Epidemiological studies of active smokers have shown that the duration of smoking has a much greater effect on lung cancer risk than the amount smoked. This observation suggests that passive smoking might be much more harmful than would be predicted from measures of the level of exposure alone, as it is often of very long duration frequently beginning in early childhood. In this paper we have investigated this using a multistage model with five stages. The model is shown to provide an excellent fit to data on the incidence of lung cancer among smokers, ex-smokers and non-smokers in a cohort of male British doctors. Contrary to our expectation the model predicted only a slight increase in relative risk with increasing duration of passive exposure. Allowing for exposures early in life does not therefore explain the discrepancy between the relative risk of about 1.5 calculated from epidemiological studies of lung cancer and the low levels of exposure indicated by cotinine measurements in those passively exposed.


Journal article


Br J Cancer

Publication Date





825 - 831


Aged, Humans, Lung Neoplasms, Mathematics, Models, Biological, Prospective Studies, Risk Factors, Smoking, Tobacco Smoke Pollution