Doctor of Philosophy
Graduate School of Public Health
Wu, Jian, Respiratory health of coke oven workers: an epidemiological study, Doctor of Philosophy thesis, Graduate School of Public Health, University of Wollongong, 2003. http://ro.uow.edu.au/theses/1890
Studies on non-fatal health indicators represent positive developments for occupational disease prevention. The indicators include disease incidence, symptoms, and impaired physiological function. While cancer mortality of coke oven workers has been extensively investigated, respiratory health of coke oven workers is a poorly investigated area. Very few studies have ever been reported. Almost all of these are cross-sectional. This study utilised data extracted from a lung function surveillance system on coke oven workers in a steelworks to investigate the association between respiratory health and exposure to coke oven emissions, particularly the exposure effects on lung function and respiratory symptoms.
Coke oven emissions are volatile materials escaping from the coke producing process. The emissions contain various complex chemical compounds, some of which are suspected to have hazardous health effects. The major pollution sources of coke oven emissions are emission and leakage during the coke producing process. Chemical components of coke oven emissions are absorbed mainly through respiratory and dermal routes.
A few cross-sectional studies have reported that decreased lung function indices such as FEV1 and of FEV1/FVC were associated with exposure to coke oven emissions. Coke oven workers were also reported to have higher prevalence of bronchitis symptoms such as cough and phlegm.
Longitudinal data from a lung function surveillance system on coke oven workers in a steelworks were available for the period from 1978 to 2000. Some modifications were made to the surveillance system in 1990. There were 1851 individuals with valid data during the surveillance period, and they constituted the study population. Repeated lung function test results, respiratory symptom information, and smoking information were extracted from the surveillance database. Detailed occupational exposure histories were traced back from the payment system and human resource department in the steelworks. Four exposure groups (‘Operation’, ‘Maintenance’, ‘Electricity’ and ‘Other’) based on the perceived exposure levels were used to classify the subjects according to their detailed occupational exposure histories.
The association between lung function and exposure to coke oven emissions was investigated cross-sectionally and longitudinally. While a longitudinal analysis provides direct assessment of effects on individuals, a cross-sectional analysis is also important since not all subjects in the surveillance system had longitudinal data. Cross-sectional analyses on the pooled last (or only) sets of lung function tests were conducted by using multiple regression. Longitudinal analyses on the time trends of lung function change over time were conducted by using a two-stage analysis technique and mixed model. Logistic regression was used to assess the impact of risk factors on respiratory symptoms. The 1990 modifications made it necessary to produce separate analyses on the pre- and post-modification data to fully utilise the information.
The surveillance data set provides a valuable resource for assessing cross-sectional and longitudinal effects of exposure on lung function. Particular strengths of the data are the relatively large sample size, long follow-up time (average around 8 years), and no evidence of recruitment associated selection bias between exposure groups.
A cross-sectional study was conducted on the data collected before the 1990 modifications. The last (or only) sets of lung function tests for 1353 male coke oven workers were pooled for the analyses. For all subjects, the most exposed ‘Operation’ position was associated with reductions in FVC of around 141 ml (p=0.014, 95% CI: 29 ml to 254 ml) and in FEV1 of around 115 ml (p=0.030, 95% CI: 12 ml to 218 ml); each year of working in ‘Operation’ was associated with a reduction in FVC of around 6 ml (p=0.027, 95% CI: 1 ml to 12 ml). For subjects with more than one year of coke oven history, each year of working in ‘Operation’ was associated with reductions in FVC of around 12 ml (p=0.002, 95% CI: 4 ml to 19) and in FEV1 of around 9 ml (p=0.006, 95% CI: 3 ml to 16 ml).
A longitudinal study was also conducted on the data collected for the 580 male workers with at least two sets of lung function measurements before the 1990 modifications. Each year of working in the ‘Operation’ group was found to increase the FVC decline by around 0.7 ml/year (p=0.02, 95% CI: 0.1 ml/year to 1.3 ml/year). After the exclusion of 111 subjects without detailed work history, the above finding was confirmed and the reduced data set also indicated an effect of exposure on FEV1. Each year of exposure in ‘Operation’ was found to increase the FEV1 decline of around 0.8 ml/year (p=0.03, 95% CI: 0.1 ml/year to 1.4 ml/year).
Data for respiratory symptoms and lung function were collected repeatedly from the 764 male coke oven workers after the 1990 modifications. Again, the last (or only) sets of data for all subjects were pooled and analysed cross-sectionally. Working in ‘Operation’ increased the risk for cough (OR=2.37; 95% CI: 1.48 to 3.81), phlegm (OR=2.55; 95% CI: 1.63 to 3.99), ‘cough or phlegm’ (OR=2.40; 95% CI: 1.62 to 3.54) and shortness of breath (OR=1.52; 95% CI: 1.01 to 2.27). Each year of working in ‘Operation’ was associated with a reduction in FVC of around 20 ml (p=0.0001, 95%CI: 12ml to 29ml) and in FEV1 of around 10 ml (p=0.0085, 95%CI: 2ml to 17ml).
The longitudinal information on respiratory symptoms and lung function changes for coke oven workers was analysed on the 515 subjects with two or more sets of tests collected after the 1990 modifications. Logistic regression analyses showed that each year of working in ‘Operation’ was associated with slightly higher risk of having shortness of breath (OR=1.02, 95% CI: 1.002 to 1.03), but not on other symptoms. Each year of working in ‘Operation’ was found to increase the annual FVC decline by around 1.8 ml (p=0.0017, 95% CI: 0.7 ml to 2.9 ml) and increased the annual FEV1 decline by around 1.5 ml (p=0.0012, 95% CI: 0.6 ml to 2.4 ml).
It is believed that exposure in the early years involved higher exposure levels than that in the late years. To explore whether a difference in effect size (or dose-response relationship) can be observed in different exposure periods, lung function tests from the pre- and post-1990 modifications were combined and analysed longitudinally for the 981 male workers with at least two sets of tests using a mixed model. A mixed model includes more than one random term and takes into account the dependence of the response variable when modelling the repeated measurements. During the period from 1973 to 1983, each year of working in ‘Operation’ was associated with a reduction in FVC of around 40 ml (p<0.0001, 95% CI: 32 ml to 49 ml) and in FEV1 of around 31 ml (p<0.0001, 95% CI: 23 ml to 39 ml). In the period after 1987, each year of working in ‘Operation’ was associated with a reduction in FVC of around 6ml (p=0.02, 95% CI: 1 ml to 11 ml) and no exposure effect was found on FEV1.
In summary, various studies all indicated broadly the same conclusion. Work duration in the most exposed position in the coke ovens had an adverse impact on the respiratory health of coke oven workers. These findings are similar to those found in the literature, even though the only available information is from cross-sectional studies.