Moderate and severe malnutrition, including underweight (low weight-for-age z score (WAZ) and wasting (low weight-for-length z score [WLZ]), are major contributors to child morbidity and mortality in Bangladesh and other settings with similar severe resource constraints (1). WAZ is commonly used as a summary marker of children's nutritional status, and this indicator has been applied by the hospital service of the International Center for Diarrheal Disease Research, Bangladesh (ICDDR,B), to identify children who require more intensive nutritional therapy following treatment of other acute illnesses. More recently, the World Health Organization (WHO) has recommended using either WLZ or mid-upper arm circumference (MUAC) to identify acutely malnourished children for nutritional rehabilitation (2) because these children have an elevated risk of mortality (3). Regardless of the criteria used to identify malnourished children, information on the most feasible methods for treating these children is needed to ensure adequate nutritional recovery and thereby reduce their risk of dying.
In Bangladesh, the prevalence of underweight (WAZ <−2 with respect to the international growth standard) among children younger than 5 years old (under-5) is among the highest in the world. The latest national nutrition survey (4) found that 41% of under-5 children were moderately or severely underweight (WAZ <−2), 12% were severely underweight (WAZ <−3), 17% were moderately or severely wasted (WLZ <−2), and 3% were severely wasted (WLZ <−3). Previous studies have found that severely underweight children have an ∼8-fold increased risk of mortality (5). Stunting, severe wasting, and intrauterine growth restriction together are responsible for 2.2 million deaths per year and 21% of disability-adjusted life-years for children younger than 5 years (3).
A previous study of hospitalized Bangladeshi children found that the posthospitalization death rate within 3 months of discharge was 14 times higher among severely underweight children than among their better-nourished peers (6), indicating the need for close follow-up of such children. Several investigators have shown that as soon as associated infections and other complications of severe malnutrition are adequately treated in the hospital setting, subsequent nutritional rehabilitation can be completed under supervision through outpatient services or day care centers, using centrally processed or home-prepared food mixtures, supplemental micronutrients, and appropriate parental counseling and psychosocial stimulation of the children (7–15). Many families, however, are unable to participate in day care programs or hospital-based follow-up services because of competing time demands for income generation and child care responsibilities, so alternative forms of community-based service delivery are needed.
For these reasons, we developed community-based nutritional rehabilitation and follow-up regimens, and assessed the effects of providing food supplementation and/or psychosocial stimulation (PS) as additional components to the current hospital-based follow-up regimen. We hypothesized that children ages 6 to 24 months with uncomplicated severe underweight (WAZ <−3), with or without accompanying wasting (WHZ <−2), who are managed in the community (following an initial period of hospitalization for treatment of diarrhea and other acute illnesses) would return for scheduled follow-up visits more often and gain more weight during the 3 months of treatment compared with children in the control group who received the routine, hospital-based outpatient management of severe malnutrition. We further hypothesized that provision of food supplementation and/or psychosocial stimulation would enhance these outcomes.
PATIENTS AND METHODS
The study was a randomized, nonmasked intervention trial to compare the effects of different community-based treatment regimens and the current hospital-based follow-up program.
Description of Study Site
The study was conducted from June 2005 to June 2007 at the Dhaka Hospital of the ICDDR,B and 4 community clinics in 4 outlying police districts (thanas) of the city (Demra, Gulshan, Sabujbagh, and Mirpur thanas), where ∼70% of the severely malnourished patients admitted to the Dhaka Hospital reside. These communities are located from 8 to 15 km from the hospital, and travel times to the hospital generally range from 30 to 60 minutes, using public transportation, at a cost of 35 to 70 taka (US $0.50–$1.00) per round trip. The ICDDR,B Hospital provides treatment for ∼100,000 patients per year, ∼60% of whom are under-5 children. Each year ∼3000 severely underweight (WAZ <−3) under-5 children are admitted to the hospital's specialized nutrition unit. These children are managed according to a standardized treatment protocol based on WHO recommendations (16,17), as described previously (18,19), and they are subsequently studied as outpatients at the hospital's nutrition follow-up unit (HNFU). As part of the current project, community-based nutrition follow-up units (CNFU) were established within existing community clinics in each of the 4 above-named thanas. A female health worker (HW) was trained by the research team and assigned to each CNFU, where she conducted a follow-up clinic from 9 AM to 4 PM 5 days per week.
