Contents - Previous - Next

This is the old United Nations University website. Visit the new site at http://unu.edu

A consideration of allowable fibre levels in weaning foods

G. Richard Jansen
Department of Food Science and Nutrition, Colorado State University, Fort Collins, Colorado, USA

INTRODUCTION

In recent years dietary fibre has become of considerable interest in human nutrition because of some demonstrated and some hypothesized beneficial attributes. However, these relate primarily to the nutrition of adults or, at least, school-age children. In contrast to the situation for adults, fibre in relation to infants and pre-school children raises important nutritional issues, primarily in relation to its potential disadvantages, rather than advantages.

Fibre is a particular concern in the low-cost extrusion cooking (LEC) programme at Colorado State University sponsored by USDA/AID. In this programme which has been described elsewhere (1), the emphasis is on the manufacturing of weaning foods in a developing country using low-cost extrusion technology and locally grown food crops. The lower-cost technology being used has some significant limitations in the efficiency of dehulling cereal grains such as corn or sorghum. A requirement to dehull corn to the same extent as is done in producing commercial degerminated corn grits would have at least four undesirable consequences for the LEC concept: (a) the capital costs of degerming equipment are much higher than a scourer-aspirator currently used, thus adding to the cost of an LEC plant; (b) related to the first point is the fact that the equipment available to make degerminated corn grits is of larger capacity and above the capacity of many LEC systems; (c) the technical expertise required to operate and maintain milling equipment of this nature is at a higher level than has been found to be available in typical LEC plants; (d) to mill corn to achieve the very low fibre levels present in degerminated corn results in the relatively low extraction rate of 65 - 75 per cent. This means that a large amount of grain intended for human food would need to be diverted to animal feed. Not only would this be counterproductive to the nutritional objectives of LEC projects, it would raise the cost of the product and it is not clear what market exists to absorb the by-product of the milling operation.

The above are reasons why specifications for allowable fibre levels in locally manufactured extruded food products intended for use in supplementary feeding programmes should be based on the best data and most current thought about nutrition available. It is interesting that, in spite of the importance of supplementary weaning foods to the weaning process, few guidelines or hard data are available on allowable fibre levels. This review will include a definition of dietary fibre, methods of fibre analysis, and analytical data on corn/soy blends made using a LEC system. After outlining potentially beneficial and deleterious attributes of fibre in human nutrition, published studies carried out in infants and young children will be reviewed and discussed. Finally, recommendations will be made relative to dehulling requirements and allowable fibre levels in LEC-blended foods.

II. DEFINITION OF FIBRE

Fibre is not a single chemical entity; it is an exceedingly complex mixture of poorly characterized constituents. Although our current understanding about fibre is in a state of development, the most useful concept at present is that of "dietary fibre" (2). Dietary fibre is considered to be plant-cell skeletal remains that are resistant to digestion. Although primarily plant cell walls consisting of cellulose, ligin, and the hemicelluloses, dietary fibre also includes soluble polysaccharides such as pectin, plant gums, and mucilages. What has been referred to by some as "unavailable carbohydrate" is closely similar but not identical to the broader concept of dietary fibre. Fibre is not totally carbohydrate because it includes ligin, which is a non-carbohydrate polymer. In addition, fibre is not totally unavailable, because it is now known that a portion of dietary fibre is metabolized and converted to volatile fatty acids in the gastrointestinal tract of man as well as of animals.

In contrast to the current concept of dietary fibre, the older concept of "crude fibre" has serious shortcomings. The method for determining crude fibre consists basically of measuring the residue remaining after acid and basic hydrolysis. It is known that the crude fibre method covers 50 80 per cent of the cellulose, 10 - 50 per cent of the legnin, and 20 per cent of the hemicelluloses (3). It is clear that there is no fixed relationship between crude fibre and dietary fibre because plant cell walls vary in the proportions of their basic constituents. Unfortunately, the data base on the dietary fibre content of foods and food ingredients is very sketchy at present. Also, the methodology for analyzing dietary fibre is still being developed, and no consensus on a single method for use with human foods has yet emerged. Therefore, it is necessary to continue to use crude fibre values when discussing fibre in human nutrition. However, the limitations of the method should be held in mind. In the next section, two major alternative methods for dietary fibre analysis will be outlined and contrasted with the crude fibre methodology.

