1 Supplementation significantly increased height and weight gain at all ages;
² number of children. From Gopalan et al. (1973)

Inexplicably, neither a high-energy (300 kcal/day), low protein snack with or without a vitamin-mineral supplement containing iron and zinc, improved rates of height or weight growth in 1-5 year old stunted and wasted Thai children (Gershoff et al., 1988).

1.2. Protein

Many investigators have observed an improvement in the linear growth of stunted children after supplementation with protein-rich foods such as milk, either alone or with other foods or nutrients, while others have not (Beaton & Ghassemi, 1982). However, there is little information on the specific role of protein in linear growth, because low protein intakes are usually accompanied not only by low energy intakes but by inadequate amounts of important micronutrients that are contained in dietary protein sources. These include iron, zinc, copper, calcium and vitamin A.

In the INCAP longitudinal intervention study in Guatemala there was no additional benefit for child growth of providing supplements that contained good quality protein (as atole, a mixture of dry skim milk and cereal) compared to supplemental energy alone (provided in a non-protein beverage called fresco). Energy intake was the important predictor of linear height (and weight) increments (Martorell & Klein, 1980). However, it is not certain that the beneficial effects of the atole were due to its energy content alone, because the dried skim milk and cereal in this supplement probably provided higher amounts of some micronutrients compared to fresco (Allen, in press a).

An important point is that in the INCAP intervention study severe growth stunting occurred even when protein and essential amino acid intakes were probably adequate. While atole supplements did improve linear growth, the infants and children receiving this supplement remained severely growth-stunted even though the energy intake of some of them exceeded the FAD/WHO requirement values, and protein intakes were two to three times higher than requirements in the majority of children. The INCAP investigators have stated that "although effects on length were clear, they were less than might have been expected given the level of supplement provided" (Martorell & Habicht, 1986). This suggests that other nutrients than protein or energy were inadequate in the atole. In the Nutrition CRSP, linear growth faltering was common in Egyptian, Kenyan and Mexican preschool-children even though their protein and essential amino acid intakes were adequate (Beaton, Calloway & Murphy, 1992).

The studies described above suggest that linear growth faltering, or subsequent failure to recover, can occur when intakes of energy and/or protein are adequate. This raises the question of the extent to which micronutrient deficiencies impair linear growth of human populations.

1.3. Zinc

1.3.1. Identification of zinc deficiency in the Middle East The effects of zinc deficiency on human growth were first described in Iran and Egypt (Prasad et al., 1963; Prasad, Halstead & Nadimi, 1972; Ronaghy et al., 1974; Halstead et al., 1972). In these locations zinc deficiency was associated with high phytate, high fiber, low protein diets, and with hookworm and schistosomiasis (Halstead et al., 1972), situations which may also have created other micronutrient deficiencies. The earliest zinc intervention trials were not successful in improving the linear growth of school-age boys (Carter et al., 1969; Mahloudji et al., 1975). Similarly, Ronaghy et al., (1969) found no growth benefit of providing 28 mg zinc/day plus iron, or 40 mg zinc/day plus iron and food, to short prepubertal Iranian boys, although it did improve the rate of maturation of their genitalia.

Because the supplemental zinc may have been poorly absorbed when consumed with the high fiber, high phytate Middle Eastern diets, or the studies were not of sufficient duration, Ronaghy et al. (1974) increased the amount of supplemental zinc to 40 mg/day for 13 year old moderately growth-retarded Iranian boys. Supplementation produced significant increases in height (at 182 days and 536 days), weight (at 536 days) and bone age (almost doubled at 536 days). This supplement also included 10 g/day of egg white protein, corn oil, minerals and vitamins, which failed to improve linear growth in the absence of zinc.

