Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Food science

Iron absorption from maize bread
Delivery of oral doses of vitamin A to prevent vitamin-A deficiency and nutritional blindness


Iron absorption from maize bread

Carlos Martinez-Torres, Peter Taylor, Irene Leets, Eleonora Trapper, Jos Ramrez, and Miguel Layrisse

Maize is the main food of the lower-socio-economic populations living in Central America, the southern part of North America, and the northern part of South America. In some regions it represents more than 80% of daily food consumption [1]. iron absorption from whole maize in several culinary preparations has been shown to be low [215], less than 2% in healthy persons and less than 5% in irondeficient persons [14]. Iron absorption from maize is in general lower than that from other cereals [16]. This article provides a comprehensive review of published data on maize iron absorption, as well as information collected during the past four years, with special emphasis on iron absorption from precooked maize flour eaten alone and with meals.

Materials and methods

The data are from 547 healthy volunteers living in rural areas of Venezuela. Results from 438 participants have been published in several articles [3, 6-15]. The remaining subjects were studied over the past four years. Blood haemoglobin concentration [18] and unsaturated transferrin saturation [19] were measured in all volunteers, and serum ferritin concentation [20] was determined in 321.

Preparation of foods

Maize bread

Four types of maize bread were prepared: whole-maize and polished-maize arepas (thick pancakes ), Mexican tortillas, and arepas made from precooked maize flour. The whole-maize bread was prepared by boiling the grain in aluminium pots containing four times the weight of the corn in water. After the grain became swollen and soft, a dough was made and rolled into balls of about 100 g each. These were slightly flattened into very thick cakes and grilled. For the polished-maize bread, the grain was freed of pericarp and then treated in the same fashion as the whole maize. For tortillas, whole maize is boiled in water containing 6 g quicklime per 100 g maize, formed into thinner pancakes, and grilled. Precooked maize flour is produced by an industrial procedure in which the grain is first liberated from the pericarp and the germ, and the endosperm is then exposed to humid heat followed by dry heat. The flour requires only the addition of water to prepare the dough, which is then grilled before being eaten.


Four typical breakfasts consumed by adults from low-socioeconomic populations of Latin America were tested. Breakfast no. 1 is the diet of the Andean and central regions of Venezuela. It contains about 100 g whole-maize bread accompanied by 100 or 200 ml milk and a small portion of cheese and butter. Number 2, consumed by the population of Carabobo and Yaracuy states of Venezuela, contains about 200 g precooked maize bread accompanied by 100 or 200 ml milk, coffee, and a small portion of cheese and butter. Number 3, eaten in So Paulo, Brazil, and Lima, Peru, consists of about 60 to 100 g white wheat bread. It is accompanied by 250 ml milk and coffee in Lima, and by coffee only in So Paulo. Number 4 is typical in Santiago, Chile, and Salta Province, Argentina. About 100 g white wheat bread is eaten with tea. Details of these breakfasts are given elsewhere [7, 8, 1 0, 1 5]; their nutrient contents are shown in table 1.

Standard meals

In our iron-absorption studies over the last 16 years we have frequently tested two types of standard meals consumed by persons of low socio-economic status, one without and one with meat. The first standard meal consists of three vegetable foods, in the following amounts after cooking: 180 g rice, 125 g maize bread, and 90 g black beans. This meal prepared with whole-maize bread contains 450 calories, 15 g protein, 3 g fat, 90 g carbohydrate, 2 mg ascorbic acid, and 4.7 mg non-haem iron. The same meal, but using bread made from precooked maize flour, has essentially the same quantities of nutrients except that the iron content is only 3.9 mg.

The other standard meal contains beef as hamburger and a smaller proportion of the vegetable foods. The amount of each food, after cooking, was as follows: 70 g beef, 90 g rice, 63 g maize bread, and 60 g black beans. The nutrient content of the meal prepared with whole maize is 355 calories, 1 6 g animal protein, 9 g vegetal protein, 7 g fat, 49 g carbohydrate, 1.6 mg ascorbic acid, 1.2 mg haem iron, and 3.6 mg non-haem iron. With precooked maize, the amount of non-haem iron is reduced to 3.2 mg.

Radioactively labelled material

Some of the iron-absorption tests used maize intrinsically labelled with 55Fe or 59Fe, and some used extrinsically labelled maize in which radioactive iron was mixed with the dough. Radioactive wheat bread was prepared by mixing 59Fe or 55Fe with the dough.

