INTERACTION
OF
VITAMIN
C
AND IRON*
Sean
R.
Lynch and James
D.
Cook
Division
of
Hematology
Deportment
of
Medicine
University
of
Kansas Medical Center
Kansas
City,
Kansas
66103
and
Veterans Administration Hospital
Kansas City, Missouri
64128
In early studies
of
food iron absorption, test meals consisted
of
single foods that
had been biosynthetically labeled with radioiron.‘ Absorption varied considerably
with individual food items, and in general, iron derived from animal tissue was more
available than iron of vegetable
rigi in.^-^
More recently, it has been shown that the
percentage absorption
of
iron from a given food item differs substantially when it is
eaten in a meal containing two
or
more additional foods.’-‘ For example, soybean
iron is much better absorbed than black bean iron if each is eaten separately, but a
similar percentage
of
iron is absorbed from both bean sources when they are eaten
t0gether.j Absorption therefore appears to be a property
of
the overall composition
of
the meal rather than
of
the single food item. These studies led to a very important
observation. If
a
small quantity of soluble inorganic iron is added to a biosyn-
thetically labeled vegetable food just before it is eaten, percentage absorption
of
the
two forms of iron
is
virtually identical.’ The same is true of meals containing more
than one vegetable food.’-’ Therefore, it appears that when several foods are eaten
together, nonheme iron destined for absorption behaves as though it were derived
from a single common pool. Soluble inorganic iron salts used in fortification
also
enter this pool.’,’ Iron absorption from it can be measured by determining the size
of
the
pool
and the percentage absorption
of
a
soluble radioiron tracer added
to
the
meal.’-’
Similar observations have been made for heme iron, which comes principally
from hemoglobin and myoglobin
of
meat. Heme compounds enter a second com-
mon
pool
and share a set
of
properties different from those that govern nonheme
iron absorption.’ For example, heme is not degraded to any degree in the lumen
of
the gut but is taken up by the mucosal cells with iron still within the porphyrin ring.1°
Heme iron absorption is relatively independent
of
other components
of
the meal;
ascorbic acid has no effect.”.” Absorption
of
heme iron can be measured by adding
a
small quantity
of
labeled hemoglobin to a meal just before it is eaten.9 Virtually all
dietary iron is absorbed from these two pools
of
heme and nonheme iron.”
In attempting
to
combat iron deficiency, most investigators have focused their
attention on nonheme iron, which makes up over
90%
of
the iron in an average
diet
.I3
The percentage absorption
of
nonheme iron is only about one-tenth that
of
heme. Unlike heme iron, the availability
of
nonheme iron is critically dependent
on
several factors, some
of
which enhance and others
of
which inhibit its uptake.“
Many chemical ligands present in food are known to have
a
powerful influence. For
This
work
was
supported
by
contract DSAN-C-0045 from the
U.S.
Agency
for
Interna-
tional Development and by Veterans Administration Interagency Agreement
IGA
V
589
(134)5-79033
from the
U.S.
Food and
Drug
Administration.
32
0077-8923/80/0355432 S01.75/0
0
1980,
WAS
Lynch
&
Cook:
Vitamin
C and
Iron
33
example, carbonates, oxalates, phosphates, fiber components, and tannates impair
absorption, while ascorbic acid, citric acid, tricarboxylic acids, amino acids, and
sugars promote absorption. Of these, ascorbic acid has the greatest influence on iron
assimilation.
Initial evidence suggesting that ascorbic acid can enhance iron absorption from
certain foods was obtained in studies using meals eaten with and without citrus
fruits.l*lJ-zD For example, when
100
ml orange juice containing
40-50
mg ascorbic
acid was added to a meal consisting
of
bread, butter, jam, tea or coffee, and an egg
that had been biologically labeled with '9Fe, absorption increased from
3.7
to
10.4%.18
More recently, Rossander
et
01.
studied a common Swedish "continental
breakfast" and demonstrated that when orange juice containing
70
mg ascorbic acid
is consumed with coffee and two wheat rolls, absorption is
2%
times higher than that
from the same meal without orange juice."
Moore and Dubach were the first to identify ascorbic acid as the cause of the in-
creased absorption seen with meals containing fruit.' They demonstrated a similar
increase in absorption with equivalent quantities of ascorbic acid. Their findings
have subsequently been confirmed by others. Layrisse
et
ul.
demonstrated that
nonheme iron absorption from a maize meal was increased severalfold when papaya
was added to the meal.19 The same degree of enhancement was achieved by adding
ascorbic acid equivalent in amount to that contained in the papaya. Thus, although
it is possible that there are factors other than ascorbic acid in fruit juices that
enhance absorption, there is little doubt that ascorbic acid accounts for the major ef-
fect.
