Common misinterpretations of the "linear, no-threshold" relationship used in radiation protection.

Absorbed dose D is shown to be a composite variable, the product of the fraction of cells hit (IH) and the mean "dose" (hit size) z to those cells. D is suitable for use with high level exposure (HLE) to radiation and its resulting acute organ effects because, since IH = 1.0, it approximates closely enough the mean energy density in the cell as well as in the organ. However, with low level exposure (LLE) to radiation and its consequent probability of cancer induction from a single cell, stochastic delivery of energy to cells results in a wide distribution of hit sizes z, and the expected mean value, z, is constant with exposure. Thus, with LLE, only IH varies with D so that the apparent proportionality between "dose" and the fraction of cells transformed is misleading. This proportionality therefore does not mean that any (cell) dose, no matter how small, can be lethal. Rather, it means that, in the exposure of a population of individual organisms consisting of the constituent relevant cells, there is a small probability of particle-cell interactions which transfer energy. The probability of a cell transforming and initiating a cancer can only be greater than zero if the hit size ("dose") to the cell is large enough. Otherwise stated, if the "dose" is defined at the proper level of biological organization, namely, the cell and not the organ, only a large dose z to that cell is effective.