African Journal of Food Agriculture Nutrition and Development, Vol. 3, No. 2, November, 2003
WEANING FOODS AND PRACTICES IN CENTRAL UGANDA: A CROSS-SECTIONAL STUDY
Code Number: nd03012
Breast milk is the natural first food for infants and should be fed alone for the first 4 to 6 months of life. After 6 months however, breast milk alone is not sufficient, in quantity and quality, to maintain the growth and development requirements of the infant. Appropriate foods, referred to as weaning or complementary foods, need to be introduced while continuing breast feeding up to 24 months. This is the weaning process. Exactly when to wean, how to wean and what to wean with is a subject of great importance which in Uganda, has not received adequate scientific attention. Consequently, there is very little documented information on weaning foods and practices of the infants and young children in the country. The present study was therefore undertaken to assess and document what foods are fed to infants and young children, the weaning practices and the influencing factors. The study was a household cross-sectional survey utilising stratified multistage random sampling methods. The participants were mothers/caretakers of 261 young children aged 3-28 months. Diet assessment was done using conventional methods. The findings revealed that while breast feeding was universal at birth, early weaning with watery, energy- and nutrient-poor staples was widespread in this rural area of Uganda. The negative weaning practice of introducing complementary foods too early was highly prevalent with almost half of the children (44.1%) having started complementary feeding before the age of four months. Older mothers significantly (p≤0.0001) breast fed their children for longer periods than the younger mothers. Over half of the children (62.1%) commenced the weaning process with cow's milk. The weaning foods were dominated by the green cooking banana (matooke) which is known to be bulky with low nutrient content. Children from the rural areas consumed significantly more papaya (p=0.014), pumpkin (p≤0.0001) and matooke (p=0.007) than children from the urban areas while urban children consumed significantly more cows' milk (p=0.005), rice (p=0.008), sweet potatoes (p=0.018) and pineapples (p≤0.0001) than rural children. Consumption of animal protein, fruit and vegetables was found to be very low among this cohort of children, a possible risk factor for the high levels of micronutrient malnutrition among under-five children in Uganda.
Keywords: Weaning, infants, young children, complementary foods, exclusive breast feeding
Exclusive breast feeding for the first 4 months and if possible 6 months is the World Health Organization (WHO)’s recommended method of feeding full-term infants by healthy, well nourished mothers . The advantages that breast milk accords to the baby are well - documented [1-4]. However, after 6 months, breast milk alone is not sufficient, both in quantity and quality, to meet the nutritional requirements of the child especially for energy and micronutrients notably iron, zinc and vitamin A [5-7]. As the child grows older, therefore, it is necessary to supplement the breast milk with other foods, which start as liquid foods and slowly progress to solid foods. This is the weaning process: the introduction of foods other than breast milk into an infant’s diet while slowly reducing breastfeeding [1,2].
The introduction of supplementary foods is often accompanied by stress and ill health for infants in developing countries, mostly because the foods are not properly tailored to the infant needs [7,8,9]. Many traditional weaning foods in Africa are only a slight modification of adult foods [9,10], involving only mashing and dilution without taking into consideration the special nutritional requirements of young children. Adult diets, especially in developing countries, consist of highly starchy staples which are bulky and unless properly modified, unsuitable for infants and young children with their small gastric capacities . During the complementary feeding period, children require foods that are soft, hygienic and energy- and nutrient-dense to meet their high nutritional requirements [1,2,11,12]. Moreover, weaning foods in developing countries are usually prepared under un-hygienic conditions using water from unprotected sources thus exposing the child to weanling diarrhoea. This presents a dilemma to both mother and infant; to wean or not to wean, which is termed the “weanling’s dilemma”. Exactly when to wean, how to wean and what to wean with is a subject that has pre-occupied mothers and scientists alike, for a long time.
