Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Food politics wikipedia , lookup
Stunted growth wikipedia , lookup
Probiotics in children wikipedia , lookup
Saturated fat and cardiovascular disease wikipedia , lookup
Gastric bypass surgery wikipedia , lookup
Malnutrition in South Africa wikipedia , lookup
Vitamin D deficiency wikipedia , lookup
European Journal of Clinical Nutrition (1999) 53, 50±59 ß 1999 Stockton Press. All rights reserved 0954±3007/99 $12.00 http://www.stockton-press.co.uk/ejcn Food consumption of young stunted and non-stunted children in rural Zambia JLA Hautvast*1, LJM van der Heijden2, AK Luneta3, WA van Staveren2, JJM Tolboom1 and SM van Gastel2 1 Department of Paediatrics, University Hospital Nijmegen St Radboud, PO Box 9101, 6500 HB Nijmegen, the Netherlands; Division of Human Nutrition and Epidemiology, Wageningen Agricultural University, PO Box 8129, 6700 EV Wageningen, the Netherlands; and 3National Food and Nutrition Commission, PO Box 32669, Lusaka, Zambia 2 Objective: The aim of this study was to assess the quality of diet of rural Zambian pre-school children, and to compare the dietary intake of stunted and non-stunted children. Design: Cross-sectional study, in which dietary intake was assessed with a 24-h recall method. Height and weight were measured according to standard procedures. Setting: Twelve villages in Samfya District, Zambia. Subjects: Children aged 6 ± 9 months (`infants') and 14 ± 20 months (`toddlers'), attending Mother-and-Child Health Clinics, were eligible for study. Excluded were: 12 wasted, and 18 for other reasons. In total 106 infants and 99 toddlers were included. Results: In infants and toddlers, total daily intake of energy, calcium, iron, and vitamin A was insuf®cient compared to recommended daily intakes. Only infants had insuf®cient protein intake. Compared to intake from weaning foods, breast milk was the main source of energy and most nutrients for infants. For toddlers, weaning foods were more important. Stunted infants and toddlers tended to have lower intakes of energy compared to non-stunted age-mates. Daily energy intake per kg bodyweight showed no difference between stunted and nonstunted children. Conclusions: Overall quality of weaning foods was inadequate. Stunted infants and toddlers showed a tendency of lower energy intakes compared to non-stunted age-mates. Sponsorship: This study was supported by the Jan Dekker en Dr. Ludgardine Bouwmanstichting, and by Nutricia Overseas Division, Zoetermeer, the Netherlands. Descriptors: energy intake; diet; height; preschool children; Zambia Introduction In rural Africa the prevalence of linear growth retardation, also called stunting, among children is generally high (de Onis et al, 1993). Environmental factors, such as infections, inadequate food intake, quality of care, and seasonal in¯uences contribute to the development of stunting in a child (Waterlow, 1992). This may lead to serious functional complications, such as lower mental development, reduced work capacity in adulthood and increased obstetric risk (GranthamMcGregor, 1995; Haas et al, 1996; Martorell et al, 1992). In Zambia, a less developed country in south-central Africa, national ®gures show 40% stunting in children under the age of ®ve years (Gaisie et al, 1993). A study in Samfya District, a rural area in the north-east of the country, showed an increasing prevalence of stunting with age: 11% from birth to three months of age to 55% from six to twelve months (Hautvast et al, in press). Because linear growth retardation in this area occurs in children in an ageperiod when breast milk intake declines and supplementary *Correspondence: Ms JLA Hautvast, MD, Department of Paediatrics, University Hospital Nijmegen St Radboud, PO Box 9101, 6500 HB Nijmegen, The Netherlands. E-mail: [email protected] Received 16 March 1998; revised 25 July 1998; accepted 18 August 1998 foods are introduced in the child's diet, the relationship of growth retardation with energy and nutrient intake warrants further investigation. Generally, studies on the effect of intake of energy or one selected nutrient on linear growth retardation have not shown consistent results (Allen, 1994). Several studies have shown a positive effect of energy or protein supplementation on linear growth (Gopalan et al, 1973; Mora et al, 1981; Walker et al, 1991; Schroeder et al, 1995; Neumann & Harrison, 1994). However, since another study suggested that linear growth retardation may occur even when intakes of energy and protein are adequate (Beaton et al, 1992), it is questioned to which extent micronutrient de®ciencies impair linear growth of human populations (Allen, 1994). Studies examining the role of single micronutrients on improvement of linear growth show con¯icting results. This might be explained by the fact that not one single micronutrient is responsible for the development of linear growth retardation, but that poor quality diets supply insuf®cient amounts of multiple micronutrients (Allen, 1994). With regard to rural Africa in particular, accurate data on food consumption are scarce and detailed information is required to examine the relation with linear growth. The aim of our study was to describe the quality of diet of rural Zambian preschool children and to compare the diet of stunted and non-stunted children. Food Consumption of Zambian Children JLA Hautvast et al Methods Study area The study was conducted in Samfya District, a rural area in the north-east of Zambia, which covers an area of 10 329 km2 with 107 486 inhabitants (Central Statistical Of®ce, 1990). A large part of the district is situated adjacent Lake Bangweulu. This lake borders a swamp area in the south-east, whereas the mainland of the district, with woodland vegetation and marshy ¯ood plains, is located on the western side. The main sources of food and income for the majority of the people are (subsistence-) farming and ®shing (Gould, 1989). Quality of health care facilities is poor, illustrated for instance by an infant mortality rate of 148 per 1000 births in Luapula Province (Gaisie et al, 1993), in which Samfya District is situated. Maize, cassava, sweet potatoes, pumpkins, groundnuts, beans and leafy vegetables are common agricultural products (Moore & Vaughan, 1994). The staple dish in Zambia is `nshima', a thick porridge made of water and ¯our (mostly from maize or cassava). It is supplemented with vegetables, ®sh or meat. Since laboratory facilities, which were needed for our main study, were located in the district headquarters, only villages situated within a 2-hour drive from this location were included. Thirty-seven health posts (villages with a Community Health Worker, but otherwise without health care facilities) are situated on the mainland of the district, within a two-hour drive from the district headquarters. From these 37 villages, 12 were selected at random for recruitment of children in our study. Design