-14~~~~~~~~~~~~~~~~~~~~ Public Disclosure Authorized

LA~~~~~~~~~~~~~~A

~~~~ .0 C) A~~~~~~~~~~~~~~~~~~~P Public Disclosure Authorized

M7~~~~~~ Public Disclosure Authorized Public Disclosure Authorized ENHANCING THE CONTRIBUTION OF THE EEALTHSECTOR IN THE PREVENTION

OF MNAJUTRIflON

Review of Project Appraisalsin Africa

Edited by Tonia Marek

DIVISIONOF HUMAN RESOURCES TECHNICAL DEPARTMENT AFRICA REGION WORLD BANK NOVEMBER1993

Translated from the French by Alison SondhausCarroll

With support from the Afica Bureau of The U.S. Agencyfor InternationalDevelopment ARTSIHealthand Human Resources Analysis for Africa (HIRAA) Support for Analysis and Research in Africa (SARA) To obtain a copy of this report or to send comments, please contact:

Tonia Marek The World Bank 1818 H St., NW Washington, DC 20433 USA

Tel: (202) 473 4086 Fax: (202) 473-8216

This document was written for the benefit of project designers, implementors, and specialists in the nonindustrializedcountries, and in developnmentagencies.

Each chapter may be consulted independentlyfor assistance in setling up a program or mnakingdecisions related to the experiences sumnarized here.

The author(s) of this documentbear(s) euclusiveresponsibilityfor the observations,interpretations, and conclusionsexpressed here. They do not representthe vews of the World Bank, its affiliated institutions, the membersof the Board of (Executive)Directors, or the countries they represent. CONTENTS

FOREWORD ...... ii

SUMMARY...... iii

ACKNOWLEDGEMENTS ...... iv

1. INTRODUCTION 1......

A. Choice of interventionstrategies 1......

B. Impact on operations design ...... 2

II. THE HEALTH SECTOR AND NUTRITIONINTERVENTIONS IN AFRICA ...... 4

A. Benefits already provided by the health sector ...... 4

1. Prevention of diarrheal diseases ...... 4 2. Prevention of intestinalparasites ...... 5 3. Prevention of malaria ...... 5 4. Control of preventable childhooddiseases (). ... : ...... S5

B. services that could easily be provided by the health sector ...... 6

1. Prevention of vitamin A deficiency...... 6 2. Prevention of iron deficiency ...... 18 3. Prevention of iodine deficiency ...... 27 4. Promotingbreastfeeding ...... 39 5. Dietary managementof infection ...... 53

C. Nutrition services that could be provided by the heal sector if a radical change in approach were adopted ...... 73

1. Food supplementation ...... 73 2. Nutritional rehabilitation ...... 88 3. Growth monitoring and promotion ...... 101

D. Nutrition education ...... 122

Im. CONCLUSIONS ...... 139

BIBLIOGRAPHY...... 140 NOTES ...... 180

i FOREWORD

Malnutritionin Africa has many causes,but most of them ire linked to poverty. The World Bank, in its efforts to reduce poverty in Africa, has identifiedsix interventionpoints':

1. Macroeconomicpolicies to ensuregrowth and improvementin income,better distributionof inicometo the poor, and more efficientallocation of resources.

2. Measures to ensure that the poor have access to food, and to improve their use of that food with a view toward improvingtheir nutritionalstatus.

3. Measures Lo ensure that the poor have access to physical and financial resourcesand social servicesand to improvethe efficiencyand effectiveness of these services.

4. Programs to shield the poor from the hamful effects of seasonal variations in income,drought, or macroeconomicshocks.

5. Prograns aimed at increasing the participation of the poor, particularly women, in the socioeconomicchoices that affect thei.rlives.

6. Programs aiuned at eliminating discriminationagainst certain groups, in particular womenand the poor.

This documentis airnedspecifically at improvingmeasures related to number three above.

it does not pretend to be an exhaustivestudy; therefore, certain topics, such as the nutritionalstatus of refugeesand displacedpopulations, are not addressed.

This documentwas first publishedin French under the title "CnmmentAni6liorer la Contribution du Secteur de la Sant6 dans la Lutte Contre la Malnutrition: Revue d'Evaluations de Projets en Afrique". Copies in both languagescan be requested.

1. Source:Foreword by EdwardV.K. Jaycox, Vice-President, Africa Region. The World Bank, in Salmen,Lawrence, 1992. REducingPoverty. Povcrty and SocialPolicy Series, PaperNo. 1. WorldBank. ii SUMMARY

Thisdocument reviews many assessments of interventionsin the healthsector in the area of nutrition,principally in Africa. It proposesto illustratewhat may be this sector'smost effectivecontribution to the preventionof malnutrition. It is intendedfor those responsiblefor the design and implementationof projects,project directors, and nationaland intermationalspecialists.

This reviewof publishedand unpublishedproject appraisals underscores the health sector's decisiverole in the preventionof malnutrition,but it also emphasizesthe necessity for rigorouslydefined interventions to avoidfailure and wastedresources. As a result, this documentrecommends that the healthsector (a) continuewith the benefitsit alreadyprovides: vaccinations,therapeutic treatments (for malaria,infections, dehydration), which also have a considerableimpact on malnmtrition;(b) undertake new and easilyprogrammable services thatrequire only a smallsupplementary investment of resources:promotion of breast-feeding and preventionof iron, vitaminA, and iodinedeficiency; (c) take on more complexand/or demandingactivities (for example, nutritional recuperation, the monitoring of growth, and food supplementprograms) only through a more community-basedapproach and subjectto the availabilityof all the necessaryresources and a detailedprogram; if theseconditions are not met, such activitiesshould not be undertaken,and efforts insteadshould be focusedon the first two types of benefitsto increasecoverage and improveits quality;(d) undertake work in the area of nutritioneducation only in conjunctionwith communicationsprofessionals and only when enough resources are availableto train staff, along with other requirements.

iii AUTHORS

This document was conceived, compiled, and edited by Tonia Marek, nutritionist, AFTHR. The chapter authors are as follows:

a Dr. Bernard Maire, researcher in nutrition at ORSTOM, Montpellier (France), recruited as a consultantby AFTHR to write the sections on preventingiron, vitamin A, and iodine deficiency, nutritional rehabilitation, the monitoring and promotion of growth, nutritional education, and food supplements.

* Mrs. Margaret Kyenkya-lsabirye, consultanton child feeding, UNICEF-New York (USA); Mrs. Gabrielle Fahmer, consultantto UNICEF-NewYork, and Mrs. Helen Armstrong, training coordinatorfor the Baby Friendly Initiative, UNICEF-New York (USA), wrote the section on promotionof breast-feeding.

* Dr. Ellen Piwoz, consultantto the project "SupportforAnalysis and Research in Africa (SARA),Academyfor EducetionaiDevelopment,Washington (USA), wrote the section on dietary treatment of infectionwith support from the Africa Bureau (HHRAA Project) at USAID.

* The section on activitiesalready undertakenby the health sector was taken from the document, "Malnutritionalid Infection: A Review" (with the authorizationof the ACC- SCN department of the United Nations), published by Andrew Tomkins and Fiona Wilson in October 1989.

ACKNOWLEDGMENTS

This report benefited greatly from the commentsof numerous colleagues, including specificallyDr. Mary Ann Anderson,specialist in breast-feedingat USAID; Dr. John Mason at ACC-SCN; Dr. Nicholas Cohen at WHO; Mr. Dick Heyward at UNICEF; Dr. Suzanne Prysor-Jones of the SARAProject at the Academyfor EducationalDevelopment; Dr. Hope Sukin, nutritionist in the Africa Bureau at USAID-Washington;Drs. Jacques Baudouy,Yves Genevier, Jean-Louis Lamnboray,and Maryse Pierre-Louis, specialists at the World Bank; Dr. Michele Lioy, IEC specialist at the World Bank; Dr. Judith McGuire, nutritionist at the World Bank; and Mr. Timothy Stone, independent expert on micronutrients.

The report was typed by Mr. Jim Shafer of the World Bank. Its presentationwas designed by Ms. Marit Hammond, design consultantfor the SARA Project, Academy for International Development, Washington (USA), with support from the Africa Bureau (HHRAA Project) at USAID.

iv I. INTRODUCTION

*l Choiceof interventionstrategies

Beforeidentifying any interventionstrategy to anything else? Or is it a problem of access to prevent malnutrition, each program and project ? Or of lack of knowledgeabout how leader passes through the following decision to raise a child? These three causesoften coincide stages, illustrated in the "Triple A" cycle in Africa. popularizedby UNICEF: * What does one do to resolve these Figure 1. The "Triple A" cycle problems? Each sector (particularlyagriculture, education, health, industry) can and should contributeto their solution(action). EVALUATIONOF THESITUATION This docunent deals essentiallywith actions (ASSESSMENT) relevant to the health sector, and defines its contribution to resolving these problems by reviewinglessons leamed from health projects in ANALYSISOF Africa. However,this documentdoes not address THECAUSES the meansby whichthe fundamentalcauses of the ACTIONOTHEPROBLEM malnutrition problem may be resolved, as $\IN \~ identifiedin the conceptualframework developed by UNICEF (below); this conceptualframework depicts schematicallythe multiple factors that Source:UNICEF. affectnutritional status and that are the domain of the projectand program leaders (targeting,equity, and policy measures).

* Once a malnutritionproblem is identifiedas In Afriu, the World Bank's health sector a result of studies and surveysthat have evaluated fiSances health and nutrition, food security and the prevalenceand extentof the problem(situation nutrition, and stand-alone nutrition projects. assessment),the interventionstrategy is selected. More of the projects could includenutrition, but due to the complexityof the problem's causes, * Beforedeciding on an interventionstrategy some project and program leaders hesitate to get to prevent malnutrition, it is essential to ask involved in prevention of malnutrition. This questions about the underlying causes of the documnentshould allow these decision makers to problem (analysis of the situation): is it a chooserationally the type of nutritionintervention problem of householdfood insecuritymore than that could be included in a health project or program.

1 Introduction

Figure 2. Conceptual framework of the causes of malnutrition and mortality

a and bMnlfddstuns

insufklentfowd Iineuv us

mm_ moh.andcarchild* andsanato causesue

Govenimntl andnangavenuetl IntiXtutons

Plitcal andIdeolgical superstkucur

Economcst=rcture

Source:UNICEF. * Impact on operationsdesign

When a health sector interventionin nutrition The first three sectionsof the next chapter are is designed,the desired goal is to provide an array closelyrelated, and the interventionsare classified of the most efficient services at each level of the acrording to the effort level and resources they while respecting a certain require to be integratedin most health services. chronological order: first, it is imperative to Benefitsalready providedby the health sector in provide the services that meet a nutritionare indicatedin section A. If a health demand-especially therapeutic care (for service would like to intensifyits efforts, and if it infections, malaria, dehydration) and has sufficient resources, it may undertake the vaccinations-before introducing services for activitiesindicated in section B. But to take on which the demand is not as evident, such as the activities described in section C, a service nutritionor famnilyplanning. must fundamentallymodify its current approach and be able to mobilizesubstantial resources. A

2 Impacton opcrxdonsdesign health service that is not in a position to meet feeding, dietary managementof infection through these requirements should refrain from changes in diet, micronutrients (iron, vitamin A, undertaking these activities, since the project and iodine). appraisals show that they will have no impact. It would be more prudent to use additional available SECTION C. Nutrition services that could be resources to extend the coverageof benefits listed proided by the health sector if a radical change in section A or to take on those described in in approach were adopted (moving from an section B.2 approach based on a health center to a comnunity-based approach but with the technical In sum: support of a health center), if the supplementary resources are available, and if the program is SECTION A. Benefits already provided by the correctly planned. If not, such activities shoid health sector, which have a positive effect on not be undertaken (including monitoring and nutrition should continue to be provided: promoting growth); instead, resources should be infection control, EN, prevention of malaria, concentrated on the first two types of benefits ORT, and . For this type of (sections A and B). benefit, efforts would be directed toward increasing coverage and quality. SECTION D. Nutrition education is included in most health benefits. It can be provided SECTION B. Nutrriionservices that could easily efficiently only if designed by communications be proWded by the health sector, which require professionalsand if the necessary resources for its few supplementaryresources: promotion of breast- implementationare made available.

2. As nutritioneducation (section D) is includedin most benefits, we will considerit separately;its requirements, however, resemble those of the benefits mentionedin sectionC.

3 II. THE HEALTH SECTOR AND NUTRITION INTERVENTIONS IN AFRICA fl Benefitsalready provided by the health sector'

In Africa, the majority of health services already contribute to preventionof malnutritionby 1. Prevention of diarrheal illnesses interveningdirectly to improve inadequatepublic health conditions,one of the inunediate causes of malnutrition; these interventionsand their impacL Diarrhea is probably one of the leadingcauses on nutritional status are briefly described in the of mortality in infants. For example (Tompkins following paragraphs. It should be noted that and Watson, 1989), in an urban community in preventive and therapeutic interventions modify The Gambia, more than 35 percent of deaths mothers' behavior and make them more receptive among children between the ages of zero and to nutrition and family planningproblems, as long three were attributed to diarrhea accompaniedby as they are given some assurances about their malnutrition. Diarrhea in a well-nourishedinfant children's health status. will not have the same impact as diarrhea in an infant who has practicallyno nutritionalreserves. In general, infections reduce food absorption This is one of the reasons that diarrhea is ot cause anorexia. The latter may be more or less responsiblefor more deaths in Africa than in the pronounced, and its consequencesare sometimes industrializedwerld. exacerbated by the feeding practices imposed by parents on a child with a fever and which vary Diarrhea also has a direct impact on nutrition considerablyamong cultures. since it affects intestinalabsorption (see Tomkins, 1981 for a complete overview of this subject). Furthermore, infectionsthat are accompanied While duringdiarrhea's acute phasethe absorption by fever increase a child's basic metabolic of macronutrientsremains at a high level (MolIa demands thereby increasing his calorie et al., 1983). this level mav well fall considerably requirements; they may also lead to appetite loss in the case of chronic diarrhea. Dehydration in or nutrient malabsorption. Gastrointestinal the case of acute diarrhea also leads to a drying infections occur most frequently among infants; out of the buccal mucous menbranes, which their principal symptom is diarrhea. Almost as makes absorption of solid food difficult. common are childhood illnesses (particularly measles) and parasitic illnesses (such as malaria The effort to prevent diarrheal illnessesshould and intestinalparasites). be continued and even intensified; the main elements of prevention are the following:

3. To a great extent this section is excerpted from A. Tomkins and F. Watson. October 1989.

4 Bencfltsalready provided by the healthseaor

* maintainingbreast-feeding for the first four to six months of life to decreasetho risk of 4. Control gl preventableclhldhood Il11es. diarrhlea; (immunizations)

* providing appropriatedietary treatment for diarrhealepisodes (see our suggestionsonthis a) Vc.cinatgRgd nstamaff az topic in ChapterIl.B.5); one of the principal elementsof dietarymanagement is continuous A child who has been immunizedagainst breast-feedingduring the diarrhea; measleswill have a smallerrisk of malnutrition thana childwho has had measles. The impactof * maintainingbreast-feeding beyond six this diseaseon nutritionalstatus appears: months of age with the progressive introductionof weaningfoods; and a In relationto the growthcurve: significant weight loss has been noted in West Africa * usingoral rehydrationimeasures to prevent (Morley,1969) in childrenwho have had measles, dehydration,which may have a favorable and this disease has aggravatedthe rates of impacton maintainingappetite. infectionobserved in Nigeria(Laditan and Reeds, 1976) in. children suffering from wasting or kwashiorkor. 2. Preventionof intestinalparasites * In relation to dietary intake, greatly depressedin cases of anorexia,dehydration, fever, Someparasites are associatedwith a loss of andbuccal lesions, local practices in certain places appetite,malabsorption, nutrient loss, and anemia. prohibitthe intakeof certainfoods in measles Althoughimproving sanitary conditions is the cases. It shouldbe notedthat the measlesvirus long-termsolution, vermifugal treatment programs may inflict sufficientmucosal injury to the for vulnerablepopulations are used in the short intestineto causemalabsorption and proteinloss term. (Dosseterand Whittle, 1975).

-______a * In relation to immunity,greatly lowered in 3. Prevention of malaria measles cases, which opens the door to other infections, especially diarrheal and respiratory, and contributesto a deficientnutrition cycle linked The impact of malaria on nutritionalstatus to measles. This depressed irmnunitycan persist varies according to the age and immunityof the for three or four months. child as well as the degree of infection. It has a significantimpact on newborns(low birth weight, b) VACCINATIONAGAINST WHOOPrNG iron and folic acid deficiency). The child's COUGH suppressedimmunity encourages the incidenceof other infections, which may in turn lead to In the same way, this vaccinationcontributes malnutrition. to improvingchildren's health by eliminatingthe symptomsassociated with this disease (anorexia, fever, pain, vomiting, and diarrhea).

5 The health sector and nutrtion intervenions in Africa

flaNutrition services that could easily be provided be the health sector

measles, the enrichment of appropriate foods if 1. Prevention of vitamin A deficiency they are available, improvementin the production of food rich in carotene, and nutrition education on an individualand communitylevel. a) Findingsand reconmmendaions i) Long-term solutions * Awarenessof the importance of vitamin A deficiency in Africa is recent. In 1964 only five * Programs based on increasing the countries were considered at risk; this number production and consumption of foods rich in grew to 11 in 1974 and has now reached 18. vitamin A are the only ones that have shown Following the Harare conference in 1988, the themselvesto be viable. Vitamin A distribution African Council of Science, Nutrition, and Food programs (by capsule) will only have long-lasting included this concem in its guide to food and effects in the case of the treatment of children at nutrition policy. Some countries have developed risk if they are integrated into other therapeutic a national vitamin A policy, the first stage in the care and MCH services, thereby making them rational development of any sort of long-lasting routine and efficient. system of treatmentand prevention. ii) Short-term solutions * Possibleinterventions are well codifiedand understood. The problem involves choosing * Childreniat risk should be treated at all among the different solutions available. The health centers; this impliessufficient staff taining optionsmay be revised in the near future based on and the ongoing availabiiiLy of vitamin A trials now underway. The United Nations and capsules. One solution that might facilitate this several NGOs are particularly knowledgeablein availability is to include vitaniin A in the vitamin A and may be useful advisors to country's list of essential . govermmentsmaking the - choices. One must be attentive, however, to the sometimes excessive * The principal stumblingblock t' capsule pressure brought to bear by some pharmnaceutical distribution plans is the difficulty in obtaining companies; programs should undertake a good coverage,particularly in the countries of fte preliminary systematicassessment of the situation Sahel where the scattered populationand the level to confirm the selection of proposed options of training of connunity health staff and workers (distributionof capsules or education). are significantconstraints.

a To ensure that a strategy covers the short-, x One alternativeis to link the distributionof medium-, and long term in a coord.natedfashion, vitamin A to the broad vaccinationprogram with programs need to include all of the following: which it shares a certain number of capsule distribution, infection prevention, characteristics:same targetpopulation, compatible at a high rate of coverage against time frame, a similar simpleand quick procedure,

6 Prevendonof vitaminA deficiency the necessity to provide the highest rate of result of its outstanding ability to stock the coverage possible, and suppliesfrom a central vitamin, the body may tolerate perf:ctly well . prolongedperiods of time without food intake containingvitamin A or its precuisors. Depleted * A published cost assessment is not vitaminA reservesmay also be restoredthrough available. Th-, raw materialsfor treatmentare lughdoses given at intervalsof severalmonths (up relativelyinexpensive: 4 cents per child treated to four to six months; after this period, the per year with vitamin A capsulesprovided by vitaminA status of children returns to its level UNICEF. The distributioncost, however, varies before vitamin A supplements were given) significantlydepending on the strategy adopted (Stolzfus and Habicht, 1993). Vitamin A (targeted or universal), and whether the deficiencyis generallyattnrbuted to insufficient distibution and education are conducted consumptionof the vitaminitself (retiaol from the separately, in associationwith vaccinations,or consumptionof animal products: liver, milk, withinthe normal functioningof the local health buuer, cheese, eggs, fish, etc.) or of its system. precursors(carotenoids contained in certainfruits, vegetables,lea*t greens, roots, and tubers). It b) VITA1IV A DEFICIENCYIA AFRICA may also be due to impaied intesdnalabsorption (diarrhea, parasites), imaired transit (protein- i) Consequences of vitamin A deficiency energy malnutrition), or to an increase in requrements(growth, infecions). Severevitamin A deficiencyin a youngchild leads to blindnessand increasedrisk of death. Vtamn A and morbidty-: Vitamin A deficiencyleads to a heightenedsensitivity to * Severe viamin A defidency and infection"(FAOIWHO, 1989). The impact on xerophthalmia: vitamin A acts as a visual vision, albeit dramatic, is the last manifestationof pigmentthat absorbslight. A lack of vitaminA the deficiency. Histologicalstudies have shown compromises the fonnation of rhodopsin, a that problems may first occur related to the nediatingpigment of nightvision. A methodthat increase in bony tissue; the development of is sensitive to detecting vitamin A deficiency immunity; or the development of epithelial consists, therefore,of measuringan individual's formations in the respiratory, urinary, or ability to adapt to darkness(FAO/WHO, 1989). gastrointestinaltract. Preventingmorbidity has When the deficiencyis ongoing,eye tissueswill becomea secondarygoal after the preventionof be affected to differing degrees, leading to blindness, the principal goal of programs to xerophdtalmia,and finally, if treat is not monitor vitamin A deficiency until recendy. providedquickdy, blindness. WHO believesthat Vitmin A deficiency,even at a marginallevel, vitamin A deficienryis a public health problem may make an illness become more severe that may be severe, moderate,or mild according (diarrheal and acute respiratory illnesses in to the criteriain TableI. particular).At the same time infectiousdiseases leadto an increasein vitanin A consumption,and Vitamin A is a fat-solublevitamin that is a deficiencymay appear in the case of marginal stored in the liver; it is redistnrbutedto peripheral scams: measles,diarrheal illnesses, meningitis, tissues as neededthrough the bloodstream. As a

7 The health sector and nutritioninterventiwv in Afica

Table I. Epidemiological criteria for evaluating the severity and extent of Vitamin A deficiency in a popudation Citeria Public health problems On chitdren I to 6years from a suraeyof 10.000 subjecs): Severe or moderate Low or ar risk Grade XN: Night blindness (hemeralopia) > 1%

Xerophthalmia:

* grade X1B (Bitot spots) > 0.5% one of thesesympboms/ * grades X21X3AlX3B (comeal > 0.01% signs present in the xerosis/ulneraionlkerotomalacia) communityat lower rates, or in hospitalized • grade XS (comeal scar) > 0.05% subjects

Rareof plosmaticretinosis (< 10 ig/dl) > 5%

Source:FAO/WHO. 1992. Preventig Specffc Micronutrnmjcs. ThemePaper No. 6. ICN, Rome.

malaria were 'ited (Inua et al., 1983; Sommer et A deficiency based on results from Indonesia: the at., 1987; Galan et al., 1990; ...). This is why risk of death seerned to increase significantly at children with measles, respiratory illnesses, and marginal and moderate deficiency levels (Sommer diarrhea are included in the risk groups. et al., 1983). A first supplementation trial with elejrted doses aimed at reducing xerophthalmia It should be mentioned, however, that this and blindness unexpectedly reduced by almost 30 issue is at the center of a debate among percent t'e mortality of young children. Since nutritionists because some studies (Rabuathulla et then, three additional trials of this sort have taken al., 1991, among others) did not demonstrate the place. Two trials conducted in southern India effects of vitamin A supplements on morbidity. (low weekly doses to approximate food intake adequate in vitamn A) and Nepal (high doses) In Africa Ghana's VAST study followed have shown spectacular results in terms of the 1,500 children from age six to 59 months for one reduction of mortality rates due to the year. One group received a supplement every treatment-54 percent and 30 percent respectively, four months while another was given a placebo. (Sonuner et al., 1986; Vijayaraghavan et al., The results demonstrated that supplementation had 1990; Rahmathullah et al., 1990; West et al., an impact on the severity of illness but not on its 1991). That the impact was specifically on prevalence: medical visits were reduced by 12 mortality and not morbidity in most cases has yet percent and hospital admissions by 38 percent in to be explained; this may be linked to vitamin A's the group receiving the vitamin A supplement. antioxidant nature, which enables it to avoid the harmful impact of an excess of free radicals * VitaminA and mortality: Since 1983 A. resulting from infection. Sommer has drawn attention to the potential for a more important-than-anticipated role for vitanin Preventionof vitaminA deficidny

The VAST study in Ghana (Fred Binka, * Breast-feedingwomen; WHO recommends Navrongo, to be published) recently showed a that pregnant women not receive vitamin A 20% decrease in mortality in children who supplements when massive supplementation at received supplements, compared with those who high doses occurs, since they may be teratogenetic did not. This result is probably explained in part early in pregnancy; it is recommended, however, by vitamin A's role in decreasing the severity of that mothers receive supplements, but no illness. longer than one month after delivery (one month is the period during which it is known that the One should not lose sight of the fact that such woman is not pregnant again). a reduction in mortality may be meaning'ul only if one effectively reaches the highest risk iii) Treatment and prevention individuals, which implies making an effort to provide almost complete coverage. It is likely The practical rules and modalities of treatment that in the current state in which a number of and preventionare abundantlydiscussed in various health servicesoperate, the impact would probably readily accessible recent works (WHO/UNICEF/ be nil (Costello, 1991). The associationwith EPI IVACG, 1988; Amndde-Manesmeet Demaeyer, proposed by WHO would be an adequate solution 1989; Storms and Quinley, 1988; etc.) There are at least initially in that this intervention in Africa five basic interventions in this area: urgent as elsewhere currently achieves one of the best medical treatment, supplementation, fortifying levels of coverage of the target population. food, inereasing production of foods rich in vitamin A, and nutrition education. ii) Risk groups Every child affected by xerophthalmia, * Young children: A child's needs are regardless of how far it has progressed, shouldbe proportionally greater than those of an adult and systematicallytreated with vitamin A. It may be vary radically as a function of his growth. administered orally or through an intramuscular Xerophthalmiais most prevalent generallyaround injection; these two methods were considered two to three years of age, at the weaning stage. equally effective following a trial conducted in This dependson the duration of breast-feedingand Indonesia, but the applicability of these results nutritional status. The deficiency may precede elsewhere, notably in Africa, is in dispute. birth if the mother herself is also very deficient; Testing appropriate to each country or region is in moderate deficiency cases, the child's needs advised. Doses are well codified according to age during pregnancy and lactation are favored. The and physical condition. A moderate deficiency in concentration in breast milk varies among vitamin A is thought to be exacerbated by countries and regions; on average, it is believed infectious diseases, including first and foremost that breast milk can meet the needs of the nursing measles, where elevated doses are strongly infant for six months. Children with measles, reconnended to avoid serious eye complications, acute respiratory illness, and diarrhea have a in particular in the case of associatedmalnutrition. heightened need for vitamin A (see above The use of high doses during diarrheal episodes or paragraph. "Vitamin A and morbidity"). pneumonia is less well established; toxicity is a risk, and the use of one dose is recommended.

9 7he health sector and nutntion interventions in Africa

Vitamin A deficiency disappeared over the Mozambique,Niger, Nigeria (north), Uganda, and course of time in the industrializedcountries as a Chad" (FAOJWHO, 1989). It is clear that These result of a natural trend toward varying dietary countries are not all aXfectedto the same degree in intake; it was definitively eradicated with the regional and seasonal terms (periods of drought) systematic supplementationof certain foods such and that interventions should not necessarily be as milk, margarine, and foods for special dietary the same throughout the region. In Nigeria recent use for children (Amntde-Manesmeet Demaeyer, observations have suggested that vitamin A 1989). There are currently three ways to deficiency might be a problem not only in the encourage this evolution in the nonindustrialized Sudaneseregions, the driest in the north, but also countries still facing this problem. A short-term in the more humid regions in the south. Although approach would be for health services to palm oil, fruit, and vegetables are more abundant administer periodically capsules containig high there, their consumptionby young children, as in doses of vitamin A, the effect of which continues Asia, may be limited as a result of local cultural for several months due to the body's ability to practices regarding weaning, in particular related store it in the liver. In the medium term, certain to low-fat foods (Sight and Life, 1990). foods that are widely consumedby populations at risk of vitamin A deficiency could be fortified. Considering that the prevalence of cases of Only a program to develop production of foods marginal, moderate, or severe deficiency affects rich in vitamin A and to educate all levels of on average about 15 percent of the taget society to encourage their consumption of these population of preschool-age children, the health foods can resolve the problems associated with services throughout the continent will handle vitamin A deficiency in the long run. approximately eight million children (Sight and Life Newsletter, 1990). Reliable data on the iv) Distribution in Africa actual extent of vitamin A deficiency are still lackingin most African countries; there is a need "In Africa, most cases of blindness in young for prevalence and simple monitoring surveys like children are attributed to measles, but it is very those conducted in TANZANIA (Foster et al., likely that measles only precipitates an acute and 1986; Pepping et al., 1988). severe xerophthalmiain children whose vitamin A intake is marginal. Countries in which a lack of c) LESSONS LEARNED FROM TrE vitamin A and xerophthalmia are considered a ASSESSMENTS significant public health problem are found in the Sahel and sub-Sahel regions (Benin, BurkinaFaso. i) Prograimsset up in Africa Mali, Mauritania) and in eastem and southern Africa (Ethiopia, Kenya, Malawi, Sudan, The first essential step is to make a concerted Tanzania, Zambia, ... ). Moreover, in a certain effort to resolve the pmblem for the long term, number of countries for which there is no direct which means improving the production and evidence based on official observations, numerous consumptionof foods rich in carotene and vitaniin indications suggest that the lack of vitamin A is a A. fortifying food, and incorporating distribution significant public health problem. In this category of vitamin A to groups at risk into therapeutic are the countries of Angola, Ghana (north),

10 Preventionof WtaminA defidency consultations. For the short term, vitamin A calorie malnutrition) when they come to a health distributionprograms may be considered. center (hospitalsand dispensaries). An alternative proposed by WHO is to associate the Few program assessment results in Africa are administration of vitamin A with the vaccination available. Currently, results concerning the of young children and breast-feeding mothers treatment of individualswith measles and the first (during one month following childbirth). It is treatment and prevention programs in Niger and essential to record carefully which individuals BurkdnaFaso, which are representative of the receive high-dose supplements to avoid giving Sahel countries, are available as are those from them several doses close together as a result of the Malawi in eastern Africa. different programs or the combinationof different strategies. With the help of recently perfected Published cost-effectivenessstudies are not yet dose devices (atomizers), one can administer low available; the models evaluated in the Asian doses orally on a regular basis; the distribution programs (Tielsch and West, 1990) have not yet can then be carried out in the context of programs been applied in Africa. that provide more frequent contact with the children: growth monitoring, preventingdiarrhea, ii) Distribution of vitamin A capsules nutrition education sessions. This strategy has also proven to be effective (Underwood, 1990). This intervention has proven its effectiveness The problems posed by this approach concern in good testing conditions. The capsules are coverage and contance. mainly provided by UNICEF, at a cost of two U Assessing operationsin a Sahel country: In cents per capsule. Generally, the intervention Niger the program began in 1987 in four districts consists of a single high dose (200,000 units) distributed at an interval of several months according to three different or combined Doses currently reconmended by WHO strategies. (FAO/WHO, 1992) are the following: The "universal" strategy is directed at the entre target population, generally children between the or prevetion ages of one and five; it is implemented every child 6 to 23 1 dose of 100.000Ut oally every months: 3 to 6 mondhs threeto six months by specific personnel (vertical child2 to 6 years: I doseof 20.000 IU orallyevery programs) or by health center personnel 3 to 6 months (horizontal programs). The other strategy, known breast-feeding I dose of 200.000 IU oniceduring as the "target" strategy, consists of treating woman: te 4 weeks following childbirth children systematically in zones or comunnities Fortreatenr ]known to be particulary deficient; this method, childover one yearwith signs of deficiencyor measles: which is less costly, enables one to reach specifically the individuals at -at diapois 200.000 IU orally highest risk. One could also target chlfdren at the one daylater. 200.000IJ orally most vulnerable ages (one to three years). 4 weekslatr: 200.000I orally Finally, the "medical" strategy consists of treating all children at risk (hemeralopia, measles, for childrenunder one year, decreasedosage by half respiratory and diarrheal infections, protein-

11 Thehealth sector and nurition interventionsin Africa

of the Tahoua and Maradi departments with the the distance to the health centers nor to the size of aim of preventing nutrition-linkedblindnessand of the villages. The system set up routinely by the treating xerophthalmiathrough the distribution of ministry of health at die health centers or taught to megadoses of vitamin A. It was launched in a the village health workers to evaluate the number coordinated effort by the health ministry, the of capsules distributed proved effective. It was national program for the prevention of blindness, not possible to evaluate the impact on the and HKI (Helen Keller International). The first prevalence of blindness or xerophthalmia since phase consisted of operational research to they were not measured at the beginning of the determine the best way to distribute the vitamin A project. capsules for treatment and prevention. The treatment consisted of distributing three capsules Several problems were identified related to to children with signs of night blindness, capsule distribution. The capsules are provided xerophthalmia, measles, and severe malnutrition regularly to the , and the registry of orders and one capsule to children with chronic diarrhea. and deliveries is largely in order. Capsule Prevention consisted of distributing one capsule delivery to health workers, however, was every six months to children under six years and inconsistent (inadequate in 50 percent of the to women who had just given birth (half a dose cases), and the records were incorrect These for children under one). workers must be supplied with capsules by the rather than by those running the project. Two strategies were tested in two strictly They also need more training in keeping records. comparable regions: treatment and prevention were provided by medical centers and rural Professionalhealth workers have a good level dispensaries in one region (target population of knowledge after appropriate training, but their estimated at 64,000 people), and prevention was turnover is high and in the absence of periodic overseen by village health workers in the second training, the level of trained individualsfalls again region (target population of 54,000 people). The quickly. The village workers are more stable, but two strategies were then compared (Cohen, 1989; their training is too theoretical, and it does not Helen Keller International, 1989). allow them to encounter all the practical problems with which they will be faced; they need shorter For treatment, rates of coverage increased but more frequent sessions that meet their quickly and remained high throughout the first expectations. year of observation, between 50 percent and 100 percent. As for prevention. the rate of coverage This project also included a training fell in one year from 19.9 percent to 14.1 percent component on the national level for health in children and from 20.2 percent to 17.8 percent professionals, distribution of appropriate in mothers in the regions covered by the health educational materials, and intense publicity to centers. It increased from 22.8 percent to 60.9 improve consumptionof foods rich in vitamin A. percent for children and from 27.7 percent to 55 Knowledge of vitamin A deficiency among the percent in mothers in the region where the public improved considerably. Health workers distribution was the responsibilityof village health did not relay the recommendations to improve workers. The rate of coverage was not linked to production and consumption of foods rich in

I? Preventionof vitaminA deficiency

vitamin A, however, and this risks reinforcingthe four months (450,000 children and 240,000 notion of the necessity of medical intervention nursing mothers), therapeutic treatment programs while downplaying the paramount importance of at health centers for the most high risk children, the role of food intake. and educational programs in schools and preschools were set up. In the opinion of one of the evaluators (Cohen, personal communication), this project is Preliminary comparisons of the rates of a good exnampleof the possible d fTicultiesand xerophthalmia at the end of 18 months between success of a program to prevent vitamin A the project zone and the zone that was not covered deficiency in a country in the Sahel region. The demonstrated a significant impact on coverage attained by the village health workers is xerophthalmia,which fell to a rate lower than the significant, but it represents the most that can danger level defined by WHO. Night blindness currently be done given the constraints of the reportedly fell from 2.22 percent in 1986 to 0.39 terrain and functioning of the health systems; in percent in 1989. Coverage rates for young fact, coverage is still inadequate. It could be children (between one and 10 years of age who increased for women who have just given birth by had received one capsule during the previous six having older women (traditional midwives) months) were on average 55.6 percent in the distribute the capsules. For the moment, regions covered, which is considered satisfactory however, extending the project nationally is still given local logistics and the difficulties difficult to foresee; vitamin A is on the list of encountered (Helen Keller International, Burkina essential medicines in Niger, but bottlenecks Faso, 1989). appeared in the distribution system to which the individuals running the projects should be * Assessing treatmea in children with attentive. In fact, the project has often had to measles: The association between measles and appeal to independent suppliers to function. xerophthalmia has been described frequently. Furthermore, this activityis poorly integrated into Measles increases vitamin A requirements, while other activities in the health system, which makes the consumption, absorption, and transit of the its cost high (figures not available) and its vitamin is diminished. Measles may result in viability uncertain over the long term. The damage to the cornea. In general, the incidence alternative is to link distributionof capsules to the of corneal damage foUowingmeasles infection in broader vaccinationprogram in which coverage is Africa is estimated to reach 4 percent (editorial, advancing significantly. Lancet, 1987). Foster and Somrner (1986) demonstratedwith surveys conducted in schools in * Assessing imnpac in a Sahel country: In various African countries (Ethiopia, Malawi, Burkina Faso, HKI launched the largest pilot Nigeria, and Tanzania) that half of blind children program in West Africa in 1988. Following a had measles shortly before they became blind. prevalence survey conducted in 1986 (in the According to many clinicians, these cases were a Yatenga, Passore, and Sourouet Provinces) which direct result of measles, local treatment, or a indicated that vitamin A deficiency constituted a concomitant herpes simplex infection; according public health problem in these regions, training to others, the process was linked to or exacerbated programs (1,800 trainers and health workers), by a latent vitamin A deficiency (Inua and Ross, distnrbution programs to provide capsules every 1983; Markowitz et al., 1989).