FIG. 1. Crude Fibre Method (4)

III. METHODS OF DETERMINATION OF FIBRE

Three methods of determination of fibre content in foods are currently commonly used. These are: (a) the crude fibre method (b) the "unavailable carbohydrate" method of Southgate, and (c) the neutral detergent fibre method originally developed by Van Soest.

Crude Fibre Method

Most published values for the fibre content of foodstuffs were obtained using the so-called crude fibre method (4). The procedure as it is currently used is outlined in figure 1. In this method, crude fibre is the residue left after sequential hot digestion by 1.25 per cent sulfuric acid and 1.25 per cent sodium hydroxide. It has already been pointed out that crude fibre values underestimate the amount of plant materials indigestible by the enzymes of the human digestive tract. In addition, the degree of indigestibility varies considerably, depending on the fibre source.

Southgate Method

In 1969 Southgate (5) described a fractionation procedure for isolation of the main components of what he called "unavailable carbohydrates." A current schematic outline of the procedure is given in figure 2. Southgate recommended the use of amyloglucosidase at 37° C for 18 hours to remove starches from food samples. In this procedure, using hydrolytic techniques unavailable carbohydrates are measured by the liberation of sugar components. The method gives values for two types of non-cellulosic polysaccharides: those that are water-soluble and hydrolysable with dilute acid, and the cellulose and other associated non-cellulose pentosans and lignin.

Neutral Detergent Fibre (NDF) Method

The neutral detergent fibre (NDF) method was developed by Van Soest and Wine (6) for the purpose of determining the indigestible component of forages. The method measures, in essence, total cell wall material. In some cases it may underestimate total dietary fibre because of a loss of water-soluble polysaccharides (3). The procedure is outlined in figure 3.

Van Soest (7) reported that hemicellulose, cellulose, and lignin in forages could be recovered using a solution of sodium lauryl sulfate buffered at pH 7 with EDTA-Borate. Over 75 per cent of the nitrogen present in the sample of forage is solubilized using this method, and the NDF recovered represents essentially the total cell wall material (3).

Because of the high starch content of human foods in comparison to forages on which the method was tested, an overestimation of cell wall content occurs with the NDF method due to starch remaining in the residue (3). A modified procedure was developed by Robertson and Van Soest (8) in which a detergent stable bacterial alphaamylase is used to hydrolyse the starches.

FIG. 2. Southgate's Fractionation Method (4)

FIG. 3. Modified Neutral Detergent Method (4, 8)

The Southgate procedure was originally developed to measure unavailable carbohydrates, while the Van Soest procedure was initially developed for forage analysis. With the modification of the NDF method for human food analysis by the addition of an amylase, both give estimates of dietary fibre that are superior to the old crude fibre method. However, crude fibre values as well as dietary fibre values need to be considered when evaluating the fibre levels of weaning foods. This is especially true because current specifications for fibre in weaning foods are in terms of crude fibre values. Also, the values obtained in crude fibre or dietary fibre analysis can vary with particle size and composition differences in a food product independent of the fibre itself, so fibre values do not have the same degree of precision as do some other measures such as Kjeidahl nitrogen values.

IV. EFFECTS OF FIBRE IN HUMAN NUTRITION

During the last ten years, the effects and importance of dietary fibre in human nutrition have attracted considerable attention on the part of nutritionists and clinicians. Essentially all this activity has related to demonstrated or alleged benefits of high fibre intakes in adults (2, 9). A brief summary of the current state of knowledge is as follows. High fibre intakes are beneficial in preventing constipation and diverticulosis (10). Evidence is accumulating that dietary fibre improves glucose tolerance and is beneficial in treating maturity onset diabetics (1, 11).

Less well established are the reputed benefits of high fibre intakes in lowering plasma cholesterol and in decreasing the incidence of colon cancer. In the case of cholesterol, plasma-lowering depends on the source of the dietary fibre (13). For example, results with wheat bran, a popular source of dietary fibre, have been consistently but not uniformly negative. In the case of cancer, some epidemiological and experimental data suggest that diets high in animal foods, high in fat, and low in dietary fibre lead to a higher incidence of colon and breast cancer (14, 15). It is thought that if this relationship holds, the role of diet is as a promotor of cancer rather than as a cause. It is not completely clear what the role of fibre is in this complex situation.

