induce different types of leukemia.24,25 The neutron-weighted curve (Hiroshima) resembles a linear one, whereas the other curve does not.

Sensitive though the Japanese study may be, is it sensitive enough to estimate the risk in the regulatory range of particular interest today, below 5 rads, and especially below 1 rad? This region of dose effect is poorly defined statistically by the two curves in Figure 9–1 owing to the small numbers of leukemia cases below 50 rads. In fact, the two curves are not significantly different statistically. They illustrate the uncertainty of extrapolating from higher, relatively well-established regions of the dose-effect curve to the lowest regions, which are of greatest interest. The problem of allowing for natural background radiation in the very low dose region, where effects even in very large populations have been undetectable, is discussed later in this section of chapter 9 under “Fission.”

Finally, the nature of the population at risk must be examined. Is it of uniform sensitivity or does it contain subpopulations whose response in fact determines the overall results? In the case of the Japanese data, age at exposure determined how rapidly leukemia first occurred (Figure 9–2).26 As a result, early reports concluded that children under 15 were the most sensitive, but after 10 years, disease began to occur in those who had been exposed at 45 years of age or older, and this group can now be seen to be the most affected. The occurrence of other forms of cancer only began some 15 years after exposure (in all age groups) and is still increasing after 30 years.


Chemical Agents The problem is much more complex for chemical agents, since the dose to the target tissue is known only under exceptional circumstances. Even the exposure level (concentration in air or water) may not be known quantitatively, and it is rarely specified for individuals on the basis of where they spend their time, e.g., indoors or outdoors. The parameter of exposure in air pollution studies is almost invariably a measurement taken at some distance from those at risk, for example, from a single monitoring station in a metropolitan area.

Additional confusion may result from the imprecise use of the term “dose.” Dose might refer to the ambient exposure, or as in medicine, it might specify the amount of agent entering the body (by mouth or lung) but not necessarily reaching the target organ. (This could be true of radioactive substances that emit alpha and beta particles.) Thus, the agent may be inactivated in the digestive tract or excreted; if absorbed, it may be inactivated by the blood or liver, or it may be excreted by the kidneys. All these mechanisms tend to reduce the effect per unit quantity of agent or even to establish a threshold of exposure below which there is no damage.27

However, some innocuous substances are activated in the body. In the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement