Approximately 25% of all younger than 5-year-old children in the world and up to 40% of those living in the least developed countries are undernourished (1). In addition to causing acute morbidity and adverse long-term sequelae, malnutrition contributes significantly to global childhood mortality. Children with severe malnutrition carry the highest risk of morbidity and death as individuals, but the greatest population level burden of the condition lies on the more prevalent mild and moderate malnutrition (2). To manage the problem on a population level, there is a need to develop new and inexpensive interventions to prevent the development of childhood malnutrition and to treat it early, when hospitalization is not required.
Recently, Briend et al. (3,4) developed the concept of highly fortified spreads that can be stored for months even in the hot climatic conditions of sub-Saharan Africa. Spreads need no cooking before consumption and bacteria do not grow on them, making them convenient to use when food preparation facilities are limited, and hygiene is suboptimal. A therapeutic spread named ready-to-use therapeutic food (RUTF) has proven beneficial for severely malnourished children both in famine and nonfamine situations (5-7). In several clinical trials in Malawi, our own study group has shown that RUTF is safe and effective in the rehabilitation of severely malnourished children, both on an inpatient and outpatient basis (8-11). Recently, we also documented improved weight gain among underweight and stunted 3- to 4-year-old ambulatory children supplemented with RUTF (12). Thus, spreads seem to have potential in community-level management of malnutrition. Until now, however, information on their use among moderately malnourished younger than 2-year-old infants is scarce, although this is the age of highest risk for malnutrition development in low-income countries (13,14).
To provide first data on the acceptability and tolerability of fortified spreads (FSs) among infants and young children and to establish the range of its recipients expected medium term growth and change in blood haemoglobin (Hb) concentration, we carried out a randomized dose-finding trial among moderately underweight, ambulatory 6- to 18-month infants in rural Malawi. Besides using several different doses of FS, we also tested 2 different formulations: 1 in which the main source of protein was cow's milk powder and another 1 in which it was soy flour. Soy is a low-cost source of protein, and it may replace more expensive cow's milk powder for spread preparation if well accepted and biologically as effective. The ultimate aim of this descriptive study was to provide data that would facilitate the design of more focused trials in similar target groups, with predefined hypotheses on a possible causal link between FS supplementation and growth and hematological outcomes.
The study was carried out between November 2002 and March 2003 in Lungwena, an approximately 100-km2 rural community in Mangochi district of Malawi, southeastern Africa. In this area, exclusive breastfeeding for 6 months is almost nonexistent, and infant diet is typically complemented with thin maize porridge ("phala," 10% dry weight of maize flour) from 2 to 6 months of age and thick dough known as "nsima" (28% dry weight of maize flour) from the age of 9 to 2 months. Underweight and stunting are common. In a recent prospective cohort study of 813 newborns, underweight (weight for age [WAZ] z score less than −2) and stunting (height for age z score less than −2) prevalence was 10% and 50% by 6 months of age and 40% and almost 80% by 18 months of age, respectively (15). Little previous data exist on infant anaemia prevalence or micronutrient status from the area. In 1 study on 120 severely malnourished, hospitalised patients (average age 24 months), 92% had anaemia (Hb less than 110 g/L) and mean blood Hb concentration was 84 g/L. Both dietary factors and infections were judged to contribute to anaemia levels, but the relative importance of the 2 factors was not assessed (16).
To maximise the power of the trial, it was carried out in the rainy season (November to March) when nonintervention infants were expected to grow most slowly. During this season, food security in the area is typically at its lowest and weight and length gain of children poorer than during the rest of the year (15). The main staple, maize, is normally harvested after the rains, that is, from April to May.
Infants 6 to 17 months of age were eligible to participate in the study, if their guardian gave a written informed consent and if they were underweight, as defined by a weight that was below the green area when plotted on the Malawian road to health card (corresponding approximately to a weight-for-age z score less than −2 of the World Health Organisation (WHO)-adopted National Center for Health Statistics reference curve) (17). Infants were not eligible if they had any of the following: weight below 5.5 kg, weight for height z score less than −3, severe medical condition requiring hospitalization, adverse reaction within 30 minutes after a test dose of 15 g of FS, or likelihood to move out of study area during follow-up.