In Uganda, it has been reported that almost 70% of the children are already on supplementary foods by the 6th month of life, although breastfeeding continues well into the second year for most children [9,13,14]. The traditional weaning foods and weaning practices in Uganda [9,15] and indeed in many developing countries  are reported to be inadequate and may contribute to the high levels of malnutrition among the children.
Despite these reports, the subject of weaning foods and weaning practices has not received adequate scientific attention in Uganda. Consequently, there is very little documented information on weaning foods and practices of infants and young children in the country. We report results of a study undertaken to assess the weaning foods and practices in a banana/maize/bean-eating area in Central Uganda. Specifically, the study set out to assess and document the weaning/complementary foods fed to infants and young children in Central Uganda, the weaning practices and influencing factors. The results obtained can be applied to other parts of Africa with similar food and eating patterns.
MATERIALS AND METHODS
Study site, design, inclusion criteria and sample size
The study was conducted in five parishes in Busimbi sub-county located in Mityana county, Mubende district, Central Uganda. Stratified multistage random sampling design was employed to select the county, sub-county, parishes and the primary sampling units in this cross sectional study. The primary sampling unit was the household. A household was defined as a group of people living, cooking and eating together. Twenty one clusters/villages; 18 from the rural and 3 from the semi-urban environment (trading centre) were randomly selected for the study. Only the children who fell within the required age range (3-28 months), whose mothers or caretakers were present at the time of the interview and children who had started complementary feeding were randomly selected to take part in the study. Only one child was studied per household. A total of 261 children, 183 (70 %) from the rural area and 78 (30%) from the semi-urban environment, were studied.
The data were collected using a pre-coded questionnaire which was administered to the respondents in their home setting. The majority of the respondents (over 90%) were the mothers of the children. The questionnaire, specifically designed for this study, was validated by the jury method . This involved circulating the developed questionnaire to nine key informants in the country and beyond, knowledgeable in the area of childhood nutrition. Their comments were incorporated and a final version obtained. This questionnaire was then pilot tested in an area with a similar setting to that of the final study area.
The diet of the children was assessed using standard methods. The usual diet of the children was assessed using the diet history method whereby the mothers were asked if they usually fed listed foods to their children. The mothers were also asked to name the foods they fed the children right from the start of the weaning process and the corresponding ages of the children. In addition, the mothers were asked to name the most preferred weaning food, if there was no problem of availability and money constraints. A qualitative 24-hour food recall was also employed to capture all the foods and beverages consumed by the child 24 hours prior to the interview. In order to estimate the frequency at which the children consumed the foods, a qualitative food frequency questionnaire (FFQ) with 30 foods was also included in the main questionnaire.
The data were analysed using SAS (Statistical Analysis System, SAS Institute, Inc., Cary, USA). Frequency and percent distributions and descriptive statistics of means and standard deviations were carried out. Two-way and three-way cross tabulations were also carried out on some selected variables in order to determine the factors influencing the weaning practices of the children. Statistical significance was determined using Pearson’s chi-square test and correlation analysis was done using Pearson correlation coefficients. The food frequency scores of the food frequency questionnaire (FFQ) were analysed using SAS’ NPAR1WAY analysis of variance. This is non-parametric one way analysis of rank scores that tests the distribution of selected class variables between groups (in this case; rural vs. urban). The Kruskal-Wallis Test (Wilcoxon Scores) of chi-square approximation was employed to interpret significant differences in this analysis.
Child Demographic Factors
The children studied were between the age range of 3-28 months with a mean of 15.7 ±6.4SD months. For ease of sub-sequent analysis, the children were divided into 3 age categories; Group I was the infants (<12 months), group II the young toddlers (12-18 months) and group III the older toddlers (>18 months). The majority of the children (38.3%) fell in the 12-18 month age group, followed by those above 18 months (33.0%) and lastly the infants at 28.7%. There were slightly more female than male children (52.1 vs. 47.8%).
