Smoking and Occupational Lung Disease Epidemiology

Abstract

Epidemiology studies of lung cancer which may be the result of workplace exposure contrast the risk ratios or SMRs of the exposed population with a control group. If a dose-response relationship can be detected then this suggests a causal relationship between the exposure and its effect. Smoking is clearly a substantial confound ing factor but unfortunately adjustment for this is often limited by lack of accurate data. Most commonly only 'ever' versus 'never' smokers numbers are available and even then this data may only be available for some of the cohort. In the various studies of silica and lung cancer it is important that the smoking history of both silica- exposed and control groups is known. If the prevalence of smoking was similar in the two groups this would remove the confounding, but differences in the groups are usually present. Since the SMRs from a number of studies of silica and lung cancer are about 1.5 it is sug gested that these might be close to smoking-adjusted SMRs. By contrast, where the SMR is >2.0, smoking adjustments are likely to affect the magnitude but not the statistical significance of the result.