Abstract
The relation between occupational silica exposure and lung cancer has been addressed in numerous epidemiologic studies, especially during the past decade. Despite the considerable attention given to this topic, no scientific consensus has been reached regarding the potential carcinogenicity of silica in humans. The principal reasons for uncertainty are inadequacies of exposure assessment and incomplete control of confounding from other workplace agents and from cigarette smoking. This article is a review of the biases that have hindered the interpretation of epidemiologic research on silica and lung cancer. Exposure assessment is discussed in terms of the completeness and validity of exposure quantification for silica and other potentially relevant workplace exposures. The advantages and limitations of inferring exposure levels from radiographic evidence of pulmonary fibrosis are also considered. Confounding bias, from cigarette smoking and workplace exposures other than silica (e.g., radon), is examined in reference to the magnitude of effect estimate distortion that is likely to occur in epidemiologic research. Emphasis is also placed on the need to distinguish confounding from effect modification, manifested as synergy between silica and other exposures. Published studies of South African gold miners, Vermont granite workers, and California diatomaceous earth workers are used to illustrate these methodological issues. Checkoway, H.: Methodological Considerations Relevant to Epidemiology Studies of Silica and Lung Cancer. Appl. Occup. Environ. Hyg. 10(12):1049–1055; 1995.
