In ternationa l Scholars Journa ls African Journal of Food Science Research ISSN 2375-0723 Vol. 8 (2), pp. 001-009, February, 2020. Available online at www.internationalscholarsjournals.org © International Scholars Journals Author(s) retain the copyright of this article. Full Length Research paper Investigation of the potential of fortified instant Matooke flour (ITF) in rehabilitation of malnourished children (Part 2) Florence I. Muranga1,2*, Miriam Kanyago1,2, Fabian Nabugoomu3 and James M. Ntambi4 1 Department of Food Science and Technology, Makerere University, P. O. Box 7062 Kampala, Uganda. 2 Presidential Initiative on Banana Industrial Development (PIBID), P. O. Box 32747 Kampala, Uganda. 3 Faculty of Science and Technology, Uganda Christian University, P. O. Box 4 Mukono, Uganda. 4 Department of Biochemistry and Nutritional Sciences, University of Wisconsin-Madison, Wisconsin USA. Accepted 12 September, 2019 This study was done to establish potential of instant Tooke (cooking banana) flour (ITF) as a vehicle food for malnutrition intervention. Hypotheses generated were: (1) There is a significant difference in response with respect to weight gain between malnourished subjects rehabilitated using fortified IFT and Mwanamugimu Dietary Regime (MDR); (2) There is a significant difference in response with respect to serum albumin, ferritin and retinol levels between fortified ITF and MDR. To test the hypotheses, 100 malnourished children admitted at Mwanamugimu Nutrition Unit in Uganda, were randomly allocated to a test (IFT diet from part I of study) or control (MDR diet) group. Growth rates and food intake were measured daily, while blood samples were withdrawn during recruitment and before discharge for testing serum albumin, ferritin and retinol levels. Means were compared using SPSS version 11. The test and control diets were adequate in protein but relatively lower for energy against RDA. Protein and energy intake in the test group was significantly (P < 0.05) higher for porridge, while protein intake in the control group was significantly (P < 0.05) higher for kitoobero. Weight gain in the test was significantly (P < 0.05) higher than the control group. However, both groups showed no significant difference in the bio-chemical indicators. The superiority of fortified ITF formulation over the MDR was proved with respect to weight gain but not for the biochemical indicators. Hypothesis one was accepted but two was rejected. Key words: Instant tooke flour (ITF), matooke, malnourished, rehabilitation INTRODUCTION Bananas are an important staple food for more than 70% of Uganda’s population. Cooking bananas also known as ‘matooke’ (East African highland cooking bananas) are grown by 72% of farmers on an average garden size of one hectare, which is 12.3% of the world total (Karugaba and Kimaru, 1999). The high levels of production and *Corresponding author. E-mail: director@pibid.org. Tel: +256772363271, +256312265789, +256414371050. Fax: +256414340413. consumption of matooke notwithstanding, there are still high levels of malnutrition in Uganda, with 50% of children fewer than five years in rural areas either stunted or underweight (UBOS and ORC Macro, 2006). Malnutrition is a condition resulting from an inadequate (under-nutrition) or excessive intake (over -nutrition) or impaired utilization of one or more nutrients in the body. There are two types of under nutrition, macro and micronutrient malnutrition. Macro nutrient malnutrition in children occurs as protein energy malnutrition (PEM) pre- senting clinically as kwashiorkor, marasmus or marasmic- kwashiorkor (Waterloo, 1993). Symptoms of kwashiorkor may include oedema, sparse brown hair, rashes, misery and apathy, while those of marasmas include retardation, under weight, absence of subcutaneous fat and wasting of muscles (Cameron et al., 1983; Latham, 1989). Marasmic-kwashiorkor is a combination of kwashiorkor and marasmas symptoms (Cameron et al., 1983). It has been observed that PEM is often accompanied by vitamin and mineral deficiencies especially Vitamin A and Iron deficiencies. The underlying factors for malnutrition have been identified to include low energy density of weaning foods and lack of ready to prepare formulations, which lead to infrequent feeding due to the heavy workload of the caregivers especially among the rural poor (Kikafunda, 1996). In response to the