and subjects In this study a selected group of children of 6 ± 9 months of age (infants in this study) and 14 ± 20 months of age (toddlers in this study) were examined in the period July ± September 1994. The study team consisted of the principal investigator (JLAH), two quali®ed staff from the Ministry of Health, and a trained research assistant. All children of described age groups attending the mobile Mother-and-Child Health (MCH) clinics in selected villages during the study period were eligible for participation in the study. Age of the children was either copied from their clinic card (after verbal veri®cation) or estimated by comparison with age-mates with known date of birth. For practical and ethical reasons children who were not accompanied by their mother or caretaker were excluded. HIVinfection is known to be highly prevalent in Zambia (Fylkesnes et al, 1997), and may therefore have acted as a confounder in our study on linear growth retardation. Since an important symptom of established HIV-infection (AIDS) is severe weight loss, we excluded children who were wasted (weight-for-height below or equal to 72 s.d. of the median of the National Center of Health Statistics (NCHS) reference population) (Hamill et al, 1979). All communication with the mothers of the children was carried out by the local staff of the study team, who were ¯uent in the local language, Cibemba. After having received explanations about the objectives of the study by a member of the study team, mothers were asked to give their informed consent for participation of their child in the study. Out of the initially recorded group of 235 eligible children, 12 were wasted, 13 were excluded for other reasons (mother or caretaker absent (5), or refusing to participate (7), or other reason (1)), and 5 children were absent during the interview on food consumption. Finally, a total of 205 children was included in this study, from whom anthropometric measurements were taken, and a 24-h recall questionnaire was administered. Anthropometry Anthropometric measurements were performed according to standard procedures (WHO, 1983). Weight of the children, wearing light clothes only, was measured to the nearest 0.1 kg, using a portable hanging weighing scale (Salter, England, model 235 6S, 0 ± 25 kg), which was calibrated daily. Recumbent length of the children was measured to the nearest 0.1 cm, using a horizontal measuring board with sliding foot piece. Length was measured with children being barefoot and after removal of any headgear, and was done by two members of the study team cooperatively. Weight of the mothers was measured using bathroom scales with 0.5-kg accuracy. Standing height was measured to the nearest 0.1 cm using a lengthmeasuring rod with sliding head bar ®xed against a wall. Members of the study team were intensively trained in order to minimize inter-observer variation in measurements. For all children height-for-age- (HAZ), weight-for-age(WAZ), and weight-for-height-Z-scores (WHZ) were calculated using the NCHS reference (Hamill et al, 1979). Children were classi®ed into two groups according to their height-for-age Z-score: stunted (HAZ472) and nonstunted (HAZ > 72) (WHO Working Group, 1986). Mothers of children with a Body Mass Index (BMI) < 18.5 kg=m2, were considered as having chronic energy de®ciency (James et al, 1988). Food consumption 24-h recall questionnaire: A 24-h recall questionnaire, adapted for use in the local setting, was used to assess the food consumption of the children under study (Cameron & van Staveren, 1988). Consultation of local key-informants demonstrated that, in general, children start eating from a shared dish only after the age of 2 ± 3 y; before that age they are served from a separate dish. Furthermore the traditional feeding pattern appeared to be uniform throughout the week. Mothers were asked to recall all foods and beverages consumed by their child during the previous 24 hours (beginning from morning meal on the day before up to the morning meal on the day of the interview). They were then asked to both estimate the quantities of the mentioned foods, as well as to state the ingredients used to prepare the dishes and to estimate their quantities. To facilitate the quantity estimates, standardized household measures (18 cups, plates and bowls in total), marked with several levels to indicate different contents, were provided by the interviewer. Frequency and size of separate foods were recorded. Raw ingredients were converted from household measures to weight estimates by taking reference weights, calculated on the basis of the average of triple weights on a digital weighing-scale. To estimate the weight of the cooked dishes, standard recipes were cooked and weighed, according to reported quantities. For the calculation of the ratios of the raw ingredients consumed by each child, the 51 Food Consumption of Zambian Children JLA Hautvast et al 52 weights of the standard recipe most similar to the reported dish were used. This questionnaire was pre-tested in comparison with a weighed observed record (WOR) in a group of 15 children (aged 14 ± 17 months). For both methods the intake of energy, protein, fat, carbohydrate, calcium, iron, vitamin A and vitamin C were calculated and compared. The group was rather small, therefore no statistically signi®cant differences (a 0.05) were found between the mean intakes from the 24-h recall and the WOR (paired t-test). However, when the differences between the methods were expressed as a percentage of the WOR, energy, protein, fat, iron and vitamin A tended to be overestimated with the 24-h recall in comparison with WOR, whereas intake of carbohydrate, calcium and vitamin C were relatively underestimated. All differences between intakes were within 20%, except for fat intake, which showed 73% overestimation with the 24-h recall, in comparison with the WOR. This was mainly a result of reported high intakes of groundnuts. Since consumption of groundnuts for young children is promoted by local nutrition education programmes, mothers or caretakers may have tended to give socially desirable answers. In contrast, consumption of food items with low status, such as cassava, were under-reported. The sample size of our pre-test was too small to determine reliability of the recall method. But we estimated within-person variation of the 24-h recall during the pre-test; this was 1394 kJ for the energy intake for the repeated 24-h recall. Two members of the study team were trained to administer this questionnaire. Comparison of the results on intake of energy and nutrients (as listed above) from both interviewers, as were derived from the overall study, revealed no signi®cant differences (Student's t-test, a 0.05). Breastmilk volume and composition: Some of the mothers of the children studied were living at rather a large distance from the study site in each village (up to 1 ± 2 h walking). Since