The present study was approved by the research review committee and ethical review committee of ICDDR,B and the institutional review board of the University of California, Davis. Written informed consent was obtained from a parent or guardian of each of the participating children. All children 6 to 24 months of age from the 4 communities indicated above who were identified as being severely underweight at the time of hospital admission were reweighed and measured when discharged from the hospital. The children were then offered admission to the outpatient follow-up study if they fulfilled the following inclusion criteria: resolution of associated acute illnesses such as diarrhea and pneumonia, WAZ <−3 in relation to the US National Center for Health Statistics (NCHS) reference, and consent granted by the guardian. Although the patients were originally admitted to the study using criteria based on the NCHS reference data, the results presented herein are reported in relation to the new WHO growth standards (20), which became available only after the study started. Likewise, although the ICDDR,B admission criteria at the time of the study were based on WAZ (using the NCHS reference data), WHO has since recommended using WHZ and presence of edema to classify acutely malnourished children in greatest need of treatment (2), so we also present the results for children with WHZ<−2 separately. Children were excluded from the study if they had persistent anorexia, severe wasting (WHZ <−3, using the NCHS reference), edema, clinically apparent congenital or acquired disorders that may affect growth, or other acute or chronic diseases requiring continued hospitalization. Families that lacked a fixed address were also excluded (to avoid difficulties in tracing for follow-up examinations), as were children whose caregivers stated that they were not able to provide time for child care and/or psychosocial stimulation.
Randomization and Interventions
Eligible children were stratified by thana (administrative area under a police station) and randomly assigned to 1 of 5 outpatient treatment groups for a period of 3 months, as described below, using a computer-generated block-randomization scheme, with permuted block lengths of 5 and 10. Randomization was done by an ICDDR,B staff scientist not involved in the study. The follow-up appointments were scheduled at either the HNFU or the CFNU in that child's catchment area, depending on the group assignment.
The caregiver of each child was encouraged to report to the assigned follow-up site (HNFU or CNFU) fortnightly, and all of the families were asked to return to the hospital unit at the end of the third month, regardless of study group. The purposes of the follow-up visits were to reassess the child's health and nutritional status, provide hospital referral when necessary because of either intercurrent illnesses requiring inpatient care or nutritional relapse, and administer the assigned treatments. We emphasized to the caregivers the importance of taking the child to the assigned unit for all of the scheduled appointments, and we also encouraged them to return to the treatment center sooner in case of intercurrent illnesses. The specific components of the assigned treatment packages were as follows:
1. Hospital control group (group H-C): Children assigned to this group received the standard hospital-based, outpatient treatment package, which consisted of fortnightly follow-up for growth monitoring, health and nutrition education, and micronutrient supplementation. During each follow-up visit, caregivers were counseled by an HW on the importance of breast-feeding, use of safe water and hygienic practices, and preparation of nutritious, low-cost diets using locally available food items.
2. Community-based follow-up (group C-C): The standard community-based follow-up package was identical to that provided to the hospital-based control group, except that the follow-up visits took place at the nearest CNFU rather than at the HNFU.
3. Community-based follow-up plus supplementary food (group C-SF): Children assigned to this group received the same treatment package as those in group C-C, except that SF packets and preparation instructions also were provided at the time of each follow-up clinic visit for consumption at home in addition to the children's usual meals. In particular, children 6 to 11 months old received an allotment of 1 food packet (known as “pusti packet” in Bangladesh) per day and children 12 to 24 months received 2 food packets per day. Each packet contained 20 g toasted rice powder, 10 g toasted lentil powder, 5 g molasses, and 3 g soybean oil (total energy per packet ∼150 kcal, with 11% of energy from protein and 20% of energy from fat). Caregivers were asked to mix each packet of the precooked foods with 30 mL (∼6 teaspoonfuls) plain water, which resulted in a porridge-like product with a final prepared energy density of ∼2.2 kcal/g. The caregivers were asked to return all of the unused and/or empty packets at the time of each follow-up visit to permit assessment of compliance. Under-5 siblings of the index children in this group also received the SF (amounts according to age, as indicated above) to minimize any possible sharing.