13 The heatkhsector and nutritioninterventions in Afrca

In Tanzaia IfO children with measles iii) Fortfyingfood received two doses of 200,000 units over two days or a placebo: the subsequent mortality was seven This implies pinpointing with the appropriate percent compared with 13 percent in the reference survey the basic foods consumed by the entire group (Barclay et aL, 1987). The difference was population at risk. It is necessary to find one food significant, however, only for children under two among them that meets the following years (2 percent versus 17 percent). This led specifications: it may be fortified at a central WHO and UNICEF to recommend systematic location during the course of one stage in its treatment of children with measles with vitamin A processing; it must be possible to conserve in regions where the risk of death for this group vitamin A in a stable form in the food without is higher than I percent. This decision was changing its flavor or appearance; the enriched criticized to some extent at the time, but other food must remain reasonably priced and be easily evidence has been provided since then. The accessible to the poorest sectors of the population. experience was repeated in South Africa with a The principal difficulty to date has been to find double blind trial (Hussey and Klein, 1990); 189 the proper medium: monosodium glutamate was children hospitalized for complicationsassociated used in Indonesia and sugar in Central America; with measles also received 400,000 units of tea was proposed in Pakistan, wheat in vitamin A or a placebo. Children who received Bangladesh, and rice in the Philippines the treatment recovered more quickly and were (Underwood, 1990). This type of intervention hospitalized for a shorter period of time; their requires the collaboration of numerous agencies, mortality rates fell by half compared with the and the benefits are evident only after several reference group. In this region, however, in years. It should be supported by intensive contrast to Tanzania, severe vitamin A deficiency nutritional education for populations at risk to is rare. A new study conducted in Durban in the popularize the consumptionof the fortified foods. case of measles with respiratory infection Although to a great extent the health sector has complications demonstrated a clear improvement failed to grasp the process, health services should in morbidity scores after vitamin A was be involved from start to finish: identifring the administered (in accordance with dosages problem and its importance, choosingthis solution recommended by WHO), with the role of the as it relates to other steps already taken or to be different covariants having been carefully taken, participating in the choice of food to be controlled in the analysis (Coutsoudis et al., enriched and the various fortification trials, and 1991). educating the population. It should be noted, however, that although this solution was used The best way to prevent the adverse conse- successfully in the industrialized countries, quences of measles is still vaccination- For those adopting it definitively and without interruption who contract the disease, vitamin A treatment has has not yet been possible in a country where been shown to avoid some of these consequences. vitamin A deficiency is currently significant. One of the best foods to be fortified is undoubtedly breast milk: if the nursing woman absorbs some vitamin A. it enters her milk and benefits the nursing infant.

14 Prevntion of vitamin A deflcency

Forrifiedfoods in Africa: The list of foods programs are not specifically oriented toward commercially vitamin A fortified in various production of food rich in vitamin A but rather African countries is rather limited, as is their are part of agricultural development programs accessibility. An exampleis the"Maggi cube' by with broader goals. A recent table of the Nestld (1400 RU/cube weighing 4 grams), compositionof African foods gives the vitamin A manufactured in Abidjan and distributed content of the most common of these (West et al., throughout West Africa but mainly in the large 1988). cities; margarine manufacturedby Unilever also in Abidjan; cookies manufactured by VAP in Zaire Following establishment of a United Nations and Henry's in Morocco (Sight and Life, 1989). assistance plan for countries seeling to prevent Informationis lacking on the different foods used vitamin A deficiency, the PAO has been for weanLingthroughout the continent; in theory, particularly concerned with this problem. Until there are recommended international standards then, no such wide-ranging program had ever (Codex Alimentarius, CACIRS 72, 1976). Last, been set up in Africa. But Benin, Burkina Faso, foods used by the World Food Program (WFP) Chad, Maawi, Mali, Mauritaia, Niger, were fortified with vitamin A (powdered milk, Tanzania, and Zambia have formulated projects corn-soy-milk flour, cheese, and oil). The of this sort since 1988, most of which started in specific impact on vitamin A deficiency was not 1991 (PAO activity report, 1991). We do not yet addressed by the studies. have an assessment of these projects. The FAO also tried to set up regional networks that grouped iv) Improving theproduction offood rich in the responsible health and agriculture officials in carotenoids and vitmin A East and West Africa along the lines of what was already developedin Asia. The availability of foods containingretinol is extremely poor in some African countries, Most of these projects involve an increase in particularly in the Sabel: about 180 micrograms production of fruits and vegetables rich in (meg) is available on average per person on a carotene in family, connnunity, and school daily basis in Buridna Faso, while the mininmm gardens and an increase in their consumption requirement is 250 mcg for a young child and thanks to the education of school children and the more than 700 mcg for an adult (FAO food public in nutrition matters. In Tanzania, faced balance sheets, 1980). Livestockand fish are the with a high demand for palm oil (Kavishe, 1987, principal sources of vitamin A, with agriculture cited in West and Sommer, 1987), a specific and fruit harvests furnishing foods rich in beta- projectdealing with productionandtransformation carotene. Although previously quite common of this oil was set up. It is still too early to assess along the banks of the Senega river, vitamin A the project's results. deficiency has decreased considerably as a result of a policy to plant mango trees and distribute In Niger(Ministry of Healtb/AED/HK 1992) mangoes to markets. The other step generally a small social marketing project covering 26,000 taken is to develop small gardening. The people in the Tahoua department resulted in the constraints are the same as those for all design of six messages in the form of plays to agricultural activity in the country: access to land improve consumptionof food rich in vitamin A. and the availability of water. Most of the These messages were targeted at different

15 T'hehealth sector and nutritionintervemions in Africa

segments of the public: pregnant and nursing 1987 with the assistance of USAID in two women, their husbands, gardeners, community regions, one in the center (32,000 inhabitants)and leaders, and development workers. The the other in the north (40,000 inhabitants). assessmentdemonstrated that about 90% of adults Distribution of capsules to children at risk and who had seen the plays consumed more liver, and every six months to all village children with the 60% gave more liver to their children. Likewise, highest rates of xerophthalmia was set up (six 89.6% indicated that they were eating more leafy months-six years); mothers were not included in greens, and 60% said they gave more of them to the project. Family educational activities via their children. (About half of the children under women's groups and school children and the six years either do not eat leafy greens or eat them establishment of bmily gardens were also less than once a week; some parents think that a initiated. The project was integrated into a child between the ages of zero and 12 months is community development project in the two not yet old enough to eat them.) Coverage was regions. better for men than women: 61 % of men and 34% of women said they had seen at least one Three years later the program was assessed play. (Save the Children, USA, 1990). The first problem was the absence of a precise formulation v) Nurridon educadon of the objectivesand under what conditions they would be realized; this is what probably led to the Every activity aimed at reducing vitamin A uneven implementationof different parts of the deficiency (or promoting the maintenance of a project. It resulted in particular in confusion good level of vitamin A) should be accompanied about who should be responsible for the project, by education appropriate to the target groups, who should be registered and how, and who either from within the community or through the should collect and analyze the data. The capsule media. The purpose is to teach families about the distribution program was disappointing. Not dangers of vitamin A deficiency, foods that may enough capsules were provided, and in view of be consumed to prevent it, and the usefulness of the low coverage, it would have been better to monitoring in the mass capsule distribution target the highest risk villages more. In this type campaigns. This education initiative should of project, it is necessary to set up a well- always be preceded or accompaniedby adequate identified supply network. It was not possible to training for health personnel who are often still provide the three doses in a rigorous fashion ignorant of the dangers of vitamin A deficiency. because different people were involved, and no In the long run, it should make clear to all who appropriate form existed that could have allowed work in development that health initiatives are some follow-up. This was the case even though only palliative and that the solution is to be found the registration was recorded each time on the in the production and consumptionof appropriate growth form kept by the mothers and on a special foods. form kept on site at the project. There was a high rate of participation in the family education integrated program assessinent in eastern activities conducted by the women's groups, but Africa: In Malawi. the Promotion of Nutritionin the content was focused more on home economics the Family and the Community project began in and vegetablegardens while neglectingtopics such

16 Preventionof vitaminA deficiency as nutrition education itself, feeding a young provided for the training of medical ophthalmic child, identifying malnutrition, and using the assistants, health monitoringassistants, and village health and nutrition rehabilitation centers. The health workers in the vitamin A area. An women's groups achieved very encouraging extension to create vegetable gardens is underway results in the establishment of vegetable gardens, (Vitamin A News Notes, spring 1991). but evaluating them in quantitative terms was difficult because the number of gardens was not In Mali, in the context of an integrated health originally inventoried. A strategy to create a seed project with the participationof the SCF, a survey fund to encourage these gardens was short-lived on the prevalence of vitamin A deficiency and the due to the community's lack of a involvement in consumptionof food containing this vitamin was the process. Althoughthe school gardensworked, conducted in 1990 (Kolondi6bacircle). Twenty- they ran into problems due to a lack of water and four foods rich in vitamin A were identified, the seed, and a significantpart of the harvest was sold market prices of which were compatible with by the school teachers householdpurchasing power. The risk of vitamin A deficiency, clearly identified locally, is linked The project was successfu! at promoting to the late introduction of weaning foods. production of agriculturalproducts rich in vitamin Following this survey, the "medical" strategy for A, particularly during the dry season, and some children at risk (infectious disease, malnutrition) consumption; the goal is sustainability involving and training of health personnel to promote an groups active in the community(women's groups, adaptedsupplement earlier was proposed (Vitamin developmentassistants) and the broad coordination A News Notes, spring 1991). Training activities of different agricultural and health services, were also undertaken in the two other health despite a lack of training, supervision, and follow projects administered by the NGOs (CARE and through at the highest levels (district and Africare) that incorporateda vitamin A component national). A targeted educational campaign with in Macina and Dioro. A social marketing strategy incentives to consume foods rich in vitamin A is being elaborated for food rich in vitamin A. to directed toward the individuals with the highest stimulatecommunity interest in the problem. risk-pre-school-age children and pregnant and nursing women-was lacking. Also, no cost The results of the survey conductedwithin the assessment was performed in this study. framework of the Macina project (Vitamin A News Notes, winter 1990) demonstrate that in this A program was implemented in another region improvingthe consumptionof foods rich in region of th.s country (IEF, lower Shire valley) vitamin A is possible under the following that included the distributionof capsules every six conditions: sauces may easily be fortified in months to children and their mothers. It was vitamin A by increasing the quantity of green supported by women volunteers chosen by the leaves, liver, or-squashused; it is necessary to see community, each of whom was responsible for to it that children eat this sauce regularly. about 50 families; this program was at the halfway Although the results of education based on these point in 1991. It has achieved coverage of 94 principles have not yet been evaluated, it already percent of children between six months and six seems that the level of acceptance is high, which years and 88 percent of mothers (during a two- was not the case with the medical measures that month postnatal period). The project also were initially stymied by the population's

17 T7hehealth sector and naurifon intervenrionsin Africa skepticism about their effectiveness. Tests con- apply their accumulated experience to decision ducted on market days based on the purchase of making. food rich in vitamin A were designedto reinforce the practical application of nutrition education * Whether a program is sustainable depends messages. initially on resolving the problems of coverage of a target population and securing an adequate d) coS supply of iron tablets. Providing iron tablets should be made part of the national program to According to Horton (1992, p. ZB), the cost supply essential medicines and should be of fortifying sugar with vitamin A in Guatemala rigorously managed. was 14 cents/person for one year of protection (U.S. 1987 dollars), while the gdistributionof a Another problem that affects a program's capsules cost between five cents and 68 cents per effectiveness is patient compliance with the full person for a year of protection (in Bangladeshand treatment program. Providing health workers Haiti). with informationand communicationstraining and broadly informing communityleaders in this area The capsules purchased by UNICEF cost 2 should contribute to improving compliance. cents per unit (1993). U There is more than one solution to the problem of iron deficierncy;a variety of resources 2. Prevention of iron deficiency should be used in combination to resolve the problem, beginning witi providing iron supplements to risk groups, followed by a) FINDINGSAND RECOMMEDATIONS preventing, infections and parasites, educating people to increase their intake of iron-containing * Although iron deficiency is a widespread foods, and fortifying basic foods with iron or health problem in Africa, preventive measures are increasing iron's bio-availability to provide a rarely implemented on a national level. Most definitive long-term solution to the problem. African countries provide pregnant women with iron supplements, but coverage of the population * In the short term, countries shculd set up at risk is generally inadequate. In addition, oral supplementation programs for pregnant measures to prevent malaria and parasitic diseases women, the most important risk group; in most are rarely implemented. cases, a multietiologicaltreatment program should be provided at the same time (including * The first obstacle to setting up such antimalarial and anthelminthic measures, folate programs is the decision makers' lack of supplementation,etc.). information concerning the relatiDnshipbetween the expected benefit of a program and its a Success depends on providing personnel relatively low cost. This problem a.ffects many with a clear plan of action which includes institutions. but few federative bodies are able to treatment and prevention criteria, dosage levels, referral criteria, criteria for maintaining a good

IR Preention of irondefidency level of compliance, supervisorymeasures to marrow. The body uses these sourcesonly as a reducethe risk of overdose,etc. reserveto replacephysiological losses.

I A small-scale pilot test under local ii) Iron absorpdlon conditions is advised to ensure that the recommended measures are effective before The absorptionof iron from food sources applyingthem on a largerscale. The transitionto dependson the typeof food and meals as well as a large-scale program, however, should be reserve levels. Diets based on cereals or roots monitored closely, and its impact should be and tubersand that includea negligibleamount of evaluatedon a regularbasis. meat, fish, or foodsrich in ascorbicacid have a high level of nonheminiciron, but with a low * In the case of young children, when degreeof bio-availability(absorption on the order supplementationis plannedonly for regionswhere of 5 percent or less) (FAOAWHO,1989). In anemia is very prevalent, interventionwill be theory, the Africandiet overallis rich enoughin limited clsewhereto encouragingbreast-feeding iron to meetdictary requirements. In practice,the for as long -.r possibleand favoringthe intakeof iron has a low levelof bio-availability,since the ascorbic acid on a regular basis through the diet is poor in foods of animalorigin except for consumptionof fresh fruitsand vegetables,which hunters, fishermen,or people who consumean promotesthe absorptionof iron. The problemof abundantamount of vitaminC (Hercberget al., iron deficiencyshould be addressed in every 1987; Fleming, 1989). It is not unexpected, educationalprogram dcaling with child feeding therefore,that iron deficiencyis widespread. practices. Thebio-availability of severalfood dishes was * Iron deficiencyis a structuralproblem measuredrecently in Zaire, Benin, and Senegal given the type of food consumedin most African (Galanet al., 1990;Galan et al., 1991;Guiro et countries. Whilesupplementing pregnant women al., 1991); it varied from 1.1 percent to 6.6 with iron maybe characterizedas the mosturgent percent. The longcookLng Limes usually destroy step to be taken, solutionsthat addressthe core of the ascorbicacid content. Addingascorbic acid to the problem should also be sought, stch as the dieton a regularbasis alonecan increaseiron fortifyingone or more foods and implementing absorptionto about 10 percent. anthelminthicprogra-ms. Given the benefits expectedfor childrenin terns of their intellectual Althoughthe average iron contentof breast developmentand adultsin ters of theirphysical milk is rather low, bio-availabilityis very high ability, these measures should be a priority in (50 percent); breast-feedingthus protects the cotntries equippedfor this type of intervention. n.ursing infant against developing an iron deficiencyin the first four to six months of life b) IRON DEFICIENCY IN AFRICA (FAO/WHO, 1989). The bio-availability of iron in cow's milk is lower: it rangesfrom 10 percent i) Consequencesof iron deficiency to 20 percent.

Iron is foundin two areas of the body: in the a In adults, iron deficiency leads to a blood and stored in the liver, spleen, and bone diminishedability to work and lowerproductivity,

19 Thehealth sector and nuition interventionsin Afrca even when anemia is not present. The deficiency numberof studies have shown that iron deficiency usually goes undetected, although the individual at a young age results in ohanges in motor may feel a certain fatigue. Studies have development, learning, and behavior (inattention, demonstrated the impact of supplementationand fatigue) in addition to anemia (see Scrimshaw, its excellent results in terms of cost effectiveness 1991). (Scrimshaw, 1991). There are no references availablefor Africa in particular, and there is little iii) The prevalence of awenda and iron data concerningthe prevalence of iron deficiency deficiency in Africa in men (see Benefice et al., 1980, for data on anemia). The only data available from health centers concern the prevalence of clinical anemia (not * In pregnant women, anemia is known to very reliable) or laboratory-determinedanemia cause death or complications during delivery (according to thresholds arbitrarily assigned by (prematurity, placenta hypertrophy) and to result WHO). Etiologicalstudies conducted to date have in a higher incidenceof low birth weight; its exact shown that anemia frequently has multiple causes role in Africa, however, cannot be assessed. A in Africa, among them malaria, bilharzia or pregnant woman may develop anemia during ancylostomiasis, iron and/or folate deficiency, a pregnancy even if she is not anemic at the hereditary hemoglobin production anomaly beginning; conditions during labor and delivery (drepanocytemiaor thalassemia), protein-energy may also have a significant impact on matemal malnutrition,and - even more recently - AIDS anemia (postpartum henorrhage) (Maire et al., (Hercberget al., 1987; Fleming, 1989). Dakar, 1982). A premature newborn will become iron deficient more quickly. Although the Publislheddata were reviewed by DeMaeyer newborn does not need iron for the first six and Adiels-Tegznan (1985), Hercberg et al. months of life, there is a link between the (1987), and WHO (1989): in 1980 in Africa, mother's iron reserves and those of the infant approximately165 million people of a population (Ajayi, 1988. for example), and an infant bom to of 405 million could be considered anernic. a deficient mother will need iron supplements Prevalence varied from 3 percent to 100percent, before the age of six months. For this reason, depending on the region, age, and sex, with pregnant women as a whole are considered a averages on the order of 27 percent to 60 percent. priority group. Young womienwho have not yet The data should not be considered representative, reached maturity who are giving birth for the first and the actual prevalence of iron deficiency, time are considered particularly at risk. which is generally based on the prevalence of anemia, is certainly underestimated. * In children: During the first year, a child triples his body weight and doubles his iron Representatives of 10 countries met in reserves. The reserves present at birth are thus September1989 at an African regionalmeeting to completely mobilized for growth at 4-6 months. discuss the prevalence and causes of anemia in After six months, the child needs to increase his Africa and to propose a course of action, iron intake, but the cereal-basedporridge provided particularly for pregnant women (WHO, 1989), at that age is rather low in bio-availableiron. A with the express purpose of "eliminating severe

20 Preventionof iron deficency

Table II. Epidemiological criteria for evaluating the severity of anemia in a population Prevalence in the population by percentage such that the problem is considered:

Criteria Severe Moderate Light

Moderate anemi or henatocrit >40% 10-39% 1-9% <33% Severe anemia (Hb <7 g/dl) >10% 1-9% 0.1-0.9% Serum ferritin (mcg/i) < 12 < 12 a. Hb <11 gIdi pregnant woman 11 gidl young child 12 g/dl school-agechild or adult woman 13 g/dI man Source:FAOIWHO. 1992. PreveaiUigSpecific Micronutrent Deficiencies.Theme Paper No. 6. ICN, Rome.

anemia by the year 2000 and reducing the this step should be limited to hospital-type prevalence of moderate anemia." Their report institutions. Several teams have shown that an illustrates the extent to which data are lacking for intramuscular injection given once or in several the continent; Fleming believes that in addition to doses (Ogunbode et al. in Nigeria, 1980; Kaisi in the nmmbers resulting from isolated measures, Tanzania in 623 pregnant women, cited in WHO, which vary a great deal, the prevalence of anemnia 1989) is completely effective and well tolerated. is probably on the order otf 50 percent to 60 Tenkinson in Zambia (1984) compared giving a percent for pregnant women as a whole. Most single injection of iron dextran at the begining of African countries provide an iron supplement to pregnancy (133 women) to giving iron sulfate pregnant women, but coverage of the population supplements orally (142 controls): hemoglobin at risk is generally low; additional measures to was elevated in 84 percent of the cases in the first prevent malaria and parasitic diseases are not group compared to 56 percent in the control always implemented. No assessment of the group. Due to the risk of AIDS, however, the current programs is available. tendency is to limit injections in areas where cost considerations make guaranteeing the usage of c) LQSSONS DEAD MMTR disposable syringes difnfcult (DeMaeyer et al., ASSESSMENTS 1989).

i) Trials on emergency treatment ii) Trials on supplementation effectiveness

In cases of emergency (i.e., pregnant women Direct iron supplementation through who schedule prenatal visits late in their medication has the benefit of leading to a very pregnancy, who demonstrate a lack of compliance rapid correction in iron deficiency, and it may be with oral treatment programs, or who are severely used easily to target risk groups. It is also anemic), treatment by injection may be initiated; administered in a very systematic fashion: via

21 ITheheakh seaor and nutftion interventionsin Africa ferrous sulfate tablets (possibly in conjunction and hemoglobinlevels. Brunengoet al. (1988) in with folic acid) given to pregnant women, older Niger measured the iron reserve levels of children, and adult men (or a dextran iron pregnantand nursing womenbefore and after they injection in the most severe cases) and a liquid were treated with supplementsof moderate levels preparationfor very young children(DeMaeyer et of iron given orally (400 mg of fumarate per day) aL, 1989; WHO, 1989; ACC/SCN, 1990). for one month. The level of iron deficiency in the group went from 22 percent to 8 percent, but only U In pregnant women: A number of half the women agreed to participate in the study controlled trials were conducted in Africa to throughits completion. These resultsillustrate the verify the effecdtveness of various proposed extent of iron deficiency among the populationof treatments: the Sahsi and the need for a systematictreatment program. They also underscore the difficulty of In Liberia, Jackson and Latiam (1982) successfullyimplenenting such a program. provided oral iron supplementsof varying dosage levels to 621 pregnant women during their third In northem Nigeria, Fleming et al. (1986) trimesters (for 12 weeks) (1 x 60 mg or 3 x 60 compared the respective effects of different mg per day). In every case, the treatment programs that included iron and/or folate anti- moderately increased hemoglobin levels and malaria! treatments with a group receiving no decreased the prevalence of anemia from 78 treatment. They concluded that pregnant women percent to 45 percent. In Nigeria, Ogunbode et in the Sudan and Sahel regions should be al. (1980) also showed that a 40 mg dose of iron systematicallytreated with antimaaia drugs and (200 mg of ferrous sulfate) given orally on a daily an iron-folatemixture to reduce the risk of anemia basis seemed to be sufficient to prevent iron during pregnancy, preterm delivery, and low iron deficiencyduring pregnancy. Higher doses do not -and folate stores in the nursing infant. provoke a more rapid hematologicalresponse and Approximately 70 percent of pregnant women may lead to gastrointestinalproblems. Fleming with anemia responded to a combined treatment (1987) believes, however, that it is advisable to program of iron and folate supplements and double or triple these doses if one wants to antimalarial drugs in Nigeria (Fleming, 1989). replenish substantially iron reserve levels, as in According to Fleming, as a result of resistance the case of northem Nigeria, for example. problems in a number of African countries, the aporopriate preventive measure is to use Hercberg (personal communication)recently proguanil. supervised a supplementation trial of pregnant women (in their sixth month) in Niger in Women's compliance with long-term accordance with international guidelines. The treatmentis an essentialcomponent of the success number of women with anemia and with of a treatment and prevention program: 55 insufficient iron reserves fell. The improvement percent of women in the study of Brunengoet al. was still detectable three months after delivery in (1988) and 64 percent in the study of FRerninget this group when compared with a control group al. (1986), for example, agreeu to continue with receiving no supplementation. The newboms in the treatmentuntil it was completed. Nyazemain both groups, however, had identical iron reserve Zimbabwe (1984) demonstrated that this is

22 Preventionof irondeflciency

fundamentalto most long-termmedical treatnents: not demonstrate any variations in parasitemia or patients need to know something about their specificantibodies compared with a control group illness, why and how they are being treated, what during the three-month trial or the six-month to do if treatment is momentarilyinterrupted, etc. period following it (Chippaux et al., 1991). It is essential that health workers intervene to There seem to be no contraindications to advise them throughout the treatment process. supplementing young dhildren in the entire endemic malaria zone. Smith et al. (1989b), * In young children: Few iron supple- however,observed the oppositeeffect in Gambia. mentation trials have been conducted in young In their view, it is still prudent to administer an children in Africa. Smith et al. (1989a) treated antimalarial treatment in conjunction with iron anemic young children in Gambia in 13 villages treatment. duringthe rainy season when malaria transmission is at its highest rate. The mothers were asked to * In school-ge children: Older children no give 3-6 mg/kg of iron per day to each child in longer constitute a risk group. Children in the form of syrup or tablets dissolved in fruit school, however, are a captive populationthat is juice. workers supplied the easy to give supplements to with teacher weekly doses to each mother and advised them participation. WHO recommends this course of about storage and avoiding overdoses. The group action. A study of the impact of this type of was supplemented for 12 weeks while a control supplementationin children in primary school in group received a placebo. Anemia in the control Kenya (Latham et al., 1990) is available. group worsened, but the complete blood profile Supplementswere administeredfor 32 weeks on and iron reserve levels of the group receiving a daily basis (excludingweekends and holidays), treatment improved. This study demonstratesthe and the presence of bilharziasis was carefully eff:ctiveness of treating young children with oral ruled out. Although the number of children supplements and the possibility of having the involvedwas smal (29 received supplements,and mothers themselvesadminister the treatmentunder 26 were controls), there was a significantincrease the supervisionof communityworkers- in hemoglobinlevels and improvement in growth in terms of height and weight, thereby confirming It is frequentlysaid that iron supplementsrisk results obtainedelsewhere. A shorter study (four activating infection; Murray et al. observed a weeks) was conducted in Benin (Preziosi, 1990) higher incidence of a variety of infections in in which 347 infants in the COTONOU district Ethiopia, including tuberculosis and malaria received an elemental iron supplement of 50 mg among Ethiopian nomads after 30 days of iron five dayslweek. The impact on hemoglobin and treatment (260 mg orally). Hussein et al. (1988) iron levels was relatively limited. It is possible supplementedthe preschool children (2-6 years) that the dosageswere too low, the study too short, of 250 families in the CAIRO area (with a control or that infections that are prevalent during the group) for 10 weeks with 25 mg of ironlday; rainy season interfere with iron levels; additional hemoglodin and iron reserves increased studies suitable for each environment are significantly while the frequency of diarrheal necessary. A second study was conducted in episodeson a monthly basis fell. As for malarial Benm the aim of which was to increase vitamin C infection, a supplementationtrial conducted on intake to promote iron absorption; participants young children aged six to 36 months in Togo did consumed one orange daily for a period of two

23 The health secrorand nutritioninterveions in Africa months. There was no dramatic impact on important,a large number of women mentioned hemoglobinlevels, although ferritin levels were voluntarily restrictingtheir intake out of fear of significantly higher at the end of treatment having an oversized baby. A much broader (Hercberg, personalcommunication). In this age educationalapproaca to nutritionproblems linked group, an intervention to prevent intestinal to pregnancyis necessaryin such cases. This lack parasites may be very effective in decreasingthe of complianceis partly a function of the dosage: prevalenceof iron deficiencyin conununitiesthat low doses are better tolerated. To offset the have been greatly affected (Bundy, 1991) and discomfortcaused by the supplements,tablets with consequentlyin reducing the incidenceof anemia. slow gastric release (gastric delivery system or HBS; hydrodynamicallybalanced system; ACC/ iii) Analysis of the effectiveness of routine SCN, 1990) were marketed; there are no studies supplementaton availableon this topic in Africa.

The effectivenessof routine supplementation A lack of compliancewas noted even for the provided along with prenatal care in Africa is placebo group; however, noncomplianceis not rarely evaluated. Okafor et al. in Nigeria (1985) related to the discomfortlinked to the iron tablets examinedthe bone marrowof 31 pregnantwomen alone. This goes back to the more generalstudies who delivered children at the same time in a that were conductedon this issue (Morrow, 1990; communityin whichanemia due to iron deficiency Galloway, 1990). Logisticalproblems are one is conunonand where iron and folate supplements factor regularly mentioned in every program in are given orally on a systematic basis a. the almost every region of the world, including a beginningof pregnancy. Ninety seven percentof shortage of tablets and the distance to the health the patients were still iron deficient, and thirty center. But Taylor and Mutambu (1987) in five percent danonstrated megaloblasticchanges Zimbabweshowed that compliancewith the entire in their marrow despite the treatment. According treatmentdid not exceed 52 percent in a malaria to the authors, although many explanationsare prevention program, which was otherwise very theoreticallypossible (absorptioninhibitors in the well executed by a trained and motivated staff. diet, additional food sources low in iron and The other difficulty is to convincean individual folate, dosage levels that were too low), it seems who no longer feels very weak to continue with clear that noncomplianceis the issue. Even if treatment for several months; only repeated pilot supplmnentation trials have largely explanationsof the necessityto do so will counter demonstrated the effectiveness of this type of noncompliancein this area. Elegbe et al. in intervention,the effectivenessof making it part of Nigeria (1984) demonstrated that listening programs on a routine basis should be carefully carefullyto womenin the program, in conjunction analyzed. This is the case even when targetgroup with an effective and clearly explainedtreatment coverageis satisfactory. program designed to respond to the mothers' initial perceptions, can contribute to changing In referenceto the above, Gove et al. in 1987 participants' attitudes. Gove observed that the in Somalia (cited by Morrow, 1990) cited the distributionof iron tablets by trusted traditional fiollowingfactors: stomach problens and the bad midwives improved compliance. Greenwood taste or lack of the tablets' availability; most

24 Preventionof iron dejidency

(1989) made the same observation for a malaria the tablets is between $0.25 and $5 per person. preventionprogram. Tablets with slow gastric release, which are more effective, are also more expensive (between two According to the 1989 WHO report, among and 10 times more dependingon their origin). At the constraints on running effective programs is a tablet cost that is twice as high, the treatment the training of personnel (Benin, Gambia), the cost remains the same, since it requires half as lack of supervision(Congo), and most important, many tablets. No data are -availableon the costs an inadequate supply of medicines (particularly of a large-scale program in Africa. Providing iron and folate) (Benin, Gambia, Tanzania, pregnant womenwith supplementsat the time of Zambia), whether due to the lack of available their prenatal visits, which are already scheduled, financing (Tanzania) or logistical problems certainlydoes not involve any additionaloperating (Gambia). Infrastructureproblems do not seem costs aside from the purchase of the medication- to be an issue, although there are frequent staff shortages. Coverage constraints seem to be 8 Cost-effectiveness: Based on data collected surmountable as far as the target group of in 1980, Levin (1986) calculatedthat an increase pregnant women is concerned,although coverage of 10 percent in hemoglobinlevels in a population is still inadequate in areas where a plan of action following the consumption of the appropriate has not been vigorouslyimplemented: it has gone foodstuffsfortified with iron in Kenya could lead from 22 percent to 90 percent in the JNSP (Joint to a theoretical gain of $43 of additional annmal Nutrition and Supplementation Program) in revenue per person; an increase in hemoglobin Mozambique, to 90 percent in Gambia, and to levels of 25 percent following an oral 85 percent in Tanzania. Coordinationas alluded supplementationprogram would lead to a gain of to above is a key problem: in Tanzania, $107, also according to Levin- These numbers numerous pilot trials and research efforts are are based on productivity gains in agricultural carried out in a disorganizedfashion, due to the workers during experimentalstudies availableand numnbersof institutionsinvolved and the lack of are a theoreticalview of the problem. No actual centralization of efforts and results. Nigeria, data exists. confrontedwith the difficult task of coordinating prevention efforts in a variety of areas, set up a iv) Fortifyingfood specific committee on the federallevel to combat noninfectiousdiseases, includinganenia (WHO, * Increasing energy intake whenever it is 1989). inadequate is the simplest way to augment available iron intake; augmenting food intake * 7he costs of supplementation; The cost of alone will lead to an increase in iron intake. Even iron tablets (60 mg) and folate tablets (250 mcg) if the iron is not very bio-available,when a large supplied by UNICEF in 1989 was $1 for a quantity is absorbed, it may serve to meet package of 1,000 tablets originating in requirements. The shortest-termintervention is to Copenhagen. The cost on the open market for guaranteesufficient energy intake. In most cases, doses of 60-100 mg of iron was $4 to $10 per however, simply fortifying a basic food is an 1,000 tablets (in Zimbabwe, the cost was $Z 7.3 essential step. Young children are at highest risk per 1,000). With a basic course of treatment during the weaning stage, between six and 18 consisting of about 250 tablets, the total cost of months. Commerciallyprepared weaning foods

25 77Tehealth sector and nutritioninterventions in Africa

are often enriched with iron and ascorbic acid; the of diarrheal illnesses) represents a possibility to majority of the population, however, does not limit iron loss, especially in young children. The have access to such foods, and they are also problem with preventing helminthiasis relatively expensive. The CSM (Corn, Soya, (bilharziasis, ancylostomiasis, trichocephaliasis, Milk), a weaning food that is widely used in etc.) is that anthelminthic medicines are expensive USAID and CRS (Catholic Relief Services) and their effectiveness is short-lived if steps are supplementation programs in Africa, is enriched not taken to eliminate the sources of infestation in ferrous fumarate. elsewhere. Even if parasites are not eliminated, iron supplmnents are more effective than * Increasingbio-availability is another way anthelminthic treatment alone in increasing to increase the amount of iron that children absorb hemoglobin levels. Few studies conducted in from food. It may be achieved through malting or Africa are available, particularly for school-age fermentation (DeMaeyer, 1989), both of which children (see Stephenson, 1987). result in an increase in vitamin C content and a lowering of tannin and phytate content. Svanberg * Nutrition education: Taboos about food and Sandberg (1989) have shown in Tanzania that may have a significant impact on iron levels; in a sorghum-based diet, with a low level of bio- Jackson and Jackson (1987) noted that while food availability, procedures like soaking, germination, rich in folate is widely consumed by pregnant or fermentation activate phytase to such a point women in Liberia, there are still restrictions on that a moderate to high degree of bio-availability consumption of game and food rich in ascorbic is achieved (an effect similar to adding meat or a acid. Ojofeitimi et al. in Nigeria (1982) large amount of ascorbic acid to the diet). conducted research on the extent to which nutrition education was able to modify harmful * The cost offorryingfood is initiallylower eating habits during pregnancy, particularly the than that of supplementation: recurrent costs are tendency to reduce energy intake because of the also low. Therefore, it is certainly the most cost- fear of delivering an oversized baby. They effective measure (DeMaeyer et a., 1989). In showed that it was possible to change this practice countries with widespread iron deficiency, the cost with appropriate education aimed at countering of supplementing a large segment of the these fears. population would be a significant health care expense, which would be unjustifiable in terns of * As for young children, the simplest plan of other priorities; as such, only food enrichment is action involves maintaining breast milk intake for currently possible. Trials to fortify salt with iron as long as possible (iron that is highly bio- and iodine were conducted in Asia with available) and incorporating vegetables or fruits encouraging results. rich in vitamin C on a regular basis. Achieving a good level of compliance with these v) Other prevention methods recommendations again involves health workers who provide repeated explanations and * Preventionof parasites through improved encouragement. These recommendationsmay be : Preventing infection of any kind integrated easily into nutrition education programs (anthelminthic measures, wearing shoes, control

26 Prementionof iodine deftcency designed for mothers and their young children; obstacle to countering this disorder. Most setting up special programs is not necessary. important, the task of informing the public has been neglected. Informingpopulations at risk is * Family planning: Delaying the first a priority; this is the only way to maintain the pregnancyuntil a prospectivemother is physically persistentpressure on authoritiesthat will lead to mature, and spacing children to allow iron implementing and maintaining preventive reserves to be replenished, are logical and measures over time (and thus is the only way to effectivesteps that justify linking family planning ensure their success). to the preventionof iron deficiency. However, no studies have demonstratedthe real effect of such * The following six stages were outlined measures for a given populationin Africa. basedon an analysisof the difficultiesencountered in programs conducted to date throughout the world. They are sununarized by Hetzel (Hetzel, 3. Prevention or iodine deficiency 1988) from the development of a prevention program for Iodine DeficiencyDisorders (IDD): a) FINDINGSAND RECOMMENDATIONS 1 Choice of strategy based on the extent of the problem: In general, when local conditions Iodine deficiency is currently "the most include disorders causedby marginal to moderate avoidable cause of delayed intellectual iodine deficiency, iodizingsalt or other foodstuffs development globally" (Stanbury, ref. idem as quickly as possible shouldbe recommended;in Hetzel, 1988: 116-119). Of the deficiency cases of severe deficiency, treatmentwith iodized diseases, it is certainly among thosethat could be oil shouldbe initiated. eliminated relativelyquickly. One must estinate first the exact extent and * Science and improved technique have severity of the iodine deficiency in the country. produced all the answersneeded to provide for the At the very least, a representativesurvey (on the effective control of this deficiency disease district level) shouldbe conducted of a reference inexpensively,at no risk to health, and based on group, such as childrenbetween the ages of 10-15 simple techniques that are easy to implement. It years, including the collection of reliable is mostly a question of political will. The epidemiologicaldata accompaniedby water, urine, obstacles encounteredto date include,first, a lack and, if possible, circulatingthyroid hormone (T4 of understanding of the problem by political level) analyses. Laboratory facilities are useful leaders who consider it more of an aesthetic issue for preparing and tracking the program; they are (goiter) than a serious public health problem, not absolutely necessary to initiate it, however. although even a marginal iodine deficiencyhas an Expensiveequipment may be purchased together impact on intellectualcapacity. Second, iodized for use on a regional level ($30,000 worth of salt programs have had problems that discouraged investment for automatic dosages of urnary health officialswho were not managingthem well, iodine) (WHOQNICEFIICCIDD, 1989). The lack of trained personnel capable of addressing theaissuesof laboeeratoresupabero o It is also necessary to study the salt produc- addressing the Issues of laboratory supervision, tinadisrbuoniutonnthcutyad salt iodizing techniques,etc., constitutesanother

27 Thehealth sector and nutrition interventdonsin Africa to identify the best iodizing method to be able to individualswho are charismatic, very dedicated, compare the possible benefits linked to each type and able to inspire and sustain enthusiasmduring of intervention and their respective costs. For the entire course of the project. countries that import most of their salt, it is very important to ask the government to take the a It is particularlyimportant not to establish necessary legal steps so that all imported salt is a verticalprogram to avoid artificiallyprolonging iodizedand to ensure that the laws are obeyed. these programs. In fact, integrating a program into the expandedprogram of immunization(EPI), 2. These results should then be incorporated food inspection services, or other programs in into a large publicity campaign aimed at health existence avoids the problem of vested interests workers and the public. that may arise when, for example, the capsule distributionprogram must be reduced because salt 3. A national action plan should be is currently iodized. developed involvingall sectorsafter discussionof the different possible altematives and cost * Programs that work best from the considerationsthat takes into account possibleaid beginning have a solid central organization; one from specialized international agencies. A should still aim to integrate them into existing national committee to prevent iodine deficiency structures to minimize costs and guarantee their disorders should be establishedand invested with durability. The integration of the iodized oil the authority necessaryto fulfill its missionbefore distributionprogram into the vaccinationprogram any broader programs are implemented. The with which it shares some characteristics is health sector should push for its establishment, possible; nevertheless,the target group is not the and the commitee shruld includeall sectors so as same and the treatment should be limited to risk to involve the maximum number of potential zones, unlike the vaccinationprogram. One must actors in the program to be established. also make sure to avoid the simultaneous distribution of vitamin A and iodized oil (given 4. Effectivepoliicalsupportshouldbesought orally) because there may be some at the highest levels to guarantee that the incompatibility:a massive iodine dose is capable program's leaders have the necessaryauthority. of destroyingvitamin A's double bonds.