ASCORBIC
ACID
AS
A
NONHEME COMMON
POOL
LIGAND
Early observations involving fruit juices led to a series
of
studies of the effect of
ascorbic acid on the absorption
of
various forms
of
nonheme ir~n.'~.*~-~~ In two
of
the more recent investigations, ascorbic acid was added to meals prepared from
maize, wheat, soya, and ri~e.".~' The cereal component
of
each meal had been labeled
intrinsically with ''Fe by hydroponic cultivation.
'9Fe&
used as the extrinsic label
for the nonheme pool, was added to the meal in trace quantities either immediately
before the meal was eaten or to one
of
the ingredients before cooking. Supplemental
iron in the form of
a
soluble iron salt, such as
2-5
mg
ferric ammonium citrate
or
4
mg ferrous sulfate, was incorporated into some of the meals. Ascorbic acid produced
the same increase in absorption of both extrinsic and intrinsic radioiron labels over a
wide range of absorption values and ascorbic acid concentrations (FIGURE
1).
The
composite mean absorption ratio
of
extrinsic to intrinsic radioiron was
1.13,
a
value
very close to that
of
1.10
reported by Cook
et
ul.
for meals containing no ascorbic
acid.' These observations strongly support the conclusion that ascorbic acid acts
as
a
ligand for the common nonheme iron pool and thus enhances absorption
of
all iron
compounds in this pool to an equal extent. An important corollary is that extrinsic
tagging allows valid comparisons to be made between meals eaten with and without
supplementary ascorbic acid.
A number of ligands in food have
a
powerful inhibitory effect on absorption
from the nonheme iron pool. The interaction
of
ascorbic acid with two
of
these, tan-
nates and calcium/phosphate, has been ~tudied.~*.~~ Ascorbic acid reverses the strong
inhibitory effect that tea exerts on nonheme iron absorption from cereal-based
meals." The degree
of
reversal of inhibition is directly proportional to the quantity
of added ascorbic acid; absorption in the presence
of
ascorbic acid and tea may ex-
ceed absorption from the basal meal containing neither. The effect
of
ascorbic acid
34
Annals New
Y
ork
Academy
of
Sciences
41.3.43.4
7.
565,3727;
/
-/’
.
,/’
:
,/
0
S
I0
b
2025
30
3540
%
ABSORPTION
OF
INTRINSIC
IRON
IN
FOOD
FIGURE
1.
Nonheme iron absorption from test meals tagged simultaneously with intrinsic
and extrinsic radioiron labels (see text
for
explanation).
on the inhibitory influence
of
calcium/phosphate on iron absorption was studied
us-
ing a semisynthetic meal.3’ When inorganic calcium/phosphate was omitted from
this meal, absorption increased about twofold, but the mean ratio of the meal with
ascorbic acid
to
that without
was
similar
whether or not calciurdphosphate
was
pres-
ent. Apparently, ascorbic acid and known inhibitors
of
iron absorption have an op-
posite, but additive, effect on absorption from the nonheme common pool and can
therefore be considered to act
as
competitive ligands.
The influence of ascorbic acid
on
iron absorption from meals containing other
absorption promoters is
less
well defined.
For
example, meat and fish increase
nonheme
iron
absorption when added to a vegetable meal. Cook and Monsen com-
pared iron absorption from
a
standard meal containing beef, potatoes, cornmeal,
peaches, ice milk, bread, and margarine with that
of
a
semisynthetic meal
of
similar
composition prepared with dextrimaltose. corn
oil,
and ovalb~min.~’ Nonheme iron
absorption from the standard
meal
was much higher owing
to
its content
of
meat.
Absorption from both these meals was enhanced by adding
100
mg of ascorbic acid.
However, the mean absorption ratio
of
the meal with ascorbic acid to that without
was
3.19
for the semisynthetic meal and only
1.67
for the standard meal, indicating
that the enhancing effects of meat and ascorbic acid are not additive.
A
similar
observation has
been
made with fish and ascorbic acid. Whereas
100
g fish increased
the absorption
of
nonheme iron about threefold,”
the
percentage absorption of iron
from
a
maize
meal
eaten
with
100
g fish and fresh papayas containing approximately
70
mg ascorbic acid was the same
as
that from the meal eaten with the papayas
alone.”
Although
it is not possible to draw definitive conclusions from this evidence,
these two studies suggest that ascorbic acid and beef or fish are complementary
rather than additive in their enhancing properties and that the addition of ascorbic
acid to a meal containing meat
or
fish may have
a
relatively small influence on the
quantity
of
nonheme iron absorbed.
Lynch
&
Cook:
Vitamin
C
and
Iron
35
Accumulating evidence indicates that the total intake of dietary iron is less im-
portant from
a
nutritional standpoint than the availability
of
that iron.
