated with a wider range of occupations than were asthma symptoms. These occupations included bakers, farmers, coalminers, construction workers, painters and chemical industry processors. A wide variety of occupations in the processing of metal and electrical products were associated with raised odds of chronic bronchitis, in particular welding, metal treatment and the motor vehicle trades.Most of those who reported symptoms of asthma generally reported that those symptoms first occurred in the few years prior to the postal survey. Therefore, very few cases or their matched controls started new jobs in the short period after the onset of symptoms until questionnaire completion. There was no evidence that cases started a greater number of jobs after the onset of symptoms compared to their age-matched controls over the same period. Cases who did start a new job after the onset of symptoms generally did not experience a change in the Social Class of the occupation, and there was no strong evidence, given the small numbers involved, that the movements between Social Classes was different compared to the controls. The time spent in employment after the onset of symptoms was only slightly lower for asthma cases than for their matched controls over the same period. The difference amounted to 1 year less for male case-control pairs and 2 years less for female case-control pairs, when only pairs with a minimum of 10 years follow-up (mean 19 years) were used in analysis. Neither of these differences was large in relation to the degree of variation among case-control pairs.Conclusions: The reporting of symptoms of asthma, and of chronic bronchitis, was associated with employment in a number of occupations held previously, even after allowing for different smoking habits. There was overlap between the occupations associated with symptoms of asthma and with symptoms of chronic bronchitis due to the difficulty in diagnosing the two conditions using a self-administered questionnaire. However, asthma symptoms were particularly implicated with employment in the food processing, catering and textiles manufacturing industries. Chronic bronchitis symptoms were associated with a wider range of occupations than were asthma symptoms, and, in particular, with employment in known dusty jobs such as coalmining, and occupations in the metal processing industries and manufacturing of electrical products. Both work-related and non-work-related asthma cases separately showed similar associations with occupations. Therefore, given also the generally poor agreement with clinical assessments in the earlier prevalence study, the use of the self-administered questionnaire by itself may not have been a reliable method of determining the work-relatedness of asthma symptoms. Based on a paired comparison of career progression in prevalent cases and controls with limited follow up, there was no strong evidence that the onset of asthma symptoms had later adverse effects on employment prospects.