What is a 'low dose' of radiation?

Although the expression of radiation-induced biological effects and responses may be at either the cell, organ or organism level, induction of some of these phenomena (e.g. cancer of clastogenic and genetic effects) can have their origin in the interaction of a single charged particle with the target-containing volume (TCV) of the cell, e.g. the cell nucleus. However, the independent variable now used in both organ and cell population studies, the absorbed dose to the organ, provides no information directly on particle-TCV interactions. Even if calculated as a mean to an organized population of cells, the absorbed dose becomes a composite and confounded quantity, (FzN), in which F is the fraction of TCVs 'hit' by a particle during a given exposure, z is the mean value of z1, the energy absorbed in the TCV in a single hit, and N is the mean number of hits per affected TCV. Scientific precepts demand the avoidance of such confounded variables by achieving their isolation. The needed separation can be effected by the use of microdosimetric techniques, which make it possible to hold one component quantity constant while the others are varied. As an example, low-level radiation exposure (LLE) can be used to hold F at a constant value of 0.2 where, on average, there is but one hit per TCV. The probability of a cellular quantal response, as a function of z1 only, can then be determined by use of LLE to radiations covering a wide span of LETs. Conversely, the effect of varying only the fraction of cells hit can be studied by holding z constant. This can be accomplished by working within a narrow band of LET, but only in the LLE range. The effectiveness of preirradiation altering cell sensitivity as a function of the number of hits per TCV can be determined by working within, and somewhat above, the LLE range. In either risk assessment or the application of radiation as a pretreatment, minimal consequences can be assured only if very low-level exposure is employed in order that F will be small, and if the exposure is in a field of radiation of very low LET so that z1 will be as small as possible. That is to say, exposure conditions with low consequences cannot be specified in terms of any single quantity.