Contents - Previous - Next

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

The acceptance and intake of bulk-reduced weaning foods: The Luganga village study

Alexander C. Mosha and Ulf Svanberg


Child mortality rates in Tanzania were reported to be 137 per thousand live births in the latest population census data, compiled in 1987. Analysis of community survey data from the last ten years reveals that protein-energy malnutrition (PEM) and diseases are the immediate causes of this high mortality rate [1]. According to Waterlow's classification [2], the level of severe PEM was 4%-6%, and moderate PEM was 40% 60% in pre-school children, with the severe form reaching a peak between 13 and 29 months of age. Most children between the ages of 6 and 30 months also experienced recurrent illnesses. Infectious diseases, even those of mild character, have been shown to be a common cause of anorexia and to cause a significant reduction in food intake, weight loss, and impaired physical growth [3].

It is obvious that the problem starts when supplementary feeding is introduced at about 4-6 months of age and continues until complete weaning, which in most areas normally occurs between 18 and 30 months of age. Usually the preferred weaning food is prepared as a thin liquid (uji) using the main staple of the area. This most often means a gruel made of maize, sorghum, or millet flour, but in some areas mashed potatoes, bananas, or cassava are also used. Depending on the season and the area, a variety of beans, peas, and vegetables are added. The energy contribution from fat is low.

The liquid gruel usually contains around 5% dry matter, which results in an energy density of 0.2 kcal per gram of prepared gruel. For children above one year old, mothers may prepare thicker gruels. The upper limit for dry matter, however, is normally 20%, because, beyond this concentration, the gruel becomes difficult to stir. This provides an energy density of about 0.7-0.8 kcal per gram.

Evaluation of food-intake studies in young children [4-6] indicates that an energy density of the weaning diet of about 1.0 kcal per gram is needed in order to satisfy energy requirements [7]. Human milk has an energy density of about 0.7 kcal per gram [8], and it is therefore reasonable to assume that the energy density of a weaning diet should at least exceed 0.7 kcal per gram and preferably should be about 1.0 kcal per gram.

The liquid gruel will obviously be too voluminous for a young child; four to five litres would be required to meet the child's energy needs. The thicker gruel has an acceptable energy density, but its consistency makes it less easy to consume, especially when the child is sick and prefers a liquid gruel [9].

This high-volume/viscosity character of a diet is usually referred to as "dietary bulk" and can be a major constraint in providing children with enough food in areas where starchy staples are the main foods [10; 11]. The concept of dietary bulk as a limiting factor for energy and nutrient intake in pre-school children has been conceived, reviewed, and developed in a series of papers [12-15].

In attempts to find solutions to this problem, the use of the traditional food-preparation procedure of germination, or malting, has been shown to reduce the dietary bulk of cereal-based weaning foods considerably [14]. Flour of germinated sorghum - "power flour" - can also be used as an additive to already-prepared thick porridges of ungerminated flour [16].

The mechanism behind this effect is that amylolytic enzymes (amylases) developed and activated in the germination process rapidly break down the starch in the cereal, which is the main constituent of the gel structure in the porridge, reducing its water-holding capacity. As a result, the water trapped in the gel structure is released, producing a more liquid gruel.

The Luganga village study had two objectives: first, to measure how well the power-flour method for the preparation of bulk-reduced weaning foods was adopted by mothers and, second, to measure actual intakes of weaning foods with different dietary-bulk properties in pre-school children.

The study


Luganga village is located in the Iringa Region of Tanzania, near the Little Ruaha River, which is the local source of water. The population at the time of the study was about 1,300, in 250 families. Social services included a dispensary with a mother/child health (MCH) unit, a primary school, and a village shop.

The major crops grown were maize. millet, sorghum, and rice under irrigation near the river. Legumes (cowpeas and beans) and oil seeds (groundnuts and sunflower) were also grown to some extent, and seasonal vegetables and fruits were available. Most food produced was consumed in the household.