A
model has
recently been proposed for estimating dietary iron absorption from any given meal
based on the iron status
of
the individual, the content of heme and nonheme iron,
and the content of enhancing
constituent^.^'
Of the many factors that have been said
to influence iron absorption, only two were considered important enough for inclu-
sion in the method for estimating available dietary iron. Meals were defined as hav-
ing low, medium, or high iron availability based
on
the content
of
meat, poultry,
or
fish and
of
ascorbic acid. In this model,
1
g meat was considered roughly equivalent
in its enhancing effect to
1
mg ascorbic acid. Estimates
of
nonheme iron absorption
in subjects with
500
mg iron stores are shown in TABLE
1.
The model defines the ef-
fect
of
these substances when added separately to a meal. It is less clear whether their
effect is additive when both are contained in the meal. Presumably, the effect of
30-90
g
meat, fish, or poultry
and
25-75
mg ascorbic acid would be roughly additive
and therefore would place the meal in the high iron availability category, but
as
previously discussed, this point is not clearly established.
INTERACTION
OF
ASCORBIC
ACID
AND
CONTAMINANT
IRON
Not
all
the iron in
a
meal is necessarily intrinsic to the food. Chemical assays
of
Indian diets have consistently revealed iron contents considerably in excess of those
determined from food composition tables." The disparity is believed to represent
iron contamination either from the soil
or
from cooking utensils. The high prev-
alence
of
iron deficiency in these geographic areas of the world suggests that this iron
is poorly absorbed. Therefore, the potential influence that ascorbic acid has on these
less soluble forms of iron could be
of
considerable nutritional importance. Evalua-
tion
of
ascorbic acid's effect is difficult because the chemical nature of the extraneous
iron has not been identified and because suitable radiolabeled material is difficult to
obtain. One study has addressed the problem indirectl~.)~ Two likely food con-
taminants, ferric oxide, which is the major component
of
rust derived from cooking
utensils, and ferric hydroxide, which forms much
of
the iron found in soil and un-
purified water sources, were radiolabeled and added to
a
maize porridge meal; both
were poorly absorbed, with mean values of
0.01%
and
1.5%,
respectively. The addi-
tion
of
100
mg ascorbic acid increased absorption from ferric oxide to
0.5%
and
from ferric hydroxide to
6.7%.
Although the relative increase was pronounced, ab-
sorption was still inappropriately low in the context
of
the meal and the iron-
deficient population studied. This suggests that a significant proportion of each
compound remains insoluble and unavailable to the common nonheme iron pool.
Moreover, ferric hydroxide was absorbed only half as well as the intrinsic iron con-
TABLE
1
DIETARY
IRON
AVAILABILITY*
Meat,
Fish,
Meal Iron Poultry
Availability
(g)
Nonheme
Ascorbic Iron
Acid
Absorption
(mg)
VOO)
Low
<
30
and
<
25
3
Medium
30-90
or
25-15
5
High
>90
or
>
75
8
*
Based
on
data
of
Monsen
et
at.''
36
Annals New
York
Academy
of
Sciences
tained
in
maize porridge. Presumably most
of
the contaminating iron contained in
food remains relatively unavailable for absorption despite the presence of ascorbic
acid.
QUANTITATIVE
ASPECTS
While many observers have noted that percentage nonheme food iron absorp-
tion rises with increasing ascorbic acid content
of
a meal, the quantitative effect
of
ascorbic acid has been examined rigorously in only two studies, which yielded similar
In one of them, ascorbic acid-ranging in amounts from
25
to
1,OOO
mg-was added to a semisynthetic meal containing
4.1
mg elemental iron.
A
linear
relationship was observed between the absorption ratio of the meal with ascorbic
acid to that without and the quantity
of
added ascorbic acid when both values were
plotted on a logarithmic scale.
A
somewhat different relationship is seen when
percentage absorption is plotted against the molar ratio of ascorbic acid to iron using
linear scales
(FIGURE
2).
The response curve has two components. With the molar
ratios below
7.5,
which corresponds to approximately
100
mg ascorbic acid, there is
a proportionately greater effect than with molar ratios above
7.5.
It is interesting
that when ascorbic acid is taken with pharmacologic doses of ferrous sulfate in the
fasting state, a similar two-phase dose-response pattern occurs, although the degree
of
enhancement and the actual quantity
of
iron absorbed are very different in the
two studies
(FIGURE
3).’,
The factors responsible for the two components
of
the dose-response curve are
unknown, but it is tempting to speculate that the initial component may represent
the effect that ascorbic acid has in rendering ferric iron soluble at the alkaline pH
of
the duodenum
(see
below). This would account for the relatively small increase in ab-
sorption seen when ascorbic acid was added
to
ferrous sulfate, since a large quantity
‘I
5
ASCORBIC
ACID:
IRON
RATIO
FIGURE
2.
The effect
of
ascorbic acid on iron absorption from a semisynthetic meal.
Geometric mean percentage iron absorption is plotted against the molar ratio
of
ascorbic acid
to iron in the meal. (From the data
of
Cook
and Monsen.)”