bioavailability of dietary iron in man

7/27/2019 Bioavailability of Dietary Iron in Man

http://slidepdf.com/reader/full/bioavailability-of-dietary-iron-in-man 1/26

An~Rev. Nutr~ 1981. 1:123-47

Copyright©1981 .v ‘4nnualReviews nc..4ll rights reserved

BIOAVAILABILITY OF

DIETARY IRON IN MAN

,1300~

Leif Hallberg

Departmentf MedicineI, Sahlgren ospital,Universityf G~teborg,413 45 G&eborg,weden

CONTENTS

INTRODUCTION......................................................................................................... 123

MEASUREMENTSF DIETARY IOAVAILABILITYF IRON ..................... 124

Earlier ethods......................................................................................................... 124ExtrinsicagMethod............................................................................................... 125

BIOAVAILABILITYFNON-HEMERON........................................................... 126

Individual actorsnfluencing ioavailability......................................................... 126Irontatus............................................................................................................. 126Pregnancy............................................................................................................... 127Diseasetates......................................................................................................... 128

Dietary actorsnfluencingron Bioavailability..................................................... 128Bioavailabilityf iron n singleoods......................................................................... 128Pool onceptnd onceptf bioavailability................................................................. 129Diet actors hat particularlynfluence ioavailabilityf non-heineron ......................... 130Bioavailabilio~f iron n mealsrom ifferent ountries............................................... 132Bioavailabilityf fortificationron ............................................................................. 133

BIOAVAILABILITYFHEMERON..................................................................... 140Individualactors..................................................................................................... 140Dietaryactors......................................................................................................... 141

ASSESSMENTOF BIOAVAILABILITY OF IRON IN THE WHOLEDIET ...... 141Earliertudies........................................................................................................... 141Implicationsf the Pool oncept............................................................................. 142Estimations ased nBioavailable utrientDensity............................................... 143

CONCLUSIONS............................................................................................................. 144

INTRODUCTION

In recent years, knowledgebout ron absorption rom he diet has in-creasedmarkedly.Newmethods avebeendevised that for the first time

havemadet possible to measurehe absorption f iron fromwholemeals,andseveral componentsn the diet havebeen dentified that promote r

1230199-9885/81/0715-0123501.00

www.annualreviews.org/aronlineAnnual Reviews

A n n u .

R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e v i e w s . o r g

b y E T H N O L O G I S C H E S S E M I N A R o n 0 1 / 1 3 / 0 8 .

F o r p e r s o n a l

u s e o n l y .

http://www.annualreviews.org/aronline





124 HALLBERG

inhibit iron absorption. Thus, he bioavailability of iron in a meal s theresult of several known nd unknownietary factors.

There are two kinds of iron compoundsn the diet with respect to the

mechanism f absorption~heme ron (derived from hemoglobinand myo-globin) and non-heineron (derived mainly romcereals, fruits, and vegeta-

bles). Theabsorption f these twokinds of iron is influenced ifferently bydietary factors. Heine ron formsa relatively minorpart of iron intake. Evenin diets with a high meatcontent it accounts or only 10-15%f the totaliron intake. Diets in developing countries usually contain negligibleamountsof heine iron. Non-heineron is thus the mainsource of dietary

iron.Some f the iron present in the diet maybe in a chemical orm hat is

poorly or not at all absorbable.It maybe present as a very stable complexor as iron compoundsith a lowsolubility in the gastrointestinal contents.Examplesre, ironically, iron compoundshat have beenused for fortifica-tion of foods such as mostforms of reduced ron or ferric orthophosphate;

both these compoundsre only partly available for absorption. Foodsmayalso be contaminatedwith dirt and dust composed f poorly soluble ironcompoundshat usually originate from the soil.

Theabsorptionof iron is determined ot only by dietary factors, but also

bypropertiesof the subjects studied, especiallyby their iron status. Severaltimes more iron is absorbed by the severely iron-deficient than by theiron-repletesubject.

Themain actors that influence the absorptionof iron from he diet are(a) the amounts f heme nd non-hemeron, (b) the content of the dietaryfactors influencing ron bioavailability, and (c) the iron status of the sub-jects. Heine nd non-berneron are affected differently, not only by dietaryfactors but also by the subjects’ ron status. Thereforehe bioavailabilityofthese two groupsof iron compoundss discussed separately.

MEASUREMENTS OF DIETARY BIOAVAILABILITYOF IRON

Earlier Methods

Variousmethods ave been used to study the dietary bioavailability of iron

in man see Moore 71)]. Thechemicalbalance echnique s the only methodthat directly measures ron absorption from the wholediet. That method,

however, s insensitive, imprecise, and time consuming, nd it gives noinformation about iron absorption from different meals.The ntroduction of radioisotopes made t possible to label single food

items biosyntheticallywith radioiron (72). Studies with labeled foods haveshownhat absorption rom ndividual foods differs markedly.Thesediffer-ences n bioavailabilityare apparently elated to differences n solubility and


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w

n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e v i e w s . o r g

b y

E T H N O L O G I S C H E S S E M I N A

R o n 0 1 / 1 3 / 0 8 .

F o r p e r s o n a l u s e

o n l y .






BIOAVAILABILITYF DIETARY RON 125

dissociation of the chemicallyuncharacterized ron compoundsn foods. Itwas oundearly on that one food could interact with the absorptionof ironfrom another food (56). Thus, knowing he iron absorption from singlefoods does not provide a valid estimate of the absorption of iron from a

whole meal.In another type of study a trace amountof inorganic radio-iron was

ingestedas a drink with various"standardmeals" o get an index of the ironabsorption n normal ubjects and in clinical disorderssuch as achlorhydria

and iron deficiency. Subsequent tudies have shownhat this methoddoes¯ not measure ccurately he total dietary iron absorption,because t overesti-mates he absorptionof non-heine ron, to a large and varyingextent, andit does not measure eine iron absorption.

Extrinsic Tag Method

In recent years someunexpected bservations providedan important break-through and led to the devdopmentf the extrinsic tag method nd to the

introduction of the pool concept n food iron absorption 23, 33, 37). Whensingle foods biosynthetically abeled with radioiron (intrinsic tracer) werecarefully mixedwith a trace amountof iron salt labeled with anotherradioiron isotope (extrinsic tracer), the observation wasmade hat theabsorption of the two tracers, fromsuch doubly abeled foods, wasalmostidentical. Themagnitude f the absorption wasdifferent from differentfoods and in different subjects, but the absorptionfrom he extrinsic andintrinsic tracers was he same n eachsubject. Based n these findings, theconcept of a commonon-heme ron pool was introduced. This conceptassumes hat the non-hemeron compoundsn different foods in a mealcan

be uniformly abeled by an extrinsic inorganic adioiron tracer and that theabsorption of non-hemeron takes place from this commonool (33).

Heine ron cannotbe labeled by an extrinsic inorganic tracer. Themain

heme ron sources in the diet are hemoglobinnd to a lesser extent myo-globin. Thesimilarity in structure of the hememoietyof these two com-poundsmade t reasonable to assume hat they are absorbed in the same

way, and hence he absorption of hemoglobiniosynthetically labeled withradioiron gives a true measureof the absorption of all heme ron fromameal(37, 51). Hemeron is absorbed s an iron-porphyrin omplex irectlyinto the mucosal ells and his is apparentlyan entirdy different pathwayfrom non-heme ron mucosal entry (45, 85, 87). Thus, heme iron andnon-hemeron form eparate pools of iron in the gastrointestinal tract. Use

of two different radioiron isotopes enables these two pools to be indepen-dently and simultaneously abeled with biosynthetically radioiron-labeledhemoglobinnd with an extrinsic inorganic radioiron tracer (12, 37).