the design of our study was not homebased, we decided not to include collection of breast milk data. We therefore estimated the assumed quantity and composition of intake from breast milk for those children who were reported to breast-feed, from the literature. Since breast-feeding in Zambia is done on demand and since cessation of breast-feeding is a gradual process, we did not attempt to differentiate between day- and night-feeding vs night-feeding only. Data from a study in neighbouring Zaire (presently: Demographic Republic of Congo) were used to assess breast milk volume: for the children who were 6 ± 9 months old, breast milk intake in 24 hours was assumed to be 553 ml and for the children who were 14 ± 20 months old, 509 ml (Hennart & Vis, 1980). Values for composition of breast milk were derived from a study in the Gambia (West Africa) (Prentice et al, 1986). If no information on a speci®c nutrient was available, values from George and De Francesca (1989) were used, unless other studies had shown a clear relation between nutrient content in breast milk and maternal nutritional or demographical characteristics. Breast milk in our study was therefore assumed to contain per 100 ml 313 kJ (calculated by JLAH), 4.2 g fat, 1.3 g protein, 7.7 g lactose, 3.4 mg vitamin C (Prentice et al, 1986), 38 RE vitamin A (Newman, 1994), 21 mg calcium (Steenbergen van et al, 1981), and 0.05 mg iron (George & De Francesca, 1989). Calculation of energy and nutrients The energy, protein, fat, carbohydrate, calcium, iron, vitamin A and vitamin C content of the weaning foods consumed by the children was calculated. These nutrients were selected because literature has shown that they are known to be possibly involved in the linear growth process in children (Allen, 1994; Waterlow, 1994). Furthermore they were selected because reliable values of these nutrients in food items, as consumed by the children studied, were available in the literature. Nutrients were calculated using a Nutrient Calculation program (KOMEET 2.0, Division of Human Nutrition and Epidemiology, Wageningen Agricultural University, the Netherlands) and a specially developed food composition table for Samfya District. The food items included in the food composition table were based on interviews with three informants and on the food times eaten by the children under study. The nutrient contents of the different food items were based on literature values (West et al, 1988; Leung et al, 1986; Platt, 1962; National Food and Nutrition Commission, 1987). To assess the total weight of the diet of the children, the weight of all food items consumed was calculated. To estimate the total energy and nutrient intake of all children, assumed values of breast milk volume and composition were added to the values of energy and nutrient intake of the children. Food groups: To get insight in the dietary sources of the nutrient intake the different food items were assigned to one of the following food groups: (1) ®sh; (2) cereals (maize, millet, rice); (3) meat & dairy; (4) roots & tubers (cassava, (sweet) potatoes); (5) legumes & nuts; (6) fruits & vegetables; and (7) a rest group (this group consisted of food items such as water, sugar, oil, lemonade and biscuits). Recommended daily intakes: Recommended daily intakes (RDIs) for both age groups were calculated for energy, protein, calcium, iron, vitamin A and vitamin C intake using FAO=WHO recommendations (FAO=WHO, 1962; FAO=WHO, 1970; FAO=WHO=UNU, 1985; FAO=WHO, 1988). If the original reference provided ®gures of requirements per kg bodyweight, median weights of the NCHS reference population were used to calculate the recommended daily intake. For all nutrients, except energy, safe level of intake was calculated. For energy intake average requirement was calculated. Calculation of the safe level of dietary protein intake was adjusted for protein digestibility (assumed to be 85%) and amino acid composition (assumed to be 67%) (FAO=WHO=UNU, 1985). Table 1 presents the calculated RDIs for infants and toddlers. Ethics Ethical approval for this study was obtained from the Ministry of Health and the National Food and Nutrition Commission, Lusaka, Zambia. Data analysis The software package Epi Info, version 5.01b (Dean et al, 1990) was used to calculate HAZ, WAZ, and WHZ. SPSS for Windows, version 6.1.4 (SPSS Inc., Chicago IL) was used for further statistical analyses. Because the data on energy and nutrient intake were skewed, the median and 25th and 75th percentiles were Food Consumption of Zambian Children JLA Hautvast et al Table 1 Recommended Daily Intakes (RDI) of energy and nutrient intake for children aged 6 ± 9 months and 14 ± 20 monthsa Age RDI for intake of: 6 ± 9 months Energy (kJ) Protein (g)b Calcium (mg) Iron (mg) Vitamin A (RE) Vitamin C (mg) 3200 23 500 ± 600 21 350 20 14 ± 20 months 4850 23 400 ± 500 12 400 20 Source FAO=WHO=UNU, 1985 FAO=WHO=UNU, 1985 FAO=WHO, 1962 FAO=WHO, 1988 FAO=WHO, 1988 FAO=WHO, 1970 a For all nutrients except energy, safe level of intake is presented. For energy, average requirement is presented. In the situation that the original source provided ®gures of requirements per kg bodyweight, median weights of the NCHS reference population were used to calculate the recommended daily intake. b Safe level of dietary protein intake was adjusted for protein digestibility (assumed to be 85%) and amino-acid composition (assumed to be 67%). calculated. Mann-Whitney U-test was used for comparison of energy and nutrient intake between stunted and nonstunted children, and for comparison of proportional contribution of speci®c food groups between stunted and nonstunted children. For the analysis of the contribution of selected food groups to energy and nutrient intake, breast milk was also considered as a food group. Results Table 2 shows baseline demographic and anthropometric characteristics of the study population for the two groups: infants and toddlers. Mean HAZ, WAZ and WHZ scores of the infants were signi®cantly higher than those of toddlers. The proportion of mothers with chronic energy de®ciency (BMI < 18.5 kg=m2) was signi®cantly lower in the infant group (14%) compared to the toddler group (28%) (P 0.02). Tables 3 and 4 describe the dietary intake of the selected infants and toddlers. Table 3 shows the estimated median values of daily energy and nutrient intake derived from weaning foods only and from weaning foods and the assumed intake of breast milk together. All infants were still breast-feeding, whereas nine of the toddlers were not. Table 2 Baseline demographic and anthropometric characteristics of a group of infants and toddlers, and anthropometric characteristics of their mothers, Samfya district (mean (s.d.)) Infants n Sex male (%) female (%) Age (months) minimum maximum Height (cm) Weight (kg) HAZ WAZ WHZ % mothers with BMI < 18.5kg=m2 (%) a 106 47 53 7.3 (0.8) 6.0 8.9 64.4 (2.9)b 7.0 (0.9)b 71.71 (0.98)b 71.27 (0.88)b 