86 CHAPTER 4 Risk Analysis and Environmental Health Hazards
Environmental contaminants are linked to many serious
health concerns, including other diseases, birth defects,
damage to the immune system, reproductive problems,
and damage to the nervous system or other body systems.
We focus here on risk assessment as it relates to cancer,
but noncancer hazards are assessed in similar ways.
The most common method of determining whether
a chemical causes cancer is to expose groups of labo-
ratory animals, such as rats, to various large doses and
count how many animals develop cancer at the different
levels. This method is indirect and uncertain, however.
For one thing, although humans and rats are both mam-
mals, they are different organisms and may respond dif-
ferently to exposure to the same chemical. (Even rats and
mice often respond differently to the same toxicant.)
Another problem is that lab rats are exposed to massive
doses of the suspected carcinogen
relative to their body size, whereas
humans are usually exposed to
much lower amounts. Researchers
must use large doses to cause can-
cer in a small group of laboratory
animals within a reasonable amount of time. Otherwise,
such tests would take years, require thousands of test ani-
mals, and be prohibitively expensive to produce enough
data to have statistically significant results.ÃiqÀiëÃiÊVÕÀÛiÃÊUÊ}ÕÀiÊ{°£ÓÊÊ
each year—as a way of estimating their toxic potential. It
is generally assumed that a chemical with a low LD 50 for
several species of test animals is also very toxic in humans.
The effective dose–50 percent, or ED 50 , measures a
wide range of biological responses, such as stunted devel-
opment in the offspring of a pregnant animal, reduced
enzyme activity, or onset of hair loss. The ED 50 is the dose
that causes 50 percent of a population to exhibit what-
ever response is under study.
To develop a dose–response curve, scientists first test
the effects of high doses and then
work their way down to a thresh-
old level, the maximum dose that
has no measurable effect (or, al-
ternatively, the minimum dose
that produces a measurable ef-
fect) (Figure 4.12). Scientists as-
sume that doses lower than the threshold level are safe.
A growing body of evidence, however, suggests that
for certain toxicants there is no safe dose. A threshold
does not exist for these chemicals, and even the smallest
amount causes a measurable response.
Cancer-Causing Substances
Because cancer is so feared, for many years it was the
main effect evaluated in chemical risk assessment.
dose–response
curve In toxicology,
a graph that shows
the effects of different
doses on a population
of test organisms.carcinogen Any
substance (for
example, chemical,
radiation, virus) that
causes cancer.0255075100Threshold
level
Percentage of population showing responseED^50Logarithm of dose (mg/kg of body weight)Low High
Logarithm of dose (mg/kg of body weight)0Percentage of population showing response255075100Low HighED 50 for AToxicant BToxicant AED 50 for B050b. Dose–response curves for two hypothetical
toxicants, A and B. In this example, A has a
lower effective dose–50 percent (ED ) than B.
However, at lower doses, B is more toxic than A.a. This hypothetical dose–response curve
demonstrates two assumptions: first, that the
biological response increases as the dose is
increased; second, that harmful responses occur
only above a certain threshold level.