Limiting values for the RBE of fission neutrons at low doses for life shortening in mice.

We have analyzed recently published data on the effects of low doses of fission neutrons on the mean survival times of mice. The analysis for single-dose exposures was confined to doses of 20 rad or less, while for fractionated exposures only total doses of 80 rad or less were considered. We fitted the data to the frequently used power function model: life shortening = beta D lambda, where D is the radiation dose. We show that, at low doses per fraction, either (1) the effects are not additive or (2) the dose-effect curve for single exposures cannot show a greater negative curvature than about the 0.9 power of dose. Analysis of the data for gamma rays showed that an exponent of 1.0 gave an acceptable fit. Taken together, these findings indicate that the RBE for neutrons cannot change more rapidly with neutron dose than about RBEN approximately = k/D0.1N. This conflicts with the more widely accepted relationship, RBEN approximately = k/D0.5N. Because of the inherent implausibility of exponents less than 1.0 for the neutron dose-effect curves at low doses we conclude that at neutron doses of 20 rad or less the RBE for life shortening is constant and ranges from 13 to 22 depending on mouse strain and sex.