This is the old United Nations University website. Visit the new site at http://unu.edu
Methods
Results
Discussion
Conclusions and recommendations
References
M. M. Meme, W. Kogi-Makau, N. M. Muroki, and R. K. Mwadime
M. M. Meme is affiliated with the Department of Foods, Nutrition and Dietetics in the Faculty of Home Economics in Kenyatta University and is a tutor at the Machakos Technical Training Institute in Nairobi, Kenya. W. Kogi-Makau, N. M. Muroki, and R. K. Mwadime are affiliated with the Applied Nutrition Programme in the Department of Food Technology and Nutrition at the University of Nairobi.
Abstract
The dietary intake and nutritional status of 162 children in a school with a lunch programme (the feeding-programme group) and 163 children in a school without a lunch programme (the no-feeding-programme group) in Nyambene District, Kenya, were compared. The relationship between such child growth determinants as income sources, per capita weekly food expenditure and consumption frequency, per capita energy and protein intake, and the nutritional status of the children was also compared between the two groups. Daily caloric consumption in the group with a feeding programme was significantly higher than in the group without a feeding programme: 1,590 kcal, or 86% of the recommended daily allowance (RDA), versus 1,457 kcal, or 76% of the RDA (p <.05). The protein intake was mainly of plant origin. Although not significantly different between the two groups, it was higher for children without a feeding programme (62 g; 238% of the RDA) than for those with a feeding programme (56g; 216% of the RDA). The prevalence of wasting among children with a feeding programme (9%) was significantly higher than among those without a feeding programme (2%) (p <.05). The level of stunting was about the same in both groups: 24% in the group with a feeding programme and 25% in the group without a feeding programme. There was no significant difference in the prevalence of underweight between the two groups. Overall, the nutritional status of girls was better than that of boys, although the difference was not statistically significant. It is evident that children participating in the feeding programme did not have a nutritional advantage over non-participants. Thus, there is need to evaluate school feeding programmes in Kenya to identify and address the weaknesses that curtail their impact.
Introduction
Protein-energy malnutrition, mainly due to inadequate dietary intake, is the major form of malnutrition among school-age children in Kenya [1]. The negative impact of malnutrition on learning here as well as in other countries has been well documented [2-5]. A survey by the Central Bureau of Statistics showed that malnutrition is widespread among schoolchildren in Kenya and is consistently higher among boys than among girls in Kwale and Kitui districts [6]. In another study in Kenya, the nutritional status of schoolchildren was shown to deteriorate with age [7]. Similar observations were reported for Tanzania [8].
In addition to improved academic achievement, improved nutrition as a result of school feeding programmes has been shown to improve school enrolment and attendance [9-11]. School feeding programmes have been initiated in a number of countries [12]. Most of these programmes have three main objectives: to improve school attendance, school performance, and the nutritional status of the children. However, past experience has shown that many nutrition interventions, including school feeding programmes, are usually not based on information from research findings. Further, in many cases the implementation process has met with a number of problems. Consequently, the results, in terms of cost, effectiveness, and magnitude of impact, have been disappointing [13]. In an evaluation of a school feeding programme in India, its irregularity was found to be responsible for poor school enrolment and attendance [14]. Observations in Kenya have shown that school feeding programmes do not necessarily improve enrolment levels nor do they improve the educational performance or the nutritional status of the children. In some cases, however, the programmes may improve school performance, nutritional status, or both. A National School Feeding Council of Kenya school feeding programme was shown to improve the nutritional status of participating children [15].
A Government of Kenya/World Food Programme school feeding programme was introduced in 1981 and is still in operation. The objectives of this programme were essentially similar to those given above. It covers arid and semi-arid areas as well as divisions of the less arid and low-potential areas in which food production is low. These areas, which have food deficits, are poor and have school enrolment levels that are below the national average of 87%. During the first school term of 1995, the programme covered 19 districts in the Rift Valley, Coast, Northeastern, and Eastern Provinces, with a total enrolment of 360,000 children.
