GASTROENTEROLOGY 2009;137:S29 –S35
Manganese in Parenteral Nutrition: Who, When, and Why Should We
Supplement?
GIL HARDY
Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand
Micronutrient requirements are not fully under-
stood. Parenteral nutrition (PN) usually contains the
trace element (TE) manganese (Mn) from fixed-con-
centration TE supplements. Multiple TE formula-
tions may not be optimal in pediatric and home PN.
Moreover, most PN products contain Mn as a ubiq-
uitous contaminant. Excessive Mn can lead to Parkin-
son-like symptoms resulting from hypermanga-
nesemia. A survey of 40 Australasian hospitals that
contributed data on 108 patients to the annual home
PN register and a systematic review of the literature
were conducted to establish the scope of the potential
problem of Mn toxicity in PN patients. Exposure to
Mn doses 5– 6 times current daily requirements, to-
gether with the TE contamination that is reported in
PN products, can lead to neurotoxicity. Whole-blood
levels are more accurate for monitoring and correlate
well with signal intensity of magnetic resonance im-
aging. Current TE formulations restrict prescribing
options. The regulatory mechanisms of Mn ho-
meostasis are bypassed via the parenteral route so
elimination via the hepatobiliary system is impaired,
resulting in tissue or brain accumulation. Published
dosage recommendations may be excessive and offi-
cial guidelines require revision. Variability in clinical
practices necessitates that individual TE additives are
more widely available and multiple TE products re-
formulated. More frequent monitoring for any brain
accumulation is recommended. The scarcity of PN-
associated Mn deficiency, plus the growing evidence
for Mn toxicity, leads to the conclusion that it is
unnecessary for Mn to be prescribed routinely for
pediatric or long-term PN patients.
M icronutrient requirements for parenteral nutrition
(PN) are not well understood and guidelines for
supplementation are outdated. PN admixtures usually
contain manganese (Mn) as part of a fixed-concentration
trace element (TE) supplement, but current TE formula-
tions restrict prescribing options. As home PN (HPN)
becomes used more routinely, commercial supplements
of multiple TE in fixed formulations may not be suitable
for long-term use. Excessive doses of TE such as Mn have
been associated with liver cholestasis and can lead to
symptoms that include insomnia, headache, increased
forgetfulness, anxiety, rapid hand movements, and loss of
coordination resembling Parkinson’s disease. Our recent
survey of 40 Australian and New Zealand hospitals con-
tributing to the Australasian Society of Parenteral and
Enteral Nutrition Annual HPN Register,1 together with
our own clinical experiences with HPN over many years
and a systematic review of the literature, confirm Mn
toxicity is a potential problem with adult and pediatric
PN patients.
Metabolic Functions of Manganese
Mn is considered to be an essential trace element
required as a catalytic cofactor for a variety of important
enzymatic reactions. An average adult body has approx-
imately 10 –12 mg Mn incorporated into the active center
of the various metalloenzymes; arginase, glutamine syn-
thetase, and most notably Mn superoxide dismutase
(SOD), which are located mostly in the mitochondria.2
Routine administration of Mn and other TEs in PN
admixtures is recommended by most authoritative guide-
lines but there is now growing concern that the fixed-
dose multi-TE supplements for neonatal/pediatric pa-
tients and adults on long-term HPN (⬎30 days) may lead
to toxicity from chronic overexposure. In the past 2
decades there have been numerous incidents of Mn tox-
icity. More than 50% of HPN patients may have increased
blood levels leading to hypermanganesemia, which often
is associated with clinically significant cerebral and he-
patic complications.3 Notwithstanding its essentiality for
important enzyme functions, exposure to high levels of
Mn can lead to neurotoxicity.
Induction of oxidative stress by free radicals has been
implicated as a causative factor of neurotoxic damage
associated with exposure to Mn (and other TEs). It is well
known that Mn accumulates in astrocytes, and in vitro
experiments with astrocyte cultures have shown that pre-
treatment with Mn inhibits the uptake of glutamine, and
Abbreviations used in this paper: ASPEN, American Society of Par-
enteral and Enteral Nutrition; AuSPEN, Australasian Society of Paren-
teral and Enteral Nutrition; ESPEN, European Society for Clinical Nu-
trition and Metabolism; HPN, home parenteral nutrition; Mn,
manganese; MRI, magnetic resonance imaging; PN, parenteral nutri-
tion; SOD, superoxide dismutase; TE, trace element.
