Human Health Risk Assessment - Dealing with Contaminated Sites: From Theory Towards Practical Application - page 235

than the LOAEL, the NOAEL is a conservative estimate for the NAEL, while the
LOAEL is a non-conservative estimate for the NAEL.
In general, there are two options for using experimental animal data as a bench-
mark in Human Health Risk Assessment. The first is comparing actual human
exposures to the NAEL, NOAEL, or the LOAEL from animal studies. Depending
on the appropriate margin between exposure and effect dose (MOE: margin of
exposure), the contamination at the site will receive a risk classification. However,
one should make an a priori decision about that value of the MOE which would
represent an acceptable health risk for the contaminant considered, as in Yoshida
et al. ( 2000 ) who evaluated the human health risks of different population groups
exposed to dioxins, in Japan. This decision will relate to the extrapolation from ani-
mals to humans, the variability within the human population and the quality of the
database. A priori, a decision on the MOE is needed not only for its use in soil qual-
ity assessment, but also for consistent Risk Assessment practices. When the actual
exposure exceeds these effects measurements, the risk associated with the contam-
inated sites usually is labelled as 'unacceptable human health risk'. The European
Food Safety Agency recommends the use of a margin of exposure (MOE) approach
for non-genotoxic carcinogens (EFSA 2005 ).
The second option is more practical and relates to the transformation of the
NAEL, NOAEL, or the LOAEL into effects measurements applicable to humans
through the use of assessment factors . In a more advanced way, the full dose-
response data are used to derive a benchmark dose (BMD). The BMD corresponds
to a certain level of effect (e.g., 10%). The 95% lower confidence interval (BMDL)
is then used as an NOAEL (International Programme on Chemical Safety 2010 ).
Different names are used internationally for the resulting Toxicological
Reference Value, for example, Benchmark exposure, Tolerable Daily Intake (TDI),
Acceptable Daily Intake (ADI), Reference dose (RfD), Toxicity reference value
(TRV), Reference concentration (RfC), TCA (Tolerable Concentration in Air) and
Predicted No Adverse Effect Level (PNEL).
5.4.3.2 Assessment Factors
Assessment factors are used to make effect data on test animals applicable to effect
data for human beings. These assessment factors are often assumed to include
a whole scale of phenomena such as conversion, extrapolation, adjustment and
uncertainty (Vermeire et al. 1999 ). Use of the overall assessment factor is aimed
at converting a threshold dose measured in experimental animals to a safe dose
in the human population, including sensitive individuals. Conversion, for exam-
ple, relates to interspecies differences, that is, differences between the test animals
and human beings such as differences in body size and weight, physiology (e.g.,
related to the bioavailability in the body of organisms and humans) and suscepti-
bility. Extrapolation covers the variation in exposure time between the test animals
(often subchronic) and human beings (e.g., lifelong), the differences in exposure
conditions between the test animals and human beings, the difference between
a high dose and a much lower dose, and the exposure pattern (e.g., continuous