Forty pre-school children were selected at random from different households to represent all the village administrative cells. Eight of these forty were excluded from the final analysis because of frequent absence or migration from the village. The age range of the children was from 5 to 65 months. Weight-for-age data from MCH records at the Luganga dispensary indicated that about 68% of the children were classified as underweight. No one participating in the study was classified as severely malnourished (below 60% of the weight standard). The final test group consisted of 14 girls and 18 boys.


By mutual agreement between the villagers and the research team, it was decided to use the most readily available food items - namely maize grain or flour and ground-nuts, which were available in most households during the study or were provided from the village shop.

Three different diets were prepared using the same mixture of maize and ground-nuts (95:5). The first diet was a thin gruel containing 5% solid matter boiled in water for about 20 minutes, which was the common type of weaning food fed to infants. The second diet was a thick porridge with 20% solid matter and a viscosity of about 50,000 cP. The third diet was prepared like the second one, but, after the porridge was cooled to about 40 °C, germinated sorghum flour (power flour) was added (1% of the entire mixture) and mixed thoroughly until the porridge thinned down to a semi-liquid consistency with a viscosity about 3,000 cP. Viscosity measurements were performed for each diet in six households, using a Brookfield viscometer with a No. 6 spindle and a spindle speed of 20 rpm.

The preparation of the germinated sorghum or millet flour was similar to the well known traditional household technology for preparing malted flour, which is widely used locally for making beer. The seeds were soaked overnight and allowed to germinate for two to three days in the dark, covered by wet cotton cloths. After drying in the sun for two days, the seeds, with their vegetative portion, were ground into a flour with a pestle and mortar.


Adoption of the power-flour method

The technique of using the power flour to reduce the consistency of a thick porridge was demonstrated to the mothers at the dispensary on six occasions over a period of two months. On each occasion different ingredients were tried and tested by the mothers, and finally the maize/ground-nut mixture was the one selected. During the demonstrations, information was given on the importance of preparing nutritionally adequate weaning foods to be served at least three or four times a day. The children were also fed as a group with a weaning food treated with power flour at each meeting. This type of limited promotion programme was felt to be essential if the technique was to be extended on a national level without requiring too many promotional resources.

The mothers then recorded their acceptance and use of power flour in the home for a period of three months. They were asked to record every time they used power flour in making porridge for their children. For those who did not keep records, the research team used the recall method monthly. Most mothers made their own germinated flour or borrowed it from their neighbours.

Weaning-food intake

Initially group feeding was carried out at the village dispensary. Each child was served a portion of the maize/ground-nut porridge prepared as previously described, weighed with a Salter top-balance scale. The very young children who were unable to feed themselves were fed by their mothers or child minders, but the older children were left to eat on their own by cup or spoon under the supervision of their mothers. In all cases, care was taken not to force the food on the child. If the child wanted more food, it was served in a weighed portion. After the child finished eating, left-overs, if any, were weighed and recorded. Subtracting the left-overs from the recorded amounts served gave the food intake during the meal.

The group feeding was carried out on six days over a two-month period in order to demonstrate the use of power flour in weaning-food preparations to the mothers and to familiarize the children with the bulk-reduced diet.

Food-intake measurements were conducted in the homes after completion of the period of the adoption study described above. They were carried out on three consecutive days once a month over a period of six months, as follows: On the first day, in each test household, the children were served the liquid gruel (5% flour). On the second day, they were served the porridge containing 20% flour, treated with power flour to thin it down to a semi-liquid consistency. On the third day, they were served the thick porridge (20% flour). All ingredients were weighed and left for the .mother to prepare. The order of the servings was varied randomly. The liquid porridges were served in a cup, and the thick porridge was taken by hand, either by the child itself or with help from the mother. The meal was served between 9:00 and 10:00 in the morning, which implied that the younger children would have been breast-fed earlier in the morning. Food-intake measurements were carried out only if the children were free of any infection.

Statistical levels of significance in food intake between the diets were determined using Wilcoxon's non-parametric method with signed rank test [17] as described by Colton [18].