Breast Feeding Practices
Initiation of breastfeeding at birth with colostrum was found to be universal in both rural and semi-urban mothers. Over sixty percent (65.5%) of the children were still breastfeeding by the time of the study. The frequency of breast feeding was slightly higher in the day than in the night as almost half the breast fed group of children (47.8%) were breast fed over 10 times during the day compared to 36.9% in the night.
Thirty four percent of the children had been removed from the breast by the time of the study. The majority (43.3%) of these children were removed from the breast between the age of 18 and 24 months (fig. 1). A small proportion of children (10.0%) were removed from the breast quite early (before one year) while a relatively large proportion of children (26.7%) had been removed from the breast quite late (beyond 24 months) (fig. 1). Only one child was removed from the breast before 6 months of age. The reasons given by the mothers for termination of breastfeeding were variable. None of the mothers gave “inconvenience” as the reason for termination of breast feeding. Some mothers (10.7%) removed their children from the breast because they felt that breast milk was not sufficient, 8% because they were pregnant and 8.8% because the child was old. Only 3 mothers stopped breastfeeding because they had to go back to work and 9 stopped because they were ill.
Age of Introduction of Weaning/Complementary Foods
All the children in the study had started the weaning process, in that , they had been introduced to foods other than breast milk. Almost half the children (45.2%) started the weaning process between age 4-6 months (fig. 2). A large proportion of the children (42.5%) had been introduced to complementary foods too early, before the age of 4 months of age (fig 2).
The mothers were asked to name the foods they fed their children during weaning, in the order the foods were introduced, with the corresponding ages of the children. The majority of the children (62.1%) commenced the weaning process with cow’s milk followed by the East African Highlands green cooking banana locally known as matooke and cereal porridges (table 2). This is phase I of weaning (<6 months). Thereafter, matooke became the most important weaning food during phase II (6-9 months) and III (>9 months) of the weaning process followed by Irish potatoes, sweet potatoes and cereal porridges (Table 1).
Assessment of the Usual Diet of the children using the Diet History Method
Over sixty percent (66.3%) of the children consumed cows' milk as their usual diet (Table 2). Thirteen percent (13.4%) of the children consumed the milk undiluted and 42.9% consumed it diluted with water, usually in a 1:2 (milk to water) ratio. Only one household used fermented cow’s milk for child feeding. Eighty nine percent (89.1%) of the children were given maize porridge as their usual diet (Table 2) while 70.1 % were given millet porridge. Almost all the porridges (89.1%) were fed in the un-fermented/un-germinated state. In most cases (93.3%), the porridge was usually fed to the children with a “thin” consistency and in only 1.1 % was the porridge of a “thick” consistency. The majority of the mothers (72.3%) added water to reduce the viscosity to “drinking” consistency. To make the cereal porridges “richer” or more nutritious, the majority of the mothers (88.9%) reported that they added sugar while 69.0% added milk. Fifteen percent added soya, while 0.4% added groundnuts. Addition of oil/fat was done by 4.6% of the mothers and 5.4% added eggs.
The majority of the children (96.2%) took the porridges from a drinking cup. Only 2 children took their porridges from the recommended bowl/plate with a spoon. Four children took the porridge with a feeding bottle with a nipple. A very small proportion of children (0.8%) took their porridge through hand or force feeding. Most of the children consumed the porridges once to three times a day. The medium size of cup (200-250 ml) was the most commonly used size of cup and each child had his/her own cup set aside from the older children’s utensils. It was, therefore, possible to estimate the quantity of porridge the child consumed per day. The average quantity of beverages (porridge, milk and tea) consumed per child per day, estimated from the size of the cup and the number of times the child was given the beverage(s) per day, was found to be 763.7ml±381.7 (mean±SD) with a range of 0.0-2000 ml. This wide range with the corresponding high SD show the variability of the children’s beverage consumption. There was a positive and significant correlation between the amount of beverage consumption per day and the age of the child (r=0.272; P≤0.0001).