Low WAZ was chosen for entry criteria because it is the standard indicator for monitoring child growth in the defined age group in Malawi. It also seems to predict long-term adverse childhood outcomes, notably mortality at least, and other anthropometric indicators like weight for height or length for age (18).
Two different formulas of fortified spread (milk-based and soy-based FS) and 3 dosing regimens for each (25, 50, or 75 g/day) were tested. Additionally, 1 group of infants received 5 g/day milk-based FS and another 1 received no supplement. Thus, in total, 7 different supplementation groups and 1 unsupplemented group (that received no sham spread) were included in the study. The energy contents of the daily supplement varied from 96 kJ (23 kcal, 5 g milk-based FS) to 1.7 MJ (397 kcal, 75 g milk-based FS). Soy-containing formulas tasted slightly sweeter than the milk-containing ones, but otherwise the look, taste and packing of different formulas were identical.
The milk-based spread was prepared by mixing dried skimmed milk, peanut butter, oil and sugar as described for the RUTF (9). In the soy-based formula, the dried skim milk was replaced by defatted soy flour. Both FSs differed from the RUTF because of added micronutrients, the amounts of which were based on recommended daily intakes for normal infants (19) and not on the WHO formula for severely malnourished infants (20).
Micronutrients were added to the spreads as dry mix during the FS preparation. The fortification schemes were identical to all supplementation doses, so that for most micronutrients, the daily supplement contained approximately the recommended daily intake (Table 1). The small differences in micronutrient contents (notably calcium and magnesium and, to a lesser degree, iron) between the supplements were related to the micronutrient content of the spread before fortification. Micronutrients were not protected in any special way in the spreads, but quality control analyses have documented that factory-prepared spreads retain their micronutrient contents and activity for at least 12 months (Michel Lescanne, Nutriset Inc., personal communication). The phytate/iron and phytate/zinc molar ratios in the supplements varied from 0.2 to 3.5 and from 0.4 to 8.3 for iron and zinc, respectively, suggesting low bioavailablity of iron and moderate bioavailablity of zinc (21-23). Ascorbic acid/iron weight ratio in the supplements varied from 2.3 to 2.7, suggesting that iron absorption was not greatly affected by vitamin C (24). The molar ratios were calculated using molar weights of 660, 65.4 and 55.8 for phytate, zinc and iron, respectively (23).
For this trial, the spreads were produced industrially by Nutriset Inc., (Malaunay, France) and packed in foil-covered, air-sealed sachets, each containing a ration for 1 to 3.5 days (depending on the group). The production process is rather simple and can be reproduced locally (9).
This was a randomised, controlled, parallel-group, investigator-blind, dose-finding study, analogous to phase 2 clinical trials in drug development. After enrollment, infants were randomised with 2 computer-generated random number lists to receive, for 12 weeks at home, 1 of 8 alternative food supplementation schemes. The supplements were delivered home weekly for the first 4 weeks and fortnightly thereafter. The spread was given to the infants by their guardians, who were counselled that it was food rather than medicine. The guardians were advised to serve the supplement during the infant's normal eating times and otherwise to feed the infant normally.
The importance of giving the full ration to the intended beneficiary was emphasized at enrollment, and the advice was repeated at each distribution. To monitor consumption of the supplement, the empty sachets for the previous week were collected at each home visit. Other dietary intakes of the participants were not monitored or analysed.
All families were visited fortnightly and interviewed about the morbidity and possible adverse events. Families receiving food supplements were also questioned about the acceptability and tolerability of the product. Further information on the same were collected during a medical examination at the end of the trial.
Assignment and Blinding
Potential participants were initially screened by trained research assistants during a 2-month period at clinic-based or outreach under-5 clinics from Lungwena Health Centre. Possibly eligible subjects (6-17 months of age and weight below or almost below the green area in the under-5 card) were in the same week invited to a more detailed interview and physical and anthropometric assessment by a research nurse and the study nutritionist (H.K.). Those meeting the enrollment criteria were immediately randomised to the study, but the start of the intervention and follow-up was delayed until a sufficiently large number of individuals from the same living area were enrolled. This was done to coordinate the food delivery. In practice, the interval between randomisation and start of the intervention ranged from 1 to 10 weeks.