Anumber f studies have been made o validate the pool concept (7, 8,10, 15, 16, 28, 80-82). For example, the non-heme ron pool has been


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






126 HALLBERG

labeled with intrinsic and extrinsic tracers under different experimental

conditions, and the results were compared. The two tracers were absorbed

to the same extent, even when, for instance, inorganic iron in various

amounts, ascorbic acid, or desferrioxamine was added to the diet (7, 23).The extrinsic and intrinsic tracers gave identical plasma radioactivity

curves for several hours when giving doubly labeled foods (16). The twotracers were absorbed to the same extent when a doubly labeled food was

given alone and when he food item formed part of a composite meal (16).

It is probable that the reason for the identical absorption of the extrinsic

tracer and the native iron, intrinsic tracer, is a completeand rapid isotopic

exchange within the commonool of non-heine iron. There are a few known

exceptions in which this isotopic exchange is incomplete: (a) unmilled,

unpolished flee, in which it is likely that the outer dense aleurone layer of

the rice grain impairs the diffusion of iron (16); (b) ferdtin (54, 63);

(c) somepoorly soluble iron compounds sed to fortify foods, such as most

forms of reduc~l iron (14) and ferric orthophosphate (25). The validity

using radioiron-labeled hemoglobin to measure heine iron absorption from

meat has also been studied (37, 51).All of these studies verify that under most conditions the absorption of

both heine and non-berne iron can be determined simultaneously in the

same meal. The absorption of non-heme iron draws the most interest,

because it forms the main part of the dietary iron intake and because its

absorption is muchmore varied due to the marked effect of both dietary

factors and the iron status of the subjects. A "one pool-one radioisotope"

model s then applied.

In the original studies on food iron absorption from whole meals, great

efforts were made o uniformly distribute the extrinsic tracer through all

food items (12, 23, 37, 53). It has been demonstrated subsequently that

more simple technique can be used in which only one bulky component of

the meal is labeled (13). Moreover, it has been shown hat the non-heineiron absorption can be reliably measured from freely chosen meals, pro-

vided the exact intake of foods is recorded and the radioiron is administered

in some bulky component. This method is of great importance in field

studies (40).

BIOAVAILABILITY OF NON-HEME IRON

Individual Factors Influencing Bioavailability

IRONSTATUS he absorption of non-heine iron is markedly influenced

by the iron status of the subject--more iron is absorbedby the iron-deficientand less by the iron-replete subject. This leads to a markedsubject-to-


A

n n u .

R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 . D o w n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e v i e w s . o r g

b y E T H N O L O G I S C H E S S E M

I N A R o n 0 1 / 1 3 / 0 8 .

F o r p e r s o n a l u s e o n l y .



BIOAVAILABILITYF DIETARYRON 127

subjectvariability, which’makest difficult to determinewhether ifferences

in absorption between est meals studied in different groups of subjectsrelate to properties of the mealsor to the iron status of the subject.

Theeffect of differences in iron status among ifferent subjects can beadjusted by obtaining an independentmeasure f their absorptive capacity.This is accomplished y determining he absorption from a standard doseof inorganic radioiron given at physiological levels under standardizedconditions. Theconceptof a reference dose of radioiron was ntroducedbyLayrisse et al in 1969 49). An nformal agreement as been reachedamongworkersn different countries o use 3 mgof iron as ferrous "ascorbate" or

this referencedose.In a groupof subjects with varying ron status there is a high correlation

between he absorption from he reference doses and the non-heine ron inthe meals. Thus, the slope of a regression line betweenhe two absorptionmeasurementsmeals/referencedoses) is an index of the bioavailabilitythe non-heine ron in a meal(13).

In iron-balance calculations it is important to have a measurementfbioavailability that is more oncrete and absolute than the slope of such aregression ine, and which an be related to a certain iron status. Ameaning-ful measureof bioavailability of non-heine ron in a meal wouldbe the

absorption n subjectswho aveborderline ron deficiency, .e. subjects withabsent iron stores but whohave not yet developed nemia. Wehave foundthat the absorption of the reference dose in such subjects is about 40%.

Magnussont al proposed hat the bioavailability of non-hemeron in ameal should be expressed as the absorption value that corresponds to areference dose absorption of 40% 57). To increase the accuracyof thisvalue, it is important to include in each study subjects who ange widelyin iron status.

There s a goodcorrelation between ron stores and serum erritin, and

it has been shown hat there is a goodcorrelation between erum erdtinand non-heine ron absorption 4, 24, 47). Therefore, erum erritin can alsobe used as an alternative to the reference dose absorption. However,erumferritin is only an indirect measuref an individual’sability to absorb ron,and extraneous actors such as minor nfections may ffect iron absorption

and serum erfitin in opposite directions. Referencedoses therefore arepreferable and should be used whenever ossible (34).

PREGNANCYhe bioavailability of dietary iron increases during preg-nancyand is roughlyparallel to the increased iron requirements. n earlypregnancy, however, recent studies have shown hat iron absorption ismuch ower than wasexpected; the explanation for this initial low ironabsorption is not yet known84).


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






128 HALLBERG

DISEASETATESn gastric achlorhydria, the absorption of dietary non-

heme ron is reduced in relation to the absorption from a ferrous iron salt

(reference dose) (8, 22). Studies have shown hat the solubilization of

heine food iron by gastric juice in vitro is pH dependent and in vivo isrelated to the gastric acidity (5). After partial gastrectomy,a decrease in the

bioavailability of non-hemedietary iron is often observed. The magnitude

of the decrease dependson the type of gastric operation performed 58, 59).

In idiopathic hemochromatosis, the absorption of food iron is markedly

increased in relation to the size of the iron stores (6).

Dietary Factors Influencing Iron Bioavailability

BIOAVAILABILITYOF IRON IN SINGLE FOODS Absorption studies

that use biosynthetically labeled single foods were first made by Moore&

Dubach n 1951 (72). Several workers extended these studies to more sub-

jects and different foods. A markedvariation in absorption of food iron was

observed not only between different foods, but also for any given food

betweendifferent investigators and among ifferent subjects. Iron-deficient

subjects tended to absorb more iron than normals, but as a rule there was

an overlap between these groups. A review of the earlier studies has been

published (71).The use of a reference dose of ferrous ascorbate, mentionedabove, greatly

facilitated food iron absorption studies (49). The purpose was to relate

each subject the absorption of food iron to the general ability to absorb iron.

Thebioavailability of iron in various foods was expressedas the ratio of food

iron absorption to the absorption from a reference dose. The main findings

in studies on single foods, with the exception of iron in eggs, which s only

1-3% absorbed, were that iron of animal origin was better absorbed than

was vegetable iron. Thus, in normal subjects the absorption of iron in rice

and spinacb was only about 1-2%, but in animal foods such as beef andveal liver it amounted o 10-20%.A review of studies on single foods was

published in 1971 (50).

Recent studies on single foods have shown hat milling of cereals greatly

affects their iron bioavailability. Iron in milled polished rice, for instance,

is about four times better absorbed than is iron in unmilled rice (16).