4. Community-based follow-up plus PS (group C-PS): Children assigned to this group received the same treatment package as those in group C-C, except that they were also provided with PS. PS consisted of child stimulation and parental counseling conducted by trained HWs. Child stimulation included a half-hour play session with the children and their mothers, which was supervised by the HW during each CNFU follow-up visit. The play materials included low-cost, culturally appropriate homemade toys, and the caregivers were encouraged to continue providing similar stimulation to their children at home. During clinic visits the HW also provided half-hour structured lessons (group sessions) on child development and showed parents how to encourage optimal development in their children.
5. Community-based follow-up plus SF and PS (group C-SF + PS): Children assigned to this group received the same treatment package as those in group C-C, except that they were also provided with both SF and PS, as described above.
The counseling sessions indicated above were conducted at either the HNFU or CNFU, according to the children's treatment group assignment. The sessions were convened for groups of mothers, and lasted for 15 to 20 minutes, during which the aforementioned topics were reviewed. The multivitamin drops provided at the time of recruitment and each follow-up visit contained vitamin A palmitate 5000 IU, vitamin D 1000 IU, thiamin 1.6 mg, riboflavin 1 mg, pyridoxine 1 mg, nicotinamide 10 mg, calcium D-pantothenate 5 mg, and ascorbic acid 50 mg/1 mL dose (Square Pharmaceuticals Ltd, Dhaka, Bangladesh). Half of a 20-mg zinc dispersible tablet (to provide 10 mg zinc as zinc sulfate; Acme Laboratories Dhaka, Bangladesh) was given for daily oral administration at home. Iron and folic acid (as fraction of tablet) were given during weeks 2 to 12 to provide daily doses of 3 mg elemental iron (as ferrous fumarate) and 10 μg folic acid per kilogram body weight, which is the standard practice of ICDDR,B. Caregivers were asked to return any empty containers or leftover micronutrient supplements or medicines at the time of each follow-up visit as a measure of compliance.
Other components of the standard treatment package offered to children in all of the study groups included immunizations, according to the Expanded Programme of Immunization guidelines (21), deworming by using a single dose of albendazole if such treatment had not been provided in the last 6 months, and parental counseling on birth spacing and contraceptives, all of which were provided free of charge. Children with suspected serious illness (eg, dehydrating diarrhea, pneumonia, unexplained fever, severe anorexia and/or weight loss leading to severe wasting [WLZ <−3]) were readmitted to the hospital and treated accordingly. At the end of the study, children having a WLZ <−1 were invited to continue attending the nutrition follow-up clinic (regular service component) of ICDDR,B. Children in the respective treatment groups were scheduled to attend the clinics on different days each week, so only 1 set of routine treatments was provided on a particular day.
At baseline and at the end of the third month of study a research assistant measured the children's nude weight, using a digital scale with 10 g precision (Seca, model-345, Hamburg, Germany); recumbent length to the nearest millimeter, using a locally constructed length board; and MUAC to the nearest mm with a nonstretchable insertion tape. The research assistant also measured the mothers’ weights and heights, using standard procedures (22). All of the measurements were taken twice and the average was recorded; if the measurements varied by more than 100 g for weight, 5 mm for length/height, and 2 mm for MUAC, then a third measurement was done and the average of the nearest 2 measures was recorded.
Tracking of Defaulters
On the day of recruitment into the study (ie, the day of original discharge from the hospital), a study assistant accompanied the child and mother to their home to be able to identify the location in case future home visits would be needed. If a child failed to attend the HNFU at end of the third month, then the assistant revisited the home and, when possible, accompanied the child back to the hospital for the scheduled anthropometrics and data collection. If the child could be not traced, then the assistant recorded whatever relevant information could be obtained from family members or neighbors.
The primary outcomes were the rate of attendance at the scheduled follow-up visits and change in weight during the period of treatment. Secondary outcomes were changes in WLZ, WAZ, length, length-for-age z score (LAZ), MUAC, percentage of children who had ≥15% weight gain, severe illnesses that needed rehospitalization, and mortality rate.