To avoid that some participants would receive more than 50% of their daily energy requirement from the supplement, we used 2 different randomisation lists. Infants greater than 7.5 kg of weight could be included in any of the intervention groups, whereas infants between 5.5 and 7.0 kg could be randomised to all other groups but not the largest dose one (75 g/day). Infants less than 5.5 kg were not eligible to participate in the study.
An office assistant performed the randomisation and food allocations. The researchers and the laboratory assistant participating in measuring the outcome variables remained blind to the group allocation until the end of data collection. Fortnightly home visits to deliver food and obtain information on its safety and tolerability were made by office assistants not participating in the anthropometric outcome measurements.
The primary outcomes for the study included changes in weight, length and blood Hb concentration during the 12-week supplementation period. Weights were measured using an electronic paediatric weighing scale (SECA, model 834, Chasmors Ltd., London, UK) and recorded to the nearest 10 g. Length was measured to the nearest 0.1 cm by an infantometre (Kiddimetre, Raven Equipment Ltd, Essex, UK). Haemoglobin values were measured by a laboratory assistant with a HemoCue instrument (HemoCue AB, Ängelholm, Sweden) and recorded to the nearest g/L. Recorded values for weight and height were averages of 3 individual measurements by the study nutritionist (H.K.). The standard error of baseline measurements was 39 g for weight and 2.8 mm for height. The percentage reliability was 0.13% and 0.23%, respectively.
Secondary outcome variables included changes in WAZ, length for age (HAZ) and weight for length z scores, calculated with EPI-Info 2002 program using the WHO-adopted NCHS reference standards (17).
Safety of the supplements was assessed by counting the episodes of intolerance or other adverse events, reported by the infants' guardians during the fortnightly home visits or at the final medical examination.
Sample Size and Other Statistical Considerations
In the absence of earlier experience of FS use in this target group, no data existed on the right dose or formula for the spread to be used. Therefore, the trial included a rather wide range of intervention groups, rather than focusing on just 1 intervention and 1 control group. To make such a dose-finding trial statistically conclusive would, however, require 800 participants even without allowance for multiple comparisons or loss to follow-up (assuming 80% power, 95% confidence and 0.5 standard deviation [SD] difference in outcomes between the control and intervention groups). Inasmuch as such an approach was not feasible, we designed the study as a preliminary descriptive trial that had neither preset hypothesis on the outcome nor power to be statistically conclusive. No mathematical formula was used for sample size calculation, but the number of participants (16-20/group) was chosen as a convenience sample to provide distribution data on growth and other outcome variables. Such data can later be used for the sample size calculation of a more focused trial with preset hypothesis.
In a retrospective power calculation, the chosen approach and group size provided the trial with 80% power to detect a mean difference of approximately 500 g in weight gain and 1.5 cm in length gain between the unsupplemented and any of the intervention groups.
The data were entered with a Microsoft Excel 2000 programme and analysed with SPSS 11.0 software package. The results are shown with descriptive terms (means and SD). Comparison of linear outcome data between 2 groups was done by calculating the difference in means and its 95% confidence interval (CI). When more than 2 groups were compared, we used post hoc hypothesis testing with analysis of variance (ANOVA) and Bonferroni correction. However, because there was no predefined trial hypothesis, other comparisons were not accompanied by hypothesis testing.
The research protocol was reviewed and approved by the ethics committees at the University of Malawi, College of Medicine, Malawi, and Tampere University Hospital, Tampere, Finland. For each participant, written informed consent was obtained from the guardian escorting the infant to the enrollment session. All participants who were randomised into the unsupplemented or the 5 g FS/day group during the trial received a 12-week supplementation with FS after the trial completion (50 g of milk-based FS/day).
Of the 224 infants initially screened, 69 were not eligible (61 infants WAZ greater than −2, 4 infants older than 18 months, 3 infants less than 5.5 kg and 1 infant severely ill) and 27 declined participation, leaving 128 infants who were randomised into 8 different supplementation schemes. Three additional infants (twin siblings to randomised participants) were included in the trial (in the same group with their sibling) when their guardians brought them forward after the randomisation but before the onset of the intervention. Five enrolled participants, all from the FS 5 g/day group were unable to start the intervention (4 moved away and 1 guardian died) and 1 discontinued participation immediately after the onset of the intervention. Thus, a total of 126 infants started the intervention, and 125 infants completed the follow-up (Fig. 1).