Reference dose

The reference dose of ferrous sulfate was prepared according to a previous publication [2]. It contained 3 mg of iron as ferrous sulfate, and ascorbic acid at a 2:1 ascorbate: iron molar ratio.

Iron absorption

All the meals were eaten in the morning after an overnight fast. No food or drink was allowed for three hours after the meal. Approximately 0.7 Ci of 59Fe or 2 Ci of 55Fe was used in each meal. In general, one meal was given on day 1 and another meal on day 2, labelled with different isotopes. Blood was drawn 15 days after the administration of the meals to determine the haematological profile of the subjects and to measure radioactivity. The subjects were fed again with another meal on day 15 and the reference dose of ferrous sulfate on day 16.

TABLE 1. Nutrient content of breakfasts tested for iron absorption

Breakfast Diet Energy
Fat (g) Carbo-
1. Bread from
whole maize,
milk a
Venezuelan Andes,
average consumption
309 3.4 6.8 14.4 48.1 1 1.64
Central Venezuela
average consumption
429 6.1 10.2 15.9 52.1 2 2.9
2. Bread from
maize flour,
milk, coffee b
Carabobo State,
Venezuela, adults
606 10.1 7.9 11.7 100.8 2.8 1.41
Carabobo State,
Venezuela, average
730.9 7.3 8.2 24.8 11.8 1.9 1.52
Yaracuy State,
525.3 8.4 7.3 11.8 100.8 1.0 2.21
3. Bread from
white wheat
flour, coffeeb
So Paulo State, Brazil,
279 0.0 2.9 8.8 47.4 0.0 1.86
Lima, Peru, adults 530 0.0 17.1 16.7 78.4 2.4 2.34
4. Bread from
white wheat
flour, tea, sugarb
Salta Province,
Argentina, adults
280 0.0 5.0 0.5 65.0 0.0 0.90
Santiago, Chile, adults 342 0.0 6.4 4.8 70.0 0.0 2.6

a. Nutrient contents of breakfasts of Andes and Central Venezuela diets [8, 9] recalculated according to Venezuelan nutrient content tables.

b. Details of ingredients given in previous publications [11, 15].

FIG. 1. Comparison of iron absorption from the reference dose of ferrous sulfate and from whole-maize bread in 276 subjects

Blood was drawn on day 30 to measure radioactivity. Duplicate 10-ml samples together with the radioactive food and reference standard dose were prepared for counting following the technique of Dern and Hart [21, 22]. iron absorption from food was calculated using an estimation of blood volume based on sex, weight, and height [23]; no correction was made to determine total iron use.

Chemical analysis

The total iron content of the vegetable foods and meat was determined by the digestion method [24], non-haem iron by the method of Miller et al. [25], phytate by the method of Wheeler and Ferrell [26], and tannate by the method of Price and Butler [27].

Statistical analysis

The mean and standard error of the values of absorption and serum ferritin concentration were calculated using logarithms. The comparison of absorption of the two meals performed used Student's t test for paired data when the meals were given to the same

individuals, and the test for unpaired data when the meals were given to different subjects. In both cases the t values were calculated from the logarithms of the absorption data.


Iron absorption from whole maize

Figure 1 shows the comparison of iron absorption from whole-maize bread prepared as arepa and from the reference dose of ferrous sulfate. The 276 subjects tested were grouped according to absorption from the reference dose. Absorption from maize iron ranged from 1.2% to 2.3% in subjects whose absorption from the reference dose was less than 35%. There was an abrupt increase from 2.3% to 4.5% in maize iron absorption between subjects whose absorption from the reference dose was 30.0%-34.9% and those with absorption of 35.0%-39.9% (p< .02). Maize iron absorption remained at the same level in the subjects with reference dose absorption levels of 40.0%-59.9%. There was a further increase, although not statistically significant, in those whose absorption from the reference dose was above 60%.

Iron absorption from the reference dose was 35% or greater in 114 of 276 subjects tested, and, of those, only 15 had transferrin saturation lower than 16%. As the majority of the absorption tests from whole-maize iron were performed before 1975, when we started to determine the serum ferritin concentration, only 50 subjects were tested for this variable. Six of 26 subjects whose absorption from the reference dose was 35% or greater had a ferritin concentration below 12 g per litre.

