![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
,2,**
Helen Keller International,
* Jakarta, Indonesia,
Singapore, and
** New York, NY;
National Institute for Health Research and Development, Jakarta, Indonesia; and
United Nations World Food Programme, Rome, Italy
2To whom correspondence should be addressed. E-mail: remoench{at}singnet.com.sg.
 |
ABSTRACT |
|---|
 TOP ABSTRACT SUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
6-mo intervals for 2.5 y. The poor were found to be appropriately targeted, and program participation ranged from 4 to 18 mo. The proportion of households with debts ranged from 32 to 70%; although it was higher among beneficiaries than controls, it increased among controls, but not beneficiaries. However, only among urban poor mothers in Surabaya were the odds of anemia at endline lower when participating in the FFW program (0.60, 95%CI [0.40–0.89]). Other risk factors for anemia in mothers and children included nutritional status (anemia at baseline, low BMI, receipt of vitamin A capsule, child age) and socioeconomic status (maternal education, having official residency in the area, income level). Thus, postcrisis FFW programs had limited effect on anemia, the main identified nutritional problem. Closer attention is required to the potential for affecting nutritional outcomes through FFW, including food aid quality and quantity and complementary nonfood interventions. Micronutrient deficiencies should be addressed directly via supplements and fortified foods.
KEY WORDS: • food aid • food for work • micronutrient deficiency • anemia • household expenditure
The Asian economic crisis, which started in 1997 and coincided with an El Niño-influenced drought, struck Indonesia with full force in early 1998 and resulted in a change in the Indonesian government in May 1998. These interacting events posed major challenges to poorer segments of the Indonesian population to meet their security, health, and nutritional needs. A country that enjoyed economic growth, stability, and remarkable national coherence was faced, nearly overnight, with internally displaced people, interethnic violence, inflation, increased unemployment, and increased prices.
One of the main effects of the crisis was a reduction in household purchasing power, resulting in reduced consumption of micronutrient-rich foods (1,2). Affected households spent less on items such as eggs, fish, and meat (2,3). This led to an increase in micronutrient deficiencies, such as iron-deficiency anemia and vitamin A deficiency (3,4). Although child anthropometry was not sensitive enough to detect the effect of these changes in food intake [using the Government of Indonesia/Helen Keller International/Nutrition and Health Surveillance System (GOI/HKI NSS)3 data], maternal night blindness, hemoglobin concentration (Hb), and maternal BMI were sensitive (2,4–6).
In 2000, the Bureau for Democracy, Conflict and Humanitarian Response/Office of Food for Peace (FFP) of the U.S. Agency for International Development initiated a 3-y Transitional Activity Program (TAP); the program was conducted by 5 international nongovernmental organizations (NGOs) and comprised mainly Food for Work (FFW) programs and, to a lesser degree, supplementary feeding programs. This postcrisis food aid was targeted to vulnerable groups hardest hit by the 1998 economic crisis, and subsequent fall-out of inflation, lower wages, political and religious conflicts, and food insecurity. Because GOI/HKI NSS data showed that a reduction in purchasing power resulted in reduced quality of diet (fewer micronutrients), and that a reduction in energy intake affected mainly adult women (6), we hypothesized that an injection of food aid would increase energy intake, particularly among adult women, and free up household cash to permit increased consumption of micronutrient-rich foods, thereby reducing the prevalence of micronutrient deficiencies, particularly iron-deficiency anemia.
It should be mentioned that FFW activities are rarely assessed in terms of their effect on nutritional status. More often, they serve to bridge postcrisis conditions with longer-term development (7) and, although food is a resource transfer (an in-kind wage), its value is conventionally measured in terms of household income substitution (in relation to prevailing wage rates) rather than household consumption per se (8). Thus, having a short-term nutrition effect is usually of lower priority than smoothing consumption via generated income, linked to access to employment with an in-kind wage, and the creation of community assets, such as rural roads or improved soil and water management technologies. However, because the TAP/FFW programs aimed to alleviate the effects of the crisis, including the effect on nutrition, it was important to evaluate their nutritional effect.