need for energy and protein dense foods, the national children’s rehabilitation centre (Mwanamugimu) at Mulago, the main referral Hospital in Kampala, Uganda, developed a series of multi-mix recipes also called kitoobero from locally available foods. The recipes incorporate local starchy staples mainly matooke or rice. The Mwanamugimu has developed a standard dietary regime which is code named Mwanamugimu dietary regime (MDR) comprising a total of seven meals: one from porridge (with maize cereal as energy base) lunch and supper from kitoobero and in between four meals of a high energy milk (HEM) with energy enhancement from oil. The porridge and HEM diets are sweetened and supplemented by a complex mineral vitamin (CMV). The process of preparing maize porridge and kitoobero are, however, laborious and time consuming, therefore their rate of adoption by the caregivers, after discharge is limited. Advances in processing of matooke have succeeded in reducing the relative bulk of matooke starch by pre- gelatinization (Muranga, 1998 and Muranga et al., 2007). Muranga (1998) reported that matooke has high starch content ( dB : Dry basis 81.8 - 85.7%) and subsequently has a high potential as a calorie resource base. Muranga et al. (2007) comprehensively reported on the effect of processing techniques on the bioavailability of matooke starch. Further in the same study using mice, Muranga et al. (2007) demonstrated the significance of available carbohydrate in the diet at the weanling stage of mice. The authors subsequently concluded that solubilisation of matooke starch through heat processing improved its bioavailability and enhanced availability of other nutrients in the diet. The pregelatinised cooked form of matooke has since been branded as Instant Tooke Flour (ITF). Additionally, pilot studies revealed relatively high Protein Efficiency Ratios (PER), when ITF fortified with soybean and sesame was fed to weanling rats (Muranga et al., 2003). The above studies recommended testing of the pregelatinised flours in combination with high protein mixes in malnourished children in a clinical setting prior to adoption for application as a weaning food. In our first paper we established the optimal level of fortification of ITF with soy and sesame as 65:16:19. The objective of this study therefore was to establish whether the pregel form of matooke, as ITF brand, fortified at the above ratio with soy and sesame and micronutrients can serve as a vehicle food for malnutrition intervention using human subjects. Two hypotheses were thereby tested in this part of the study namely: (1) There is a significant difference in response with res- pect to weight gain between subjects rehabilitated using fortified IFT and Mwanamugimu dietary regime (MDR). (2) There is a significant difference in response with respect to serum albumin, ferritin and retinol levels between fortified ITF and MDR. METHODS Study subjects A baseline of the age, sex, heights and weights was taken before commencing the study. Table 1 shows the general characteristics of the study population. The study was carried out at the national children’s rehabilitation centre (Mwanamugimu nutrition unit) in Mulago hospital, Kampala Uganda, where 100 severely malnourished children aged 6 - 59 months were randomly allocated to either a test or control group of equal size using an open labeled randomized trial design. Sample size was determined using a formula reported by Kirkwood, (1988). Generally age range was between 6-59 months of which 50% had kwashiorkor, 31% marasmus and 19% marasmic kwashiorkor. 55% were otherwise free from secondary infection while the rest suffered malaria or other secondary infections during the study. A blood sample was also taken from subjects during recruitment and before discharge. The test group was fed on fortified Instant Tooke Flour (ITF) diet, while the control group was fed on the Mwanamugimu dietary regime (MDR). Both diets were fortified with a therapeutic complex mineral vitamin (CMV) manufactured by Nutriset Malaunay, France. The diets consisted of three definite menu categories, porridge or Uji, High Energy Milk (HEM) and Kitoobero (main menu) for both the test and control groups. The test porridge consisted of ITF supplemented with soy and sesame (65:16:19), sugar and milk and fortified with CMV. The control was similar in formulation but had the ITF