Failures and Critique of the Beir III Lung Cancer Risk Estimates

Publisher Website
Google Scholar
Abstract
It is shown that the BEIR III prediction of lung cancer incidence due to environmental radon is more than twice the actual incidence of all lung cancers among non-smokers. Histological evidence from autopsy studies of radiation victims and of non-smokers is reviewed, and it is shown that no more than 10% of all lung cancers in non-smokers can be due to radon. When alternative causes of lung cancer are considered, it is concluded that the BEIR III estimates of lung cancer risk due to environmental radon are too high for non-smokers by at least a factor of about 40. The evidence for synergism between smoking and radiation induced lung cancer is reviewed, and it is concluded that the BEIR III estimates for smokers are too high by at least a factor of about 10. The bases for the BEIR III model are reviewed, and it is concluded that it must be based almost entirely on the Czechoslovakian uranium miners; its estimated risks are higher than would be obtained from essentially all other studies. The possibility that there are other causes than radon for lung cancer among miners is discussed, and it is shown that the Czechoslovakian data are at least as consistent with the hypothesis that al! excess cancers are due to other causes (proportional to length of employment) as with the hypothesis used in BEIR III that they are all due to radon. It is concluded that other causes can easily explain all cases where lung cancer rates are high, whereas it is very difficult to explain the much lower rates for U.S. uranium miners (5 times lower than BEIR III estimates) and other similar situations. It is pointed out that the BEIR III lung cancer model is strongly supra-linear (i.e. concave downward), contrary to the bulk of experimental and theoretical evidence on that question, and contrary to other statements in the BEIR III Report. It is concluded that the BEIR III lung cancer risk estimates are probably too high by a factor of 2–5 for miners, and because of different smoking habits, the risk to the public may be another factor of 2 lower. The remainder of the over-estimate for effects of environmental radon must be due to non-linearity in the dose-response relationship.