Are atomic-bomb dose-response data from ABCC/RERF reasonable for assessment of radiation risk?

Ever since ABCC was established in 1948, the Unified Program, conceived in 1955 and a fixed population sample (Life Span Study extended) was selected from the A-Bomb survivors Supplementary Schedules of 1950 National Census, originally consisted of approximately 110,000 persons in Hiroshima and Nagasaki. Since 1958, the AHS, a fixed sub-sample of LSS-extend sample, originally consisting of nearly 20,000 persons, has been followed for long-term clinical examinations for any late ionizing radiation effects of the A-bombs. AHS participants are thus provided complete physical examinations and laboratory tests during their biennial ''cycle'' visits to the ABCC/RERF clinics. The AHS sample includes persons NIC as the control groups. On the basis of the survey of the fixed population sample, ABCC/RERF have published many papers upto the present. Those data became the basis for reports of ICRP (1979, 1990), UNSCEAR (1977, 1933), and BEIR (1990). The author would like to raise a question whether the use of ABCC/RERF data was reasonable or not. Based on the cancer incidence in Hiroshima and Nagasaki A-bomb survivors, LNT model that radiation risk is always proportional to dose, no matter how small, was adopted by ICRP, UNSCEAR, and BEIR for the assessment of risk at low doses and for recommendation of dose limits. RERF reported the cancer incidence in A-bomb survivors, 1958-1987 (1994). A linear dose-response relationship was expressed for all solid cancers. In the ABCC/RERF study, however, abscissa of dose-response curves is done from A-bomb with no consideration of dose-rate in spite that the dose-rate is a great factor for the incidence. Incidence is affected not only with radiation dose but also with radiation dose-rate both of which are in inverse proportion to distance from the A-bomb explosion center. The patients were exposed to the radiation at the different dose-rate depending on the distance. Real animal experimental data from HTO administration by Yamamoto et al. (1998, 2000) showed different incidence of tumors at different dose-rates. Dose on abscissa of data from ABCC/RERF has to be corrected to (Dose)dose-ratefactor. The contaminated medical radiation dose is not included in ''dose'' in ABCC/RERF reports. Our medical X-ray dosimetry program was begum in 1962. A report has been published by Yamamoto et al. (1988) as the initial analysis for potential contamination of A-bomb doses by medical X-ray doses received by the members in the AHS sample through the end of 1982. For the lowest A-bomb exposed group, the contamination rate is 25% in average. Sixteen percent of subjects has more than 100% contamination and a few percent has 250-500% contamination. Such contamination rate for the lowest A-bomb exposed group is very serious for assessment of radiation risk. For the middle A-bomb exposed group, the contamination rate is still effective, though it is very low for the highest A-bomb dose group. The contamination rate at the present has to be much more. The medical X-ray doses should be taken into consideration for the correction at least to the subjects exposed to A-bomb dose less than 0.1 Gy. (author)