© 2009 by the AGA Institute
0016-5085/09/$36.00
doi:10.1053/j.gastro.2009.08.011
S30 GIL HARDY
hence expression of the messenger RNA coding glu-
tamine transporters. A fascinating theory by Aschner et
al2 that requires further investigation proposes that al-
terations in glutamine/glutamate cycling in astrocytes,
induced by oxidative stress, may be a key mechanism as
to how Mn exerts its neurotoxicity.4
Mn absorption from the gastrointestinal tract inversely
correlates with dietary content. Absorption in neonates
and children is greater than in adults, and females appear
to absorb more than males. In healthy individuals less
than 5% of orally ingested Mn is absorbed. This main-
tains Mn homeostasis so that high dietary intake is not a
problem in the general population.5 Thus, the amount of
Mn in the diet influences the amount absorbed, as well as
its elimination in the bile. Once absorbed, Mn is trans-
ported to the liver where a small proportion is bound to
transferrin but the majority is bound to ␥-globulin and
albumin. The main pathway for Mn excretion is via the
hepatobiliary system: in the liver Mn is removed from the
blood and then conjugated with bile, which is secreted
into the intestine where a small fraction of Mn is reab-
sorbed and the remainder is excreted in the feces. More
than 90% is excreted via the bile but biliary excretion is
poor in neonates and may contribute to increased deliv-
ery of Mn to the brain and other tissues. Low bile excre-
tion, therefore, can increase the potential for toxicity.
Hepatic dysfunction and cholestasis are suspected risk
factors for increased Mn accumulation in the brain,2
whereas patients with biliary atresia display hyperman-
ganesemia without any increase in dietary Mn intake.6
Parenteral Nutrition, Cholestasis, and
Manganese
In PN patients the normal intestinal regulatory
mechanism is bypassed and the amount of Mn delivered
via the intravenous route is 100% bioavailable. In addi-
tion, the normal pathway of elimination via the hepato-
biliary system frequently is impaired because of PN-asso-
ciated biliary stasis and obstructive jaundice. This may be
especially important for parenterally fed infants who pass
little or no stool and often show evidence of hepatic
dysfunction and cholestasis.2 It also predisposes long-
term PN patients to tissue accumulation and/or brain
deposition of Mn, resulting in neurologic symptoms.
However, a clear cause– effect relationship between PN-
associated cholestasis and neurotoxicity has not been
established and data about the temporal relationship
between the dose and duration of Mn supplementation
and increased Mn levels have been to some degree con-
tradictory.7
In Australia, Ali et al8 evaluated the serum Mn levels of
8 long-term adult HPN patients over a 4-month period to
determine if there was any correlation with liver cholesta-
sis. The patients were receiving a commercial multiple TE
preparation providing 0.3 mg (5 ␮mol) Mn per day 3
times per week. Serum levels of liver enzymes, such as
GASTROENTEROLOGY Vol. 137, No. 5
alanine aminotransferase (ALT), ␥-glutamyltransferase,
bilirubin, urea, creatinine, and a calculated glomerular
filtration rate were obtained at the same time. Occurrence
of cholestasis was determined by the senior gastroenter-
ologist when increased bilirubin greater than 70 ␮mol/L
and increased ALT, ␥-glutamyltransferase, and aspartate
aminotransferase were considered as biochemical evi-
dence of cholestasis. Half of this small cohort of patients
showed hypermanganesemia irrespective of the presence
of cholestasis and/or duration of PN. One patient had
high serum Mn levels with very mild increases of liver
enzyme levels, 2 patients displayed increased serum Mn
levels with normal liver enzyme levels, and 1 patient had
below normal Mn levels with biochemical evidence of
cholestasis. Mn supplementation was discontinued in 3
patients with increased serum levels. Subsequent blood
tests showed serum levels eventually decreased and nor-
malized over a 5- to 6-month period. No correlation
existed in these adults between increased Mn and the
indices of cholestasis.
These clinical experiences are supported by the recent
infant study of McMillan et al.9 Cholestasis was not
deemed to be a significant predictor of Mn status in their
assessment of 54 pediatric PN patients. Of the 20 pa-
tients with cholestasis, 13 had normal serum Mn levels,
but 13 of 21 patients with high Mn levels had no cho-
lestasis. The investigators found no correlation in their
regression model comparing cholestasis as a predictor of
high, low, or normal Mn (P ⫽ .7732). Nevertheless, they
strongly recommended that all pediatric patients who
develop cholestasis should have Mn levels determined
regularly.
