The Validity of Risk Estimates of Leukemia Incidence Based on Japanese Data

On the basis of all available pertinent radiobiological evidence it must be expected that the RBE of neutrons for the induction of human leukemia is a function of absorbed dose. An analysis of epidemiological and dosimetric data obtained at Hiroshima and Nagasaki yields results that conform to this expectation. It must, therefore, be concluded that the shapes of the dose-effect curves for the two Japanese cities are different and among the simplest assumptions that may be applied a linear dose dependence for neutrons appears to be the most plausible. Risk estimates based on a linear dependence for gamma radiation must, therefore, be excessive.