Comments

1. The results indicate that the diet used in this study satisfied the protein needs of healthy, well nourished, pre-school children. In contrast with the most common dietary practices among people of low socio-economic conditions in developing countries, the preparation and administration of the foods differed in three important ways: (a) the increment in energy density achieved by the addition of oil and sugar to some foods; (b) the availability of staple foods in sufficient amounts to satisfy the children's appetites; and (c) the preparation and administration of the foods in the same way as for adults (e.g., feeding whole beans to small children instead of only a watery bean soup). These three conditions can be met by an increment in local food resources, including fats, combined with an educational programme for the use of such resources.
2. It is not necessary to use large amounts of proteins from animal sources or of commercially processed foods, as these tend to be more expensive and less regularly available and may imply greater changes in food preparation and eating habits than the mere addition of fats and carbohydrates to staples. The diet used in this investigation included small amounts of animal proteins on some days. It remains to be shown if these small additions of high-quality proteins are essential to other diets of exclusively vegetable origin. Our previous investigations with diets based on corn and black beans indicated that such additions or other protein fortification measures are not necessary, provided that the vegetable foods are eaten in such quantities and proportions that a good complementary vegetable protein mixture is absorbed by the children. In any event, the present study demonstrated that it is not necessary for such high-quality protein "complements" to be ingested with every meal or every day.
3. The use of mixed diets with varied menus and free choice of intake departed from the traditional patterns of metabolic studies. However, it resembled everyday life since children do not eat the same foods or the same amounts day after day, and all the nutrients in their diet are not always digested and absorbed simultaneously or in similar time sequences. This experimental design also accounted for some of the possible interactions of various nutrients and other food components, such as the effects that the presence of different amounts and types of dietary fibres may have on nitrogen and fat absorption on different days.
4. Complete collections of daily excrete might yield more accurate metabolic-balance results, but this is feasible only with older children and adults. The 4-day interval surrogate used in this study was adequate, as shown by the replication of the 28-day periods. The results did not vary unless the children became sick for several days with fever and marked anorexia; illness and mild symptoms did not produce any changes in metabolic balances.
5. Growth rates were also similar in the two consecutive periods. Therefore, a 28-day period might be adequate to assess the effects on body weight and metabolic balance of local diets eaten in the usual patterns if the children do not become ill. Otherwise, 28-day replications are necessary. The evaluation of other anthropometric changes, such as length, lean arm diameter, and subcutaneous skin-fold thickness, requires a longer period (42 to 56 days).
6. There is no satisfactory explanation for the apparent high nitrogen retentions. This is known to occur in children and adults with protein intakes well beyond their requirements. The important observation derived from this study was that nitrogen balance remained positive and sufficed to cover the estimated needs for growth and insensible nitrogen losses. This is supported by the stability of the CHI and the small, but consistent, increment in lean arm diameters.
7. The energy expenditure and balance data confirm our previous observations that mean net energy intakes (i.e., dietary intakes measured by bomb calorimetry and corrected for faecal losses) of 82 to 85 kcal/kg/day are adequate for active, well nourished children two to four years old. The energy not available from the protein absorbed, calculated from urinary nitrogen excretion, was about 1 kcal/kg/day. Consequently, the net energy intake is similar to that calculated from food composition and the Atwater energy factors for proteins, fats, and carbohydrates. Since the coefficient of variability of the mean net intake was 5 to 6 per cent, it seems safe to recommend energy intakes of 92 to 95 kcal/kg/day (mean + 12 per cent) for well nourished children of this age group.
8. The question remains open as to whether diets of the types used in this investigation will allow children with mild-to-moderate nutritional deficiencies to catch up. This is highly relevant, since such children make up the majority of the pre-school age population of low socio-economic groups in the developing world. The catch-up in length and high rates of weight gain observed in this investigation suggest that this might be possible. The answer can be explored through a similar study on apparently healthy children with small weight deficits (5 to 15 per cent below that expected for their heights) who do not have clinical signs of malnutrition except for their leanness. A longer period of time (e.g., 84 instead of 56 days) may be necessary to assess the rates of catch-up growth and to account for the impact of disease that might occur.