Measuring and Monitoring Levels of
Manganese
Regular monitoring of patients receiving fixed
doses of Mn over prolonged periods is recommended, but
plasma and serum analyses are poor indicators of body
stores. Balance studies are problematic because of the
rapid excretion of Mn into bile, and because studies over
short periods do not give results proportional to Mn
intakes. A battery of potential biomarkers including lym-
phocyte Mn-SOD and/or arginase activity has been pro-
posed, but there is no readily available indicator of whole-
body Mn status.10 The recent prospective observational
study in pediatric PN patients concluded that there was
no correlation between copper and Mn levels and no
evidence to support the practice of dosing Mn based on
the more reliable determination of serum copper values.9
Reference intervals of what is considered a normal Mn
level vary considerably. Because 60%– 80% of Mn is con-
tained in red blood cells,5 erythrocyte or whole-blood Mn
concentrations appear to be the most accurate and re-
producible parameter. According to The Auckland refer-
ence laboratory, normal Australasian values can range
November Supplement 2009
between 7 and 10 nmol/L (0.38 –1.1 ␮g/L) in serum and
140 –220 nmol/L (7.7–12.1 ␮g/L) in whole blood. These
ranges are similar but tighter than the US values and the
range quoted by the Scottish Trace Element and Micro-
nutrient Reference Laboratory,11 which in our opinion is
the most useful published reference for a normal range.
Mn deposition in the brain can be detected with mag-
netic resonance imaging (MRI), which correlates with
increased signal intensity on T1-weighted images. Whole-
blood Mn correlates with MRI signal intensity. In 2
recent HPN case reports with increased plasma Mn from
Canada, both showed signs of Mn brain deposition by
MRI but returned to normal after 9 months of Mn-free
PN.12 Although routine cerebral MRI scans may not
always be clinically feasible, whole-blood Mn determina-
tion in combination with the occasional use of MRI to
monitor Mn accumulation in long-term HPN patients
seems to be an acceptable compromise and a good prac-
tical recommendation.5
Is Manganese Deficiency Clinically
Relevant?
Apart from experimentally induced cases, there
are virtually no published reports of Mn deficiency in
hospitalized and/or PN patients. Only one isolated case
of a short-bowel patient with Mn deficiency has been
reported, which was corrected with oral supplementa-
tion.13 This remedial action supports the experience
from the Australasian HPN Register, with 108 regis-
tered HPN patients (20% children), that most HPN
patients who are encouraged to eat a little food in line
with current guideline recommendations do not show
signs of Mn deficiency.5 On the other hand, high Mn
levels are particularly problematic in long-term HPN pa-
tients.14 Hypermanganesemia also has been reported in
acute care and in pediatric settings. A review of 389 cases
of hypermanganesemia (265 adults and 124 pediatric
patients)3 and a recent analysis of post mortem data15
confirm the cumulative effect of Mn supplementation in
long-term HPN patients. Although increased whole-
blood or plasma Mn levels are fairly common in HPN,
overt clinical signs and symptoms are not observed in all
patients with hypermanganesemia. Unfortunately, elimi-
nating the fixed-ratio TE products from HPN regimens
means that selenium, zinc, and copper levels concurrently
decrease to below normal. The most recent experience from
the United States with short-term trauma patients who
were supplied PN containing 300 ␮g/d Mn confirmed
that there is minimal loss in the continuous venous–
venous hemofiltration ultrafiltrate, suggesting that exces-
sive amounts of Mn are being retained.16 Overall, it
appears that supplementation with fixed formulations of
TE additives can lead to potentially toxic levels of Mn,
but it is difficult to predict individual clinical responses.5
MANGANESE IN PARENTERAL NUTRITION S31
Posology, Formulation, and
Contamination of PN Products
The 2006 Nutrient Reference Values for Australia
and New Zealand recommend oral Mn intakes of 5.5
mg/day for men (5 mg/day for women). However, the
adequate intake value established in 2001 by the US
National Academy of Sciences is much lower at 2.3 mg/
day for men and 1.8 mg/day for women. Current Aus-
tralasian Society of Parenteral and Enteral Nutrition
(AuSPEN) PN Guidelines recommend 275 ␮g/day for
adults and 1 ␮g/kg/day for infants and children.17 More
recent guidance from American Society of Parenteral and
Enteral Nutrition (ASPEN)17 and European Society for
Clinical Nutrition and Metabolism (ESPEN)18 recom-
mend 60 –100 ␮g/d or 1 ␮g/kg/d for adults with a max-
imum of 50 ␮g/d for children receiving long-term PN.