Matooke (cooked banana) was ranked as the most popular weaning food as a large majority of the children (96.6%) consumed it as part of their usual diet (Table 2). The other popular weaning staples included Irish potatoes, cassava and sweet potatoes (Table 3). The common sauces eaten with the staples were fish, beans (Phaseolus spp.) and groundnut. Consumption of meat and vegetables was very low (Table 3). The majority of the children (88.9%), usually consumed food from the family meals while a small proportion had food especially prepared for them. However, almost all the children had both options at one time or the other. Of those mothers who prepared food especially for the child, 34.5% prepared it freshly but the majority (46.0%) prepared the food once a day and a small proportion (1.5%) prepared it once in two days. Inconvenience and lack of time were the major reasons given by the mothers for not freshly preparing the child’s food (29.5% and 21.1%, respectively) while a small proportion (9.2%) reported that lack of fuel was the main reason.
The majority of children were fed solid foods three times a day (44.8%) followed by those who were fed two times a day (41.3%). A small proportion of children (13.4%) were fed only once a day. Usually, the mother mashes some food with her fingers, while she is eating, which she softens with sauce and feeds the child. In most cases, there is no specific time set aside for cooking or feeding the child.
Thirty one percent of the mothers, mostly in the semi-urban areas, were found to have used locally processed Baby Soya at one time or other during the weaning process. The only imported commercial baby food found in use was Cerelac® and this was used by only 2.3% of the mothers. Most mothers (82.0%) reported that there were no foods forbidden to children. A few (18.0 %), however, reported that some foods namely: mud-fish, yam, liver and pork were forbidden to children. Most children (94.3%) were fed and looked after by their own mothers. The role played by relatives, grandmothers, ‘house helps’, and older siblings was small: less than 3% in each case.
Assessment of the diet of the children using the Food Frequency Questionnaire (FFQ)
The FFQ had 10 frequencies: 1-7 for the number of days per week a food was consumed and 8 and 9 if a food was consumed once a fortnight or once a month, respectively. Score 10 was for foods which had never been consumed by the children. About one fifth (22.6%) of the children consumed cow's milk every day while an almost equal percentage (23%) consumed milk rarely at a frequency of once a month (Table 3). Eleven percent (10.7%) of the children had never consumed cow’s milk since they were born. Almost one quarter (24.5%) of the children consumed maize porridge while 12.3% consumed millet porridges, on a daily basis.
Matooke (cooked banana) was by far the most frequently consumed staple food with 21.5 % of the children consuming it daily, followed by cassava, maize meal, sweet potatoes, Irish potatoes and rice (Table 3). For the sauces, fish and beans dominated with 20.7% and 20.3%, respectively, of the children consuming these food items every day. The frequency of consumption of groundnut and meat was low with 5.7% and 1.1 %, respectively, of the children consuming them daily. The frequency of consumption of vegetables was very low with less than 5 % of the children consuming dark green leafy and coloured vegetables daily. The consumption of fruits was also very low except for passion fruit, commonly consumed as juice (Table 3).
The FFQ data was subjected to a one way analysis of rank scores in order to find out if the distribution of foods consumed by the children from the rural and the semi-urban environment differed significantly. The results showed that children from the semi-urban consumed significantly more rice (p=0.008), sweet potatoes (p=0.018) and pineapples (p≤0.0001) than rural children. Rural children, on the other hand consumed significantly more matooke (p=0.007), papaya (p=0.014) and pumpkin (p≤0.0001) than their semi-urban counterparts. There were significantly (p=0.005) more children from the rural areas who had never consumed milk than from the urban areas.
Assessment of the Usual Diet of the Children using the 24-hour Food Recall
A total of 40 different foods had been consumed by the children on the day preceding the interview. The majority of the children had consumed matooke and beans on the day preceding the interview (Table 4). Again, the consumption of fruit and vegetables was very low among this group of children.
