Definition of exposure in epidemiological studies on foundry workers.

Epidemiological analytical studies investigating lung cancer risk among foundry workers are reviewed in relation to the available information on exposure. Nine studies (eight historical cohorts and one case-control) have been included. Overall, the information on exposure was poor. In none of the studies were there industrial hygiene measurements for investigating dose-response relationships, while duration of employment was the most common proxy of exposure utilized. None of the studies made an attempt to construct job-exposure matrices. The interpretation of the evidence in causal terms appears mainly based on the consistency across studies of the increased lung cancer mortality rather than on the relationship between the exposure in the working environment and the excess mortality reported.