Sample Size Calculation
We estimated that a sample size of 75 children per group would be adequate to permit detection of a 33% difference in the follow-up attendance rates and a 325-g difference in weight gain in any of the intervention groups compared with the control group during the 3-month follow-up period, with a significance level (type I error) of 5% and power (1-beta) of 80%, considering 5 treatment groups. To ensure adequate numbers of children for the comparisons of weight gain, we allowed for 25% attrition and therefore planned to enroll 100 children in each group.
Data were entered and analyzed using SPSS for Windows (version 10.2, SPSS Inc, Chicago, IL). Baseline characteristics were compared among the treatment groups. Categorical variables were compared by χ2 test, and the Fisher exact test was applied when the expected number in any cell was ≤5. For normally distributed continuous variables, means were compared by analysis of variance followed by the Tukey test for multiple pair-wise comparisons. Variables not normally distributed were compared by the Kruskal-Wallis test followed by the Mann-Whitney U test. The outcome variables were compared both within and between groups from enrollment to discharge. A 3-factor analysis of covariance was done to assess the main effects of SF, PS, and HNFU versus CNFU on anthropometric outcomes, controlling for the respective baseline anthropometric values. Factorial analysis (SF vs PS) on follow-up attendance rate also was completed. Results from all of the randomized children were included in the analysis on an intention-to-treat basis. P < 0.05 was considered statistically significant.
A total of 553 children were assessed for eligibility, of whom 32 were not enrolled because they failed to meet inclusion criteria, 12 were excluded because their caregivers were not able to provide time for child care and/or psychosocial stimulation, and 2 were excluded because they lacked a fixed address (Fig. 1). Thus, a total of 507 children (44% girls) were enrolled in the trial. The mean (± SD) initial age of the children enrolled in the study was 12.6 ± 4.0 months, and their mean anthropometric status indicators were WAZ −3.83 ± 0.6, LAZ −3.46 ± 0.99, and WLZ −2.71 ± 0.76, in relation to the WHO 2006 growth standard. In addition to being severely underweight, 337 (66.5%) of the children also were severely stunted, 138 (27.2%) were moderately stunted, and 412 (81.3%) were moderately wasted (WLZ <−2). At baseline, there were no significant groupwise differences in the children's mean age, anthropometrics, or family and household characteristics (Table 1), including their early child-feeding practices, immunization status, parental education and employment, and other indicators of socioeconomic status (data not shown). After 3 months of study, a total of 131 children defaulted, including 1 who died (Fig. 1). By inquiring of their neighbors, we learned that most of these families (∼90%) migrated from their initial urban residence to their home villages. The prevalence of breast-feeding was maintained at nearly the same levels as baseline in all 5 treatment groups.
A total of 60% to 81% of the children attended the first scheduled follow-up visit, depending on treatment group. As early as the first follow-up visit, the children in group H-C were less likely to return for their appointment than the children of groups C-SF and C-SF + PS (Fig. 2). The rate of spontaneous attendance gradually decreased with time following hospital discharge in all treatment groups, but children in group H-C had a consistently greater default rate. The attendance rates at follow-up visits were always greatest in groups C-SF and C-SF + PS, followed by groups C-PS, C-C, and H-C (Fig. 2). Likewise, the mean number of follow-up visits during the 3-month study period differed significantly by treatment groups in the same order: C-SF (4.8 ± 1.6) = C-SF + PS (4.5 ± 1.9) = C-PS (4.1 ± 2.0, but less than C-SF) > C-C (3.1 ± 2.1) > H-C (1.9 ± 1.8) (analysis of variance). By factorial analysis it was found that children who received SF completed a significantly greater number of follow-up visits than children who did not receive SF, but there was no difference in follow-up attendance rate in relation to the provision of PS (Fig. 2).
No particular child or family characteristics predicted which children were more likely to attend the scheduled follow-up visits. The children who were continuing in the study at the end of 3 months did not differ significantly from those who had defaulted with regard to their initial age, nutritional status, or other individual, parental, or household characteristics, regardless of treatment group (data not shown). Similar comparisons were done among the children who attended the last follow-up visit spontaneously (n = 246) versus those who attended only after 1 or more home visits by the study assistant to encourage participation in the final measurements (n = 130), and there were no significant differences for any of the baseline characteristics (data not shown). The children who spontaneously attended the appointment at 3 months gained more weight and WLZ, however, compared with those who did not attend without special encouragement: median (interquartile range) 0.85 (0.57–1.23) versus 0.74 (0.43–1.07) kg, respectively, P = 0.015; and 0.81 (0.48–1.35) versus 0.62 (0.17–1.14) WLZ score; P = 0.001.
