Contents - Previous - Next

3. Dietary intakes of children in developing countries

The diets of disadvantaged peoples of the developing world vary from one community to another, but in many regions the main components are cereals and plant foods. Consumption of animal milks and meat is often limited (Bansal et al., 1964; Walker, 1972; Prentice, 1991). It would be predicted, therefore, that the Ca and Zn intakes in these communities would be low. Intakes of P and Mg would be expected to be relatively high as cereals and plants are rich sources of both these minerals (Department of Health, 1991). Culinary practices, such as the use of plant ash [e.g. Nyasaland Bantu (Walker, 1994); American Indians, Papua New Guineans (Golden, 1988)], lime in the making of tortilla [Mexico (Golden, 1988; Cerqueira et al., 1979)], and dried baobab leaf in preparation of steamed millet [The Gambia (Prentice et al., 1993)] could make valuable contributions to mineral intakes in some societies.

In general, only 30-40% of dietary Ca and Zn are absorbed from typical Western diets, while absorption of P and Mg is somewhat higher at 40-60% (Department of Health, 1991; Schwartz, 1990). Ca absorption from predominantly plant-based diets is thought to be considerably reduced due to the chelation properties of compounds such as phytates (inositol hexaphosphates) and oxalates (Allen, 1982).

The Ca:P ratio, per se, appears to be relatively unimportant (Department of Health, 1991) except that in developing countries a low Ca:P ratio is likely to imply a high phytate intake. Zn absorption also appears to be affected by phytate, dietary fibre and by geophagia (King & Turnlund, 1988; Prasad, 1991). An important consideration in Zn bioavailablity is thought to be the phytate and Ca content of the diet, with a high Ca X phytate product being associated with an increased formation of insoluble complexes (Bindra, Gibson & Thompson, 1986; Sandstead, 1991; Xu et al., 1992). However, a number of studies have suggested that in practice the effects of food phytates on Ca absorption may not be large (Walker, Fox & Irving, 1948; Bronner et al., 1954; 1956; Irving, 1964) and there is controversy over the effect of phytates on Zn absorption (Hambidge, Casey & Krebs, 1986; Anderson, Gibson & Sabry, 1981). Hydrolysis of phytate occurs in the human intestine (Xu et al., 1992) and phytate P is well absorbed particularly when Ca intakes are low (Walker, Fox & Irving, 1948; Irving, 1964). In addition, methods of food preparation may be important in diminishing the effects of phytate, such as action of phytases during baking, fermentation and germination (Golden, 1988; Murphy, Beaton & Galloway, 1992; Irving, 1964). Other components of the diet can increase or reduce mineral bioavailability by altering absorption or excretion, for example protein, salt and other minerals (Department of Health, 1991; Allen, 1982) and high intakes of non-haem iron have been shown to depress Zn absorption (Hambidge, Casey & Krebs, 1986; Solomons & Jacob, 1981). Gastrointestinal illnesses, common in children in many developing countries, can have deleterious effects on mineral absorption.

The extent to which individuals habituated to low mineral intakes have adapted to their diets is unknown. Positive balances have been reported in children and adults with low Ca and Zn intakes (Begum & Pereira, 1969; Ziegler et al., 1989; Hegsted, Moscoso & Collazos, 1952; Luyken & Luyken-Konig, 1961), indicating that absorption efficiency must be high and losses low. It has been suggested that colonic absorption, after degradation of fibre and phytate by bacteria, may play an important role in this context (Fraser, 1988a, b; 1991). Adaptation of Ca metabolism to diets containing high amounts of phytate has been observed (Walker, Fox & Irving, 1948). Mg and Zn restrictions are also accompanied by increases in absorption efficiencies and decreases in endogenous losses (Widdowson & Dickerson, 1964; Department of Health, 1991; Schwartz, 1990; Taylor et al., 1991).

The diets of children in developing countries may differ from those of adults. Many infants and toddlers are breast-fed for prolonged periods and receive specially-prepared weaning foods. Older children may receive preferential or reduced amounts of certain foods, such as milk or fish, relative to adults. It is, therefore, important to judge the likely adequacy of the mineral supply during childhood by using dietary data collected from children. However, there have been comparatively few detailed investigations of mineral intakes of infants and children in the Third World. A selection of studies are summarised in Table 3.

Table 3. Examples of measured mineral intakes of children in disadvantaged communities in developing countries

Mean values for boys and girls, seasons combined.
^a High Ca associated with consumption of tortilla.
^b Mineral intakes from breast-milk plus weaning foods.
^c Plus Prentice, Bates & Paul (unpublished).