In contrast to the situation for adults, in the nutrition of infants and pre-school children, emphasis is placed on the importance of keeping fibre intakes low. However, few hard data are available. The maximum allowable level for crude fibre in PL 480 Title II blended foods ranges from 2.0 to 2.5 per cent (16, 17). The Protein Advisory Group (PAG) of the UN suggests an upper limit of 5.0 per cent crude fibre in supplementary foods (18). The Codex Alimentarius standard for infant cereals does not establish an upper limit for fibre (19).

Possible undesirable aspects of high fibre levels in weaning include, a. increased bulk and lower calorie density, b. irritation of the gut mucosa, c. reduced digestibility, and d. reduced vitamin and mineral availability. The calorie density of a cereal gruel will be lowered by fibre, but this effect is relatively small compared to other factors such as the amounts of fully gelatinized starch, sugar, milk solids, and fat in a cereal/legume blend. Knowing the nature of dietary fibre, it is reasonable to be concerned about possible undesirable local effects of fibre on the intestinal mucosa in infants, especially those recently recovered from malnutrition. Unfortunately, few data based on untoward local effects of fibre or other hull constituents in the intestine are available to establish limits on either the types or amounts of dietary fibre in weaning foods.

It is well known that digestibility of cereal and legume blends is less than that of fibre-free animal foods. Since dietary fibre is to some extent, but not totally, unavailable, it is clear that higher fibre levels will reduce the digestible energy of a food by at least the amount of fibre that is not digested. The digestibility of NDF in man ranges from 10 to 70 per cent (3). What is not clear is whether or not, and to what extent, dietary fibre interferes with the availability of protein or total food energy from other dietary constituents. (The available data on protein digestibility from selected studies in infants and young children will be considered in section VI.)

Several recent studies in swine are of interest with regard to the effects of dietary fibre from these sources on digestibility in young, growing non-ruminant animals. Kornegay (20) studied the feeding value and digestibility of soybean hulls in swine. Corn-dehulled soy diets containing 2 - 24 per cent soybean hulls were fed to 240 pigs weighing 16 to 30 kg each. The soybean hulls contained 47.0 per cent crude fibre and 67.2 per cent dietary fibre. Average daily weight gain was maximal on diets containing 6 per cent hulls. Feed per gain was constant for animals on diets with up to 6 per cent hulls but increased with 12 to 24 per cent hulls in the diet. Based on digestibility studies carried out in older swine, 50 per cent of the dietary fibre was digestible. It is interesting that average daily gain in these young growing pigs was depressed only slightly, if at all, by up to 24 per cent hulls in the diet. Knabe et al. (21), in studies with sorghum-cottonseed meal diets, also found the crude fibre had minimal effects on feed intake and weight gain, although the feed-gain ratio was increased at the higher fibre levels.

Although there is still considerable debate concerning the relative importance of fibre versus physic acid in depressing mineral availability, there is considerable evidence that fibre can bind cations such as zinc, copper, iron, calcium, and magnesium and reduce their availability (22 - 26). Although few studies have been done on the effect of fibre levels on mineral availability in young children, zinc deficiency resulting in poor growth and hypogonadism in young boys in the Middle East has been attributed to fibre and physic acid in whole meal unleavened breads (27). The effect of fibre on vitamin availability has apparently not been widely investigated (28).

V. ANALYSIS OF EXTRUDED CORN-SOY BLENDS (LEC-CSB)

Three samples of LEC-CSB (70 per cent corn, 30 per cent soy) were made at Colorado State University for metabolic evaluation in young children. All were extruded on a Brady extruder at 160° C. These samples were made from (a) commercially degerminated corn/dehulled soy, (b) whole corn/dehulled soy, and (c) whole corn/whole soy. The soybeans were dehulled on a Sturtivant scourer-aspirator. (The metabolic results will be discussed in section VI.) Crude fibre levels for these three products are shown in table 1. Neutral detergent fibre values were not measured. The crude fibre levels ranged from 0.8 per cent for the degerminated corn/dehulled soy product to 2.3 per cent for the whole corn/whole soy blend. Only the latter product exceeded the PL 480 "Food for Peace" Title II specifications of 2.0 per cent for CSB, CSM, or ICSM. Subsequently, three samples of LEC-CSB varying in fibre content were again made at CSU. These samples were analysed before and after extrusion for both crude fibre and neutral detergent (dietary) fibre. These samples were made from 70 per cent corn and 30 per cent soy as follows: (a) dehulled corn/dehulled soy, (b) whole corn/dehulled soy, and (c) whole corn/whole soy. In this experiment, both the corn and soy were dehulled using a Sturtivant scourer/aspirator. This is the type of dehulling equipment currently recommended for use in LEC plants.

