Age specific interactions between smoking and radon among United States uranium miners.

United States uranium miners who smoked have death rates from lung cancer that are intermediate between the rates predicted by the additive and multiplicative models (on a ratio scale) across all age groups. Age specific patterns of interaction have not been thoroughly examined, and most analyses have been internal ones in which there was no truly non-exposed group. Here age specific death rates of lung cancer among ever smoking uranium miners have been examined for conformity with the additive and multiplicative models. The multiplicative model fits well for the youngest and oldest categories, but poorly for the middle age ranges. In the middle age range, predicted rates under the multiplicative model were quite high, surpassing the corresponding United States death rates for all causes combined. If the multiplicative model is assumed to hold across all ages, one hypothesis that might explain the observed age specific patterns is that the full expression on the multiplicative model might not be seen at certain ages due to a limited pool of miners susceptible to lung cancer. These data, however, have several limitations such as small numbers of deaths from lung cancer among never smokers, the use of qualitative rather than quantitative smoking and radon exposure data, and ignorance of the underlying biological mechanisms of interaction.