The data demonstrate that, as predicted, Ca intakes of children in many disadvantaged communities do appear to be very low. The figures in Table 3 can be compared with Ca intakes of British toddlers and older children which average 600-700 mg/d and 700-900 mg/d respectively (Department of Health, 1989; Prentice & Paul, 1990). In contrast, the measured intakes of zinc are not dissimilar to those in affluent societies (Table 3; Sandstead, 1985; 1991), due to the fact that unrefined cereals and plant foods contain reasonably high amounts of zinc (Department of Health, 1991). The measured intakes of P are moderate to high in older children, resulting in low Ca:P ratios, 1:2 - 1:3 mg/mg. Estimates of P intakes of young children in Western countries are 800-900 mg/d (Fomon, 1974). Children who are exclusively or partially breast-fed have lower P intakes and the Ca:P ratio of their diet is higher (Table 3; see below). In the limited number of studies in which Mg ingestion has been quantified, intakes of children in developing countries appear to be of the same order of magnitude as adult intakes in the UK (Department of Health, 1991).

Breast-milk is a major dietary component for many infants and young children in the developing world. Breast-milk production by mothers from disadvantaged communities is similar in general to that of mothers in affluent societies and can be maintained for prolonged periods (Prentice et al., 1986). Mineral concentrations in mature breast-milk decline as lactation progresses in women from both developed and developing countries (Bates & Tsuchiya, 1990; Laskey et al., 1990; Laskey, Dibba & Prentice, 1991). This is particularly striking for Zn (Krebs et al., 1985). There is no evidence of regional variations in P, Mg and Zn concentrations in breast-milk (Hambidge, Casey & Krebs, 1986; Bates & Tsuchiya, 1990; Laskey, Dibba & Prentice, 1991), but Gambian women have been shown to have Ca levels that are 20% below those of British women, and low Ca concentrations have been reported from a number of other developing countries (Laskey et al., 1990). Whether this is a consequence of low maternal Ca intakes is not known. In general, average Ca, P, Mg, Zn intakes of exclusively breast-fed children at 3 months are approximately 200 ma, 100 ma, 23 mg and 1.8 mg respectively, although it is important to realise that there are wide variations in breast-milk mineral outputs between individual mothers (Bates & Tsuchiya, 1990; Laskey et al., 1990; Laskey, Dibba & Prentice, 1991).

The contribution that breast-feeding makes to mineral intakes during partial breast-feeding in the developing world has been little studied. A detailed investigation in a poor rural village in The Gambia, where children are breast-fed for about 2 years but receive weaning foods from 3 months, showed that breast-milk still provided 40%, 20% and 15% of their Ca, P and Zn intakes respectively at 15-18 months of age (Table 4; Prentice & Paul, 1990). In a Thai study, 58% of Ca intakes were provided by breast-milk for children at 1 year (Chusilp et al., 1992). The value of the contribution may well be greater as the bioavailability of minerals from breast-milk is higher than from a mixed diet. However, the mineral density of breast-milk is relatively low when compared with animal milks and with components of the adult diet (Table 5). Calculations using the Gambian data indicate that, for children over 12 months old, replacing breast-milk with an isocaloric amount of the other foods in their diet would not alter calcium intakes materially and would increase P and Zn intakes by 30% and 38% respectively (Prentice & Paul, 1990; Prentice, Bates & Paul, unpublished). Similar data from other communities in the developing world are not available but the measured intakes of breast-fed and not breast-fed children (Table 3) suggest that a similar pattern would be expected in other areas. Breast-milk, therefore, although making substantial contributions to the intakes of many children in the developing world, is not an especially rich source of these minerals.

Table 4. Contribution of breast-milk to measured mineral intakes in rural Gambian children

Data are from Prentice & Paul (1990) and Prentice, Bates & Paul (unpublished). Children in this community breast-feed on demand for approximately two years, with weaning foods introduced from about 3 months of age. The local diet is based on rice, millet, groundnuts, fish and leaves (Prentice et al., 1993).

^a Values are the percentage of total mineral intakes contributed by breast-milk at each age. The calculations take no account of the differing bioavailability of minerals from breast-milk and other foods.

Table 5. Mineral density of some Gambian foods

Values are mg/100 kcal. Data from Prentice et al. (1993), Prentice & Paul (1990) and Prentice, Bates & Paul (unpublished).

^a Composition of Gambian breast-milk [Bates & Tsuchiya (1990); Laskey et al. (1990); Laskey, Dibba & Prentice (1991)]

^b Mineral density of the diet at 12-18 months minus breast-milk.

^c Contains dried baobab leaf.

A comparison of measured intakes (Table 3) with theoretical deposition rates (Table 2) demonstrates that the Ca intakes of many children in developing countries are very close to the biological requirement. This is before making any allowances for losses in body fluids, for incomplete absorption or for illness. Measured P and Zn intakes of children who are no longer breast-fed are substantially greater than theoretical accretion rates (approximately 5-10 times). However, Zn supply may approach the biological requirement when losses and absorption are considered (Department of Health, 1991; Hambidge, Casey & Krebs, 1986). The Zn, and to some extent the P intakes during breast-feeding are close to the biological requirement particularly in older children receiving little supplementary food (Krebs & Hambidge, 1986; Sandstead, 1985). Mg intakes are considerably higher than theoretical accretion rates at all ages (approximately 10-50 times).

Contents - Previous - Next