Abstract
Linear non-threshold (LNT) model is a basic theory for radioprotection, but the adaptability of this hypothesis to the risk at very low doses or at very low dose rates is not sufficiently investigated. In this study a method for quantitative risk assessment in epidemiological studies investigating dose rate effects was proposed. For describing quantitatively the association between risk and dose rates, we applied some statistical models to experimental data at low dose with statistically sufficient accuracy. Data on micronucleus formation and [3H] thymidine uptake in human cells were analyzed using the Median Effective Dose (MED) as a measure of the risk. In this experimental data, it was obvious that mathematically estimated MED increased with longer irradiation time and that the biological response depends on not only cumulative dose but also irradiation time. We should point out that there is a dose rate range over which risk depends on both dose and dose rate, and that risk depends primarily on dose rate at very low dose rate. Furthermore the proposed approach was applied to some epidemiological studies that have been already reported. It was suggested that the model can describe the relationship between the risk and dose rate under some hypotheses in epidemiological studies. One of the advantages of the model is that we can subsequently test predictions for other experimental or epidemiological data because the model can describe a wide dose-rate range extending from background to experimental level radiation.