Similarly, the bioavailability of wheat iron in bread baked with flour of

different extraction varies in proportion to the bran content (9).

The bioavailability of iron in human reast milk needs a special comment.It has been shown n both adults and infants that the bioavailability of iron

in human milk is about twice as high as that in cow’s milk (67, 79).

Simulated humanmilk with an identical concentration of protein, fat, car-

bohydrate, iron, total mineral content, calcium, and phosphorus also has


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






BIOAVAILABILITYF DIETARY RON 129

a lower ron absorption(68). The eason for the higher bioavailabilityiron in humanmilk is unknown. hedifferences in bioavailability of humanmilk and cow’smilk have also been demonstrated ndirectly by comparingcalculateddifferences n total body ron in growingnfants fed different milk

regimens 78).Although tudies on the bioavailability of iron in single foods and the

early studies on the interaction of foods affecting iron absorptionwereof

great theoretical interest, as mentionedarlier, this kind of data did notaccuratelypredict the absorptionof iron from he total diet.

POOL CONCEPT AND CONCEPT OF BIOAVAILABILITY Some of the

non-heine ron compoundsn the diet are probablyonly partly soluble or

partly dissociated. Whenn inorganic radioiron tracer is added o a meal,an isotopic exchange etweenhe tracer and the native iron probablyoccursin a commonctive part of the pool--a pool of isotopic exchange. t isreasonable to assume hat the iron from this pool is transferred to the

transport system f the mucosal ells. Thus,differences n bioavailabilityofiron in different single foodsmay e considered s differences n the relativesize of their pool of exchange 33). Bymixing wo foods with a differentbioavailability of their non-heine ron, for instance eggs and white wheat

flour, the bioavailability of iron in the omeletmaderom hese foods s not

a simple meanvalue of the two foods 06).Anumber f dietary factors (see below)have been showno influence the

bioavailability of non-hemeron. Factors hat increase he absorption e.g.ascorbic acid) will incrase the non-heine ron absorption from all foodsincluded n a meal. Such actors maybe consideredas increasing the sizeor availability of the pool of exchange. actorsdecreasing ron absorption,suchas tea, phytates, or certain fibers, may educe he pool of exchange y

forming more insoluble or undissoeiated iron compounds.Thus, the

bioavailabilityof iron in a meal s not the sum f the absorption f iron fromthe single foodscontainedn a meal,but rather a net effect of all food tems,and heir constituents, increasingor decreasingnon-heineron absorption.

There s a differencebetween ioavailabilityof iron in a foodand sotopicexchangeability f its iron with an extrinsic tracer. For example, on-hemeiron in wheatbran or eggs (37) rapidly and completelyexchangeswithextrinsic tracer, but it has a low bioavailability (a small pool of isotopicexchange). n comparison, oods such as white wheatflour also have com-plete isotopic exchangewith the tracer, but this iron has a several-fold

higherbioavailability.Therehavebeenattempts o study the bioavailability of food ron in vitro

by measuringhe fraction of iron liberated by incubationwith hydrochloricacid or mixtures of hydrochloricadd and pepsin to simulate gastric juice


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






130 HALLBERG

(48, 83). In another study, the ionizable iron (i.e. the fraction that reacts

with o., tt’-dipyridyl) was measuredafter an initial peptic digestion at pH

1.35 followed by an increase of the pH o 7.5 to simulate duodenalalkalinity

(75). The validity of these manoeuvers equires further evaluation.

DIET FACTORS THAT PARTICULARLY NFLUENCEBIOAVAILABIL-

ITY OF NON-HEME RON

Ascorbic acid It was shownearly on that ascorbic acid or orange juice with

a high content of ascorbic acid markedly increases food iron absorption

(72). This effect is due to a promotion of non-heme ron absorption (2,

18, 23, 82), and there is no effect on the absorption of heme ron (45, 85).

The absorption increase is related to the amountof ascorbic acid. A signifi-cant effect has been observed with only 25 mg of ascorbic acid, which is the

amountpresent in a third of a glass of orange juice. In one study on the

absorption of iron from maize, the ratio of iron absorption was 2 and 6,

respectively, when the amount of ascorbic acid was increased from 25 to

200 mg 10). In another study, ascorbic acid was increased from 25 to 1000

mgand the absorption ratios were 1.7 and 9.6, respectively (27). Orange

juice containing 70 mg of ascorbic acid increased iron absorption from a

breakfast meal 2.5 times (77). The addition of cauliflower, which also

contains about 70 mg of ascorbic acid, to a vegetarian meal increased the

absorption of non-berne iron three times, from 0.32 to 0.98 mg 44). Similar

effects were observed in studies on meals containing rice, maize, wheat,

or soy (28, 80-82). In one study on corn flour meals, a six-fold increase

absorption was obtained both with 70 mgof ascorbic acid and with papaya,

which contains about the same amount of ascorbic acid (53).

In summary, ascorbic acid is a very potent promoter of non-heme ron

absorption. Crystalline ascorbic acid and native ascorbic acid present in

foods have about the same promoting effect. Cooking and baking can de-stroy the ascorbic acid and hence its effect on iron absorption (82). The

effect of ascorbic acid seems o be independentof the effect of other promot-

ers of iron absorption, such as meat. However, when two promoters are

present, the relative size of the effect of either will be smaller. In the

presence of an inhibitor of non-heme iron absorption, such as tea, the

relative enhancingeffect of ascorbie acid on iron absorption seems o be the

same. The absolute increase in amountof iron absorbed, however, is less,because of the lower original absorption (29, 77).

The effect of ascorbic acid on food iron absorption maybe related both

to its reducingeffect, preventing the formationof insoluble ferric hydroxide,and to its effect on forming soluble complexes with ferric ions, which

preserve the iron solubility on the more alkaline duodenal pH (19, 21).


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






BIOAVAILABILITY OF DIETARY IRON 131

Meat andfish An enhancing effect of meat and fish was first reported byLayrisse et al in 1968 (56). This observation has since been confirmed

manyother studies (23, 26, 38, 41, 42, 49, 53, 60, 61). The absorption-

promoting ffect of meat is dose related (44). Several investigators have tried

to clarify the mechanism f this meat effect (11, 26, 39, 60). It is not due

to protein per se, as egg albumin has no promotingeffect, nor is it due to

the aminoacids present in meat. It is unlikely that the effect is related tothe content of nucleoproteins, as calf thymus, which is exceptionally rich

in such proteins, has no greater promoting effect than does beef (11).

special interest is the observation that meat has about the samepromoting

effect on the absorption of heme and non-heme ron (39; see below).

Tannates Recently it was reported that tea markedly reduced the absorp-

tion of non-heme ron from foods. The absorption from bread was reduced

to one third and from a vegetable soup to one fourth whenserved with tea

comparedwith water (29). In a Western-type breakfast, the absorption was

reduced about 60%by tea (77). This effect has been ascribed to the forma-

tion of iron tannate complexes 20). It has also been suggested that tannins

maybe partly responsible for the low bioavailability of iron in many egeta-

ble foods (30). Tannatesare also present in coffee. In a study on the effect

of various drinks on the absorption of non-heme ron from a hamburgermeal, tea reduced the absorption by 61%and coffee reduced it by 33% 44).

It is possible that the inhibiting effect of coffee is due to tannates.