All of the guardians reported that their children avidly ate the spread and swallowed it without difficulty. Research assistants confirmed this for each participant by occasionally observing the FS eating session. All of the empty FS sachets were returned to the study team. At weekly interviews, no intolerance or other adverse reactions were reported.
The guardians usually reported having given the food to their infants with a spoon or with a finger.
Growth and Change in Blood Hb Concentration
Figure 2 describes the changes in weight, length and blood Hb concentration for each individual participant, and Table 3 summarises the data according to the intervention group. Compared with totally unsupplemented participants, weight gains were on average higher among infants receiving daily at least 25 g of FS, and length gains and change in blood Hb concentrations were on average higher among infants receiving any dose of FS. In an unadjusted ANOVA, however, group level differences approached statistical significance only for the change in Hb (Table 3). Applying post hoc Bonferroni correction to the ANOVA indicated further that the only group demonstrating statistically significant difference to the control group was the one receiving 50 g/day milk-based FS (data not shown).
All average gains were largest among infants receiving 50 g of FS daily. During the 12-week supplementation, the mean difference (95% CI) in the gains between infants receiving 50 g/day of milk-based FS and nonsupplemented controls was 290 g (range, −130 to 700 g), 0.9 cm (range, −0.3 to 2.2 cm), and 17 g/L (range, 0-34 g/L) for weight, length and blood Hb, respectively. Accordingly, the WAZ improved on average 0.2 z score units, and HAZ fell only 0.05 units among the supplemented (50 g/day), whereas WAZ remained unchanged and HAZ fell more (0.3 units) among the unsupplemented infants. Weight for length was little affected in all groups (Table 3).
No marked outcome differences were observed between the groups receiving comparable doses of soy-based or milk-based FS (Fig. 2, Table 3). In a pooled analysis, comparing outcomes among all infants receiving 25, 50 or 75 g/day of milk-based FS with those among participants receiving 25, 50 or 75 g/day of soy-based FS, the difference in means (95% CI) was 0.0 kg (range, −0.5 to 0.7 kg), −0.1 cm (range, −0.6 to 0.3 cm), and 1 g/L (range, −5 to 7 g/L) for weight, length and blood Hb concentration, respectively.
This study was conducted to examine the feasibility of using FSs as food supplements to moderately underweight, ambulatory infants during the complementary feeding period. It confirmed, as was previously noticed in therapeutic programmes treating severely malnourished children, that FSs are well accepted and tolerated, and they can be eaten on a sustained basis at a dose of up to 75 g/day from the age of 6 months. In this specific sample in rural low-income setting where growth faltering was common and malaria was highly endemic, underweight infants receiving 25 to 75 g FS for 12 weeks gained more weight and length and had a higher increase in blood Hb concentration than infants not receiving any intervention. Infants receiving soy protein-containing FS had comparable outcomes to those receiving milk powder-containing FS.
The results of the current trial-a mean difference of 290 g and 0.9 cm in 12-week weight and length gain, respectively, between unsupplemented infants and the supplemented group with highest gains (50 g milk-based FS/day)-are comparable to those obtained with fortified blended flour supplements (25). This suggests that spreads may be as effective as blended flour supplements to break the process of stunting and to accelerate recovery from moderate undernutrition in infancy. It should be noted, however, that infants in the current sample had a high prevalence of anaemia and malaria at enrollment. Although a comparable prevalence has been found elsewhere in sub-Saharan Africa (26), the current results may not be applicable to all other conditions in which malaria prevalence may be lower. The HIV/AIDS epidemic, on the other hand, is not likely to significantly modify the population level effect of FS in this age group. Human immunodeficiency virus tests were not done in the current trial, but in a comparable sample of 180 underweight 6- to 15-month-old infants from the same area, less than 3% had HIV infection (John Phuka, personal communications).