5. Steps to organize staff training programs, * The medical solution is of interest because necessary resources, and program implementation it does not require one to wait for decisions from should be taken. other sectors, which often involves a significant delay; it may be implementedquickly. Although 6. Upon completion,the program should be the only long-term solution to the problem of appropriately assessed. iodine deficiency is fortifying salt or another foodstuff with iodine, the medical professional Too often, some steps are not followed,which should not ignore the problem simply because it makes it difficult to complete a program cannot resolve it fully. efficiently and successfully. One of the determining elements underscored in most * The distribution of iodized oil by mouth programs is the presence of one or more has proven to be a good solution; it provides only

28 Preventionof iodtinedeficiency one year of coveragebut does not require highly zones of iodine deficiency may also be specializedstaff; as such, it may be integratedinto diminished.From this perspective,the prevention a treatmentand preventionprogram by community of iodine deficiencyin pregnant women is vital. healthworkers or withinthe commnunity(schools). It is less expensivethan intmuscular injectionsif c) EoL staff not hired for that specificpurpose are used; if not, the number of visits necessarymakes fthis The principalcause is an iodine deficit in the a more costly solution. environment, which results in a low level of iodine in drining water and locally produced * From the beginning of the process, the food. Some vegetables also containgoitrogenic projects should consider all of the treatment elements(corn, sweet potato, cabbage, and some 'chapters" and solicit appropriatefunds: iodizing typesof cassavathat are rich in linamarine,whose salt, megadoses, training, supervision, main metabolite is thiocyanate, the goitrogenic communication,and research. Standard plans effect of which has been widely shown in the were preparedby an appropriateinstitution, which north of Zaire [Delange and Abluwalia, 1983]). is able to furnish high-level technical advice But if iodine intake is adequate, the thyroid can (ICCIDD). adapt to an overloadingof thiocyanatewithout developinghypothyroidism or goiter (Cliff et al., b) CONOUEICES OF IODINEDFCIEANCY 1986in Mozmbique), and in most cs iodine supplementation corrects the disorder. The Hypothyroidismthat results from an iodine presenceof goitrogenicfoods is one more reason deficiencyis manifestedin adults by apathy and a to insiston a high level of iodizationin salt. decrease in mental functioning; this may be corrected with iodine supplements. In children, d) PREVALENCEiN AFRICA from the momentof conception,iodine deficiency may result at the very least in a decreasein IQ in Almostall Africanrcountries (41, accordingto additionto a variety of forms of crefinismlinked Hetzel, 1989) have zones that are known to have to a permanent change in the bamin,which is a high prevalenceof goiter. A broad region of characterized by spastic diplegia, mental iodine deficiency varying in severity extends retardation, and deafnessalone or in conjunction across the entire center of the continent, from with muteness. This deficiency is oftun more Nigeriato Senegalin the west and from Ethiopia, prevalent among women, probably for hormonal Tanzamia, and Kenya in the east; the entire reasons. An iodine deficiency in a pregat southern plateau is an area of moderate woman may lead to the death of her fetus or deficiency, including vast zones in Zimbabwe, newborn or to a congenital deformity. Botswana, Mozambique, and Madagascar. in Postneonatalmortality is also elevated in zones 1987. Tanzani Kenya, Ethiopia, Sudan, where the deficiencyis common. Zambia, and Zimbabwe were the only countries that had data on a nationallevel (Ekpechi, 1987). When the deficiency is marginal, it may go There is still little data available on the Sahel undetected. It may alter the growth of childrenin region or other countriesin eastern Africa, except the community,however, thereby reinforcingthe for the centers of severedeficiency that have been impact of malnutritionon growth. The mental tat persist despite abilit of apparentlyabilty f nonnalaparetlynoral childrenhilden livingivig ~ i recognizedfor some time and

29 The health sector and nurition interventionsin Africa

Table m. Epidexologial criteria for evaluatng the severit of iodine deficiency in a population Crteria Severe Moderate Ughr A. Incidenceof goirer(%) (inscooL-age chifdren, 6 to 12years): Taral(grades I o 3) (totalgoiter) >50 20-49 119 sibe fgrades2to3) (visibkgoirer) >10 5-9 1-5 B. Uinaryiodie (median,megI) <20 2049 50-99 C. Thyroidsdmadadng hormone (7Ff): % > 5 rg/i fcr*teriaare currendybeing retsed) (adudtsor infants,not neaal) D. Prevalnceof crednism(%) >I <1 0 Source: FAWWHO.1992. Preveng Specific MicronutrientDeficiencies. Theme Paper No. 6, ICN Rome.

sporadic or isolated interventions (Brown et al., prising three subgroups: West Africa (Pr 1990, Zaire). Ekpechi, Nigeria), Central Africa (Pr Lantum, Cameroon), and East and SoutheemAfrica (Pr e) PRESHOJND 7NMAW7MA2W Kavisbe, Tanzania). A regional strategy and action plan was established at the same time There are two main methods of preventing (WHO, 1986). iodine deficiency; they have amply proven their effectiveness. In the short term, for either i) DsbUt Wionof iodized salt emergencysiuations or isolatedcommunities that cannot easily reach sources of fortified food, Salt is the only foodstuff that is consumed in iodized oil is administered. In the longer run, the a more or less constant amount throughout the solution is to fortify foods that are broadly and year by all members of society regardlessof social regularly consumed, which usually means salt. class. Production sites for salt are among the The strategy officially adopted by the regional most centrally located, compared to those for WHO bureau for Africa is to fortify salt with other foods. Finally, salt is one of the rare foods iodine wherever possible, while providing zones that reaches even the most remote regions. that are prevalent in iodine deficiencywith iodized oil treatmn programs. To date, the lack of Ini practical tenns, the problem is in making poilitrcalwlnentprogms. rtboate, the plograms this product available to populations isolated in political will needed to ensure that promountainous or semidesert areas. Other run well has -been an obstacle, resulting constraints include preserving the iodine content particularly in a lack of essential coordination of salt, especially in humid climates (regular among various sectors. An important step was inspection is essential), and verifying treatment taken with the creation of a task force for the effectiveness. prevention of iodine deficiency in Africa, com-

30 Pmr tion Of iodine deficiency Althoughthese issues are outsidethe technical iii) Administeringiodized oil indihduaUyby domain of the health sector, their consideration injecton should underlie decisionsmade concemingthe implementationof a distributionprogram. The A single intramuscularinjection is given, health sector should encocrage efforts to which preventsdeficiency for three years (one coordinatethe varioussectors to implementthe injectionof 0.5 ml for childrenunder one year of distributionprogram successfully,inform and age and 1.0 ml for those over one year). The educate consumers,verify that the program is target groupcomprises all femalesfrom birth to working, and measure the health advances age 40 and all males from birth to age 20 in achieved. communitiesaffected by the deficiency. Dosage amountsvary with age; after age 40, injections The cost of iodizingsalt at a salt treatment are not given to avoid possible thyrotoxicity. center that has alreadybeen set up involvesin Aside from the immediatediscomfort, there are general a surchargeof $5 to $12/ton, that is, no notablesecondary effects. The only problem about$0.024 to $0.06annually per person. Local is that injectionsmust be givenat a healthcenter, marketprices, however,may differ: in Congo, which contributes to their higher cost. An iodizedsalt sold in the endemiczone is about20 increasingnumber of health center officialsare percent more expensive (AfricanTask Force, concerned,however, about the heigbtenedrisk of 1990). the transmissionof viruses, such as the AIDS virus, and for this reason, there is renewed ii) Administeringiodized oil by mouth interestin oral supplements.

A capsuleis givenby mouthonce every 12 In centralAfrica, the cost of oil, syringes, months. For this method,staff do not need as and needlesrepresents 41 percentof the totalcost much training as in the case of injections. In (equipmentand distribution). Administering addition,the risk of viral transmissionlinked to capsulesby mouth is clearly less costly than injectionsmay be avoided. Currently,to avoid intramuscularinjections; but since this method the possible toxic effects of administrationby providesa shorterperiod of protection,especially mouth,using small quantities (0.1 to 0.25 ml for at low doses, the annualcost per person may be nursinginfants and 0.5 ml for all others),which as high. provide about one year of protection, is recommended. Tests conducted in Kirotshe, iv) Othermethods Zaire wereconclusive (WHO/UNICEFIICCIDD, 1989)and are beingcarried out on a largerscale. Otherpossible approaches are modilfyingand diversifyingeating habits by incorporatingfoods A recent alternative,which is even less originatingin zones that are not deficient in expensive than iodized oil capsules (by 20 iodine,for example,or newfoods; supplementing percent),is administeringvaporized iodized oil by food and water (Lugol's solution)consumed by mouth(ORIODOL, Guerbet Laboratories). This people in the iodine-deficientzone; providing technologyis relativelyrecent and has not yet medicationthat is made from products rich in been implementedon a large scale. iodine(iodine tablets, which should be availableto pregnant women in dispensariesin endemic zones);providing iodine treaunent of agricultural

31 7Tcheakh sector and nutritionintrvwntions in Africa lands and cattle. Aside from salt, many other endemic regions. One should be aware, however, foodstuffs are fortified, among them a variety of that providing iodine supplements to a woman condiments, sauces, flours, bread, milk, etc. who is fir along in her pregnancy risks causing Fortifying weaning foods has also been proposed, neonatal hypothyroidism, which is transitory in as has supplementing food aid, which is the case most cases; for this reason, low doses of 0.5 ml for vitamin A and powdered milk. Ongoing per patient are advised. supplementation trials of bottled water were conducted, particularly for wells and water * Children zero tofive years of age: If the exploration sites in Mali. There have been no mother is nursing and she is receiving iodine, her complaints about the water's appearance or taste; breast milk will contain it. Giving iodized oil to no cases of hyperthyroidism have been reported; children beginning at six months of age is and the impact was rapid. The cost of the recommended to avoid developmental problems diffusers used to iodize a well serving 800 people resulting from iodine deficiency. is about $150 per year; the cost of educating village residents so that they accept and use this * Childrenfrom six to ISyears of age: This method must also be included. As for goitrogenic group is not a priority, but it may be targeted foods, if their consumption cannot be eliminated, through schools. Studies have shown that various procedures to detoxify them may be supplementing this group may enhance learning applied, such as soaking and retting some types of ability. bitter cassava root in central Africa. 3 The risk of thyrotoxicity has been v) Target groups demonstrated i.nindividuals who are over 45 years of age, iodine deficient, and have consumed U Married women of childbearing age: This iodine-rich foods. For this reason, it is is the target group, since protecting the fetus as recommended that individuals over 45 not be early as possible is desired, which means included in iodization programs. involving women even before they become pregnant. It is not always easy, however, to Current recommended doses are the following reach this group without a vertical program. If (FAO/WHO, 1992): this is not possible, one should try to target child of 6 to 200 mg orally (I capsule)per year or pregnant women. 12 months: 240 mg via injection(0.5 ml of Lipiodal)every 2 years

* Pregnant and nuing women: It is between 1 400 mg orally (2 capsules)per year or and45 years via injection(1.0 ml of Lipiodol)every relatively easy to reach this group through the of age: 3 years health system if it provides good prenatal coverage. Iodine deficiency in the mother has an impact on the various stages of development of the fetus; it may lead to neurologicalcretinism in f) ASSESSINGPROGRAMS CONDUCTED IN AFRICA the second trimester (or perhaps even in the firs! trimester) and to myxedematous cretinism in the Of the 32 countries classified as having iodine third trimester. For this reason, experts agree deficiency problems, only seven have established that pregnant women should be treated in highly significant treatment and prevention programs.

32 Preventionof Iodinedeficiency

Followinga scientific meeting specificallyon theoreticalcapacity to distribute500 doses per iodine deficiency,held under the aegis of the day. Tha goal was to treat the entirepopulation OAU in Addis Ababn in 1980, severl over five years (500 doses x 200 days/yearx 3 governmentsmade an effort to initiateprograms teamsx 5 years = 1,500,000people treated) for to iodizesalt, but wiLhoutmuch success. In the a costof $0.35per treatedindividual or $0.07per opinionof Ekpechi(1987), the main reasonsfor personper year of protection($150,000 for oil, the small number of programs implementedto $60,000for supplies,$125,000 for coordination date in this area are (a) an acute lack of andanalysis, $172,500 for the three mobileteam, informationon the exact situation (out-dated totaling an estimated$507,500 for the entire surveys,inadequate modalities), and (b) a lack of programin 1977)(Thilly et al., 1977). politicalwill, sincedecision makers have not yet realizedhow seriousthe problemof this endemit. Despite the delay in putting together the diseaseand its consequencesare. teams,360,000 people were treated by two teams during the first four years in an area of 1,200 Since 1987, a number of nationalprojects square kilometers,and more an one million were initiated, but mostly in anglophoneEast people were treated between 1977 and 1984 Africa(WHO/UN1CEF/ICCIDD, 1937). (Htzel, 1988). The program was gradually transferredto routinelocal health services, while i) Assessitngiodized oil injectionprograms a coordinationand evaluation committee led by an epidemiologistat a central agency followedthe Zaire: The first treatmenttrials in Africa evolutionof the programclosely (rhilly et al., analyzedwere held in Zaireabout 15 yearsago in 1977). Beginningin the second year of the isolatedmountain areas where the prevalenceof program,teams that werenot part of the program endemicgoiter had reached a very high level treated42,000 people with a level of coverage (Thillyet al., 1977). estinmtedat morethan 70 percent(Thilly, 1981). Thisprogram was accompanied by an educational Giventhe delays requiredto developa salt campaign,specifically on rettingcassava and the iodization program in Zaire (legislative, value of eominuous breast-feeding since economic, logistical, and administratlve thiocyanateis not concentratedin breast milk. problems),launching a medicalprogram using One of the mostdramatic effects of the program, iodized oil injections was judged more aside from the decrease in the prevalenceof appropriate.The progran beganwith a pilottrial goiter, was rue decrease by half of perinatal involving30,000 people on the islandof ldjwi in mortalityand the increaseof an avnrageof 200 the easternpart of the countryin 1970(Thilly et grams in birhtiweight for childrenborn to first- al., 1977). Irn view of the dramaticresults time motheis, comparedwith children born to obtained, an eradication program was mothersreceiving a placebo. implementedin the UbangiMongala region in the northwest for a population of 1.5 million Zairecurrently has oneof the world's highest inhabitants.The programwas carriedout using rates of endemicgoiter. In 1987,out of a general teamsand a logisticalplan d:. had alreadyproven population of 33 million inhabitants, it was effective for a vaccinationprogram. Three mobile estimated that 13 million lived in iodine-deficient teams were set up, each consistingof three zones,withfiveto 10millionlivingin areas with nurses, a secretary, and a driver, with the

33 The healthsector and nutririoninterventions in Africa high levels of endemic goiter, while approximately is likely that the populationthat is most at risk is 1.2 million showed various signs of cretinism not reached. Specific targeting of these zones is (IDD Newsletter, 1988; Hetzel, 1989). Fortified required, which clearly increases the cost of a true by previous successes, the country launcheda six- eradication program. year treatment and prevention program in 1988 through the existing health care system for 50 Ideally, injectionsshould be administeredwith percent of the population in risk zones. In terms one sterile syringe and a disposable needle. This of priorities, pregnant women and newborns are poses two problems: the risk that soiled needles targeted first, followed by school children, and are passed around and high costs. In many cases, finally adult men with goiter. This program health servies don't have the means to meet these includes all the components of an integrated conditions, and they reuse the equipment after program: training of personnel, laboratory boiling i!, which is not ideal. facilities, complementaryepidemiological studies, and salt iodization trials. Application methods Tanzania: Tanzania carried out a single vary according to regional conditions; the intramuscularinjection pilot program (480 mg) in program as whole is directed by a national a region with a high level of goiter, and the commnittee. A final evaluation will be conducted results were analyzed after three years. The in 1993. The estimated cost per person treated is program had a positive effect, but an interval of $1.00, includingadministrative fees (the oil costs three years was judged to be too long for children about $0.40); this amounts to a total cost of $1.2 whose needs are greater than thnsc of adults million (IDD Newsletter, 1988). These numbers (Waichter et al., 1986). According to the should be evaluated carefully, given changes in researchers, the interval between the two the value of Zaire's currency. injections should be reduced to two years. Pregnant women had a satisfactory level of One of the difficulties in a program of this thyroid function, as did their children who were scope is knowing to what extent it may be breast feeding; on the other hand, as soon as decentralized. Experience has shown that the breast-feeding ceases, it is necessary to provide choice of the population to be treated, the treatment for the children. There were no toxic organizationof the program's schedule, the supply effects noted, in contrast to some previous studies. of iodized oil doses, and the ongoing analysis of epidemiologicaldata and statistics should be done In MALAWI, between 1984 and 1986, by a central agency. The pilot trial's success is 63,000 people receivedan injection of iodized oil, what led to the crucial support of the authorities of a population at risk totaling betweer 150,000 throughout the program's implementation. and 350,000 (IDD Newsletter, 1987). A program to import iodized salt has not yet been instituted. The easiest way to ensure the continuity of a program over many years is to integrate it into the ii) Assessingprograms to administer iodied expanded program of immunization insofar as oil orally these programs achieve a significant level of coverage. Coverage at the level of the 85 percent Algeria: The first trials of iodized oil given obtained in Zaire, however, is still insufficient; it orally in Africa were conducted in Algeria.

34 Preventionof iodinedeficency

According to Ben Miloud (WHO/UNICEF/ routes of administering iodized oil were still ICCIDD, 1987), the procedure provides effective similar (decrease in the prevalence of goiter and protection for two tc three years. More recently, normal biological indicators). The quantity of after pilot trials, doses of 0.5 ml (240 mg) were iodine in breast milk was increased satisfactorily judged adequate for 18-24 months of protection; during a period of at least 30 months (Phillips et 30,000 women and children were successfully al., 1988; Phillips & Osmond, 1989). treated; there were few secondary effects, no toxic effects on the fetus in pregnant women, and only Malawi: Since 1989, Malawi has modified one case of exophthal-micgoiter. This simple its program to introduce treatnent by mouth technique, which is easy to integate into primary (African Task Force, 1990). health care services, has been extended to the target population at large throughout the risk zone iii) Assessing iodide tablet distribution since 1987. programs

Sudan: In the Darfour region, the possibility An attempt was made to distribute potassium of using iodized salt seemed unlikely, given iodide tablets in schools in Sudan (in the Darfour transportation difficulties, the distance to the Red region); after supply and distribution Sea, and the preference of the populationfor local irregularities, however, the program was salt; iodized oil injection trials were thus suspended (AbdelWahab et al., 1984). Phillips et conducted on an experimentalbasis in 270 school al. (1988) in Zaire compared the administration children- The satisfactory results showed that an by mouth of various dose levels of potassium injection every four years could be an acceptable iodide (0.5 g, 1.0 g, 2.0 g) to the administration strategy. A controlled trial was then conducted to by mouth of iodizedoil. There were no long-term compare the effects of administering by mouth effects (eight months) for any of the potassium two doses, at one-year intervals, of 400 mg of iodide dose levels in contrast to the iodized oil. iodine with those of a single injection over two The daily dose thus produces a prophylactic years of 1 ml (475 mg of iodine) in 2,316 school effect; although very easy to implement on a children. The reduction in the prevalence of conummity level, it quickly runs into supply goiter and the normalization of biological problems on a larger scale. It should be available indicators was similar in both groups (Eltom et to pregnant women in dispensaries in endemic al., 1985). zones, however, when no other alternatives are available. Zaire: In 23 villages in the eastern part of Zaire (Phillips et al., 1988), Phillips comparedthe iv) Analysis of iodineforificadon of salt effects of administering iodized oil by mouth (2 mnl)with a placebo and with the injection of Lesotho: In Lesofefo,the presence of visible iodized oil (2 ml). Eight months later, the goiter in adults fell from 15920 percent to about prevalence of goiter was still unchanged, but 5 percent betwveen1950 and 1977. This decline is biological indicators were nomtal in individuals attributed to the importation of iodized salt from who had received the iodized oil by either route South Africa (IDD Newsletter, 1987), although whothadreer iveconductedaiodizedoithe hendrofute neidter the content of the salt nor its distribution Anothearsanalysis conduchatrinedthe byothetwo was really subject to supervision. It is not klown years showed that the results obtained by the two

35 The health scnor and nutrition intervenrionsin Africa how much tine is required to achieve this result; Fortification levels are often still too low; this is a concrete example, nevertheless, of the experience has shown that the level should be at long-term effectiveness of the importation of least 100 mg/kg to account for the losses that iodized salt. usually occur before distribution. Packaging poses a problem: preparing small packets Algeria: The first iodization trials of enclosed in a good-qualityenvelope preserves the industrial salt produced locally were conducted in iodine content but increases the price. Machines Algeria in 1967 (WHO/UNICEF/ICCIDD,1987); exist that can handle the packaging of small the program was limited by a number of quantities and can be adapted for use almost constraints: insufficient production; an iodine anywhere. With these machines, treatment could conten; that was too low for the consumer; a be limitedto zones with a low number of endemic higher price than ordinary salt, although it was cases, but experience in this area is still lacking. produced primarily fbr low-income consumers; The principal problem to date is that the program and a lack of supervision. It was necessary to is not economicallyfeasible. substitute a distribution program of iodized oil by mouth. Zambia: A treatment center that was established in 1982 had to close after three years Kenya: A salt iodization program set up iu (Mannar, 1989) after excessive corrosion and a 1970did not lead to an appreciabledecrease in the lack of spare parts; it is this sort of program prevalence of goiter due to the availability of limitation that requires a degree of political and other sources of salt in the country, and the economic commitment on the part of the prograrn was not continued (Van den Haar & government in the long run. Kavishe, 1986). In fact, setting up an iodization program should be preceded by an analysis of all Regional Programs: For countries that do salt suppliers, their distribution networks, and the not produce their own salt and are thus obliged to iodine content of their salt. In 1987, Kenya import it, the most effective step is to verify finally strengthened its legislationconcerning the through legal mechanisms that imported salt is importationof iodized salt and the minimal iodine iodized; laws in this area are often not applied, content required, and it instituted strict inspection however, even when they are on the books. The procedures (Alnwick. 1988). national conmnitteeto prevent disorders related to iodine deficiency is responsible for making sure Ethiopia: Ethiopia, with about 12 million that the law is applied and that analyses are cases of goiter (which rules out only distributing conducted on a regular basis. Not all countries iodized oil), opted for distributing iodized salt. have passed laws of this type yet. The government set up iodizationcenters in Assab and Massawe in 1987. The country quickly In fact, analysis of production on a continent- increased its production levels, and it could wide level (see WHOIUNICEFIICCIDD, 1989. become an exporter in time. Political leaders are for example) shows that some countries supply still not sufficiently aware of the need for ongoing their own neighbors with salt. For example, programs in thi. area (Madinger. 1988). Senegal supplies salt to about 20 countries in West and Central Africa. The production site at Kaolack is private. After almost 20 years of

36 Prevefrionof iodinedeficiency ignoring the problem, the head of this operation northwest part of Mali in four villages (about has just agreed to negotiate the possibility of 1,000 people) in a highly endemic zone (the iodizing salt. Cameroon has a large salt prevalence of goiter surpassed 50 percent). The treatmnentcenter in Douala that supplies a good prevalence of goiter fell rapidly, and biological part of the country's needs as well as those of the indicators gradually normalized. A study of the Central African Republic, Chad, Gabon, and population's acceptance of this program was also Equatorial Guinea. The salt could be iodized at a conducted by a sociologist. The comp3ny is low cost; the price would be $0.02 annually per capable of producing large quantities of diffusers person (Mannar, 1989). GHANA exports salt to quickly (100,000 in 1993!), which would permit about 12 countries in West Africa. large-scale tests. As long as the cost of the diffusers is not too high, this solution certainly has Following renewed interest in the issue, a bright future in cases where the sources of the Kenya and Tanzania have become exporters of water supply are not too spread out (an ideal iodized salt and are able to satisfy the needs of solution for "well-water exploration sites"). The surrounding countries (Uganda,Rwanda, Burundi, predicted cost is estimated to be $.10-$.15 Eastern Zaire, Zambia, and Malawi). Malawi annuallyper person. The interestingaspect of this and Tanzania are in the process of establishing technique is that it allows one to iodize livestock additional production sites (Mannar, 1988). and lands under cultivationand could be adapted Botswana also possesses a large salt production for other nutriments, like vitanin A or iron, for site; negotiations have been underway for several exmple, or even for the distribution of larvicide years to begin iodizing salt produced there. It and bactericide. The project is currently being could then be~exported to neighboring counttnes extended to the entire district of Kita with support (African Task Force, 1990) from the World Bank (see WHOIUNICEF/ ICCIDD, 1987; WHOIUNICEF/ICCIDD, 1989; At this stage, a concerted regional strategy is African Task Force, 1990). One of the problems essentialto keep the pressure on producers so that to be resolved for a large-scale program is tht of they make necessary changes and to guarantee changing the diffuser annually when the pump is them a market so that they will accept the serviced. changes. It will then be possible to bring the iodine deficiencyproblem under control quickly in vi) Informationlendon foenuicaon a large part of the continent. According to Martin, a questionnaire v) Fortifyingwater distributed to country representativesattending the Dar-es-Salaam conference in March 1990 Mali: This solution was adopted because of demonstrated that practically nothing had been the number of entry points for salt originating in done in this area (African Task Force, 1990). the Sahara, making it impossibleto iodize all salt This program element is clearly neglected. If supplies. Thanks to support from the Rh6ne making a vaccination program work depends on Poulenc Foundation, a silicon polymer-based being able to convince the mother to bring in her iodine diffuser, which captures sodium iodide and child, and if the success of an ORS program releases it in small quantities into the water over depends on the ability to teach her a new the course of one year. was developed. This technique, the success of a program to prevent technology was tested successfully in the iodine deficiency implies that all levels of a

37 Te health scnor and nutflon intennwsionsin Africa populationhave been informedand participate especiallyin countriesin West Africa. Health since preventionis based on changingcommon workersare obviouslypoorly trained, and for this behavior(avoiding or treatinggoitrogenic foods, reason cannot activelysupport the programsir. using differentsalt). All significantmeans of place. On a national level, there is a lack of communicationshould be broughtinto play, such informationexchange among countries; active as radio, meetings, etc. This aspect, which is regional groups should be established. often neglected, is essential, but it frequently poses difficult problems. In Nigeria, 70 percent vii) An integratedprogram of inhabitants are believed to be isolated from access to the media (Hetzel, 1989). A campaign Tanzania: Since 1983, Tanzania has of this sort would not appear to be necessary if implemented a comprehensive program, after iodized salt is not competingwith other salts on establishinga nationalcommittee to prevent iodine the market availableto the population. deficiency. The program comprises a public outreach component through the radio and In Zimbabwe, the consumer association newspaper, a training component for health refused to approve legislation concerning the workers in the zones at risk, studies and importation of iodized salt because it did not laboratoryanalyses, and iodized oil injections in provide the consumer with enough choice; zones where the prevalence of goiter exceeds 60 information about disorders related to iodine percent. twentythousand injections were given as deficiencyin the countryand the fact that using a result, whilea distributionprogram of 700,000 iodizedsalt does not pose any risks is importantin capsules (by mouth) has been underway since countering such opposition (IDI)DNewsletter, 1987, specificallyin the context of the JNSP in 1987). Iringa. A salt iodizationpilot program was finally set up in a district; the percentage of school UNCFmd-ies nEs fiat children with goiter decreased from 60.9 percent provide Lhepublic with information; this format to 30.4 percent in six months, and to 6.0 percent has proven itself to be very useful in promoting 12xmonths.in a an te were in 12 months. Several iodization centers were the acceptance of' programs and public establishedto provide supplementationthroughout participation. Only films that have been made the country(Van den Haar & Kavishe.1986). locally have an appeal to people. however: they are not interested in those made elsewhere. Two problemsdelayed the program in its first Educating political leaders at all levels is phase: the first was the initial lack of infbrmation on the extent of the problem. and the second was essential, since the vital step of setting up a . m . a . national committee to prevent iodine deficiency the coordinationof various aspects involved in the begins with their awareness of the problem: the program. Since 1986. almost 20 districts have absence of political will until recently is a result been integrated into the iodized oil distribution of this lack of information. Playing this role program; the level of coverage is 50 percent should be a priority of the health sector. (African Task Force, 1990). Doctors in the district were made responsiblefor the distribution, In terms of training, there is a clear lack of but each district organized the distribution brochures and pedagogicalmaterials available, independently through health centers and

38 Promotingbrasfefeding dispensaries, primary schools, NGOs, mobile practices, it is necessary to initiate campaigns to teams, or even political leaders in the villages. promote bti and to ensure that The principal current problen is supplying the regulations in this area are enforced. capsules, which makes targeting distributon necessary in the first phase. * Improving breastfeeding practices in maternity associated with the Baby The Tanzanian example has been an Friendly Hospital initiative leads to a notable indisputable success; the program included the improment in children's health. This sort of important component of raising awareness about initiative may be implementedonly by sensitizing the problem and education. The circumstances decision makers and training and are rather unusual in Tanzania: it has a national paramedicalworkers in the correct way to manage language, literacy rates that reach 85 percent, breastfeeding. universal primary school attendance; and an omnipresent and active political party structure. * Community initiatives may increase the During the initial study, discussions with prevalence of breastfeeding, and they have a professors and students were held in schools; positive impact when they are coordinated on a students also served as information agents for the national level by a multidisciplinaryteam. population at large. The results of the study were discussedwith political leaders in each district and U Highly motivated individuals should village. Radio shows based on interviews of coordinate the projects, and the govemment women with goiter were broadcast and meetings should take a stand on the issue in support of were held with local journalists to encouragethem these efforts. to pass on the information. - As soon as an initiative gets underway, a data base should be created; compilation and 4. Promoting breasfeedig rigorous evaluation of the data is necessary throughout the project. WHO/UNICEF indicators shouldbe used (WHO, 1991). a) FiMNNGSA RECOMMENDA7ONS * Since breastfeeding improves the outcome i) Recommendations of other interventionsin the health sector, budgets for these interventions should protect, promote, * In a traditional society, where breastfeeding and support breastfeeding. For example, since is very prevalent, a coordinated policy backed by breastfeeding improves the effectiveness of regulations to protect this practice produces immunization, it should be encouraged by results. The tendency to decrease brmstfeeding immunizaton programs. A "withdrawal" froin an may be reversed. which helps to protect it, as is EPI budget could be used for a national campaign the case in Papua New Guinea. But the recent to promote breastfeeding. A water advent of cable and satellite television advertising decontaminationand quality improvement project threatens these efforts and poses a problem that could set aside part of its budget to protect may only be resolved on an internationalleveld To breastfeedingto have the greatest impact. Family maintain policies to protect breastfeeding planning budgets should include something for

39 The healthsector wnd nutrition interventions in Africa protecting breastfeeding. The team responsible for practicing what is recommended, how these encouraging breastfeeding should have enough constraints have been overcome over the last three financing and not be dependent on external decades, and recommendations for future actions. sources of funds, but it is also in the interest of people running the EPI, decontamination, and The above report was presented to a family planning programs to make a contribution. WHO/UNICEF meeting of high-level government policymakers from industrialized and developing ii) Background countries, and 10 UN and bilateral agencies, held in July 1990 in Italy. After reviewing the report, That promotion, protection, and support of the meeting produced the Irmocenti Declaration on breastfeeding is essential to attain the Protection, Promotion, and Support of Breast- and nutrition goals for children during the 1990s feeding (1990), which is now policy for WHO and is no longer in dispute. The challenge is to UNICEF. The principles are also embodied in identify, then implement specific actions that will USAID's 1990 policy and strategy for breast- result ini increased prevalence of exclusive feeding promotion. Participating governments and breastfeeding through a minimum of four months agencies committed themselves to specific actions and that of complemented breastfeeding through and targets to attain optimal breastfeeding. two years or beyond, as illustrated in Table IV. This pattern of feeding is known as OPTIMAL iii) Pracdces in Africa BREASTFEEDING. Figure 3 illustrates the extent of the problem The best and most concise review of the in some African countries: in Mali, for example, global situation regarding breastfeeding is only 10 percent of children 0-4 months of age are presented in a 1990 WHOIUNICEF report breastfed exclusively, which is the ideal feeding entitled, BREASTFEED!NG IN THE 1990s: method. In Ghana, more than half of children Review and linplicationsfor a global strategy under four months of age are given a bottle, (WHO, 1990). The report presents the results of which senselessly exposes them to the dangers of a three-year systematic assessment and analysis of diarrhea. the status of breastfeeding, constraints to Moreover, among children who are six Table IV. Feeding recommendations by age months and older, many are undernourished, as in group the case of Mali, for example, where 55 percent Agein of children are given only breast milk and other months Recommendations liquids (Figure 4). 0-4 100 percent of inliantsshould be exclusively breast-fed iv) Breasifeeding: impact on morbidity and

4-6 Two-monthtransitional period moral

6-24 All infants should be receiving The optiml feeding patter described above complementaryfood has become rare. Even where frequent and Soarce: WHO 1992. prolonged breastfeeding is customary, early use of

40 Promotingbreas.feeding

Figure 3. Infants 0-4 months who are exclusivelybreastfed and those who are supplemented with a bottle' in Sub-Saharan Africa, DHS 1986-1991

A. ExaluIely bromtted B.Recve botUe-fedhng

100 T- 100

90 90

.60 71 U~~~ 603

* 0 C 50 ~~~~~~~~~50 * IeU~~~~~~~~~~ 7 40 E 01 70402 33 14

Notes: 1. The information on feeding practices is based on thie 24 ho-urs preceding te suxwey.

2- WHO recommends that all children receive only bTeast milk until 4-6 monthis of age.

Source: USAID/MACROXIMPACT.1993. Vuirkion of rnfantr sad Young Children in MaL. Africa Nutrtion Cha4t04oks. Wash0gton, D.C. March. water, cornmercial baby mnilks and foods, or can and does inflict on infat' nutrition status and homemade porridges or cows' mrfikis conumon physiological development is fullly explained in practice (Dimond and Ashworth, 1989). The WHO's publication, She Physiological Basis for situation will continue to deteriorate unlcss time, Infant Feeding (Akre, 1989)- Figure 5 illustrates money, and skills are invested in repromotion of the dramatic increases in mortality and morbidity, the optimal feeding pattern at global and country due to diarrhea and respiratory infections. directly levels, and within families. The consequences of related to suboptimal breastfeeding (Victora, et this deterioration are reflected in increased costs al., 1987). The consequences also tranlate in of health care because of increased prevalence and more births and increased demand for family severity of commnon infant and child illnesses, in plarming services. TWableV illustrates this inVac. particular diarhea and respiratory infections, andI' higher mortality and morbidity rates in developing Breastfeeding also protects againlst ; countries- The harm that early supplementation diabetes; necrotizing enterocolitis; and parasitic,

41 7z heakh sector and iautrtion interventons in Africa

Figure 4. Infants 6-9 monthswho receive no Exclusivebreastfeeding is the orly human other rood other than breast milk in selcted activitythat fulfills-up to the first six monthsof countries in Sub-Saran Africa, DHS life-the three necessary conditions for good 1986-1991 nutrition: food, health, care. Breastfeeding 60 55 significantlyenhances almost all other health 49 n interventions.Its diminutionor absencemay even cancel out the benefits of other interventions. For 40 34 example,in Malaysia,a declinein breastfeeding 3 31_ offset the impact of water and 23 24 improvements(DaVanzo et al., 1985). 20 14 10 6 v) Brctfeeding my&FYs

o Myths, fallacies,and misinfrirmationabout

4 X X ,X infant feeinghave helped to create an artificial c D E feedingculture. Womenare no longerconfident of the capabilityof their ownbodies to be the sole Nbte: 1. WHO recommendsthat begiuning at six source of nourishment for their young. In the months of age, all childrcn should be ven developing world, where custom facilitates solid food in additionto breast milk. 2. Thisincludes liquids. breastfeeding,rejection of colostrum,prelacteal

Source: USAID/MACROIIMPACT.1993. Nurion feeds,and inappropriatesupplementation are now of Inf°r and YoarugCUldren in Mali Afnic evident in many regions (Dimondand Ashworth, Nutrition Chanbooks.Washington, D.C. MLarch. 1987). Medicaltraining textbooks are devoidof any informationon the processof lactation,and fungal, and urinarytract infection.Breastfeeding how to manageit and breastfeeding. In turn and the accompanyingamenorrhea protects health workers have misguided parents and womenfrom breast, ovarian and epithelialcancers mismanagedbreastfeeding so badly that 'not (Cunningham,Jelliffe, and Jelliffe, 1992). Those enough milk," the most common reason for concernedwith health in developing countries may supplementingbreastfeeding in the early weeks,is see protectiononly in terms of infectiousdiseases, now considered an iatrogenic disorder but it is importantto be aware that the less (Proceedingsof UNICEF/WHOworkshop at the commonconditions cited above may also occur in InternationalPediatricians Congress, Rio de poor settingswhere healthservices and families Janeiro, September1992). It is only through can Hl affordto be overburdenedwith 'Westernm extensiveand intensiveretraining of these health as well as the familiardiseases. WHO's CDR workersthat countrieshave managedto improve programmehas estimatedthat about 1.5 million their knowledge,attitudes, and practices and, infantsdie every year from diarrhea and acute subsequently,breastfeeding practices. One of the respiratoryinfection (ARI) caused by inadequate most concise general publications answering or no breastfeeding. questionsor commonmyths and practiceswas

42 Promoting breastfeeding

Figure 5. of Mortality, Brazil

A. Duet. dinha maedoingto feading B. Due to acub prasimy nf_eoo mahod cAording to feeding nitha

20 4 18 ~~~~~~~~~~3.5 16 3 14 ~~~12 ~~~~~~~~~2.5 ! 10 2 _ 12 2i1 £1.5

4 2 2lo ,o .g 6 \iEE 0.52 e 0 0 ;4~~iiijn ~~~IE E g zI!