The changes in the children's weight, length, and related anthropometric indices during the period of treatment are shown by treatment group in Table 2. All of the treatment groups demonstrated a positive change in median weight and WLZ (P = 0.001). The absolute and percent weight gains and changes in WAZ and WLZ were greatest in groups C-SF, C-SF + PS, and C-PS, which did not differ significantly from each other, lowest in group H-C, and intermediate in group C-C. The changes in length were greater for all 4 groups managed in the community compared with group H-C, but the former groups did not differ significantly from each other. Notably, the median LAZ declined significantly in all of the groups, although the groupwise order of change paralleled the length data. Despite the improvements in weight-related indicators of nutritional status noted above, the anthropometric results indicate that the children were still underweight and severely stunted at the end of the treatment period. Recovery from severe underweight (WAZ >−3) was significantly more common in group C-SF + PS (53%) than in the other 4 groups (24 to 40%); P<0.05.
The specific effects of food supplementation, PS, and location of follow-up care (hospital vs community) on changes in anthropometrics were assessed by 3-factor analysis of covariance, controlling for the respective baseline anthropometric values. There were no interactions among these main effects. Gains in weight (kg), WAZ, and WLZ were significantly greater (P = 0.009, P = 0.018, and P = 0.02, respectively) among children who received SF compared with those who did not receive SF; and gains in length and LAZ were significantly greater (P = 0.048 and P = 0.021, respectively) among children treated in the community compared with those who were managed at the hospital. There were no significant effects of PS on any of the anthropometric outcomes.
Children who attended at least 5 of the scheduled follow-up visits gained more in weight, length, and all of the anthropometric indices than those who attended fewer of the scheduled appointments (Table 3). Children who were wasted (WLZ <−2) at the start of treatment had a greater change in WLZ than nonwasted children, (median [interquartile range] 0.83 [0.44–1.42] vs 0.43 [0.4–0.79], P < 0.001), but there was no significant difference in change in WAZ. On the contrary, the wasted children experienced a greater decline in LAZ than the nonwasted children (median [interquartile range] −0.29 [−0.59 to 0.02] vs 0.02 [−0.34 to 0.25], P < 0.001) (Table 4), and there was a significant positive relation between the initial WLZ and subsequent change in LAZ by regression analysis and controlling for treatment group (difference in LAZ = 0.23 + baseline WLZ * 0.23, P < 0.001).
Twenty children reported to the follow-up units for unscheduled visits because of intercurrent illnesses (mostly cough, fever, and/or diarrhea), and 21 children were readmitted to the hospital because of diarrhea plus fever (n = 11), diarrhea plus pneumonia (n = 5), diarrhea plus weight loss (n = 2), diarrhea plus urinary tract infection (n = 1), and typhoid fever (n = 1). There were no significant differences in the number of unscheduled clinic visits or hospitalizations by treatment group, and all of these intercurrent illnesses were successfully treated. One child (group H-C) died at home 26 days after enrollment. This child did not attend any of the scheduled follow-up appointments, and the parents did not report the child's associated illness symptoms (fever and cough), which were only discovered after the HW's home visit at 3 months because of the missed final appointment.
Compliance with micronutrient supplementation among the children who attended the follow-up visits was reportedly 98% to 100% during the previous 2 weeks. Compliance with food supplementation was ∼90% in group C-SF and ∼92% in group C-SF + PS.
The results of this trial indicate that previously hospitalized, severely underweight children whose follow-up appointments were scheduled at community nutrition clinics were more likely to continue in treatment than children who were managed at the hospital follow-up clinic, and the children managed in the community had greater weight gains and linear growth. Moreover, provision of food supplements at the time of each return visit further enhanced the attendance rates, weight gain, and recovery from wasting. The effects of the interventions on the children's psychosocial development are under analysis and reported separately.