Apart from inappropriately high dosages, Mn contam-
ination of intravenous products is a concern. For more
than 20 years studies from the United States have shown
that PN solutions without Mn supplementation can con-
tain 5–38 ␮g/L, especially from calcium gluconate, mag-
nesium sulphate, or potassium chloride.19 Buchman et
al20 have reported Mn contamination up to 310 ␮g/L in
amino acid solutions and various standard PN formulae,
and more recently Japanese investigators estimated their
HPN patients received up to 6 ␮g/day of Mn from glu-
cose, amino acid, and electrolyte solutions.21 As with all
TE determinations, the need for meticulous blood col-
lection, sampling, and analytic methods rightly has been
emphasized by Shenkin (see Question and Answer Ses-
sion). Certainly some of the earlier literature reports do
not specify the laboratory procedures used and therefore
may be questionable. Nevertheless, the reports of Mn
contamination have been so widespread over the years
that there is no doubt there is a problem that needs to be
addressed by manufacturers. In our own investigation of
products used to compound HPN solutions in New Zea-
land, after taking reasonable precautions in line with
Good Laboratory Practices to minimize external contam-
ination, we found Mn levels ranging from 45–90 ␮g/L.5
Five of 6 HPN patients receiving this level of contamina-
tion, in addition to the 275 ␮g/day Mn as part of their
HPN prescription, showed higher than normal blood
levels ranging up to 33 ␮g/L, with accompanying in-
creases in liver function tests (ALT, ␥-glutamyltrans-
ferase) until Mn was withdrawn. This variability coupled
with the ubiquitous Mn presence in needles, tubing, and
so forth makes it difficult to ensure consistent intake and
further supports the argument that recommended doses
still may be excessive.
The optimal dosage for long-term supplementation for
adults is likely to be no more than 1 ␮mol/day (55
␮g/day) and 1 nmol/kg/day (1 ␮g/kg/day) in infants.5,21
This recommendation also can be regarded as a maxi-
mum dosage for the critically ill and others on short-
term PN (5–10 days) in whom contamination and Mn
S32 GIL HARDY
accumulation is less of an issue. Unfortunately, in coun-
tries where only fixed-TE combinations are licensed and
individual TE products are not readily available, it is
impossible to adjust the concentration of Mn in the PN
formula without reducing the other essential TEs. In view
of the numerous reports of toxicity in pediatric and
long-term patients (⬎14 days) there is a real need to
reformulate these multi-TE products with a lower Mn
concentration and to manufacture some products that
exclude Mn altogether. There is also a good case for
regulatory agencies to follow the Food and Drug Admin-
istration example for aluminum and insist that PN prod-
uct literature state the maximum possible content of Mn
and other TE contaminants.22
Conclusions and Recommendations
There is now a persuasive argument for not rou-
tinely adding Mn to PN admixtures. In HPN patients the
intestinal regulatory mechanism is bypassed and intrave-
nously delivered TE are completely bioavailable. Mn de-
ficiency is virtually unknown in human beings, apart
from some reports of subclinical deficiencies, but the
growing evidence for toxicity suggests PN supplementa-
tion remains too high. Neurotoxic damage may result
from oxidative stress induced by Mn accumulation in
astrocytes. Ubiquitous Mn contamination appreciably in-
creases the actual amount delivered, and together with
dietary sources could be sufficient to cover most PN
requirements.
●
Routine whole-blood measurement of Mn in com-
bination with MRI is recommended to monitor po-
tential accumulation.
● Posology guidelines and TE formulations need revi-
sion, and current international guidelines still may
be excessive for long-term HPN.
● Fixed-dose multiple TE formulations restrict pre-
scribing options and make it difficult to adjust Mn
levels without reducing the other essential TEs.
● Manufacturers should reformulate these multi-TE
products with lower and/or zero Mn content and
make individual TE products available.
● All PN products should be labeled with a maximum
allowable TE contamination level.
Variability in clinical practices necessitates updated
micronutrient guidelines and reformulated products for
HPN. The scarcity of PN-associated Mn deficiency, in
combination with the growing body of evidence for Mn
toxicity, leads us to conclude it is unnecessary for Mn to
be added routinely to HPN regimens.
