Article Text

Download PDFPDF

Oral Session 15 – Cancer: industrial cohorts

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


M. Sim1, J. Hoving1, L. Fritschi2, D. McKenzie1, G. Benke1, A. Del Monaco1, E. MacFarlane1, N. deKlerk3, A. W. Musk4.1Department of Epidemiology and Preventive Medicine, Monash University, Australia; 2School of Population Health, University of Western Australia; 3Centre for Child Health Research, University of Western Australia, 4Dept of Respiratory Medicine, Sir Charles Gairdner Hospital, Western Australia

Introduction: Previous studies in the aluminium industry have found excess rates of lung and bladder cancer in workers in smelters using the Soderberg process. We report findings on a cohort study of workers in several different sectors of the aluminium industry, namely bauxite mining, alumina refining, and pre-bake smelting.

Methods: The cohort comprised 11 446 employees, employed for at least 3 months between 1 January 1983 and 31 December 2002. These workers were matched against the Australian national death and cancer registries for cause of death and occurrence and type of cancer. Expected numbers of deaths and cancers were calculated from national population rates.

Results: Overall mortality in males was significantly reduced with an standardised mortality ratio (SMR) of 74 (95% confidence interval (CI) 68 to 81) and similarly reduced for all major categories of death, including respiratory disease, circulatory disease, cancer and injury. Cancer incidence in males was significantly reduced with a standardised incidence ratio (SIR) of 91 (95% CI 83 to 99). There were higher than expected rates of kidney cancer (SIR 198; 95% CI 117 to 335), stomach cancer (195; 115 to 329), and mesothelioma (241; 100 to 578) in a combined group of smelting, rolling mill, and powerhouse workers. In workers involved in bauxite mining and alumina refining, there were elevated rates of melanoma (130; 100 to 168) and mesothelioma (349; 181 to 670). Linking these latter mesothelioma cases to the relevant state mesothelioma registry, which independently collected asbestos exposure data for all reported cases, we found that almost all had had occupational or environmental asbestos exposure prior to their work in the aluminium industry.

Conclusion: Workers in the alumina and aluminium industry have lower than expected mortality rates, consistent with the healthy worker effect. While the overall incidence of cancer is similar to that in the Australian population, the excess for some types of cancer needs to be explored further to determine the role of work and other factors.


K. Ulm1, P. Gerein1, J. Eigenthaler2, S. Schmidt2, H. Ehnes2.1Institute for Medical Statistics and Epidemiology, Technical University Munich, Germany; 2Steinbruchs-BG, Langenhagen, Germany

Introduction: Workers compensated for silicosis outside the mining industry are at an increased risk for developing lung cancer. In published meta-analyses, no data from Germany are involved. Furthermore, exposure data are necessary to assess a threshold value in order to reduce the risk for silicosis and for lung cancer.

Method: A cohort study among workers compensated for silicosis between 1988 and 2000 from the stone and quarry industry was initiated. The cohort was followed up until the end of 2001. From all workers a detailed description about their jobs was assessed.

Results: There were 440 workers enrolled in the study. During the follow up, 144 workers have died compared with 74.35 expected cases based on the mortality rates of the general population from German,y leading to a standardised mortality ratio (SMR) of 1.94 (95% confidence interval (CI) 1.63 to 2.28). Lung cancer was the cause of death in 16 cases (2.40; 95% CI 1.37 to 3.90). All workers had a peak exposure above 0.15 mg/m3, the current threshold value in Germany. The cumulative exposure was above 2 mg/m3 years and the average exposure was 0.10 mg/m3 or larger. No association between the exposure and the risk of developing lung cancer could be observed.

Conclusions: Workers from the stone and quarry industry compensated for silicosis are at an increased risk for developing lung cancer. In order to reduce that risk the exposure has to be lowered to a peak exposure below 0.15 mg/m3 and an average exposure below 0.10 mg/m3.


I. T. S. Yu1, L. A. Tse1, T. W. Wong1, C. C. Leung2, C. M. Tam2, A. C. K. Chan2.1Department of Community & Family Medicine, The Chinese University of Hong Kong, Hong Kong; 2Pneumoconiosis Clinic, Tuberculosis and Chest Service, Department of Health, Hong Kong

Introduction: Lung cancer has been the main focus of studies looking into the carcinogenic effects of silica dust exposure, but a number of studies have shown increased mortality risk of oesophageal cancer among workers with silicosis. We studied the standardised mortality ratio (SMR) of oesophageal cancer among a cohort of workers with silicosis in Hong Kong, taking into consideration potential confounding by known risk factors.

Methods: The mortality of oesophageal cancer was investigated for both caisson (heavy silica dust exposure and radon exposure) and non-caisson workers (moderate silica dust exposure) during the period 1981 to 1999 in a cohort of 2789 workers with silicosis in Hong Kong. The SMR was calculated using the age specific mortality rates of Hong Kong men as the external reference. Axelson’s indirect method was used to adjust for the confounding effects of cigarette smoking, alcohol drinking, and place of origin (reflecting dietary practice), using conservative estimates.

