Contents - Previous - Next

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

Contaminant iron in water and soil samples and in vitro availability studies - G. Oluyemisi Latunde-Dada


The iron content from 24 well water samples in tropical southern Nigeria ranged from 0.05 to 3.1 µg per millitre. Iron extracted from soil samples in deionized water, 0.1 M NaOH, 0.1 N HCI, and 0.1 N HCl-ascorbic acid averaged 79. 7, 146.3, 809.8, and 1,268.1 µg of iron per 100 g of soil sample respecfively. Extraction of iron from the soil samples was significantly affected (p < .05) by the addition of ascorbic acid. The dialysable iron in some meals containing contaminant iron from the samples compared favourably with a standard meal with ferric chloride. Dialysable iron was influenced significantly (p<.05) by the components in the meals. Meat and ascorbic acid enhanced the dialysability of contaminant iron sources in both the semi-synthetic meals and some tropical food items. The contribution of these contaminant iron sources to the diet of Nigerians in relation to the high frequency of iron-deficiency anaemia is discussed.


A significant proportion of the iron intake of people in developing countries is accounted for by the contaminant iron in their diets [1]. Contaminant iron can cause iron overload, as in the case of the siderosis in the Bantus of South Africa, who derive excess dietary iron from a traditional fermented maize beverage brewed in iron pots [2]. The iron in readily available reduced ionic state is absorbed through the gastrointestinal tract. The iron content of foods cooked in iron utensils increases significantly [3-5]. The bioavailability of iron from apples with nails inserted to rust [6] and from food cooked in iron pots [7] is relatively high.

Contaminant iron in the diet of people in developing countries may also be derived from particles of soil adhering to food items that are not properly washed before consumption [8]. In Ethiopia a significant proportion of the high iron intake from the cereal teff (Eragrostis tef) has been found to be accounted for by contamination with soil during threshing under the hooves of cattle [9].

Generally, agricultural products in these countries are exposed to significant contamination with soil particles during processing and retailing. Moreover, these food items are often rinsed in water obtained chiefly from wells that supply drinking water for the rural populace. Both the water and soils may contain iron. We attempted to quantify the amount and in vitro availability of this mineral in samples of soil and well water obtained from locations in south-west Nigeria.

Materials and methods


Water and soil samples

Duplicate samples of water were taken from 24 wells in different locations at Ago-lwoye, Ogun State, Nigeria. The water was filtered and stored for subsequent analysis.

Ten samples of soils from different farm lands were also taken at a depth of six inches below the surface. These samples were sieved to remove the stones, oven-dried at 50°C to constant weight, and stored in polythene bags for analysis.

Iron was extracted from triplicate 5-g soil samples by washing three times with a total volume per sample of 50 ml of a solvent containing 0.02 g of ascorbic acid in 100 ml of 0.1 N HCI, and the extracted iron was analysed for soluble iron. Iron extracted from the soil samples in the solvent was concentrated for the in vitro dialysable iron assay.

TABLE 1. Composition of the experimental meals (grams per 100 g)



Meal 2


Meal 4











Egg white










Soya oil





















Semi-synthetic meals were prepared using refined ingredients as described by Miller et al. [10] (table 1). Tropical food items such as maize, yams, cassava, and plantains were purchased from a local market.

Reagents for in vitro digestion studies

A solution was prepared from 16 g of pepsin powder (porcine stomach mucosa, Sigma Chemical Co., Poole, Dorset, UK) in 100 ml 0.1 M HCI pancreatinbile extract mixture. Four grams of pancreatin (porcine pancreas, Sigma Chemical Co.) was suspended in 0.1 M sodium bicarbonate and the volume brought to 1 litre with 0.1 M sodium bicarbonate.

A solution was prepared from 100 g of trichloracetic acid, 100 g of hydroxylamine HCI, and 100 ml of concentrated HCI dissolved in deionized water and brought to a volume of 1 litre.

A solution was prepared from 250 mg of bathophenanthroline sulfonate dissolved in 2 M sodium acetate and brought to 1 litre with 2 M sodium acetate.