Fig 13 K;: F_ : m : Fining.a.E r.. met-o

Source:Victora, C. et al. 1987. "Evidencefor protecon by breastieedigagainst infant deaths from infectiousdiseases i Brazil.kLancez. 3(9):322. published in 1991 by the Academy for Educational comparisonwith exclusive artificial feeding. This Development. and othermethodological flaws compromisemuch data, particularly from industrialized countries. There is an importantdifference between full, Consequently, cohorts of truly breastfed babies, exclusive breastfeeding and the wide range of which can match in other variables cohorts of feeding practices that include some breastfeeding artificially fed babies, have been hard to find. or breast milk. Figure 6 elaborates on these Many studies claiming to compare exclusive differences. Exclusive breastfeeding means breastfeedingwith other feeding pattems ignore or nothing should enter the infant's mouth exceptthe discount the use of water and "non-nutritive" breast. fluids (Traore in Hill, 1990).

Early evidence of the advantages of b) EXPEMEmNs IN rHE PROMOfOx OF breastfeeding has been sullied by poor study BREASTFEEDING design: for example, the definition of breastfeeding is an importantstep that was omitted Few assessments in this area are available for from many studies. In too many studies, any Africa; therefore, most examples are taken from breastfeeding pattern from exclusive to mainly other continents. artificial with a few restricted breastfeeds was counted as breastfeeding for the purposes of

43 The health sector and nutritioninterventions in Africa

Table V. Contraceptive Prevalence Required to Maintain Current Fertility Rates If Duration of Breastfeeding Were to Dedine Contraceptive Contraceptive Contraceptive PrevalenceNeeded PrevalenceNeeded Country Prevalence with 25% decline with 50% Decline Ghana (1988) 13% 24% 34% Senegal (1986) 11% 23% 35% Morocco (1987) 36% 41% 47%

Source:"Demogmaphir and Health Survey" data using the aggBegatmodel ftom Bongiarts.John. and Robert Poter. 19S3.Ferriiiy. Biology and Behavior:An Analyszs of the Pxirmte Dnennizew. New York,Academic Press.

i) Awareness campaigns impedimentsto optimal breastfeeding.

Some tried methods of breastfeeding a A plan for ensuring long-term protection, promotion, and support are described sustainabilityof breastfeeding promotion below. A comprehensivereview of breastfeeding campaigns. promotion programs in more than 25 developing and industrialized countries (Green, 1989) a A well-designedmedia plan with messages concludesthat breastfeedingcampaigns have often and media appropriate to the target been geared toward "motivating" mothers to audience(s)- breastfeed by working on their perceptions and attitudes-but have often overlookedother factors n Interpersonal support systems such as that in the end may make breastfeedingdifficult, health workers and counselors, however motivated the mother is. U Sound administrative and financial General messages about the benefits of managementof such campaigns. breastfeedingare ineffectiveif they do not address the major impedimenmsto optimal breastfeeding These principles could well apply to any practice. Media campaignsand other advocacy health interventionprograrn, but breastfeedingis strategies must therefore be an integral part of a particularly dependent on a fliliy integrated broader program. approach.

Green describes small-scale (booklets and Theseprograms use differentmethods ranging posters) and wide mass media (TV. radio, from a simple outpatient card in the Dominican newspapers) initiatives. She states that the most Republic to an imaginativemultimedia campaign successfulhave the followingcommon qualities: in Brazil. Campaigns are most effective if carefully targeted. and pretesting of material and a An overall communicationstrategy based messages is invaluable. What works for one on an in-depth analysis of the main society may be disastrous in another. For

44 PromoingbreaWeadt Figure 6. Percentageor Intrnts 0-4 months In introduced, bottlefeeds were phased out, and selectedAfrican countriesaccording to the type mothers were provided with support and of breastfeeding information.The resultswere a virtualeliminadon 100 of diarrheaand a 95 percentfall in mortality. No 90 cases of lactationfailure bave occurredsince the 80 start of the program. Helpingstaff to breastfeed their own inants was an integral part of the 70 program- *60 The WHO/UNICEFBaby Friendly Ilitiative 6 40 l launched in June 1991 is aimed at accelerating changes in hospital practices. The Intiative 30 - promotes retraining of health workers in 20 breastfeeding management;transformation of 10 - hospitalsand health care facilitiesto implementa set of WHOIUNICEFguidelines to support

' * .= T _ a * T e. T Twomen initiate ad sustain exclusive

a - a , Oc gObreastfeeding,andeliminationofhospitalpractices X= ~ Sthat underminebreasffeeding, promote artificial feeding, and prper ae misinfrmation among

*Em 0 uasti Wid *1 C Chealth workers. Included in the lastcategory is the b,.a.t-nins War practice of receiving and using free and low-cost Source: Dcno5rapbiGHealth Survcys 19861991. supplies of breastmilk substitte. eiample, media advocacyfrom a Braziliansoccer This initiativehas acceleratedcountry-level star was successful,yet a plan to use well-known action to promote and protect breamtfeeding. footballerswas rejectedin Hondurasbecause the Within 18months of its beinglaunched more than pretest group pointed out their reputation as 100 goverments had committd to tdang some womanizer and lack of credibility. action in this area and in particularto prohibiting distribution of free and low-cost supplies of iii) Baby Frendly InihVCe HospitaLs breastmilk substiutes in health care facilites.

Among the best documentedof this type of The criteria for becominga baby friendly interventionis Dr. NatividadReluclo-Clavano's hospital are based on the followingTen Steps to pioneer reform program at Baguio General Successful Breastfeding: Hospital in the Philippines in the 1970s (Relucio-Clavano, 1981 in Jelliffe and Jelliffe, * Have a written breastfeedingpolicy that is 1988). The effort focusedon reeducatinghospital routinelyc icatedto all health care staff who took bottlefeeding, mother-child staf. separation, and baby food industry influence within the hospitalfor granted. Rooming-inwas

45 7Thehnlth sectorand nutritionInterventions InAfrica

* Train all health care staff in skills drop in total infection mortality between 1975 and necessary to implement this policy. 1983 (Mata, 1988).

* Inform all pregnant women about the In Indonesia the Sanglah Hospital project benefits and management of breastfeeding. provided facilities for staff to breastfeed, so they could serve as community models. Formula * Help mothers initiate breastfeeding within industry personnel and promotional materials were a half-hour of birth. banned from health facilities. Rooming-in was established, as well as lay volunteer support of * Show mothers how to breastfeed and how mothers. Mortality from infection dropped from to maintain lactation, even if they should 21.4 per 1000 to 8 per 1000 within six months of be separated from their infants. rooming-ir. Mortality rates due to infection dropped, particularly for acute otitis media (which * Give newbom infants no food or drink fell from 106.9 per 1000 to 8.4 per 1000). other than breastmilk, unless medcally Formula purchase fell from 105.6 to 25.6 tins a indicated. month. Average hospital stay decreased from 3.2 to 1.8 days. These data cover a one-year period * Practice rooming-in-allow mothers and from 1984 to 1985 (Wellstart, 1983). infants to remain together-24 hours a day. In 1986, in a satellite town of Braziia, Taguatinga Hospital, which serves a population of * Encourage breastfeeding on demand from poor worker families, set the following policy: the child. hunan milk only, exclusive breastfeeding starting at birth, and rooming-in. All contact with * Give no artificial teats or pacifiers to formula companies was proscribed. The neonatal breastfeeding infants. mortality rate fell from 25 per 1,000 to 9 per 1,000 in four years and there has been no diarrhea * Foster establishment of breastfeeding since the policy began (De Oliveira, 1990). These support groups and refer mothers to them maternity facility reforms have become the on discharge from the hospital or clinic. backbone of broader initiatives and wherever the plan has been repeated, not only are morbidity and mortality reductions dramatic, but the hospital In the Puriscal project in Costa Rica, a team shows significant cost-savings. What is in a large (San Juan de Dios remarkable is that despite irrefutable and Hospital) gradually introduced the reforms, which consistent evidence of dramatic improvements in included early, exclusive breastfceding. elimi- health and cost-saving outcome, it has proven so nation of prelacteal and supplementary feeds, and difficult to alter practices. rooming-in. Neonatal mortality from diarrhea. sepsis, bronchopneumonia. and meningitis were An evaluation of the impact of the BFHI in its almost eliminated, and there was an 83 percent first two years was planned for 1993.

46 Promoringbreasgfeeding

iii) Trainfngof healthworkFers addressesthe coursecontent and the practicalities, and stressesthe value of one- to two-weekfull- Mosthealth workers have received training in timecourses. Thereis a clear distinctionbetween infant bottle feeding managementthat damages training sessions and informationdissemination their capacityto help mothersto breastfeed. seminarsthat do not impart skills. The latter cannotbe expectedto resultin concretechanges in Health workers often have greater difficulty practices. In Phillips,Feacham, and Mills(1987) than other groups in learningabout breastfeeding an assumptionis made that to educate hospital after their training, and the revelationthat they staff, all that is needed is a week to prepare a mighthave been doingharm to thosein their care pediatrician,who will thenrun a seminarat .hich can be a painful one. It seems appropriatethat each nurse spendshalf a day! This is a serious health workersshould be the best-informedabout underestimateof what is neededto "detrain"the the benefitsof breastfeeding,yet this knowledge pediatrician,then impart some new knowledgeand alonemay havelittle effecton practices. A study demonstrate some of the skills essential to from Nigeria (Ojofeitimi, 1982) showed that supportingbreastfeeding mothers. To assumethat althoughnurses were more knowledgeablethan a doctorwill be equippedto train othersafter only teachersabout the benefitsof breastfeeding,they a week of exposureto this new information,and breastfedtheir owninfants for significantly shorter prior to puttingit in practiceand experiencingits durationthan the teachers. impactis very risky, and a formula for failed trainingefforts. So, just as with media campaigns,training must not be viewedas merelyanother passing on The Phillips model assumes hospital-based of the "breastis best" nostrum,but as a handing training. Eventhen, more time and resourcesare over to health workersthe tools withwhich to do needed (UNICEFBFHI Programmeguidelines, theirjob effectively. Thistraining demands more 1992). thanjust passingon of information;it demandsan interactiveprocess whereby the health worker The IBFANapproach is to organizetraining gains the confidenceand skills to help mothers. sessionsas describedabove, adapt the training An exampleof innovativeand effectivetraining forun to the circumstancesin each country (or comesfrom the Universityof WestemAustralia, region), and have a training team that includes where training of medical students routinely local and outside experts. Team of health includesbreastfeeding mothers in the classroom professionalsare trined at the USAID-sponsored (Harunann, 1990). WesternAustralia now has progranune,Welistart in San Diego, California. higher than nationalaverage breastfeedingrates During 1992a similarservice wasestablished by with more than 90 percent leaving the hospital the Instituteof Child Health in London. The breastfeeding(Hitchcock et al., 1984). Lusophone countries plan to use the Santos LactationCentre in Brazil. One pool of trainingexperience comes from IBFAN Africa. The lessonslearned have been Trainingof trainers is an integralpart of the succinctlydocumented by Helen Annstrong (in process. AmongWellstart, IBFANAfrica, and Labbok and McDonald, 1990). Armstrong now Instituteof ChildHealth, an importantgroup

47 The health sector and nutritioninterventions in Afnica of experts (master trainers) in this subject now providelactation support staff in healthinstitutions exists in Africa. This justifies an investmentin to reeducate the existing health workers rapid capacity-buildingwithin the region. Within (Anmingsjhklpen,1990). each countrythere shouldbe a LactationTraining and ResourceCentre. The master trainers should iv) "Mother-to-monther"networks be seconded or recruited to work full-timewith these centers. When this is accomplished, rraining of lay counselorsis a sensitive area progress in the whole area of breastfeedingis because care needs to be taken to ensure that the accelerated. Trainingcurricula for these courses, lay counselor and health worker feel a sense of for all levels, are being distributedby UINICEF complementarityand mutualsupport. This is not and WHO as part of the Baby FriendlyHospital always the case, especially if informational Initiative. support to the mother is contradictory. One of their most importantcontributions is in the use of The necessityfor a prolongedcourse lies as counselingtechniques in breastfeedingsupport. In much with the emotional content as with the Belize lay counselorsplayed a key part in the quantity of information to be absorbed. successfulnational breastfeeding campaign (at four Armstrong identifies four phases of a course: monthsbreastfeeding rates rose from 18 percent in "ennui," when people question the need for the 1983 to 49 percent in 1989). Of i0 counselors course at all; 'resistance," when challenge to trained, 67 percent were communiitywomen and current practicesarouses strong negativefeelings; the remaindernurses, conmunity healthworkers, "absorptiop." when the content engages the and TBAs. Despitevolunteer status, the retention participants and they warm to each other; and rate vds 62 percent. Average work load was "looking forward," when the inspiration of the about eight bours a month, helpin: an average of course motivatesparticipants to seek and plan of nine women. Counselorsmet three to six times a ways to change. This necessarygroup dynamic year for trainingand support(Huffinan and Steel, cannot be achieved in a short course or in a 1992). course of weekly seminars. Kyenkya-Isabiryeand Magalhaes(in Labbok Training health workers as part of and MacDonald, 1989) describe the historical breastfeeding intervention programs will be developmentof mother-to-methersupport and the necessary and difficult as long as bad basic groups involvedin its provision, and propose its trainingpersists in medicaland healthschools and adaptationto healthcare services. textbooks, and that incorrect information materials. sometimes funded by baby food In many societies,female relativesand TBAs companies, continue to be readily available. have given the support, encouragement. and Currently, no integrated program can function advice. In industrializedsocieties the role of the without prioritizing the retraining of health lay counselor emerged from the mother support workers. It is of note that in Sweden,where basic groups founded by breastfeedingwomen in the training and breastfeedingpractices are relatively 1950s. '60s. and '70s. Their influencehas been good. the governmentstill deems it necessaryto important. The turnaboutin breastfeedingtrends

48 Promotingbreastfeeding in industrializedcountries has been attributed to similar disaster situations trigger commercial their infldence more than to medical edict exploitation. It is therefore imperative to view (Morley, 1980; Helsing in lelliffe and Jelliffe, situationswhere artificial feeding is taking place 1988). La Leche League in the United States, as in need of extra vigilance in terms of Amningsjhlilpenin Sweden (and kindredgroups in implementingmarketing regulations. the other Scandinavian countries), The Nursing Mothers of Australia, the National Childbirth In Africa, programs to improve infant feeding Trust in the U.K. and many more have provided practices need to include efforts to promote a pool of information and support. A key breastfeeding through development and characteristic of these groups is their dynamic enforcement of national marketing regulations approach to new information and their ongoing (IBFANAfrica, personal communication).IBFAN learning from mothers and babies. has the expertise and commitment to assist governments in this effort, and UNICEF and v) Monitoring adverdsing for breast mik WHO are mandated to work in this area. substitutes Establishing sustainable monitoring and enforcementmechanisms is essential and has not The InternationalCode of Marketingof Breast been successfullypursued. Th-bcewith skills do milk Substitutes (WHO, 1981) remains the basic not have adequatefinancing to act. One exception and minimum standard to guide countries in is Mauritius, where an independentorganization regulating the way infant feeding products (infant has taken on this task over the last seven years, formula, breast milk substitutes, and feeding and the process continuesto benefit breastfeeding. utensils) are marketed. The code containsarticles on how products should be labeled, displayed, Papua New Guinea took measures to protect sold, and stored. The code also specifies what breastfeeding before the concept of an healthworkers and manufacturersand distributors international code had ever been publicly of products under the scope of the code should discussed. Between 1962 and 1972 breastfeeding and should not do. The code is a recommendation rates in the capital city, Port Moresby, declined to memnberstates to take action to implementit. from 94 percent to 78 percent and hospital admissionsand deaths due to diarrhea doubled. A One key point regarding the code: its main 1976 survey found a significant difference in goal is protecting breastfeeding,but its attendant malnutrition rates (35 percenrt and 26 percent goal is protecting infants who must be fed between bottlefed and breastftd children. The artificially. It is in the best interests of the infant govex. :.ent, the health professions, and a mother and his caretakers that the choice of feeding support group (Susu Mamas)mounted a campaign method and product is impartial. The current targeting health workers, the conunuity. and HIV is leadingto an increase in children schoolchildren. There was widespread mnedia who are not being breastfed, either because of coverage. Traders were asked to voluntarily ormhanhood(and no readily availablewet nurse), restrict the sale of feeding bottles but did not decision not to breastfeed, or inability of the comply. Legislation was passed to control the mother to care for her child (including sale and distributionof breast milk substitutes and breastfeeding) because of illness. These and to require a prescription for the sale of feeding

49 The health sector and nutnrtioninterventions in Africa bottles and teats. Health workers were required adopted as a minimum standard. The few to ensure that conditionswere appropriatebefore countries that have achieved more or less full writing a prescription. Advertisingof all these implementationhave shown good results. Sweden products was banned. By 1979 breastfeedingin and Norwayhave more or less effectivevoluntary Port Moresby had increasedand malnutrition rates codes; breastfeedingrates have risen since code fallen dramatically. adoption but in the context of a multifaceted strategy. The Code DocumentationCentre, based The Papua New Guineaexperience has shown in Penang, Malaysia, specializes in training how a simple set of initiatives can protect government officials in code analysis, breastfeeding and reverse a trend. Key factors development,enforcement, and monitoring. The appear to be coordination between different center has multilingual staff and has provided sectors, political will for implementationof the support and technical assistance to a number of law, and good behavior on tie part of the baby governments preparing national marketing food companies. In contrast to many other regulations. Recent examples include India, countries, there were no entrenched vested Nepal, Brazil, Swaziland, and Bangladeih. interests and the companies had no established market to lose (Biddulphin Jelliffe, 1988). Also, Othercountries that have establishedeffective the emphasis of restriction focused on feeding codes are Brazil, Guatemala, and Kenya. All bottles, teats, and pacifiers rather than on infant three have madc code legislation and milk products. implementationpart of a coordinated national breastfeeding campaign and all have shown Recent events show that breastfeeding improvements in breastfeeding trends. All protection should be continuous, as long as there countries suffer from external pressures. is money to be made through the sale of Guatemala is particularly vulnerable to TV substitutes. A 1989 report showed a reversal of advertisingfrom neighboring countries. the successful trend, with a 60 percent increase in artificial feeding in the Port Moresby sample. The United States has the most aggressive Flouting of the legislation has occurred mainly marketing. There is public (including TV) because of a lack of monitoring. Employed advertisingof baby milks, free offers and samples mothers find insufficient time allowed for are sent directly to pregnant women, and breastfeeding. Pacifiers are on sale and external "dischargepacks" are common. After Carnation TV advertising cannot be controlled (Minei. (Nestle) and Gerber launched TV advertising 1992). This erosion of effectivenessindicates the campaigns. there was a 15 percent increase in need for vigilance and strengtheningof existing overall sales of artificial baby milk and a 15 measures. The problem of external advertising percent fall in breastfeeding rates (Lawrence, can only be addressed at the internationallevel. 1991). There have been various breastfeeding promotioncampaigns in the United States (Green, Despite 11 years since the code's adoption. 1989)but the trend is diminishing. Reports from there are very few countries where it is the Brazilian campaign emphasize that implemented in its entirety even though it was

50 Promotingbreasfeceding promotionalmessages can only work if industry there is planneddiscrimination against mothers, promotionis controlled(Rea, 1990). employment conditional on sterilization, or long-tenncontraceptive injections (Mitter, 1986). vi) Rightsof mothers Claiming the right to breastfeed at work is impossiblefor mostwomen unless their requestis Measuresexist to provide entitlementsfor backedby governmentsupport. employedmothers. VanEsterick's Women, Work and Breasffeeding(1992) provides a current Thereis no evidencethat breastfeedingin the reviewand analysisof this aspectof the subject. workplacehas everdisrupted production; there is What scarceevidence there is revealsthat where even evidence that when breastfeeding is breastfeedingis valued,it is accommodated.Van facilitated,workers return to their jobs early Esterickcites data from Mali showinga society (Ketcherand Lanese,1985). Van Esterick lists where it is taken for granted that babies advantagesfor the employer,including earlier accompanytheir mothers everywhere-to the return,less absenteeism,fewer training costs, less office,the fields,or on a train. employeeturnover, and betterproductivity due to lowerworker anxiety. There is great regionalvariation in national legislationon maternityprotection. The Clearing Provisionof crechesalone is not the answer Houseon InfantFeeding and MaternalNutrition to enabling women to combine work and (APHA, 1989)reports on the current status of breastfeeding. regulationscountry by countrybut does not give a pictureof implementaiin. It is oftendifficult A comparison of two factories in for womento claim their rights. Unionsare too Mozmbique found that one, a cotton factory, concernedwith job protectionand fair pay to deal had a pleasant-lookingcreche where babies were withsomething seen as a minorissue. Brazilhas bottlefed. Motherswere forbidden to leavetheir some of the best legislation; however, posts to breastfeed and high rates of infant enforcement is a problem. As with infant malnutritionand diarrhearesulted. At a cashew products marketingregulations, mechanisms to factory, with less attractivefacilities, mothers enforcematernity entitlements are essential. wereallowed to organizetheir work to breastfeed and their children stayed healthy. Mothers TheILO Conventions(Nos. 3 and 103)which workedin pairs so as not to disruptthe machine lay down standards for maternity protection managementif one needed to feed her baby (includingthe right to two paid nursingbreaks a (Zinkin, 1983). A similar situation was day),are ratifiedby only25 countriesand are not encounteredby one of the authorsin Mauritiusin promoted. In Honduras enterprisesemploying 1985. more than 20 women must provide a suitable place for breastfeedingand in Uruguay public At CanipinasUniversity in Brazil all female sectorworkers can work half-timefor sixmonths. employeeswere entitledto free day-careas long Most poor womenwork in the informalsector as they were breastfeeding,which positively where there is little controlover their working increasedduration. It is pertinentthat the dean conditionsand rights. Even in the formalsector wasfully committed to thisproject (Hardy, 1990).

51 Tlh health secrorand nutritioninJerventions in Africa

In Ethiopia, at a fruit-growing cooperative, Including breastmilk in the list of foods women left their children at a day-care center produced in the country is likely to start a process administered by a Children's Affairs Committee of assigninga (money) value to this valuable food of the coop. An innovative feature of the resource. When the cost of its replacement is program has been to provide work credits to calculated, countries that have to import infant women for time off to breastfeed. This, together feeding products are likely to choose to institute with other aspects of the day-care center, have means to ensure breast milk's continued been linked with a reduction in disease-related production, including adequate maternity deaths (Hargot, 1989). Creches are one of the entitlementsand resources. most practical ways of facilitating continued exclusive breastfeeding. In a study in a large vii) Integraing breastfeedingpromotion into health institution in Nigeria, half the women other heath programs interviewedsaid they would have breastfed longer if there had been a creche (Bamisaiye and Promotion of breastfeedingshould be part of Oyediran, 1983). very important measures aimed at preventing diarrhea and reducing the rate. A practical and imnediate action for programn managers is to routinely review how any proposed Examples and experiences of how program or project is affecting or will affect breastfeedingpromotion and support can and have participating women's capacity to breastfeed and been integrated into other health programs were to care for their young infants. Such a review shared at a UNICEF workshop in 1988. The should address the short- and long-termeffects of proceedings are documented in a UNICEF the possible infant feeding options available to the publication edited by Baumslaug and distributed women participating in the program. The review by UNICEF headquarters. Programs include should lead to adjustmentsin the program to make inmnunization,control of diarrheadiseases, family exclusive breastfeeding for zero to four-to-six planning, safe motherhood, and growth months and continued through two years the promotion/monitoring. easiest and most practical option. c) Cons This should apply to any program or project whose objectives are to improve health and the There is little information on the cost of status of women, to generate income for families breastfeeding. Phillips et al. (1987), however, and individuals, to improve education, etc. Cost estimated the cost of a "packageof breastfeeding analysis of the above option should go beyond the promotional ac.tivities," which includes: i) economic interests of the goverrfnent, and changing hospital routines; ii) individualized employer and include those of the families and instruction;iii) a media promotioncampaign; and communities as well as the total well-being of iv) legislation to limit the commercializationof mother and child. breastmilk substitutes at between $1 and $10 per mother exposed to this "packageof activities."

52 Dietary managementof infecdon child illness. The fonnative research included 5. Dietarymanagemen of infection ethnographicstudies, nutritional assessments, and household trials of new behaviors and food recipes.The resultinginterventions focused on a) BACKGROUND providingmothers with specific 'guidelines on feeding frequency,food quantities,and recipe Illnessadversely affects children's nutritional preparations.This informationwas intendedto well-beingthrough increasednutrientrequirements buildmothers' self-onfidence and it was used to and losses,and child-drivenreductions in dietary encouragemothers to uke an activerole in child- intake(anorexia). In Africa, the effectof illness feedinginteractions. on nutrition is exacerbatedby children's poor nutritionalstatus, due to large deficitsin their Results of household hials indicated that nutrient intakes during convalescence and mothers were willing to change their feeding post-recovery, and mothers' reluctance to practicesif they perceivedpositive benefits for encouragetheir childrenactvely to eat. theirchildren and themselves. The most frequently accepted behavior changes involved small Mothers' reluctanceto take an activerole in modificationsof existing practices, such as child feedingstems from a traditionalview that enriching a traditional weaning porridge, or learninghow to eatthe family'sstaple food is part increasing the quantity and/or frequency of of a child's socializationprocess. It is generally feedingother solid foods. However,when new believedthat the purposeof eating is to fiUlthe foods or practices were adopted they often stomach,and that a childknows best whenhe is replacedrather than complementedthe traditional hungry and when he is full. Providingtoo mch diet. Obstaclesto trying new behaviors included guidancemay spoil a childand causehim to be perceivedtime and other resource constraints. greedy for foods that are not available to Mothers'continuation of the new practiceswas everyone.When children are ill, however,African usuallydetermined by their children'sreactions to mothers are usually wiling to prepare special them. foods and more activelyencourage their children to eat. Therefore,future interventions should stress the benefitsto mothersand childrenof adopting Therefore, periods of illness and improvedfeeding practices, and providemothers convalescenceare seen as opportunetimes to with alternativemethods for overcomingtheir provide informationand to encourageAfrican children'sresistances to new foodsand behaviors. mothersto adoptnew feedingbehaviors. Recommendedbehavior changes should build on existing practices. The importance of b) CONCLUSIOISAND RECOMMEIV7DTJONS breastfeeding and continuing traditional beneficialpractices should also be stressed. * Theprograms reviewed in thisdocnent all includedintensive community-based (formative) * Once programs moved from formative researchprior to the design of interventionsto research (i.e., householdtrials) to implemen- improvefeeding practicesduring and following tation (i.e., conmmnity-basededucation) their

53 7he health sctor and nutrition intervenions in Africa

results have been less encouraging.For exanple, examine changesin behavior over the appropriate mothers in Cameroon had improved knowledge (long-term)time intervals. but there were few measurable improvementsin feeding practices after less than one year of Therefore, programs should incorporate the program implementation.Within two months of methods employedduring the householdtrials (of being trained, more than 50 percent of mothers in .ndividudelizt'd care, decision-making, and Nigeria knew how to prepare eko ilera, but less negotiation) into their strategies to change than 20 percent indicated that they would prepare mothers' feeding practices. Program evaluations and feed it on a continuousbasis. should be designed, and adequate resources allocated,to measure changesin feeding practices The failure of programs to live up to the and nutritionalstatus over time. promise suggestedby the householdtrials, and to result in changes in feeding practices, is believed U Developingeffective programs to improve to be due to a combinationof factors. Mothers feeding practicesduring illnessand convalescence were willing to adopt new practices during the requiresknowledge of health providers', mothers,' trials because of the individualized care and and other caretakers' beliefs and practices, attention provided by the field-workers, and availableand acceptablefoods, and the nutritional because they were active participants in the quality of the local diet. Research should also process of deciding what behavior changes to determine feasiblefeeding changes, how mothers adopt. can be motivatedto adopt them, and how they can overcomeresistances from their children to pursue During implementation,however, the intensity them. During this research, it is important to and personalizednature of these interactionswere understandthe health personnel's attitude towards not sustained. Mothers who received individual nutrition education, as well as to gather counselling were more likely to have improved information to improve their group and feeding knowledge, yet this knowledgemay not interpersonalcounselling skills. have resulted in the adoption of new practices without mothers' active participationin deciding Therefore, involving health providers, what those practices should be. mothers, and children in the intervention design process, through careful formative research, is In addition to the above explanation, it is feasible in Africa. The goals of this research are generally believed that changing behavior in a two strategies: one for improving mothers' populationis a long-term process. Whereas some feeding practices and another for enhancinghealth members may adopt a new practice imnediately providers' nutrition educationskills. after it is introduced, there are others who will accept it only after it is already well-establishedin U To date, nutrition programs have not the community. The programs examinedmay not focused on improvingmothers' feeding practices have produced measurable changes in feeding and children's dietary intake during periods of practices because evaluation designs did not convalescence - when appetite returns and compensatory(catch-up) growth is possible - or

54 Dietarymanagement of infection on developingsustainable strategies for engaging to increasesin basalmetabolic rate and require- mothers in pro-active child feeding practices. menufor immunecell production(Akre, 1989). These two areas merit further attention. At the same time, nutrients are lost during the sweatingcaused by fever. Therefore, as a starting point, future nutrition programs should encouragemothers to improve - Diarrhea has reported average incidence child diet during convalescence. Such efforts rates of 4- to 8 episodes per year in African shouldtake advantageof mothers' concerns about childrenunder 5 years of age. Incidenceis highest child feeding during illness, as well as their (up to 10 episodesper year) in the first two years willingness to take special measures to increase of life (PRITECH, 1989; 1993). Diarrhea results their children's dietary intake during this time. in decreased nutrient absorption (mainly carbohydrate and fat) due to mucosal injury, Health services could easily implement reductions in the concentrationsof bile salts and measures on dietarymanagement of infections,but digestiveenzymes, and increasedgastrointestinal also on dietary prevention of infection as transit times. illustrated by Table VI (adapted from Gillespie and Mason, 1991). - Measles: in Sub-Saharan Africa it is estimatedthat less than 60 percent of infants have c) OVERVIEWOF THE PROBLEM been imunized (ACC/SCN, 1992). Measles causes mucosal injury, impaired nutrient i) The nutriionalconsequences of infection absorption, immunosuppression,and often results in severe diarrhea, with its attendant nutritional The relationship between nutrition and costs. infection has been described as synergistic: frequent infections reduce the nutritional - Intestinalparasites, includinghookworms, well-being and growth rate of young children, schistosomes, ascaris, and whipworms infect while malnutritionsuppresses host resistanceand millions of African children and adults. They results in more severe infections. affect nutrition by competing for their host's nutrients while producinv fever, diarrhea, * Common childhood illnesses such as vomitingand anorexia, altenng nutrient synthesis diarrhea, respiratoryinfections, malaria, and other and transport, impairing the absorption of fat, parasitic infections adversely affect nutrition protein, and vitamin A, and causing blood loss through increasedmetabolic demands and nutrient and iron deficiencyanemia (Stephenson,1 987). requirements, decreased nutrient absorption, and/or decreased dietary intake due to anorexia * 'Although infrequently documented and (depressedappetite). rarely quantified, perhaps the greatest nutritional consequenceof childhoodillness occurs as a result - Fever is common during episode 3 of diar- of reductions in dietary intake that accompany rhea, respiratory infection, and malaria. It and followthe frequent infectionsdescribed above increases the body's energy requirementsby 10- (Bentleyet at, 1993). The reductions may occur 15 percent per [° C rise in body temperaturedue when mothers deliberatelyrefrain from feeding a

55 The health senor and nutnrtioninterventions in Africa

Table VI. Examples of actions that c- be implemented by health services and of those necessitating additional resources Actionsthat can easily be implementedby Actions that require additional health services recources Dietary managementof a Promote exclusive breastfeedingduring * Provide supplementary infections illness; feeding; * Maintain sufficientfood intake during a Control intestinalparasites. illness; * Give vitamin A during mcasles, acute respiratory infections,and persistent diarrhea; * Provide Oral RehydrationTherapy; * Maintain adequatefood intake during chronic illness; * Give iron and anti-malariadrugs during malaria episodes; * Provide Growth monitoringand promotion.

Dietary preventionof U Give colostrumto the baby; U Give supplementaryfeeding infections U Exclusivelybreastfeed for 3 to 4 to the child and the mother; months; U Establish conmnunity * Continuebreastfeeding for two years; programs for growth monitoring * Provide growth monitoringand and promotion. promotion; * Provide pre- and post-natalmonitoring of mothers; * Give vitamninA as a preventive measure in highly deficient areas;

Source:Adapted from Gillespic, S.. andJ. Mason.1991. Nutrilion RelewvntActions:rpericncefrom the Eightiesand Lessons far dte Nineties. et de ACCISCN. 1991. Some Optionsfor ImprovingNutrition in te 1990s. child who is vomiting, or when they refrain from pological studies throughout Africa. In general, encouraging an anorectic child to eat when he or becoming accustomed to food and learning to eat she is not interested in doing so. The nutritional the family diet is an important aspect of childhood consequences of these reductions are especially socialization in Africa. Mothers may initiate great in parts of Africa where children's usual feeding when their children exhibit cues of hunger intakes during healthy days are well below their (e.g., looking at or reaching for food), but estinated requirements (e.g., 60 to 70 percent in children are usually assumed to know when they Dickin et al, 1991), and thus no opportunities to are full and thus when eating should stop. replace nutrient losses or attan catch-up growth are possible- In times of illness as well as health, children in the diverse populations of Cameroon (CARE a The relative importance of mothers' reluc- et al, 1989), The Gambia (Samba and Gittelsohn, tance to encourage their children to eat - and the 1991), Ghana (Ministry of Health, 1989a), Niger influence it has on children's nutritional (Keith, 1991a), Mali (Toure. 1991; Dettwyler, well-being - is suggested in numerous anthro- 1989), and Swaziland (National Nutrition

56 Dietarymanagemem of infection

Council, 1988) are observed to regulate their (Coulibaly,1989). In-depthstudies of the nutrition dietary intake by deciding when, what, and how messagesgiven to mothers found that providers' much they will eat of the foods and liquids they advice was usuMly vague and non-specific. are offered. With a few notable exceptions, Resultsof similarstudies in Senegal (Sene, 1991), including Yoruba-speaking mothers in Nigeria Niger (Youssouf, 1992), and Burldna Faso who forcefeed young children who are unwilling (Roisonet al, 1990) were comparable. to eat (Oni et al, 1990), mothers and other caretakers throughout Africa generally do not * Qualitative assessments of nutrition actively encouragetheir childrento eat when they communications skills in several countries, refuse or show no interest in food. including Burldna Faso (Duran-Bordier, 1992) and Cameroon (CARE et al, 1989), found that ii) The quality of and health workers have difficulty using educational indiviual counselling materialsand visual aids duringgroup discussions. In addition, health promoters were found to * Evaluationsof programs to improve child require practicaltraining in how to listen, develop feedingpractices during and followingillnesses in solutions, and provide tailored feedback to Africa have underscored the need to improve mothers, especiallyin areas where these types of health providers' knowledge concerning proper exchangesare uncommon. When this was done in nutritional management of illnesses as well as Mali, the results were impressive: mothers their interactivecounselling skills. residing in villages where health promoters were specificallytrained in interactivecounselling skills in Uganda, for example, 80 percent of health had better feeding practices and fewer workers surveyed (N=35) believed that solid underweight children than mothers residing in foods (e.g., sorghum, millet, potatoes)and fluids villages where health workers were not similarly (e g. , fermenteddrinks) shouldbe withheld from trained (Rohdeet al, 1993). children with diarrhea (Konde-Lule et at, 1992). These beliefs were contrary to the nation's iii) Conclusions Control of Diarrheal Diseases (CDD) policy that feeding should continue as usual during diarrheal U Thus, programs to improve feeding episodes. practices during and following childhood illness must develop strategies to improve the dietary In Mauritania, a survey of health facility intake of children with depressed appetites. personnel in charge of diarrheal disease control Effective educational messages can only be and counselling(N = 243) found that only 62 and developedwith an understandingof local feeding 49 percent of CDD workers knew program practices and diet, as a well as mothers' messages about feeding during and following motivations and resistances to changing their diarrhea, respectively. During 91 clinic feeding style and behavior. Health providers' observations, 76 percent of the CDD workers knowledge, beliefs, and nutrition education gave "advice to mothers on feeding following practices must also be studied in order to identify diarrhea, but only 33 percent (N =30) verified training needs and improve their community that mothers understood what they had said' educationand interpersonalcounselling skills.

57 The health sector and nutrition interventionsin Africa

This chapter describes the design and impact kcal/kg per day after recovery from prolonged of programs that have attempted to address these diarrhea, compared to average daily intakes of 70 needs in Africa. kcal/kg during their illness (Tomkins, 1 983).