0.13 (0.74)b 14a Toddlers 99 60 40 16.5 (1.7) 13.5 20.0 72.9 (3.5) 8.5 (1.2) 72.37 (1.09) 71.99 (1.02) 70.78 (0.78) 28 Statistically signi®cant difference between infants and toddlers, P < 0.05. Statistically signi®cant difference between infants and toddlers, P < 0.001. HAZ Height-for-age Z-score; WAZ Weight-for-age Z-score; WHZ Weight-for-height Z-score; BMI Body Mass Index. b Three infants had only consumed water in addition to breast milk during the period covered by the 24-h recall questionnaire, and one infant had only consumed breast milk. The mean number of meals consumed, including snacks, was 2.6 (range 0 ± 5) and 3.0 (range 1 ± 6) for infants and toddlers, respectively. Intake of energy and nutrients from weaning foods only was higher in toddlers compared to infants. In both groups breast milk provided a substantial part of the energy and nutrient intake. In the group of infants breast milk adds substantially to the intake of fat, vitamin A, and vitamin C. In the group of toddlers intake of fat, vitamin A, and vitamin C was also increased by breast milk consumption, but relatively less than in the infant group. As expected, iron intake was hardly increased by intake of breast milk in both groups. In infants as well as toddlers, the median energy intake from weaning foods was mainly derived from carbohydrates (70% and 69%, respectively). Fat contributed 12% and 13%, protein 11% and 15% to the energy intake of weaning foods in infants and toddlers, respectively. Inclusion of breast milk in the diet of infants and toddlers showed a considerable increase in energy percentage derived from fat, with a decrease in energy percentage derived from carbohydrates. Comparison of the total daily intake of both weaning foods and breast milk of infants with recommended daily intakes for that age group, demonstrates that intake of energy, protein, calcium, iron and vitamin A was too low. Median energy, protein and vitamin A ranged from 57 ± 80% of RDI. Median calcium intake was 26% of RDI and median iron intake 10% of RDI. Intake of vitamin C in infants was adequate. For the toddler group total intake of energy, calcium, iron and vitamin A was too low compared to the recommended daily intake. Median energy and vitamin A intake were both 70% of RDI. Median calcium and iron intake were 40% and 42% of RDI respectively. Intake of protein and vitamin C in toddlers were adequate. Table 4 shows the proportional contribution of different food groups to the daily energy and nutrient intake, for infants as well as toddlers. In infants the main source of intake of energy and all nutrients, except iron, was breast milk, as was estimated from literature. In toddlers the assumed intake of breast milk contributed to > 50% of the intake of fat, calcium, vitamin A, and vitamin C. For both infants and toddlers important sources of energy intake from weaning foods were cereals, and roots and tubers. For the infants, important sources of protein from weaning foods were ®sh and cereals, while cereals and 53 Food Consumption of Zambian Children JLA Hautvast et al 54 Table 3 Estimated daily intake of energy and nutrients at baseline of weaning foods only, as assessed by the 24-h recall, and of weaning foods and the assumed intake of breast milk (from literature), of a group of infants and toddlers, Samfya district; median (25th ± 75th percentile) Infants (n 106) Intake of: Total weight food (g) Energy (kJ) (kcal) Protein (g) energy % Fat (g) energy % Carbohydrates (g) energy % Calcium (mg) Iron (mg) Vitamin A (RE) Vitamin C (mg) Toddlers (n 99) WF WF BM WF WF BM 256.(153 ± 441) 829.(438 ± 1636) (197) 6.(2 ± 13) 11.(8 ± 19) 3.(1 ± 6) 12.(8 ± 22) 32.(19 ± 63) 70.(59 ± 83) 28.(13 ± 62) 1.7.(1.0 ± 3.2) 30.(5 ± 82) 2.(0 ± 8) 809.(706 ± 994) 2560.(2169 ± 3367) (610) 13.(9 ± 20) 9.(7 ± 13) 26.(24 ± 29) 35.(29 ± 40) 74.(61 ± 106) 55.(48 ± 61) 144.(129 ± 178) 2.0.(1.3 ± 3.5) 240.(215 ± 292) 20.(19 ± 26) 499.(355 ± 690) 1804.(1317 ± 2740) (430) 18.(11 ± 30) 15.(10 ± 22) 6.(3 ± 13) 13.(9 ± 19) 75.(43 ± 108) 69.(56 ± 77) 80.(49 ± 185) 4.3.(2.9 ± 7.1) 87.(46 ± 181) 10.(2 ± 38) 964.(834 ± 1189) 3397.(2910 ± 4333) (809) 24.(16 ± 37) 13.(9 ± 18) 26.(24 ± 33) 26.(20 ± 31) 114.(79 ± 142) 61.(53 ± 68) 182.(146 ± 289) 4.6.(3.1 ± 7.3) 278.(228 ± 353) 26.(18 ± 54) WF weaning foods; BM breast milk. Table 4 The proportional contribution of different food groups (assessed with the 24-h recall) and breast milk (estimated from literature) to the intake of energy and nutrients at baseline of a group of infants and toddlers, Samfya district (%)a Food groups Fish Cereals Meat & dairy Roots & tubers Legumes & nuts Fruits & vegetables Restb Breast milk Infants (n 106) Total weight food Energy Protein Fat Carbohydrates Calcium Iron Vitamin A Vitamin C 2 3 19 2 0 7 12 7 0 3 12 11 3 19 2 25 1 0 1 1 3 1 0 1 3 2 0 2 9 2 0 17 7 26 2 5 1 3 6 6 1 1 6 0 0 2 1 1 0 2 4 5 6 12 23 4 2 3 5 1 4 0 0 67 66 56 84 56 76 20 82 82 Toddlers (n 99) Total weight food Energy Protein Fat Carbohydrates Calcium Iron Vitamin A Vitamin C 4 6 33 6 0 15 16 14 0 6 21 17 7 32 5 31 2 0 1 1 5 2 0 2 3 2 0 4 13 3 1 22 12 23 3 16 1 5 7 11 1 2 6 1 0 4 3 3 0 4 11 10 17 28 36 5 3 5 6 3 5 1 0 45 47 29 67 34 51 7 61 56 a b Figures in one row do not always total up to 100%, due to rounding off. Rest includes amongst others water, sugar, oil, lemonade, biscuits. roots and tubers were important sources of carbohydrates and fruit and vegetables major sources of vitamin C. Infants mainly derived their iron from cereals, and roots and tubers, while breast milk provided a third source of iron. For toddlers, in addition to breast milk, important sources of fat were legumes and nuts, of calcium, ®sh, of vitamin A, ®sh and fruits and vegetables, and of vitamin C, roots and tubers and fruit and vegetables. The intake of protein in the toddler group was mainly derived from ®sh, with breast milk as a second source; carbohydrates were equally derived from cereals and breast milk. The most important sources of iron for the toddlers were cereals and roots and tubers, while breast milk contributed less. Figures 1a, 1b, and 1c show the proportional contributions of speci®c food groups to the intake of energy, protein and iron in infants and toddlers, respectively. Figures 1a and 1b clearly demonstrate that in infants of 6 ± 9 months of age, the assumed intake of breast milk is the main food and the most important source of energy and protein. For toddlers the assumed breast milk intake is clearly still an important contributor to energy and protein intake. In both infants and toddlers, cereals and roots and tubers are important contributors of energy from weaning foods. Fish and cereals are important contributors of protein from weaning foods in both groups. Cereals and roots and tubers are important sources of iron from weaning foods. To be able to compare the dietary intake of stunted and non-stunted children, Table 5 presents median energy and nutrient