TABLE 1. Fibre Content of Corn-Soy Blends (%)

Note: Corn-soy blends made from 70% corn, 30% soy (W:W).

a. These samples sent for metabolic evaluation in young children (35). The degerminated corn was obtained from commercial sources and the soy was dehulled at CSU using a Sturtivant scourer/aspirator.
b. Both the corn and soy samples were dehulled at CSU using a Sturtivant scourer/aspirator.
c. Data of Al-Hooti (29). d. Standard errors based on eight separate samples including both raw and extruded samples. Extrusion did not affect the recovery of CF or NDF.

The fibre levels for these samples are also presented in table 1 (29). Extrusion did not significantly affect fibre levels when measured by either method, so the data for both raw and extruded samples of the same blend have been averaged together. In this case, the crude fibre level of the product made from dehulled corn/dehulled soy (1.37 -+ 0.13) was not significantly lower than that for the product made from whole corn/dehulled soy (1.46 + 0.05). This indicates that very little dehulling of the corn took place with this equipment. The sample made from whole corn/dehulled soy was 0.56 per cent lower in fibre than in the product made from whole corn/whole soy, indicating that the fibre level of whole soy was reduced by about 1.9 per cent or about 40 per cent of the level originally present. The crude fibre level of the completely whole grain product in this experiment (2.0 per cent) was lower than that observed in the first experiment (2.3 per cent). Whether this was caused by a difference in the fibre content of the ingredients or methodological problems in crude fibre analysis is not known at present. It is interesting to note that neutral detergent fibre (NDF) levels in these samples were 4 to 5 times higher than the crude fibre (CF) levels. This difference is consistent with other reported values for CF and NDF (10). As discussed in section 11, the NDF values are much better indicators of the amount of dietary fibre than crude fibre values.

VI. STUDIES IN YOUNG CHILDREN

The effect of fibre levels on protein and energy digestibility in young children does not appear to have been investigated systematically. However, an extensive series of evaluations of the protein quality of a wide variety of weaning foods in a single metabolic unit offers an excellent opportunity to examine the effect of fibre on nitrogen digestibility and retention in young children. Drs. Graham and Maclean and their associates have reported on protein quality studies in infants on wheat-soy blends and oat-soy blend (30), corn soy milk (31), sweetened instant corn-soy milk (32), whole wheat and white flour (33), whole corn and corn endosperm (34), and three corn-soy blends varying in fibre content made at Colorado State University using a Brady extruder (35) as described in section VI.

Pertinent data from these various experiments are summarized in table 2. In interpreting these data, several points should be kept in mind. In comparing different blends, the strongest inferences can be made within a single reference. For convenience, the references for the various blends are indicated in table 2. In making comparisons between blends reported in separate publications, apparent absorption and retention values expressed as percentages of the casein control values are likely to have more validity than absolute values. Since nitrogen absorption values are reported as "apparent" absorptions uncorrected for endogenous faecal nitrogen, it is difficult to compare samples fed at different nitrogen intakes.

The data reported in table 2 for the two wheat-soy blends (WSB), oat-soy blend (OSB), corn-soy milk (CSM), and instant corn-soy milk (ICSM) do not allow any strong inferences to be drawn on the effects of fibre on protein digestibility or nitrogen retention because of a limited number of subjects and considerable variability. It is interesting to note, however, that WSB 2 with a crude fibre level of 2.7 per cent made from 63 per cent whole wheat and 35 per cent defatted soy flour, resulted in apparent absorption as good as that with CSM in which the crude fibre level was 0.8 per cent. Also, WSB-2 showed essentially the same digestibility and considerably higher retention than did WSB-1 (84.2 vs. 68.7 per cent of casein) in spite of a higher crude fibre level (2.7 vs. 1.9 per cent).

In other experiments whole wheat was compared with whole flour and whole corn was compared with corn endosperm. In the case of wheat, several varieties of wheat differing in protein content were fed as both whole wheat and white flour. Since the results did not vary significantly as a function of variety, the data are presented in table 2 for all varieties together. In the case of corn, both normal and opaque 2 corn were fed as whole grain and as endosperm. Because these data did vary as a function of variety, they are presented in table 2 for normal and opaque-2 corn separately.