Iron absorption from various kinds of maize bread

Table 2 shows the iron absorption from various kinds of maize bread. Data from the two studies were normalized according to iron absorption from the reference dose of ferrous sulfate. The arepa made from precooked maize flour showed the highest absorption, about 100% higher than from the Mexican tortilla, about 40% higher than from arepa made from whole maize, and about 30% higher than from arepa made from polished maize.

The phytate content of the different kinds of maize showed a relationship with the absorption data. The phytate content of whole-maize grain, polished maize, and precooked maize flour was 772, 537, and 251 mg per 100 g respectively. The phytate content of the Mexican tortilla, whole-maize arepa, polished-maize arepa, and precooked-maize arepa was 565, 318, 213, and 120 mg per 100 g respectively.

Iron absorption from breakfasts

Table 3 shows iron absorption from four different breakfasts, which varied according to the type of bread and beverages consumed. Due to the large number of subjects tested, it was possible to divide them into three categories according to their iron status:
1. Normal: Characterized by a normal haemoglobin concentration, serum transferrin saturation above 16%, serum ferritin concentration above 12 g per litre, and absorption from the reference dose less than 35%.
2. Moderate iron deficiency: The same characteristics as normal but with iron absorption from the reference dose greater than or equal to 35%.
3. Severe iron deficiency: Normal or low haemoglobin concentration, serum transferrin saturation 16% or below, serum ferritin concentration 12 g per litre or less, and generally iron absorption from the reference dose above 35%. There was no statistically significant difference in iron absorption between severely iron-deficient subjects and those with normal or low haemoglobin.

The highest level of non-haem iron absorption was observed in those who ate breakfast no. 2: 3.5%, 6.3%, and 10.0% in subjects in the normal, moderate iron deficiency, and severe iron deficiency categories respectively. The worst breakfast was no. 4, iron absorption being significantly reduced in relation to the other breakfasts in subjects in all categories of iron status. In each breakfast, subjects with moderate iron deficiency showed a twofold to threefold higher rate of absorption than healthy persons, and in three breakfasts of four, there was no significant difference in absorption between subjects with moderate and severe iron deficiency.

The non-haem iron content of the breakfasts tested was relatively low-between 0.9 and 2.9 mg. For example, in the case of breakfast no. 2, consumed by adults in Yaracuy State in Venezuela, in which the iron intake was 2.2 mg, the iron absorbed was 0.08, 0.14, and 0.23 mg in persons with normal haemoglobin, moderate iron deficiency, and severe iron deficiency respectively. For breakfast no. 4, consumed by adults in Santiago, Chile, in which the iron intake was 2.6 mg, the iron absorbed was 0.02, 0.07, and 0.08 mg respectively.

Because the consumption of precooked maize flour has increased in the last 10 years in various Latin

TABLE 2. Absorption of iron from various kinds of maize bread

  Study 1
(22 subjects)
Study 2
(13 subjects)
Mean SE Mean SE
Blood haemoglobin concentration (g/100 ml) 13.2 0.6 13.2 0.5
Serum transferrin saturation(%) 23 2 29 2
Serum ferritin concentration(g/litre) 19 1 30 1
Absorption of iron from maize (%)a
A. whole-maize tortilla 2.8 1.2 -  
B. whole-maize arepa 4.0 1.3 3.6 1.2
C. precooked-maize-flour arepa 5.8 1.2 6.7 1.2
D. polished-maize arepa -   5.5 1.2
Absorption of iron from
reference dose of ferrous
sulfate (%)b
29.7 1.2

a. Normalized according to the absorption of iron from the reference dose by each individual and that by all subjects tested in the two studies.

b. Mean and standard error of absorption by all individuals in the two studies. The significances of the differences in level of iron absorption between the different breads were as follows in study 1:
A vs. B. p<.05 A vs. C, p<.01 B vs. C, p<.05
In study 2: B vs. C, p<.01 B vs. D, p<.05 C vs. D, p<.05

American countries, two studies were carried out enriching the breakfast with eggs and ascorbic acid (table 4). The mean iron absorption from the basic breakfast from the two studies was 6.3%. Enriching the breakfast with one or two eggs did not significantly alter non-haem iron absorption. In the second study, enrichment with 30 mg of ascorbic acid led to a threefold increase. Enrichment of the basic breakfast increased the total iron availability from 0.13 mg to 0.23 and 0.39 mg when consumed with one or two eggs respectively, and to 0.42 mg with 30 mg of ascorbic acid.


Contents - Previous - Next