This article presents findings of an evaluation by HKI of the nutritional effect of the FFW programs of the TAP/FFP initiative, focusing on anemia among young children (3–59 mo old) and mothers.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
1500 TAP beneficiary and 1500 control households were selected (more details below). Baseline (BL) surveys were conducted immediately before the start of the TAP program implementation and food distribution in each site. Subsequent survey rounds were conducted at 6-mo intervals.
Sample size.
A sample of 1200 beneficiary and 1200 control households at each site would enable detection of a meaningful increase of maternal BMI from 21.8 to 22.3 kg/m2, assuming a SD of 3.4 kg/m2 (from GOI/HKI NSS data), with a power of 0.90 and 
of 0.05. The sample sizes required to detect meaningful changes in other outcome indicators, specifically maternal Hb, child anthropometry, and child Hb, were similar or lower. An additional 300 households (25%) were added in anticipation of drop-out.
For child and maternal Hb, a sample size of 95 subjects per group would be sufficient to detect a difference of anemia prevalence of 15% with 90% power and of 0.05. To be able to make subgroups by age or length of participation, for example, we decided to take a blood sample from as many subjects as possible. However, the urban control group was selected from the GOI/HKI NSS, in which Hb samples were routinely taken from 20% of all households. Furthermore, among urban beneficiaries, and among rural beneficiaries and controls, not all households had children < 5 y old. In addition, at one urban site, a communication error led enumerators to collect Hb only from a subsample of beneficiaries. Subjects with Hb < 80 g/L were referred to the health center for treatment and not included in the analysis.
Selection of beneficiary and control group.
Before the beginning of the FFW programs, each implementing NGO provided a list of all households that would participate. Each TAP NGO set criteria for participating households; these varied somewhat among NGOs, but comprised socioeconomic, health, and administrative criteria (see also Table 1). From these, 1500 households were randomly selected using interval sampling, which selects every xth household based on the size of the eligible population compared with the desired sample size. For example, when 6000 beneficiaries participated, every 4th household would be selected, starting from the 1st, 2nd, 3rd, or 4th household. The latter was determined by random selection. Throughout the course of the evaluation, as FFW beneficiaries "graduated" (usually because the program moved to a new area) or dropped out of the program (e.g., due to moving or finding other work), new beneficiary households were added to the evaluation sample. In urban areas, control households were drawn at baseline from the coinciding round of the GOI/HKI NSS, which is a random sample of urban poor. Each round of NSS sampled 2400–3600 households per city. To avoid confounding, the evaluation identified all GOI/HKI NSS households that were not participating in social-safety net programs, then randomly selected 1500 households as controls. As the evaluation progressed, if the number of urban controls fell below 1200 households due to drop out, new control households were recruited from the corresponding round of GOI/HKI NSS to maintain a sample size of 1500 households.
In rural areas, 1500 households were randomly selected from all or half of the program villages, respectively. In CJ, an equal number of beneficiary households was selected from each of 8–11 interfaith committees with which the implementing NGO was partnered, and the beneficiaries were then randomly and equally sampled from all of the 1–6 villages that a committee worked with (31–38 villages/y). In EK, 29 of 58 villages were selected because it was logistically very difficult to visit all villages. For each selected intervention village, the partner NGOs helped to identify comparable villages that met the criteria for program implementation but were not involved with the program. Within those villages, HKI applied the selection criteria of the respective NGO to identify potential control households and then randomly selected an equal number of control households from each village.
The evaluation collected data that were very similar to those collected by the GOI/HKI NSS, and because it was conducted as part of the same agreement with the GOI, all permissions given by the Ministry of Health (MOH) for the GOI/HKI NSS also applied to this evaluation. In accordance with GOI/MOH ethical guidelines, written informed consent was obtained from subjects or caretakers before drawing blood for Hb assessment, respondents were free to discontinue the interview at any time, and data were used solely at the group level and for program evaluation.
Data collection and quality control. To better understand the "income effect" of the FFW programs, household expenditure data were collected by asking the respondents for the amount spent in the previous week on >20 items and on categories of expenditure.
Blood was collected by finger-prick to measure Hb concentration using the HemoCue system. Because the Hb of children < 3 mo old does not reflect iron status (Hb can be low while stores are adequate), data on infants < 3 mo old are not reported. Vitamin A intake data were collected with the semiquantitative 24-VASQ method (9). For vitamin A intake, BL data are based on one interview, whereas intervention data are the median value for all interviews during the intervention period.