intake from weaning food only, in infants and toddlers. For infants all comparisons were non-signi®cant (P-values were all > 0.30); for toddlers all comparisons, except vitamin C, were non-signi®cant (P-value vitamin A 0.09; all other P-values > 0.15). In toddlers intake of vitamin C from weaning foods was statistically different between stunted and non-stunted children, with higher Food Consumption of Zambian Children JLA Hautvast et al (FAO=WHO=UNU, 1985). Both in infants and toddlers, the overall results on intake of calcium, iron, vitamin A and vitamin C in stunted and non-stunted children, were not consistent with the ®nding of lower energy intake in stunted children. Total weight of foods, as consumed by infants and toddlers, showed no difference between stunted and nonstunted children (data not presented). Comparison of the proportional contribution of the speci®c food groups to the weight of the consumed weaning food between stunted and non-stunted children did not show any difference. Figure 2 shows the results of the proportional contribution of speci®c food groups from weaning foods only, to the energy intake of stunted and non-stunted infants and toddlers. No statistically signi®cant differences were observed between the groups, except for the energy intake from the rest food group of the toddlers. Discussion Figure 1 Percentage of energy (1A), protein (1B) and iron (1C) as derived from breast milk (estimated from literature) and speci®c food groups (assessed with 24-hr recall), of a group of infants and toddlers, Samfya district. Figure 2 Percentage of energy as derived from speci®c food groups (assessed with 24-hr recall), of a group of stunted (ST) and non-stunted (NST) infants and toddlers, Samfya district. values in the group of stunted children (P 0.012). In infants and toddlers, stunted children appeared to have, though not statistically signi®cantly, lower energy intake from weaning foods compared to non-stunted children. However, no difference was observed between stunted and non-stunted children in energy intake per kg bodyweight and no difference was observed in energy percentage distribution. The median total daily energy intake per kg bodyweight from weaning foods and the assumed intake of breast milk was 352 kJ and 243 kJ for stunted and nonstunted infants, respectively, and 411 kJ and 373 kJ for stunted and non-stunted toddlers, respectively. This indicates that for all groups, total dietary energy intake per kg bodyweight is insuf®cient when compared to the recommended daily energy intake per kg bodyweight, which is 400 kJ and 445 kJ for infants and toddlers respectively Our study in Samfya District (Zambia) demonstrates that daily intake of energy, protein, calcium, iron, and vitamin A from weaning foods is an insuf®cient supplement to the assumed breast milk intake in infants in order to meet the requirements needed for their body maintenance and growth. In toddlers, the diet, including the assumed intake of breast milk, gives an insuf®cient supply of energy, calcium, iron and vitamin A. We observed that stunted infants and toddlers tended to have lower energy intakes from weaning foods, though not statistically signi®cant, in comparison to non-stunted age-mates. Intake of other (micro) nutrients did not show a relation with being stunted. Children enrolled in this study were recruited through MCH clinics, and therefore sampling was dependent on attendance of mothers or caretakers and their children at these clinics. According to data from the Zambian Ministry of Health, our study recruited 75 ± 85% of the children in the speci®c age groups in each catchment area. Furthermore, since both ill and healthy children were equally absent at the MCH clinics, we assume to have covered a group of children representative for the mainland of the district. Since a previous study (Hautvast et al, in press) had shown that the feeding pattern of children living in the Bangweulu Swamps of Samfya District was less adequate in comparison with those living on the mainland, the results of this study are only valid for the mainland. Since the mainland of the district is comparable to most rural areas in Zambia with respect to its ecology, land-use and other population characteristics, the results from this study may be extrapolated to other rural areas of Zambia. Regarding the validity of our measurements, we note that a source of possible error may be the use of a 24 h recall method. With respect to this method, information bias might have occurred, since it depends on information given by the mother. Because the mothers were not accustomed to estimating quantities, this was facilitated by providing a wide variety of commonly used household measures, all marked with various levels to indicate different contents. However, as was shown in the pre-test, intake of most nutrients tended to be overestimated with the 24 h recall in comparison with the WOR; fat intake was overestimated the most. In the present study, however, the mean intake of fat was almost similar to the mean fat intake observed with the WOR during the pre-test. 55 Food Consumption of Zambian Children JLA Hautvast et al 56 Ferguson et al (1994) and Dop et al (1994) showed in their studies in Ghana and Senegal, respectively, that, the repeated 24-h recall gives results more similar to the weighed observation, in comparison with one recall only. However, they also demonstrated that although the 24 h recall is not suf®ciently accurate to assess individual food intake, it is of use in dietary surveys in groups of children. The relative validity of the 24 h recall at group level applies to macronutrients and depending on the applied data base, also to micronutrients. Since the ®rst aim of our study was to provide descriptive data on the dietary quality of the children studied, we consider our results on intake of macro- as well as micronutrients at group level, valid for description in this paper. Further, we may query our assumptions on the amount and composition of breast milk. Prentice et al (1994) showed that breast milk volume is only negatively affected by maternal body mass index in situations of (near) famine. In our study, 14 ± 28% of the mothers had chronic energy de®ciency (BMI < 18.5 kg=m2), which, though a serious situation, does not indicate presence of a famine. The observed difference in the proportion of mothers with chronic energy de®ciency between infants (14%) and toddlers (28%), may be due to a decrease in maternal weight during prolonged lactation, which was shown by Adair and Popkin (1992) to occur after 10 ± 14 months of lactation. Since our assumptions on breast milk volume were based on Zairian mothers (Hennart & Vis, 1980), whose height and weight were fairly similar to our mothers, and are genetically closely related to the ethnic group included in our study, these estimates will have been acceptably realistic. However, our assumptions on breast