Mortality and Cancer Frequency Among Military Nuclear Test (Smoky) Participants

To the Editor.— In analyzing their data concerning follow-up of nuclear test participants and in drawing their conclusions, Caldwell et al1 have failed to take account of current knowledge of the epidemiology of human radiation exposure. It is known that the bone marrow has a high relative sensitivity to the induction of cancer and that leukemias may appear within five years of radiation exposure; solid tumors, however, generally appear later, and excess risk may not become apparent earlier than 15 to 20 years from exposure.1 Since the reason for undertaking this study was the report of leukemia in one participant and because leukemia is known, a priori, to be caused by exposure to ionizing radiation, it seems reasonable that the null hypothesis for this study should be that leukemia rates among test participants are the same as in the US population; this hypothesis is convincingly rejected by the