iv) The traditional dietary management of Episodesof illness are reported to cause great the sick and anorectic child in Africa concern amongAfrican mothers, and mothers who are normally passive feeders may take on a more Programs to study and improve the dietary active feeding style when their children are sick. management of the sick child have found that In Mali, for example, 80 percent of mothers while African mothers may hold specific beliefs reported that they would encourage a sick child to about foods to avoid during specific illnesses eat, compared to only 45 percent who said they (e.g., Konde-Lule et al, 1992; Odebiyi, 1989), would similarly encourage a healthy child (Toure, they do not generally stop all solid-feeding, or 1991). On the other extreme, Yoruba-speaking refrain from providing most liquids and breastmilk mothers in Nigeria were observed force-feeding (a during childhood illnesses (Bentley et al, 1991; fermented pap) during 71 percent of days when Keith, 1991a; Sene, 1991; Spain, 1991; Toure, children were ill with diarrhea, compared with 50 1991; Dettwyler, 1 985). As noted above, it is percent of days during convalescence and 33 usually the child who regulates his or her intake. percent of days during post-recovery (Brown et a), 1988). In Nigeria, where energy intakes during periods of health, diarrhea, and convalescence Mothers in several African countries have also were compared, 5-28 month old children reported that they prepare special foods for their consumed less solid food but similar quantities of children when they are sick. In the Arua and breastmilk and liquid paps during diarrhea Kabale districts of Uganda, for example, mothers compared to days of health. Total energy intake, prepared a fermented cassava, millet, or particularly from solid foods were not increased sorghum-based food for children with diarrhea. during convalescence (Dickin et al, 1991), as is These foods' distinctive taste was believed to help currently reconmnended by WHO (1992). restore the appetite and thirst of anorectic children (Sserunjogi and Tomkins, 1990). Fermented millet Likewise, no concepts of convalescent feeding and sorghum porridges are also specifically fed to - i.e., the need to increase intake following sick children in Rwanda and the Sudan illness - or catch-up growth were reported by (Ashworth and Draper, 1992). mothers in Senegal (Sene et at, 1 992) , The Gambia (Samba and Gittelsohn, 1991), Mali One of the greatest obstacles to reducing the (Toure, 1991), Niger (Keith, 1991a), Ghana nutritional consequences of childhood illness in (Ministry of Health, 1989d), Cameroon (CARE Africa is the fact that the reductions in intake that et al, 1989), and Swazland (National Nutrition occur with infection are small in comparison to Council, 1988) where feeding practices during and the deficits observed on healthy days relative to following childhood diarrhea were studied. In The age-specific dietary requirements. In Nigeria, for Gambia, however, children consumed up to 120 example, children's average daily energy intake

58 Dietay managementof infecion during diarrhea was 85 kcalAkg,an 11 percent v) Progrms to inprove dietay intake and reductionfrom their average intake of 96 kcal/kg feeding practices in Afrca on healthy days. Consumption on healthy days, however, was less than 70 percent of the As noted earlier, efforts to reduce the children's estimated daily energy requirements nutritional consequences of childhood illnesses (Dickin et al, 1991). must be based on an understandingof local beliefs and practices, as well as individuals' motivations When this situation occurs it is generally and resistances to changing behavior. Projects believed that the ideal approach to reducing the includedin the review were selectedbecause they nutritional consequences of infection is first to studied these conditions prior to developing focus attention on improvingchildren's usual diets specific interventions. and mothers' day-to-day feeding practices (Behrens et al, 1990). By improving intake on a Although the projects were carried out by a daily basis, children's overall nutritional status variety of institutions ranging from university should be enhanced and the severity and research centers to govermment ministries to nutritional impact of subsequent infections private non-governmental organizations, each reduced. project employeda combinationof qualitativeand quantitative research techniques in roughly five Encouraging African mothers to take a more stages: active role in child feeding on a daily basis may be difficult, however, given financial, time, and U A review of pre-existing information on food availability constraints. In addition, and as feeding practices, diet, and illness beliefs, noted above, the process of learning to eat the practices, and . family food is considered to be an importantpart of children's socialization - the time when they U A brief (6-8 week) or in-depth (up to 1 learn to become independentand satisfiedwith the year) ethnographic study of health providers', food that is available (Keith, 199la; Toure, 1991; community members' , and mothers' health, National Nutrition Council and The Manoff illness, and nutrition beliefs and practices, and Group, Inc., 1992). their sources of infonnation on the same.

Although many African mothers may be U A nutritional assessment of existing diets reluctant to modify their daily feeding practices in and practices, their potential for enrichment or fear that their children will become selfish or improvement,and possible resistances or obstacles greedy, most appear willingto ukrespecial steps to to improvingdiet quality and feeding practices. insure that their children eat when they are ill. Thus, programs to improve children's diet may * Individualand group trials of new feeding therefore find it beneficialto use times of illness practices, foods, and recipes to determine the and recuperation as opportunities for providing most feasible alternatives for improving dietary mothers and other caretakers with knowledgeand intake and peoples' reactions to new products skills to improve child diet and feeding practices- (e.g., foods and recipes) and behaviors.

59 The health secrorand nutition intervenrionsin Africa

* Development of an overall strategy for growth monitoring sessions with copies to take improvingchild feeding in the population, based home, and general information posters for on the findings from these "fornative research" communityviewing. activities. Anirnateurs visited each project village All of the projects described below were monthlyto lead a discussiongroup and, with the carried out by multidisciplinary teams that assistanceof local midwivesand mothers, to carry included social scientists, health professionals, out a cooking demonstrationon how to prepare nutritionists, and community members. Projects the traditional weaning pap enriched with either focused on improving the feeding and diets of milk, egg, or peanut butter. During a second visit healthy as well as sick children. Many of the the aninateurs held growth monitoring sessions projects are still in progress and informationon with individual-levelnutrition counselling. Home their nutritionalimpact is not yet available. visits wverealso scheduled for children who were moderatelymalnourished (low weight-for-age). Cameroon RESULTS: A post-intervention AP survey was DESIGN: This projec. was carried out from 1985 carried out by Tulane University in 23 to 1989 in 37 villages of the Extreme North interventionand control villages after six months Province by CARE/Cameroon, in collaboration to one year of program implementation. The with the Center for Nutrition (IMRMP), the evaluationfound increasedcoverage and mothers' Ministriesof Health and Agriculture,and USAID. participation in nutrition-related activities and Technical assistance was provided by the Manoff improvementsin mothers' nutrition knowledge in Group, Inc. and the Educational Development interventionconmmunities. Knowledge scores were Center (EDC) under The Weaning Project. highest among mothers who received individual counselling. A significant increase in the percent The methodologydescribed above resulted in of modhers feeding enriched porridge to their the development of an education and training children(from 44- percent to 50 percent) was also program designed to improve feeding practices of observed in some villages (p < 0. 02). children less than 36 monthsof age, using locally available resources. Emphasiswas also placed on Specific knowledgeand attitude gains related improving the communicationsskills of CARE's to the dietary management of the sick child communityhealth workers (animateurs). included an increase in the number of mothers who 1) knew to feed-more followingillness, from Educationalmessages, targeted at parents and 39 percent at baseline to 55 percent at follow-up advising them on the appropriate diet, feeding (p < 0.0001), 2) knew to continuefeeding a child frequency, and meal compositionfor healthy and with diarrhea, from 86 percent to 93 percent(p c sick children of different ages (0-3, 4-9, 10-15, 0.0001), and 3) felt they would be able to and 16-36 months), were developed. Educational encouragea sick child to' eat, from 74 percent to materials included picture cards for use in groups 82 percent (p < 0.005). The small sample size discussions, counselling cards for use during and limited duration of the interventionprior to

60 Dietary mawgememof Infection

the survey were noted as weaknesses of the The project followed the five basic stages evaluationdesign that limitedmeasurement of the described above. Because the original intention program's potential imnpact(Tulane University et was to develop a food to be used during diarrheal at, 1989). episodes,a survey of feedingpractices (N-2655), market surveys of available foods (N = 4), From an institutionalperspective, the positive longitudinaldiarrheal surveillance and weighed benefits of the project were that CARE staff dietary studies (N = 45), and randomizedclinical learned research and inter-personal trials measuring the acceptabilityand digestibility commnunicationsskills that they continueto use in of various foods (N = 69) were also performed. their daily work. They leamed to appreciate the importanceof involvingprogram beneficiariesin These studies resulted in the development of pre-intervention design (formative) research. an educational program to promote the Supervisory staff also learned how to plan, consumption of a fbrtified pap (eko iCera) to manage, and monitor educationalactivities. The complementbreastmilk and other foods that were need to streamline and simplify the formative regularly consumedby children less than 3 years research, particularly if it is to be carried out by of age. The pap was made using the traditional persons untrained in the social sciences, was also fermented maize or sorghum paste (ogi), and indicatedduring the project (CARE et at, 1 989). addingtoasted cowpea flour, palm oil, sugar, and locally-germinatedsorghum (malt)flour to reduce COSTS:The costs to mothers of implementingthe viscosity. eko itera had a total energy density of proposed feeding practices changes were not 85 kcal/100 g (when cooked), compared to 25 calculated. The total budget for the project, kcal/100 g for the traditional ogi (Guptill et at., however, was estimated at $700,000 for five 1993). years. This sum included$500,000 for all external technicalassistance (staff time, travel, per diem, Comnunity health workers (CH2W) were communications,evaluation expenses, overhead, trainedby the projectstaff to prepare the eko ilera etc.), and $200,000 in local expenses (field ingredientsand recipe, and to use the supporting workerstravel, per diem, research costs, materials educationalmaterials. The CHW's then trained 10 development and production, etc.) (Griffiths, teaching mothers, who selected and trained 10 1992). additional mothers of children less than 3 years old during cooking demonstrationsand follow-up Nigeria visits. Mothers were given a three-day supply of cowpeasand sorghum (Esrey et al, undated). DESIGN: This project was carried out from 1986 to 1989 in 11 communitiesof Kwara State RESULTS:Trained mothers were able to prepare by the University of Ilorin in collaborationwith the recipe properly, using at least 90 percer.t of the Johns Hopkins UniversityDivision of the ingredients in appropriate quantities. Sugar, Human Nutritionand the Academy for however,was added at less than the reconmended EducationalDevelopment (AED) as part of the amount, and 12 percent of mothers did not add AMD-fundedDietary Managementof Diarrhea any sugar because it was believed to cause jedi Project (DMD). jedi, a form of diarrhea. Mean energy density was

61 7hehealth sector and nutritioninterventions in Africa

91 percent of the original recipe (Cuuptillet al, contintueto prepare and feed the recipe. Mothers 1993). who completed their primary education, who worked as wage eamers, and who had children Knowledge, trial, and 2doption of the eko who were not yet able to feed therselves were ilera recipe were evaluated among 295 mothers also more likely to be classified as eko ilera within two to eight weeks of their training, and adopters. Adoption rates declined, however, with among 301 mothers who were not trained by time since the last cooking demonstration (Guptill teaching mothers. The evaluation found that 57 et aL, 1993). percent of trained mothers new how to prepare eko ilera (all ingredients, quantities, and cooking Evaluation of dietary intakes among 43 steps), 48 percent knew the recipe and had tried it children living in intervention communitiesand 45 at least once using the ingredients provided by ibe children living outside the project area project, and 17 percent knew the recipe, had ("controls") found no significant differences in prepared it at least once with homemade total energy intake between the two groups prior ingredients, had all the ingredients in the home to or following the intervention. Intakes were also during the evaluation visit, and indicated that they not significantlydifferent for sick (with diarrhea, would continue to prepare and feed eko ilem to fever, or respiratory illness) and healthy children their children (adopters) (Guptill et al, 1993). during the baseline and follow-up measurements. Children living in the project area, however, Most mothers (85 percent) who knew the increased their consumption of paps, whereas recipe had tried it at least once, yet only 29 control children increased their consumption of percent of those who knew and 34 percent of solid foods from the baseline to the follow-up those who tried the recipe also adopted it. studies. This suggested that the enriched weaning Knowledge, trial, and adoption of the recipe were pap, eko ilera, was replacing rather than negatively related to mothers' perceptions of its complementingthe consumptionof additional solid cost and preparation time, but it was not foods among study children (Guptill, 1 990). determined whether these perceptions prevented mothers from using the recipe, or if mothers who Thus, although trained mothers were able to tried and adopted the recipe subsequentlydecided prepare the enriched weaning pap promoted by the that it was not costly and time consuming to project, and more than 50 percent of mothers in prepare (Esre et al, undated). the project area had tried the recipe at least once, the brief intervention - - a minimum of two In addition to perceptions about cost and cooking demonstrations per mother - did not preparation time, adoption of the recipe was also result in significantly increased energy intakes related to mothers' parity, education, wage among study children during and immediately earning status, and her usual method of feeding followingthe interventionperiod. Greater success, the child. Whereas mothers with only one child however, might be expected if additional time and were more likely to know and try edo itera once, attention were given to promoting the high parity mothers (with more than five children) consumptionof other solid foods by children also were more likely to indicate that they would consuming eko iler-.

62 Dietary mwaagenm of infection

COST: Eko ilera required approximately 30 and children were fed, in six denonstrations minutes to prepare when all ingredients were carried out over two months- Mothers were asked already available, compared to 20 minutes for the to record their use of the power-flour recipe over traditional porridge given to children. Preparation a three-month period. After this period, of a two-week supply of the cowpea flour required consumption of each recipe by participating an additional three hours (for soaking and children was measured monthly, on three sun-drying), and a four to six-week supply of malt consecutive illness-free days, for six months required approximately two to three hours over a (Mosha and Svanberg, 1990). minimum of three days for processing (Guptill et al, 1993). The total cost of eko ilem was RESULTS: Results of trials indicated that the estimated at $ 0. 13/100 g in 1987, compared to percentage of mothers willing to prepare the $ 0.05/100 g for the traditional porridge and $ power-flour porridge almost every time they 0.501100 g for comnercial infant cereal (Bentley prepared food for their child increased from 13 et al, 1991). percent to 28 percent, and the percentage who used the germinated flour at least 25 percent of Tanzania the time increased from 48 percent to 85 percent, during the course of the three-month trial. The DESIGN: This project was carried out in 1983 as dietary studies found no differences in average a pilot intervention implemented by the Tanzania intake (g per meal) for the three recipes by Food and Nutrition Center in Luganga village in children less than 12 months old. Among cAildren the Iringa region. It included local research and 12-24- months, the average consunption per meal household testing of the use of a commonly of kimea was significantly higher (p < 0. 05) available amylase-rich flour (ARF) to reduce the than consumption of the thick porridge recipe (347 viscosity and increase the energy density of a versus 277 g per meal). The average energy traditional weaning food. The result of the pilot density of each of the recipes was not reported, study were subsequendy umplemented by the but researchers noted that the kimea preparation government of the United Republic of Tanzania in contained four imes more energy per unit volume seven regions, through the Joint WHO/UNICEF than the traditional uji recipe (Mosha and Nutrition Support Program (JNSP). Svanberg, 1990).

For the pilot study, 40 children aged 5-65 Use of kimea for child-feeding was months were randomly selected to participate :n subsequently promoted (interpersonally and an extended feeding trial of three recipes using a through radio broadcasts) in at least 168 villages mixture of maize and peanuts (95:5). These as part of an integrated primary health care recipes were: 1) a thin gruel containing 5 percent program that also included tri-monthly growth solid matter boiled in water (the traditional monitoring, feeding and,hygiene education, on-site porridge or uji), 2) a thick porridge containing 20 feeding for severely malnourished children, percent solid matter, and 3) the thick porridge immunizations, and other maternal and child thinned with germinated sorghum flour health services. A 1988 evaluation of the ("power-flour" or kimea). Participating mothers expanded (JNSP) program found that 93 percent were trained in methods for preparing the recipes, of mothers interviewed (N =443) knew about

63 7Thehealh sector and nutrition intenentions in Africa

kimea for child feeding but only 8 percent COSTS: No estimates of the costs of preparing reported using it on a daily basis (Government of kimea-thirned uji have been identified. It is Tanzania/WHOIUNICEF, 1988 cited in Ashworth important to realize, however, that the potential and Draper, 1992). benefit of using malted flours arises from the fact that they enable young children with limited A 1989 evaluation in one project district gastric capacities to consume greater quantities of (Kyela) found that 42 percent of mothers knew a less bulky food. Thus, use of kimea and other about kimea, but less than half of these mothers germinated grains to thin (but not dilute) (45 percent) reported using it. Among mothers traditional porridges necessarily implies greater who knew about kimea, only 37 percent could costs to families: the end result of their correct also correctly describe how to use it to thin use is that more raw ingredients are consumed children's porridges. Difficulties in correctly (Ashworth and Draper, 1 992). preparing kimea-thinned porridges were believed to be due to the common misconception that the The Gambia germinated "power-flour" was the source of energy in the recipe (Tomkins et at, 1990). DESIGN: Two projects to improve feeding practices during and following child illness have Additional constraints to the successful been implemented. The first project was promotion and adoption of "power-flour" for child implementedby the Ministry of Health (MOH) in feeding included 1) confusion created because 1981 to 1984, with finding and technical mothers were previously advised to increase the assistance provided by the AID-assisted Mass energy density of their children's porridges by Media and Health Practices (MMHP) and adding sugar or oil, 2) germinated grains are HEALTHCOM Projects. The second project is traditionally used for brewing alcoholic beverages currently being implemented by The Gambian and mothers disapproved of their use for Food and Nutrition Association (GAFNA), a child-feeding, and 3) mothers may not have consortium of organizations working in nutrition considered weaning food improvement to be a in the country, with technical assistance and priority (Tomkins et al, 1990). funding provided by PRITECH.

Note that although the use of kimea was not The first project (MOHIMMHP) was specifically advocated for feeding children during implementednationwide, beginning in 1 982 and and following illness, this project is included in lastingapproximately two years. A combination of this review because the end result of the use of qualitativeand quantitative research methods were malted flours - the production of a less viscous used to develop radio programs, and to design and but more energy-dense (kcal/g weight) traditional produce simple printed educational materials that porridge - is both nutritionally beneficial as well were used in five promotional campaigns as potentially attractive to mothers who dilute (Rasnuson and Booth, 1985). The objectives of foods offered to sick children and children the program were 1) to teach rural mothers about suffering from depressed appetites. oral rehydration (ORT) and how to monitor children with diarrhea, 2) to discourage

64 Dietwaymantagement of infecaion mothers from withholding foods during diarrhea, months after the intensity of tell educational and 3) to promote feeding during and after campaign subsided, however, only.44 percent of diarrhea episodes. The target audiences were rural mothers reported feeding more during recovery, mothers, grandmothers, and the older siblings of while 33 percent and 23 percent reported feeding children under five (Rasmuson et al, 1990). the same or less food, respectively. Baseline feeding practices during recovery were not Approximately 180 health workers were reported (Rasmuson et al, 1990). trained for five days in diarrhea treatment and management methods. An additional 650 COSTS: The costs to mothers of implementing volunteers were trained to assist mothers with the recommended feeding practices were not ORT preparation and child feeding. The program reported. Feeding recommendations were. focused on feeding during diarrhea in the first formulated based on whether they could be year, and after a mid-term review the emphasis implementedby mothers and not on nutritional shifted to increased feeding of enriched solid considerations alone (Rasmuson et al, 1990). foods during convalescence when appetite was restored. Specific solid foods and recipes were DESIGN: The second project (GAFNA/ promoted as sources of power and weight gain for PRITECH) is being carried out in several recovering children. Additional materials were communities (named below) representing five produced and about 100 health workers received ethnic groups throughout the country. Formative three days of additional training on using the research began in 1 990., with an ethnographic revised messages (Rasmuson et al, 1990). study of feeding practices during and following diarrhea in Fass Njaga Choi, Sintet, liffarong, RESULTS: The evaluation included repeated Sarra Kunda, and Kulari. This study included key surveys of 800 to 1000 women living in 20 infonnant interviews (N=65), direct observations villages, implemented over two to six month of feeding (N=57), and other methods such as intervals for two years. After one year of free listing, ranking, and triad sorting of foods, implementation,66 percent of women interviewed food combinations, and recipes (N = 111 knew how to prepare ORT and 47 percent respondents) (Samba and Gittelsohn, 1991). reported using it, but only 21 percent reported adopting the program's advice of giving solid Recipe trials were carried out in three phases. foods during and after diarrhea. Shortly after the The first phase included the development of a revised messages were delivered, 55 percent of nutrition counselling chart for children 0-3, 4-6, mothers reported that they continued feeding solid 7-9 and 10-24- months of age. This was followed foods during diarrhea, compared to only 14 by group discussions and cooking demonstrations percent at baseline (Rasmuson et at, 1990). in three communities. Household trials of new recipes and feeding practices were implemented in Likewise, at the end of the revised campaign, Sarra Kunda, Sintet, and Sankule Kunda among 58 percent of mothers reported giving more food 60 mothers (four visitsfmother over two. weeks). to their children during recovery from diarrhea, Of these mothers, 21 had children wit diarrhea while 26 percent reported giving the same amount during the trial (35 percent), 22 had children and 16 percent reporting feeding less food. Five classified as undernourished (37 percent), and the

65 The healthsector and nutrition interventionsin Africa remainder (N = 17 or 28 percent) had children Ghana classified as healthy. DESIGN: This project was carried out from 1986 RESULTS: The main results of the trials are to 1989 in seven rural and peri-urban communities presented below: in the northern savannah (N=2), forest (N=2), and coastal savannah (N = 3) agro-ecological zones * 57 percent of mothers with children less by the Nutrition Division of the Ministry of than 4- months (N=7) were willing to breastfeed Health with funding from USAID/Ghana. The excln .vely. No mothers who had already purposes were to conduct formative research on intrcd4uced other foods and liquids were willing to infant feeding practices and to develop a resume exclusive breastfeeding. comprehensive strategy for improving them that could later be implemented nationally by the * 68 percent of mothers with children 4-24 Ghanaian government. Local collaborators months old (N = 3 1 ) were willing to add peanut included the Ghana Education Service and the paste to their traditional millet pap (ogi). Most National Council for Women and Development. mothers found that their children liked the taste of Technical assistance was provided by The Manoff the enriched pap and ate it well, but five mothers Group, Inc. , under The Weaning Project. were unable to continue the practice because they could not afford to purchase the peanuts. Four A streamlined version of the five stages mothers added either bean flour, butter, milk, or described above were used to study well-nourished dried fish instead of peanuts. and undernourished children's diets (N= 101 for the latter) and feeding practices, and to test the * Mothers of children 4-24 months were acceptability of behavioral changes that were willing to increase feeding frequency, either by anticipated to improve their dietary intake (N = cooking enriched ogi more often (at 4-6 months), 105). These data were supplemented with adding adult foods to the child's diet (at 7-9 information obtained through 16 focus group months), or adding nutritious snack foods (at discussions with mothers, grandmothers, fathers, 10-24 months). local food vendors, and community health workers (Ministry of Health, 1989a-d). * Mothers of sick and undernourished children were eager to try new feeding practices, This research resulted in the development of and to continue them if their children ate well or a broad-based strategy to improve infant feeding, liked themn(Samba Ndure, 1993). which included communications, legislation, income generation, child care, and other activities. COSTS: The costs to mothers of implementing The planned communications activities included the proposed behaviors changes were not the use of individual (interpersonal) counselling disclosed. and mass media, with messages targeted to mothers with children less than 2 years old, mothers who work, and mothers with ill or

66 Dietarymanagemenrt of infection recuperating children (Ministry of Health, included $ 170,000 for all external technical 1989d). assistance (staff time, travel, per diem, communications, overhead, etc.), and $60,000 in RESULT: No data are available on the impact of local expenses (field workers travel, per diem, activities in Ghana because the planned materials research costs, esc.) (Griffiths, 1992). and interventions to improve infant feeding are gradually being implemented by the government. Swaziland Results of the household trials, which formed the basis of the educational messages developed for DESIGN: This was a nationwide project carried the strategy, found that over a four to seven day out from 1986 to 1989 by the Swaziland National period mothers were willing to: Nutrition Council, in collaboration with the Ministries of Agriculture and Health, UNICEF, * stop (80 percent) or reduce (20 percent) the and the Swaziland Infant Nutrition Action amount of water given to breastfed infants less Network (a local NGO). The purposes of the than 4- months old (N= 10), project were to conduct formative research, develop a strategy to improve feeding practices, * use both breasts for nursing young infants and to develop prototype materials that were later (90 percent; N = 10). produced by UNICEF and used by the Swazi govermnent. Local funding was provided by * to thicken the consistency (79 percent; UNICEF. Technical assistance was provided by N=29) or enrich the traditional cereal porridge the Manoff Group, Inc. , under The Weaning given to children greater than 4 months old with Project. either legunes, fish powder, or oil (8 1 percent; N = 52), Formative research included the use of focus group discussions (N =20) among mothers with * to increase the amount of other solid foods children less than 24 months, first time fed to children greater than 7 months old (82 (primiparous) mothers, mothers who work away percent; N=65), from home greater than six hours per day, grandmothers, and fathers. These were followed * increase staple or porridge feeding by a series of household depth interviews and frequency by at least one meal per day (83 observations in three communities, among 43 percent; N=35), and families with well-nourished and undernourished children, and interviews with traditional and * feed fruit to their children (82 percent; modem health providers (N=34-). Household N =28) (Ministry of Health, 1989b). trials of behavior changes were carried out with families totalling 31 children less than 24 months COSTS: The costs to mothers of implementing of age (National Nutrition Council, 1987; 1988). the proposed feeding practices changes were not calculated The total budget for the project, Results of the formative research were used to however, was estimated at $230,000 for formative develop a communications program to 1) create research and strategy development. This sum awareness about the importance of proper child

67 T7hehealth sectorand nutntion interventios in Afnca feeding during illness, recuperation, and health, 2) * 71 percent of mothers asked (N = 24-) improve mothers', other caretakers' , and health were able to feed their children at least one providers' knowledge, attitudes, and practices additional meal or snack each day, regarding child feeding and general child care, and 3) promote specific products (e.g., powdered U 67 percent of mothers asked (N = 18) were malt, a child-feeding bowl) to be used to realize able to feed their children a larger quantity of specific feeding practice improvements(Swaziland food at each feeding, and National Nutrition Council and The Manoff Group, 1992). * 67 percent of mothers asked (N=9) were also able to measure their children's food before Educational materials that were developed serving it. included computer-generated, individual counselling cards and a flip-chart for use with Several children were said to be sick or groups. A local theater group wrote and produced recuperating from illness during the trial period, a drama about child feeding, and radio programs although the exact number was not reported. were written and recorded. Other promotional Infants were given the same recommendations, materials included a project logo (showing the regardless of their illness or recuperation status. family food pot and the saying "Feed the Nation Failure to accept a recommendation because of of Tomorrow"), banners, and tee-shirts. Project illness was not mentioned, and illness was cited implementatioi began in August 1989 by training only once as a reason for discontinuing a l.ome economists and clinic nurses to use the reconmmended behavior (National Nutrition educa, onal materials. promoters and Council, 1988). male extension workers were trained subsequently. A baseline study of child feeding practices RESULTS: No formal evaluation of the project was conducted by the University of Swaziland in has been completed to date. Resul.s from the 1990. Implementation of the feeding practices week-long household trials, however, showed that improvement strategy is ongoing. The benefits almost ali mothers (89 percent) were willing to from the project to date include 1) an increased improve at least one feeding practice. The most understanding of young child feeding and its successful trials involved improving the quality of potential for improvement in Swaziland, and 2) the weaning diet: training of local professionals in qualitative research methods, communications, and * 90 percent of mothers asked (N = 10) were computer-generatedgraphics techniques. able to add germinated sorghum malt to the traditional adult maize porridge (liphalishi), COSTS: The costs to mothers of implementing the proposed feeding practices changes were not * 86 percent of mothers asked (N =22) were calculz-ed. The total budget for the project, able to add one or more energy dense food however, was estimated at $300.000 for formative (relish, oil, or peanut butter) to the liphalishi, research, strategy development, and materials development, testing, and production. This sum

68 Dietarymanagement of infection included $150,000 for all external technical and sugar had an estimated energy density of 83 assistance (staff time, travel , per diem, kcal/100 g, which was comparable to the density communications, overhead, etc.), and $ 150,000 of eko ilera promoted in Nigeria (85 kcal). The in local expenses (field workers travel, per diem, energy density of recipes using sugar and peanut research costs, materials development and solids or sour skinuned milk was lower, at about production, etc.) (Griffiths, 1992). 5 1 kcalf100 g. Protein densities for all three recipes ranged from 1.3 to 1.5 g/100 g. The Niger energy and protein densities of the traditional porridge were 22 to 24- kcalIOO g and 0.3 g/100 DESiGN: This project was carried out from 1990 g, respectively (Hung, 1992). to 1992 in eight villages in the Konni and Dosso areas by the Niger National Diarrhea Control RESULTS: The results of this research are Project. Technical assistance was provided by an currently being analyzed to make anthropologist and a nutritionist, under the recommendations for strategies to improve child PRITECH project diet and feeding practices and reduce the nutritional impact of diarrheal disease. As with the The formative research included household previous two projects, only the results of the depth interviews and observations of 64 mothers household trials can be reported at this time. Of with children 4-24- months old, at least half of the 116 mothers completing the week-long trial: whom had diarrhea in the previous two weeks. This information was supplemented by key * 73 percent made the agreed upon recipe at informant interviews with modern and traditional least once per day during the week-long trial health care providers, other mothers, and fathers period, and an additional 11 percent made the (Keith, 1990). Trials, which focused on the recipe during half of the days. Less than 4- preparation and feeding of enriched weaning percent did not make the special recipe even once. recipes, were implemented in 1 16 households. In most of the trial households children were ill (42 * 25 percent modified the recom-mended percent), recuperating from illness (50 percent), recipe by leaving out one of the ingredients or were believed by their mothers to have lost because it was unavailable, the child did not like weight recently (5 percent). In less than 3 percent it, or because it was perceived to cause or of households were children believed to be totally aggravate diarrhea (i.e. , sugar or peanuts). healthy (Keith, 1991). * 12 percent prepared a special gruel for their During the household trials one of five basic children but did not enrich it with any of the recipes was recommended. Each recipe had a reconmnended ingredients because the child millet-flour base which varied regionally by its refused to eat them or did not like their taste. degree of fermentation, dilution, and consistency. To this base, mothers were advised to add either * 91 percent of mothers indicated that they peanut solids (kulikuli). sour skimned milk, or a would continue to prepare and feed the enriched fried bean, millet, or wheat cake. Sugar was to be gruel after the intervention trial period (Herman added to all recipes. Recipes using the fried cakes and Keith, 1993).

69 The health secor and nutrtion intervemions in Afrca

Food frequency studies completed on the first interacting with mothers in the clinic setting. and last days of the trials suggested that the Interviews obtained information on feeding consumption of the enriched gruels resulted in a practices during diarrhea. Observations decrease in the consumption of drinks such as documented health agents' ability to assess fura. Changes in the frequency of consumptionof children's clinical states, and the advice they gave non-milletsnacks and foods on the first versus last to mothers concerning feeding during and day of the trial varied (average range: - 0.6 to following diarrhea (Sene, 1991). 1.1/day), depending on the village studied. Children' s total dietary intakes (including Phase II included seven focus group - breastmilk, millet-based gruels and family foods, discussions with mothers who had at least two as well as non-millet "snack foods") were not children under five ytars and one child under systematically evaluated, making it difficult to three years of age. The acceptability of various estimate the nutritional impact of the proposed foods for feeding during diarrhea was discussed. recommendations during periods of illness, Several recipes were prepared and administered to convalescence, and health. The costs of preparing children. The nutritional characteristics of the various recipes were not estimated. conmmonpreparations were also evaluated (Sene, 1992). Phases I and It were completed in urban COSTS: The cost of preparing the different communities near Dakar and rural villages in the recipes was not estimated. Field work for the Thies region. formative research cost $5,000 in local expenses. The use of outside consultants required an Phase m was designed as an operations- additional $65,000(including fees, travel costs research effort to compare three nutrition and other expenses). Administrativeand technical education interventions: one using health facility support were approximately $20,000. staff only, another using traditional birth - attendants, and the third using representatives of SENEGAL local women's groups as nutrition promoters. Ministry of Health personnel from the central DESIGN: This project began in 1991 and is still level (Nutrition and Health Education Services) being carried out in nine villages in the Fatick and regional and district levels participated in the region and district. It is a collaborative effort research. between the Ministry of Health (SANAS), the Organization for Food and Nutrition Research in Phase m field work inc'.udedfour home visits Africa (ORANA). and local health providers and to 1) identify children 6-36 months who had or mothers working in the project communities. were recuperating from diarrhea, 2) interview Funding was provided by PRITECH. mothers and observe feeding and related child-care practices during a 12-hour period. 3) introduce The project has been carried out in three one or more feeding practice changes based on a phases. Phase I includedinterviews with 30 health rapid assessment of observed practices, and 4) agents and 33 mothers with children suffering evaluate mothers' reactions to the new feeding from diarrhea, and observations of health agents behaviors and whether they were likely to

70 Dietarymanagement of infection continuethem in the future. Fifty-fburchildren 2) 77 percent of mothers who agreed to from nine villagesparticipated in this stageof the enrichdteir children's weaningporridge (N=31) research: 39 (72%) had diarrhea and 15 were with either peanut butter, milk, butter, or oil recovering from diarrhea during the household actually tried this reconumendationand all but trials. three of these mothers indicatedthat they would continuethis practice. Milk and butter were the A training module and series of nine most acceptableingredients to add to porridge. counsellingcards were developedon the basis of Butterwas more commonlyadded to the porridge research findings. Eleven health post personnel of infantsaged 6-11 months;milk was most likely (e.g., nurses and monitrices) were trained in to be tried and continued. nutritional concepts and appropriate feeding practices during and following diarrhea, the 3) 43 percentof motherswho agreedto feed contentof the counsellingcards, and how to use their children more frequently during the day them with mothers. Traditionalbirth attendants (N=7) tried this recommendation,and all said and nine nutritionpromoters from local women's they plannedto continmethe practice. groups were also trained, in separatesessions, in how to implementgrowth monitoring activities 4) 80 percentof motherswho agreedto feed and in the use of the preparedcounselling cards. their childfrom a separatefeeding dish insteadof the family pot (N= 15)tried the recommendation, RESULTS: The trainingsessions described above and all but four (67 percent) said they would were carried out in April-June, 1993, and continuethis practice. information on the program's impact at the communitylevel is not yet available.The rapid 5) 87 percentof motherswho were asked to assessments characterized the major feeding feed their sick children with a miture of milk, problems,in descendingorder of frequency,as 1) oil, and sugar (N=30) tried the preparation,and insufficientfood quantity offered to the sick or 62 percent of those trying it felt they would recovering child, 2) insufficientfood quantity continuewith the practice. consmnedby the child, 3) low energy densityof the diet, 4) low freqency of feeding, and 5) Mothers said they would continue the failure to feed the child from a separate bowl reconmmendedpracties becausethey were easyto (Diene, 1993). implement, their children liked them, they promoted weight gain, and they gave their Duringthe trials: children energy and appetite. The reasons for resistanceto trying and continuingnew practices 1) 71 percentof motherswho agreedto feed werefailure to rememberthe recomnendation,the their childrenmore food each day (N-7) actually ingredientswere not available(e.g-, peanutbutter) tried this recomnendation.All motherstrying the or were expensive,and their childrendid not like reconmnendationsaid they would continue the them. Although feeding practices were not practice. perceivedas major caises of health and nutrition problems, mothers were motivated to accept

71 Thehealth sectorand nurition interemionh in Africa nutritionaladvice that wasnew and believedto be gether. Such coordinationhelps to insure that effective(Diene, 1993). research findings are used to benefit program implementation. Lessonslearned from the projectincluded the understanding that health providers require COSTS: Onlyimplementation costs for PhaseIII intensive and practical training in how to of the project have been estimated and they communicate, persuade, and negotiate with represent $35,000 for local expenditures. No mothers to change their behavior. Operational internationalconsultant was hired. Monitoring research efforts are facilitated when research and supervisioncosts werearound $15,000. institutionsand implementingagencies work to-

72 *X Nutrition services that could be provided by the health sector if a radical change in approach were adopted

(therapeutic and/or preventive care), the 1. Food supplementation' appropriate groups are targeted:

* if the aim is to prevent growth deficiencies, a) FINDINGSAND RECOMMENDA7TONS all children under two years are included. Geographical targets may also be i) Food supplements for pregnant women incorporatea.

Although MCH services are considering the * if the goal is to provide therapeutic care, option of dietary supplements, targeting pregnant malnourished children (criterion is women rather than children with this intervention height/weight ratio) should be targeted may be more productive and cost-effective, individually. because the period during which supplementsare necessary (the third trimester) is shorter and the U Using weaning foods that are child-specific, number of beneficiaries is smaller. culturally acceptable, and have a high energy density. ii) Food supplements for children : As a counterpart to correct targeting measures, Repeated observationsin Africa and elsewhere which can help to improve cost- effectiveness, have proven the worth of providing supplements make sure that the target group is adequately to enhance the nutritional status of children at covered. critical ages when their family situation and customary eating habits do not provide for the U Integrate other types of services aimed at this nutritional needs of a young child. target group (nutrition education, basic health care, etc.) An overview of the assessments suggests the following principal factors that lead to success a As is the case for many other interventions, (this is not an exhaustive list): community participation appears to be crucial. Only if the community understands and accepts a An initial evaluation that correctly analyzes the the principle of targeted supplementation will situation (extent, distribution, and causes of children at risk receive the supplements directly nutrition problems), with goals adopted and without their being resold, shared, or substituted interventions outlined according to the program's for the customary food ration. The conununity needs. should form a local conmnittee cat makes the original request for supplementsand ensures that * Depending on the situation (for example, the the program is managed at a community level. predominance of growth-deficient and/or The committee should be kept informed of the underweight children) and the objectives program's progress at regular intervals so that it understands fully any changes made.