As a result of a shortage of funds at both the Ministry of Education and District headquarters, there has been no monitoring and evaluation of the programme’s performance. The Ministry of Education has been unable to meet the agreed 50% of the internal transport, handling, and storage costs [16]. The nutritional impact of the school feeding programme has not been evaluated [17]. The objective of this study was to assess the effect of a school lunch feeding programme on the nutritional status of the children and school attendance levels in Nyambene District. Household factors likely to influence child nutritional status were controlled for in the analysis. The hypothesis was that children in a school participating in the feeding programme would, by the end of the school term, have better nutritional status and school attendance records than those in a school without a feeding programme. The relationship between some of the child growth determinants and nutritional status was also examined and compared with that of children in the school without a feeding programme.
Study site, population, and subjects
The study subjects were children aged 5 to 10 years in a randomly selected school (Lukununu primary school) in Antuambui, Laare division of Nyambene District, eastern Kenya, participating in a World Food Programme school lunch feeding programme (feeding-programme group) and children in a school (Mwerongundu primary school), purposely selected from a number of schools without a feeding programme (no-feeding-programme group) in the same area.
Nyambene District is 400 km from Nairobi and has an area of 12,000 km2. Low and unreliable rainfall together with low altitude and high temperatures classify this area as semi-arid. This is a constraint to the production of staples (maize and beans), which has made the district dependent on food from other districts [18]. Antuambui and Laare have areas of 181 and 1,112 km2, respectively. Antuambui has a population of 18,864, a population density of 104 per square kilometre, and 3,452 households [19]. The paradoxical wealth of Nyambene is derived from the sale of Catha edulis, i.e., khat or miraa (as it is known locally), a slightly intoxicating herb.
Statistics from 1992 indicate that miraa is the most important cash crop in this district. It covers 5,200 hectares around Maua Hills, which include the divisions of Tigania, Ntonyiri, Laare, and Antubetwe. In Laare, miraa covers approximately 1,000 hectares [18]. Most of the miraa is sold in urban areas in Kenya, while some is exported to Middle East countries. An average household earns about KSh 12,000 (US$200) per month from the crop, as compared with an estimated gross national product of US$295 per capita per annum based on 1998 figures. Miraa is traditionally a business for men, most of whom make merry with the proceeds from it in towns and shopping centres. Thus, despite the high earnings from miraa, the money does not contribute to food security and is of limited benefit to women and children. Casual jobs related to the crop are readily available.
Field study instruments
A structured questionnaire was developed to elicit information on demographic and socio-economic characteristics of the households. The questionnaire also sought information on foods or meals consumed by the household and by the index child during the previous 24 hours, the amounts of ingredients used for preparing the foods or meals, food expenditures, and the sources and frequency of intake of various foods: cereals (maize, millet, and sorghum), legumes (kidney beans, peas, horse beans, and green gram), vegetables (cabbage, kale, green English peas, green pigeon peas, green beans, spinach, and carrots), livestock products (meat, eggs, and milk), fruits (orange, avocado, tomato, and passion fruit), root and tuber crops (Irish potato, sweet potato, and arrowroot), and plantains.
Weights were measured with a 100-kg capacity bathroom Salter scale with increments of 100 g. A vertical measuring rod with a length of 175 cm and a precision of 0.1 cm was used for measuring height. School records for the 65-day term were used to obtain information on the level of attendance.
Sample size and sample determination
The sample size (169 children from each school) was determined according to Fisher et al. [19] using a prevalence of stunting among older children (5-10 years) of 30% reported by the Central Bureau of Statistics for the area [20] and an assumption that the difference in malnutrition between the two groups would be significant at the .05 level.
Conduct of the study
Sample selection
Nyambene District was selected randomly from among 19 districts with World Food Programme school feeding programmes. Central Laare was then randomly selected from the three World Feeding Programme-designated feeding zones in the district, Igembe South, Igembe North, and Central Laare. Finally, Lukununu primary school was randomly selected from the three primary schools in Central Laare with a feeding programme.
A primary school without a feeding programme (Mwerongundu primary school) near the Lukununu primary school was selected to ensure that its characteristics were similar to those of the school with a feeding programme.