Analytical methods

Meal samples were homogenized in 50 ml of deionized water. The pH of the homogenates was adjusted to 2 with 6 M HCI and the final weight adjusted to 100 g with 0.01 M HCI. Extracted and concentrated iron from the soil samples was added to the meals at a level of 1 mg per 100 g and the final weight adjusted to 100 g with 0.01 M HCI. The standard meal contained ferric chloride at the same concentration.

Three millilitres of the pepsin solution was added to the homogenized sample and incubated in a shaking water bath for two hours at 37°C. Twenty-gram aliquots of the pepsin digest were transferred to conical flasks. Segments of dialysis tubing containing 25 ml of deionized water and an amount of sodium bicarbonate equivalent to the titratable acidity (determined according to Miller et al. [10]) were put into each flask, and the flasks were incubated in the shaking water bath for about 30 minutes or until the pH was 5. Five millilitres of the pancreatin-bile extract mixture was then added, and the incubation continued for another two hours.

After pancreatic digestion, the contents of the dialysis tubes were quantitatively transferred to a 25 ml volumetric flask and made up to volume with deionized water.

Dialysable iron (%) =

(m g iron per ml dialysate x 25 x 100)/[m g iron per ml sample x weight of sample (g)]

The in vitro calorimetric analysis of dialysable iron as an estimation of availability described by Miller et al. [10] has the advantages of cost, speed, and reduced variability, and correlates well with human in viva studies [11].

Meal samples were dry-ached, and iron was estimated calorimetrically as described by Schricker et al. [11], modified to use ferrozine (R) colour reagent (3-(2-pyridyl)5, 6-diphenyl-1, 2, 4-triazone-P, P-1-disulfonic acid, Sigma Chemical Co. ) as described by Carter [12]. Total iron concentration was calculated from a standard curve of ferric chloride solutions. All glassware was rendered ironfree by soaking in 6 N HCI and rinsed thoroughly with deionized water before use.

Statistical analysis was done using analysis of variance. The least significant difference (LSD) test [13] was applied when treatment F was significant at the 5% level of probability.

Results and discussion

The iron content from the 24 well water samples ranged from 0.05 to 3.1 m g per millilitre of water. The iron content of tap water from four sources averaged 0.06 m g per millilitre; that from rain water was negligible (table 2). The well water therefore contributes contaminant iron to the diet of those who consume it.

The solubility of the iron from the 10 soil samples in deionized water ranged from 25 to 140.7 m g per 100 g of soil (table 3). For some of the samples, the solubility increased minimally in 0.1 M NaOH (pH 11.5). There were, however, significant increases in the solubility in 0.1 N HCI 1.5 and 0.1 N HCI-ascorbic acid solution.

The stable forms of iron in an aqueous environment are the ferric form, Feł+, and its reduced ferrous form, Fe2+. These are hydrated as Fe(H2O)ł+ 6 and Fe (H2O)˛+ 6 in solution. The corresponding insoluble iron hydroxides, FeOH2 and FeOH3, are formed as pH is increased. At a neutral pH of 7, Fe2+ precipitates and has a solubility of about 10-t M, whereas Feł+ has a solubility of about 10-6 M [14]. This probably explains the low solubilization of iron in well water (pH 6.8-7.5) and from the soil samples in deionized water. The increased solubility of iron from some of the soil samples in the alkaline medium was similar to the observations of others [15, 16].

TABLE 2. Iron content of well, tap, and rain water (micrograms per millilitre)





of variation

Well watera





Tap waterb





Rain water





a. Samples from 24 wells.
b Samples from 4 sources.

At room temperature and on oxidation in the presence of alkali, Fe2+ can convert to the insoluble and stable oxyhydroxide FeOOH, but Feł+ cannot form the stable FeOOH or Fe2O2 without heat. Therefore, Feł+ forms a less stable intermediate that is more susceptible to resolubilization in an alkaline medium [16]. Ascorbic acid can solubilize Feł+ by forming complexes and/or by reduction to Fe2+ [17]. The chemical reduction of Feł+ to Fe2+ by ascorbic acid has been well documented in food and model systems [18].

In general, the total extractable iron from the soil samples was much lower than the value of 2.27 g per 100 g from dry clay soil [19]. Variation in mineral solubility can be due to the soil type, the season of the year, the amount of organic matter, and the presence of various chelates [20]. The soils sampled for the current study during the planting season were loamy.