Phytates, phosphates, andfibers Several studies have shown that sodium

phytate decreases iron absorption in man(45, 66, 85). The lower fraction

of iron absorbed from brown bread compared with white has been at-

tributed to the high content of iron phytates in bran (71). In a study on wheat

bread containing various amountsof bran, it was found that the iron absorp-

tion decreased in relation to the increasing bran content. Most of the

phytate, however, was broken down during leavening and baking of the

bread, with a corresponding ncrease in content of phosphate (9). The finalcontent of phosphate in the wheat bread was not of such a magnitude that

it could explain the decrease in iron absorption with increasing amountsof

bran. It is possible, as has been suggested, that the inhibiting effect of bran

is partly due to its content of fiber components9).Monoferric phytate, prepared from wheat bran, has been found to have

a high bioavailability for the rat (73). Phosphates have been shownanimal studies to reduce iron absorption; however, recent studies in man

indicate that a reduction is only observed whenboth calcium and phosphate

are added (69). Studies on the effect of various fiber materials areprogress, but so far no reports on studies in manhave been published.


A

n n u .

R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 . D o w n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e v i e w s . o r g

b y E T H N O L O G I S C H E S S E M

I N A R o n 0 1 / 1 3 / 0 8 .







132 HALLBERG

In summaryeveral conflicting results have been reported on the effectof phytates, phosphates, nd fibers on the bioavailability of non-hemeron.Further research is badly needed n this area, considering he importanceof cereal and vegetable ron for iron nutrition in man.

Egg, milk, oxalate, succinate, and cystein Eggs have been reported todecrease the absorption of iron froma breakfast meal (31). Eggyolk wasfound o decrease he absorptionof iron froman inorganic iron salt givento rats (17). In a recent study in which he effect of various componentsbreakfasts were compared,eggs were found to cause a decrease in thepercentage absorption of non-heme ron, but the actual amountof ironabsorbed increased slightly (77) due to the higher iron content of the

breakfast containingegg iron.Milk has been found to decrease iron absorption from meals with a low

bioavailability. In a recent study of a hamburger eal,however, o decreasein iron absorption was seen whenwater was exchanged or milk (44).

In earlier reviewson factors that influence iron absorption, oxalate hasoften been mentioned s decreasing ron absorption (19). Recent tudiesmeals served with and without extra oxalic acid (100 mg) have failedshow uch an effect (44). Succinic acid, which ncreases the absorptioniron from pharmaceuticaldoses of iron, had about the same35% bsorp-

tion-promotingeffect on dietary non-heine ron in a standard hamburgermeal, whengiven in an amountof 150 mg(44). For cystein, divergentresults have been obtained: Some tudies showed o effect (11), whereasother studies have shown marked bsorption-promotingffect (60). Inrecent study, an effect wasonly obtained f cystein wasaddedafter cookingthe food, an observation hat may xplain the divergent results, and whichmaybe related to the oxidation of cystein into cystine by cooking 64).

In summary, everal factors enhanceor inhibit non-hemeron absorp-

tion. Continued tudies on diets from different parts of the world willprobablydentify additional factors.

BIOAVAILABILITY OF IRON IN MEALS FROM DIFFERENT COUN-

TRIES To compare he bioavailability of non-heme ron in a variety ofwholemealsstudied by different investigators it is necessary o relate theiron absorption from hese meals to somecommontandard, preferably, asmentioned bove, to the absorption froma dose of ferrous ascorbate. Thissection is limited to studies in which eference dose absorption has been

measured.Results obtained by different workershave been recalculated tocorrespond o a reference dose absorption of 40%.Theassumption as beenmade hat there is a linear relationship between he absorptionfrommealsand reference doses and that the regression lines go through he origin.


A n n u .

R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w

n l o a d e d f r o m a r j o u r n a l s . a n n u a l r

e v i e w s . o r g



o n l y .






BIOAVAILABILITY F DIETARY RON 133

Table1 shows hat bioavailability of non-hemeron in mealsstudied indeveloping ountries. In Venezuela,mealscharacteristic of three differentareas werestudied. Meals epresentative of the central areas werecomposed

mainly f cereals (maize)and vegetables, with the addition of 50 g of meatat lunch. Meals rom he Andes ad a very similar composition.Meals romthe coastal area contained 50-60 g of fish and 150 g of papaya 60-70 mgof ascorbic acid at supper) (53). The mportance f fish, meat, and ascorbicacid for the iron absorption s evident fromTable 1.

Thai meals were simple rice-based containing boiled vegetables andcurry. Themealswereprepared rom carefully washed oods to ensure thatthey werefree fromcontaminating ron fromdirt and dust. This explains

why he iron content of this diet was ower han is usually reported n foodsurveys rom his area. Thehigher iron content and higher iron absorptionfrom he third mealshowneflects a higherenergy ntake (40). Theadditionof fish to the Thai meal markedlyncreased the non-hemeron absorption(38, 41, 42). TheBurmesemeals also demonstrate he great importancefish on ron bioavailability 3).

The bioavailability of non-hemeron from different types of Westernbreakfastmeals s graphed n Figure1 (77). There s a sixfold differenceabsorption between hese breakfast meals, even though the iron content

only varied from2.8 to 4.2 mg. The owest absorption wasobtained froma continental breakfast with tea (0.07 mg), and the highest came romcontinental breakfast with coffee and orange uice (0.40 mg).

Studies of lunch and dinner meals have also shown marked ariationin iron bioavailability. Table 2 shows he non-hemeron absorption from

nine dishes containing fish or meatand from four dishes with negligibleamounts f meator fish. In the first groupof nine meals he two mealswiththe highest absorptionfigures (0.81 and 0.90 mg)had one thing in common,namely high acidity. ThepHof the borscht soup and the sauerkraut meal

were5.2 and 3.6, respectively. In the vegetarianmeals, he veryhigh absorp-tion from he one containingcauliflower could reflect its high content ofascorbic acid (74 mg). n total, these 13 mealsshowed seven-folddiffer-ence n bioavailability of their non-hemeron (44).

All 10 meals llustrated in Table3 havean energycontent of about 1000

kcal. In these meals, here wasan almost ix-fold difference n bioavailabil-ity (44).

In summary, ll these studies demonstrate he marked nfluence by the

composition f the mealson the bioavailability of the non-hemeron.

BIOAVAILABILITY OF FORTIFICATION IRON Two factors determine

the bioavailability of the iron used o fortify foods: the composition f the

meals n which he mainpart of the fortification iron is included, and he


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






"


A

n n u .

R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3

- 1 4 7 .


b y E T H N O L O G I S C H E S S E M I N A R o n 0 1 / 1 3 / 0 8 . F o r p e r s o n a l u s e o n l y .






I I I


A n n u . R

e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e

v i e w s . o r g



o n l y .







A n n u .

R e v . N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .


b y E T H

N O L O G I S C H E S S E M I N A R o

n 0 1 / 1 3 / 0 8 .

F o r p e r s o n a l u s e o n

l y .






138 HALLBERG

Basal reakfastwithdifferent rinks

coffea

milkcho¢olate.......................

coffeaBasa~eeakfastith

no addition

and ornflakes milk

oneegg

one gg~- bacon

orangeuice

Basalbreakfastwith

teano addition

andone egg÷ bacon

orangeuice

0 0.i0 o. o o. o.