Recent graduates in nutrition, public health, agriculture, and related fields were hired as evaluation enumerators. Enumerators were carefully selected, trained, and retrained before each survey round in questionnaire administration and methods for taking anthropometric and blood measurements. Many enumerators also had previous experience working with the GOI/HKI NSS. Each team of 4 was supervised by 1 field supervisor.
For quality control, a special team reinterviewed 10% of households that had already been visited by enumerators. Enumerators that were found to perform badly were dismissed from future data collection rounds. Further details about quality control procedures can be found in de Pee et al. (10). An additional level of quality control, based on a comparison of height data collected during the ongoing and the previous round, was employed to ensure that the same households and key household members were being interviewed and measured from round to round.
Data entry, management and analysis. Data were entered using SPSS Data Entry Builder for Windows, which verifies the validity of values as they are being entered, and 10% of the data were reentered. After data entry, the consistency of the data was checked and further cleaning was performed.
Throughout the course of the program, data were organized according to chronological round of data collection in each site. Baseline surveys were considered round 1 (R1), and subsequent rounds round 2 (R2), round 3 (R3), and so on. These rounds were referred to as "true" chronological rounds. However, over time, households graduated or left the program, and new FFW households were enrolled in the evaluation at various true rounds. To optimize the evaluation, the dataset of true rounds was reorganized in a way that allowed all new households to be set to the same baseline or "virtual" round 1. For the analysis of nutritional effect, the virtual rounds were used, using the following terminology. "Baseline" refers to virtual round 1, "round 2" to virtual round 2, and "endline" to the last round for which data are available for a subject. By using virtual rounds, all children and all mothers that joined the evaluation at any time during the intervention period could be analyzed together. Thus, for this analysis, the assumption was made that the effect of participating in FFW depended on duration [length of time between BL and endline (EL)], but not on the specific period of participation, whether in y 1, y 2, or other period of the evaluation. This assumption is valid because NSS data showed that the prevalence of childhood anemia had remained relatively stable during the period of the evaluation. Furthermore, at some sites, the median duration of follow-up for the beneficiaries was different from the controls (shorter for some and longer for others). To take that into account in the analysis, a selection was made such that control and beneficiary households (including women and children) in any one site had a comparable length of duration of follow-up.
Statistical tests used included 
2 tests to assess differences of proportions between BL and EL within a group or between beneficiaries and controls at BL or at EL; nonparametric tests (Wilcoxon Signed Ranks test and Mann-Whitney U-test) for comparison of nonnormally distributed values; and multiple logistic regression analysis to assess whether being a beneficiary reduced the odds of being anemic at EL while taking into account nutritional status (anemia at BL, stunting, BMI for mothers, child age, child received vitamin A capsule), socioeconomic status (maternal education, income, ownership of the family registration card), and program characteristics (duration of participation for beneficiaries and duration of follow-up for comparison mothers or children, and being a beneficiary). Odds ratios (ORs) are reported for being anemic at baseline and for being a program beneficiary, for a multiple logistic regression model that also included other factors with a significant P-value. P-values < 0.05 were considered significant, but those < 0.10 are also reported. Data management and analysis were conducted using SPSS for Windows 11.5. EpiInfo software was used for calculation of anthropometric Z-scores for children.
|
RESULTS |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
Vitamin A intake from retinol-rich foods (animal foods and fortified foods) did not differ between beneficiaries and controls at either baseline or round 2, among mothers or children (data not shown).