composite replaced by maize flour. The High Energy Milk (HEM) diet basically consisted of milk, sugar and CMV with the energy enrichment coming from IFT for the test diet and from oil for the control. The Kitoobero test diet (a multi-mix recipe cooked together in one pan) consisted of ITF-soy-sesame formulation, salt and water while the control consisted of a triple mix of rice, fish powder and groundnut paste. The test and control groups were fed daily on their respective diet regime to provide calories ranging from 150 kcal to 220 kcal per day, with 150 kcal being the minimum. The calorie (energy) requirement given was calculated basing on the body weight of the subject, and this increased as the weight and appetite of the patient increased. According to WHO (2003), a vigorous approach is required to achieve very high intakes and rapid weight gain of >10 g gain/kg/d. The formula used in determining the quantity of feeds per meal for each subject is as shown below. Each subject was given seven meals per day. Amount of food served/meal = 150 Kcal x body weight 7 (1) The feeding regime therefore required administration of porridge at 6 am, HEM four times at 9 am, 3 pm , 9 pm and 12 midnight) where as Kitoobero diets were administered for both lunch and supper at 12 noon and 6 pm respectively. Meal times were observed by the Table 1. General characteristics of the study population. Characteristic Test group (%) Control group (%) Overall (%) Sex of child Male 28 20 48 Female 22 30 52 Age (months) 6-12 13 14 27 13–24 34 25 59 25–36 4 7 11 37–59 0 3 3 Malnutrition type Kwashiorkor 23 27 50 Marasmus 18 13 31 Marasmic-kwashiorkor 10 9 19 Secondary infection Malaria 22 17 39 Diarrhea 3 2 5 Septic wound 0 1 1 No infection 24 31 55 Occurrence of infection Week 1 12 5 17 Week 2 17 17 34 Week 3 1 4 5 Length of hospital stay 1 week 15 17 32 2 weeks 21 24 45 3 weeks 14 9 23 Outcome Discharged (blood withdrawn) 43 47 90 Runaway 6 0 6 Discharged (blood not withdrawn) 1 3 4 study nurse or investigator to avoid sharing of meals. Every meal given to the subjects was weighed before feeding and the leftover food was also weighed to determine the actual intake of food by difference. The actual food intake by difference was calculated by subtracting the leftover food from the food served to each child and was used to calculate the protein (P) and energy (E) intakes per meal follows: Protein intake per meal = Food intake x Protein content per diet (2) The energy intake was also calculated as in equation 2 while Table 2 shows a summary of the calculations. The adequacy of the meal was benchmarked against the Recommended Daily Allowance (RDA) for the ages of the children used in the study. The adequacy was compared with respect to the percentage increase in protein and energy requirements over the standard RDA. The increase was calculated as shown thus: % increase in RDA for Protein = Actual Protein intake (g/day) - RDA of age group for Protein x 100 Protein RDA of age group (3) Percentage increase in RDA for energy was also calculated as in equation (3) but replacing energy for protein. The subjects were fed for a period ranging between 1 and 3 weeks depending on how long they took to attain the median weight or at least 85% of the median weight. This was determined by the initial weight, and secondary infections suffered, such as diarrhea and malaria. The Discharge criteria included meeting the weight gain requirement as stated above and being on the study for at least two weeks (If the child had met the weight requirement and had been on the study for two weeks, they were discharged to avoid congestion due to limited space in the ward). However, the ability of the caretaker to adopt and maintain behavior change Table 2. Calculation of the energy and protein contents of the diets. Group Diet Energy (Kcal) Calculation of energy intake for each child Protein (g) Calculation of protein intake for each child Test Porridge (ITF + sesame +soybean + milk + sugar) 898.1 Intake*898.1/800 26.3 Intake*26.3/800 Kitoobero (ITF + sesame + soybean) 1004.2 Intake*1004.2/700 22.2 Intake*22.2/700 High energy milk (HEM) 1002.02 Intake*1002.02/1000 43.09 Intake*43.09/1000 (milk + ITF + sugar) Control Porridge (maize flour + milk + sugar) 837.0 Intake*837/1000 28. 5 Intake*28. 