These results are not surprising because management of severely malnourished children in the community causes fewer disruptions of family life, less interference with breast-feeding, and reduced exposure to other potentially infected children. Community-based treatment also costs less. Earlier studies of edematous and/or severely wasted children in Bangladesh showed that home-based/community-based management was effective (9), cost less than one-fourth of hospital-based care, and was preferred by the children's mothers (23). Another recently completed study at ICDDR,B (24) also confirmed that home-based management was effective. Although both of the foregoing studies found beneficial results with home-based/community-based management, the health personnel in those projects completed frequent home visits, which is not likely to be economically feasible in large-scale programs. Therefore, we eliminated home visits in the present study and requested that the parents take their children to the community-based or hospital follow-up clinics instead. Clearly, the community-based approach was more successful, possibly because of the distances involved and related transportation and time costs. Although there was a fairly high percentage of children who failed to complete the study (25.8% overall), this could be explained partially by demolitions that were being conducted in the study communities, which forced people back to their village homes.
Providing food supplements also encouraged greater attendance at the scheduled follow-up appointments, which is not unexpected in view of the families’ precarious economic, and presumably food security, situations. Children who received food supplements also gained more weight, which may be a direct effect of the supplements per se or an indirect effect of the children's greater access to other components of the treatment regimen. For example, children who returned more frequently received more micronutrient supplements, including zinc, which has been shown to enhance weight gain during recovery from severe malnutrition (25–27). This may also explain why children who were managed in the community had greater rates of growth than those who were required to return to the hospital for their follow-up visits, even when they did not receive food supplements.
Other factors that were associated with greater weight gain were more frequent clinic attendance, regardless of treatment group. As noted above, the beneficial effect of clinic attendance on weight gain may have been related to specific components of the treatment regimen, such as the SF and/or micronutrients. Alternatively, it is possible that the frequent attendees had more concerned or knowledgeable caregivers who were able to devote more time to child care activities. The available information does not allow us to distinguish among these possible explanations. An earlier study likewise found that children whose parents regularly attended nutrition education sessions had greater growth rates (28). Children in the present study who were wasted at the start of treatment had less linear growth than those who were not wasted. This observation is consistent with a previous report, which found positive relation between weight-for-height and linear growth in 4 longitudinal growth studies of young children (29).
Whereas exposure to PS also increased the frequency of follow-up visits, this had no further additive effect of the children's growth during this 3-month period. This observation is consistent with the results of earlier studies in Jamaica, which indicated that PS had no effect on the growth of 9- to 24-month-old stunted, Jamaican children (30); however, other possible benefits of PS on the children's mental development index and longer-term growth are being assessed.
A previous study at ICDDR,B (24) reported considerably greater weight gain (∼7.5 g/kg body weight/day) among than was observed during the present study. In the previous study, the children were severely wasted (WH <70% of NCHS reference), and they received 1 week of inpatient care and intensive feeding at the ICDDR,B Nutrition Rehabilitation Unit before starting weekly outpatient monitoring and micronutrient supplementation for the following ∼30 days until achieving WH >80% of the reference median. During the inpatient period, the children's mean energy intake was ∼200 to 250 kcal/kg/day, and their mothers received hands-on training on preparation of energy-dense local recipes of rice, lentils, and vegetable oil (khichuri, 1.4 kcal/g; and halwa (2.5 kcal/g). Possible explanations for the differences in weight gain reported for these 2 studies are the greater degree of wasting among children enrolled in the former study, the inclusion of an intensive inpatient treatment period, and the more intensive dietary counseling that was provided.