Among children aged 24–59 mo (who eat more family food and in larger quantities than younger children), there was a decline in the prevalence of anemia among beneficiary children in JKT compared with control children (61 to 48%, P < 0.001 and 46 to 45%, P = 0.76, respectively). The same trend was observed in SBY (47 to 38%, P < 0.10 and 43 to 42%, P = 0.86, respectively). In MKS, a significant reduction of anemia occurred between BL and EL among beneficiary children, but a similar change occurred in control children (46 to 37%, P < 0.01 and 48 to 41%, P < 0.12, respectively). In CJ, anemia prevalence was 34–42% and in EK 44–54%, and no significant changes or differences were observed at either site. The main factors that led to higher odds of being anemic at EL were being anemic at BL (all sites), having a mother with anemia at BL (3 sites), and being younger (3 sites) (see Table 5). Being a beneficiary did not affect the risk of being anemic. Among children aged 3–23 mo, risk factors did not differ, but in MKS and CJ, beneficiaries had higher odds of being anemic than controls.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
This evaluation found no effect of the FFW programs on reducing child (24–59 mo) and maternal anemia, except in SBY, where program participation prevented an increase in maternal anemia. For younger children (3–23 mo) in MKS and CJ, the OR of being anemic was higher among beneficiaries than controls. The pathway for the protection from anemia among beneficiaries would be the ability to substitute expenditure away from rice and into more micronutrient dense foods and/or an ability to pay for health interventions or supplements. Possible reasons for the protective effect of FFW program participation among mothers in SBY include the following: 1) it was 1 of the 2 sites with the longest duration of program participation (18 mo); 2) it was an urban site with greater access to micronutrient-rich foods than in rural areas; and 3) the proportion of beneficiary households with debts was lower than in the other 2 urban sites. However, it is somewhat surprising that the effect was observed only among mothers in SBY, not among children.
Although this evaluation found virtually no effect of the FFW programs on anemia, this does not imply that improved nutrition should not be expected from FFW programs. A recent study in Ethiopia showed that receiving food through FFW activities improved children’s weight-for-height, particularly in areas affected by serious drought (12). Thus, the poor effect on nutritional status of these Indonesian FFW programs, implemented in a situation nutritionally less severe than the Ethiopia program, warrants further consideration, in particular because the TAP under which the FFW programs were implemented aimed to alleviate the effects of the economic crisis, including effects on nutrition.
In Indonesia, the main nutritional problem aggravated by the economic crisis was micronutrient deficiencies indicated by anemia levels (2,4). To have a measurable effect on anemia, a minimum implementation time is required. For example, it takes 3 mo of highly supervised consumption of an iron supplement or food vehicle fortified with highly bioavailable iron before effects on the prevalence of child anemia can be observed (13). If iron-fortified flour is used, it will take longer for measurable effects to occur because the time depends on levels of consumption, level of fortification, and type of iron used as the fortificant. Thus, a reduction in anemia via indirect means (dietary change effected by means of shifts in expenditure patterns at household level) will take longer to occur and may not be very large. In fact, 3 of the NGOs enrolled households in FFW for <8 mo, whereas only 2 enrolled households for 18 mo (median duration).
Also, to affect anemia, micronutrient intake should increase (14). The food provided as wages in TAP/FFW programs was mainly staple food, with one NGO complementing the rice with oil and another with oil and pinto beans (see Table 1). These rations were provided as a supplement to other self-obtained sources of food, and the direct contribution of these FFW foods to iron intake was limited. The mechanism of substituting expenditure away from rice into more micronutrient-dense foods or supplements seems to have been limited because the available funds were spread across other competing demands (e.g., clothes, fuel, transport), a phenomenon also observed in a different setting (15). Although the food from FFW did at least have a "buffering" effect on the household economic situation because the proportion of FFW participants with debts did not increase, increased gross income was too little to go beyond maintaining basic levels of energy consumption and meeting other household needs, at the expense of increasing micronutrient consumption. This is not very surprising when we consider that the FFW ration was worth US$6–20/mo per beneficiary household (accounting for fluctuations in exchange rate, market prices, and variation in food basket). Although the published minimum monthly wages of the GOI in TAP areas ranged from 240,000 to 630,000 rupiah (US$25–75)/mo (16), actual incomes were even less (see Table 2). Thus, the relatively high price of micronutrient-rich foods appears to have considerably limited the ability of households to acquire enough vitamins and minerals from their own resources.
Humanitarian agencies try their best to provide appropriate food commodities (17) and have learned from negative outcomes of imbalanced food assistance, particularly in regard to micronutrient deficiencies (18–21), increasingly tailoring the basket of foods to both the micro- and macronutrient needs of vulnerable populations (22). Because FFW programs are often implemented in marginal areas in which health statistics are poor and micronutrient deficiencies can be expected among FFW participants, the share of micronutrients derived from nonfood aid sources must be assessed systematically and deficiencies addressed.