5/1000 Kitoobero (rice + g/nuts + fish powder) 1178.4 Intake*1178.4/780 39.78 Intake*39.78/780 High Energy Milk (HEM) (milk + oil + sugar) 1006.8 Intake*1006.8/1000 38 Intake*38/1000 250 Test Control 200 in ta k e 150 F o o d 100 50 0 Porridge intake Kitoobero intake HEM intake meals served Figure 1. Comparison in food intake per meal between the test and control group. learning how to feed the child) was taken into consideration in order to avoid a relapse of malnutrition. Serum albumin, ferritin and retinol analyses The serum albumin content was determined using Bromo Cresol Purple (BCP) method a WHO standard method while, ferritin con-tent was determined using a photometric method by a procedure as reported by Tietz (1987). Retinol content was determined using Colorimetric method based on Carr-Price Reactions using Trichoro Acetic Acid as reported by Underwood and Stekel (1984). DATA ANALYSES All the data was analyzed using SPSS version 11. The analysis for each response variable was based on a three factor analysis of variance model in which the groups were one of the factors. For each type of meal there was a three factor ANOVA which included group (test versus control), age and time (length of stay on study). The response variables included protein and energy intake, weight gain, serum albumin, ferritin and retinol levels. RESULTS Food intake Figure 1 shows the comparison in food intake per meal between the test and control groups. The tests here for ITF include ITF -soy-sesame vs. maize flour in porridge, ITF vs. oil in HEM, and ITF-soy-sesame vs. rice and groundnut in the kitoobero diets. There was no significant diff- errence (p > 0.05) between the test and the control groups in the intake of porridge, kitoobero 12 Test Control 10 (g ) 8 in ta k e 6 P ro te in 4 2 0 Porridge Kitoobero HEM Figure 2. Comparison in protein Intake per meal between the test and control group. and HEM. However, in all cases, the control group took in more food compared to the test group. Protein intake Figure 2 shows the comparison in protein intake per meal between the test and control groups. There was a significant difference (p<0.05) between the test and control groups in the protein E n e rg y I n ta k e ( K c a l) 350 Test 300 250 200 150 100 50 0 Porridge Control Kitoobero HEM Figure 3. Comparison in energy intake per meal between the test and control groups. Table 3. Comparison of adequacy of Test (ITF) and Control (MDR) diets for protein and energy with recommended dietary intake (RDA) for the respective subjects in the study. Age GP Protein (g/day) Energy (Kcal/day) Test GP Increase in RDA (%) Control GP Increase in RDA (%) Test GP Increase in RDA (%) Control GP Increase in RDA (%) 6-11 16.41 49.18 16.98 54.36 596 -13.96 535.86 -22.64 12-23 17.15 31.92 20.2 55.38 616.14 -39.94 630.79 -38.51 24-35 19.16 47.38 21.07 62 684.54 -33.28 659.99 -35.67 36-59 - 30.31 78.29 - 960.02 -34.01 intake for porridge. There was indeed a signi- ficant difference (p < 0.05) between Control and Test groups in the protein intake for kitoobero but no significant difference (p > 0.05) in the protein intake for HEM. Energy intake Figure 3 shows the comparison in energy intake per meal between the test and control groups. There was a significant difference (p < 0.05) between the test and control groups in the energy intake for porridge but no significant difference (p > 0.05) in the energy intake for kitoobero and HEM. The test porridge was therefore better than the control porridge with respect to energy, which credits the IFT formulation, but the energy intake for kitoobero and HEM for the two groups was comparable. Table 3 shows the relative adequacy of the food intake for the different subject groups against standard recommended daily allowance for normal subjects (National Academies Press, 2005). The adequacy was compared with respect to the percentage increase in protein and energy requirements over the standard RDA. There was no significant (p > 0.05) difference between the test and the control groups within the different age groups. Except for age 12 - 23 months The food intake for both the test and control diets was Test Control W e ig h t g a in ( g /k g b o d y w e ig h t/ d a y Figure 4. Comparison in weight gain between the test and control groups over study period. adequate for protein since it was close or well above the 50% recommended increase for subjects under reha- bilitation. Energy intake was however, low for both the test and control groups since