Abstract

OBJECTIVES--To design a job-exposure matrix for epidemiological studies of men who had worked in a Norwegian aluminium smelter between 1922 and 1975. METHODS--Jobs held by cohort members were identified from personnel records. Tasks and their locations were determined for all jobs, and information was gathered about changes in exposure conditions over time. The jobs were combined into categories thought to have experienced similar exposure conditions, and time weighted average exposures were estimated on a relative scale. The results were reviewed by a panel of former smelter employees and an experienced industrial hygienist. RESULTS--96 different jobs could be identified from the cohort members' work histories. These were grouped into 18 categories, and relative exposure intensities were estimated for 31 different combinations of category and period. The most prevalent exposure in the cohort was pot emissions (fluorides, sulphur dioxide, and carbon monoxide; 74% ever exposed), followed by magnetic fields and heat stress (65-68%), asbestos (40%), and coal tar pitch volatiles (33%). CONCLUSIONS--Although the use of this job-exposure matrix in the subsequent epidemiological studies may result in some misclassification of exposure, this is unlikely to seriously attenuate true risks in a stratified analysis based on cumulative exposure.