Despite the positive responses to treatment, 43% of the children were still wasted and 47% were severely underweight at the end of the 3-month treatment period. Even in the groups that received food supplements, 32% to 37% had WLZ <−2 at the end of the study, despite apparently high compliance with recommended supplementation. One reason for this may be that supplements provided a relatively small amount of additional energy. For example, if the children's average initial weights were ∼6 kg and the recommended level of energy intake for children recovering from acute malnutrition is ∼150 to 200 kcal/kg/day (16,31), then these children would have required an average energy intake of 900 to 1200 kcal/day. The mean energy intake from breast milk of partially breast-fed children in developing countries is estimated to be 379 kcal/day in 9- to 11-month-old children and 346 kcal/day in 12- to 23-month-old children (32). Thus, if we assume that the average energy intake from breast milk was 363 kcal/day in these 6- to 23-month-old children, there would have been an energy deficit of 538 to 838 kcal/day, which would have to be provided by supplementary/complementary food. One study from Bangladesh found that the average energy intake from complementary food in 9- to 12-month-old children was only 147 kcal/day (33). Thus, the additional amount of energy provided by the food supplements in the present study may not have been sufficient to meet the children's total daily energy needs. Moreover, the supplemented food may have replaced some of the home food, which was not measured in the present study. Notably, the food supplements and amounts that we distributed in the present study are the same as those that are provided to growth-restricted children by the Bangladesh National Nutrition Program, although we also provided micronutrient supplements, which the national program does not (34). The results of the present study suggest that a greater amount of food or longer duration of supplementation may be necessary to promote greater rates of recovery. Moreover, better counseling would be another option to enhance the desired outcome.
In conclusion, community-based provision of follow-up services increases adherence to the follow-up schedule and physical growth, and provision of SF–with or without PS–enhances the rate of nutritional recovery. All of the interventions provided in the present study can be delivered through existing primary health care services in the local communities. This would require a small amount of additional funding for the SF (each packet of 150 kcal costs 3.25 taka, ∼5 US cents) and micronutrient supplements (20 taka, ∼30 US cents/child/2 weeks), and to cover the expense of training and supervising HWs so that they are able to provide the relevant services, including growth monitoring and PS. Thus, community-based service delivery, including SF, could be scaled up through existing community clinics to permit better nutritional rehabilitation of greater numbers of severely underweight children at reasonable cost.
We sincerely appreciate the statistical assistance of Janet M Peerson, statistician of the Program in International and Community Nutrition, and the advice of Dr Rahman Azari of the Department of Statistics and Dr Lucia L. Kaiser of the Department of Nutrition of the University of California, Davis.
1. Pelletier DL, Frongillo EA Jr, Schroeder DG, et al. The effects of malnutrition on child mortality in developing countries. Bull WHO 1995; 73:443–448.
3. Black RE, Allen LH, Bhutta ZA, et al. Maternal and child undernutrition: global and regional exposures and health consequences. Lancet 2008; 371:243–260.
4. National Institute of Population Research and Training (NIPORT), Maitra and Associates, and Macro International. 2009. Bangladesh Demographic and Health Survey 2007. Dhaka, Bangladesh and Calverton, Maryland, USA: National Institute of Population Research and Training, Maitra and Associates, and Macro International.
5. Pelletier DL, Frongillo EA Jr, Schroeder DG, et al. A methodology for estimating the contribution of malnutrition to child mortality in developing countries. J Nutr 1994; 124:2047S–2081S.
6. Roy SK, Chowdhury AK, Rahaman MM. Excess mortality among children discharged from hospital after treatment for diarrhoea in rural Bangladesh. Br Med J (Clin Res Ed) 1983; 287:1097–1099.
7. Heikens GT, Schofield WN, Dawson S, et al. The Kingston project. I. Growth of malnourished children during rehabilitation in the community, given a high energy supplement. Eur J Clin Nutr 1989; 43:145–160.
8. Brown KH. Appropriate diets for the rehabilitation of malnourished children in the community setting. Acta Paediatr Scand Suppl 1991; 374:151–159.
9. Khanum S, Ashworth A, Huttly SRA. Controlled trial of three approaches to the treatment of severe malnutrition. Lancet 1994; 344:1728–1732.
10. Heikens GT, Schofield WN, Dawson SM, et al. Long-stay versus short-stay hospital treatment of children suffering from severe protein-energy malnutrition. Eur J Clin Nutr 1994; 48:873–882.
11. McDonald K, Grantham-Mc Gregor S, Chang S. Social stimulation of the severely malnourished child: a home training programme. Indian J Paediatr 1989;56:97–103.
12. Grantham-McGregor S, Powell C, Walker S, et al. Nutritional supplementation, psychosocial stimulation and mental development of stunted children: the Jamaican study. Lancet 1991; 338:1–5.