Adoption of the power-flour technique

Table 1 shows the frequency of using the power-flour method to prepare weaning food over period of three months by 40 mothers in Luganga village. The group that used the method "almost every time weaning food was prepared" increased from 13% to 28% over the period. The percentage of the women who used the germinated flour a quarter of the time or more increased from 48% to 85%.

TABLE 1. Frequency of use of power flour by mothers in preparing weaning food (percentages)


1st month

2nd month

3rd month

Almost every time




Half the time




A quarter of the time




Several times




Do not remember








Food intake in pre-school children

Table 2 shows the food intake of the three test diets for each child. Each value is an average of six measurements, taken once per month over a period of of six months. Overall there was wide variation in individual food intake within each age group, and, as was to be expected, the older children were able to eat more per meal.

Table 3 presents a summary of the food-intake data by age groups and a statistical evaluation.

In the group 5-12 months old, which had only four children, the average food intake was about the same for all three diets (range 153-163 g per meal).

In the group 12-24 months old the intake of the porridge treated with power flour was significantly higher (p c .05) than that of the thick porridge with the same flour concentration: 347 + 94 g compared with 277 + 93 g per meal. It was even slightly higher than the overall mean intake of the liquid gruel with 5% solids, 330 + 118 g per meal.

In the 24-48-month group the overall mean intake of the liquid gruels - that with 5% solids and that with 20% solids treated with power flour - was significantly higher (p < .05) than that of the thick porridge.

The 48-65-month group had less variation in individual food intake, and there was no significant difference in food intake between the three diets, which averaged 542 g per meal.

TABLE 2. Intake (grams per meal) of maize/ground-nut porridges with different dietary-bulk properties by pre-school children

Age group and child's number Thin gruela Thick porridgeb Liquefied porridgec

5-12 months

1 63 68 58
2 187 143 151
3 123 158 133
4 241 243 310

>12-24 months

5 355 415 383
6 266 314 346
7 271 228 279
8 230 161 287
9 201 213 213
10 352 199 335
11 568 396 522
12 399 292 409

>24-48 months

13 156 226 307
14 466 294 296
15 318 309 288
16 282 333 320
17 626 525 608
18 751 673 746
19 702 528 629
20 455 441 465
21 700 276 399
22 458 444 478

>48-65 months

23 527 481 727
24 535 487 451
25 548 466 553
26 320 324 420
27 520 429 604
28 599 646 590
29 630 510 548
30 616 723 699
31 523 535 484
32 622 472 576

Mean individual values of six single-meal measurements taken once per months for six months.
a. Gruel with 5% solid matter.
b. Porridge with 20% solid matter.
c. Porridge with 20% solid matter with 1% germinated sorghum flour added.

TABLE 3. Statistical summary of food-intake data by age groups Intake (grams per meal)

Age group (months) N Thin gruela Thick porridgeb Liquefied porridgec
5 - 12 4 154 ± 77d 153 ± 72d 163 ± 106d
> 12 - 24 8 330 ± 118d,e 277 ± 93d 347 ± 94e
> 24 - 48 10 491 ± 201d 405 ± 141e 445 ± 172d
> 12 - 48 18 420 ± 184d 348 ± 135e 401 ± 148d
> 48 - 65 10 544 ± 90d 517 ± 107d 565 ± 99d

a. Gruel with 5% solid matter.
b. Porridge with 20% solid matter.
c. Porridge with 20% solid matter with 1% germinated sorghum flour added.
d, e. Wilcoxon signed rank test for paired samples [17]. Mean values within the same row not having a common superscript are significantly different at p < 05.


Acceptance of the power-flour technique

The use of power flour to prepare a bulk-reduced liquid weaning food was accepted by almost all the mothers, possibly because we used an indigenous technology to tackle a recognized problem. On the other hand, only half of the mothers used the power-flour technique frequently - meaning at least once a day - over the three-month period. It was recognized by the village health committee that a more frequent use of power flour could be expected if the village nutrition extension worker were trained with special emphasis to promote it and if it were used in preparing the meals served at the village daycare centre. It was felt important that the children should be weighed regularly and their weight recorded on a growth chart available to the mothers so that they could observe the benefits of proper feeding of their children.

