Radar commentary: Use of linear no-threshold hypothesis in radiation protection regulation in the United States.

Radiation protection recommendations advanced by the International Commission on Radiological Protection and National Council on Radiation Protection and Measurements, and many times adopted into regulations by the United States Nuclear Regulatory Commission, need to be based on scientifically justified assumptions and conclusions. The linear no-threshold model assigns risk to every radiation exposure above zero dose and is the current basis for setting radiation protection standards worldwide. This hypothesis is vigorously challenged by many individuals but just as vigorously defended in spite of the uncertainties surrounding health effects at low dose levels. It is clear that at radiation doses below 100 mSv, the effects, if any, are so low as to be unobservable and perhaps, therefore, unknowable. However, the linear no-threshold hypothesis is used routinely to formulate regulatory dose limits for workers and the general public and to derive stochastic radiogenic risk estimates at low doses. This note will show that while the linear no-threshold hypothesis may play a legitimate role in setting radiation protection standards and operating policies, such as establishing dose limits or as part of an operational "as low as is reasonably achievable" (ALARA) policy, it is inappropriate for use in estimating possible cancer risks associated with low-level radiation exposures. It will also demonstrate that the raising, not lowering, of current regulatory dose limits is more solidly supported by the actual observed data on radiation dose and effects. The authors submit that the misuse of the linear no-threshold model for predicting radiation effects in exposed individuals and populations should be discontinued.