As shown in table 2, apparent absorption and retention of nitrogen were slightly, but not significantly, less with whole wheat than was the case with white flour. Considering that the whole wheat contained four times more crude fibre than did white flour, the slight difference in protein digestibility is noteworthy. Nitrogen absorption from whole wheat was 89.6 per cent of that from casein.

The comparison of whole corn with corn endosperm is even more interesting. Apparent nitrogen absorption from whole corn was at least as good as absorption from corn endosperm, for both normal and opaque-2 corn. Nitrogen retention from whole corn was better than nitrogen retention from corn endosperm for both normal and opaque-2 corn.

The experiments carried out with corn-soy blends (CSBs) made on the Brady extruder were designed to give information on the protein quality and digestibility of these blends as a function of fibre level. In the first experiment CSB was made from commercially-degerminated corn and soy dehulled using a Sturtivant scourer/aspirator. This product had about the fibre level of ICSM, and apparent nitrogen absorption was found to be similar to that reported previously for ICSM. Nitrogen retention on this CSB sample tended to be slightly lower than for casein, although not significantly so.

TABLE 2. Summary of Nitrogen Balance Studies in Infants

a.40.7% winter wheat flour, 40.5% wheat protein concentrate, 16.4% toasted defatted soy flour (30).
b.62.9% whole wheat flour, 35% defatted soy flour (30).
c.62.9% whole groat oat flour, 35% defatted soy flour (30).
d.68% partially gelatinized cornmeal, 25% toasted defatted soy flour, 5% nonfat dry milk (31).
e.40% completely gelatinized cornmeal, 38% full-fat soy flour, 5% nonfat dry milk,15% sugar (32).
f. Ref.33.
g. Ref. 34.
h. 70% degerminated corn grits, 30% dehulled soy—extruded at 163° C (35).
i. 70% whole core, 30% dehulled soy—extruded at 163°C (35).
j. 70% whole core, 30% whole soy—extruded at 163°C (35).
k. Mean ± standard error.

In the second experiment apparent nitrogen absorption and retention were determined for three samples differing in crude fibre. Sample 1 was made from degerminated corn and dehulled soy, sample 2 from whole corn and dehulled soy, and sample 3 was made from whole corn and whole soy. In this experiment protein was fed at 8.5 per cent of calories. As shown in table 2, apparent nitrogen absorption and retention were significantly lower in both higher fibre products than was the case for the lowest fibre CSB made from degerminated corn and dehulled soy. When whole corn was used, apparent nitrogen absorption and retention were not significantly different whether the blend was made with dehulled or whole soy. It should be noted that in these experiements with extruded CSB, the degerminated corn grits and the whole corn were from different sources, thus the differences in nitrogen utilization noted cannot be totally ascribed to the degermination process.

VII. DISCUSSION

In developing countries, the weaning child needs to make the transition from breast milk, which is high in energy, highly digestible, fibre-free, and essentially sterile, to the "family pot" that consists for the most part of cereals with some legumes, root crops, fruits, and vegetables and that is low in energy, relatively poorly digestible, and usually high in fibre. The water supply is often contaminated. Because of the relatively poor nutritional status of many lactating women, it is generally felt that infants in developing countries should be given supplemental foods by the age of three months (36). The weaning period is the time when infants are most likely to suffer from diarrhoea and die. In the following discussion, it is assumed that fibre in a supplemental food becomes a more significant nutritional issue the younger the child is. Specifically, infants younger than one year represent the population group for whom it is most important to establish an upper limit for fibre. The blended foods being made as part of the low-cost extruder (LEC) programme may be used as weaning foods by infants as young as three months. They can also be used by older children and adults, especially pregnant or lactating women. The specification for fibre in LEC-blended foods will be discussed primarily in terms of the use of these blended foods as supplemental foods for infants less than one year of age, since it is in this age group where the specification needs to be set at the lowest level.

The data reviewed in section VI are helpful but not yet definitive in establishing the fibre specification for LEC-blended foods. The data on the three CSBs varying in fibre level demonstrate an advantage in using degerminated corn and dehulled soy as ingredients in terms of absorption and retention of nitrogen. However, this advantage needs to be balanced against milling losses and higher processing costs. Also, absorption and retention of nitrogen were comparable in these higher fibre blends to what had previously had been observed for Title 11 CSM. It seems somewhat surprising that the whole-soy blend was digested as well as the blend made from dehulled soy, but this lack of effect of a low level of soybean hulls on digestibility is consistent with the previously cited study in young growing swine (20).