Factors Influencing the Weaning Practices
Two weaning practices: age of introduction of complementary foods and age of termination of breastfeeding were selected for in-depth study. These variables were cross-tabulated with various selected demographic and socio-economic variables such as; age, education and occupation of the mother, number of children below 5 years, area of residence and economic status of the household, in order to assess their influence on the weaning practices.
Age of introduction of complementary foods
Education of the mother had a non-significant influence on the age of introduction of complementary foods with a larger proportion of mothers with a higher level of education introducing supplements earlier than their less educated counterparts (Table 5). The area of residence also had a non-significant influence on the age of introduction of complementary foods. The biggest proportion of mothers from the semi-urban areas introduced supplementary foods to their children very early (before 4 months of age) while the biggest proportion of mothers from the rural areas introduced supplements between 4 to 6 months. The age of introduction of supplementary foods was not influenced by the age or occupation of the mother, the number of children below 5 years a mother had nor the economic status of the household.
Age of termination of breastfeeding
The age of termination of breast feeding was significantly influenced (P 0.0001) by the age of the mother. None of the older mothers (>40 years) terminated breast feeding before 12 months and none of the teen-aged mothers breast fed their children beyond 24 months (Table 5). Education of the mother had a non-significant influence on the age of termination of breast feeding with the majority of mothers with no formal education breast feeding beyond 24 months and none terminating breast feeding before one year (Table 4). The occupation of the mother, the number of children below 5 years a mother had, the area of residence and the economic status of the household had no influence on the age of termination of breast feeding.
In this study, we sought to establish what foods were used for weaning, the weaning practices and factors influencing the foods and weaning practices in central Uganda.
Breast feeding and Weaning practices
The rate of breastfeeding was found to be very high in this part of Uganda as indeed it is for the rest of the country [13,14]. Initiation of breastfeeding at birth was universal, a health promoting practice as the first milk, colostrum has all the nutrients, energy and anti-bodies an infant needs for a good start in life. The momentum of breast feeding was found to be maintained as by 12 months, over 80% of the children were still being breastfed. These results are in agreement with the results of 1988/89 UDHS which found that the median duration of breast feeding was 19 months with 82 % of the women still breast feeding their children at 10 months, 42% at 20 months and 13% at 24 months . In the current study, the age of the mother had a significant association with duration of breast feeding, with younger mothers breast feeding for shorter periods than the older mothers in agreement with Ugandan  and Saudi studies . Education of the mother had a non-significant influence on the duration of breast feeding as mothers with no formal education breastfed their children for longer periods than mothers with secondary or higher education, consistent with previous Ugandan , Nigerian  and Kuwaiti  studies.
A large proportion of children (70.0%) were breastfed for prolonged periods (>18 months) in agreement with Vella’s work  from Northern Uganda. The practice of early termination of breastfeeding (before 12 months) was found to be negligible. Although it is recommended that breast feeding should be continued up to 2 years of age, some workers [19-21] have found that prolonged breast feeding beyond 18 months has been associated with malnutrition of the children. Nonetheless, the question of prolonged breast feeding beyond infancy is still controversial with some encouraging it , while others discourage it .
In this study, almost half of the children (44.1%) started complementary feeding before the recommended age of 4 months. This age is too early for weaning, as the child’s alimentary system is not yet ready for the highly starchy foods commonly used for weaning in developing countries [24,25]. Although there might be a small proportion of children who might need supplemental feeding at 2-3 months , caution is needed when introducing food to an infant earlier than 4 months under unhygienic conditions prevailing in developing countries [8,26].
A randomised controlled study by Cohen and co-workers  established that children who are exclusively breast fed from healthy, well nourished mothers may not benefit from complementary foods before the age of 6 months. The question of the length of exclusive breast feeding is quite controversial. Although WHO recommends exclusive breast feeding for the first 4-6 months, they would actually prefer it to be the first 6 months . However, in the African setting, the completely exclusively breast fed infant, especially for a full 6 months, is rare. Mothers usually give some water, juice, cow’s milk, cereal porridges etc in addition to breast feeding. Indeed our results show that almost 90% of the children (89.3%) started complementary feeding before the age of 6 months.