Question and Answer Session
DR BUCHMAN: Both cholestasis and liver disease
have been associated with manganese toxicity; do you
GASTROENTEROLOGY Vol. 137, No. 5
think that these patients developed liver disease and then
manganese accumulates both there and in the brain, or
does manganese contribute to liver disease? I’m curious
because in one of the slides you showed, I think 13 out of
20 people with cholestasis didn’t have an elevated man-
ganese. How do you think this works?
DR HARDY: I guess that’s the million-dollar question;
I honestly don’t know. I think there are many factors that
can be associated with cholestasis. The main factor is
probably that patients are being fed intravenously rather
than enterally. I think this is one of the areas that we
need more research.
DR NIELSEN: I was fascinated by the pictures of the
brain that you showed of children with high manganese.
You said the investigators went to a “manganese-free
solution.” There’s absolutely nothing that is totally free
of manganese, so I assume there was some manganese in
that second solution. Does this study suggest that some
of the brain changes in these children were reversible?
DR HARDY: That appears to be the case. It was a
Japanese paper, and no reference was given about how
they made their “manganese-free” parenteral solution;
one can withhold added manganese from the parenteral
nutrition regimen but some is there as a contaminant. If
I remember correctly, there were some very small lesions
in the so-called “manganese-free” group. In terms of
reversibility the lesions did slowly regress once the added
manganese was withheld.
DR SHENKIN: Thanks, Gil. Very interesting presen-
tation. I’d like to just put a little bit of a health warning
if I could on some of the data which you are presenting.
I’m very concerned about some of the numbers which are
in some of the contamination studies. One of the easiest
ways to get a wrong result in manganese is in the sample
collection procedure. If one collects a sample using a
needle to aspirate from a bag, and if one puts it into a
container which has not been acid-washed, the results
will be uninterpretable. Manganese, as you’ve pointed
out, is ubiquitous; it’s everywhere. When we are collecting
samples from patients’ PN bags we make sure to never
use anything metal and everything which is used has
been acid-washed. If you look at some of the studies in
the literature, you can in fact get very low manganese
contamination. The Japanese study that you mentioned,
done by Takagi, had a maximum of 6 mcg/day of man-
ganese contamination. You see other studies quoting
hundreds and hundreds of mcg/day of manganese. If you
look at the collection procedures which have been used,
very few researchers point out precisely how they col-
lected their samples, whether they used needles or not,
and whether they used acid-washed containers or not. If
they did not follow these sampling precautions, then the
studies are invalid. I’m sorry to belabor this point, but
this is fundamental to every study on manganese.
DR HARDY: I accept your point, Alan. I think the
same issues arise for all trace element studies.
November Supplement 2009
DR SHENKIN: There are some studies which have
used scrupulous sampling techniques. Alistair Forbes
here did an excellent study using proper sampling and we
have done some ourselves, but many studies have not
taken that degree of trouble.
DR HARDY: I have to admit that in our studies we
didn’t use acid-washed containers. We didn’t use metal
needles, we tried to avoid any metal contact, and we did
as much as we could to validate the methodology to
exclude potential manganese contamination, but I agree
with your general criticism of the literature.
DR BIESALSKI: You mentioned neurotoxicity and
that manganese increases reactive oxygen species. I agree
only in part with you. Manganese SOD is an enzyme
located in mitochondria and if you have overexpression,
it depends on the selenium status whether you get reac-
tive oxygen species or not. I think this is important with
respect to antioxidants; we have to keep in mind they
form a network, so where you have low selenium with a
high manganese concentration, you probably have a
higher toxicity. I think this is an important aspect.
DR HARDY: Yes, very important, and of course there
are other forms of SOD, incorporating nickel, zinc, cop-
per, or iron, so I think the point about selenium is
interesting.
DR JEEJEEBHOY: Manganese is a different kind of
element in that first of all it’s distribution is very specific;
it’s a metallo-cofactor for certain enzymes primarily in
the mitochondria. As you know, mitochondrial turnover
is different from tissue turnover. Mitochondria are ex-
tremely stable and act almost like an independent or-
ganelle sitting in the body. I wonder whether there’s truly
a manganese requirement. Since once the mitochondria
are formed the metallo-coenzymes in the mitochondria
are very interchangeable as pointed out, and selenium,
copper, and other metallo-cofactors might have the same
effect as manganese. It seems to me that under the rules
of Cotzias, manganese doesn’t appear to be actually a real
requirement. It seems to me to be like cadmium, al-
though more toxic. Once there’s a certain amount of
manganese in your body, that’s all that matters. Do we in
fact have a requirement for manganese? There’s no defi-
ciency disease that’s ever been described in humans.
