Chemistry Reference
In-Depth Information
differences that need to be examined. It is hoped this
analysis will lead to further refi nement of the method
for EMs discussed in this chapter. The fact that a few
signifi cant differences in recommendations occurred
is not surprising given the fact they resulted from
the deliberations of fi ve different groups of experts
meeting over a span of 20 years.
For copper, the SRPMI (WHO, 1996), PRI (SCF,
1993), and RDA (IOM, 2001) were not markedly dif-
ferent. The variation in recommendations arises from
the fact that the SRPMI and the PRI were based on
normative and average requirement of 0.8 mg/day
and a CV in intakes of 20%. The RDA of 0.9 mg/day
was derived from an EAR of 0.7 mg/day assuming a
CV of 15%. This evaluation was made on the basis of
additional well-controlled deletion/repletion studies
not available to WHO and the SCF experts. The most
signifi cant difference in the AROIs for copper is the UL
recommended. The upper boundary of the SRPMI of
12 mg copper/day was derived from an analysis of the
database on adverse effects of copper in food, where
no effects were seen at intakes of around 10 mg/day.
The SCF (2003a) and the IOM/FNB (IOM, 2001) both
considered potential liver damage as the critical effect
and evaluated the same small double-blind study in
humans given 10 mg copper/day as a soluble sup-
plement. No adverse effects on liver function were
reported. In the professional judgement of the SCF,
an UF of 2 was required to take into account potential
variability of response within the normal population.
The IOM/FNB concluded an UF of 1 was adequate,
given the large international database indicating a lack
of adverse effects in humans consuming 10-12 mg/
day of copper in foods and the rarity of observed liver
damage from copper exposures in human populations
with normal copper homeostasis.
Iron is an example of an EM in which the intakes
within the general population are skewed by the
requirements for growth in children and adolescents,
as well as the losses in females through menstrua-
tion. Also, the bioavailability of the iron consumed
will markedly affect the determination of the AROI
boundaries. Taking these facts into account, very simi-
lar lower boundaries of the AROI were recommended
by SCF (SCF, 1993), WHO (FAO/WHO, 2002), and the
IOM/FNB (IOM, 2001). All determined the basal losses
and modeled the components of iron requirements to
determine an average requirement. Based on a CV in
losses of 15% and assuming 15% as the upper limit of
iron absorption, both WHO and the SCF recommended
9 mg iron/day as the lower boundary of the AROI. The
value of 8 mg/day, derived by the IOM/FNB results
from the assumption of 18% as the upper limit of iron
absorption. The UL of 48 mg/day recommended by
WHO in 1983 was derived by the examination of data
on the gastrointestinal effects from consumption of
soluble iron supplements by humans, whereas the UL
of 45 mg iron/day recommended by the IOM/FNB in
2001 was based on the determination of a LOAEL for
gastrointestinal distress of 70 mg iron/day (60 mg as
iron fumarate) in a controlled double-blind study and
applying an UF of 1.5 to account for extrapolation from
a LOAEL to a NOAEL.
Manganese is an example of an EM in which there
are limited data of suffi cient quality to support the
development of an AROI. As shown in Table 3, WHO
concluded that data on manganese requirements and
toxicity were not suffi cient to set a SRPMI, and the SCF
(1993) concluded it was preferable to give a safe and
acceptable range of intakes rather than a defi nitive
AROI. Given the limitations of the human data and the
fact that NOAELs are lacking for critical toxic endpoints
in animal studies (particularly neurotoxicity after oral
intake), the SCF (2000b) concluded it was not possible
to set a UL of intake for manganese. The IOM/FNB con-
cluded that balance studies for manganese were inad-
equate to set an EAR and thus a RDA. The mean intake
of manganese in adult men within the United States
is 2.1-2.3 mg/day (IOM, 2001). Because symptoms of
manganese defi ciency are not found in this population,
the recommendation of an AI of 2.3 mg manganese/day
was appropriate at this time. The UL of 11 mg/day was
set on the basis of no adverse effects found in humans
consuming Western-style diets containing up to 11 mg
manganese/day and supported by a LOAEL of 15 mg/
day reported from a study in humans where signifi cant
increases in serum manganese were found after 25 days
of supplementation. A UF of 1 applied to the NOAEL
was considered appropriate.
As shown in Table 3, the AROIs for selenium
derived by the three organizations are very simi-
lar. All three groups considered the enzyme plasma
glutathione peroxidase as an appropriate indicator
of minimal selenium defi ciency and for toxicity, the
threshold where biochemical disturbances in sele-
nium metabolism were seen in humans (850
g/day).
The differences shown can be explained by different
scientifi c judgments made with regard to whether
maximal peroxidase activity was the most appropriate
indicator of nutrient need and the level of uncertainty
in the epidemiological data that indicated a thresh-
old of toxicity at intakes of 850
µ
g selenium/day. The
lower boundary of the AROI recommended by WHO
was 40
µ
g selenium/day based on the intake required
to provide two thirds of the maximum glutathione
peroxidase activity and a 16% interindividual varia-
tion in selenium intakes. However, experts advising
SCF (2000c) and the IOM (2000) considered it more
µ
