The influence of follow-up on DS02 low-dose ranges with a significant excess relative risk of all solid cancer in the Japanese A-bomb survivors.

Determinations of the lowest colon dose, D min, below which there is a statistically significant excess relative risk of all solid cancer, when analyses are restricted to the range [0, D min], are of current interest in research related to radiation protection and risk assessment. In reviewing recent cancer mortality reports on the Life Span Study (LSS) of Japanese A-bomb survivors, reported D min values were found to vary between different reports. The report 12 (follow-up: 1950-1990) found a D min of 50 mGy, but the most recent report 14 (follow-up: 1950-2003) found a D min of 200 mGy. There were small dosimetry changes between report 12, which used DS86, and report 14, which used DS02, but these changes are unlikely to account for a difference in D min of a factor of 4. This short communication examines the reasons for this difference in D min by presenting further investigations into D min using different trial values for D min and various follow-up time spans, all with the same DS02 dosimetry. Magnitudes of the low-dose risks in different dose ranges are also presented. It is shown here that the main influence on D min comes from the length of follow-up and a D min of 50 mGy may also be obtained with the most recent LSS mortality data and DS02, if a restricted follow-up is analyzed. A systematic trend was evident of lower D min values for earlier mortality follow-up periods, consistent with information from earlier LSS reports. Although it may seem surprising that the D min increases with longer follow-up and better statistics, this systematic trend appears to be a consequence of decreasing mortality risks with longer follow-up, even though the error bars on the risks are getting smaller with increasing follow-up. These systematic trends also persisted after accounting for differences between baseline cancer rates for two groups of survivors who were either proximal or distal to the A-bomb hypocenter. Similar systematic trends, although much less pronounced, were also found in the LSS cancer incidence data. Some evidence is provided here that results on low-dose radiation risks from earlier follow-up periods should not be ignored by radiation protection authorities, once the results from the new extended follow-ups are published. This is because of the possibility that the new data for extended follow-up beyond a certain calendar time, which pertain to very long times since exposure, may be contributing to an overall reduction in radiation related risks per unit dose compared to analogous risks determined from earlier follow-up periods, because of the risk effect modification of time since exposure.