DR SHEARER: I was interested by the fact that you
said that 90% of manganese was excreted in the bile. I
wondered whether there have been any studies on the
molecular mechanisms of this excretion and how cho-
lestasis and biliary obstruction could affect these ex-
cretion mechanisms. Are there any transport proteins
known?
DR HARDY: That’s an interesting point; I’m afraid I
don’t know what would be involved, but one would
assume a transport protein is involved.
DR BUCHMAN: I’m trying to remember back to our
data from 16 years ago. I believe that when we measured
contamination we found increased manganese deposi-
MANGANESE IN PARENTERAL NUTRITION S33
tion in both the kidneys and brain in a rat model of
parenteral nutrition. My question is actually for Lyn: did
you evaluate manganese concentrations in the brain tis-
sue of your autopsy patients and, if so, did you find any
increased amount as opposed to other organ tissues?
DR HOWARD: The answer in terms of the published
data is that we didn’t. However, we did collect autopsy
brain samples from these 8 long-term PN individuals and
when we have good control data we will be glad to
publish our findings.
DR JAKSIC (Children’s Hospital in Boston): I have 2
questions. We often cut the trace minerals in half in kids
and in adults who have cholestasis. I was wondering
specifically has there been evidence of neurotoxicity, per-
haps confirmed by an abnormal MRI in patients who
have been treated in this fashion?
DR HARDY: Yes, I think Takagi in Japan has looked
at this. Your approach is the same as normally taken in
Australasia. If high levels of manganese are seen, first the
supplement is reduced by half and if the levels stay high
the added manganese is discontinued. We haven’t looked
at MRIs in such patients in Australasia.
DR JAKSIC: So even in patients where this has been
done, manganese levels continue to be high?
DR HARDY: Yes, because in our experience and in
other reports in the literature it can take 6 to 9 months
for the manganese levels to return to the normal range
even after excluding a manganese supplement from the
patient’s parenteral nutrition. These are, of course, long-
term PN patients.
DR JAKSIC: The second question is a little more
theoretical. In regards to your glutamine synthase argu-
ment, I can see that by increasing manganese you reach a
point of adequacy. Conversely, I see that manganese de-
ficiency will cause less activity of glutamine synthase, but
once you reach adequacy, why should increasing the
amount of manganese cause further change in that en-
zyme’s activity? It seems to me it’s like the classic analytic
biochemistry experiment where students add more cata-
lyst trying to increase yield. Once you’ve got enough
enzyme activity I don’t see why further increasing the
metallo-cofactor would increase the activity of that en-
zyme.
DR HARDY: You are correct; it may be that once the
enzyme’s activity is optimal, then the manganese deposits
in some other form. I don’t think that’s been looked at,
but it appears from the work of Aschner and colleagues
that excess Mn causes ATP loss in astrocytes; that distorts
glutamate/glutamine homeostasis, which could lead to a
build-up of ammonia in astrocytes and hence neurotox-
icity, but it’s speculative. There is some work from France
looking at encapsulating glutamine synthase to admin-
ister it in a more concentrated fashion to reduce ammo-
nia levels in animals, and those preliminary data seem
interesting.
S34 GIL HARDY
DR SALTZMAN (Tufts, Boston): This is a question
about use of trace element blood levels in research stud-
ies and in the clinical setting. If we measure levels of the
micronutrients that are provided in parenteral nutrition
solutions in milligram amounts we certainly would be
measuring both what was in the solution as well as what
reflected the patient’s tissue level. With this problem in
mind, does it matter if blood levels are measured during
the infusion and, if it does, how long after the infusion
should one wait to measure blood levels?
DR HOWARD: That is a $99 question. Do you want
to answer it, Dr. Hardy?