Registration of all children between 5 and 10 years of age in the schools was then obtained. In each school, 169 children were randomly selected for the study, ensuring that, in the case of the school with a feeding programme, these had participated in the feeding programme for at least a year. Sixty-two (38%) of the children in each school were also randomly selected for the 24-hour dietary recall study.
Training of interviewers and pilot study
Two enumerators from the study area who spoke the local language were trained in interviewing techniques to collect biodata on the children, anthropometric measurements, and dietary intake data. The enumerators were used to pretest the questionnaire, which was adjusted accordingly before the definitive study began.
Definitive study
The definitive study was conducted during the school term from January to March 1995. The field assistants were closely supervised by the researchers, and the questionnaires were examined every day to check on the validity and reliability of the information collected.
Administration of the questionnaire
The selected children gave directions to their homes so that the enumerators could administer the questionnaire to their mothers and, where applicable, record the 24-hour dietary recall information. Since the enumerators were from the area, they easily found the homes. Household measures and food models, as described by Cameron and Van Staveren [21], were used to assess the amounts of food prepared or consumed.
Intake at lunchtime in school was observed for children in the feeding programme (Lukununu primary school). All the ingredients used for the meal and their weights were recorded. The amounts consumed by individual children were determined using the “observed weighed technique” as described by Cameron and Van Staveren [21]. This technique involved weighing the food before serving and weighing the food remaining on the plate afterwards. The amount of food consumed by each child was expressed as a proportion of the total amount of food that had been prepared for all the children, to calculate the amounts of the ingredients in the food for each child.
The information on the amount of food consumed at lunch by the children in the school without a feeding programme (Mwerongondu primary school) was collected from their mothers. Intakes from breakfast and supper meals for both groups of children were assessed at their homes.
Anthropometric measurements
Weights and heights were measured as described by the United Nations National Household Survey Capability Programme Manual [22] and recorded to the nearest 100 g and 0.1 cm.
School attendance
Each child’s attendance was determined by reviewing the school records at the end of term.
Data entry and analysis
Preliminary data cleaning was done in the field with the assistance of the enumerators. At the end of each day, the researchers checked all the questionnaires to ensure that all the data had been collected; if any were missing, the households were revisited. The questionnaires were then sent to the Applied Nutrition Programme of the University of Nairobi where the data were entered and analysed with the Statistical Package for the Social Sciences (SPSS).
Estimation of caloric and protein intake
The Eastern, Central, and Southern Africa (ECSA) Technical Center for Agriculture and Rural Cooperation (CTA) food-composition tables for Eastern and Central Africa [23] were used to calculate the total daily intake of calories and proteins by household and by child, using data from the 24-hour dietary recall and, for children attending the school with a feeding programme, the ingredients of the school lunch. The Kenya food-composition tables compiled by Sehmi were used for any food items not found in these tables [24]. The combined caloric and protein contributions of breakfast and supper (two meals consumed at home by all children) were also computed according to the same food-composition tables. The adequacy of protein and caloric intake was expressed as the proportion of the recommended daily allowance (RDA) according to the Food and Agriculture Organization/World Health Organization/United Nations University [25]. Household per capita energy and protein intake was then computed.
Assessment of nutritional status
The Centers for Disease Control ANTHRO computer programme was used to compute Z score deviations of the weights or heights from the weights and heights of the National Centre for Health Statistics (NCHS) reference children of the same height (or age). The indices obtained were weight-for-age, height-for-age, and weight-for-height, as indicators of underweight, stunting, and wasting, respectively. The children were considered to be malnourished when the respective Z scores were below -2 SD from the median for NCHS reference children.
General characteristics of study households and parental characteristics
The general characteristics of the study population are shown in table 1. The study groups were similar in household size and dependency ratio, and the sex ratio in both areas was 1:1.
