The carcinogenic effect of low doses: the validity of the linear no-threshold relationship

Abstract

The linear no-threshold relationship (LNT) has been advocated by the International Commission on Radiological Protection (ICRP) for assessing the carcinogenic effect of low dose irradiation (below 200 mSv). Initially, the main argument of its proponents was that risks were unlikely to be underestimated with LNT. Subsequently, it has been argued that: the traversal of a cell nucleus by one track can cause a double-strand break and since the probability of disrepair is never zero there is no inferior limit to the use of LNT; the repair probability of the complex DNA lesion caused by radiation is much smaller than that of the lesion caused by cell metabolism and is not much influenced by dose and dose rate; and this relationship is compatible with epidemiological and experimental data. Gradually, what was initially an hypothesis has become a dogma. The arguments of the opponents to LNT are given. In conclusion, for doses between 200 and 20 mSv the linear extrapolation could provide the superior limit of the risk although the actual risk might be much smaller. Assessment by LNT of the carcinogenic effect for doses below 10 mSv is not scientifically justified and should be banned. Doses below 1 mSv/year should not be taken into account for risk assessment.