It is of particular importance to evaluate the importance of dehulling corn on digestibility in connection with the local manufacture of weaning foods in developing countries. In making CSM or ICSM in the United States for the Title 11 programme, commercial degerminated corn is used. This is made by a degerming process that is of high capacity and results in an extraction rate of 65 - 75 per cent. The milling fractions removed are diverted primarily to animal feeds. In developing countries the cost of such milling equipment would increase processing costs and could negate some of the economic advantages of the LEC concept. In addition, it is not clear whether it is desirable to: a. divert such a large amount of corn away from human consumption and b. whether the market for animal feeds is sufficiently large to absorb the mill fractions if the corn were degerminated. The dehulling equipment currently being used in LEC plants is capable of dehulling soy reasonably well, but it is much less effective in dehulling corn.

The CSB made with whole corn was clearly less digestible and resulted in lower nitrogen retention than observed with the blend made from degerminated corn. These results need to be considered, however, in the context of the studies in which whole wheat was compared to white flour and whole corn was compared to corn endosperm. In the case of wheat, the whole-wheat protein was digested as well as protein from white flour in spite of a much higher crude fibre level. More relevant to the question of the use of whole versus degerminated corn are the data reported for whole corn versus corn endosperm. For both normal and opaque-2 corn, the protein of whole grain was digested as well as the protein from the endosperm. Whole-corn protein was retained significantly better than endosperm protein was, most likely because of the presence of the higher quality protein from the germ. These studies with whole wheat and whole corn are especially interesting since feeding these lower protein cereals at 6.4 - 6.5 per cent of calorie intake would of necessity result in higher average fibre intakes than was the case with CSB, even when the latter was fed at 8.5 per cent of calories.

Considering all the data available, there does not appear to be any reason to require a lower maximum crude fibre level for weaning foods than the 2.0 per cent currently specified for Title 11 CSB. This fibre level can be achieved in CSB made from 70 per cent whole corn and 30 per cent dehulled soy. Sufficient data are not yet available to specify the allowable fibre levels for weaning foods in terms of neutral detergent fibre (NDF). It would appear that an NDF level of approximately 8 per cent will prove to be appropriate, but more data are required to establish this limit. In view of the difficulty of dehulling corn in an LEC factory and the considerable cost savings that will result from using whole corn, it is recommended that whole corn and dehulled soy be used as ingredients in making extruded corn-soy blends in an LEC plant. Use of whole corn has the additional advantage of retaining germ protein and oil in the final product. It should be noted that whole corn has been used in supplemental foods fed to pre-school children with apparent success for many years under the generic name Incaparina (37).

The above recommendation on dehulling is made only for corn. Since soy is relatively easy to dehull, whole soybeans have not traditionally been fed to infants, and, as in dehulling soy one is not lowering protein quality or reducing the oil content, it is recommended that soybeans continue to be dehulled. The data reported by Graham and MacLean (35) for LEC-CSBs suggest that soybeans are adequately dehulled using the low-cost Sturtivant scourer/ aspirator. Minerals can be added to these blended foods in ample amounts after extrusion, so use of whole corn should not pose a particular problem for mineral availability. However, it would be desirable to carry out experiments on experimental animals and infants on this point. It is possible that somewhat more iron and zinc in the fortification mixture may be desirable if whole corn is used. Additional experiments on the effect of extrusion and particle size on digestibility of the whole cereals that are used, or that potentially will be used in LEC-blended foods are warranted. These studies should include both corn and sorghum as components of blends with dehulled soy.

In this discussion, dehulling requirements have been considered primarily in terms of the resulting fibre levels. However, it is recognized that in dehulling other important substances, such as physic acid and tannins, as well as surface contaminants, are removed. This consideration reinforces the desirability of additional studies on mineral availability if whole corn is used, and also the desirability of continuing to dehull the soybeans until definitive data are available that indicate that this is not necessary. Use of whole corn will result in an added requirement that the grain-cleaning operations be carried out carefully, meeting all specifications that would be established.

Because sorghum and millet contain variable but higher levels of tannins that reduce digestibility, it is clear that considerably more work is needed before the use of whole sorghum or millet can be recommended in making extruded weaning foods for use in developing countries. The same is true for the use of whole soy, as mentioned above, and whole legumes such as dry beans.

