The Dose Makes the Poison

All things are poison, and nothing is without poison; only the dose permits something not to be poisonous.

— Paracelsus

When I was in elementary school, conversations on the playground often took a fatalistic turn. Perhaps it was just an echo of the Cold War era, but I can recall chatting with my school chums about chemical compounds and death. We'd exclaim, "If you breathe too hard, it could kill you," or "You could drink so much water that you would die!"

Today, the concern is less about the lethal quantity of these relatively benign substances, but rather about pollutants in our food, water, and air. Yet despite our lack of sophistication, my young friends and I weren't actually so far off the mark. We didn't realize it at the time, but we were channeling a sixteenth-century physician, Paracelsus. Considered the father of toxicology, Paracelsus is credited with the first and most important tenet of the field, the idea that the dose makes the poison: "All things are poison and nothing is without poison; only the dose makes a thing not a poison." In other words, seemingly benign substances like water as well as obviously dangerous ones like arsenic can be deadly when administered in excess.

Paracelsus's groundbreaking idea centers on the dose–response relationship: the fact that in most cases the greater the dose, the greater the adverse, or toxic, response. While humble in its simplicity, the concept provides a thematic platform upon which modern regulatory toxicology is based. Furthermore, the relationship is actually more interesting than it would first appear, as both dose and response are surprisingly nuanced.

When a chemical, toxic or benign, contacts a biological organism, the contact is known as an exposure. The exposure dose is the quantitative amount of a chemical that a person (wittingly or unwittingly) is exposed to, and this quantity can be either directly or indirectly measured. For common chemicals that are deliberately administered, such as pharmaceuticals, the route of administration is direct, and generally occurs via oral consumption or injection. For exposures of this type, the dose is generally given in terms of the mass (in grams, g, or milligrams, mg) of the chemical being administered. For example, a regular-strength aspirin pill, one of most commonly consumed pharmaceuticals, contains 325 mg of the active ingredient, acetylsalicylic acid. The tablet also contains a number of other inert chemicals, but the dose refers to the amount of the active ingredient. For injections, the dosage is expressed in the same way. An epinephrine auto-injector, for example, widely self-administered by individuals with food allergies, will administer a dose of 0.3 mg of epinephrine to the individual despite the fact that the injected solution contains other chemical compounds.

In the examples given above, the exposure route is direct and easily quantifiable, but what if, on the other hand, the exposure is indirect? Indirect exposures would include the exposure that results when a fish ventilates contaminated water across its gills, or a person inhales secondhand smoke into their lungs. In these cases, the quantitative dose of the chemical exposure is much less certain, and much more difficult to measure. Rather than determining the exposure dose, it is far easier to quantify the concentration of the chemical in the "environment" (the water that the fish is ventilating, or the air that the animal or person is breathing). Furthermore, since the amount of the compound that the animal ventilates or inhales is not known exactly, the exposure cannot be quantified in terms of mass, but rather is quantified in terms of its concentration (the amount of chemical found in a specific volume of air or water) in the local environment.

Regardless of the direct or indirect source of the exposure, the response of an animal to a chemical exposure is also generally expressed in one of two broad categories, either discrete or continuous. Organism death is the ultimate discrete response, in that animals can only be found in one of two states, dead or alive. While perhaps somewhat gruesome, death provides a very valuable (and oft-times used) endpoint for toxicological studies. In contrast, variable responses to an exposure can also occur. For example, the impairment of cognition due to alcohol consumption is a classic example of a continuous variable. The response to alcohol is not all-or-none, but rather increases in its impact as the administered dose increases. This is also true for other types of toxicological impairment, such as changes in genetic expression or alterations in the activity of proteins.

Interestingly, the way that an exposure dose is expressed, whether indirect or direct, and the way that the response is measured, whether discrete or continuous, do not affect the overall shape of the dose–response relationship. In the majority of cases, the shape of the dose–response curve remains sacrosanct regardless how the dose and response data are represented within it.

Quantifying the Dose–Response Relationship

The dose–response relationship is a very powerful tool, frequently used by regulatory agencies. A common approach used to test new chemicals, or chemicals used in novel ways, begins with the generation of dose–response relationships. Generally, the first battery of toxicity testing evaluates the capacity of a chemical to produce the discrete endpoint, death, which is exacted upon a population of experimental laboratory animals, such as mice.

A dose–response curve does not really focus upon death, but rather mortality. Death is the response of an individual organism, and clearly each individual can be in only one of two states: dead or alive. In contrast to death, mortality is the response of a population of individuals. The mortality rate describes the proportion of a population that dies in response to a calamitous exposure to toxic chemicals. To graphically illustrate the mortality of a group of animals that are exposed to the same dose of a toxic compound, we use the discrete dose–response curve. At one extreme of the toxicology curve, animals exposed to low doses survive (mortality rate is zero), whereas at the other extreme all of the animals exposed to higher doses of a chemical die...