TABLE 3. Iron extraction from soil samples (micrograms per 100 g)



0.1 M

0.1 N

0.1 N HCI





























































Values are means of triplicate sampling.

The dialysable iron from the synthetic meals was significantly affected by the addition of contaminant iron (fig. 1). The dialysability of soil iron compared favourably in some cases with the standard meal containing ferric chloride. The variability of dialysable iron from the soil samples could be accounted for by the chemical forms of iron and the presence of intrinsic chelates. Dialysable iron also increased significantly in the presence of ascorbic acid.

Dialysable iron was significantly enhanced in some food items when ascorbic acid was added. This enhancement was increased when contaminant iron from soil sample 10 was added to the food. The exogenous iron from the contaminant soil was mixed with the intrinsic food iron and was thus subjected to the enhancing and inhibitory influences of food components. Interpretation of the results from composite meals becomes complicated because of the contribution of iron from the various food sources and the effects of varying ratios of iron and of enhancers and inhibitors.

FIG. 1. Dialysable iron from meals containing iron from contaminant sources. The meals are: A, with casein; B. with casein and ascorbic acid; C, with egg white; and D, with meat. The bars marked 0 represent the standard ferric chloride meals, and those marked 1, 2, 4, 5, 7, and 10 represent meals containing iron from the correspondingly numbered soil samples listed in table 3. The small vertical lines represent the least significant differences.

The concept of a common pool of non-haem iron [21] from the meals and the contaminant source was measured calorimetrically. studies have found exchangeability of contaminant iron with a radio iron pool of 70% [19] and of 0%-35% [8], depending on the soil type.

The contribution of contaminant iron to the iron nutrition of people in developing countries is therefore strongly influenced by the presence of enhancers of iron availability. Unfortunately, these are often not appreciably consumed by a significant proportion of the people. Hallberg and Bjorn-Rasmussen also concluded that contaminant iron could be a good source of dietary iron [8]. Similarly, the iron added to food by cooking in iron utensils is as available as native food iron [5].

It is evident that enhancers of iron availability, such as the haem iron of meat and ascorbic acid, play important roles in iron nutrition, and low intake of these is one of the main reasons iron deficiency is prevalent in developing countries despite the relatively high intake of dietary and contaminant iron.


The funding of this study by the International Foundation for Science (IFS), Sweden, is gratefully acknowledged. The assistance of Mr. Tunde Odusanya in the collection of samples is also appreciated.


1. Galan P. Cherouvrier F. Zouhoun 1, Zouhoun T. Chauliac M, Hercberg S. Iron absorption from typical west African meals containing contaminating Fe. Br J Nutr 1990;64:541-46.

2. Gordeuk VP, Bacon BP. Brittenham GM. Iron overload: causes and consequences. Ann Rev Nutr 1987; 86:485-508.

3. Burroughs AL, Chan JJ. Iron content of some Mexican-American foods. J Am Diet Assoc 1986;86: 897-99.

4. Brittin HC, Nossaman CE. Iron content of food cooked in iron utensils. J Am Diet Assoc 1986;86:89799.

5. Mistry AN, Brittin HC, Stocker BJ. Availability of iron from food cooked in an iron utensil determined by an in vitro method. J Food Sci 1988;53(5):1546-48.

6. Rosanoff A, Kennedy BM. Bioavailability of iron produced by corrosion of steel in apples. J Food Sci 1982;47:609-716.

7. Martinez FE, Vannucchi H. Bioavailability of iron added to the diet by cooking food in an iron pot. Nutr Res 1986;6:421-27.

8. Hallberg L, Bjorn-Rasmussen E. Measurement of iron absorption from meals contaminated with iron. Am J Clin Nutr 1981;34:2808-05.

9. Hofvander Y. Hematological investigations in Ethiopia, with special reference to a high iron intake. Acta Med Scand Suppl 1968;494:1-74.

10. Miller DD, Schricker BR, Rasmussen RR, Van Campen D. An in vitro method for estimation of iron availability from meals. Am J Clin Nutr 1981;34:224856.

11. Schricker BR, Miller DD, Rasmussen RR, Van Campen D. A comparison of in vivo and in vitro methods

for determining availability of iron from meals. Am J Clin Nutr 1981 ;34:2257-63.

