again and prospective studies initiated, especially since recent legislation has led to major improvements in mining conditions. It should not be automatically assumed that the exposures experienced in mining will increase the cancer incidence rate. Significant evidence from the United Kingdom shows that lung cancer is diminished for unknown reasons in underground coal miners.20
In the case of uranium miners, a lung cancer mortality rate of about 0.2 per GWe-plant-year can be estimated from the doses and cancer-induction factors given later in this chapter (1400 person-rem×(2×10−4) cancers per person-rem).*
In dealing with the quantitative assessment of risk due to emissions and wastes, the dose-effect curve is used whenever feasible. For the discussions that follow, it is important to consider the advantages and limitations of the method, particularly in the examples detailed below.
Two kinds of toxic agents are generated in energy cycles, artificially radioactive elements (fission products or elements activated by neutron adsorption) whose half-lives may be short or very long, and chemicals. Much more is known (for present purposes) about the mode of action of the ionizing radiations than that of the chemical agents at the low levels of dosage that are of primary concern. It appears that radiation primarily induces late effects (e.g., cancer), whereas the chemicals produce more immediate effects, although they might produce late effects as well.
Ionizing Radiations The systematic study of the Japanese atomic bomb survivors has provided an outstanding example of the knowledge that can be gained from a large-scale epidemiological study, and the time and effort required to achieve it.
Results for leukemia are illustrated by the dose-effect curves in Figure 9–1, for Hiroshima and Nagasaki. The death rate (mean annual rate for the period 1950–1974) is plotted against radiation dose (per person) received in 1945.21 For the entire study period, there was a crude excess of 67 cases of leukemia in the total population of about 83,000 persons. The curves illustrate the great sensitivity of the epidemiological method at its best, when a specific class of rare disease due to a specific cause is fully