The fathers and mothers of the children with a feeding programme were significantly older (44.3 ± 9.4 and 35.3 ± 7.2 years, respectively) than the parents of the children without a feeding programme (41.6 ± 8.1 and 32.7 ± 6.4 years; p <.01 and p <.005, respectively). In both groups, the level of education of the parents, as measured by the number of years of schooling, was generally low, and the fathers had more years of schooling than the mothers. The fathers and mothers of the children without a feeding programme had significantly more years of schooling (4.0 ± 4.5 and 2.3 ± 3.4 years, respectively) than the parents of the children in the feeding-programme group (2.7 ± 3.5 and 1.7 ± 2.8 years; p <.005 and p <.05, respectively).
Characteristics of the study subjects
The distribution of the children by sex and age is shown in table 2. The mean age of the children was about the same for both groups (8.2 ± 1.2 years for children with a feeding programme and 8.3 ± 1.2 years for children without a feeding programme). Absenteeism of children in the school with a feeding programme (8.6 ± 7.5 days per 65-day school term) was significantly higher than that of children in the school without a feeding programme (3.1 ± 3.9 days; p <.001).
Table 1. Demographic characteristics of the study households (mean ± SD)
|
Characteristic |
School feeding programme |
t |
|
|
Yes (n=162) |
No (n=163) |
||
|
Household size |
8.2 ± 2.6 |
8.3 ± 2.5 |
-0.3 |
|
Dependency ratio |
1.2:1.0 ± 0.7 |
1.0:1.0 ± 0.6 |
1.6 |
|
Father’s age (yr) |
44.3 ± 9.4a |
41.6 ± 8.1 |
2.7 |
|
Mother’s age (yr) |
35.5 ± 7.2a |
32.7 ± 6.4 |
3.4 |
|
Father’s education (yr) |
2.7 ± 3.5 |
4.0 ± 4.5a |
-3.1 |
|
Mother’s education (yr) |
1.7 ± 2.8 |
2.3 ± 3.4a |
-1.8 |
Household income sources and food expenditure
Table 3 shows the distribution of households by sources of income, ranked in order of importance and also by per capita household food expenditure. The main source of income was miraa, which was reported by the same proportion of households in both groups (60% and 61% in the programme and non-programme groups, respectively). There was also no significant difference in the proportion of households who derived part of their income from business (23% and 22% of households whose children attended schools with and without feeding programmes, respectively). The sale of food crops was reported by a small proportion of households (4% of feeding-programme households and 6% of no-feeding programme households) and was not significantly different between the two groups. Employment was an important source of income for only a small number of households in both groups (9% of the feeding programme households and 10% of the no-feeding-programme households). Sale of animals and their products was the least important source and involved very few households: 3% and 1% for the programme and no-programme households, respectively.
Table 2. Distribution of the children in the schools with and without feeding programmes according to age group, mean age, and school absenteeism rate
|
Age group (yr) |
Feeding programme |
No feeding programme |
||||
|
Boys |
Girls |
% |
Boys |
Girls |
% |
|
|
5-6 |
17 |
19 |
22 |
14 |
15 |
18 |
|
7-8 |
31 |
43 |
46 |
32 |
42 |
45 |
|
9-10 |
22 |
30 |
32 |
28 |
32 |
37 |
|
Total |
70 |
92 |
100 |
74 |
89 |
100 |
|
Age (yr)-mean ± SD |
8.2 ± 1.2 |
8.3 ± 1.2 |
||||
Table 3. Distribution of households according to source of income and food expenditure
|
Variable |
School feeding programme |
Total |
|
|
Yes |
No |
||
|
Source of income-no.(%) |
|
|
|
|
Miraa |
97 (60) |
99 (61) |
196 (60) |
|
Business |
38 (23) |
35 (22) |
73 (23) |
|
Casual employment |
11 (7) |
8 (5) |
19 (6) |
|
Sale of food crops |
7 (4) |
9 (6) |
16 (5) |
|
Permanent employment |
4 (3) |
11 (7) |
15 (5) |
|
Sale of animals and animal products |
5 (3) |
1 (1) |
6 (2) |
|
Weekly household food expenditure (KSh)-mean ± SDa |
45.6 ± 37.2 |
52.2 ± 42.2 |
t = -1.5 |
Food expenditure was low: KSh 45.6 ± 37.2 in households whose children participated in the school feeding programme and KSh 52.2 ± 42.2 in households whose children did not participate. Food expenditure did not differ significantly between the two groups.