12. Carter P. Spectrophotometric determination of serum iron at the submicrogram level with a new reagent (ferrozine). Anal Chem 1971 ;40:450-52.

13. Steel RGD, Torrie JH. Principals and procedures of statistics. 2nd ed. New York: McGraw Hill, 1980.

14. Lee K, Clydesdale FM. Iron sources used in food fortification and their changes due to food processing. Crit Rev Food Sci Nutr 1978;11:117-53.

15. Saltment P. Hegenauer J. Christopher J. Tired blood and rusty livers. Ann Clin Lab Sci 1976;6: 167-76.

16. Eyerman LS, Clydesdale FM, Huguenin R. Zajicek OT. Characterization of solution properties of four iron sources in model systems by solubility studies and IR/ VIS reflectance spectrophotometry. ] Food Sci 1981; 52(1):201-09.

17. Clydesdale FM. The effects of physiochemical properties of food on the chemical status of iron. In: Nutrition: bioavailability of iron. American Chemical Society, 1982:5583.

18. Rizk SW, Clydesdale FM. The effect of ascorbic acid, pH and exagenous iron on the chemical iron profile of a soy protein concentrate. J Food Biochem 1984;8:9195.

19. Hallberg L, Bjorn-Rasmussen E, Rossander L, Suwanik R. Iron absorption from southeast Asian diet: 11. Role of various factors that might explain low absorption. Am J Clin Nutr 1977;30:539-48.

20. Elwood PC. Iron content of soil samples. Proc Nutr Soc 1965;24:112.

21. Hallberg L. The pool concept in food of iron absorption and some of its implications. Proc Nutr Soc 1974;33:285-91.

FNB note Errors in early infant nutrition in the world and in Czechoslovakia: The risk of development of hypertension in adolescence and adulthood - Rudolf Kohn

The optimal diet of newborn mammals, including humans, is maternal milk. The nature of the placenta and its permeability for nutrients and other substances differs considerably among mammals. This applies to nutrients and other naturally occurring substances in the diet and to infant metabolism. It is also true for substances that may have an unfavorable effect, e.g. trace elements, some of which are extremely important for the development of the individual but are toxic in excess. In this brief note I cannot deal with contaminants associated with civilization, although some of them have a very adverse impact on the development of individual subjects.

Under the influence of Czerny's and Finkelstein's German doctrine, we have used a very inadequate substitute for breast milk: diluted cow's milk. Nevertheless, it had a favourable effect on the reduction of infant mortality and infantile diarrhoea. At the time, we were not aware of the adverse sequelae associated with this diet.

The sodium load, as well as the load of other salts, was excessive. The total protein load was also too large. Moreover, human milk contains albumin, while in cow's milk casein predominates. Many authors have provided unequivocal evidence of the excessive sodium load of formula-fed infants. As evidence I am presenting only an illustration (fig. 1). A cow's-milk-formula diet causes a considerable increase in sodium intake. On a mean daily intake of 10 g of sodium chloride in adults (recent recommendations are 6 g day) the intake per kilogram of body weight is only 2 milliequivalents of sodium. The osmotic load of a formula-fed infant is thus double the load of adults. If given soup, the infant's osmotic load may be even treble. This means that at the age of about four months our infants have the highest salt intake in their lives.

To this we have to add, under conditions in Czechoslovakia and other eastern European countries, adverse ecological conditions. Because of the high nitrate content of drinking water, it was recommended to dilute tap water in Prague and elsewhere with mineral waters at a 1:1 ratio. This increased the salt load even more. In Germany water with a low sodium content was recommended.

I maintain that this sodium load in early infancy is excessive. It is especially unfavorable in genetically disposed infants and those with infections or who are otherwise debilitated. In the case of renal immaturity it can be one of the main causes of risk for the development of hypertension during adolescence and adulthood. The kidney in early infancy is an immature excretory organ which can easily reach the borderline of its excretory capacity in the case of a large supply of protein-breakdown products and salts.

I believe that the contemporary high morbidity and mortality from cardiovascular diseases in Czechoslovakia and other countries is evidence of errors we have committed.

FIG. 1. Intake of sodium by infants from diets of cow's milk formula and human milk (according to Janowsky)

Contents - Previous - Next