Absorption (mg)

Figure 1 Absorption of non-heme iron from different breakfast meals. The values represent

absorption n subjects with a 40% bsorptionfromoral reference dosesof ferrous ascorbate

(3 mgof Fe) (77).

properties of the iron compoundsed. For example, f flour is used as thevehicle for iron fortification and bread is mostlyeaten at breakfast, the

composition f the breakfast will have a determiningnfluence on the extraamount f iron absorbed s a result of iron fortification. Thus, he bioavaila-bility of fortification iron is determined y the composition f the meals nwhich he fortification iron is includedand not by the vehicle carrying the

iron (25, 33, 38, 52).

The ron fortification compoundsed mayonly be partially dissolvedunder he conditionsprevailing in the gastrointestinal tract. Reducedron,for instance, varies greatly in bioavailability. This variation is mainly ueto differences n rate of dissolution,whichn turn seemso be closelyrelatedto surface area per unit weight 14, 25). Verysoluble iron salts, suchferrous sulfate, completelymixwith the non-heine ron pool. Sodiumronpyrophosphate nd ferric orthophosphate re muchess available for ab-sorption (25), and indeed ferric orthophosphate,which s frequently usedby the food industry, varies markedlyn bioavailability between ifferentcommerciallyvailable preparations (44).

Thebioavailability of ferric orthophosphate, hen sed o fortify salt, canbe markedly mprovedby the addition of sodiumacid sulfate and sodiumhexametaphosphate76). Thecost of fortification, however,will increase


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .







significantly, and urthermoret will not be possible to simultaneouslyfortify the salt with odine.

Sodiumron ethylenediaminetetraacetateNaFeEDTA)s an iron com-

pound ith promisingroperties or iron fortification. Comparisonf equalamountsf elemental ron added o identical mealsshowedhat more ronwas absorbed roma meal fortified with NaFeEDTAhan with ferrous

sulfate (55, 65, 88). Furthertudies are neededo evaluate ts advantages

as an ron fortificantand ts possible nteractionwith he absorptionf othertrace elements n the diet. Field studies are currentlyunderwayn Guate-malafor NaFeEDTAortification of sugar and in Thailand for NaFe-EDTAortification of fish sauce, which s widelyused n SoutheastAsia.

Theeffect expected f an iron fortification programan be measuredn

several ways.By using radioisotopes t is possible to measure owmuchextra ron is absorbedt different evels of iron fortificationfrommeals nwhich he vehicle of iron fortification is frequently onsumed.his kindofstudyhas beendonewithdifferent diets. Asshownn Figure , results fromdifferent studies emphasizehe markednfluenceof different mealson the

bioavailability of fortification iron. All studies shownn Figure2 wereperformedith ferrous ulfate as the fortificant (38).

Absortion increasemg

1.2- at 40% ef.dose bs.)

1.0-

0.5

o.310.2

0.I

15mgAmountf iron added

to meal

oThaibasal meal

oThaibasalmealwith fish

¯ Swedish ixedmeal

~ Swedish eal: no meat,lowas¢orbic cidcontentAUSmealwith meat/cook}

,~US meal nomeatetal

x Maizemeal,Layrisse t al

Figure 2 Absorption ncrease at different levels of iron fortification for different types of

meals. Iron was added s ferrous sulfate in all studies [Thai studies, Hallberget al (38);

meals, Cook t al (25); Swedishmeals, Bj/Srn-Rasmussent al (14); and maizemeals, Layrisse

et al (52)].


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






140 HALLBERG

Anothermethodor evaluating he bioavailability of fortification iron is to

study the effect of a fortification program n the change n hemoglobinndplasma erritin levels in a population. t is strongly recommendedhat sucha field fortification trial be carried out beforea nationalfortification pro-gram s implemented90). To evaluate the effectiveness of the program,

fortification trial has to be run for a long time, probably coupleof yearsat least, to obtain statistically significant hemoglobinhangesn comparisonto a control group. Thereare several reasonsfor the lowsensitivity of suchtrials--first, probablyonly a small segment f the populationhas an iron-

deficiency anemia nd can be expected o respond; second, a great propor-tion of these individuals have just mild anemiawith only a small potentialresponse; hird, the amount f extra iron administereds usually small; and

further, in anemic ubjects the treatmentof iron deficiency is accompaniedby increased iron losses when he hemoglobinevels increase, for example,by increased menstrual ron loss.

Clearly, with the extrinsic tag methodt is possible o make preliminary

evaluation f the feasibility of fortifying a certain typeof diet with iron andto estimate the extra daily amount f iron needed o obtain a reasonableeffect (36). Radioisotopetudies also providea bases or designing ortifica-tion trials by delineating an appropriate dose of iron compound,nd byindicating the observationperiod needed 90).

BIOAVAILABILITY OF HEME IRON

Individual Factors

Asmentioned n the section on measurementf bioavailability of dietaryiron, the absorptionof heme ron into the mucosal ells is independent fnon-berneron. It is taken up into these cells as an iron-porphyrin omplex,

which s split in the mucosal ells by a specific enzyme86). Non-hemeron

and heme iron then seem to have a commonathwayout of the mucosalcells into plasma.This difference in absorptionmechanismf the twokindsof iron maybe the probable reason whyheme ron absorption is much essinfluenced by the iron status of the subject. If a meal contains a normalamount f heme ron (5 mgor less), there is little if any influenceof thesubject’s iron status on heine iron absorption.Ifa very large amount f hemeiron is eaten, for instance in a meal containing bloodsausage, whichmaycontain as much s 50 mgof heme ron, there is a definite relationshipbetween ron status and heme ron absorption (39). The same s true when

solutions containing only a few milligramsof heme ron are given withoutfoodon an empty tomach45, 85). In contrast to inorganic iron, patholog-ical conditions hat affect the small bowelmucosa, uchas coeliac disease,do not significantly effect the absorptionof hemeron (1). Similarfindingshave been reported in patients with Billroth II partial gastrectomy 46).


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






BIOAVAILABILITYOF DIETARY IRON 141

Dietary Factors

With one exception, factors known o influence the absorption of non-heme

iron such as aseorbie acid or phytates do not affect heine iron absorption.

Meat greatly facilitates the absorption of both non-heine and heme ron (39,61). The absorption-promoting effect of meat seems to be of the same

magnitude or both kinds of iron. This fact suggests that there is a common

mechanism f action of meat on the absorption of both heme and non-heme

iron (39). The bioavailability of heme ron in foods prepared from blood but

given without meat is much ower than if meat itself is included.

Mostof the iron in liver is heme ron (about 70%)and the remaining 30%

is mainly ferritin and hemosiderin. A clear relationship between ron status

and iron absorption from liver, containing 2-4 mgof iron, has been reported(62). The bioavailability was about the same as from veal. Maizeeaten with

liver reduced the absorption of liver iron, indicating an inhibiting effect of

maize on the absorption of non-heme iver iron.

Fromstudies that use different dose levels of heme ron it appears that

the bioavailability of heme ron in meals containing meat is about 25%, and

the bioavailability of heme ron given without meat or liver has a maximum

absorption of about 10%, decreasing with increasing dose to a few percent

(39).

