Factors that can modify the lung cancer risk of Colorado plateau uranium (7440611) miners were examined. The cohort was an established group of 3,347 white males employed in uranium mines on the Colorado plateau who had been followed previously for mortality from 1950 through the end of 1982. The mean year of birth of the cohort was 1920 and the average age of the miners at first hire was 32 years (yr). The miners had spent a mean of 61 months (mo) underground. Their mean exposure to radon daughters was 822 working level mo (WLM). The distribution of radon daughter exposures of the miners was highly skewed, the median exposure being 430WLM and the median exposure rate being 10WLM. For this analysis, the cohort was followed to 31 December 1990. Exposure response models were fit to the lung cancer mortality data that incorporated potential risk factors such as smoking, age, exposure, and cumulative exposure to radon daughters. A total of 1,699 deaths occurred in the cohort by 31 December 1990. Of these, 377 were from lung cancer. The relative risk of lung cancer increased with cumulative radon daughter exposure; however, age at first exposure was no longer a significant risk factor as was observed earlier. Time since last exposure and exposure and exposure rate showed a strong negative effect on lung cancer risk. The lung cancer risk also decreased significantly with increasing age. A miner under 60yr of age had more than 10 times the excess relative risk per unit radon daughter exposure than a miner with the same cumulative exposure who was more than 70yr of age, 1.107 versus 0.97. Smoking and radon daughter exposure interacted in a submultiplicative manner as a lung cancer risk factor. Neither smoking alone nor the smoking/radon daughter exposure interaction showed a significant age dependence. The authors conclude that this most recent update of the lung cancer risk in this group of uranium miners suggests that summary estimates of lung cancer risk as a function of the level of exposure to radon daughters should not be reported. This analysis and other studies suggest that lung cancer excess risk estimates are strongly age dependent and also must be corrected for exposure rate, time since last exposure, and cigarette smoking history.
National Institute for Occupational Safety and Health, Division of Surveillance, Cincinnati, OH 45226