all the percentages were negative and therefore far below the 10% recommended energy increase for such subjects. Given the overall adequacy of the protein as per Table 2 against the RDA values for the respective age groups, focus of the test is therefore on the performance of the energy sources in the test vs. the Control diet. Weight gain Figure 4 shows the comparison in weight gain between test and control group over the period of study. There was a significant difference (p < 0.05) in the weight gain between the test and control groups over the period of study. Both the test and control groups had a progressive gain in weight in the first two weeks. However, there was a slight decrease in weight gain for the control group in the third week. The difference in weight gain between the test and control group was particularly significant in the third week as it shot up for the test group but decreased for the control group. Figure 5 shows the comparison in weight gain between the test and control groups by age groups over the period of study. There was no significant difference (p > 0.05) in weight gain by age group over the period of study between the test and control groups. However, the test group in the 6 - 11 months and the 24 – 35 months age groups, had a higher weight gain compared to the control group in week 1, 2 and 3. The 12 - 23 months age group had a comparable weight gain for both groups in week 1, in week 2, the control group had a higher weight gain, while in week 3, the test group had a higher weight gain. The test group also had a higher weight gain in the 36 - 59 months age group in week 1 and 2. Blood serum levels Serum albumin: There was an increase in the serum albumin levels for both groups. The test group had an increase from 31.095 - 36.137 g/l while the control group had an increase from 27.953 - 34.148 g/l. However, there was no significant (p < 0.05) difference between the test and control groups in increase in the serum albumin levels over the period of study, implying that the two groups did not respond differently from each other. Serum ferritin: The test group had an increase in the serum ferritin level from 13.447 - 16.056 µmol/l while the control group instead had a drop in the serum ferritin level from 15.169 - 14.704 µmol/l. However, there was no significant difference (p > 0.05) between the test and control groups over the period of study implying that the two groups did not respond differently from each other. Serum retinol: The test group had an increase in the Test Control Figure 5. Comparison in weight gain between the test and control groups by age groups over study period without outliers. Key: 1 = 6-11 months, 2 = 12-23 months, 3 = 24 – 35 months and 4 = 36 – 59 months. serum retinol level from 5.197 - 7.421 g/dl, while the control group had a slight drop in serum retinol level from 6.320 - 6.299 g/dl. However, there was no significant difference (p < 0.05) between the test and control group over the period of study, implying that the two groups did not respond differently from each other over the period of study. DISCUSSION The data revealed the relative energy inadequacy of both the test and control diets. Further the fact that the food intake for the test group though not statistically significant was less compared to the control group is unfortunate for the instant formulation since it is indicative of a significant barrier to energy emanating from inherent bulk in the ITF. In Muranga et al. (2009) the end viscosity of the ITF-soy- sesame optimal formulation was set at 191 RVU compared to that 423RVU for pure ITF which was approximately 50% reduction in viscosity due to incorporation of the protein mixes. Therefore whereas pre gelatinisation of the matooke served the reduction of the bulk to a certain extent further reduction in bulk requires to be effected to enable children in this disadvantaged health state to consume enough amounts to meet their high energy requirements. To this end Muranga et al. (2007) proposed extrusion cooking, which yields an instant flour that is virtually soluble in warm water. The above observations are however, consistent with a pre- vious study reported by Brown et al. (1995) and Bennet et al. (1999) where poor food intake was attributed to the low energy density of the control diet (Brown et al., 1995; Bennet et al., 1999). Generally lower glycaemic and the