Several options exist for compensating for the lack of micronutrients. Consistent provision of fortified blended foods (FBFs) that have been proven successful in other circumstances (22) to FFW rations would be beneficial. If the FFW programs evaluated here would have added WSB to the general FFW ration, instead of distributing it intermittently and only to some households, micronutrient effects might have been different. However, the formulation of such foods should be done carefully; when addressing anemia, iron compounds with adequate bioavailability should be chosen to avoid a lack of effect. Another approach would be to ensure distribution of micronutrient supplements to postcrisis-affected populations in addition to delivering at least a basic supply of energy and protein. It could be that adequate FFW income would free up household resources to the extent that households could purchase micronutrient supplements at a low (production/subsidized) cost—at least at a cost per vitamin and mineral that is lower than their equivalents in high-quality foods during postcrisis periods. This idea was introduced by HKI a few years ago and is currently being implemented in the form of "sprinkles" (23) together with one of the TAP/FFW NGOs in Indonesia. The advantage of sprinkles compared with FBFs is that sprinkles do not alter the taste or texture of commonly consumed and/or provided food nor do sprinkles require special preparation because they are simply sprinkled over the usual food.
In conclusion, the FFW programs evaluated here did not reduce anemia, except among mothers at 1 site. It appears that the freed-up household resources were used to cover a wide range of costs, and hence did not increase micronutrient intake as much as needed. Although food alone is rarely the answer to malnutrition, it is a key vehicle for most emergency and nonemergency postcrisis responses, and generating value-added nutrition outcomes should be pursued whenever possible (24). Other methods that can also access freed-up household resources should be explored for increasing micronutrient intake in these circumstances, including (in-home) fortification as well as supplementation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
3 Abbreviations used: BL, baseline; CJ, Central Java; EK, East Kalimantan; EL, endline; FBF, fortified blended foods; FFP, Food for Peace; FFW, Food for Work; GOI, Government of Indonesia; Hb, hemoglobin; HKI, Helen Keller International; JKT, Jakarta; MKS, Makassar; MOH, Ministry of Health; NGO, nongovernmental organization; NSS, Nutrition and Health Surveillance System; OR, odds ratio; SBY, Surabaya; TAP, Transitional Activity Program; WSB, wheat-soy blend.
4 The group that was considered to have received little consisted of those that reported at only 1 of the 6 monthly data-collection rounds that they had received WSB in the previous month. Those that reported it at none of the rounds were considered to have never received it.
Manuscript received 15 September 2004. Initial review completed 13 November 2004. Revision accepted 1 March 2005.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTSUBJECTS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1. Helen Keller International/Indonesia (2000) Nutrition and Health Surveillance System (NSS). Monitoring the Economic Crisis: Impact and Transition, 1998–2000 2000 Jakarta, Indonesia.
2. Bloem, M. W. & Darnton-Hill, I. (2001) Micronutrient deficiencies: first link in a chain of nutritional and health events in economic crises. Bendich, A. Deckelbaum, R. J. eds. Primary and Secondary Preventative Nutrition 2001:483-506 Humana Press Totowa, NJ. .
3. Block, S. A., Kiess, L., Webb, P., Kosen, S., Moench-Pfanner, R., Bloem, M. W. & Timmer, C. P. (2004) Macro shocks and micro outcomes: child nutrition during Indonesia’s crisis. Econ. Hum. Biol. 2:21-44.[Medline]
4. Bloem, M. W., de Pee, S. & Darnton-Hill, I. (2005) Micronutrient deficiencies and maternal thinness: first chain in the sequence of nutritional and health events in economic crisis. Bendich, A Deckelbaum, R. J. eds. Primary and Secondary Preventive Nutrition 2nd ed. 2005 Humana Press Totowa, NJ. (in press).
5. Kiess, L., Moench-Pfanner, R., Bloem, M. W., de Pee, S., Sari, M. & Kosen, S. (2000) New conceptual thinking about surveillance: using micronutrient status to assess the impact of economic crises on health and nutrition. Malays. J. Nutr. 6:223-232.
6. De Pee, S., Bloem, M. W., Sari, M., Soekarjo, D. D., Tjiong, R., Kosen, S., Muhilal, & Satoto, (2000) Indonesia’s crisis causes considerable weight loss among mothers and adolescents. Malays. J. Nutr. 6:203-214.