ASSESSMENT OF BIOAVAILABILITY OF IRON INTHE WHOLE DIET

Earlier Studies

The first attempts to measure he bioavailability of iron in the whole diet

were made with the chemical balance method. As mentioned, this method

is both insensitive and inaccurate. However, rom a few diets, somecare-

fully done studies indicated that about 10%of dietary iron was generally

absorbed [see Moore 71)].Indirectly, the absorption of iron from the whole diet can be assessed by

the rate of regeneration of hemoglobin in subjects with iron-deficiency

anemia. To calculate iron absorption from repletion studies, an observation

period of several months s required and it is necessary to know he basal

iron losses and to exclude pathological iron losses in the subjects studied.

A few of these indirect studies have been reported (20, 32, 74). In a recent

study in Sweden on healthy men in whom ron-deficiency anemia was

induced by repeated phlebotomies (74), the rate of hemoglobin recovery

indicated an average total absorption of iron from the diet of 3.8 rag/day.

Since the basal losses of iron are about 1 mg/day n men, the total amount

of iron absorbed during this repletion period was 4.8 rag/day. The Swedish

group have shown hat the absorption of food iron is about twice as high

in subjects with iron-deficiency anemia as in non-anemicsubjects without


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






142 HALLBERG

iron stores (57). Figuresof bioavailabilityof dietary iron obtained n anemicsubjects measuring he hemoglobinegeneration rate therefore should be

halved o be comparable ith the figures shownn TablesI-3. Byextrapola-

tion from the Swedishphlebotomy tudy, the average bioavailability ofdietary iron for non-anemic ubjects with no iron stores wouldbe 2.4mg/day.

Anassessment f the bioavailability of dietary iron can also be maderom

a large sampleof menstruatingwomen y establishing the upper limit oftotal iron losses that can be balancedby absorptionof iron from he diet

without the developmentf iron-deficiency anemia. In a random opulationsampleof about 500 Swedishwomen,his upper limit was 1.7 mg/day 43).Theaverage iron intake in these women as 12 mgand the energy intake

was about 1900kcal.The otal daily absorption of dietary iron has also been measuredn 32

young nlisted menwith the extrinsic tag method, abeling both heine andnon-hemeron with twodifferent radioiron isotopes. Breakfast, lunch, anddinner were served as homogenizeduddingsof a mixture of foods repre-sentative of an average6-week ood ntake. Themean otal daily absorptionwas1 mg 12). Recentstudies have shown hat the absorption of ironlower from homogenizedhan from identical non-homogenized eals (41).Therefore t is possible that the figure obtained s an underestimate f the

true total absorption.Layrisseet al (53) measuredhe total daily dietary iron absorption y the

extrinsic tag method rom hree Venezuelan iets. Onone day the non-heine

absorption wasmeasured rom breakfast and lunch by using 55Feand 59Fe,and 2 weeks ater from supper by using saFe. Theheme ron absorption wascalculated. These esults are shown n Table 1.

Implications of the Pool Concept

Currentknowledgebout food iron absorption mplies that the bioavailabil-ity of iron in a diet dependsnot only on its content of heme nd non-hemeiron, but to a large extent on the balancebetweenactors that stimulate andinhibit the absorptionof iron. These acts wereconsidered o some xtent

by a WHOroup of experts whomet in 1972and whoadvised that dietaryiron requirements e adjusted according o the content of animal foods inthe diet (89). Recently,a modelwasdevelopedor the estimationof avail-able dietary iron considering both the content of heme nd non-hemeronand he presenceof enhancing actors such as meatand ascorbic acid (70).With he new nformation bout he bioavailability of iron in different meals

and especially about the very marked ariation betweenmeals, moreup-to-date modelsneed o be developed or the calculation tff bioavailability ofdietary iron.


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .







Estimations Based on Bioa.vailable Nutrient Density

Thecritical iron balance ituation today, especially in womenf childbear-ing age, reflects to a large extent the present low food ntake and hus the

difficulty of absorbingsufficient iron per unit of energy consumed. heamount f a nutrient per unit energy n the diet--the nutrient density--is

a poor measure f the adequacy f iron in the diet because of the markedvariation n bioavailability of iron in different meals.A more seful measurewould e the amount f iron absorbed roma meal n relation to its energycontent. Since this bioavailablenutrient density mustbe described n terms

of the subjects’ ron status, the following efinition for iron has beenpropo-

sed: The amountof iron absorbed (in milligrams) from a meal per unitenergy 1000kcal) by subjects who re borderline ron deficient [i.e. absorb

40% rom reference doses of iron (35)].As mentioned arlier, fromstudies of hemoglobinegenerationrate (74)

the iron absorptionat a state of borderline ron deficiencywasestimated s2.4 rag/day. As he energy ntake was2675kcal/day, the averagebioavaila-ble nutrient density of this Swedish iet was0.9 mgof Fe/1000 cal. In thestudy on menstruatingwomen,he calculated bioavailable nutrient density ’from the same ype of diet was also 0.9 mgof iron/1000 kcal (1.7/1900).For the three Venezuelan iets (53), the corresponding igures would

0.81, 0.69, and 1.03. The atter represents the coastal area, whichdiffersfrom he others by containingappreciableamounts f fish (50-60 g) in allmeals, including breakfast.

Thenutritive value of a diet for a certain nutrient mustbe basedon its

ability to meet he requirements f certain target groups. To meet he ironrequirements in 90%of women f childbearing age, 2.2 mgof iron mustbe absorbed aily (95). If the averageenergy ntake is about2000kcal, theaverage bioavailable nutrient density mustreach 1.1 mgof Fe/1000kcal.Onsuch a diet women ith borderline iron status will not have any iron

stores, but they will not be anemic.Thebioavailable nutrient density ofdifferent meals s shownn Tables1-3; a very marked ariation is indicated.

It is possible to estimate the averagebioavailablenutrient density of aparticular diet by recording he meals consumeduring a certain period oftime and calculating the mean ioavailable nutrient density from he energycontribution of each meal. Ourrough calculation from a typical Swedishdiet gavean averagebioavailablenutrient density of 0.9. This figure corre-spondswell with iron bioavailability measured y the hemoglobinegenera-

tion rate studies in Swedishmenand the iron balance calculations inmenstruating Swedishwomen. he bioavailable nutrient density conceptcan also be used in the evaluation of diets in developing ountries and npredicting he effects of iron fortification or other measureso increase he

bioavailabilityof iron.


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .






144 HALLBERG

CONCLUSIONS

Within a fairly short period of time knowledge about the bioavailability of

dietary iron has grown rapidly. With the new pool concept of food iron

absorption, we are beginning to understand the interaction between differ-

ent foods in the absorption of iron and the importance of the meal composi-

tion for the bioavailability of iron. The pool concept has made us recognize

that an increased iron absorption can be achieved not only by increasing the

iron content of the diet, but also by increasing the content of items that

facilitate the absorption or by reducing items that inhibit the absorption of

iron. It is hoped future research in this area will find effective and realistic

ways of improving the bioavailability of dietary iron and hence iron nutri-

tion, especially in developing countries where iron deficiency is most severeand most prevalent.

ACKNOWLEDGMENT

This review was supported in part by grants from the Swedish Medical

Research Council B81-19X-IM721-06A.

