Dose-response studies and 'no-effect-levels' of N-nitroso compounds: some general aspects.

One major problem in the evaluation of potential carcinogenic food additives and contaminants is that of thresholds or, better, of 'no-adverse-effect-levels'. Arguments in favor of the postulated 'irreversibility' of carcinogenic effects are based on dose-response studies, single dose and multigeneration experiments as well as on the concept of somatic mutation as the first step in carcinogenesis with subsequent transmittance of induced defects during cell replication. The problem of extrapolation of results of animal experiments using high doses to low exposure and low incidences in man is not yet solved satisfactorily. Possible practical consequences include zero tolerance, acceptable thresholds at low risk and safety factors. Acceptable intakes should never be considered constants but should be changeable as soon as new facts in regard to the safety evaluation are available.