7. Barett, B. Ch., Holden, S. & Clay, C. D. (2004) Can Food-for-Work programs reduce vulnerability?. Dercon, S. eds. Insurance Against Poverty 2004:361-386 Oxford University Press New York, NY. .
8. von Braun, J., Teklu, T. & Webb, P. (1999) Famine in Africa: Causes, Responses and Prevention 1999 Johns Hopkins University Press Baltimore, MD.
9. De Pee, S., Bloem, M. W., Halati, S., Soekarjo, D. D., Sari, M., Martini, E., Kiess, L., Muita, M., Davis, D., Sakya, N. & Gorstein, J. (1999) 24-VASQ Method for Estimating Vitamin A Intake: Reproducibility and Relationship with Vitamin A Status. Report of the XIX International Vitamin A Consultative Group Meeting 1999:96 International Life Sciences Research Foundation Washington, DC.
10. De Pee, S., Bloem, M. W., Sari, M., Kiess, L., Yip, R. & Kosen, S. (2002) The high prevalence of low hemoglobin concentration among Indonesian infants aged 3–5 months is related to maternal anemia. J. Nutr. 132:2215-2221.
11. Helen Keller International/Indonesia (2004) FAM, Food Aid Monitoring Activity, Final Report 2000–2003 2004 Jakarta, Indonesia.
12. Quisumbing, A. R. (2003) Food aid and child nutrition in rural Ethiopia. World Dev. 7:1309-1324.
13. Sari, M., Bloem, M. W., de Pee, S., Schultink, W. & Sastroamidjojo, S. (2001) Effect of iron-fortified candies on the iron status of children aged 4–6 y in East Jakarta, Indonesia. Am. J. Clin. Nutr. 73:1034-1039.
14. Ramakrishnan, U. & Huffman, S. L. (2001) Multiple micronutrient malnutrition. Semba, R. D. Bloem, M. W. eds. Nutrition and Health in Developing Countries 2001:365-391 Humana Press Totowa, NJ. .
15. Reed, B. A. & Habicht, J. P. (1998) Sales of food aid as sign of distress, not excess. Lancet 351:128-130.[Medline]
16. Biro Pusat Statistik (Central Statistics Bureau), GOI (2003) Wage Statistics, Q-2/2002–Q-2/2003 2003 Jakarta, Indonesia.
17. Marchione, T. J. (2002) Foods provided through U.S. government emergency food aid programs: policies and customs governing their formulation, selection and distribution. J. Nutr. 132:2104S-2111S.
18. Berry-Koch, A., Moench, R., Hakewill, P. & Dualeh, M. (1990) Alleviation of nutritional deficiency diseases in refugees. Food Nutr. Bull. 2:106-112.
19. Wolde-Gebriel, Z., Gebru, H., Fisseha, T. & West, C. E. (1993) Severe vitamin A deficiency in a rural village in the Hararge region of Ethiopia. Eur. J. Clin. Nutr. 47:104-114.[Medline]
20. ACC/SCN (1993) Nutritional Issues in Food Aid. Symposium Report, Nutrition Policy Discussion Paper No. 12 1993 Geneva, Switzerland.
21. Toole, M. J. (1995) Preventing micronutrient deficiency diseases. University of Nairobi (1995) Report of a Workshop on the Improvement of the Nutrition of Refugees and Displaced People in Africa, December 5–7, 1994 1995:47-61 Machakos, Kenya.
22. World Food Programme (2004) Nutrition in Emergencies: WFP Experiences and Challenges. UN WFP. Executive Board Annual Session. Document No. WFP/EB.A/2004/5-A/3 2004 Rome, Italy.
23. Zlotkin, S., Arthur, P., Schauer, C., Yeboah Antwi, K., Yeung, G. & Piekarz, A. (2003) Home-fortification with iron and zinc sprinkles or iron sprinkles alone successfully treats anemia in infants and young children. J. Nutr. 133:1075-1080.
24. Webb, P. (2003) Food as Aid: Trends, Needs and Challenges in the 21st Century. World Food Programme, Occasional Paper No. 14 2003 Rome, Italy.
This article has been cited by other articles:
|
|
 |
|
  Minerva BMJ, June 18, 2005; 330(7505): 1456 - 1456. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