73 The health sectorand nutritioninterventions in Africa

* Whenever the availability of food supplies Many programs were targeted primarily at allows, limit food imports used for supplements: irnproving the feeding practices of mothers in the as with all forms of food aid, imports also have case of young children, in particular by distorting effects. When supplements are needed overcoming certain habits or beliefs; in these on a national level and external food aid is cases, the food distributed was accompanied by required, the aid program should be nutrition education information adapted for this institutionalizedquickly. The health sector should use. Because of the poverty of the families in also invite other sectors to put adapted nutrition question, the food is generally distributed free of and food policies into place that include a regional charge or its purchase is subsidized so that it may and local distribution plan to be implemented as be distributed at a nominal price acceptable to the quickly as possible and to encourage communities beneficiaries. to commit to supervising supplementation, particularly to verify that young children are c) FOODSuPPLEMENTS FOR PREGNAVATwou&v consuming the supplemnents. Research shows that it is more effective to * Supplements should not substitute for meals supply pregnant women than newborns with prepared for the family if they are adequate supplements if the desired effect is to affect the nutritionally. The duration of targeting measures child's growth (Kusin et al., 1992). In Africa, should also be limited. however, it has been shown that the practice of not eating too much during pregnancy ("eating * Nutrition education at an adequate technical down") is common because women fear a level should accompany supplementation in each complicated pregnancy if the child is too big case, otherwise supplementation acts only as a (Brems and Berg, 1989). Research conducted to palliative and does not eliminate the real causes of date on the effect of supplementation on birth malnutrition. weight indicates a relatively small difference between the weight of infants born to women who b) DEFNmON received supplemcnts and those born to women who did not. This small difference in birth Programs whose goal is to supplement the weight, however, does make it possible to reduce diets of vulnerable sectors of the population significantly the proportion of infants with a low (preschool children, pregnant women, nursing birth weight: from 23.7 percent to 7.5 percent in women; some programs are also for school-age Gambia (a difference in weight of 200 grams); a children) with foods that correct a nutrition deficit decrease of 29.6 percent in Mexico: and in and improve the nutritional status of the Guatemala the prevalence was reduced from 29 individuals concerned are grouped in the category percent to 13 percent in the poorest women. of dietary supplement programs- Some programs, however, have functioned more as instruments to According to Prentice et al., the greatest redistribute and transfer income and thus reduc_ deficiency in growth in utero appears to occur the poverty of the families at risk. after the 37th week of pregnancy. This means that supplementation during the last trimester

74 Food supplementation

Table VII. Impact on birthweight of various supplanentation programs for pregnant women Calories/day Calories/day before providedby Increasein birth Study supplementation supplements weight(in grams) Comments References Colombia 1620 133 51 birth weight increasedby Herreraet al, 18g for womenwith a 1980 low height/weightratio Guatemala 1415 149 111 Lechtiget al, 1980 Taiwan 1600-2000 276 SSg for boys Adairand Pollit, 35g for girls 1985 The Gambia 1419 430 200, +t- 53 duringthe rainy season Prentic et al, (supplementationwas 1987 ineffectiveduring the dry .______season)

could have a greater impact on birth weight than d) FOODSVPPLEMFIWS FOR CHILDREN supplementation during the first two trimesters. i) Previous assessmients Kusin and Java's research has shown that children of mothers who received supplements A distinctive feature of supplementation were taller and heavier than other children during programs is that they have already been the object their first five years; the weight difference was of several broad assessment efforts over the last considerable until 24 months and the height 10 years; these include, in particular, Anderson et difference until five years. The reason for this al (1981), who focused on the methodology and difference could be that women who received design of the programs; Beaton and Ghassemi in supplements had more milk. 1982 and updated by Ghassemi in 1989; assessments between 1980 and 1985 of the "Food So, although MCCHservices are considering for Peace" programs (PL 480, title ED (USAID, the option of dietary suppiements, "targeting 1985); Bremer-Fox et al. (1987) on PL 480 as pregnant women rather than children with this well; Kenmedyand Knudsen (1985); and a new intervention may be more productive and cost- analysis by MvIora,King, and Teller of the effective, because the period during which performance of the prograins during the 1980s supplements are necessary (the third trimester) is and perspectives for the 1990s (Morm et al., shorter and the number oi beneficiaries is smaller" 1990). (Kusin et al., 1992). It should be noted that supplementation seems to have an effect only if * The conclusions of these assessments: the woman is already malnourished, and it has no effect if she has a good diet. On several occasions, the assessments showed an improvement in anthropometric indicators, albeit a modest one; among other effects observed

75 The healih sector and nurridonintevrentions in Africa was a decrease in mortality, thanks to steps taken A brief overview of these assessment criteria to decrease the severity of episodes of illness follows. Of the 43 projects evaluated in 1982 by (Rose and Martorell, 1991). The rates of Beaton and Ghassemi, six concerned African coverage are generally very low even for the most countries; 23 included a control group, and of important programs, and these programs have a these only one was in Africa. Of 15 PL 480 great deal of difficulty in reaching the youngest program assessmentsconducted between 1980 and children (less than 24 months). The quantity of 1985 (USAID, 1985), five include data on die food provided is often very small relative to the nutritional status of the participants, two of which needs of young children; in addition, the actual are in Africa (Morocco: Gilmore et al., 1980; amount of food ingestedby the children is reduced and Senegal: Echenberg et al., 1984). Finally, even more since food is shared, sold, or the author who analyzedthe assessmentsof the PL substituted for the normal ration. Target 480 programs financed by USAID during the measures, on whatever basis (geography, age, 1980s recognizes that these assessments were not economic status, nutritional status), are usually designed to demonstratethe impact on nutrition of inadequate. Finally, the staff is often the programs: of 27 assessments, seven chose an overwhelmedby the tasks required because of a adequate sample size but none of them used a lack of training, especially in the managementof protocol such as a before/after comparison with a this type of program. Some program-sobtained control group, which is indispensable for good results when they were run in conjunction measuring accurately the impact of with other nutrition or health measures (notably supplementationon the participants (Mora et al., education) (see Gilmore et al., 1980, for 1990). example). ii) Problems wih program objectives a Program assessmentproblems: Most programs pursued many objectivesat the Most frameworks used to evaluate the samnetime, either explicitly, as title 1t of PL 480 programs were not suitable for measuring the real specifies: "...to combat mnalnutritionespecially impact on nutrition. This is why few reliable in children; promote economic and community results are currently available (see Esrey et al., developmert with the goal of reducing the need 1985, for example). for such assistance...," or implicitly, with the goa! of improvingfeeding practices, preventinggrowth in Africa, assessments of the "PL 480, title deficiencies(stunting) and helping children losing 11" programs of American aid managed by the weight to recover (wasting), and providing CRS are the most widely available. In Africa. uniform treatment in zones with a high prevalence problems often revolved around the definition of of both of these conditions. program objectives, their relation to the .ype of nutritien problems in the affected zones, and the In Africa, the anthropometric index used in indicators used to measure these changes. assessments has generally been the "weight/age" ratio. Only a few programs (none in Africa) have tried to evaluate the impact in terms of actual

76 Food suppkmentation energyintake, physical activity, energy expended, The importtnce of targeing: These results morbidity,and mortality. This is the resultof the clarifythe apparentineffectiveness observed. lack of simple, quick, and inexpensive In fact, it is not surprisingthat there is little measurementtechniques. impact in terms of the "weight/age"ratio becausethe programswere generallynot set Recouping delayed growth in terms of up for specificage groups(most often 0-60 height: Many studies have confirmedthe months), and they mosdy affected children early onset of delayed growth in the first over two years in regions that had a high year, which then proceedsat a normal pace prevalenceof delayed growth in terms of (see the resultsof the CollaborativeResearch height(and perhaps a mild energydeficit after SupportProgram conducted in Egypt,Kenya, the weaningperiod) and did llot even target and Mexico, for example;cf: Callowayet low-weightchildren. A typicalexample is the al., 1988); the chances of maldng up for supplementationof preschool-agechildren in growth deficienciesafter the second year the poorer regions of Addis Ababa in appearto be slimeven when the diet seemsto Ethiopia; this is one of the rare longitudinal be adequate, at least in terms of caloric assessmentswith a control groupusing three intake. In addition,there seemto be specific indices: 'weight/height"ratio, "heightlage" ages when growth responds to ratio, and "weight/age"ratio (Deneke and supplementation,generally during the child's Wolde-Gabriel, 1985). The program's first year (Lutteret al, 1990). completeineffectiveness was demonstrated by the fact that the inFprovementsobserved in the Recoupingweight: Severalresearch projects groupreceiving supplements were the sameas outlined the effects of supplementation those obse ied in the control grup. dependingon the ageof the child and type of According to the authors, the program mlnutrition. For example, an energy coveredbetween 25 percentand 100 percent supplementgiven to childrenbetween the ages of the children's energy needs (but with the of six and 24 monthswho were mildly and usual diversionproblems: sharing, selling, moderately underweighthad a pronounced and substituting the supplements for effect (Rivera, Habicht, Robson, 1991). customaryrations). The prgrams targeted Narangwal'sstudy in India (Kiehnannet al., childrenbetween six months and six years 1978) showed that a well-managedand (fewertan 10 percentwere under one year) correctly targeted food supplementation in a case in which46 percentof chlldrenwere program had a significantimpact on weight shorterthan normal(height/age < -2 standard and height, while isolated health measures deviations) and only 3.5 percent were taken to fight infectionhad an effect on underweight(weight/height c -2 standard morbidityand mortalitybut not on growth;on deviations)!They concludethat the practice the other hand, the two programs of givingfood to childrenwithout taking into implementedtogether had a certainsynergistic considerationtheir age and/or nutritional effect. statusshould not continue;targeting based on anthropometricmeasurements is suggested, althoughwhether this should be done using

77 The health sedor and nutntioninterventions in Africa

the 'weight/lheight" ratio or the "height/age" basis and their effectiveness should also be judged ratio is not specified. in terms of their economic impact. The income transfer aspect seems to have been significant in a Experience and new information suggest, number of programs, specifically in those of PL therefore, that the nature of the objective (and 480, title Il: thus, in Mauritania (McClelland et thus of the choice of an evaluation framework al., 1984), distributed food represented 14 percent adapted to it) is central to determining the of family income per person and 33 percent for choice of implementing factors, which will be nomads; in Morocco, between 4 percent anid 24 examined next (coverage, type of targeting percent (Gilmore et al., 1980); in Kenya, between measures, size and type of ration, role of 5 percent and 18 percent, depending on the health services, and integration into other number of rations in the richest rural families, and initiatives). between 22 percent to 72 percent in the poorest families (Fleuret, 1985); in Senegal, this * If the program's objective is to treat represented on average $6, or about 18 percent of underweight children currently at risk in regions the family's income (Echenberg et al., 1984); in where this problem is prevalent, priority should be Lesotho, it represented $31.50 per family each given to putting targeting or even screening month (USAID, 1988). measures intc place based on anthropometric criteria (such as "weight/height" ratio). This iii) Program operations framework would be particularly appropriate in zones subject to chronic or acute seasonal food Supplementation provided directly to shortages. househols or at health centers:

* If the program's objective is to conribute U At health centers: supplementation to inproving long-term conditions by changing provided at health centers ensures that the feeding practices and preventing growth desired ration will be consumed. It offers the deficiendes (to go back to an old idea, a sort of advantage of providing for the delivery of 'vaccination" against malnutrition), priority other services: outpatient care, nutrition should be given to targeting based on geographical education, and growth monitoring. On the location and age (for children under two) other hand, experience has shown that considerations. This framework seems to be supplementation reaches few children under adapted most to zones with a low prevalence of two years and that the ration substitutes in underweight children but a moderate to high rate part for the customary diet. It is costly of children with growth deficiencies in terms of (infrastructure, personnel) and provides for a height. low level of coverage in a limited area. The drawbacks of coming daily are evident as are * If the objective is first to reduce poverty the risks of dependence, relieving the mothers through an income transer with only the hope of of their responsibilities, and social stigma. indirectly affecting nutritional status. targeting measures should be implemented on an economic

78 Food supplemewraion

* To households: the distributionof food to Supplementation programs have had be consumed at home offers the possibilityof particularly low rates of coverage. During the providing greater geographic coverage, 1970s, Beaton and Ghassemi (1982) note that because a single center can cover several programs reached 6 percent to 10 percent of the hundred children. Oriented toward the populationof preschool children. These are rough family, it also reaches children under two estimates and do not reflect the rate of actual more easily. Mothers have the advantage of participation. Some African countries have not needing to show up as frequently. It is similar rates of coverage: Morocco in 1979 not as expensive and provides for the (Gilmore et al., 1980) had 9 percent (11 percent possibility of delivering other services at of poor families and 6 percent of malnourished health centers. On the other hand, the risk families); Tunisia in 1976, 15 percent; Kenya in that food destined for the target group may be 1980 (USAID, 1980) averaged 1.9 percent but diverted (shared, sold) is greater. The varied between 1.3 percent and 4.9 percent distribution center may also be too far away. depending on the region; Togo, 5.6 percent for And because distribution takes place less children between zero and five (Stephens et al., frequently, educational actions may be 1984); Senegal, 10.3 percent of children under provided with less regularity. three (Echenberg et al., 1984); Gambia, 13 percent of children under three and only 8 percent Growth or recuperation is believed to be more of the malnourished in this age group (UJSAID, rapid with supplementationprograms administered The Gambia, 1988). Botswana, with 83 percent from health centers. There are not many data to of children under five years eligible in 1985-86 verify this, however. In a controlled experiment covered, is an exception in Africa (Quinn et al., in Chad, Stefanek and Jarjoura (1989) found no 1986). differences in the effect on growth in the short run (average weight gain after 60 days) between two Such rates suggest that these interventions groups of malnourishedchildren (weightlheightC were not effective in reaching young children 80 percent of the reference median), with one throughout the country: they exclude many receiving supplements at a health center and the sectors of the populationthat have a real need for other receiving the same ration to be consumedat food supplements. This brings up the issue of home. In every case, growth was significantly geographically targeted measures. In fact, often greater an a control group receiving no a program is not weli located relative to the supplements. geographicaldistribution of the populationand the prevalence of malnourishedchildren. Kenya is a Coverage: good example of this problem: in 1979, 62 percent of childrenreceiving rations from the CRS Supplementation programs in Africa are lived in the central and eastem provinces, which providing increasingly less coverage; there were represent only 33 percent of Kenya's total 2.1 million beneficiaries in 1983 and only 1.3 population. In the central province, the nunber million in 1989 (Mora et al., 1990). of malnourishedchildren who could benefit frm a food ration supplement reached 21.4 percent, but in the other provinces this number varied

79 T7hehealth sector and nutntion interventionsin Africa between 2.1 percent and 8.4 percent. The total reach children under two years in Africa and rate of coverage for children between six and 60 elsewhere. months (the ages eligible for the program) did not exceed 2 percent (USAID, Kenya, 1980). For preventive purposes in areas where the prevalence of children with growth deficienciesin Logistical problems and inadequate coverage terms of height is moderate to high and the by the health care system are an obstacle to prevalence of underweight children is low, all the solutions to the problem of coverage in Africa; factors used currently, including cost-efficiency, they also limit the effectivenessof programs on a argue for age-specifictargeting - under two years national level. This is a peculiarity of Africa and as early as possible, beginning at four to six compared with a large number of Asian and Latin months - possibly axcompaniedby geographical American countries, which provide a much higher targeting. For therapeutic purposes in areas level of health coverage. Improving the overall where the prevalence of underweight children is effectiveness and the actual impact of these moderate to high, individual targeting programs clearly begins with a change in their (anthropometric)seems necessary and jusdfied in scale and targeting measures. This may mean terms of cost efficiency. In Zaire, individual new costs, but it also means that the way in which targeting based on need was preferred for the health sector functions must change. participants, depending on their nutritional status, Botswana's exceptionalcoverage rates (Quinn et as it offset the subsidy and thus the cost: $0.15 al., 1986) are attributablespecifically to the high per kilogram of food for well-nourishedchildren, rate of coverage that the health care systemn half the cost for first- and second-degreechildren provides the population (575 permanent centers accordingto Gomez (weight-age),and free for the and 175 mobile ones for a population of 218,000 most malnourishedchildren (Zeitlin et al., 1987). children under five years, or about 300 children The program also included geographical targeting per center; the distributionof food reached 13,700 of zones with low socioeconomicindicators. The children on a daily basis who were < 80 percent limitation is still the low level of participation of their age-appropriateweight). above a certain age in growth monitoring programSwith which the supplementationprogram An important part of the success of is associated: in the case of Zaire, participation Zimbabwe's program (Mason and Tagwireyi, fell considerably after nine months. 1989) is also attributable to its ability to reach remote zones, thus permittingchildren to be fed a The case of Senegal (Echenberg et al., 1984) short distance from their homes, which is essential amply illustrateshow the absence of geographical, to maintainiiaghigh levels of participation. economic, or age-specific targeting measures decreases significantly the actual coverage of the Targtding: target population. The program was mainly focused on the rural population (<10 percent One of the main lessons of the extensive were city dwellers), but the distribution centers previous assessments is the failure of progrms to were not situated in the least-developed zones: they were dispersed like the health centers. In

80 Food supplementation

general, participants were at a slightly lower Ration size and type of food: economic level than the population at large; but 10 percent of the highest-risk families had not For the programs involvingfood supplements heard of the program and were excluded from it. delivered to households, sharing food with other Ninety percent of children were enrolled before members of the family or selling a portion of it is they were two, but 20 percent stayed beyond age an ongoing problem. Beaton and Ghassemi three, thus decreasing the chance that other (1982) report that in the programs they evaluated, children at risk could enter the program. As only 40 percent to 60 percent of the food such, coverage of the population actually at risk distributedreached the targeted children. Because was certainly less than the 10 percent publicized. the programs provided food aimed at decreasing In most of the countries of the Sahel, seasonal the energy deficit by 40 percent to 70 percent, the targeting measures should be considered (Mora et result was that in fact only 10 percent to 25 al., 1990). percent of this deficit was made up on average.

In Botswana (Quinn et al., 1986), the To avoid the problem of. inadequate food program provided general coverage without any rations, many programs included an estimate of specific targeting measures in rural areas during the impact of sharing and resale in their the most acute part of the drought period. When calculation of ration size. Rations of 300-400 conditions were less severe, targeting measures calories/day were increased to 600-800 based on medical criteria were implemented; the calories/day. As such, in Kenya, presuming that focus was initially on children whose weightlage 50 percent of the food distributed was consumed ratio was less than 80 percent, but this was soon by the child, coverage of the child's needs varied changed to include children whose weight had between 23 percent and 47 percent, depending on fallen rapidly in the previous three monfts. This his age. This practice has its limits, however: adjustment was made to exclude those whose low assessments made between 1980 and 1985 of the weight relative to age was attributablemainly to a PL 480 progran show nh'at in Gambia, in significant growth deficiency in terms of height; transition periods, title II monthly rations were the result was to enhance the program's consumed in two to three days. In normal effectiveness. Targeting was implemented on a periods, in Ghana and >'auritania, food geographicalbasis at the same time: first, remote distributed for one month was gone in two weeks villages were chosen and then small farms; during (USAID, 1985). In Senegal and Cameroon, drought, for exanple, priority was given to cattle initial assessments estimated that 6 percent to II breeders with fewer than 40 head of cattle. In percent of calories were actually consumedby the Zimbabwe, targeting was implementedbased on beneficiaries (Mora et al., 1990). In Senegal, for medical criteria that were thoroughlyexplained to examnple, according to the report done by the population: arm width < 13.0 centimeters Echenberg et al. (1984), rations did not last for (Mason & Tagwireyi, 1989). longer than two weeks; although in theory the goal was to provide 7.5 kg of food monthly per child, each child received 5.53 kg per month on average in 1981 and 3.£4 kg per month on average in 1982.

81 Thehealrh sector and nutritioninterventions in Africa

The sale of title II rations sometimes Integration into other health programs: represented a significant amount of a family's income. Food donations are seen as a new source All the assessments indicate that the impact is of income for the family rather than as a way to greater when a combination of direct and indirect change children's customary eating habits. A way interventions is used. The example of Morocco to avoid or limit the diversion of food rations is to (Gilmore et al., 1980) seems to demonstrate the provide food that is specifically for children. impact of the addition of an educational component (11 percent of children are It is necessary, therefore, to provide malnourished when supplementation is nutritionally appropriate and socially acceptable accompaniedby nutrition educationcompared with food supplements for the critical age period in 33 percent when no educational component is which the weaning process occurs; it is irmportant included). One must recognize, however, that the to remember that this is a period of rapid growth assessment framework does not really allow for - from six months on (and probably earlier for formal conclusions. Growth monitoring seems to some children), nutrition requirementsexceed the have played the role of encouraging nutrition nutrition breastfeeding can provide - and that the education and has served as a means of integrating diet customarily consumed does not have the this education into supplementationprograms. In energy density necessary to meet requirements. Senegal, the addition of a growthimonitoring and Thus, supplementationshould be carried out using educationcomponent did not have the same effect new weaning foods with a high energy density (Echenberg et al., 1984); the educational that are a combination, depending on the situation, component, however, did not seem to be very of local foods (from domestic, small shop, or well developed in this case. The quantity of food semi-industrial sources). It should also be distributed in the two cases was not the same, accompaniedby nutrition education. with the ration size distributed in Morocco to a large extent having taken into account the possible Zimbabwe has attempted to provide an problem of sharing (three rations per family per energy supplement based on corn, oil, peanuts, month of 500 calories per day); the education and local beans totaling 530 calories per child (or program implemented in Morocco was also about half of the daily requirement of a two year carefully developed over several years by old child). But production was insufficient, at professionals in this area. least at the local level; this is what led to the push to develop the initiative by giving land and other Nutrition education, development of inputs to women specifically in charge of the appropriate weaning foods, and monitoring of problem (Mason & Tagwireyi, 1989). Whenever growth as instruments to follow up a program demand increases as a result of insufficient seem to be the key service areas into which production due to environmental problems supplementation in conjunction with primary (drought). this solution is not appropriate. health services (prevention of diarrhea! diseases, vaccinations, prevention of parasites) should be integrated. Education is vital to ensuring the

82 Food supplemenation maintenanceof breastfeedingduring the first ran the program at the local level. Their monthsof supplementation(Marchione, 1990). perception of the role and value of supplementationwithin the context of clinical The Iringa program in Tanzania is an medicalpractices and public health has been a exampleof a successfullyintegrated program in criticalfactor (Burkhardt, 1980). which the feeding practices affecting young childrenwere analyzed on the family,village, and Botswana(Quinn et al., 1986)demonstrated regional levels. Various solutions were the extent to which a good food distribution implementedto addressdifferent short- and long- programmay act as a powerfulforce to attract term situations: local or commnercialproduction groups at risk to the health system; preschool of weaningfoods and increasingfood security for children came four times more frequently to householdsand villages by improvingcultural health centers with these programs (40,000 in practices. Growth monitoringand educational 1980, 140,000in 1985). This leads to a marked sessionson the weaningprocess completedthe increase in vaccinationcoverage and a general process. Youngchildren were provided with food (and rather unexpected) improvementin the supplementsin day-care cernterswhose resources children's overal health as a result of their were providedentirely by what was availablein cominginto contactwith the health systemon a the villages and not through external food aid more regularbasis. This wasthe case during the (Seenappa& Ljunqvist,1988; WHO/UNICEF, worst droughtconditions when the demandfor 1988). food aid was very high; it is still true in many cases in which the food situation is poor and Roles played by the healh center, NGO, and healthcoverage is low. One shouldnot lose sight the conumunity: of the fact that food supplementscannot be providedon an individualbasis or for an entire * At rie healthcenter level, two issuesseem populationindefinitely; the risk is thenthat people to be fundamental: extendingservices provided suddenlystiop coming when food is no longer by the centers and employing competent, available, as has been noted in maay CRS motivated,and community-orientedpersonnel. programs(see Diene in Senegal.1989; Nyiribibi in Rwanda, 1990;etc.), althoughthis is mostly Assertingthe need for integrationbrings up determinedby the qualityof the health services againthe questionof healthcoverage and thus of offered(Morm et al., 1990). increasing the coverage of health centers in Africa. Simplyadding a supplementationprogran a NGOs play an important role in to the activitiesof the healthcenters under current supplementation programs; encouraging an conditions in most African tLuntries is no integratedapproach in conjunctionwith public guaranteeof success. Ghassemi(1989) reports healthservices still producesthe most convincing that in fact the two elementsthat are key to the results; in Zimbabwe, the NGOsworked under success of supplementationprograms have been the ongoingsupervision of the Ministryof Health the extentto whichthe healthcenters have used an and a committeecomposed of representativesfrom active approachto extend the interventionand varioussectors, including Agriculture, Education, follow it up and the participationof doctorswho and Rural Development. Previous supple-

83 Thehealth seaor and nutritionintern'etions in Africa mentation program assessments, as is the case It is difficult, however, to sustain motivation with other programs, illustrates the necessity of and provide supplies over the long run; the long- political, material, and financial support at all tent strategy should be modified to ensure more levels, from central government agencies to the autonomy. This is why Zimbabwe set up a broad conmunity itself. program to produce peanuts and beans at the community level to encourage supplementation; Local managcment of the programs must be unfortunately, the first trials held at about 500 very flexible: the ratio of food purchases to centers took place during a succession of drought transport costs must be readjusted based on local yars, and the results have not yet reached conditions to reach the most remote villages even expected levels (Mason & Tagwireyi, 1989). if fewer children will receive food supplements as Tanzania seems to have a successful production a result. Programs that are too centralized do not and distribution program at the village level (Van have the resources necessary when these choices der Haar, 1983). are made to carry them out effectively. One risk is the temptation to use a strong vertical Effectiveness of some programs: organization, given the sums involved. In Zimbabwe, administrative structures that date The program conducted in Lesotho by CRS in from the struggle for independence were used; the PL 480 title U framework (USAID, Lesotho, they proved to be very efficient in setting up the 1988). considered exceptional in its inability to screening process, the food preparation and improvethe nutritionalstatus of childrenreceiving distribution programs. A national coordinating food, was the subject of a detailed study. The committee had already prepared a manual and study recommended that ti,: program be training program adapted to the different terminatedand offered numerousexplanations for intervention levels in various sectors (Mason & its failure, including: inadequate rations because Tagwireyi, 1989). of sharing and resale, the absence of an educational component, inappropriate targeting U Communztyparticipation has often appeared measures, and inadequate coverage. Food given to be a critical factor in the success of intervention to the mothers represented in many cases an programs. It is even more important in income transfer that allowed families to purchase supplementation programs in terms of consumer goods having nothing to do with organization, local resources, and the commitment nutrition. Another important factor was that the and approval of the community: confidence in food did not generally reach the children most at and acceptance of new foods, regular attendance, risk because inclusion in the program at the and supervision through home visits. In Africa, dispensary level was based on the criterion CRS programs tried to encourage parents' nweightlage ratio," regardless of age. In a commitment through a formal contract. In country in which acute and severe cases of Zimbabwe, local authorities were involved in mnalnutrition are rare, but where growth supervisingyoung children's consumptionof food deficiencies in terms of height are very frequent, supplements and teaching the monthersabout the the type of targeting adopted resulted in the value of local high-energy foods. inclusionof the oldest children (those who had a

84 Foodsuppiementotion low weight/ageratio becausethey were growth Two title ll, PL 480 programsin Morocco deficientin terms of height, datingusually from and Senegal were the object of an in-depth the end of the first year). As such,the preventive nutritionevaluation. In Senegal(Echenberg et al. objectivesof the programwere not fulill&. In 1984), growth (weight-age) in participating addition,since childrenremained in the program childrenunder two years was closely correlated to for many years, coveragewas not as broad as it the levelof participationand to the durationof the couldhave been. program. This is a very raw result; in fact there was no differencein growthbetween participants It is interestingto notethat the conclusionsof andnonparticipants of the same ageand economic the Lesotho program's final assessmentconcur status from dte same villages. There does not withan analysisthat couldbe madebased on the seem to have been any impact on growth; but understandingof this problemthat evolved during mortalityrates for the mostvulnerable age groups the 1980s (see above): in fact, it has been were lower for participants than for proposedthat the only criteria for inclusionin nonparticipants. Accordingto a survey of the future programsbe that childrenbe betweensix mothers,the food distributedfor the monthlasted and 24 months of age, that supplementation about two weeks; it was probably used for consistof a specifichigh energydensity weaning purposes other than feeding young children. food, that an educationalcomponent be added, Furhermore, the amountof food distributedwas and thatthe programin its entiretybe coordinated not alwaysconsistent with recommendations either with scheduledvaccination times. This should becauseof managementproblems or as a result of result in increased prevention, especially for repeatedshortages. childrenbetween six and 12 months,a period in which growth slows down in Basothobabies; a In Morocco (Gilmore et al., 1980), a marked increase in energy intake, with the comparisonbetween childrenwho were in the specificfood limiing resale and sharing; and a programfor the previoustwo years and thosewho higherdegree of coverage,with the age linit of hadjust enteredshowed a reductionof 69 percent 24 monthsincreasing the possibilitiesof rotating in severeand moderatemalnutrition (weight-age in other children. < 80 percent). An educationalcomponent was addedduring the course of the program. In 1975, Duringthe 1980s,a CRSprogram in Burkina 33 percentof childrenwho receiveda supplement Faso (USAID, 1981)indicated that 3 percentof were still mnalnourished;in 1978, with the children had a weight/ageratio lower than 60 combinationof supplementationand education, percent of the standard value after participating only 11 percent were malnourished. This for two to three years in the program,compared program,which was exemplaryin its contentand with 10 percentof childrenwho had just entered execution,was put in place at the same time as a the program. It shouldbe remembered,however, numberof socialchanges (some emancipation for that this type of frameworkmay be less valid women).which may have had a determinantrole, becauseof self-selectionand bias related to the were taldngplace. children'schanging ages. Moreover,these were cases of severemalnutrition. The supplaemntationprogram implemented in Zimbabwe did not produce satisfactoryresults

85 The health sector and nutrition imerventeionsin Africa immediately (MOH, Zimbabwe, 1982). At the there seems in fact to have been a decrease after beginning, the objectives were not well defined, the program had been running for several years. and the program lacked precise instructions. But Botswana's experience can serve as an example successive readjustments were made following an for countries that face frequent weather-relatedor evaluation of the program after three years, and economicproblems. the program has produced better results. Comparison with a control group showed a In Gambia, a supplementationprogram for significant impact; the amount of weight gained young children (rations were taken home) was set doubled, a direct result of the number of meals up in 1981 in 42 centers and was accompaniedby received. The program was targeted at children a nutritioneducation program, growth monitoring, who had an arm width less than 13 centimeters and primary health care; it will remain in place (they were underweight). This is an exemplary until 1994. In 1987, there were 104 functioning program in that it progressed from an emergency centers. There is a high level of participation (97 aid program to a comprehensive and relevant percent), and coverage rates are about 18 percent educationalprogram. It led villagers to consider (rural areas). The program was targeted at useful changes in their cultural practices, children in risk zones under five years whose especially an emphasis on the use of peanuts. It weightiage ratio was less than 60 percent, but led to a project to distribute land for cultivation, after an evaluation it was recently refocused on seeds, and fertilizer to women with the goal of pregnant and nursing women and their children providing health centers with basic foods (peanuts, under two years. The food assistance program beans) for the supplementation programs for was also modified to substitute locally produced young children. weaning foods for imported foods (Lotfi and Mason, 1989). In Botswana (Quinn et al., 1988; Lotfi and Mason, 1989), the success of the supplementation iv) Costs program was to limit the impact of the 1985-86 drought on malnutrition levels. Because The average cost of programs conducted in householdfood availabilityproblems under normal the 1970s was between $15,000 and $25,000 conditions remained, the system was annually (1976 U.S. dollars) per recipient for institutionalized on a permanent basis: the daily rations of between 300 and 400 calories. program's success is due precisely to recognition Food costs represent about 70 percent of this of the existence of the problem and putting inmto figure in general. This is relatively expensive place a supplementation system (interministerial given the projects' apparently low degree of committee, creation of a department of food effectiveness and compared with other types of resources). The supplementation program was intervention (for the record, $0.05 to $3 for complementedby a special program to rehabilitate nutrition education, $15 to $50 for integrated children and a "food for works program. The programs, according to Anderson et al., 1981). main result was an exceptionallevel of coverage; Duration and the level of supplementationare two as a consequencethere was no sudden increase in important determinants of the cost. the rate of malnutritionduring drought years, and -

86 Food supplemaeation Accordingto Horton'scalculations (1992), the beneficiariesreduced the unit cost to $4.60. costs of 58 prograns conducted worldwide Extendinga programis alwaysless expensive than differedaccording to the program's target: the setting it up, especiallyif targetingefforts have cost of distributingsupplements of 1,000calories alreadybeen completed. In Botswana, the cost per day to each recipient annually is $75 for fell from $62 in 1985-86 to $38 in 198748 untargetedprograms, $64 for targetedprograms, following an increase in the number of $74 for programsrun throughschools or maternal beneficiaries(the cost in this case is an average and child health services, and $134 for highly estimatethat includessupplementation for children targetedprograms (1988 U.S. dollars). On the and the 'food for work' component)(Lotfi & other hand, Hortoncalculated that programsthat Mason, 1989). were moderatelybroad in scope (between 100,000 and 500,000 beneficiaries) were the least In Tanzania, in the context of the Iringa expensive. JNSP(Ghassemi, 1989; WHOIUNICEF, 1988) in 1988,the program'sannual cost (includinggrowth The cost of programsin Morocco between monitoringfour times a year, therapeuticcare, 1975 and 1979 reached$35 to $42 annuallyper and food distributiontargeted at childrenbetween beneficiary,depending on the value of the fbod zero and 36 months) was $17 per beneficiary. ration (which varied between 500 and 800 This includes$3.60 for the preparationphase, calories,depending on the year). Programsalso $5.30 for the extensionphase, and $8.05 for the included a significant educational component routine phase. It is probablyone of the best (motherswere asked to contributeabout $7 per programsin termsof cost-effectiveness,as would year); the cost of the food representedabout half be expectedin an integratedprogram, since the of the total cost (Gilmore et al., 1980). In managementand personnel costs are divided Kenya, in 1980 for high-energyrations (770 amongthe differentactivities. calories),the annual cost per person was about $68; the mothers' contribution,although it was The cost of Africanprograms is generally10 modest(five Kenyan shillings per child)was high times more than programs in Latin America, for the poorest families(Hoorweg, 1989). In regard]ess of the cost of food and its Lesotho in 1987 (USAID,Lesotho, 1988), the transportationto the countriesinvolved. This is estimatedcost was $49.97 per beneficiary,$97 linked to transportation costs within these per child under five years, $97 per child under countries,although the numberof beneficiariesof twoyears, and $386 per malnourishedchild; food each progran is generally much lower; the representedabout 60 percentof the totalcost. In problemsof a widely scatteredpopulation and Zimbabwe in 1980-81 (M.O.H., Zimbabwe, accessibilityare often considerablein Africa 1982),the program's cost was $12.84 per child (Bremer-Fox,1987). The c3sts are higher for annually (62 percent for food, 38 percent for donor countries, the NGOs that manage the transportationand administrativecssts). distribution(21 percentcompared with 9 percent elsewhere),the countriesreceiving the food aid In Gambia, for a populationof beneficiaries whosecontribution is also proportionallyhigher, totaling312,000, the cost was on average$8.90 and the beneficiariesthemselves: $1.46 on per beneficiaryannually; the extensionto 600,000

87 The health sector and nutrition interventionsin Afrca average compared with $1.01 (Bremer-Fox et al., treatment; a high level of medical supervision and 1987). regular evaluation of the treatment results are key to a program's success.

2. Nutritional rebabilitation * Centers should not be treating more than 20 to 30 children at a time; they should have specific admission, recuperation, and discharge a) FNDINGSAND RECOMMENDATIONS requirements; they must have a registry, even if only a simplified one, that allows for regular * The success and failure rates of home-based evaluation of results. rehabilitation are comparable to those of health centers; it is the most attractive alternative * Centers should be reserved for children who wherever primary health care is available and are severely malnourished. Treatment of children there is an adequate level of medical supervision; with a moderate level of malnutrition at centers is it is not expensive, and the family and health care less effective. system is fully involved. Home-based rehabilitation is probably the only approach that * Indispensable to the long-nn success of these can treat all malnourished children in the long programs is their integration into a coherent run. Initiatives of this type should be evaluated package of prevention, screening, reference, and quickly and carefully in the future. follow-up programs.

* Inpatient or outpatient rehabilitation centers are b) DErINMON a better choice than the hospital, which should be a last resort only. Nutritional rehabilitation is concerned with the process of directly addressing cases of severe * Inpatient centers are effective when the malnutrition to ensure recovery and population density in the surrounding area is low convalescence. or malnutrition is not very prevalent. In urban areas or densely populated rural areas, outpatient In additiou, the aim of such rehabilitation is to centers are the most appropriate and economical prevent any relapse and to eliminate possible scars solution. in the medium to long term. Quite often, it is also used for children with less than severe cases * The prevalence of malnutrition in a given area of malnutrition. Finally, it frequently includes should determine the location of th'e centers to other types of interventions; traditionally, the avoid their underuse. Transportation probl-- s t:za tys of nutrition programs that target young limit the impact of the centers to their immed, : hit;L' a:e considered separately: nutritional area. -%Aittir i. supplementation, and nutrition educati. *.ve. mne, however, each program There should be a health facility nearby that has tended ._ nrrow elements from the other can provide a minimal amount of medical two, as Hoc.h, points out (1988), althougn it is

88 Nuiritional rehabilitation not always clear which is the most effective The cost of rehabilitationunder these hospital element in a given program. This series of conditionswas so high that it was impossible to programs, which is not limited to therapeutic envisage providing rehabilitation in this form on measures, also includespreventive ones discussed a broader basis. Thus, centers specializing in below. nutritionalrehabilitation that were adapted to local conditionsand able to handle a larger number of c) JuSTIFICnONs children with severe or moderate malnutrition at a reasonable price were developed. ]3engoa Preventive nutrition measures should be proposed the general principle on which these promoted initially. There is no doubt, however, centers were based in 1957. They began to enjoy that given the somewhat limited success of the real success when it became apparent that, with measures implemented, severe malnutrition will the treatment methods used at the time, the still affect a small percentage of preschool mortality rate was much lower in these centers children in Africa for, a long time to come, than in hospitals. independentof any unusual circumstancessuch as famine or food shortages. Although the number That these centers could play a role in the of children involved is low, it is still considerable education of mothers during the course of their in absolute tenns and is sufficict to fill all children's treatment became evident after a short available beds in the pediatric ward if care for time, and as a result, they took on a new these children were left to hospitals and other dimension; the goal was no longer, just to similar institutions. It is clear that these children rehabilitate severe cases of malnutrition at a low need immediateand appropriate treauncnt, and if cost but also to avoid relapses, which were rather they do not receive it, they are likely to die frequent in children who were treated at hospitals. quickly. Becausethe centers were increasinglylocated The dramatic 'revival" of these children is closer to the population,emphasis was placed on often a good launching point for nutrition increasing the participation of the mothers. The initiatives at a community levei; the population centers began to operate more frequently in iarge, pays more attention to prevention advice if the integrated spaces that provided preventive and/or ability to treat those who were ill has been therapeuticcare. Efforts also were made to make demonstrated. the mothers themselves dLr=cdyresponsible for severely malnourished children in their homes d) BACKGROUND under the looser supervisionof health personnel.

