Estimating the risks of respiratory symptoms amongst workers in the UK wool industry. Final report to the HSE project 1/HPD/126/144/91

A study of the respiratory health of 2153 UK textile workers took place in 1985. Previous analyses had shown that prevalences of symptoms of chronic bronchitis, breathlessness, persistent rhinitis and persistent conjunctivitis were all clearly related (after adjustment for age, sex, smoking habit and ethnic background) to concentrations of inhalable dust in current jobs. These analyses included men with missing exposure data for some of the occupations in which they had worked; gaps in their exposures were filled by making assumptions which were considered reasonable but were untestable.Recent consideration of possible control limits for the industry has focussed attention on predictions of health effects at low concentrations. While the early analyses had produced models which generally fitted the data well, the predictions at the lower concentrations were not examined separately. In order that decisions on control should be informed by the strongest possible evidence, we have reanalysed the data after omission of those workers who had gaps in their exposure data. Particular attention has also been paid to the effect of changes in the form of the prediction model on predictions at the extremities of the data.The new results confirm the earlier broad conclusions that each of the symptom complexes is dust-related, but have led to some revisions of the prediction models and to improved predictions of risks at low concentrations. These new predictions show that, in a workforce containing similar proportions of smokers and non-smokers, men and women, and Europeans and Asians to the study population, we would expect 8% of workers exposed to an average dust concentration of 0.5 mg.m-3 to show symptoms of chronic bronchitis compared to 14% of workers exposed to 5 mg.rrr3 and 16% of workers exposed to 10 mg.m-3.Expected prevalences of breathlessness among Europeans, under similar assumptions, ranged from 5% at 0.5 mg.m-3 to 8% at 5 mg.m-3 and 9% among workers with an average exposure of 10 mg.m-3. Breathlessness among Asians did not appear to be related to dust concentration, although this may have been due partly to the jobs done by Asians, the majority of which entailed exposure to low levels of dust only. Expected prevalences of rhinitis were in general higher than the othersymptoms, with 17.5% of workers exposed to 0.5 mg.m-3 expected to show symptoms, rising to 24% at 5 mg.rrr3 and 26.5% at 10 mg.m-3.For conjunctivitis, expected prevalences varied between European and Asian workers and, among Asian workers, between those who were interviewed in English and those who were interviewed in Urdu. For both European workers and English-speaking Asians, 5% of workers exposed to 0.5 mg.m-3 would be expected to show symptoms of conjunctivitis; however, at higher dust concentrations expected prevalences among Europeans were higher than among Asians. At an average concentration of 5 mg.m-3, 14% of Europeans would be expected to report symptoms of conjunctivitis compared to 8% of Asians, while at 10 mg.m-3 the expected prevalences were 18% and 9.5% respectively. Expected prevalences among Asians at the higher dust levels, however, appear to overestimate those observed in the data, perhaps due to the lower dust concentrations experienced by this group and the consequent uncertainty in extrapolating to higher dust levels. These exposure-response relationships are illustrated in Figures 4.3 and 4.4 of the main report.Because we have observed differences in the patterns of results recorded in different ethnic and smoking groups, and between those speaking different languages, it was not appropriate to make overall predictions for working populations other than the study group without specifying in advance the proportions of smokers and non-smokers, males and females, Europeans and Asians.However, for each of the symptoms, there was strong evidence for a dose-response relationship with current dust concentration, which extends well below the nuisance dust level of 10 mg.m-3 and is observable to very low levels. No evidence was found of a dust threshold below which risk was not increased above the natural background. “”