Literature Cited

1. Anaud,B. S., Callender,S. T., Warner,G. T. 1977. Absorptionof haemoglobinand norganic iron in eoeliae disease.Br. Z tlaemat. 37:409-14

2. Apte, S. V., Venkataehalam,. S. 1965.Theeffect of ascorbic acid on the ab-sorption of iron. Ind. Z Med. Rex53:1084-86

3. Aung-Than-Batu,Thein-Than, Thane-Toe. 1976.Iron absorptionfromSouth-east Asian dee-based meals. Am. J.Clin. Nutr. 29:219-25

4. Bezwoda,W. R., Bothwell, T. H., Tor-

rance, J. D., Mac hail, A. P., Charlton,R. W., Kay, G., Levin, J. 1979. Therelationship between marrow ronstores plasma erritin concentrationsand iron absorption. Scand. £ Ha-ematol. 22:113-20

5. Bczwoda,W., Charlton, R., Bothwell,T., Torrance, J., Mayet,F. 1978. Theimportance f gastric hydrochloric cidin the absorptionof nonhemeood ron.J. Lab. Clin. Med.92:108-16

6. Bczwoda,W. R., Disler, P. B., Lynch,S. R., Charlton,R. W.,Torrance, . D.,

Derman,D., Bothwell, T. H., Walker,R. B., Mayer, . 1976. Patterns in foodiron absorption n iron deficient whiteand Indian subjects and in venesectedhaemochromatoticatients. Br. £ Ha-emat. 33:265-76

7. BjSrn-Rasmussen,E. 1973. Foodironabsorption n Man. II. Effect of ironsalt, aseorbie acid anddesferrioxamineon the isotopic exchange etween ativefood iron and an extrinsic inorganiciron tracer. Scand. Z ttaeraatoL 11:391-97

8. BjiSrn-Rasmussen, . 1973. Food ronabsorption n man. V. Validity of theextrinsic tag two-poolmethodor mea-surementof dietary non-heine ron inpatients withvarious linical disorders.Scand. Z Gastroenterol. 8:645-50

9. Bjfrn-Rasmussen, E. 1974. Iron ab-sorption from wheatbread. Influenceofvarious, amounts f bran. Nutr. Metab.16:101-10

10. BjSrn-Rasmussen, E., Hallberg, L.1974. Iron absorption from maize.Effect of ascorbicacid on iron absorp-tion from maize upplemented ith fer-rous sulphate. Nutr. Metab.16:94-100

11. Bj~irn-Rasmussen, E., Hallberg, L.1979. Effect of animalproteins on theabsorption of food iron in man.Nutr.

Metab. 23:192-20212. Bj6rn-Rasmussen,E., Hallberg, L.,Isaksson, B., Arvidsson,B. 1974. Foodiron absorption n man.Applicationofthe two-pool extrinsic tag method omeasure hemeand non-heme ron ab-


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w


b y


R o n 0 1 / 1 3 / 0 8 .


o n l y .



BIOAVAILABILITY OF DIETARY IRON 145

sorption from the wholediet. Z Clin.Invest. 53:247-55

13. Bj~rn-Rasmussen,E., Hallberg, L.,Magnusson,B., Rossander, L., Svan-

berg, B., Arvidsson,B. 1976.Measure-mentof iron absorptionfromcompositemeals. Am,~ Clin~ Nutr, 29:772-78

14. Bj6rn-Rasmussen,E., Hallberg, L.,Rossander,L. 1977. Absorption f for-tification iron. Bioavailabilityn man fdifferent samplesof reduced ron, andpredictionf the effectsof iron fortifica-tion. Br. J. Nutr. 37:375-88

15. Bj~irn-Rasmussen,E., Hallberg, L.,Walker,R. 1972. Food ron absorptionin man. I. Isotopic exchange etween

food iron and inorganic ron salt addedto food: studies on maize, wheat andeggs. Am.J. Clin~ Nutr. 25:317-23

16. Bj~Sru-Rasmussen, ., Hallberg, L.,Walker, R. 1973. Isotopic exchange firon between abelled foods and be.twecna food and an iron salt. Am,J.Clin. Nutr. 26:1311-19

17. Callender, $. T., Marney, $. R.,Warner,G. T. 1970. Eggsand iron ab-sorption. Br. £ Haemat.19:657-65

18. Callcnder, $. T., Warner,G. T. 1968.Iron absorption from bread. Am.J.

Clin, Nutr. 21:1170-7419. Conrad,M. E. 1970. Factors affecting

iron absorption. n Iron Deficiency,Pa-thogenesis, Clinical Aspects, Therapy,ed. L. Hallberg, H. G. Harwerth, A.Vanotti, pp. 87-114. NewYork: Aca-demic

20. Conrad,M.E. Crosby,W. H. 1962. Thenatural history of iron deficiency in-duced by phlebotomy.Blood20:173-85

21. Conrad,M.E., Sehade,S. G. 1968.As-eorbie acid chelate n iron absorption:role for hydrochloriccid andbile. Gas-troenterology 5;35--45

22. Cook, . D., Brown,G. M., Valberg,L.S. 1964.Theeffect of achylia gastricaon iron absorption. :. Clin. Invest. 43:1185--91

23. Cook, J. D., Layrisse, M., Martinez-Torres, C., Walker, R., Monsen,E.,Finch, C. A. 1972. Food iron absorp-tion measured y an extrinsic tag. J.Clin. Invest. 51:805-15

24. Cook, . D., Lipsehitz, D. A., Miles, L.E. M., Finch, C. A. 1974.Serumerritin

as a measure f iron stores in normalsubjects. Am.J. Clin. Nutr. 27:681-8725. Cook, . D., Minnich,V., Moore,C. V.,

Rasmussen,A., Bradley, W. B., Finch,C. A. 1973. Absorptionf fortificationiron in bread. Am. J. Clin. Nutr.26:861-72

26. Cook,J. D,, Monsen, . R. 1976. Foodiron absorption n humanubjects. IlLComparisonf the effect of animalpro-teins on non-heineron absorption.Am.J. Clin. Nutr. 29:859-67

27. Cook, . D., Monsen, . R. 1977. Vita-min C, the commonold and iron ab-sorption. Am.J. Clin. Nutr. 30:235-41

28. Disler, P. B., Lynch, . R., Charlton,R.W., Bothwell, T. H., Walker, R. B.,Mayet, . 1975.Studieson he fortifica-tion of canesugar with iron and aseorobic acid. Br. £ Nutr. 34:141-52

29. Disler, P. B., Lynch,S. R., Chadton, .W., Torrance, J. D., Bothwell, T. H.1975. Theeffect of tea on iron absorp-tion. Gut16:193-200

30. l)isler, P. B., Lynch, . R., Torrance,.D., Sayers, M, H., Bothwell, T. H.,Charlton, R. W. 1975. The mechanismof the inhibition of iron absorptionbytea. S. Aj’k J. Meal.Sc~40:109-16

31. Elwood,P. C. 1968. The role of foodiron in the prevention of nutritionalanaemias. In Occurrence, CausesandPrevention f NutritionalAnaemias,d.G. Blix, pp. 156--65.Uppsala:Almqvist&Wiksell

32. Finch, S., Hasldns, D., Finch, C. A.

1950. Iron metabolism.Hematopoiesisfollowingphlebotomy.ron as a limit-ing factor. ,L Clit~ Invest. 29:1078-86

33. Hallberg, L. 1974. Thepool concept nfood iron absorption and someof itsimplications. Proc. Nutr. Soc. 33:285-91

34. Hallberg, L. 1980. Food ron absorp-tion. In Methodsn Hematology,d. J.D. Cook. London:Churchill

35. Hallberg, L. Submitted or publication36. Hallberg, L. 1981. Iron nutrition and

food iron fortification. Cli~ Haematol.