correspondingly high satiety responses are associated with high amylose starches (FAO, 1988) . Matooke starch, however, is a low amylose starch, but has been associated with heat mediated transformations that could affect the endotherm values (gelatinization transition temperature) and therefore the pregelatinisation process (Muranga, 1998) particularly given that the method of gelatinization was a slow steam operation. That both diet regimes met the RDAs for protein is a fact that can be in part attributable to the milk supplementation in both diet regimes but particularly in the control. The differences in the protein intake between the test and control group for the Kitoobero meals underscores the significance of animal protein in the diet. The control group consumed animal protein from the dried silver fish also known as ‘mukene’, whereas the test group only consumed plant protein (soybean and sesame) as shown in Tables 2 and 3. On the contrary lack of significant differences in the protein and energy intake for HEM after substitution of oil for ITF is significant for the ITF as a vehicle for protein/energy supplementation since oil is a more expensive ingredient. That diets for both the test and control groups did not meet the Recommended Daily Allowance (RDA) for energy is consequent to the carbohydrate bulk barrier in the control MDR from the maize porridge and the rice Kitoobero. The reason for the IFT regime is already high lighted above under nutrient adequacy. The differences observed between the test and control in energy intake with respect to porridge is therefore attributed to the relatively higher energy density and the low viscosity of the IFT (Kanyago et al., 2009). Capdevila et al. (1998) reported volume intake and energy density to be one of main factors influencing energy intake. It is, however, important to note that the RDA is calculated for a normal population and the study population was an abnormal one for which management was being done in order to restore them to normalcy. The calculation for energy requirements per meal was based on the weight of the individual subjects and this increased as the weight of the subject increased. The short fall in energy intake notwithstanding, our results served to demonstrate that the IFT formulation was a superior to the MDR as vehicle for malnutrition intervention which is evidence of a superior bioavailability of energy from the IFT formulation in contrast to the control. Muranga et. al. (2007) indeed reported on this availability after testing growth responses of mice on matooke flour formulations. Indeed the progressive gain in weight for the test group was commensurate with other studies conducted to rehabilitate severely malnourished children. These studies generally report a higher weight gain for the children receiving high-energy-density porridge compared to those receiving low-energy-density porridge (Donmen et al., 1996; Bennet et al., 1999). Catch-up growth represents accelerated tissue gain beyond that expected for chronological age; it occurs after a period of slowing or cessation of growth. Protein and energy may influence both normal and catch-up growth (Tanner, 1989). The test diet regime was therefore able to sustain both the normal and catch growth for the subjects throughout the period of study. This is consistent with what was reported by Uauy and Alvear (1991), that weight loss is usually fully reversible, since weight gain is mainly affected by energy supply in relation with energy expenditure. Nonetheless, the excellent performance in weight gain for the test group in only in the third week is indicative of a requirement for a more available source of carbohydrate at the initial rehabilitation phase in malnutrition similar to that in the weaning stage of animals as reported by Muranga et al. (2007). The control diet was able to some extent in the first week to serve a lower catch up growth but in the second week it fell under negative energy balance and hence negative catch up growth which is unfortunate given the intensity of labour and good will associated with its administration to date. The change in blood serum levels for albumin, ferritin and retinol for both the test and control groups was not significant. Albumin is a hepatically synthesized protein frequently used to estimate overall nutrition status (Marinella and Markert, 1998), while