13. Engle PL, Bentley M, Pelto G. The role of care in nutrition programmes: current research and a research agenda. Proc Nutr Soc 2000; 59:25–35.
14. Collins S, Dent N, Binns P, et al. Management of severe acute malnutrition in children. Lancet 2006; 368:1992–2000.
15. Brown KH, Nyirandutiye DH, Jungjohann S. Management of children with acute malnutrition in resource-poor settings. Nat Rev Endocrinol 2009; 5:597–603.
16. World Health Organization. Management of Severe Malnutrition: A Manual for Physician and Other Senior Health Workers. Geneva: World Health Organization; 1999.
17. World Health Organization. Management of the Child With a Serious Infection or Severe Malnutrition: Guidelines for Care at the First-Referral Level in Developing Countries. Geneva: World Health Organization; 2000.
18. International Centre for Diarrhoeal Disease Research, Bangladesh: Centre for Health and Population Research. Standardized Protocol Training Manual for the Management of Severely Malnourished Children. Dhaka, Bangladesh. 1999.
19. Ahmed T, Ali M, Ullah MM, et al. Mortality in severely malnourished children with diarrhoea and use of a standardised management protocol. Lancet 1999; 353:1919–1922.
21. Bangladesh Ministry of Health and Family Welfare (BMHFW). Expanded Programme on Immunization. EPI Programme Review 1998. Dhaka: Expanded Programme on Immunization, Directorate General of Health Services, Ministry of Health and Family Welfare, Government of Bangladesh; 1998.
22. World Health Organization. Physical Status. The Use and Interpretation of Anthropometry. Report of a WHO Expert Committee. WHO Technical Report Series 854. Geneva: World Health Organization; 1995.
23. Ashworth A, Khanum S. Cost-effective treatment for severely malnourished children: what is the best approach? Health Policy Plan 1997; 12:115–121.
24. Ahmed T, Islam MM, Nahar B, et al. Home-based nutritional rehabilitation of severely malnourished children recovering from diarrhea and other acute illnesses. Paper presented at 10th annual scientific conference, ICDDR,B, Dhaka, 2002.
25. Naheed A, Walker Fischer CL, Mondal D, et al. Zinc therapy for diarrhoea improves growth among Bangladeshi infants 6 to 11 months of age. J Pediatr Gastroenterol Nutr 2009; 48:89–93.
26. Brown KH, Peerson JM, Baker SK, et al. Preventive zinc supplementation among infants, preschoolers, and older prepubertal children. Food Nutr Bull 2009; 30:S12–S40.
27. Roy SK, Tomkins AM, Haider R, et al. Impact of zinc supplementation on subsequent growth and morbidity in Bangladeshi children with acute diarrhoea. Eur J Clin Nutr 1999; 53:529–534.
28. Fernandez-Concha D, Gilman RH, Gilman JB. A home nutritional rehabilitation programme in a Peruvian peri-urban shanty town (pueblo joven). Trans R Soc Trop Med Hyg 1991; 85:809–813.
29. Dewey KG, Hawck MG, Brown KH, et al. Infant weight-for-length is positively associated with subsequent linear growth across four different populations. Matern Child Nutr 2005; 1:11–20.
30. Walker SP, Powell CA, Grantham-McGregor SM, et al. Nutritional supplementation, psychosocial stimulation, and growth of stunted children: the Jamaican study. Am J Clin Nutr 1991; 54:642–648.
31. Ashworth A. Efficacy and effectiveness of community-based treatment of severe malnutrition. Food Nutr Bull 2006; 27:S24–S48.
32. Brown KH, Dewey KG, Allen LH. Complementary Feeding of Young Children in Developing Countries: A Review of Current Scientific Knowledge. Geneva: World Health Organization; 1998.
33. Kimmons JE, Dewey KG, Haque E, et al. Low nutrient intakes among infants in rural Bangladesh are attributable to low intake and micronutrient density of complementary foods. J Nutr 2005; 135:444–451.
34. National Nutrition Programme (NNP). Ministry of Health and Family Welfare. Government of the People's Republic of Bangladesh. Monitoring Report. December 2007. Dhaka; 2008.