Nutritional rehabilitationwas the first type of The success or failure rates and the cost of intervention implemented in most non these centers seem to be about the same as those industrializedcountries. It was initiallyprovided of other countries; this is also true in the case of in specialized hospitals in conjunction with operationalproblems and their possible so!utions. research worK,and the foundationfor the various There is really nothing uniquely "African" in this treatmentsthat were subsequentlyprovided almost area, nor is there a problem of transfer of any sort everywhere was laid within this framewoi'k. of new technology. Rather, the problen is one of

89 7he health sector and nutrtion intervenrionsin Africa a lack of exchange of experiences within and rehabilitation in a hospital, which is five to ten between countries to ensure that positive results times higher than that of a specialized center; the are achieved uniformly throughout the continent. failure, death, and relapse rates; and the lack of instructionprovided mothers during their stay the e) HOSPIrAL-BASEDREHABITATON hospital clearly represents a waste of resources.

In the first extensive assessment of the Has the situation changed since then? First, effectiveness of hospital treatment for cases of in terms of short-term results, particularly the severe malnutrition, Cook (1971) pointed out that number of deaths during hospitalization, most all of the articles published between 1956 and reports indicate similar results. Teyssier et a]. in 1969 reported a mortality rate between 8 percent 1983 in a large hospital in Dakar, Senegal, and 52 percent, with the majority between 20 reported a risk of death of 43 percent, similar to percent and 40 percent. He also noted that there the level already reached in 1964; Hazoume and had been no improvement in 15 years. Among Toukourou confirmed that hospital treatment of the 24 reports cited, 11 originated in Africa, malnutrition at the University Hospital Center indicating that the situation there was the same as (CHU) in Cotonou, Benin, in 1982 was long (30- everywhere else in the world. The reasons given 45 days), difficult, and still had a death rate of 33 at the time included: admissions that were too percent; Tolboom et al. in 1986 in Maseru, the late in a number of cases, the frequeut separation capital of Lesotho, reported a mortality rate of 25 of the child from his/her mother, infections that percent; Assimadi et al. in 1984 also mentioned a were often acquired in the hospital, the time and mortality rate of 20 percent to 26 percent at the patience recuired of hospital staff often CHU in Lom&and pointed out that the average overwhelmedby other tasks to feed children, and rate for all hospitals in Togo at the same point in the low quality of care in hospitals outside of the time was 26 percent, reaching as h.gh as 35 university or research hospital networks. percent in some cases. In 1990, Atakouma, also in Lome, still reported a mortality rate of 25 Other factors include the number of deaths percent. during the first year after release from the hospital, which is between 14 percent and 37 The cause seems to be clear; the hospital is percent of children who received treatment, or a not really the place to treat severe malnutrition. mortality rate of more than 50 percent. Many of Although during the same time pe.d, much the survivors still suffered from malnutrition in better results were achievedby a number of pilot varying degrees at the time of the most recent centers elsewhere in the world (a few percent in check up. In contrast, these relapse and mortality Jamaica, for example), and in nonhospitalcenters figures after less than five years are much lower even in Africa, many pediatricians have become for alternative institutions, such as nutrition convinced that the special characteristics of rehabilitation centers or dispensaries. malnutrition in Africa result in a mortality rate that is necessarily higher. However, Andrien Considering the number of availablebeds and notes a reductionin risk of death from 36 percent the prevalence of severe malnutrition; the cost of to 10 percent at the Bouake Hospital in C8te

90 Ntritional reiabilirauion d'Ivoire between 1970 and 1981; in 1986, Van and private hospitals. The operatingassumptions, Roosmalen et al. reported a mortality rate at which correspond to the current consensus in the hospitals in Tanzania of 9 percent; Teyssier et al. area of treatment (WHO, 1982) included the acknowledge that a change in hospital treatment following: initial stabilizaticn of dehydration, methods for severe malnutrition in Dakar, treatment of infections, and the use of a balanced Senegal, in 1984 led initially to a reduction in the and high-energy semiliquid diet. The results mortality rate from 43 percent to 20 percent. showed that in most cases, hospitals did not Success is not directly linked to recruiting less operate at a satisfactory level and improvement severe cases; in Niger, in a rural hospital setting was necessary and possible. Only 21 percent with a minimal amount of equipment, which was provided an adequate and accurately determined not part of pediatric services, Pecoul et al. (1988) diet; among the possible causes of this is the lack registered a mortality rate of less than 15 percent of basic ingredients such as powdered milk, oil, (the rate is almost 50 percent for patients who and sugar, which should be distributedto hospitals leave the hospital on the first day!) for cases of on a therapeuticbasis. severe malnutrition. An evaluation of the length of treatment in the The best results obtainedby Teyssier et al. in hospital, another limiting factor of a hospital- Dakar and those noted by Andrien at Bouake are based approach, indicates that hospital stays of related to more rigorous control of the treatment two to dtree weeks are long enough, which helps process, particularly of feeding the child, after to reduce costs; any additional time is not evaluating the results, reviewing previous justified, and it is clear that an institution more operating principles, and more effectively basic than a hospital and nearer to the family's motivatingproject leaders and the entire staff. On home should provide follow-up care. the other hand, in a zone in Zaire where forms of severe edema are common, in 1983 Janssen et al. Additional factors that lirit the success of tried to decrease the high rate of mortality hospital treatments: excessive delay in recruiting recorded in their hospital (20 percent to 30 patients and improperly selecting them. The percent) by using more sophisticated and hospital in Bouake (Andrien, 1983) attributes its aggressive medical techniques including in improvement in part to the screening by health particular parenteral treatmnentfor their patients. care institutions referring cases to the hospital The mortality rate remained urLchanged(29.8 earlier. This is an important aspect of the percent), indicating that these techniques did not hospital's role: it should be integrated into the provide the solution to the problem; in fact, this circle of health care institutions for both type of treatment does not shorten the duration of admissionsand discharges. In 1990, Atakounia in treatment or improve its quality and includes a Lome confirmedthat 25 percent of admissionsare high risk of infection and complicationslinked to not appropriate because of a lack of exact transfusions, given the conditions under which admission criteria. This is a problem at all of the most hospitals in Africa operate. rehabilitation centers: few of them have well- defined admission criteria, and this should be In 1984, Hone conducted a nationwide addressed by precise recommendations for the assessment in Zambia of the operation of public entire health care system.

91 T7hehealth sector and nutrition irnterventionsin Africa

The other major criticism of rehabilitation It is clear that, with a few exceptions (Van institutions is that their failure rate in the long run Roosmalen et al. in Tanzania, 1986; Bac in renders the money spent on thern in part an Bophutatswana, 1986 for example), there is absolute Loss. Van Roosmalen et al. in Tanzania practically no coverage of cases of severe (1986) mention a death rate of 8 percent, mainly malnutritionin the region by these hospitals. during the first year, a relapse rate of 13 percent, arl a rate of maintenance of good nutritional In conclusion, the hospital is the institutionof status of 75 percent. In 1987, Hennart et al. in last resort among health care facilities available to Zaire note a subsequent mortality rate of 15 treat malnutrition,particularly in severe cases. To percent over five years, with a rate of 9 percent limit their cost, hospital stays should be short, and for the first year; the other children remained as soon as the treatment no longer includes underweight and at a low height-for-age. measures that are specifically medical but rather Okeahialam in Nigeria (1983) notes a death rate are based on diet, the child should be referred to of 8 percent, a less than satisfactory rate of a more appropriate institution that costs less and subsequent growth for the survivors (rapid is able to provide long- term follow-up. stagnation in growth). This is the major problem faced by these institutions: what becomes of the Precise admission and discharge criteria children who are partially rehabilitated at the should also be well defined. hospital when they return home, when the mother has not received even a minimumnamount of Mortality rates should not exceed 10 percent education during the hospital stay and without any in the worst-case scenario. If rates exceed this follow-up? level, the treatnent provided should be reevaluated with reference to recommended The Bouake hospital addressed U'.equestion international methods; it is also advisable to of education during the hospitalilay, and Andrien confirm that staff members have sufficient time (1983) evaluated the results. In Andrien's view, and training to provide good-qualitycare and that providing instruction is feasible. It is difficult, supervisionof the care is competent and effective. however, because doctors in charge of pediatric care must be willing to make preventionof infant On a national level, the results achieved in malnutrition a priority, a central agency and local different regionalhospitals should be evaluated on administrative body must support the effort, a a regular basis and their practices coordinated. training program for staff in charge of nutrition education must be established, and staff members An effective methodology exists; the must be given a certain status to give weight to conditions under whiclh it should be applied so their efforts in this area. Finally, nutrition that rehabilitation is successful have also been education requires a commitment to confronting known for some time (Bengoa, 1957; 1967; the problem on an individual level and to 1976). The lack of opportunityon a national and establishing a real dialogue with the mother so international level for practitioners to compare that this brief contact is useful to her and responds their work, which is in part related to the very to her specific concerns. independent character of most of the public and

92 Nutritional rehabilitation private institutions (no systematic assessments), is animals, health and nutrition education, food the reason for this failure. distribution), or tleir degree of ntegration into the community (local management comnittee, f) NURmONAL REHABUTATION CENTE commuut' participation in maintenance of the center). There are two types of nutritional rehabilitation centers: inpatient centers, where the To assess the performance of these centers, a mother and child stay full-time during the entire number of criteria are used. First, their impact course of treatment, and outpatient centers, where on nutrition is evaluated, which is divided into the mother and child go on a daily basis for a two phases, as in the case of hospitals: the certain amount of time and return home at night recuperation phase or the center's initial There are a number of variations of the lattzr therapeutic success, which corresponds to the first model that involve changes in the frequency and month of treatment, and the long-term phase in duration of the mother's contact with the center. the subsequent years, which is a measure of the visits may be on a weeldy or monthly basis or success of the instruction provided to the mothers, several times per week or month during a period although other variables inevitably affect the of a few weeks to several montis. outcome. Another criterion is the decrease in malnutriton rate among the siblings of treated Some centers, especially the inpatient ones, children and in the commurnity of which the are attached to or near a hospital or other family is part. Fmally, the center's cost- significant health care institution; they may be effectiveness must be evaluated. closely related, as in Dabou, Cote d'voire (Bouvier, 1985; Ferre, 1991) or in i) Short-erm success Bophutat,wana (Bac, 1986) where the hospital controls the center, or loosely associated with it. * Risk of deah: It appears evident from a Other centers function independently or even in an reading of most of the reports that the centers isolated fashion. They are sometimes grouped have a risk of death during inpatient care that is together under a shared supervisory and significantly lower han most hospitals, or at least assessment structure (Whyte et al., Kenya, 1989). no higher than the most effective hospital institutions (Bengoa, 1976). Larchet et al. in Another important distinction is the presence Burkina Faso (1977) give a death rate of 4.6 or absence of long-term follow-up of the families percent; Ojofeitimi and Teniola in Nigeria (1980) and the community. Sometimes this is provided recorded 3 percent; Jansen and Verkley in Kenya by mobile teams that also ensure screening in (1986) reported 6 percent; Birem-Etchebes and villages (Bac, 1986; Ferre, 1991; Gressart, 1974). Gonzales (Cenbral African Republic, 1987) give 12 percent; Whyte et al. (Kenya, 1989) estimated They are also distinguished by other that the mortality rate among their centers varied operational characteristics: in terns of between 5.4 percent and 10 percent; finally, Beau recruitment, nature and number of activities et al. (Senegal, 1990) indicated that the average (associated medical treatment, cooking risk of death was 2.6 percent (but 11.4 percent for demonstrations, gardening and breeding of small children with kwashiorkor). This result is enough

93 7he health sector and nutnrior. interveions in Africa to justify rehabilitation in a center rather than a fully reflect the nutritional status at the time or the hospital, except fur very difficult cases. One progress actually made during the course of factor is certainly the reduction of the number of treatment. infections acquired in a hospital setting because of the small number of children gathered together in WHO (1982) recommends daily weight gains, a center and isolated from children with the types depending on weight, of 50 to 100 grams during of infectious disease normally found in the the recuperation phase. Asokumar and Enahoro pediatric ward of a hospital, in Nigeria (1991) mention daily weight gains of five to 20 grams, depending on age and the * Rateofattrition: Animportantobstacleto recuperation phase, which is low; Whyte et al. the success of these centers remains the high rate (Kenya, 1989) indicate that 26 percent of of attrition or premature departure that is noted to participants have a weight gain of more than one some degree almost everywhere, as was already kilogram for treatnent lasting more tan three the case for hospitals: Gallin and Pecoul in Niger weeks, 29 percent have a gain of between 500 (1938) in a mobile program had a 39 percent grams and one kilogran, and 34 percent have a attrition rate; Dagn6lie in Kenya (1979) had a rate gain of between 100 grams and 500 grams, while of 35 percent for recuperation treatnent lasting 1I percent lost weight: a weight gain less than two to three months; Larchet et al. in Burkina 500 grans during this period is insufficient; Faso (1977). 25 percent; Beau in Dakar (Beau et Birem-Etchebes and Gonzales (Central African a., 1990; Fontaine et al., 1984; 1987) gives rates Republic, 1987) believe that 30 percent to 45 of 6 percent to 13 percent, depending on the time percent of children are cured with a weight gain period (1984-1990), with treatment lasting an of 800 grams in 20 days; Beau et al. (Senegal, average of 12 days compared with 20 for the 1990) estimate the average weight gain to be 10 others. gramns/kilogram/day (which is approximately WHO's recommendation). Regining weight is on Some of this attrition is explained by the average much slower than in a hospital; but this is severity of the cases: the parents are convinced compensated by the longer duration of treatment that the child cannot be cured. Other cases are a in general, from at least 20 days to several result of the duration of treatnent or its cost for months. the mothers (transportation in the case of outpatient centers), ii) Long-term success

* Nutrifional saus: the criteria used most * Relapse and death rates after discharge often are weight gain and the weight-age index. from the Center: Ojofeitimi and Teniola indicate In reference to the latter, Beghin and Viteri, in a 7 percent relapse rate; Larchet et al give a their 1973 overview, consider recuperation death rate of 7.8 percent, a relapse rate of 3 effective in 70 percent to 80 percent of cases in percent, a stagnation rate of 12.5 percent, and a well-maintained centers; the failure rate is favorable progress rate of 76.5 percent; Birem- probably a result of the choice of the weight-age Etchebes and Gonzales (1987) had a subsequent index for admission and discharge. which does not death rate -of 12 percent on average during six

94 Nutritionlwrehabilatoago years of operation and a relapse rate of 20 al. in Zaire (1979; 1980) studied the effect on percent, of which half recuperatedagain under siblingsand did not considerthat the impactwas favorableconditions. In fact, this aspect is not positive; but the choice of controls is never adequatelyevaluated in mostcenters; it is difficult perfect in this type of study, and the centers to estimatethese rates after the fact, since the evaluateddid not function well (there was no center loses track of a significantnumber of improvementin the short or longterm in children participants. Centeradministrators, however, are receivingtreatment). Therefore,no conclusions certainlyincreasingly aware of this problemand can be drawn. are tryingto set up a more effectivehome follow- up programduring the first year after the child iii)Reasonsfor the effectivenessorfaikue of has been discharged. center depening on their characteristicg

U Nutritionalstanw of childrenwho have Short-termvariations are the result of many survivedand not relapsed: theseresults are also factors: difficult to assess for the same reasons. The general impression is that, following several U Inadequateration size: low energy density, monthsof rapid growth once the childrenleave infrequent meals (see, for example, the the center, their growthpatterns are identicalto assessmentof Whyte et al. of 12 centers in those of children their age who have not Kenya,1989). experiencedsevere malnutrition. Some consider this a success;others considerit a failureof sorts U The recruitmentlevel: children suffering (this probably depends in part on the nutritional from severe malnutrition regain weight at the satus of childrenof dte same age in the region). most siking rate; the effect is not as Kraut (1978) observed excellent results in the impressivein casesof moderatemalnutrition, short term by followinga treatmentsimilar to that when in some cases the treatmenthas no reconnnendedby WHO/FAO (1971) (over a effect at all. The criteria used to recruit period of three to five months),but no difference children,therefore, must be appropriatefor at all in the long run with childrenof the same the capacityof this Wpeof center. age. Rehabilitationis generallyincomplete at the time of dischargefrom the center;the durationof Otherfactors related to the centers' operating treatnent is too short to havea long-lastingeffect; conditionsinclude: and the impact of education, in view of the conditions under which it is generally provided * A significant problem {hat is frequently and the fact that it calls for foodsthat are largely encounteredis the staff's inabilityto foUow unavailablelocally, is not very convincing. As the developmentof children and male the living conditionshave not really fundamentally apprte decisionswhen a problemanses changed, it is not surprisingthat the long-term becauseof the lack of an exact "guide" to impactis negligible. recuperation.To remedythis state of affairs, Asolkmarand Enahoro(1991) propose using * Impacton the youngersiblings of children a systemto calculatea score based on the receivig treatmentand the community:Brown et variabilityin weightgain accordingto body

95 Thehealth secnor and utritioninterventions in Africa

weight. This system would be determined by Outpatient centers are always unsuccessful each center based on its average results. when the mother and child spend only a few Another method proposed by Beghinand Van hours at then: the mother should stay at the Lerberghe (1989) is the target-weight. The center for seven or eight hours each day so applications are still limited, and it is that she can participate in meal preparation, impossible to say which of the methods is educationalsessions, and other demonstrations more likely to improve the outcome of while ensuring that her child receives at least treatment provided by the centers. The need three substantial meals. This simple change exists, nevertheless, and should be addressed has improved the performance of several by the use of appropriate written materials centers (Beau, personal communication). In (follow-up guide and simplified set of rural areas, the inpatient centers generally instructions in case problems arise) that are produce better results, at least in the short widely distributed (Ferre, 1991 in Dabou, for run, than the outpatient centers. example). U The length of contact with the center: four * Korte and Patel (1974, Tanzania) compared weeks seems to be the maximum required for the two types of centers: inpatient and educationto be effective in the long run; there outpatient; the average length of treatment is is no evident benefit for longer periods of slightly shorter in the former (41 days) than contact (Schneidman et al., 1971, in in the latter (47 days), which cancels out the Uganda). Usually, the result of a several- cost difference for the same degree of week stay in a center is an improvement in rehabilitation. The inpatient centers offer the the mother's understanding of malnutrition possibility of admitting children whose and food, but little change in her habits. mothers are not able to travel, although in When there is change, it is usually reflected these cases recuperation takes much longer. in the mother's offering a greater variety of The mother's presence leads to better results foods to her child but not more food. This in the months that follow undoubtedlybecause certainly explains the lack of improvement in of the instruction she receives at the center. the child's nutritionalstatus in the longer run. The population covered by an outpatient center is necessarilysmaller than that covered An alternative that is frequendy explored: by an inpatient center, except in urban areas once the delicate first week has passed, or densely populatedrural zones. In fact, two centers ensure prolonged contact with the important factors favor inpatient centers: the mothers and their children through a weekly mothers contact with instructors is much or monthly visit to a center or in their village closer. and the child receives a sufficient to weigh the children and to provide nutrition number of meals during the day if the center and hygiene instruction, cooking is well managed. In these cases, the mothers demonstrations, and sometimes food. The recall more easily what they learned at the hope is to maintain contact with the mothers center, and home visits demonstrate that their for a longer period of time while avoiding the habits have changed. saturation period noted after several weeks at

96 Nutritionalrehabilitation

the centers. However, participationrates and shouldbe targetedon an individualbasis. dwindleas time progresses. Lemaireet al. In these cases, associated agencies must in Burundi (1990)registered a follow-uprate committo addressingthe problemdirectly in of 75 percentover an averageduration of 20 socioeconomicterms, especially in urban weeks,with good resultsfor severecases, and areas;this is whysocial workers assisting the less positiveresults for the moderatecases. centersare needed,especially in urbanareas. Wittaker et al. (South Africa, 1985) had Sometimes the distributed food requires sustainable results for the 80 sessions on specific preparationinstructions and is no average (at various intervals) over eight longer available once the recipient returns months. Verkleyand Jansen(Kenya, 1983), home. Oncethe first phaseis complete,meal who had a mixed approachthat was mobile preparationconducted at the centers must and followedby a home-basedcheck up (one include foods that are available at home. visit monthly for six months), obtained Innovations should be limited (utensils, disappointingresults: motivationwas linked cooking techniques),so that they ae more to the distributionof food; it disappearedas easilyadopted by the mothers. soonas distributionceased. U Some centers were located in relatively Gressard (1974, Rwanda), cites a mixed isolatedareas becauseof the concernabout systemof long duration;after eight days in a being close to the population. The centers center and supervisedtreatment at home for shouldbe close to a health care institutionto one month (the average recoverytime), the guarantee access to medical care, which mothersattend a "schoolfor parents" once a inevitably will be needed, and to group monthfor two years: the averagefollow-up together health care services at the same time was 28 consecutive months for 1,500 location to avoid umecessary trips for the families! Many factors may explain the mothers. Centerslocated at a hospitalhave a success of this system, including the disadvantage: recipients are sent to the prevalenceof malnutritionin the region or hospitalfor the smallestcomplication, and the even the distributionof food to poor families. frequent result is a back-and-forthwhose Educationchanged the mothers'understanding effectivenesshas not been proven. In of malnutritionand relapsesin the longterm, addition,the mothersbecome accustomed to and fewer new cases of malnutrition in the medical treatment approach to familieswith one child alreadyaffected were malnutritionand resist the idea of a center reported.Providing all the healthcare-related that does not provide medical services activitiesat the sameplace and same timefor (Bouvier,in Dabou, 1985). the mothers,keeping the distanceto the center as short as possible,and providinga concrete U After Bengoa's publication, many centers programall help to encourageparticipation. were created with the objectiveof ensuring strictly dietary treatment. The * A goodlaunching point in somecases is food "demedicalization"of the process was distribution,but this can have a distorting designedto showmothers that treatmentwas effect;it is onlyjustified for destitutefamilies possibleto a large extent under simplified

9. The health secdorand nutrition interventionsin Africa

conditions. It is now clear that treating U Several assessments mention the degree of associated infections is also necessary, integration into the connunity. If the inoculating against measles after the first few confidence of the villagers has not been won days of life is an effective measure, and beforehand, no preventive measure will be providing good medica: supervision by a effective. One way to encourage the pediatrician and possibly a nutritionist (who conmnunityto participate is to set up a local cannot be recruited for the center's needs health care conmittee or invite an already alone) are essential conditions for a center's existing committee to participate in the success. operation and management of the center (Stanfield, Uganda, 1971). The trials Among the reasons regularly advanced to described, however, were not always justify the above is the staff training required successful in this area. in rehabilitation techniques. This factor is rarely addressed in the descriptions or The centers' longevity is related to several assessments of the centers, and one may factors besides their therapeutic effectiveness. suppose that it is rarely considered. First, the source of financing is important. In Nonetheless, it is reported that staff is practice, many centers remain dependent on incompetentor insufficientlytrained. Regular external financing, which varies with economic internships in the pilot centers for all staff conditionsand is therefore very vulnerable to ups sent to the dispensaries or other institutionsto and downs in the economy. At the same time, screen or treat severe malnutrition would be external financingplays a bigger role during times an effective way to increase and consolidate of economic upheaval (Whyte et al., 1989). knowledge in this area. Dependenceon external sources of financingoften accentuates the independent nature of centers U Some caseswould suggest that encouragement comparedwith other public health institutionsthat is more effective than education (Glatthaar et balk at providing the centers with more support al., South Africa, 1985). The fact that the than that provided by their customary services. best "educators" when the children return to This source of conflict (Beau, personal their communitiesare the mothers who play a communication)must be avoided by putting the dominant role in social terms rather than the centers under the official medical supervision of mothers who have received a good evaluation the regional public health authority. The political would tend to confirm this. Introductior.of support of local officials and officials of the motivational techniques at the centers (which various ministries concerned should be sought: would require staff training) could be an this is how a coherentpolicy can be put into place additional success factor. In the same way, little by little. The problems of building on the community level, support provided by maintenance,training, and staff availability can be dynamic and motivated individuals in the addressed orny if they are considered from the conmnunityis more effective than education beginning; this implies implementing a over the long term (Lifanda, Burldna Faso, comprehensive administrative and financial 1988). strategy to attain the desired goal and coordinating

98 Nutritional rehabilitation with other nutrition and health programs in the centers is not appropriate. The Dabou center in country (Stanfield, 1971; Whyte et al., 1989). C6te d'Jvoire (Ferre, 1991)conducts a systematic screening test given the considerable increase of fl CosrsOF WE CENWRS AIDS in rent years (14 percent in 1988). The risk of death is very high: 63.6 percent compared In 1973, Beghin and Viteri cited the cost of with 10.4 percent for those testing seronegative one day at the center, which was 10 times higher (Rey et al., 1990). Whyte et al. (1990) mention (five to 18 times) than one day at the hospital. this increasing concern in various centers in That stays at a center tend to be longer than those Kenya. Any indicationof generalizedadenopathy in hospitals, however, must be taken into account. and/or tuberculosisrequires screening. Rehabilitationis also more thorough at the center, and the risks of death or relapse are lower: the g) CErNERUM)TA7TONS cost-effectiveness relationship, therefore, clearly favors the center. Although the rehabilitation phase should no longer pose any particular problem, the long term The exact operting cost of the centers is not is stil! a stumblingblock. The best results of the always mentioned. Cook indicated an average education provided in the centers are obtained in cost of $80 to $120 for a hospital stay in 1971 reasonably fertile zones with families that are not compared with $12 to $13 for a center. too poor. In the more deprived zones, feeding or Atakouma (1990) cites a figure of $130 in the supplepr.ntation programs are probably more C.HUof Lotd, Togo in 1990. Korte and Patel in effective (Hoorweg, 1988). These regions often Tanzania in 1974 estimated the cost of have the highest rates of severe malnutrition, rehabilitation in an outpatient center at $56 (47 whether temporary or permanent in nature. days), an inpatient center at $79 (41 days), and $83 for a 21-day hospital stay. Beau (1990) Instruction at the centers is an effective way esimated the cost of an outpatientcenter in Pikine to increasethe mothers' knowledge, in particular (Senegal) to be $12 (20 days) (but the center's in making them understand that the causes of the upkeep and part of the food for the mothers are illness are linked to food; this influences their provided free of cost). acceptanceof the treatment. On the other hand, the possibilityof changing mothers' attitudes and The emergence of AIDS threatens to disrupt behavior in other areas remains limitedwhen there the operation of some centers; young children is no further contact with them after they leave the contaminated by their mothers becone ill rather center. early and end up in the rehabilitationcenters as a result of their weight loss, without having yet There does not seem to be a relationship been diagnosed as seropositive for HIV. These between increasing the mothers' knowledge and children, like those infected with tuberculosis, for the child's ability to catch up during the course of example, do not recover or recover very little treatment. Rather, instuction seems to enhance during their stay. The increase in their numbers the mothers' ability to reproduce what they have may hurt the centers' effectivenessand may worry learned (Korte and Patel, 1974); based on this other participants, and their admission to the fact, it is misleadingto think that the mothers who

99 Thehealth sectorand nutritioninterventions in Africa have improved their knowledge are good at long term. Finally, he believes that rehabilitation demonstrating what they have leamed in their is technicallyfeasible at home as long as the local communities,as some programs propose. health service provides adequate supervision.

For Whvte et al. (1989), these centers alone Tellier provided the first assessment of a will never be effective on the community level; home-based treatment program in Zaire (1990). they run the risk of failing if they are set up in a It is based on an actual signed contract between vacuum as a local strategy for improving living the mother and the health care center. The health conditions. care center provides the medical treatment that may be needed during the entire course of The gap between the estimated number of rehabilitation(13 weeks) as well as a weekly exam children who could benefit from treatment at a with weighing and individual counseling. It also center and the number actually treated is furnishes a sufficientquantity of soy for the local considerable. Whyte et al. (1989) cite, for manufactureof pap, and the mother agrees to give example, 72,000 children under the 60 percent the pap four times a day to her child (soy is weight-age ratio in three provinces in Kenya in widely produced and used in this region). The 1989. The number of children actually treated is financial contribution of the family equals the around 2,000 in the FLTP centers. Even if this price of the soy provided. The mother agrees not figure were doubled or tripled to take into account to wean her child if she is still nursing when the services provided by other health institutions, malnutrition is diagnosed (this is true for 44 the result is far from the estimated number of percent of the children). This system is attractive those needing treatment. This poses a real because of its complete independence of foreign problem; the long-term solution seems to lie in sources, its integration into the local health care treatment at home under the supervision ot system, and the active participation of the family. primary health care center-. The assessment concerned about 100 cases; the mortality rates are not higher than those for other h) HOME-BASED REHAlBLITA770M types of rehabilitation; weight gain is slower. A clear improvement was observed in 67.5 percent In view of the limitations of vertical of the children at the end of the contract, a slight interventionsand the cost of centers in rural areas, ir-provement in 15 percent, and no improvement inexpensivehorizontal solutions were investigated: in 16 percent. There was only one failure, and 18 this is the case of home-basedtreatments. Beghin percent left the program. The success rate is believes that the cost of a center may represent definitely comparable to that of other centers. A one-fourth to one-third of the cost of a local lack of progress is attributable to the family's primary health care system; although this cost is motivation (irregular visits) and the fact that much less than that of a hospital, it is excessive in health care workers lose their motivation as the some poor regions. He also thinks that only a results become disappointing. Weekly visits semn rehabilitation program conducted entirely by the to produce the best results: more frequent mother can really prevent relapses and the meetings are too burdensome for staff and appearance of malnutrition in siblings over the

100 Growthmonitoring andpromotion families. Good medical supervision remains particularpossiblenervoussystemscarring. Stoch essential. and Smith (1976) in South Africa, Fisher et al. in Zambia (1972), and Hoorweg (1976) in Uganda Lifanda, in Burlina Faso (1988), gives an noted a slight permanent effect on the brain. equally enthusiastic account of a similar setup. Despite the methodological differences among For her, success is a function of individualized these studies, their conclusions are similar to those attention and regular follow-up. The mother's more recently produced by Galler et al. in motivation and patience are more important than Barbados (1987), who conducted a meticulous the food given, which more than anything else, study of fine motor skills and the intellectual serves to attract people initially. The commitment performance of young adolescents between the of the staff and important or influential people ages of 10 and 15 after their nutritional guarantee success on a conmaunity level. recuperation. Aside from minor differences noted between children suffering from kwashiorkor and i) ThE VERY LONG TERM OUTCOME FOR wasting, most of these children still have some REWABILr7 ED CHrIDREJN trouble in these two areas, confirming a slight unrecoverable impact on the nervous system. The first concern has to do with recuperation Because of associated environmental factors of low height-for-age children. In fact, while related to the family, a strict control is evidently weight begins to increase inunediately during the difficult. course of recuperation, very often height does not increase or increases very little during the first several months, and many children remain 3. Growth monitoring and promotion noticeably shorter after several months. Satge et al. (1970) studied in Senegal what became of a number of rehabilitated children in the longer a) FRDINGS ANDRECOMMENDA7TONS term; somatic recuperation seems complete from the age of four to five, with the exception of a * Given the cost and apparent inefficiency of marked delay in the bone growth. Briers et al. the various known programs, a reexamination of (]975) in Uganda and Stoch et al. (1976) in their objectives and operations is necessary. South Africa reached similar conclusions. Bowie Monitoring growth can be a screening method that et al. (1980), in South Africa, and Cameron et al. follows each child's growth, a social welfare tool, (1986) attempted to follow recuperated children or a way to bring together different initiatives to beyond the age of puberty. Their growth during inprove the nutrition situation. Programs are puberty is less but is more prolonged; although rarely able to integrate all of these aspects these children generally remain smaller than their simultaneously. reference groups in somatic terms, they are similar in size to other children of the same age in a Programs that do not really use the weight the community. gain information as a means of educating mothers should be discontinued, and an alternative use Other aspects of recuperation in the long run should be found for the time subsequently were considered in a number of studies, in available for the mothers and staff. The available

101 Thehealth sector and nutritdonhnterventonw InAfrica

resources should be dedicatedinitially to staff * Growthmonitoring should not be the first training in the nutritlon problems of young activity implementedin the hope of attracting children, methodsof treating them effectively, patienlsto the health care center, but rather the imnrovinghealth care coverage, educatingthe last, when the overall health care system is communityabout the problemslinked to poverty operating adequately and providing good for the most destitute, the lack of time for coverage,educational sessions for the community mothers,etc. Growthmonitoring is not rquired are available,and the problemsof integrationinto to achievethis result; it is a usefuladdition when - and parLicipationof - the communityhave everythingelse is in place. been resolved. At this time, a real growth monitoringsystem would give familiesthe ability a An exact strategy including possible to interprettest results, decide which interventions obstaclesto decisionmaking and implementation to undertake, and verify the effect of these shouldbe defined,along with steps for a regular changeson the growthof their children. In this evaluation: Are the targetingmeasures adequate context, weighingis only a way to monitorthe to achieve the objectives? Are the decision outcome: it may be doneperiodically, probably criteria or the method of applying the strategy on a quarterlybasis, and may serve as a method correct? Alternativesbased on the answers to of nutritionmonitoring for the community. these questionsmust be designedand tested. U Nutritiontraining and supervisionmust be When this objectiveis met, a program of strengthened,especially with the prospect of individualgrowth monitoring for all childrenmay increasing coverage by a diverse and often graduallybe put into place with the intentionof illiteratestaff. actuallymonitoring growth. Criteriafor enrolling in the program and completing it should be * Given this, promoting a single model defined. Monitoringshould be promotedfrom withoutadapting it to localconditions (such as the birth (not beginningat six months)until the child qualityof healthcare services,population density, reaches 36 months. This monitoringshould be and accessibility)should be avoided. Growth very simple, based on the principle that a promotingmeasures may also be implemented competentperson who is able to communicatethe withoutnecessarily weighing the children. appropriatelessons shouldbe availableto weigh the children and counsel the mothers. b) DEFINITONS

* If the health care system decides;to Growth monitoringcovers all activitiesthat undertake an initiativesuch as monitoringand involve following the growth of individual promotinggrowth, it mustbe integratedinto other preschool-agechildren on a regular basis to healthcare activitiesto ensure optimizationof the determinetheir health and nutritionstatus over the system's operations, costs, and effectiveness: course of time and to interveneto improve it singlehealth committee, single health care system whennecessary. information system. linking interventions to identifiedproblems, etc.

IA? Growthmonitoring and promotion This generallyimplies invitinga group of defined, the cost and feasibility of existing mothers and their young children to regularly alternatives must be analyzed: systematic scheduledweighing sessions; after recordingthe measurementsof arm width in villagesto screen weighton the appropriatechart, the evolutionof for cases of severe malnutrition, small-scale each child'sweight relative to his age is explained surveys to determine the prevalence of and various health or natritioninterventions are malnutritionor the Improvementresulting from an suggested. intervendon,etc.

