Measurement of Dose-Response Relationships

It should be clear from the earlier discussion that the measurable endpoint of toxicity could be a biochemical, physiological, or pathological change. This toxic effect will show a graded 7 increase as the dose of toxicant increases. Alternatively, the toxic effect may be an all-or-none effect such as death or the presence or absence of a tumor (which can be considered in such a way). These are also called quantal effects. In this case, an increase in the dose will result in an increase in the proportion of individuals showing the response rather than an increase in the magnitude of the effect. 

Therefore, we can identify two types of relationship with the dose of the toxicant a dose-effect relationship (graded effect) and a dose-response relationship (all-or-none effect). However, the term dose-response relationship is often used to describe both types. 

With graded effects, as the dose increases, the effect, such as inhibition of an enzyme, increases from zero to maximum. This results in a sigmoid curve when plotted (Fig. 2.4). There is, therefore, a threshold dose below which there is no effect but above which an effect is detectable. Clearly, there will also be maximal effect above which no further change is possible it is impossible to inhibit an enzyme more than 100%, for instance. 

However, the quantal type of relationship with dose will also show a sigmoid curve when appropriately plotted. In this case, the curve derives from a frequency distribution (Fig. 2.5), which is the familiar Gaussian curve. 

Those animals or patients responding at the lowest doses (Fig. 2.5) are more sensitive (hypersensitive) and those responding at the highest doses are less sensitive than the average (hyposensitive). The median point of the distribution is the dose where 50% of the population has responded and is the midpoint of the dose-response curve (Fig. 2.6). If the frequency distribution of the response is plotted cumulatively, this translates into a sigmoid curve. The more perfect the Gaussian curve, the closer to a true sigmoid curve will the dose-response curve be. 

---