Abstract
Japanese nuclear power plants (NPPs) located in 12 prefectures from Hokkaido to Kyushu islands. Cancer mortality in 1972-97 was analyzed mainly in Ibaraki prefecture where the earliest NPP was located. Standard mortality ratio (SMR) was examined for 85 municipalities which the prefecture consisted of as the end of 1997. The empirical Bayes estimate of SMR, EBSMR was also calculated for the latest 5 years period. EBSMR was less than or about 1.0 except some municipalities where it was more than 1.1 for digestive cancers. Adult T-cell leukemia (ATL) may count for 3% of leukemia deaths in the prefecture. Superficial increase after the commissioning year can be observed for cancers of lymphoid-hemopoietic tissues in females, particularly among the 75 years old or more, commonly through all municipalities with an NPP in Japan compared with selected control areas. Superficial decrease can be observed for solid cancers among the 50 years old or more. The confounding was likely to be caused by temporal changes of age specific mortality and of the number of cancer deaths in the elderly. The proportion of the latter in all cancer deaths became larger in females than in males except non-digestive solid cancers. SMR by large areas such as six blocks through Japan is relatively constant in the long period but not in the ATL endemic area for leukemia and malignant lymphoma. Estimated risk was also influenced by a short term variation in the selected municipalities; for an example, 1983-87 low mortality for non-digestive solid cancer in the area with an NPP in Ibaraki prefecture.
