Risk Assessment - Toxicology of Metals - page 284

individual is the critical concentration in that individual.
The concept of a critical concentration is only applica-
ble to effects that occur as a result of a specifi c threshold
concentration being reached, so-called deterministic
effects or threshold effects.
When a group or population is exposed to a toxic
agent and several effects are studied, a set of dose-
response curves are achieved as displayed in Figure 2.
At high doses, severe and lethal effects occur (curve D
in Figure 2). A well-known concept in this context is
the median lethal dose LD-50. Experiments in animals
determining the dose-response curve for lethality are of
very limited value in environmental and occupational
health. Such experiments should also be limited based
on ethical considerations. At doses lower than those
causing lethality, there will be toxic effects leading to
disease (curve C, Figure 2). The median toxic dose (TD-
50) for a specifi c toxic effect that may represent a clinical
disease not leading to death is shown in Figure 2 (curve
C). At slightly lower doses TD-05, the toxic dose for 5% of
the exposed group may be identifi ed. At even lower doses,
adverse functional changes without clinical effects may
be identifi ed by biomonitoring (curve B, Figure 2). Such
subclinical changes, which can be shown to have a rela-
tionship as a precursor to clinical disease, are often
taken as critical effect. The part of this curve represent-
ing relatively low doses and low response levels (e.g.,
the level when 5% of the exposed group displays this
effect, i.e., the effect dose for 5% (ED-05), is of interest in
occupational and environmental health. When a prac-
tical threshold is reached, when no adverse (critical)
effect can be observed, the “ No-observed-adverse-
effect-level” (NOAEL) is reached (Figure 2). Because

observations are made in specifi c dose groups, NOAEL
is the highest dose level at which a group of individu-
als have been exposed without any statistically signifi -
cant adverse effects being demonstrated. At even lower
doses, there may be a dose-response curve describing
(curve A in Figure 2) slight, but statistically signifi cant,
changes in enzymes for which there is a high reserve
capacity, and a small decrease does not imply any
decrease in organ function. Such effects are named
subcritical effects. Doses below this curve, at which
no effects whatsoever occur, represent the no-observed-
effect-level (NOEL) , which is the highest dose at which
no effects at all can be observed (Figure 2).
The S-shaped dose-response curves for threshold-
type effects are explained by a Gaussian distribution in
sensitivity among persons making up the group stud-
ied. This has been further discussed and explained in
Chapter 6).
As pointed out in the foregoing text, in preventive
medicine, it is of particular importance to identify early
effects. If these can be prevented, then later, perhaps
more severe, effects can be avoided. Against this back-
ground, the terms critical concentration, critical organ,
and critical effect have been established for some met-
als (Nordberg, 1992; Task Group on Metal Accumu-
lation, 1973; Task Group on Metal Interaction, 1978;
Task Group on Metal Toxicity, 1976). These concepts
are mainly applicable to deterministic/threshold-type
effect, but the critical effect concept may also be used
when stochastic/nonthreshold effects are considered.
Further description and discussion of these concepts
and examples of their use can be found in Friberg and
Kjellstrom (1981), Kjellstrom et al . (1984), Nordberg
(1992), WHO/IPCS (1992), and Jarup et al . (1998).
Effects that occur as a result of a random interaction
of the metal compound with DNA or when they depend
on a series of events such as in multistage carcinogene-
sis are named stochastic effects or nonthreshold effects . They
are often assumed to display a linear component of the
dose-response curve at low doses (curve C in Figure 3).
Cancer and some reproductive effects are considered to
belong to this category of effects (see also Section 4.2.2).

Response
(% with effect)

100

A B C D

50

4.1.3 Benchmark Dose

The concept of benchmark dose (BMD Cr ) was intro-
duced into toxicology and risk assessment by Crump
(1984) as the statistical lower confi dence limit on the
dose producing a predetermined level of change in
adverse response (the benchmark response, or BMR)
compared with the response in untreated animals.
There are several advantages in using this concept
instead of NOAEL in risk assessment, as discussed
in Chapter 6, because the variance in the observation

5

NOAEL ED 05 TD 05 TD 50 LD 50

NOEL

Dose

FIGURE 2 Spectrum of dose-response curves for deterministic
(threshold-type) effects of a metal compound. No-observed-effect
level (NOEL), no-observed-adverse-effect level (NOAEL), effect
dose 5% (ED-05), toxic dose 5% (TD-05), toxic dose 50% (TD-50), and
lethal dose 50% (LD-50) are indicated.