Besides possible questions about external validity, the major limitation of our study was its absence of preset hypothesis and its small sample size. Being the first acceptability and dose-finding evaluation of spreads in this target group, the trial included a large number of different intervention groups, and there was no specific, predefined hypothesis about the magnitude of intergroup or intragroup variation in outcomes. Hence, the observed associations and differences cannot be taken as proof of causality. The biological plausibility of the results, their consistency using several outcome variables, the magnitude of the effect and a trend for dose-dependency are, however, consistent with a possibility for a contributory link between the spread supplementation and a favourable outcome. Therefore, further trials with predefined hypotheses and adequate confidence and power are warranted in the same setting and target group. Data presented here will allow the estimation of the required sample size for such trials.
Providing all of the necessary nutrients to rapidly growing or underweight infants during the complementary feeding period has proven difficult, especially in the absence of foods from animal sources. To fill nutrient gaps during this period, the WHO recommends the use of fortified products or nutritional supplements (27). To date, the strongest evidence in favour of this approach comes from studies using multinutrient sprinkles or tablets, some of which have documented favourable effects both on linear growth and anaemia alleviation (28,29).
Fortified spreads represent a new interesting approach to provide energy and missing nutrients to underweight or growing infants at low costs, especially in conditions in which the normal intake is based mainly on cereals, with a low fat and micronutrient content (4). At a daily dose of 25 to 50 g, spreads are somewhat more expensive than micronutrient-fortified corn-soy flour, tablets, or sprinkles (US$ 0.10-0.20/day vs US$ 0.02-0.04/day), but they also provide the beneficiaries with extra energy, essential fatty acids and a wider range of micronutrients (30). Whereas this trial provided no information whether the supplementary spreads merely replaced something in the infants' diet or truly added to their dietary intake, earlier evidence suggests that the spreads affect their recipients' habitual diet less than flour-based supplements, probably because of their low volume and no need for processing before consumption (12). Transportation and storage costs are also likely to be smaller for the spreads than for the more bulky flour-based supplements. The efficacy of these new food supplements should therefore be tested and compared with existing approaches in adequate-size controlled clinical trials.
The authors thank the people of Lungwena, the staff at the Lungwena Training Health Centre and our research assistants for their positive attitude, support and help in all stages of the study. The authors also thank Ms. Heini Huhtala, MSc, for advice on statistical analyses. The study was financially supported by grants from the Academy of Finland, Foundation for Paediatric Research in Finland, and Medical Research Fund of Tampere University Hospital. André Briend was formerly a consultant to Nutriset, the company that produced and provided FS free of charge for this study.
H.K. contributed to study design, collected the data, did most of the analyses and finalised the manuscript. K.M., A.B. and M.M. contributed to study design, data analysis and manuscript writing. P.A. designed the study, raised its funding, participated in data analysis and wrote the first draft of the manuscript. P.A. is the guarantor of the manuscript.
1. United Nations Children's Fund (UNICEF). The State of the World's Children 2001: Early Childhood
. Oxford, UK: Oxford University Press, 2000.
2. Pelletier DL, Frongillo EA Jr, Schroeder DG, et al. The effects of malnutrition on child mortality in developing countries. Bull World Health Organ
3. Briend A, Lacsala R, Prudhon C, et al. Ready-to-use therapeutic food for treatment of marasmus. Lancet
4. Briend A. Highly nutrient-dense spreads: a new approach to delivering multiple micronutrients to high-risk groups. Br J Nutr
5. Collins S, Sadler K. Outpatient care for severely malnourished children in emergency relief programmes: a retrospective cohort study. Lancet
6. Diop El H, Ndour M, Ndour M, et al. Comparison of the efficacy of a solid ready-to-use food and a liquid milk-based diet for the rehabilitation of severely malnourished children: a randomized trial. Am J Clin Nutr
8. Manary MJ, Ndekha M, Ashorn P, et al. Home-based therapy for severe malnutrition with ready-to-use food. Arch Dis Child
9. Sandige H, Ndekha MJ, Briend A, et al. Home-based treatment of malnourished Malawian children with locally produced or imported ready-to-use food. J Pediatr Gastroenterol Nutr
10. Ciliberto MA, Sandige H, Ndekha MJ, et al. A comparison of home-based therapy with ready-to-use therapeutic food with standard therapy in the treatment of malnourished Malawian children: a controlled, clinical effectiveness trial. Am J Clin Nutr
11. Ndekha MJ, Manary MJ, Ashorn P, et al. Home-based therapy with ready-to-use therapeutic food is of benefit to malnourished HIV infected Malawian children. Acta Paediatr
12. Maleta K, Kuittinen J, Duggan MB, et al. Supplementary feeding of underweight, stunted Malawian children with a ready-to-use food. J Pediatr Gastroenterol Nutr
13. Shrimpton R, Victora CG, de Onis M, et al. Worldwide timing of growth faltering: implications for nutritional interventions. Pediatrics
14. Maleta K, Virtanen S, Espo M, et al. Timing of growth faltering in rural Malawi. Arch Dis Child
15. Maleta K, Virtanen S, Espo M, et al. Seasonality of growth and the relationship between weight and height gain in children under three years of age in rural Malawi. Acta Paediatr
16. MacDonald AC. The effect of zinc fortification of a therapeutic complementary food on the growth and immune function of severely malnourished children in Malawi. A PhD thesis. Ontario, Canada: University of Guelph; 2000.