Caloric and protein intake by households and children
In feeding-programme households, the per capita daily caloric and protein intakes were 3,010 kcal and 139.5 g. The values in no-feeding-programme households were slightly higher, 3,320 kcal and 148.5 g, but there was no significant difference between the two groups.
The school feeding programme was designed to provide the children with 150 g of maize, 40 g of beans, and 15 g of vegetable oil, which would provide 793 kcal (43% of the RDA) and 24 g of protein (92% of the RDA). Because of a shortage of ingredients, the school lunch programme operated for only 38 of the 65 days of the school term. The children therefore obtained only 59% of the energy and nutrients they would have if the programme had operated throughout the term.
The lunchtime caloric intake of the children with a feeding programme (860 kcal) was significantly higher than that of the children without a feeding programme (666 kcal) (table 4) (p <.05), but the protein intake from home lunch of the children without a feeding programme (31 g) was significantly higher than the lunchtime protein intake of those with a feeding programme (24 g) (p <.01). The total caloric and protein intakes were not significantly different.
A paired t test on the caloric intake for the feeding programme children showed a significantly higher intake at lunch (860 ± 1.0 kcal) than at supper (556 ± 326.1 kcal) (p <.001). No significant difference was found in protein intake for the two meals. A paired t test for lunch and supper at home for the no-feeding-programme children, on the other hand, did not show any significant difference for either caloric or protein intakes. The mean total protein consumption by children without a feeding programme (238 ± 129% of the RDA) was slightly but not significantly higher than that of the group with a feeding programme (216 ± 109% of the RDA).
Children in both groups consumed fewer calories and more protein than the RDA. Breakfast contributed less than 10% of the RDA for calories. The caloric adequacy for the children with a feeding programme (86 ± 21% of the RDA) was significantly higher (p <.05) than that for the children without a feeding programme (76 ± 27% of the RDA; p <.05) (table 4). This was because the lunch meal contributed significantly more calories for the group with a feeding programme (46 ± 3% of the RDA) than that for the group without a feeding programme (35 ± 17% of the RDA; p <.001). The contribution of breakfast and supper (two meals that children in both groups took at home) to the RDA for energy and protein did not show any significant difference between the two groups (table 4).
Table 4. Twenty-four-hour intake of energy (kcal) and protein (g) and adequacy of intake by children a
|
Meal |
School feeding programme |
t |
||
|
Yes (n=162) |
No (n=163) |
|||
|
Breakfast |
||||
|
|
Energy |
173A (9.3) |
187A (9.7) |
-0.5 |
|
|
Protein |
5.2a (20.4) |
5.8a (22.5) |
-0.8 |
|
Lunch |
||||
|
|
Energy |
860C (46.3) |
666B (34.6) |
4.5b |
|
|
Protein |
24.0b (94.3) |
30.8c (118.7) |
-2.6b |
|
Supper |
||||
|
|
Energy |
556B (30.1) |
605B (31.7) |
-1.8 |
|
|
Protein |
26.4b (101.6) |
25.2b (97.1) |
0.3 |
|
Total |
||||
|
|
Energy |
1,590 (85.7) |
1,457 (76.0) |
1.7b |
|
|
Protein |
55.6 (216.3) |
61.8 (238.4) |
-1.2 |
b. p <.05 for figures in the same row, t test.
Nutritional status of the children
The proportions of stunting (height-for-age), underweight (weight-for-age), and wasting or acute malnutrition (weight-for-height) are shown in fig. 1. The prevalence of chronic malnutrition indicated by stunting among children in the feeding programme (24%) was not significantly different from that of the children without a feeding programme (25%). This was also the case for the proportion of underweight children with a feeding programme (22%) compared with children without a feeding programme (18%). Nevertheless, significantly more children in the feeding programme (9%) showed signs of acute malnutrition (i.e., wasting) than did those not in the programme (2%; p <.05). In both groups, there was a negative and significant relationship between age and nutritional status as shown by the level of wasting (weight-for-height) (p <.05 for those with a feeding programme and p <.005 for those without a feeding programme). There was also a negative and significant relationship between age and underweight (weight-for-age) in both groups of children (p <.05 for children with a feeding programme and p <.005 for children without a feeding programme). For children without a feeding programme, there was a significant and positive relationship between age and stunting (height-for-age) (p <.001).