Food intake

The amounts of food consumed in one meal by the children in this study were within the range of values reported by others on children fed similar diets [4; 6; 19; 20]. There are, however, few reported studies comparing food intake of bulky and bulk-reduced weaning gruels. Araya et al. [21] found no correlation between the energy density (range 0.47-0.86 kcal per gram) and food intake in pre-school children served different lunch meals. Those meals, however, were not porridges or gruels; they were mainly composed of cereals, legumes, and tubers with vegetables.

In this study all the children consumed both the bulk-reduced gruel with 20% solids and the 5% gruel, each with a liquid consistency, in equal amounts. However, the thick porridge of 20% solids was consumed significantly less by the children between 12 and 48 months old. It seems, therefore, that the limiting factor was the viscosity (consistency) and not the energy density of the gruel. This is in agreement with the findings of Svanberg et al. [6] on Ethiopian children fed the sorghum-based weaning food faffa, bulk-reduced by the addition of 5% germinated sorghum flour, and of Gopaldas et al. [20], who served young children a liquefied wheat porridge with fat.

It is important to note that the children between one and two years old in this study were the ones most sensitive to the consistency of the diet. This is the age group in which supplementary feeding is supposed to cover the larger part of the energy and nutrient requirements, and also in which the malnutrition problem reaches its peak.

The weaning food usually prepared by mothers in Tanzania is a liquid gruel, and such a "rue] obtained by using power flour to thin down a thick porridge will provide four times more energy per unit volume than the traditional uji. Furthermore, the traditional feeding method in the villages is by cup or spoon, and introducing a more energy-dense liquid weaning food prepared using a traditional household technology is not likely to cause the problems associated with bottle-feeding and commercial weaning foods.

TABLE 4. Estimated number of meals required to meet 60% of the daily energy needs of a one-year-old child using maize/ground-nut (95:5) porridges with different dietary bulk properties


Volume needed (ml)

Actual intake per meal (ml)

Meals needed per day

Thin gruel a




Thick porridge b 870 277 3. 1
Liquefied porridge c 710-870 347 2-2.5

Based on a requirement of 1.080 kcal per day for a one-year-old child, according to local weight standards [22] and FAO/WHO/UNU 1985 [7] energy-requirement standard.
a. Gruel with 5% solid matter.
b. Porridge with 20% solid matter.
c. Porridge with 20% solid matter, with 1% germinated sorghum flour added.

The importance of this finding is demonstrated even more if the intake figures obtained in this study are used to estimate the actual amount of different weaning gruels needed per day in order to meet the energy requirements of young children. Table 4 demonstrates the difference in food intake needed to cover 60% of the daily energy requirements for a one-year-old child given gruels with different dietary-bulk properties.

It is assumed that breast milk will cover the rest of the requirements. More than 3,500 ml of the 5% maize/ground-nut gruel is needed per day, an amount obviously impossible for a one-year-old child to ingest. Of the thick 20% porridge, about 870 ml per day is needed. This could be eaten in three meals according to the observed average intake of 277 ml per meal. Flour concentrations of between 20% and 25% can be liquefied by the addition of germinated cereal flour. The amount of such gruels needed would then be in the range of 710-870 ml per day, and thus a 60% energy coverage could be achieved with only two or three meals a day. The importance of such a high-energy-density weaning food is emphasized by data from nutrition surveys in Tanzania indicating that children under five years of age are usually fed only two or three times a day.


The authors thank the managing director and staff of the Tanzania Food and Nutrition Centre, and the residents and leaders of Luganga village for their very good co-operation.