milk content of energy, protein, fat, lactose, and vitamin C were derived from Gambian mothers, whose nutritional status was slightly better compared to our mothers, and therefore may have been an overestimation compared to reality. Assuming that the mothers in our population were at risk for low vitamin A status, we chose to base our estimates of vitamin A content of breast milk on data from other developing countries (Newman, 1994). Therefore, if vitamin A status of our mothers was adequate, we may have underestimated the vitamin A content of breast milk. Since studies in Kenya and Zaire (Steenbergen van et al, 1981; Prentice & Barclay, 1991) demonstrated lower ®gures on calcium content of breast milk compared to developed countries, we assumed that this would also be applicable for our situation. However, if in our study conditions affecting calcium content of breast milk were different from these studies cited, our estimate of the concentration of this nutrient in breast milk may have been underestimated. Our study was performed during the end of the dry cold season and the beginning of the dry warm season, a period when food availability is substantially reduced, but not yet decreased to a minimum (Moore & Vaughan, 1994). Therefore our results on intake of weaning foods re¯ect a situation of food availability between the annual minimum and maximum. Furthermore, a study from the Gambia showed that, in relation to maternal dietary restriction, breast milk volume and nutrient content are decreased during the rainy season in comparison with the dry season (Prentice et al, 1986). Since our study was not performed during the rainy season, our estimates on breast milk volume and composition are not confounded by season. Total daily energy and protein intake of our infants and toddlers were similar to intakes of breast-fed children in Kenya and Uganda of the same age (Steenbergen et al, 1978; Rutishauser, 1974). Comparison with Kenyan toddlers (Calloway et al, 1993; Murphy et al, 1992), who were, however, not breast-fed, showed that energy, protein, vitamin A, and iron intake was in a similar range, but fat intake was higher (50%) in our toddlers, whereas carbohydrate and calcium intake was higher (24 ± 41%) in the Kenyan children. Assuming that our estimates on breast milk intake were correct, energy intake of weaning foods was 640 kJ and 1450 kJ short for infants respectively toddlers to meet the RDI for energy (Tables 1 and 3). This may ®rstly be explained by the energy density of the weaning diet, which was 3.3 kJ=g for infants and 3.7 kJ=g for toddlers. These densities are low, low ± normal respectively in comparison with the recommended energy density of 3.6 ± 4.8 kJ=g of weaning foods (Cameron & Hofvander, 1983). This was due to the high intake from the food groups cereals, and roots and tubers, which contain food products with relatively low energy values, and due to the overall low fat-energy percentage of the weaning diet. Another study in Zambia investigating weaning practices in children aged 6 ± 29 months (Hayes et al, 1994), also observed low average caloric density per feeding. Secondly, the median weight of each consumed dish was 102 and 180 g for infants and toddlers respectively: this is low when compared to the maximum volume of 110 ± 200 ml for infants and 200 ± 300 ml for 1-to-3 y olds respectively, which can be eaten at one time (Cameron & Hofvander, 1983). Associated with the overall low energy intake, de®cient intakes of other macronutrients and micronutrients may be expected, as is also observed in our population. With respect to the comparison of dietary intake between stunted and non-stunted children, we again have to consider the selection procedure. As noted before, the selection of villages for this study was con®ned to those situated on the mainland of the district, whereas villages from the swamp area were not included. However, a previous study showed similar rates of stunting in preschool children for both areas (Hautvast et al, in press). Children for this study were recruited through mobile MCH clinics, and only those who were not wasted were included. Therefore the population included in this study may have been less diverse with respect to their nutritional status. Mean HAZ-scores of the so-called non-stunted infants and toddlers were below the median of the NCHS-reference population (both groups 71.12 s.d.) (Table 5). This means that linear growth of these children was also slightly retarded, and that they may not have been the optimal control group. However, the difference in mean HAZscores between the selected study groups of stunted and non-stunted children was more than 1.5 s.d., which, if the sample size had been larger, would have been suf®ciently large to detect possible differences in energy and nutrient intakes. In the framework of a larger study, the sample size was calculated on the basis of detecting a signi®cant difference in biochemical data between stunted and non-stunted children, which required a number of at least 30 children per group. However, to detect a signi®cant difference in energy intake between stunted and non-stunted children, a sample size of more than 120 children per group would have been required (with a power of 80% and alpha of 0.05). Due to Food Consumption of Zambian Children JLA Hautvast et al 57 Table 5 Anthropometry and estimated daily energy and nutrient intake from weaning foods of a group of infants and toddlers by height-for-age group, Samfya district; median (25th ± 75th percentile) Infants Nutritional statusa WAZ HAZ Dietary intake Energy (kJ) (kcal) Energy=kg (kJ=kg) (kcal=kg) Protein energy % Fat energy % Carbohydrate energy % Calcium (mg) Iron (mg) Vitamin A (RE) Vitamin C (mg) a Toddlers Stunted (n 39) Non-stunted (n 67) Stunted (n 69) Non-stunted (n 30) 71.92 (0.7) 72.72 (0.5) 70.89 (0.8) 71.12 (0.7) 72.42 (0.8) 72.91 (0.8) 71.01 (0.8) 71.12 (0.6) 626.(378 ± 1689) (149) 96.(61 ± 235) (23) 10.(8 ± 18) 15.(8 ± 29) 68.(56 ± 83) 29.(11 ± 48) 1.7.(0.6 ± 3.5) 20.(2 ± 95) 1.(0 ± 5) 920.(462 ± 1536) (219) 112.(69 ± 218) (27) 11.(8 ± 19) 12.(8 ± 20) 71.(61 ± 84) 25.(13 ± 63) 1.8.(1.0 ± 3.1) 37.(7 ± 75) 2.(0 ± 8) 1708.(1259 ± 2702) (407) 226.(145 ± 320) (54) 16.(9 ± 23) 14.(9 ± 21) 68.(56 ± 76) 80.(52 ± 184) 4.3.(2.8 ± 6.6) 99.(66 ± 182) 13.b (2 ± 44) 2121.(1499 ± 2958) (505) 221.(157 ± 298) (53) 13.(10 ± 20) 11.(8 ± 16) 75.(60 ± 79) 82.(35 ± 198) 4.4.(3.1 ± 7.2) 60.(30 ± 169) 5.