In press37. Hallberg, L., Bj/Srn-Rasmussen, E.1972. Determination f iron absorptionfrom wholediet. A new wo-poolmodelusing two radioiron isotopes given asharmand non-harmron. Scand. J. Ha.ematol. 9:193-97

38. Hatlberg, L., Bj/Srn-Rasmussen,E.,Garby, L., Pleehachinda,R., Suwanik,R. 1978. Iron absorption from South-East Asiandiets and he effect of ironfortification, Am. £ Clin. Nutr. 31:1403-8

39. Hallberg~ L., Bj/Sru-Rasmussen,E.,Howard,L., Rossander, L. 1979. Die-tary hemeron absorption.A discussionof possible mechanismsor the absorp-tion-promotingeffect of meat and forthe regulation of iron absorption.Seand.J. Gastroent.14:769-79


A n n u . R e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w n

l o a d e d f r o m a r j o u r n a l s . a n n u a l r e v i e w s . o r g

b y E

T H N O L O G I S C H E S S E M I N A R

o n 0 1 / 1 3 / 0 8 .







146 HALLBERG

40. Hallberg, L., BjSrn-Rasmussen,E.,Rossander, L. 1979. The measurementof food iron absorption n man.Ameth-odologicalstudy on the measurementfdietary non-haem-Fe bsorption whenthe subjects havea free choice of fooditems. Br. J. Nutr. 41:283-89

41. Hallberg, L, Bjrrn-Rasmussen, E.,Rossander, L, Suwanik,R. 1977. Ironabsorption fromSoutheast Asiandiet.II. Roleof various factors that mightexplain low absorption. An J. Clir~Nutr. 30:539-48

42. Hallberg, L., Garby, L, Suwanik,R.,BfiJrn-Rasmussen,E. 1974. Iron ab-sorption from Southeast Asian diets.Am. J. Clin. Nutr. 27:826-36

43. Hallberg,L., H/Sgdahl,A.-M.,Nilsson,L., Rybo,G. 1966. Menstrual lood oss--a population study. Variation atdifferent ages and attempts to definenormality. Acta Obstet. Gynec.Scand.45:320-51

44. Hallberg,L., Rossander,L., Bj~irn-Ras-mussen,E. Unpublished tudies

45. Hallberg,L., SiSlvell, L. 1967.Absorp-tion of hemoglobinron in man. 4etaMed. Scand. 181:335-54

46. Hallberg,L., SiSlvell, L., Zedeffeldt,B.1966.Iron absorption fter partial gas-

trectomy. A comparativestudy on theabsorption from ferrous sulphate andhaemoglobin.4eta Med.Scand. Suppl.445:269-75

47. Heinrich, H. C., Briiggeman,,, Gabbe,E. E., Gliiser, M.1977.Correlationbe-tween diagnostic ~gFe absorption andserumferritin concentration in man.Zeitschrift Naturforsch. Tell c 32:1023-25

48. Jacobs, A., Greenman, D. A. 1969.Availability of food iron. Br. Med.J.

1:673-7649. Layrisse, M., Cook, . D., Martinez,C.,Roche,M., Kulm, . N., Walker,R. B.,Finch, C. A. 1969. Food ron absorp-tion: Acomparisonf vegetableand an-imal foods. Blood33:430-43

50. Layrisse, M., Martinez-Torres, . 1971.Food ron absorption: Iron supplemen-tation of food. Ping.Itetnatol. 8:137-60

51. Layrisse, M., Martinez-Torres, . 1972.Model or measuring ietary absorptionof heme ron: test witha completemeal.Am. Z Clin. Nutr. 25:401-11

52. Layrisse, M., Martinez-Torres, C.,Cook, J. D., Finch, C. A. 1973. Ironfortification of food: Its measurementby the extrinsic tag method. Blood41:333-52

53. Laydsse, M., Martinez-Torres, C.,Gonzales, M. 1974. Measurementf the

total daily dietary iron absorptionbythe extrinsic tag method.Am.Z Clin.Nutr. 27:152-62

54. Laydssc, M., Martinez-Torres, C.,Renzi,M., Leers, I. 1975.Ferritin ironabsorption in man. Blood45:689-98

55. Laydsse, M., Martinez-Torres, C.,Renzi, M., Velez, F., Gonzales, M.1976.Sugar s a vehicle or iron fortifi-cation. Am.J. Clin. Nutr. 29:8-18

56. Layrisse, M., Martinez-Torres, C.,Roche,M.1968. Theeffect of interac-tion of various oodson ron absorption.Am. Z Clin. Nutr. 21:1175-83

57. Magnusson, ., Bj/Srn-Rasmussen, .,Rossander,L., Hallberg, L. Manuscriptin preparation.

58. Magnusson, ., Faxdn, A., Cederblad,~., Rossander,L., Kewenter, ., Hall-berg, L. 1979.Theeffect of parietal cellvagotomy nd selective vagotomywithpyloroplastyon iron absorption.Scand.J. Gastroent.14:177-82

59. Magnusson, . E. O., Hallberg, L., Ar-vidsson, B. 1976.Iron absorptionafterantrectomy with gastroduedenostomy:Iron absorption rom ood and from er-rous ascorbate. Scand. J. I-laematol.Suppl. 26:69-85

60. Martinez-Torres, ., Layrisse, M.1970.

Effect of amino cids on iron absorptionfrom a staple vegetable food. Blood35:669-82

61. Martinez-Torres, ., Layrisse, M.1971.Iron absorption from veal muscle. Am.J. Clin. Nutr. 24:531--40

62. Martinez-Torres,C., Leers, I., Renzi,M., Layrisse, M.1974. Iron absorptionby humans rom veal liver. J. Nutr.104:983-93

63. Martinez-Torres, C., Renzi, M., Lay-rissc, M. 1976.Iron absorptionby hu-mansfrom hemosiderin and ferritin.Further studies. J. Nutr. 106:128-35

64. Martinez-Torres,C., Romano, ., Lay-risse, M.1981,Effectof cystine on ronabsorption n man.Am.J. Clir~ Nutr. Inpress

65. Martinez-Torres, C., Romano,E. L.,Kenzi, M., Layrisse, M.1979.Fe (III)-EDTAomplexas iron fortification.Further studies. Am. Z Clin. Nutr.32:809-16

66. McCance,R. A., Edgecombe,C. N.,Widdowson,E. M. 1943. Phytie acid

and iron absorption. Lancet 2:126-2867. McMillan, . A., Landaw,S. A., Oski,F. A. 1976. ron suflieieneyn breast-fedinfants and he availability of iron fromhumanmilk. Pediatrics 58:686-91

68. McMillan, . A., Oski, F. A., Lourie,G., Tomardli, R. M., Landaw,S. A.


A n n u . R

e v .

N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e

v i e w s . o r g



o n l y .






A n n u .

R e v

. N u t r .

1 9 8 1 . 1 : 1 2 3 - 1 4 7 .

D o w n l o a d e d f r o m a r j o u r n a l s . a n n u a l r e v i e

w s . o r g

b y E T

H N O L O G I S C H E S S E M I N A R o n 0 1 / 1 3 / 0 8 .


bioavailability of dietary iron in man

Documents