The above comments are made in the specific context of an LEC operation making weaning foods to be fed to infants less than one year of age. Nothing discussed here is meant to imply that foods intended for older children and adults, or multi-mixes made at the home or village level, should not contain more fibre than specified for these LEC blends. The numerical value specified as the value limit for fibre is probably the secondary consideration, the primary one being the dehulling requirements for the particular ingredients and process used. It is clear that considerably more work is needed before a full understanding of allowable fibre levels and dehulling requirements for the Icoal manufacture of weaning foods is achieved.

REFERENCES

1. G.R. Jansen and J.M. Harper, "Applications of Low-Cost Extrusion Cooking to Weaning Foods in Feeding Programs," FAO Food and Nutrition Quarter/y, 1980 (in press.)

2. G.A. Spiller and R.J. Amen, eds. Fiber in Human Nutrition (Plenum Press, New York, 1976).

3. J.B. Robertsons "The Detergent System of Fiber Analysis," in G.A. Spiller, ea., Topics in Dietary Fiber (PIenum Press, New York, 1980). p. 1.

4. D.A. Southgate, "The Analysis of Dietary Fiber," in G.A. Spiller and R.J. Amen, eds., Fiber in Human Nutrition (Plenum Press, New York, 1976) p. 73.

5. D.A. Southgate, "Determination of Carbohydrates in Foods. II. Unavailable Carbohydrates," J. Sci. Fd. Agric., 20: 331 11969).

6. P.J. Van Soest and R.H. Wine, "Use of Detergent in the Analysis of Fibrous Feed. IV. Determination of Plant Cell-Wall Constituents," J. Assoc. Off. Agr. Chem., 50: 50(1967).

7. P.J. Van Soest, "Use of Detergents in the Analysis of Fibrous Feeds. I. Preparation of Fiber Residues of Low Nitrogen Content," J. Assoc. Off. Agr. Chem., 46: 825 (1963).

8. J.B. Robertson and P.J. Van Soest, "Dietary Estimation in Concentrate Feed Stuffs," Paper no. 636, presented at 69th meeting of the American Society of Animal Science, University of Wisconsin, Madison, 23 - 27 July 1977.

9. Fifth Annual Marabou Symposium, "Food and Fiber," Nutrition Reviews, 35 (3): 1 - 72 (1977).

10. "Dietary Fiber: A Scientific Status Summary by the Institute of Food Technologists Expert Panel on Food Safety and Nutrition," Food Technology, vol. 33, Jan. 1979.

11. J.W. Anderson and K. Ward, "High-Carbohydrate, High-Fiber Diets for Insulin-Treated Men with Diabetes Mellitus," Am. J. C/in. Nutr., 32:2312 11979).

12. H. Trowel I, " Diabetes MelIitus and Dietary Fiber of Starchy Foods," Am. J. Clin. Nutr., 31: 552 11978).

13. D. Kritchevsky, "Fiber, Lipids, and Atheroscierosis," Am. J. C/in. Nutr., 31: 565 (1978).

14. C.T.L. Huang, G.S. Gopalakrishna, and B.L. Nichols, "Fiber, Intestinal Sterols, and Colon Cancer," Am. J. C/in. Nutr., 31: 516 (1978).

15. A.R.P. Walker, "The Relationship between Bowel Cancer and Fiber Content in the Diet," Am. U Clin. Nutr., 31: 5248 11978).

16. USDA-ASCS, "Purchase of Corn/Soy Blend for Use in Export Programs," (Announcement CSB-2, Shawnee Mission, Kan. USA, 25 July 1975).

17. USDA-ASCS, "Purchase of Wheat-Soy Blend for Use in Export Programs," (Announcement WSB-5, Shawnee Mission, Kan. USA, 25 July 1975).

18. "Guideline on Protein-Rich Mixtures for Use as Supplemental Foods," The PAG Compendium vol. E (Worldmark Press Ltd., John Wiley and Sons, New York, 1975), p. 63.

19. "Recommended International Standards for Foods for Infants and Children, Codex Alimentarius Commission CAC/RS 72/74 1976 (Joint FAD/WHO Food Standards Program, FAO, Rome, 1976).

