1 Subsequent surveys found a significant excess of papillary thyroid cancer but not of follicular, medullary, or anaplastic cancer.2,3 A straight line adequately describes the relationship between radiation dose and thyroid cancer incidence, relative risks are similar in males and females, and age at exposure substantively influences risk. Risk is highest for children exposed when younger than 10 years, and there is no significant increase in risk of thyroid cancer for those exposed after age 20 years. Radiation-induced thyroid cancers are rarely fatal, but the risk per unit dose following exposure in childhood is higher than for any other radiation-induced malignancy. 4 Radiation-induced thyroid cancer has been extensively studied in human populations other than atomic bomb survivors. These populations include patients treated with radiation for malignant and nonmalignant conditions, populations exposed to radioactive fallout from nuclear weapons tests, and populations living in the vicinity of nuclear installations such as Chernobyl in the former Soviet Union and Hanford in Washington State. 4,5 The latest research from the Adult Health Survey (AHS) in Japan reported in this issue of JAMA by Imaizumi and colleagues 6 derives from a cross-sectional study of 4091 survivors of the Hiroshima and Nagasaki atomic bombings in 1945 who were invited to a special thyroid examination more than 50 years later. The prevalences of thyroid cancer (n=87), benign nodules (n = 207), and cysts (n = 324) were significantly increased and directly related to radiation dose, and the risk decreased with increasing age at exposure; however, autoimmune thyroid diseases (positive antithyroid antibodies, antithyroid antibody-positive hypothyroidism, or Graves disease) were not linked to radiation exposure.There are several new findings of note. First, it is remarkable that a biological effect from a single brief environmental exposure nearly 60 years in the past is still present and can be detected. The radiation doses (mean, 45 cGy) were related to the distance from the hypocenters, and enormous effort was required to accurately estimate exposures for individuals. 7 The existence of dose-response relationships strongly supports the authors' assertion of a radiationrelated excess of thyroid neoplasms nearly 60 years after exposure. Second, using highly sensitive assays to detect serum levels of antithyroid antibodies and thyroid-stimulating hormone, this study was not able to confirm the findings of a smaller but similar investigation of Nagasaki atomic bomb survivors that purported to show an association between radiation and autoimmune thyroid disease. 8 The absence of a dose-response relationship for any measure of autoimmune disease in the study by Imaizumi et al is consistent with earlier studies of atomic bomb survivors 6 and a recent study of persons exposed as children to iodine 131 releases from the Hanford nuclear site.
9The study of atomic bomb survivors remains the single most important study of radiation effects in huma...