(0 ± 17) Anthropometry presented as mean (SD); b Signi®cant difference between stunted and non-stunted group; P < 0.05. the larger day-to-day variation of micronutrient intake, the group should have been even larger to detect signi®cant differences in micronutrient intake. Furthermore, if we had used a repeated 24-h recall in our study, the results on energy and nutrient intake may have had a lower coef®cient of variation, and therefore possible differences between stunted and non-stunted children would have been more distinct. We did not expect mothers of stunted children and nonstunted children to answer the 24-h recall differently. Though volume of breast milk appears to be relatively unaffected by moderate maternal malnutrition (Prentice et al, 1994), this is less clear for nutrient content of breast milk (Abrams, 1991): decreased levels of protein, and fat soluble vitamins have been observed in malnourished lactating mothers. In our study, mothers of stunted children appeared to have more frequent chronic energy de®ciency (26%) when compared to mothers of non-stunted children (15%) (statistically signi®cant). Therefore, breast milk composition of mothers of stunted children may more often have been of lower quality than of mothers of nonstunted children. Moreover, lower body mass and higher prevalence of illness in stunted children may result in weak suckling, which will subsequently reduce milk output. However, so far no studies have investigated these hypotheses, and we can therefore only speculate that these assumptions may in¯uence the quantity and quality of breast milk intake in stunted children. The observed difference in energy intake of 300 ± 400 kJ between stunted and non-stunted children, equals about 100 ± 130 ml of breast milk and represents 10% and 8% of the RDI of infants and toddlers respectively. The lower energy intake of stunted children will have resulted in reduced growth and activity levels in comparison with non-stunted children. Remarkably, despite no differences in intake of weaning foods from speci®c food groups in observed stunted and non-stunted toddlers, daily vitamin C intake from weaning foods in stunted toddlers was higher in comparison to nonstunted toddlers. We have no other explanation for this observation other than chance. In our study, energy intake per kg bodyweight from weaning foods and assumed intake of breast milk was de®cient for both stunted and non-stunted infants and toddlers. Since other studies showed lower gain in length and smaller attained length for children with lower energy intakes (Schroeder et al, 1995; Neumann & Harrison, 1994), further adequate length gain of our non-stunted, and especially our stunted children, may be threatened. Both ®ndings indicate that catch-up in height growth will most likely be impossible. In contrast to our results, an interesting study by Walker et al (1990) showed equal energy intake, though de®cient, in stunted as well as non-stunted Jamaican children of 9 ± 24 months, with higher energy intake per kg bodyweight in stunted children compared to non-stunted. However, in that study non-stunted children were de®ned as those with a height above 71 s.d. of the reference, and dietary recalls were only obtained in children who were not ill. Both aspects are different from our design which may therefore explain these different result. We conclude that despite longstanding preventive health care programmes, including nutrition education, the quality and quantity of diet in infants and toddlers of Samfya District is still insuf®cient. Within a macro-economic context, this points to structural problems at household level, all affecting intra-household food availability. In view of the low quality of weaning foods, the observed prolonged breast-feeding into the second year of life in this environment appears to be a highly recommendable practice, which should certainly be continued. The high prevalence of stunting, which is already apparent in infancy and continues into childhood, may be interpreted as an adverse consequence of this poor nutritional and environmental situation. In case the observed de®cit in energy intake in stunted as well as non-stunted children will persist, an overall insuf®cient linear growth may be expected. Moreover, for the group of stunted children whose energy intake tends to be even lower, further deterioration in growth may occur, making the prospect of catch-up growth impossible. Contributors Ð JLA Hautvast, WA van Staveren and JJM Tolboom designed the study, and are the guarantors for the integrity of the article as a whole. All authors contributed to the preparation of the manuscript. JLA Hautvast was the principal investigator, who managed the data- Food Consumption of Zambian Children JLA Hautvast et al 58 collection, was responsible for the statistical data-analysis, wrote the ®rst drafts and the ®nal version of the manuscript. LJM van der Heijden codesigned the dietary methods, and supervised the dietary data-collection. AK Luneta contributed to the design of the study and provided information on dietary habits in the country of study. WA van Staveren co-designed the dietary methods, and contributed to the interpretation of the results. JJM Tolboom contributed to the interpretation of the results. SM van Gastel performed a substantial part of the data-collection and was responsible for the dietary data-calculation. Acknowledgements Ð The authors wish to express their gratitude to individuals of the Ministry of Health and the National Food and Nutrition Commission for their assistance in the facilitation and implementation of the study. Special thanks are extended to Mr J Mpundu, and Ms E Semba for the administration of the questionnaires. UNICEF-Zambia is acknowledged for their advice and material support. We gratefully acknowledge Ms AWEM van den Wijngaart for her assistance in the pre-testing of the questionnaire. We wish to thank Dr MA van't Hof and Mr J Burema for their statistical advice, and Prof Dr LAH Monnens for reviewing the manuscript. References Abrams B (1991): Maternal undernutrition and reproductive performance. In: Infant and child nutrition worldwide. Issues and perspectives, ed. F Falkner. Boca Raton, Florida: CRC Press. Adair LS & Popkin BM (1992): Prolonged lactation contributes to depletion of maternal energy reserves in Filipino women. J. Nutr. 122, 1643 ± 1655. Allen LH (1994): Nutritional in¯uences on linear growth: a general review. Eur. J. Clin. Nutr. 48, Suppl. 