20. E.T. Kornegay, "Feeding Value and Digestibility of Soybean Hulk for Swine," 1 Animal Science, 47: 1272 (1978).

21. D.A. Knabe, T.D. Tanksley, Jr. and J.H. Hesby, "Effect of Lysine, Crude Fiber, and Free Gossypol in Cottonseed Meal on the Performance of Growing Pigs," J. Animal Science, 49: 134 (1979).

22. J.G. Reinhold, "High Phytate Content of Rural Iranian Bread: A Possible Cause of Human Zinc Deficiency," Am. J. C/in. Nutr., 24:1204 (1971).

23. J.G. Reinhold, B. Faradji, P. Abadi, and F. Ismail-Beigi, "Decreased Absorption of Calcium, Magnesium, Zinc, and Phosphorus by Humans due to Increased Fiber and Phosphorus Consumption as Wheat Bread," J. Nutr., 106: 493 (1976).

24. J.G. Reinhold, F. Ismail-Beigi, and B. Faradji, "Fibre vs. Phytate as Determinant on Availability of Calcium, Zinc, and Iron of Breadstuffs," Nutr. F/epts. Int, 12: 75 (1975).

25. J.G. Reinhold, A. Nasar, and H. Hedayatti, "Effects of Purified

Phytate and Phytate-Rich Bread upon Metabolism of Zinc, Calcium, Phosphorus, and Nitrogen in Man," Lancet, 1 283 (1973).

26. J.G. Reinhold, A. Parsa, N. Karimian, J.W. Hammick, and F. Ismail-Beigi, "Availability of Zinc in Leavened and Unleavened Wholemeal Wheaten Breads as Measured by Solubility and Uptake by Rate Intestine in Vitro," J. Nutr., 104: 976 (1974),

27. H.H. Sandstead, A.S. Prasad, A.R. Schulert, Z. Farid, A. Miale, S. Bassilly, and W.J. Darby, "Human Zinc Deficiency, Endocrine Manifestations, and Response to Treatment," Am. J. C/in. Nutr., 20: 422, 1967.

28. J.H. Cummings, "Nutritional Implications of Dietary Fiber," Am. J. C/in. Nutr., 31: 521 (1978).

29. S. Al-Hooti, "Effect of Extrusion on Digestibility and Mineral Availability in Cereal Blends," thesis, (Colorado State University, Fort Collins, Colo. USA, 1979).

30. G.G. Graham, J.M. Baertl, R.P. Placko, and A. Cordano, "Dietary Protein Quality in Infants and Children. VIII. Wheat-or Oat-Soy Mixtures," Am. J. Clin. Nutr., 25: 875 (1972).

31. G.G. Graham, E. Morales, G Acevedo, R.P. Placko, and A. Cardano, "Dietary Protein Quality in Infants and Children. IV. A Corn-Soy-Milk Blend," Am. l Clin. Nutr., 24: 416 (1971).

32. G.G. Graham, J.M. Baerti, R.P. Placko, and E. Morales, "Dietary Protein Quality in Infants and Children. IX. Instant Sweetened CornSoy-Milk Blend," Am. J. Clin. Nutr., 26: 491 11973).

33. W.C. MacLean Jr., G. Lopez de Romana, and G.G. Graham, "Protein Quality of High Protein Wheats in Infants and Children," J. Nutr., 106: 362 (1976).

34. G.G. Graham, D.V. Slover, G. Lopez de Romana, E. Morales, and W.C. MacLean Jr., "Nutritional Value of Normal Opaque-2 and Sugary-2-Opaque-2 Maize Hybrids for Infants and Children. l. Digestibility and Utilization," J. Nutr., 110: 1061 (1980).

35. G.G. Graham and W.C. MacLean Jr., "Digestibility and Utilization of Extrusion Cooked Corn-Soy Blends," report submitted to the Office of Nutrition, Agency for International Development, Washington, D.C., 29 January 1979.

36. J.C. Waterlow and A.M. Thomson, "Observations on the Adequacy of Breast Feeding," Lancet, 4 Aug. 1979, p. 238.

37. N.S. Scrimshaw, M. Behar, D. Wilson, V. Viteri, G. Arroyane, and R. Bressani, "All-Vegetable Protein Mixtures for Human Feeding. V. Clinical Trials with INCAP Mixtures 8 and 9, and with Corn and Beans," Am. J. Clin. Nutr., 9: 196 (1961).

Contents - Previous - Next