DR HARDY: I think there are serious limitations to
measuring plasma levels. We generally do stop the par-
enteral nutrition for a period of time before we take our
sample but this is a critical issue. We could discuss some
of the research we’d like to do, but until we get this basic
analytical question correct and know that we are mea-
suring a value that reflects tissue manganese, I think the
research or clinical monitoring is going to be a little bit
questionable. This is equally true for other micronutri-
ents. As an aside, I think we have to stop talking about
giving trace elements as though they were metals. We are
not giving metallic manganese, iron, or copper; we are
giving them as salts that may have different valences. At
the moment we generally use atomic absorption meth-
odology to measure the manganese or other trace ele-
ments. Copper has 2 valences, iron has 2, selenium and
manganese have 3, so I think we need to develop more
sophisticated methods which measure the actual valence
of these elements. I accept the concerns you have about
measuring plasma levels.
DR LIPMAN (Washington, DC): I have 2 questions.
The first one dovetails to this last discussion. If we’re
interested in contamination and if manganese is ubiqui-
tous, isn’t what’s most important the total amount of
manganese being delivered to the patient? I’m less inter-
ested in what’s in the solution and more in what’s com-
ing out of the bag and going into the patient’s vein. Has
anybody looked at the difference between these values?
DR BUCHMAN: I think you’re asking if in addition
to the added amount and the contaminant in the solu-
tion components, does any come from the container or
delivery set or is there adherence to the plastic tubing
like, for example, insulin and vitamin A.
DR HARDY: There may be manganese contamination
in the plastic tubing. PN is usually infused through a
silicone catheter without a metal needle. Takagi refers to
an unpublished report from Japan that looked at the
administration tubing or catheters for manganese but
found minimal contamination.
DR LIPMAN: My second question relates to Khursh’s
speculation that maybe we have no requirement for man-
ganese. This does not quite make sense to me because we
have manganese in the body; we’ve had to get it from
GASTROENTEROLOGY Vol. 137, No. 5
some place, and since there’s a mechanism for excretions
through the bile, there has to be a mechanism for replen-
ishment.
DR HOWARD: I think that’s an important point. Dr.
Shenkin, you want to address that?
DR SHENKIN: Could I just make the point that the
Japanese study which Gil’s mentioned from Takagi’s
group used solutions which were minimally contami-
nated, providing somewhere between 3 and 6 mcg/day
from contamination. He did a cross-over study giving
either zero manganese or 1 ␮mol/d (55 ␮g/d), 2 ␮mol/d
(100 ␮g/d), or 20 ␮mol (1000 ␮g/d). The groups receiv-
ing 1 ␮mol/d (55 ␮g) and above maintained whole-blood
manganese concentration. With zero manganese the
whole-blood value fell below the mean but still stayed
within the normal reference range. This is the only study
which suggests that 1 ␮mol/d of manganese may be
optimal to maintain the patient’s normal starting man-
ganese level.
DR NIELSEN: Dr. Hardy, some Japanese workers
headed up by a fellow named Neross fed a very low level
of manganese in the parenteral nutrition solution of a
pediatric patient for about 5 years. They noted osteopo-
rosis and stunted growth and when they added some
manganese to the solution these clinical problems were
overcome. This suggests you can get deficient in manga-
nese if you get down to 6 ␮g/day; that’s a pretty low
amount.
DR HARDY: I guess it comes back to the point that
we are talking about short-term acute situations and
longer-term situations. We’ve been encouraged in our
discussions with the industry in Australia that they are
agreeable to producing multi–trace elements products
with manganese and without manganese. I guess we’re
waiting for deliberations at this conference to define
what we need.
DR SHIKE: Regarding the question which came up
earlier, when should we measure blood levels, during the
infusions or between the infusions? We should keep in
mind that an infusion of PN is not like the absorptive
state of an oral diet. What the organs see during 10 or 12
hours of PN infusion is as important as what they don’t
see during the 12 hours off. If you take the case of
calcium, phosphorus, or magnesium, what we and others
have shown is blood levels and urinary excretion go up
during infusions and go down off infusions. This may
have profound consequences. For instance, in the case of
calcium, the ionized calcium certainly will suppress PTH
for a prolonged period during the infusion. Now in the
case of things that get deposited in organs like manga-
nese in the brain, if the manganese levels are higher
during the infusion, the brain for 10 or 12 hours is seeing
high levels and therefore would deposit them. So I think
this is an important question: should we measure micro-
nutrients both during and between infusions?
November Supplement 2009
References
1. Angstmann K, Ball P, Hardy G, et al. The AuSPEN HPN Register
2008 annual report. Available at: www.auspen.org.au. Accessed
May 14, 2009.