serum ferritin and retinol are used to determine iron and Vitamin A status respectively. Serum albumin for both groups was within the normal ranges at the end of the study. The normal levels range from 35 - 50 g/l (Latham, 1997) . The initial serum ferritin levels for both groups were within the normal range, which is between 9.0 - 30.0 mol/l (Tiez, 1987). This could be attributed to infection, inflammation, liver disease, and states of increased RBC turnover (e.g. ineffective erythropoiesis and hemolysis), which can falsely elevate these levels (Lanzkowsky, 2000). Also there were instances of hemolysis due to resistance from the study subjects during withdrawal blood samples. Despite the high reliability of low concentrations of ferritin in plasma or serum as an indicator of iron deficiency, concentrations in the normal range in iron deficiency are not uncommon (Garibay et al., 2001). The same author also reported iron deficient subjects to be having plasma ferritin concentrations within the normal range. The drop in serum ferritin level for the control group is not easy to explain and is an area for further study. The serum retinol results for both groups were below the normal levels. Blood-serum retinol concentrations of below 20 g/dl signify marginal deficiency and that below 10 g/dl signify definite deficiency (Lindblad et al., 1998; Saskia and Dary, 2002). The control group also had a drop in the serum retinol levels like in the serum ferritin. This also is an area for further study. Serum retinol concentration reflects an individual’s vitamin A status, particularly when the body’s reserves of vitamin A are limited, because serum retinol concentration is homeo- statically controlled and will not drop until body stores are significantly compromised (Olson, 1994) . However, serum retinol concentration is also affected by factors that affect release of holo- RBP from the liver, infection, protein status, adequacy of other nutrients and organ disease. In general, these factors lower serum retinol concentration (Olson, 1994). Overall the results of the biochemical indicators in this study showed that they were a less reliable benchmark for monitoring rehabilitation of malnourished children. The major limitation to this study was time and financial constraints. However, the cost effectiveness issues are being covered in the follow-up study. However, the ITF preparation is a simple technology in which the government of Uganda has invested to ensure effective technology transfer to all Matooke growing communities in the country. SUMMARY Our results demonstrated that there is a significant difference in response with respect to weight gain between subjects rehabilitated using fortified ITF and Mwanamugimu dietary regime (MDR) so hypothesis one was accepted. Further we also evidenced that there was no significant difference in response with respect to serum albumin, ferritin and retinol levels between sub- jects rehabilitated on fortified ITF and those on MDR and so hypothesis two was not accepted. The results further justified the significance of energy density over protein intake in protein energy malnutrition rehabilitation. We therefore recommend exploring technology for further solubilisation of matooke flours in order to optimize energy density. ACKNOWLEDGEMENT This research was possible through funding from Makerere University through the African Institute for Capacity Development (AICAD) grant. We acknowledge Prof. Kakitahi, J.T. of Institute of Public Health Medical School Makerere Uiniversity and Dr. Kiboneka, E.K. of Mwanamugimu Nutrition Unit Mulago Hospital for their input in this study. We also acknowledge the Department of Food Science and Technology, Makerere University, Mwanamugimu Nutritional Unit, Mulago Hospital for infrastructural support and the Presidential Initiative on Banana Industrial Development (PIBID) for providing the final infrastructure for writing and publication of this work. REFERENCES Bennett VA, Morales E, González J, Peerson JM., Romaña GL and Brown KH (1999). Effects of dietary viscosity and energy density on total daily energy consumption by young peruvian children. Am. J. Clin. Nutr,. 70: 285-291. Brown KH, Grinan SM, Perez F, Peerson JM, Ganoza L and Stern JS (1995). Effects of dietary energy density and feeding frequency on total daily energy intakes of recovering malnourished children. Am. J Clin. Nutr., 62:13-18. Cameron M, Hofvander Y. (1983). Manual for feeding of infants and young children. 