* p < 0.05, chi-square test, for the comparison between schools.
General characteristics of the study population
The percentage of economically inactive persons in the study area has not changed much since the 1989 survey of the Central Bureau of Statistics. Fifty-three percent were economically inactive during the study, as compared with 56% in 1989 [20]. These results indicate a high dependency ratio of dependents to wage earners: 1.2 to 1 in feeding-programme households and 1.0 to 1.0 in households not in the programme). This is characteristic of many communities in developing countries.
The significantly higher mean age of the parents in the programme group than in the no-programme group influenced the educational status of the study population. The findings on parental level of education are similar to those from other parts of Kenya [20], which show that younger persons are, on the average, more educated than older ones. This explains the higher average level of schooling of the younger parents of children in the feeding programme.
The higher literacy rate of adult males compared with females was expected [20]. However, the overall educational status of the community (based on the number of years spent in school) was very low, due to a high school dropout rate that has been attributed to the growing of miraa [20]. This would explain the low ranking of permanent employment (second to last) as a source of family income in both groups of households.
Food sources and dietary intakes
The similarity between the two groups in household caloric and protein intake and in the contribution of different meals to their total nutrient intake is not unexpected, since the community agricultural activities, family characteristics, sources of income, and expenditures on food are similar, and they share the same sociocultural characteristics. The dependence on purchased food rather than on home-produced food is mainly due to the growing of miraa, which is difficult to intercrop, as well as to the shortage of rain and water in the region. This has made the area heavily dependent on market food supplies from outside the district [18]. It is possible that the slightly higher household per capita caloric and protein intake placed the no-feeding-programme children at an advantage as compared with the programme group, whose school lunch programme was irregular.
The calories (less than 10% of the RDA) derived from breakfast, equivalent to four slices of bread for both groups, are too few to enable the children to perform adequately until lunch. This suggests that the children were hungry for a great part of the morning. The proportionately high contribution of lunch and supper to daily nutrient intake implies that more emphasis was placed on these two meals than on breakfast.
Lunch was a more important source of calories than supper for the children in the school feeding programme. The low caloric intake observed in both groups is of concern. Even with the introduction of the lunch programme, the caloric intake did not improve much, as indicated by the study data: 86% of the RDA for the feeding-programme group and 76% for the no-feeding-programme group. Others have observed that schoolchildren in Kirinyaga and Embu Districts in Kenya, similar ecological zones, have lower intakes of calories [2] and protein [15] than the RDA.
Although the total protein intake was apparently adequate in both the programme and the no-programme groups, the proteins were mainly of plant origin with low nutritional value. Food and Agriculture Organization data show that the in vivo biological value (BV) of eight proteins of plant origin (cereals and legumes) is 61.2 ± 2.65 as compared with 81.6 ± 7.05 for five animal proteins [26].
The low intake of animal protein (less than 10% for both groups) is typical for schoolchildren in Embu, Kenya [15], where children obtain more than 90% of their protein from plant sources. Although in theory, combinations of proteins from different plants in the right proportions can give amino acid profiles comparable to those of animal proteins [27], this is not the case when a high proportion of the protein comes from cereal. The community should be encouraged to use the income from miraa to keep small animals, in order to improve the amino acid balance. In addition, increased consumption of legumes should be promoted. In advising the community, it should be pointed out that the diet should be adequate in energy, since energy spares protein [28].
Nutritional status of the study children
The level of stunting of the children in both groups implies a similar past nutritional experience. The stunting levels observed were slightly higher than the 22% reported by the Central Bureau of Statistics [29] in 1989 for children under five years. Similar results were obtained in Samburu District [30], where stunting increased with age. Other studies have reported practically the same levels in similar ecological zones in Kenya for schoolchildren. A Central Bureau of Statistics survey [6] showed that 24% of the schoolchildren in Kitui District and 26% in Kwale District were stunted.