2. Aschner J, Aschner M. Nutritional aspects of manganese ho-
meostasis. Mol Aspects Med 2005;26:352–362.
3. Dickerson RN. Manganese intoxication and parenteral nutrition.
Nutrition 2001;17:689 – 693.
4. Milatovic D, Gupta R, Sidoryk M, et al. Manganese induces
oxidative impairment in cultured rat astrocytes. Toxicol Sci 2007;
98:198 –205.
5. Hardy IJ, Gillanders L, Hardy G. Is manganese an essential sup-
plement for PN? Curr Opin Clin Nutr Metab Care 2008;11:289 –
296.
6. Reimund JM, Dietemann JL, Warter JM, et al. Factors associated
to hypermanganesemia in patients receiving home parenteral
nutrition. Clin Nutr 2000;19:343–348.
7. Siepler JK, Nishikawa RA, Diamantidis T, et al. Asymptomatic
hypermanganesemia in long-term home parenteral nutrition pa-
tients. Nutr Clin Pract 2003;18:370 –373.
8. Ali A, Murdoch A, Pascoe A, et al. Can hypermanganesemia occur
without cholestasis in patients receiving HPN. Nutrition 2008;24:
508.
9. McMillan NB, Mulroy C, MacKay NW, et al. Correlation of cho-
lestasis with serum copper and whole blood manganese levels in
paediatric patients. Nutr Clin Pract 2008;23:161–165.
10. Failla M. Considerations for determining optimal nutrition for
copper, zinc, manganese, molybdenum. Proc Nutr Soc 1999;58:
497–505.
11. Trace Element and Micronutrient Reference Laboratory. NHS
Scotland. Available at: http://www.trace-elements.org.uk. Ac-
cessed May 14, 2009.
12. Stevens L, Rashid M. Manganese toxicity in patients on long term
HPN therapy. JPEN J Parenter Enteral Nutr 2008;23:197–198.
13. Norose N, Terai M, Norose K. Manganese deficiency in a child
with very short bowel syndrome receiving long-term parenteral
nutrition. J Trace Elem Exp Med 1992;5:100.
14. Howard L, Ashley C, Lyon D, et al. Autopsy tissue trace elements
in 8 long-term parenteral nutrition patients who received the
MANGANESE IN PARENTERAL NUTRITION S35
current US Food and Drug Administration formulation. JPEN J
Parenter Enteral Nutr 2007;31:388 –396.
15. Klein CJ, Nielsen FH, Moser-Veillon PB. Trace element loss in
urine following traumatic injury. JPEN J Parenter Enteral Nutr
2008;32:129 –139.
16. Russell D. The AuSPEN guidelines for intravenous trace elements
and vitamins. Available at: http://www.auspen.org.au. Published
1999. Accessed May 14, 2009.
17. Mirtallo J, Canada T, Johnson D, et al. Safe practices for paren-
teral nutrition. JPEN J Parenter Enteral Nutr 2004;28:S39 –S70.
18. Koletzko B, Goulet O, Krohn K, et al. Guidelines on paediatric
parenteral nutrition of ESPHGAN and ESPEN, supported by ESPR.
J Paediatr Gastroenterol Nutr 2005;41:S1–S87.
19. Pluhator-Murton MM, Fedorak RN, Audette RJ, et al. Trace ele-
ment contamination of total parenteral nutrition. JPEN J Parenter
Enteral Nutr 1999;23:222–227.
20. Buchman AL, Neely M, Grossie VB Jr, et al. Organ heavy-metal
accumulation during parenteral nutrition is associated with
pathologic abnormalities in rats. Nutrition 2001;17:600 – 606.
21. Takagi Y, Okada A, Sando K, et al. Evaluation of indexes of in vivo
manganese status and the optimal intravenous dose for adult
patients undergoing home parenteral nutrition. Am J Clin Nutr
2002;75:112–118.
22. Food and Drug Administration. Aluminum in large and small
volume parenterals used in total parenteral nutrition. Fed Regist
2000;6:4103– 4111.
Received May 14, 2009. Accepted August 7, 2009.
Reprint requests
Address requests for reprints to: Gil Hardy, PhD, FRSC, Faculty of
Medical and Health Sciences, University of Auckland, Private Bag
92019, Auckland, New Zealand. e-mail: g.hardy@auckland.ac.nz; fax:
(64) 9 367 7192.
Conflicts of interest
The author discloses no conflicts.