1, S75 ± S89. Beaton GH, Calloway D & Murphy SP (1992): Estimated protein intakes of toddlers: predicted prevalence of inadequate intakes in village populations in Egypt, Kenya, and Mexico. Am. J. Clin. Nutr. 55, 902 ± 911. Calloway DH, Murphy SP, Beaton GH & Lein D (1993): Estimated vitamin intakes of toddlers: predicted prevalence of inadequacy in village populations in Egypt, Kenya, and Mexico. Am. J. Clin. Nutr. 58, 376 ± 384. Cameron M & Hofvander I (1983): Manual on feeding infants and young children. Hong Kong: Oxford University Press. Cameron ME & van Staveren WA (1988): Manual on methodology for food consumption studies. New York: Oxford University Press. Central Statistical Of®ce (1990): Census of population, housing and agriculture. Lusaka: Central Statistical Of®ce. Dean AG, Dean JA, Burton AH & Dicker RC (1990): Epi Info, version 5: a word processing, database, and statistics program for epidemiology on microcomputers. Stone Mountain, Georgia: USD, Incorporated. Dop M-C. Milan Ch, Milan Cl & N'Diaye (1994): The 24-h recall for Senegalese weanlings: a validation exercise. Eur. J. Clin. Nutr. 48, 643 ± 653. Ferguson EL, Gibson RS & Opare-Obisaw C (1994): The relative validity of the 24-hour recall for estimating energy and selected nutrient intakes of rural Ghanian children. Eur. J. Clin. Nutr. 48, 241 ± 252. FAO=WHO (1962): Calcium requirements: report of a FAO=WHO expert group. FAO Food Nutr Ser no. 30. Rome: FAO. FAO=WHO (1970): Requirements of ascorbic acid, vitamin D, vitamin B12, folate and iron: report of a joint FAO=WHO expert group. WHO Tech Rep Ser no. 452. Geneva: WHO. FAO=WHO=UNU (1985): Energy and protein requirements: report of a joint FAO=WHO=UNU expert consultation. WHO Tech Rep Ser no. 724. Geneva: WHO. FAO=WHO (1988): Requirements of vitamin A, iron, folate and vitamin B12: report of a joint FAO=WHO expert consultation. FAO Food Nutr Ser no. 23. Rome: FAO. Fylkesnes K, Musonda RM, Kasumba K, Ndhlovo Z, Mlunanda F, Kaetano L & Chipaila CC (1997): The HIV epidemic in Zambia: socio-demographic prevalence patterns and indications of trends among childbearing women. AIDS 11, 339 ± 345. Gaisie K, Cross AR & Nsemukila G (1993): Zambia demographic and health survey 1992. Columbia: Macro International Inc. George DR & De Francesca BA (1989): Human milk in comparison to cow milk. In Textbook of Gastroenterology and Nutrition in Infancy and Childhood, ed. E Lebenthal, pp 242 ± 243. New York: Raven Press. Gopalan C, Swaminathan MC, Kumari KK, Hanumantha RD & Vijayaraghavan K (1973): Effect of calorie supplementation on growth of undernourished children. Am. J. Clin. Nutr. 26, 563 ± 566. Gould J (1989): Luapula: dependence or development? Vammala, Finland: Vammalan Kirjapain Oy. Grantham-McGregor S (1995): A review of studies of the effect of severe malnutrition on mental development. J. Nutr. 125, 2233S ± 2238S. Haas JD, Murdoch SJ, Riverea J & Martorell R (1996): Early nutrition and later physical work capacity. Nutr. Rev. 54, S41 ± S48. Hamill PVV, Drizd TA, Johnson CL, Reed RB, Roche AF & Moore WM (1979): Physical growth: National Center for Health Statistics percentiles. Am. J. Clin. Nutr. 32, 607 ± 629. Hautvast JLA, Tolboom JJM, Luneta AK & Staveren van WA (in press): Nutritional status and feeding patterns of rural preschool Zambian children living in different ecological areas. Ecol. Food Nutr. Hayes RE, Mwale JM, Bwembya PA, Mulunga MK & Vermoer AB (1994): Weaning practices and foods in high population-density areas of Lusaka, Zambia. Ecol. Food Nutr. 33, 45 ± 74. Hennart Ph & Vis HL (1980): Breast-feeding and post-partum amennorrhoea in central Africa. J. Trop. Ped. 26, 177 ± 183. James WPT, Ferro-Luzzi A & Waterlow JC (1988): De®nition of chronic energy de®ciency in adults. Eur. J. Clin. Nutr. 42, 969 ± 981. Leung WW, Busson F & Jardin C (1986): Food composition table for use in Africa. Bethesda: U.S. Department of Health, Education and Welfare, and Rome: FAO. Martorell R, Rivera J, Kaplowitz H & Pollitt E (1992): Long-term consequences of growth retardation during early childhood. In Human growth: basic and clinical aspects, eds. M Hernandez & J Argente, pp 143 ± 149. Amsterdam: Elsevier Science. Moore HL & Vaughan M (1994): Cutting down trees: gender, nutrition, and agricultural change in the Northern Province of Zambia, 1890 ± 1990. Social History of Africa Series. Portsmouth: Heinemann. Mora JO, Herrera MG, Suecun J, Navarro de L & Wagener M (1981): The effects of nutritional supplementation on physical growth of children at risk of malnutrition. Am. J. Clin. Nutr. 34, 1885 ± 1892. Murphy SP, Beaton GH & Calloway DH (1992): Estimated mineral intakes of toddlers: predicted prevalence of inadequacy in village populations in Egypt, Kenya, and Mexico. Am. J. Clin. Nutr. 56, 565 ± 572. National Food and Nutrition Commission (1987): Food Composition Table for use in Zambia. Lusaka: National Food and Nutrition Commission. Neumann CG & Harrison GG (1994): Onset and evolution of stunting in infants and children. Examples from the Human Nutrition Collaborative Research Support Program. Kenya and Egypt studies. Eur. J. Clin. Nutr. 48, Suppl 1, S90 ± S120. Newman V (1994): Vitamin A and breast-feeding: a comparison of data from developed and developing countries. Food Nutr. Bull. 15, 161 ± 177. de Onis M, Monteiro C, Akre J & Clugstone G (1993): The worldwide magnitude of protein-energy malnutrition: an overview from the WHO global database on child growth. Bull. WHO 71, 703 ± 712. Platt BS (1962): Tables of representative values of foods commonly used in tropical countries. London: Medical Research Council. Prentice A, Paul A, Prentice A, Black A, Cole T & Whitehead R (1986): Cross-cultural differences in lactation performance. In Human Lactation 2: Maternal and environmental factors, eds. M Hamosh & AS Goldman, pp 13 ± 44. New York: Plenum Press. Prentice A & Barclay DV (1991): Breast-milk calcium and phosphorus concentrations of mothers in rural Zaire. Eur. J. Clin. Nutr. 45, 611 ± 617. Prentice AM, Goldberg GR & Prentice A (1994): Body mass index and lactation performance. Eur. J. Clin. Nutr. 48, S78 ± S89. Rutishauser IHE (1974): Factors affecting the intake of energy and protein by Ugandan preschool children. Ecol. Food Nutr. 3, 213 ± 222. Schroeder DG, Martorell R, Rivera JA, Ruel MT & Habicht JG (1995): Age differences in the impact of nutritional supplementation on growth. J. Nutr. 125, Suppl. 4, S1051 ± S1059. Steenbergen van WM, Kusin JA, Voorhoeve AM & Jansen AAJ (1978): Food intake, feeding habits and nutritional state of the Akamba infant and toddler. Trop. Geogr. Med. 30, 505 ± 522. Steenbergen van WM, Kusin JA & Rens van MM (1981): Lactation performance of Akamba mothers, Kenya. Breast feeding behaviour, breast milk yield and composition. J. Trop. Ped. 27, 155 ± 161. Walker SP, Powell CA & Grantham-McGregor SM (1990): Dietary intakes and activity levels of stunted and non-stunted children in Jamaica. Part 1. Dietary intakes. Eur. J. Clin. Nutr. 44, 527 ± 534. Walker SP, Powell CA, Grantham-McGregor SM, Himes JH & Chang SM (1991): Nutritional supplementation, psychosocial stimulation, and growth of stunted children: the Jamaican study. Am. J. Clin. Nutr. 54, 642 ± 648. Food Consumption of Zambian Children JLA Hautvast et al Waterlow JC (1992): Protein-energy malnutrition. London: Edward Arnold. Waterlow JC (1994): Causes and mechanisms of linear growth retardation (stunting). Eur. J. Clin. Nutr. 48, S1 ± S4. West CE, Pepping F & Temalilwa CR (1988): The composition of foods commonly eaten in East Africa. Wageningen: Ponsen & Looyen B.V. WHO (1983): Measuring change in nutritional status. Guidelines for assessing the nutritional impact of supplementary feeding programmes for vulnerable groups. Geneva: WHO. WHO Working Group (1986): Use and interpretation of anthropometric indicators of nutritional status. Bull. WHO 64, 929 ± 941. 59