The failure of the school feeding programme to achieve its objective of improving nutritional status is, however, evident from the observation that wasting levels were significantly higher among children with a school feeding programme than among children without a programme. In fact, the level of wasting in the school with a feeding programme (9%) was more than three times the national level for children under five years of age as reported by the Central Bureau of Statistics [29]. A possible explanation for this could be the irregularity of the school lunch programme, coupled with lower dietary intake at home for children with a school feeding programme.
If the children had obtained all of the nutrients that were intended to be supplied by the lunch, their nutritional status would have been better. Clearly, the existence of the feeding programme as implemented did not confer better protection on the participating children, as shown by the prevalence of wasting. The observations, however, do not justify phasing out the school feeding programme, but rather improving its nutrient quality and regularity of supply. In Kirinyaga District [15] the nutritional status of children participating in a school feeding programme under the National School Feeding Council of Kenya improved when these problems were addressed.
The observation that the nutritional status of girls was generally better than that of boys is consistent with results of studies on schoolchildren in other parts of Kenya, including Samburu [30], Kitui, and Kwale [6]. Similar findings were reported in Tanzania [8]. This maybe due to the fact that customarily among the Meru people, girls are socially more actively involved in food preparation and would, therefore, have greater access to food.
The higher proportion of children in the school feeding programme who had a present and past history of malnutrition is not surprising, considering that the prevalence of wasting among children in the feeding programme was significantly higher than among children not in the programme. Moreover, the children in the programme were slightly more underweight than those not in the programme.
The negative relationship between age and wasting in the two groups is consistent with the higher proportional nutrient demand for growth and development in younger children. The positive increase in stunting with age was also found in another study on schoolchildren in Samburu District, Kenya [30].
The school feeding programme did not improve children’s school attendance or nutritional status. The hypothesis that children in the programme school would have better nutritional status and school attendance was not confirmed. The actual implementation of the lunch programme fell far short of its goal. This is a disappointing outcome. There are two obvious recommendations. One is to determine how the supplementary feeding programme can be improved. The other is to determine whether a better-implemented programme would have given a better result.
Acknowledgements
This study was undertaken with assistance from the Agency for Technical Cooperation of the Federal Republic of Germany (GTZ) and the University of Nairobi.
1. GOK/UNICEF. Children and women in Kenya-a situational analysis. A publication of the Government of Kenya and UNICEF, Kenya Country Office. Nairobi: Regal Press, 1992:93-103.
2. Pieters JJL, De Moel JPC, Van Steenbergen O, Van Der Hoeven WJM. Effects of school feeding on growth of children in Kirinyaga District, Kenya. East Afr Med J 1977;54:624-9.
3. Pollit E. Malnutrition and infection in the classroom. Paris: UNESCO, 1990:92-131.
4. Akinadewo O. The importance of health education on child feeding and development. Official Report of the First International All Africa Conference on Health Education. Lagos, Nigeria: Federal Health Education Division, 1981;406-11.
5. Levinger B. School feeding programmes in developing countries: an analysis of actual and potential impact. Washington, DC: Office of Evaluation, Bureau for Food and Voluntary Aid, Agency for International Development (AID), 1986:15-24.
6. Central Bureau of Statistics. Report on height monitoring in primary schools (Kwale and Kitui Districts). Nairobi, Kenya: Ministry of Planning and National Development, 1991:22-5.
7. Bellin F, Kogi-Makau W, Muroki NM. Nutritional status of the Samburu: baseline survey for the Samburu District Development Programme (SDDP). Hamburg, Germany: Gesellschaft für Agrarprojeckten, 1993:81-93.
8. Kimati VP, Scrimshaw NS. The nutritional status of Tanzanian children: a cross-sectional anthropometric survey report. East Afr Med J 1985;62:105-17.
9. Babu SC, Hallam JA. Socioeconomic impacts of school feeding programmes: empirical evidence from a South Indian village. Indian Food Policy 1989;14(1):58-66.
