抄録
The traditional scale for toxic responses from chemicals, proposed by Gaddum, used a log scale for dose and a linear scale for response. Gaddum’s use of this scale was merely empirical. However, a log scale for dose and a linear scale for response conforms to chemical thermodynamics because the effect of a chemical on a chemical reaction is directly proportional to the logarithm of its concentration. Regulatory agencies for carcinogenesis mostly have chosen to plot dose on a linear scale without any scientific basis for this decision. This decision is totally deceptive because it obscures the effect of low doses of carcinogens to which humans usually are exposed. When experimental data on animals for chemical carcinogens are plotted on a logarithmic scale for dose and a linear scale for response, clear thresholds are apparent for carcinogenesis for most data. Furthermore, the curve for most carcinogens and sites is linear to the lowest doses that have been observed to give an increase in tumors. That is, curiously, there is no need for a probit correction. The probit correction is used to account for biological variability among individuals. This lack of a need for a probit correction has puzzled this author since his first publication on a log-linear scale for carcinogenesis data on methyl eugenol. It is proposed that the most probable explanation is that hormesis applies to chemical carcinogenesis for some data, and at doses only slightly below the threshold for carcinogenesis, there actually may be inhibition of carcinogenesis for some agents and sites.