1.3.2. Zinc and growth of children in industrialized countries During the 1970s and 1980s there were several studies of the impact of zinc supplementation on short children in the United States and Canada (Table 4). Hambidge's group in Denver, Colorado, reported that short middle-income children had low hair zinc concentrations, impaired taste acuity and poor appetite (Hambidge et al., 1971). Later (Hambidge et al., 1976) they found that the height of 37% of children age 3-5 years, enrolled in a school program for low-income children, was at or below the 10th centile. Of these, 74% were Mexican-Americans. Hair and plasma zinc were lower than in children from a middle-income group, suggesting that inadequate zinc nutriture was common in the low-income families. In 1983 a zinc supplementation trial was performed on children 2-6 years old, with height-for-age below the 10th centile of NCHS reference values, and meeting two out of three of the following criteria: low dietary zinc, low plasma zinc, and low hair zinc. They were given a placebo or 5 mg zinc per day for a year. The zinc produced a 10% increase in height velocity which was significant within 6 months in the boys and the group as a whole, an increase in hair zinc, but no change in plasma zinc. The supplement increased food (energy) intake from 1280 to 1880 kcal/day in boys (Krebs, Hambidge & Walravens, 1984), but had no significant effect on either the food intake or linear growth of the girls.

Gibson et al. (1989) also showed a beneficial effect of zinc supplements on Caucasian Canadian boys with low height centiles. A supplement of 10 mg zinc/day was provided to 5-7 year olds, for 12 months. Only boys with a low initial hair zinc concentration responded to the supplement with faster linear growth. Energy intake and weight gain were unaffected.

Table 4. Zinc supplements and growth in industrialized countries

M = males.
¹ Significant effect of treatment;
² treatment effect is significant only in males;
³ not significant;
⁴ only males studied.

In Colorado, approximately 5 mg supplemental zinc per day for 6 months improved the weight-for-age of infants and preschool-children (age 8-27 months, mean 15 months) with the failure-to-thrive syndrome (weight below the 10th centile) (Walravens, Hambidge & Koepfer, 1989). Linear growth was not improved by the supplements. Possible explanations, apart from a lack of zinc deficiency, include selection on the basis of low weight-for-age rather than low length, and the relatively short duration of the study.

This research has been extended to answer the question of whether the smaller size of breast-fed infants in industrialized countries might be in part due to the decline in breast milk zinc concentrations that occurs during the first 6 months of lactation. Walravens et al. (1992) supplemented 4-9 month-old infants who were primarily breast-fed and born to low-income, immigrant families in France. When 5 mg zinc/day was provided, both weight and length gain increased significantly during the 3-month period. Length-for-age Z scores of supplemented infants were above NCHS reference values throughout the 3 months, while those of the placebo group fell slightly below reference values after 3 months. While some substitution of breast milk for weaning foods with lower zinc content cannot be definitively ruled out, these data are intriguing and need to be replicated in other settings. The nutritional implications of these results are not clear, because breast milk zinc content does not seem to be affected by poor maternal zinc status, nor increased by maternal zinc supplementation (Karra et al., 1989).

One intriguing aspect of the U.S., Canadian and French studies is that there was no obvious reason for the children to be zinc deficient. The U.S. children consumed quite generous amounts of animal products and a limited amount of ethnic foods, with an average pre-supplement zinc intake of 5.2 mg/day compared to 6.5 mg/day by children in middle-income families. It is possible that the onset of zinc deficiency occurred much earlier, perhaps in infancy, or that the zinc requirements of these stunted individuals were unusually high.

1.3.3. Zinc and growth of children in developing countries Data on zinc interventions to improve growth in developing countries are summarized in Table 5. In China, 1-6-year-old children with poor growth benefited from 1-2 mg zinc/day supplemented for 6 months, by significantly increasing weight and length gain (Chen et al., 1985).

A zinc intervention trial has been completed recently on children living in the Ecuadorian Andes (Dirren et al., in press). A national survey in Ecuador had found the prevalence of linear growth retardation in the rural Andes to be 67%. In addition 69% of children between 6 and 11 months were anemic, vitamin A status was marginal, riboflavin status poor and serum zinc was < 65 mg/dl in 35% of rural children. All children between 12 and 30 months of age in a community were enrolled and given 10 mg zinc/day or a placebo for 15 months. Anemic children were treated with iron. There was a significant increase in height starting at 3-6 months and becoming stronger with time. Over the 15-month period boys grew 1 cm more in the supplemented vs placebo group (P < 0.001) but girls only grew 0.6 cm more (P < 0.07). The supplement significantly increased the weight Z score of girls but not boys.

Table 5. Zinc supplements and growth in developing countries ¹