10. Roy P, Rath NR. School lunch in Orissa. New Delhi, India: New Delhi Council for Social Development, 1970. Cited by: Levinger B. School feeding programmes in developing countries: an analysis of actual and potential impact. Washington, DC: Office of Evaluation, Bureau for Food and Voluntary Aid, Agency for International Development (AID), 1986:15-24.
11. Cotten J. Evaluation research on the Plan 480 Title II School feeding programme in Haiti. Port-au-Prince, Haiti: USAID Haiti, 1982. Cited by: Levinger B. School feeding programmes in developing countries: an analysis of actual and potential impact. Washington, DC: Office of Evaluation, Bureau for Food and Voluntary Aid, Agency for International Development (AID), 1986:15-24.
12. Lasswell AB, Roe DA, Hochheiser L. Nutrition for the family and primary care practitioners. Philadelphia, Pa, USA: George F. Stickley. 1986:69-115.
13. Payne P. Appropriate indicators for project design and evaluation. in: Food aid and the well-being of children in the developing world. New York: UNICEF, 1986: 109-41.
14. CARE. School Feeding in Karnataka, India: impact on enrolment and attendance. Washington, DC: CARE, 1977:34-5.
15. Sigman M, Neumann C, Jansen AAJ, Bwibo N. Cognitive abilities of Kenyan children in relation to nutrition, family characteristics, and education. Los Angeles, Calif, USA: Society for Research in Child Development, 1989;60:1463-74.
16. World Food Programme cites woes. Nairobi, Kenya: Nation Newspaper Publication, 1995; 22 September:22.
17. World Health Organization/Food and Agriculture Organization. Kenya country position paper on nutrition. International Conference on Nutrition. Rome: WHO/FAO, 1992:55-86.
18. MoALM. Annual report. Nairobi, Kenya: Ministry of Agriculture and Livestock Marketing, Laare Divisional Office, 1995.
19. Fisher AA, Laing JE, Townsend JW. Handbook for family planning operations research and design. New York: Operations Research Population Council, 1991:43-6.
20. Central Bureau of Statistics. Kenya population census, Nairobi: Ministry of Planning and National Development, 1989.
21. Cameron ME, Van Staveren WA. Manual on methodology of food consumption studies. New York: Oxford Medical Publications, Oxford University Press, 1988: 13-31.
22. United Nations national household survey capability programme manual. How to weigh and measure children: assessing the nutritional status of young children in household surveys. New York: United Nations Department of Technical Cooperation for Development and Statistical Office, 1986:3-12.
23. West CE, Pepping F, Scholte I, Jansen W, Albers HFF. ECSA/CTA food composition tables. Wageningen, Netherlands: Wageningen University of Agriculture, 1987.
24. Sehmi JK. National food composition tables for planning successful diets in Kenya. Nairobi: Public Health Laboratories, Ministry of Health, 1994.
25. Food and Agriculture Organization/World Health Organization/United Nations University. Report of joint expert consultation on energy and protein requirements. Technical Report Services No. 724. Geneva: WHO, 1985.
26. Food and Agriculture Organization. Amino acid contents of foods and biological value of proteins. Nutritional Studies No. 24. Rome: Food and Agriculture Organization, 1970.
27. Muroki NM. Improvement of the nutritional value and organoleptic and keeping quality of unfermented and fermented rice “uji” (porridge) with haricot bean (Phaseolus vulgaris) tempe. Research Report. Bogor, Indonesia: Nutrition Research and Development Centre, 1990.
28. Passmore R, Eastwood MA. Human nutrition and dietetics. Hong Kong: Longman Group, 1986:279-91.
29. Central Bureau of Statistics. Fourth rural child nutrition survey. Nairobi: Ministry of Planning and National Development, 1994.
30. Kielmann AA, Unit of Applied Human Nutrition. Assessment of the nutritional impact of the Wamba Food Security Programme. Nairobi: Unit of Applied Human Nutrition, Department of Food Technology and Nutrition, University of Nairobi, 1988:31-6.