  1. Government of the United Republic of Tanzania and United Nations Children's Fund (UNICEF). Analysis of the situation of children and women. 1985;1.
  2. Waterlow JC. Classification and definition of protein energy malnutrition. In: Beaton BG. Bengoa JM. eds. Nutrition in preventive medicine. Geneva: World Health Organization. 1976.
  3. Mata LJ, Kromal RA, Urrutia JJ, Garcia B. Effect of infection on food intake and the nutritional state: perspectives as viewed from the village. Am J Clin Nutr 1977;30: 1215-27.
  4. Rutishauser IHE, Frood JDL. The effect of a traditional low-fat diet on energy and protein intake, serum albumin concentration and body-weight in Ugandan preschool children. Br J Nutr 1973;29:261-68.
  5. Ministry of Health. A pilot survey of the nutrition of young children in 1963. Rep Publ Health Med Subjects (London) 1958;11X.
  6. Svanberg U, Fredrikzon B, Gebre-Hiwot B, Taddesse WW. Sorghum in a mixed diet for preschool children. 1. Good acceptability with and without simple reduction of dietary bulk. J Trop Pediatr 1987;33: 181-85.
  7. Report of a Joint FAO/WHO/UNU Expert Consultation. Energy and protein requirements. Technical report series 724. Geneva: World Health Organization, 1985.
  8. Macy IG. Kelly GJ. Human milk and cow's milk in nutrition. In: Kon SK, Cowie AT, eds. Milk: the mammary gland and its secretion. New York: Academic Press, 1961:265.
  9. Church M. Quality and quantity of diet in relation to child development. Proc Nutr Soc 1979;38:41-49.
  10. Nicol BM. Protein and calorie concentration. Nutr Rev 1971 ;29:83-88.
  11. Binns CW. Food volume, a limiting factor in nutrient intake in the Papua New Guinea highlands. Proceedings of the Food Conference, Papua New Guinea University of Technology, Lae, Papua New Guinea, 1975.
  12. Ljungqvist BG, Mellander O, Svanberg U. Dietary bulk as a limiting factor for nutrient intake in preschool children. 1. A problem description. J Trop Pediatr 1981 ;27:68-73.
  13. Helström A, Hermansson A-M, Karlsson A, Ljungqvist B, Mellander O, Svanberg U. Dietary bulk as a limiting factor for nutrient intake in pre-school children. 11. Consistency as related to dietary bulk: a model study. J Trop Pediatr 1981 ;27: 127-35.
  14. Brandtzaeg B, Malleshi NG, Svanberg U, Desikachar HSB, Mellander O. Dietary bulk as a limiting factor for nutrient intake in pre-school children. III. Studies of malted flour from ragi, sorghum and green gram. J Trop Pediatr 1981;27:184-89.
  15. Karlsson A, Svanberg U. Dietary bulk as a limiting factor for nutrient intake in pre-school children. IV. Effect of digestive enzymes on the viscosity of starch-based weaning foods. J Trop Pediatr 1982;28:230-34.
  16. Mosha AC, Svanberg U. Preparation of weaning foods with high nutrient density using flour of germinated cereals. Food and Nutr Bull 1983;5(2):10-14.
  17. Wilcoxon F. Probability tables for individual comparisons by ranking methods. Biometric 1947;3:119.
  18. Colton T. Statistics in medicine. Boston, Mass, USA: Little, Brown & Co., 1974:219-27.
  19. Susheela TP, Narasinga Rao BS. Energy density of diet in relation to energy intake of preschool children from urban and rural communities of different economic status. Hum Nutr Clin Nutr 1983;37C:133-37.
  20. Gopaldas T, Deshpande S, John C. Studies on a wheat-based amylase-rich food. Food Nutr Bull 1988;10(3):5559.
  21. Araya H, Vera G, Pak N. Effect of dietary energy density on food intake of preschool children in one meal. Nutr Rep Int 1983;28:965-71.
  22. Rutishauser IHE. Heights and weights of middle class Baganda children. Lancet 1965;2:565.

Contents - Previous - Next