of subgroups that received different doses. The results of epidemiologic studies examining lung cancer mortality (Table 4.9) are discussed below.
The study by Archer and colleagues (1973a) was one of the first to examine cancer mortality as a result of exposure to uranium in a setting other than the enclosed, poorly ventilated spaces in an underground mine. There were 104 deaths during the 17-year period (1950–1967) of follow-up of uranium mill workers. The number of lung cancer deaths was 4 (4.26 expected) for an SMR of 94. The small cohort, short period of follow-up (maximum of 17 years), and small number of cancer patients limit this study’s power to detect an increase in lung cancer deaths. In a somewhat larger cohort with up to 37 years of follow-up, Waxweiler and colleagues (1983) found fewer lung cancer deaths than expected (SMR = 83).
The SMR for lung cancer in this study was 113 (90% CI 61–192). There was no trend to higher SMRs for lung cancer with increasing duration of employment (a proxy for dose of radiation) or length of follow-up. A trend toward higher SMRs with increasing duration of follow-up would be consistent with a biological effect of exposure because a longer period of follow-up reduces the influence of cancers detected during the first 10 years after employment. Most cancers detected in the first 10 years after exposure are not likely to be causally related to the exposure. The single exception was a trend for higher SMRs in black male employees who had been employed more than 20 years and followed over 20 years. There were only two cases of cancer in this category.
The study had several shortcomings. For most employees, the period of potential exposure was relatively brief. Fifty-four percent of the employees worked less than 6 months at the plant. Most employees were young, with 61 percent being between 18 and 28 years at the date of hire. There was no information on cigarette smoking status. The length of follow-up was too brief and the cohort still too young to accumulate enough deaths to draw any firm conclusions about the association between exposure and lung cancer. Additionally, the authors did not have precise information about job assignments and had no information about the level of radiation exposure. Therefore, workers could not be classified by the extent of their exposure. Further, the workers may have been exposed to other potentially hazardous compounds including radon daughters.