Results: The SMR of oesophageal cancer was 4.21 (95% confidence interval 1.81 to 8.30, based on eight deaths) for the underground caisson workers, which was considerably higher than that of non-caisson workers (1.69; 0.87 to 2.95, based on 12 deaths). The SMR of oesophageal cancer among caisson workers with silicosis was reduced to 1.36 after taking into account the confounding effects from smoking, alcohol drinking, and place of origin, with an attributable fraction of 26.5% among caisson workers with silicosis. No excess mortality risk of oesophageal cancer was observed among non-caisson workers with silicosis after the adjustments.

Conclusion: Workers with silicosis who had worked in underground caissons in Hong Kong had a 36% excess risk in oesophageal cancer mortality after adjusting for smoking, alcohol, and the place of origin, which reflected dietary preference. It was possible that the increased risk was due to the direct toxic effect of silica dust associated with the heavy ingestion of silica particles after their clearance from the upper respiratory tract. Whether radon exposure played a part would require further focused investigations.


I. Mori.Department of Hazard Assessment, National Institute of Industrial Health, Kawasaki, Japan

Introduction: Vinyl chloride monomer (VCM) is well known causative agent for angiosarcoma of the liver. Although many reports have suggested increased risk for cancers of the liver other than angiosarcoma, or cancers of the lung, brain, and lymphatic and haematopoietic tissues, recent large scale epidemiological studies have revealed that more evidence is needed to conclude carcinogenicity of VCM.

Methods: The study cohort consisted of 4523 male employees who had worked for at least 1 year before 31 December 1964 in one of 25 plants that had begun VCM and/or PVC production before 1965. The observation period was from 1 January 1950 to 31 December 1996. The vital status was confirmed by the Japanese Koseki system and, for deceased cases, the underlying cause of death was coded by ICD-9. Preliminary analysis was carried out for major causes of death by proportionate mortality ratio (PMR) and/or proportionate cancer mortality ratio (PCMR). Standardised mortality ratio (SMR) was used for detailed analysis for cancer of the liver.

Results: A total of 814 deaths occurred, and 48 subjects (1.1%) were lost to follow up. Preliminary analysis suggested significant excess for all cancers combined (PMR 158; 95% confidence interval (CI) 141 to 176), and for cancers of the liver (PCMR 197; 95% CI 152 to 251) and colon (PCMR 179; 95% CI 114 to 270). Excess for cancers of the biliary tract, pancreas, and lymphatic and haematopoietic tissues were observed but were not statistically significant. Of the 65 liver cancer cases (SMR 144; 95% CI 111 to 183), 38 cases occurred in the PVC production workers (SMR 151; 95% CI 107 to 208) including high risk groups such as reactor cleaners. Only three cases were identified as angiosarcoma of the liver.

Conclusions: The results for liver cancer were consistent with previous reports and suggested increased risk for hepatocellular carcinoma, although further analysis based on detailed exposure assessment and consideration on confounding from hepatitis B/C virus infection are necessary.


J. L. Marchand, E. Imbernon, M. Goldberg.Institut de Veille Sanitaire, Département Santé Travail, St-Maurice, Val-de-Marne, France

Introduction: A cohort of workers of EDF-GDF (the French national electricity and gas company) was constituted to study mortality patterns according to occupational risks, and particularly the associations between occupational exposures to various chemicals and cancer mortality.

Methods: The “Cohort-78” comprises every person who was at EDF-GDF in 1978 or who entered the company between 1978 and 1995. Workers who had been employed by the company for at least 1 year were followed up for mortality from 1978 to 1998. The job history in the company was known for each subject, and cumulative exposure to 30 different chemicals (such as solvents, oils, resins, and fibres) was evaluated through the subject’s entire career using a specific job exposure matrix. Standardised mortality ratios (SMR) were computed to compare cancer mortality among the exposed with the French cancer mortality, then multivariate Poisson regression was used to estimate relative risks of cancer death associated with exposures. In view of the small number of exposed women, this study was restricted to men (n  =  182 490).

Results: The subjects contributed for 3 155 402 person years at risk over the period of follow up; 6629 cancer deaths were identified, among which the main cause was lung cancer. Several exposures were associated with significant excess mortality from respiratory cancers. Using Poisson regression, the known associations between asbestos, crystalline silica, coal tar pitches, and lung cancer were observed, as well as an association between asbestos and pleural cancer. Other associations were observed, such as crystalline silica and oral cancers. For other cancers, some exposures (such as chlorinated solvents) were associated with digestive cancers, but no association was observed with urinary cancers.

Conclusions: Known cancer mortality risks associated with established carcinogens were confirmed, and other potential risks were suggested. The analyses were systematic, and some of these associations will be further investigated. The “Cohort-78” is the largest occupational cohort in France to our knowledge, and should contribute to the study of the carcinogenic effects of various chemicals to which many workers are or were exposed, both inside and outside the electricity and gas industry.