3 rd Ed. Oxford University Press, pp. 51-66. Capdevila F, Vizmanos B, and Henneberg CM (1998). Implications of the weaning pattern on macronutrient intake, food volume and energy density in non-breastfed infants during the first year of life. . Am. Col. Nutr., 17; 3:256–262 FAO (1998) Carbohydrate in Human Nutrition- FAO Food and Nutrition paper- 66. Report of a joint FAO/WHO expert consultation, Rome 14- 18 April 1997, ISBN 92-5-104114-8. Garibay E, Torres IS, Nelson SE, Ziegler EE, Rogers RR, Janghorbani M, Fomon SJ (2001). Iron absorption during recovery from malnutrition. Am. Col. Nutr. 20(4): 286-292. Kanyago M, Muranga FI, Nabugoomu F, Ntambi JM (2009). Optimal level of fortification of instant matooke flour porridge and its nutritional quality attributes. Accepted May, 2009 for publication in the African J. Food Sci. Karugaba A, Kimaru G (1999). Banana production in Uganda. An essential food and cash crop. RELMA technical handbook No. 18. Regional land management unit, RELMA/SIDA, ICRAF House, Nairobi, Kenya. pp. 1-2. Kikafunda-Kakuramatsi J (1996). Dietary risk factors for childhood malnutrition in Uganda. PhD Thesis, University of Reading, Reading, UK. pp10-15. Kirkwood BR(1988) Essentials of Medical Statistics, Blackwell Scientific Publications. Lanzkowsky P (2000). Classification and diagnosis of anemia during childhood. In: Lanzkowsky P (ed.) Manual of pediatric hematology and oncology, 3 rd Ed. Academic Press, San Diego, New York, London, pp. 1-12. Latham MC (1997). Human nutrition in the developing World. Food and Nutrition Series - No. 29 Food and Agriculture Organization of the United Nations. http://www.fao.org/DOCREP/W0073e/w0073e05. htm#P3172_360093. Accessed 2005 March 17. Lindblad BS, Patel M, Hamadeh M, Helmy N, Ahamad I, Dawodu A, Zaman S (1998). Age and sex are important factors in determining normal retinol levels. J. Trop. Ped., 44:2: 96-99. Muranga FI (1998). Composition and physicochemical characteristics of starches of different banana varieties. PhD Thesis, Makerere University Kampala, Uganda. Pp. 23-36. Muranga FI, Nampala P, Bukusuba J (2003). Investigation of potential of low bulk matooke preparations as vehicle foods for malnutrition interventions II, Poster presentation at the workshop for the launching of IBMU project. 1 st Oct., 2003. Muranga FI, Sampath H, Marlette JA, Ntambi JM (2007). Impact of processing technique on the apparent bioavailability of cooking banana (matooke) Afri. J. Biochem. Res 1(5): 072-077. Marinella MA, Markert RJ (1998). Admission serum albumin level and length of hospitalisation in elderly patients. South. Med. J., 91; 9: 851- 854. National Academies Press (2005). Dietary reference intakes for energy, carbohydrates, fiber, fat, fatty acids, cholesterol, protein, and amino Acids. A report of the panel on macronutrients, subcommittee on upper reference levels of nutrients and interpretation and uses of dietary reference intakes, and the standing committee on the scientific evaluation of dietary reference intakes. pp 168-191 Olson JA (1994). Vitamin A, retinoids and carotenoids. In Shils et al (Eds) Modern nutrition in health and disease 8th Ed. Lea & Febiger Philadelphia, PA. pp. 287-307. Saskia DP, Dary O (2002). Biochemical indicators of vitamin A deficiency In: Serum retinol and serum retinol binding protein supplement: Proceedings of the XX international vitamin A consultative group meeting. The American society for nutritional sciences J. Nutr., 132: 2895S-2901S. Tanner JM (1981). Catch up growth in man. Br. Med. Bull 37: 233-238. Tietz WN (1987). Fundamentals of clinical chemistry 3 rd edition. W. B. Saunders Company Philadelphia, London, Toronto, Mexico city, Rio de Janeiro, Sidney, Tokyo, Hongkong. P. 958. Uauy R, Alvear J (1991). Effects of protein-energy interactions on growth In: Protein-Energy Interactions Nestle Foundation I/D/E/C/G Steering Committee http://www.unu.edu/unupress/food2/UID07E/ uid07e00.htm. Accessed 26 April 2005 UBOS, ORC Macro (2006). Uganda demographic and health survey 2006. Calverton Maryland USA: UBOS and ORC Macro, pp 21-23. Waterloo (1993). Relationship of gain in height to gain in weight. Eur. J. Clin. Nutr., 48. WHO (2003). Guidelines for the Inpatient Treatment of Severely Malnourished Children. SEARO Technical Publication no. 24. p. 48.