17. World Health Organization (WHO). Measuring Change in Nutritional Status: Guidelines for Assessing Nutritional Impact of Supplementary Feeding Programmes for Vulnerable Groups
. Geneva, Switzerland: WHO, 1983.
18. Pelletier DL, Low JW, Johnson FC, et al. Child anthropometry and mortality in Malawi: testing for effect modification by age and length of follow-up and confounding by socioeconomic factors. J Nutr
19. Food and Agriculture Organization and World Health Organization (FAO and WHO). Joint FAO/WHO Expert Consultation. Vitamin and Mineral Requirements in Human Nutrition
. Geneva, Switzerland: WHO, 2002.
20. World Health Organization (WHO). Management of Severe Malnutrition: A Manual for Physicians and Other Senior Health Workers
. Geneva, Switzerland: WHO, 1999.
21. World Health Organization/Food and Agriculture Organization/International Atomic Energy Association (WHO/FAO/IAEA). Trace Elements in Human Nutrition and Health
. Geneva, Switzerland: WHO, 1996.
22. Hurrell R. Influence of vegetable protein sources on trace element and mineral bioavailability. J Nutr
23. International Zinc Nutrition Consultative Group (IZiNCG). Technical document no. 1. Assessment of the risk of zinc deficiency in populations and options for its control. Hotz C, Brown KH, eds. Food Nutr Bull
2004;25:94-203. Available at: http://www.izincg.ucdavis.edu/publications/FNBv25n1supp2zinc.pdf
. Accessed June 10, 2006.
24. Hurrell R. How to ensure adequate iron absorption from iron-fortified food. Nutr Rev
25. Caulfield LE, Huffman SL, Piwoz EG. Interventions to improve intake of complementary foods by infants 6 to 12 months of age in developing countries: impact on growth and on the prevalence of malnutrition and potential contribution to child survival. Food Nutr Bull
26. Nabakwe EC, van Lichtenbelt M, Ngare DK, et al. Vitamin A deficiency and anaemia in young children living in a malaria endemic district of western Kenya. East Afr Med J
27. Pan American Health Organization/World Health Organization (PAHO/WHO). Guiding Principles for Complementary Feeding of the Breastfed Child
. Washington, DC: PAHO, 2003.
28. Zlotkin S, Arthur P, Schauer C, et al. Home-fortification with iron and zinc sprinkles or iron sprinkles alone successfully treats anemia in infants and young children. J Nutr
29. Zlotkin S, Antwi KY, Schauer C, et al. Use of microencapsulated iron (II) fumarate sprinkles to prevent recurrence of anaemia in infants and young children at high risk. Bull World Health Organ
30. Nestel P, Briend A, de Benoist B, et al. Complementary food supplements to achieve micronutrient adequacy for infants and young children. J Pediatr Gastroenterol Nutr
31. World Health Organization and United Nations Children's Fund (WHO and UNICEF). Global Strategy for Infant and Young Child Feeding. Geneva, Switzerland: WHO, 2003.
© 2006 Lippincott Williams & Wilkins, Inc.