In the current study, the negative weaning practice of starting the weaning process late (introducing supplementary foods later than 7 months), common in many African cultures , and which has been found to be associated with child malnutrition, was found to be negligible. Although there were some foods forbidden children, food taboos were not of important magnitude in this part of Uganda.
Further analysis of the factors influencing the age of weaning revealed that the area of residence had a non-significant influence with proportionately more urban mothers introducing weaning foods to their children earlier than rural mothers. Education of the mother too had a non-significant influence on the age of weaning with the better educated mothers introducing supplements earlier than the less educated mothers, in agreement with experience from Nigeria [10, 17] and Kuwait . Mothers with a higher education level usually work away from home, hence the need to introduce supplements earlier.
The economic status of the household had no influence on the age of weaning of the children in agreement with a previous Ugandan study , but in contrast with the findings from Nigeria .
Almost two thirds of the children started the weaning process on to cow’s milk. The question of unmodified cow’s milk and weaning is controversial. Many nutritionists in developing countries recommend cow’s milk to supplement breast milk after 6 months of age (despite the low iron content in cow’s milk) mostly because starchy staple alternatives are much worse. However, in most cases, the families are too poor to afford even the cow’s milk. In fact, in the current study, the frequency of consumption of cow’s milk was found to be low with 33.7% of the children consuming cow’s milk rarely and 10.7% of the children never having consumed cow’s milk since birth. Our previous results on this same cohort of children revealed a significant association between never having consumed milk since birth and high rates of stunting 
The East African green cooking banana, matooke, was by far the most commonly used solid weaning food in phase II (6-9 months) and phase III (>9 months) of weaning. Over 90% of the children were fed matooke as part of their usual diet in addition to Irish potatoes, cassava, sweet potatoes, maize meal and rice, combined with sauces made up mostly of beans and fish. Consumption of animal protein, fruit and vegetables, rich sources of micro-nutrients, was found to be very low, in both rural and the semi-urban areas. This could be a contributing factor to the high prevalence of micro-nutrient deficiencies, particularly Vitamin A deficiency (VAD) and iron deficiency anaemia (IDA) estimated at 28% and 50 %, respectively, among under-five children in Uganda .
Cereal porridges, particularly maize porridges, were also found to be important foods in phases I and II of weaning. However, a very high proportion of the porridges (93.9 %) were fed to the children with a “very thin” consistency using a drinking cup instead of the recommended bowl and spoon . The energy density of this “watery” porridges is very low (11,31). The problem of dilute or watery porridges for children in Uganda is a very serious one, and may be contributing to the high levels of childhood malnutrition identified in many nutrition and health surveys [13,14,21,29]. To compound the problem, the porridges were not adequately supplemented with energy- and nutrient-rich supplements such as milk, groundnut and eggs. Furthermore, the food processing practices of germinating and/or fermenting the cereals used to make the porridges, which have been found to decrease viscosity while increasing the energy density of cereal weaning porridges , were not common in this part of Uganda.
The negative practice of feeding with bottles with a nipple associated with high infant mortality  was negligible in this study. Use of imported commercial weaning foods, a known risk factor for infant malnutrition in low income countries was found low in this study mainly because of the largely rural sample.
While breast feeding was universal at birth, early weaning with watery, energy- and nutrient- poor staples was found to be widespread in this rural area of Uganda. The weaning foods were dominated by matooke which is known to be bulky with low nutrient content. Consumption of fruit and vegetables was extremely low with negative implications to the micro-nutrient nutrition of children in Uganda. It is recommended that mothers and caretakers of young children be sensitized on the importance of proper nutrition to the growth and health of their children. They should also be trained and equipped with child care and feeding skills to enable them combine the variety of locally available food ingredients to make nutritious foods for their children.
Copyright 2003 - Rural Outreach Program