This type of activityhas becomean essential i) Relkionshipbetween growth and mortity element of primary health care services in a in young children numberof countries;its implementation,however, has metwith so manyobstacles that some consider The relationshipbetween malnutritionand it doomedto failureand wonderwhether it should mortalityhas been demonstratedby now classic still be implemented when more effective studies in India and Bangladesh(Kielmann and altemativesexist. Thesedoubts concern not just McCord, 1978;Chen et al., 1980;Briend et al., Africabut all of the developingcountries in which 1986).In Africa,work in Zaire (KasongoProject this initiativewas put into place (Gopalanand Team, 1983),Tanzaia (Yambi, 1988),Guinea Chatteree, 1985). Thus, the questionbecomes Bissau (Smedmanet al., 1987), and Senegai whetherthese frequently encountered obstacles are (Garenne et al., 1987) helped .o refine our a result not only of the conditionsin which the understandingof the problem. In zones where measuresare carried out (Hendrataand Rohde, mortalityis low and the nutritionduficit roughly 1988)but also of an error in the underlyingtheory hornogenous,as in the case of Kasongo,the risk on whichthey are based, or at least a gap between of deathalso seemsto be homogenous,regardless the ideas and the objectivesactually pursued, or of nutritionalstatus. In Tanzani and especially the inabilityof the proposedactions to reducethe in zonesin the Sahelwhere there are highrates of risks linked to a slowdownin growth (Lotfi, mortality and malnutrition,the relationshipis 1988;Nabarro and Chinnock.1988). Thecurrent proportionalto the deficitin nutritionindicators. approachis to refer to growth Monitoringand Linkedmostly to the weight/heiglhtdeficit in the Promotion to showthat the ultimateobjective is early years, the,risk of death may also be related not monitoring. to stunting in older children. The weight/age indicator,which combinesthe effect of the two c) WHYMONIR C?IIDREN'SCROWTN? phenomena,and arm width representthe closest linkbetween growth aid mortality. Growthis monitoredbecause it is one of the most accurate indicatorsof nutritionproblems. Studiesexamining transversal and longitudinal Growth monitoringmay also address different measurementsat more or less closeintervals (one objectives,as illustratedin this chapter. Clear to six months)have shownthat the latter werenot and realisticobjectives must be defined for this better. The relationshipis importantespecially type of activity,which requires a highdegree of for short periods of time and for significant involvementon the part of health care workers weightloss at ages whenmortality is high (six to and families; each time a priority objectiveis 24 months) (KasongoProject Team, 1983, in

103 The hcalth secrorand nutntion interventionsin Afnca

Zaire; Briend and Bari, 1989 in Bangladesh; ii) Relationship between growth and Yambi, 1988 in Tanzania; Maire et al., 1989 in morbidity Senegal); but in this case, growth monitoringchat does not include a weight record or even just the The most widely held opinion is that mother's opinion may be enough. In practice, malnutrition predisposes a child to infections, or transversal screening (such as arm width), which at least lengthens the time it takes to treat them, is less expensive and easier, will be more while at the same time infections may be effective. responsible to a great extent for cases of malnutrition. A two-step approach appears to have been somewhat successful: screening of the most Briend recently examined in the published malnourished children and monitoring of these literre the validity of these two relationshipsas children's growth on a monthlybasis to intervene far as diarrheal diseases are concerned (Briend, as soon as weight begins to fall or the child has a 1990). The first view, that is, that malnutrition severe infection. This system probably explains predisposes a child to diarrhea, seems to be self- how the SCF project in Gambia obtained a evident (in Africa as elsewhere: Tomkdns, reduction of 50 percent in the infant mortalityrate Nigeria, 1981 and Gambia, 1989; El Samani et comparedwith national levels: 83/1000 compared al., Sudm, 1988; Maire, Senegal, 1990). On with 167/1000 (as long as other factors do not the other hand, although diarrhea certainly has an intervene simultaneously)(Shorr et al., 1989). In effect in the short term on weight gain, it is not Benin, more or less regular contact with a clear that it has a long-termeffect. Other than for comnunity health care worker at home or during very difficult conditions in which food is scarce, a consultation turned out to be a factor that it seems that catch-up growth happens naturally, protected against the risk of death (Velemaet al., all the more because most diarrheal episodes in 1991); this factor alone has nothing to do with conununities are moderate (Moy, 1990). Under growth monitoring in the form in which it is these conditions, preventing infectious diseases carried out. If the desired effect is on mortality will certainlyhave a beneficial effect on children's more than anything else. :t may be better to free mortality and physical well-being, but it will have up health care workers for home visits with the little effect on improvinggrowth (edito-ial, Lancet best possible coverage than to impose growth et al.. 1991). In a study conducted in Gambia monitoring sessions on them. (Rowland et al., 1977), the quantity of food ingested by the cnild was the main limiting factor Finally. demonstrating a link hetween in the off-season: in these cases, diarrhea does malnutrition and mortality is not enough for an not make a great difference. Dickin et al. (1990) improvement in nutrition to lead automaticallyto in Nigeria, arrived at the same conclusion: a reduction in mortality; malnutrition may variations in energy intake related to diarrheal represent only one of many confounding factors. episodes are small, while the energy deficit in This link has yet to be demonstrated (Nabarro and daily intake, even in the absence of illness. Cinnock. 1988). represents the most significant problem. Black (1991) concludes that even if diarrhea has an

104 Growth monitoring and promotion unambiguouseffect, managingto provide adequate optimal tecouping of weight and height: for nutrition under ordinary conditions clearly has a Lutter et al. (1990), this would be between three more pronouncedeffect on growth than preventing and six months, for example; programs that do fever and infections, including the practice of not enroll mothers from the time their children are ORT. The relationship is obviously more born or that do not advise early supplementation meaningful in the reverse: maintaining adequate to avoid the risk of changing breastfeeding food intake may decrease the effect of infections routines and thereby increase the risk of death while preventing infections has little impact on from diarrhea cannot benefit from this window of growth. opportunity. Although similar results are not available for Africa, this infonnation should be iii) Relationship between growth and considered. Armar-Klemesu et al. in Ghana improving nutitional status (1991) dernonstrated that early supplementation with a traditional fermented pap can ensure Growth patterns observedin young childrenin excellent growth rates without significant African comunuities demonstrate that after a drawbacks as far as infectious diarrhea is rather quick take-off period, the deficit increases concerned (Mensah et al., 1988). This approach progressively around six or seven months. can make a considerable difference for children Graphs illustrating the rapidity of weight and whose mothers are no longer able to provide height increases on a curve show a steady adequate amounts of breast milk during the first accumulated delay compared with stndard six months. measurements until about 12 to 18 months; after that time. some weight is frequently gained back When the prevalence of underweight children around 24 months. At this point, the problem is is significant, as a result of a lack of food that the accumulated growth deficit in terms of availability, targeting food aid to low-growth height is still present and persists in a more or less children may be effective (Botswana program; pronounced fashion during subsequern years, pre-harvest gap periods in the Sahel). This is not depending on the circumstances (Van Lerbeghe, the case when growth deficiency in terms of 1988). height is at issue.

This has implications for growth monitoring The effectiveness of growth monitoring in based on weight gain, which is generally how improving nutritional status is difficult to verify growth is monitored. An increasingly significant insofar as it is never conducted in isolation even part of the slowdown in weight gain is related to if it is implemented independent of other the delayed growth in terms of height; this measures. At the very most, the programs with deficiency represents the most important the greatest nurmber of hWalthand nutrition quantitative indicator in many African countries components seem to have the greatest effect (cf. (national survey in Congo: Cornu et al., 1989; Iringa, in Tanzania, where the program includes national survey in Swaziland: Serdula et al., micronutrient supplements, which could have 1987a: Carlson and Wardlaw, 1989)- The contributed to its effect on growth in terms of situation becomes more complex given the fact height). A representative transversal survey that some ages seem to be better than others for conducted beforehand to evaluate this aspect

105 *hehealfh seaor and nutrition intervendons in Afnrc

seems to be the most important factor missing; discrimination to all mothers, without any this component should be systematic (if necessary, preselection. by means of simplified techniques such as arn width measurements). Limited coverage is clearly the major obstacle in every case. A national program like the one in d) GROWrHMONrTORG OnEcMrs Indonesia, analyzed by UNICEF in 1990 and considered relatively successful, has an average i) Growth monitoring as a selection coverage rate of about 40 percent. In Zaire, technique to improve the effectiveness of recent analyses of various programs indicate an health care services average coverage rate of about 30 percent (Gerein, 1988; Ceplanut, 1990). This is the case Growth monitoring is often used as a means of numerous programs in Africa (see below). A of screening children who should benefit from population selected in this way has several specific specific intervention measures: medical treatment, characteristics: a common geographical location food supplements, treatment for malnutrition, based on the distance from the gathering point, social services, nutrition and health education, etc. and the mothers selected were the most concerned To be effective, this step must satisfy very with the health of their child (generally the most specific conditions: good level of coverage for educated, especially in rural areas; see Ceplanut, the population, specific selection criteria, effective 1990) or the least overburdened (Gerein, 1988; interventions, and lower costs than those of the Lovel et al., 1984; etc.) In other words, the proposed intervention. lower the level of coverage available, the less selec ion through growth monitoring will be able An example of the problems linked to to screen for children who are most in need of the selection through growth monitoring is provided proposed services. by a project in Zaire (Gerein, 1988): the selection of "at-risk" children was based on the If growth monitoring is to be used as a tool to existence of an infection at the time of weighing provide better therapeutic and preventive and/or a slowdown in growth in terms of weight. supervision of children, its first objective should This strategy quicldy failed: 44 percent of be to increase coverage. Unfortunately, available children seen presented a problem with gaining assessments show that the use of services provided weight and another 20 percent had a health by health care centers is often not associated with problem; screening thus detected 64 percent of growth monitoring because these programs are "at-risk' children whose mothers were invited to managed completely or in part independently participate in educational sessions. Because this (Gerein, 1988; Diene, 1989; Nyiribibi, 1990). A type of screening method effectively excludes one- system that requires the mother to spend hours in third of all children, it is not efficient in theory; transport and/or in the waiting room to hear that this is particularly so because the children will all she must return on vaccination day to finish the fall under the "risk" category at different times. injections needed has an unacceptably low cost It is better to attempt to provide education without effectiveness (of the three programs analyzed by Gerein in Zaire, two deferred preventive

106 Growthmonitoring and promotion

treatment for infections to the next consultation; of children gaining or losing weight; the potential on average, 35 percent of children had an of longitudinalmeasurements has not been fully infection when they were weighed). exploited.

ii) Growth monitoringas a tool to mobiize Following an evaluation conducted in seven health care officils and health care worken counties (two of which were African), UNICEF proposes (UNICEF, 1992) not encouraging Many growth monitoring programs have put growth monitoring without integrating growth a data collection and analysis system into place to promotion measures. A three-step growth follow developments in the community's monitoring program has even been suggested: nutritional status. These results may have many uses: they may be feedbackfor communityheath U promoting growth without growth care workers and officials and a monitoring tool monitoring; for government authorities or even a means of appealing to decision makers. This activity * promoting growth with selective growth occupies a significant amount of time of the monitoring, which requires that the workers responsible for monitoring (existenceof communityrequest this and that sufficient several weight recording aids: individualweight resources to train staff to monitor are record and "master chart" in the CRS projects; up available; and to three in Gmby, Senegal, none of which, moreover, was adapted to the preparaton of a i promoting growth with comprehensive report: Diene, 1989). growth monitoring. This option reqires the most resources, and the first two Longitudinaldata are often interpretedfrom a phases should be completed before this is transversal perspective; a system of rapid undertknl transversal surveys at repeated intervals is more efficient and less costly. The problem is whether e) OBSlrArAs TO PROGRAMEFFECTVE1ESS IN the data are representative: Serdula et al. in AFRCA Swaziland (1987b) demonstated that at the country level, compared with the results of a i) Objectivesin project assessments in Africa national survey, these data were useful if only new participants were taken into account (how A survey of the literature on these programs representative they are varies inversely with the suggests a global desire to reduce morbidity and regularity of visits); regional differences do not mortality and improve the nutritional status of appear to be reliable. CRS has demonstratedthat children in the commnity. At the same time, analysis of monthly or seasonal trends based on growth monitoringin the field is assigned limited monthly weight data could be done in a objectives, which vary greatly from project to satisfactory manner, even if these data were not project and which are not always in agreement representative of the nutritional status in the with overall objectives. region in absolute terms. There have been few attemptsto analyze the trends in term of numbers

107 The healthseaor and nuintioninternetnons in Africa

Among the best assessments conducted are decision makers at various levels so that they those in several regions in Zaire since 1987 apply the necessary resources to enhance the (Gerein, 1988; CEPLANUT/UNICEF. 1990). children's growth and development. In practice, Before the 1970s, a growth monitoring program different programs emphasized different called "kilo" by the participants themselves objectives: screening for severe malnutrition, existed; it was coupled with vaccinationsessions, selecting beneficiaries for admission to various and its actual objective was to ersure that all intervention programs, community contact, participants were fully vaccinated. Inplemen- educating mothers, mobilizing decision makers tation of growth monitoring programs along with with the data collected, etc. Very few programs primary health care programs then spread to other have been able to pursue all of these objectives at parts of the country. Although for a significant once with any degree of success, as has the Iringa number of mothers, the purpose of the program is program in Tanzania. still vaccination, the goal of the health care services is primarily to screen children who are at In Ghana, in the rural Kitampo zone, growth the highest risk for malnutrition. monitoring was initiated in 1987 as a starting point in a comprehensiveprogram of health care Several initiatives related to growth and agricultural activities (Shorr et al., 1989). In monitoring added a distnbution progran of Gambia, an NGO established this activity in a weaning foods to families at the same time; this remotely accessible zone in 1985 to serve as a was the case for most of the programs supervised point of contact for various health care services by CRS, which used North American food aid to and increase community participation (ref. id.). Africa for this purpose. Participationwas closely In Senegal, in the rural areas of Fatick and linked to food distribution. In fact, the main Kaolack, several growth monitoring experiments purpose of this type of program was to select begun several years ago are still in existence groups needing supplements and to evaluate the (Diene, 1989);the health authorities are concerned effect of the latter on the children's growth. with producing an "integrativematrix" based on the preventive activities: vaccinations, malaria One of the objectives of the Joint Nutrition prevention, prevention of diarrhea, etc. For the and SupplementationPrograms set up in various time being, the objectives being pursued in most African countries over recent years (PCAN1JNSP, of the projects are screening for severe run by UNICEF, WHO, and FAO) was to reduce malnutritionand collectingthe type of information the morbidity and mortality of infants and young that will permit nutrition monitoring of the children by enhancing their growth and comnmunity. development. This generally involved a community development strategy that included The criticism generally leveled against these prominently among its activities growth programs is that the staff (and the mothers) often monitoring, with two ostensible objectives: to trust their understanding of clinical assessments allow the community to judge the extent of the and the child's approximate weight more than the malnutrition problem and find ways to address it; directionof the weight curve (APHA, 1981; Lovel and to explain the problem clearly to planners and et al. 1984; Gerein, 1988; etc.). Does this mean

10 Growthmtorneg andpromoion

that intuitively they think that this way of judging (Zaire: Gerein, 1988 and Ceplanut, 1990) or less a child's health is more relevant to the objective denselypopulated areas (Sahel zones). Rwan's being pursued? Before even considering whether national program (Nyiribibi, 1989) estimaes its the technical aspects of a program have been coverageto be less than 20 percent; as previously effectively implemented, the theoretical noted, assessmentsmade in Zaire between 1987 effectivenessof growth monitoring in achievinga and 1990 produced an average figure of 30 number of objectives could be debated, as Lotfi percent (from 12.5 percent to 41 percent). This has done (1988). figure varies between 25 percent and 35 percent in Somala (ISS, 1989). In Congo, it is 63 percent ii) Coverage, waotherobstacle on average but less than 40 percent in the north, where access is difficult (Comu et al., 1990). The problem of coverage represents a significant obstacle to growth monitoring, as it Coverage usually depends on the distance to does for most health care interventions. It is not the health post: in Zaire (Ceplanut, 1990), 37 always precisely understood. percent of participatingmothers live at a walking distance of less than 10 minutes; 52 percent, Some programs have achieved spectacular between 10 minutes and one hour; and 11 percent, results in terms of extending coverage initially: more dtan one hour. These mothers would like to Cole-King in 1972 in Malawi reports that see these activities decentralized to the village coverage of children under fnve years by the level, or even the hamlet or neighborhoodlevel health care systmnwas increasedfrom 17 percent (Diene, Senegal, 1989; Cornu, Congo, 1990). To to 40 percent in two years. Morley (1973) cites make up for the limited number of staff, the coverage levels in excess of 80 percent in Ilesha solutiongenerally adopted is to leave the weighing (150,000 inhabins) and Esa-Oke (7,000 sessions up to community health workers or inhabitants) for pioneer projects in Nigeria, with women's organizations. Another solution is to a high rate of regular attendance, and 67 percent empower other agencies to conduct monitoring in Zambia. Currently, some programs are still other than that associatedwith healdLcare: social producing good results (M. Bac in Bophutatswana development workers and Multipurpose Rural attains coverage rates of 90 percent in a very Extension Centers (CERP) in Senegal, for integrated and decentralized system; the Iringa example (Diene, 1989). This requires rigorous program attained a coverage rate of almost 100 training and constant supervision. percent: Moneti and Yee, 1989). One of the reasons for the impact of the Botswana program, Targeting children at risk may be achieved based on the distnbution of food to children through a quick transversal survey to determine determined to be at risk by the growth monitoring the conununitiesat risk; the objective is to prevent program, is the program's high level of coverage, malnutrition in these groups rather than just to which reaches almost 80 percent of the target monitor their growth. In Angola. the national population (Lotfi and Mason, 1989). But many program was initiated in a small city by drawing other programs have a modest level of coverage up a completelist of children under five and their of their target population, especiallyin the case of nutritional status and designing a map indicating large-scale projects in less accessible zones the houses of the most malnourished children.

109 T77cheakh sector wad nutrnon ntefenWtionsin Afnca which allowed them to be includedas a priority in and her proximity to the center; the regular the program (Delahaye, 1983). follow-upvisits decrease from 72.5 percent when getting to the center is easy to 38.6 percant when The mothers frequentlymention that they did it is difficult (Cornu et al., 1990). The distance not participate in the program because they did constraint is generally the main reason why not know about it. In Zaire (Ceplanut,1990), on follow-upis not conductedon a regular basis. average, 52 percent of the populationhas access to health care services, but no more than 30 Despite a very structured program with a percent of the populationuses themLThe program good level of coverage in Bophutatswana, Bac in Rwanda is attemptingto create village health (1985) notes that the average number of visits care committees to spread informationabout the yearly per child is seven after five years of program's existence,motivate the community,and operations. One of the difficultiesis related to the encourage it to participate (Nyiribibi, 1990). variety of people accompanyingthe children: in These committeesshould not be differentfor each 32 percent of the cases, it is the mother alone; in health care agency: in Senegal, growth 36 percent, it is alternately the mother, monitoringand food distribution initiativesled to grandmother, older sister, or aunt; and in 32 the creationof specif1cmothers' committees;how percent, it is someoneother than the mother who to integrate them into the health care committees accompaniesthe child each time. This is why the is currently a problem (Diene, 1989). In Ghana, entire family needsto be aware of the importance the FFH program used local information of regular follow-up. Some programs have transmissiontechniques (tam tam); they appeared, institutedhome visits the day before the sessions however, to favor one social group (Shorr et al., for the mothers who did not participate in the 1989). previous session (FFH, Ghana: Shorr et al., 1989). iii) Reguar participation, another obstacle Another essential factor is the mother's The constraint of growth monitoring perception of the benefit of follow-up; in this compared with other health care interventionsis context, followingchildren of a certain age when that it requires a significant number of the benefit of doing so is uncertain should be consultations, regardless of the urgency or avoided.and an attractivearray of services should seriousness of the child's state of health. This be offered at the same time as follow-upvisits. essential point is underestimated, however: growth monitoring makes sense and may be Many programs initiallychose to monitor the effective only if follow-up occurs on a regular growth of all children of preschool age. basis, which allows the speed of growth to be However, children over 36 months represent 15 determined. In Congo, follow-upvaries quite a percentto 25 percentof participantsat the centers bit, depending on the centers: the minimum is in Zaire, for example, according to Gerein once every four monthsover approximatelya 12- (1988). Severalprograms are currentlyproposing month period. Effective follow-upis determined to reduce the age for monitoringto 0-36 months, by the age of the mother, her level of education, as was the case in the PCAN/JNSP programs.

110 Growthmonitoring and promotfon

This seems to be a reasonable objective because 36 to 60 months(Nyiribibi, 1990). The effect on the problems related to being underweight, not participationin and organization of these services includingexceptions to be verifiedbeforehand (the has not yet been determined. The Iringa program Kivu in Zaire, for example), occur before the age in Taania, which has had some success, of three. Beyond that age, the subsistingproblem operates on the basis of quarterly visits (Moneti is mainly growth deficiencyrelated to height that and Yee, 1989). growth monitoringcannot easily address. The age groups effectivelyrepresented in practice are often Simultaneouslyproviding a variety of services fewer: most regular visits consist of children that meet the mothers' expectations in particular under 18 months or even 12 months in urban (preventingintestinal parasites is often requested: areas (often linked to the vaccination schedule). Senegal, personal observation; Guyon in Zaire, Efforts must be made, however, to persuade personal observation) or that have a dramatic mothers to continue visits on a regular basis impact (food distribution)helps to promote loyalty beyond this time frame; only 4.6 percent of to the program; the disadvantageis sometimesthat children are more than 12 months old in Congo less attention is paid to promotinggrowth as such: (Cornu et al., 1990). As such, in some urban in Zaire, 60 percent of mothersbelieved weighing areas where social welfare programs are in place sessionswere one phase of the vaccinationprocess (Dakar, Senegal), participationmay be a function (Gerein, l988)! Inversely, halting food of incentive payments made until that age, and distribution in CRS programs resulted in a huge one solution would be to increasethe age at which drop in participation (Diene, 1989; Nyiribibi, the last payment is made (this applies, however, 1990). to less disadvantagedsocial groups). Several additionalsolutions were attemptedin Monitoring should begin at birth rather than various places, but their effectivenessat the local six months, as is the case for some programs level and ability to be reproduced elsewhere have (such as Gniby, Senegal: Diene, 1989), because not yet been assessed. In Rwanda, for example, the curve's direction is determined in the first few mothers receive a 'certificate' at the end of the months, and it takes several months for the weighing and educational sessions (Nyirbibi, mother to understandall of the counselingshe has 1990). In Senegal, in the Kaolack region, several received. The initial counseling sessions should initiatives of this type were attempted: reassure the mothers that breastfeeding is going commercial (sewing, truck farming) or social well and encourage them to continue with it as activities, weighing sessions at the homes of long as possible; the curve should be the actual village chiefs who were supportive. etc. (Diene, individual guide to gradual supplemertation 1989). The problem is to ensure that health care (Greaves and Kendrata, 1990). services summarze and share their experiences and results witi other centers. Finally, a prior One alternative is to modify the spacing of survey of local customs and discussions with the visits depending on age. The program in Rwanda conmnulity itself will avoid scheduling conflicts recomnends one visit per month from zero to 12 with family or cultural ceremonies and activities months, one visit every three months from 12 to relited to agriculture, which is a reason frequently 36 months, and one visit every six months from given for missing appointments. Other reasons

111 T7hehealth sector and nuintionintemions in Africa given include mothers' illnesses and the Simondon(1991) has shown that birth weight completion of vaccinations (Shorr et al., 1989; in Senegal and Congo has a definite predictive Ceplanut, 1990). value for moderate malnutrition and growth deficiency in terms of height. Specific targeting In spite of the time lost with the sessions, based on the birth weight with the goal of which is between two and six hours in Zaire for maintainingthe highest level of growth possible in example, for many mothers these sessions are a these children could be somewhat effective; this social event that encourages them to participate in has not yet been tested. Along the same lines, she the program, even though weighing takes only noted that height at six months has a predictive two minutes, the individual consultation, two value for later growth deficiency (supporting the minutes, and the group instruction, 10 minutes. conclusions of Henry et al., 1989, and Huttly et They are afraid of criticism by their neighborsor al., 1991, in other countries); although the health care workers, however, when their introducingroutine height measurnems may not child is not well; any approach that tends to be realistic given the difficulties this would create reprimand the mothers of children at risk, for a number of programs, she believes therefore, has a negative effect on participation measurementscould be taken at certain key ages (Gerein, 1988; Ceplanut, 1990). (six and 12 months) to prevent or make up for the height deficiency that occurs between these two iv) Targeting, anotherobs.acle ages. This targeting is worthwhile, however, only if one knows how to prevent or recoup the height If the goal of growth monitoringis to promote deficiency at this stage, which is not always the normal growth, there is no reason to modify case. targetingmeasures, other than for the age problem as referred to above. In programs with limited f) OT OBSTACLESTO IMPLEMJNG THE resources, however, it may be prudent to target SERWVCS the highest risk communities; in cases where malnutritionis very prevalent, it may be necessary i) Weghing to target as a priority the highest risk children. Few examples exist, however, of this type of To measure weight gain accurately, a precise targeting; it should probably be based on polls or scale is needed (within about 100 grams); even screening conductedbeforehand- The SCF project inexpensive hanging scales are adequate. in Burkina Faso (Lifanda. 1988)proceeded along Dilapidated and poorly calibrated scales are still these lines by using very specific screening frequently used (Gerein. 1988) by staff who are criteria as a basis for its targeting. Growth not trained in how to maintain them. Weighing monitoring is more a way of verifying the and recording errors are more often the result of effectivenessof the measures implementedthan of the sessions' having been conducted too quickly promoting normal growth, even if the ultimate than of a misunderstandingof the results (some objective is to use it for the latter once the child sessions involved more than 100 children! In has attained an adequate level of growth again. Zaire. 40 to 50 children were weighed on average, with groups ranging from 19 to 162). In

112 Growlh monitonng and promotion cases like these, more sessions must be provided, any major problems after initial training, but which poses a financial problem. retraining on a regular basis is recommended.

A survey conducted by APHA in 1981 ii) Using growth charts indicated that Gambia, Lesotho, Malawi, Burkina Faso, Zaire, and Zimbabwe as well as The number of curves actually plotted is still Morocco, Tunisia, and Egypt had a national insufficient; in Congo, although 60 percent of system based on the weight curve as a function mothers participate in the sessions, at most 19.1 of age (refer to Harvard, NCHS, or Tanner). percent have a curve that has been plotted (Comu Botswana had an experimentalsystem at this time et al., 1990). One must be able to locate the based on the height/age ratio and the weight on the curve with sufficient precision weight/height ratio and sometimeson arm width. (within 100 grams); unfortunately, most available Kenya (MRC) also opted for a system including curves have successive gaps of 500 grams each, these two indicators and arm width. Programs which makes the task difficult for the health care begun by CRS in various countries were all based workers (eight of 14 charts are correct in Zaire: on the weight curve as a functionof age (APHA, Ceplanut, 1990). Griffiths introduced a "bubble" 1981). Rwanda has just decided to introduce all curve that includes two improvements: first, it of these indicators into its national program has a steeper slope to make the increases more (Nyiribibi, 1990). visible at ages when weight gain is on average rather small; second, "bubble" markers are placed Arrn width measurements give satisfactory every 100 grams and may be counted and checked and inexpensivescreening results and detect cases off (Griffiths, 1987b). A test was conducted by of severe malnutritiopwith a sufficientdegree of comparing this with a traditional linear graphic specificityand sensitivity;they are not sufficiently system, but recorded every 100 grams, in various sensitive, however, for growth monitoring or for countries (specifically in Lesotho in Africa). use in nutrition supervision. Introducing height Fewer weight-reporting errors and more age- poses an additional measurement problem that is reporting ones resulted. Trainers prefer the difficult to solve; it is easy to manufacture the traditional linear graphic system, but the health necessary equipment on site, however, which care workers like the "bubble" systen (Griffiths makes it inexpensive. The JNSP in Sudan (ISS, and Berg, 1988). The new system must still be 1989) refused to introduce the weight/height tested in routine use. indicator because of the belief that it was beyond the abilitiesof the coununity health care workers Another type of chart was introducedby CRS and midwives. In Senegal, in Gniby (Diene, in all of its African programs: it is composed of 1989), indicators used for screening malnutrition tables or contour curves that express the position (weight/height ratio or arm width, depending on in percentage of weight/age ratio (see the literacy level of the community health care weight/height ratio or arm width as in Burkina workers) are differentiated from those used for Faso, for illiterate people) (APHA, 1981). This growth monitoring (weight/age ratio). Dealing system is attractivebecause it gives a straight line with the different indicators does not seem to pose for normal growth and a negative slope if weight gain between two mements is insufficient,

113 The healthsector and nuantioninren'emions in Africa which simplifies interpretation for the health care large variations were noted depending on the workers: the slightest delay in growth is projects: 1 percent loss rate in Nigeria and inmmediatelyevident and leads to an intervention. almost 20 percent in Zambia and Malawi. Lovel The comparison between the "road to health" in rural Ghana (1984) notes that the mothers system and the percentage weightlage line system carefully store the charts and consider them an was made recently in Lesotho (Ruel, 1990 and important document; they always bring them 1991). The reaction of the health care workers along when they travel or go to the hospital. was tested first; the two systems, which were Some programs (FFFH, Ghana: Shorr et al., taught for the first time (one week), were 1989) pick up the charts the day before the generally assimilated equally well. Health care sessions and return them the following day, which workers faced with a system change are still more encourages a good level of participation; but this confused when they go from the "road to health" system works only where small groups of mothers to the percentage weight/age tables; introducing are concemed, otherwise it is too much work to such a change, as some have proposed, is thus not sort the charts. In a center in Zaire, the chart recommended. When mothers were confronted management system uses a specific person at each with the two systems, they performed better if weighing session. they were trairned in using the "road to health" system instead of the system used by CRS. iii) Interpreting growth curves Lesotho adopted the "road to health" system nationwide for this reason. This assessment Insufficient time, vague instructions, and a should be repeated in countries where the literacy misunderstanding of the cumulative effect of level of mothers and health care workers is lower repeated small slowdowns in growth are all factors than that of the Lesotho study. that mean that, in a certain number of cases, weight is not recorded on the curve or that There are numerous variations on the health children who have not gained enough weight will care record or growth charts; the charts must be not benefit from additional studies (20 percent in durable but inexpensive to produce: in most Zaire according to Gerein, 1988). An adequate cases, keeping them in a plastic covering that is physical examination of the children was wider and longer than the sheets themselves conducted in only four of the 14 sessions also maintains them during the monitoring process, evaluated in Zaire (Ceplanut, 1990). despite their being transported and stored at home. Finally, it is important not to confuse people by The principal finding is that health care using several health care records (vaccinations!). workers rely too much on the position of weight on the curve rather than on the curve's Whether the growth charts should be kept at appearance, although it is generalLy recommended home by the mothers or at the health center has that they now concentrate on weight gain rather been frequently debated. Morley (1973) had than the weight/age ratio. It should be explained confirmed some time ago that the loss rate when to them that analyzing weight gain is designed to mothers kept the charts was identical to or even improve theirjudgment and the mothers' ability to lower than the loss rate for health centers, but think critically. This aspect is often neglected or

114 Growthmonitoring and promotion is the subject of a cliched question-and-answer mothers were able to understand the charts, session; the only way to improve their although no efforts were made to instruct them. understandingof the context of children's growth Few women, however, are able to make the is to foster a dialogue, but this requires special connection between the curve and the status of training of the health care workers, which has their child's health (5 percent, according to Gerein rarely been available to date. in Zaire; 2 percent, according to Lovel in Ghana). Gerein notes that in Zaire, mothers Althoughthe absence of weight gain is easy to understandably often still have questions analyze, this is not the case for insufficientweight concerning how to change their childrearing gain. Beyond 24 months, weight gain is slight; techniquesgiven the shape of the growth curve. the accuracy of the scale and weighing, in addition to physiological variations unrelated to growth, iv) Counsding mothers often result in larger differences: daily fluctuationsin weight have a standard deviationof The percentage of mothers who are acually 200 grams for monthly weight gains of less than counseled in these programs is quite variable: 200 grams (Zerfas, 1979);Van Loon believes that from 57 percent to 75 percent in SomaLia (ISS, the conlidncc interval of 95 percent of weight 1988); in Zaire, according to Gerein, almost all measurementsin routine programs may be greater of the mothers know their children's weight, but than one kilogram without close staff supervision only 50 percent of mothers with a child at risk (Van Loon, 1987). Therefore, it is not easy to receive growth information on their child; the determine at what point growth is actually curve is not always shown or explained to the insufficient; the programs do not always provide mothers in this case (10 of 14 in the Zaire clear guidelines. The instructions for the WHO assessment nonducted in 1989 by the Ceplanut). work sheet (1986) and the training manual for Seventy seven percent of mothers of children primary health care workers (WHO, 1986) are with a low weight/age ratio return home without very basic. Steveny (Mokolo, Cameroon, 1982) counseling of any kind in Solenzo,Burkina Faso published an interesting example of the types of (Sauerbornet al., 1989). In Zaire, a counseling decisions made depending on age and the curve's session (when provided) lasted an average of two appearance; the Boga program in Zaire (Gerein, minutes. If growth monitoring is to serve as a 1988), which includes some clear instmctions, "gateway"into the health care system for mothers, was adequately implementedoverall; the KasaLgo as is the ostensible objective in several programs, program, however, had an algorithm-based a sufficientamount of time for each visit between decision p.ocess that was too sophisticated and mothers and staff members must be set aside to had not been used for very long. review all of the child's and mother's health problems. The number of mothers involved in Education level plays a role in the mothers' each session should be limited, therefore, and the ability to interpret the charts correctly; according number of staff members or home visits should be to Lovel et al. (1984), however, in Ghana, 40 increased: this system is necessarilymore costly. percent of illiterate people are able to interpret but at least it will achieve one of the defined tLem correctly after some instruction. In Boga objectives. (Zaire: Gerein, 1988), 76 percent of illiterate

115 The health sector and nutintion interventionsin Africa

If advicethat is clichedor too generalis to be malnutritionin a child, identifypossible causes, avoided,a frequentcriticism of these programs, prevent it from developing,and understandwhy understar4ingthe attitudesand behaviorof the growthmonitoring is importantand how it works. mothersat the outsetis important.This is seldom The assessmentof Unibyin Senegal(Coly, 1989) achieved. In a study in rural Ghana (1984), clearly demonstrates that the mothers who Lovel demonstratedthat conununitieshave their frequenthealth centers without growth monitoring own ways of monitoringif a child is growing know nothingabout nutrition;this differenceis normally (leather braceletsof varying sizes for madeeven clearer by the fact that 26.2 percentof arms or legs, for example),and they have their motherswithout any instructiongive a child only own solutionsthat should not be ignored. This the familymeal while 2.9 percentstill do so after study was also able to show the limits of the instruction (the effect on nutritional status, mothers' knowledgein this case: the notionof however, has not been measured). In Zaire, slowingdown of growth in terms of heightas a according to Gerein, in areas where radios, factor in malnutritionis not understoodby many newspapers,and other forms of writtenmaterial mothers(only 22 percent). The survey indicates are rare, and where there are few women's that, in fact, very few mothersmake a connection groups, informationon the nutritionalneeds of between weight and the child's growth; their children is provided basically by the growth understandingof their child's nutritionalstatus monitoring programs (13 sessions out of 14 remains "transversal." This has also been included a group educational session lasting observed in other assessments. To enhancethe betweenfive and 42 minutes). This is a positive effectivenessof growthmonitoring, mothers must outcome (the same result could be obtained understandand be persuadedthat analyzing weight withoutusing the growthcurve, whichis actually gain in accordancewith the results of successive not really used by more than 5 percent of weighingsessions can help them to monitortheir mothers). child's growth more accurately. But this also impliesthat the staff be trainedin this area and Individualizedinstruction must take into that they know how to talk with mothersin the consideration the mother's educational level; context of their local knowledgeand attitudes. group instructionshould be provided in small They shouldalso workcontinuously to identifythe groups correspondingto the-children's different most effective nutrition messages, to which age groups (which determinesthe frequencyof mothers are very sensitiveand which determine their visits). Discussionshould return frequently their subsequentparticipation (Diene, Senegal, to the same topics to ensure that they are well 1989). The directed nature of this counseling, understood and memorized,and it should be withoutinteraction, is also a result of a lack of interactiveto bring up any questionsthe mothers trainingof healthcare workersin adult education may have and the steps the communityitself can techniques; specific training in this area is take to respondto them (Gerein). essential. Burldna Faso (Lifanda, 1988) presents an Educationalsessions are essentialinitially to alternative: educational outreach sessions teach all mothers to recognize the signs of (observations,demonstrations, dialogue) lasting

116 Growth monitoring and promotion two weeksand givenin the villagesthemselves by still a problem in this program (Diene, Coly, a travelinginstructor. This type of innersion is 1989). attractive and is worth evaluating in other contexts. Trainingis very brief (one to two weeks for severalcommunity health care workers);growth Dietary recommendationsand an improved monitoringand nutrition are notspecifically taught pap recipe have been effectively passed on in many of the nurses' trainingprograms: they through educationalmaterials (Diene, Senegal, still do not considerthem "medical"in nature, 1989). Practical advice, however, concerning which is why they do not tend to initiate or weaningfoods, the appropriateage to introduce supervisesuch activitiesthemselves. Training is them, and the type of supplementaryfoods to not appropriateto the situation: nurses in Zaire provide for young children is not adequately do not mentionto mothersthat they should give provided and is often very cliched or even their childrenmore than two meals a day; they conflicting. Health care workers feel helpless believethe mothersdo not have the time to do so, giventhe difficultiesthese mothers face, and they but they do not offerthem any alternatives.They do not seem to be trained to propose any are also not trained in teaching adults (too alternatives. The results of researchin this area authoritarianand not interactiveenough), and they should be disseminatedquickly (Alnwick et al., have a great deal of difficultyin allowingtheir 1987;M.S.A.S. in Congo, 1991,for example)to knowledgeto be questioned.Finally, retraining is program authorities, in particular concerning as mucha problemas initialtraining. modificationsin the energy densityof the paps and the effect of traditional fermentation The first problem supervisors frequently techniques. encounter is how to define what they should observe: the national program in Angola v) Staff training and supervision produced a specific supervision work sheet to be filled out at the time of each visit (Delahaye, The goal of the Gniby programin Senegal 1983). It is oftenperceived as a way to checkup was to test the communityhealth care workers' on the programrather than an analysisof what is abilityto implementthe tasksrequired to prevent happening,and the resultsand goodinitiatives are malnutrition through growth monitoring and not always analyzed and passed on to other promotion activities: although initially their centers. One key role of supervision is to performancewas rather poor, at the end of one produce clear and up-to-dateguidelines and to year, the workers,including illiterate ones, had sustainstaff motivationover the long term. It is satisfactorilymastered the tasks; the monitoring up to the supervisorsto comeup withalternatives quality depends on their motivationmore than in caseswhere growth monitoring clearly does not anythingelse. An assessmentof the results that favor growth promotion in the community focusedon the Imowledgeand attitudesof the concerned(Gerein, 1988). mothers demonstratedthat they had noticeably improvedcompared with a neighboringvillage Writtenreference materials appropriate to the servingas a control. Supervisionat all levelsis contextare lacking: the supervisors'role should be to adaptthe equipmentavailable for use by the

117 the health sector and nutrition interventionsin Africa

health care workers at the centers. At the very entrust the health post managers with supervision least, the following should be defined and of the health care workers responsible for growth verified: the criteria for inadequate weight gain, monitoring and to incorporate the use of data into the types of possible additional physical the health care monitoring system (Diene, examinations, questions to be asked of the person Senegal, 1989). accompanying the child (depending on the nature of the relationship to the child) steps to be taken Aside from food supplements, few growth by the mother and the health care system, and the monitoring programs systematically include other nature of follow-up (Gerein, 1988). nutrition initiatives. The JNSP (1ss, 1989) combined more or less successfully growth h) How CANGROWTH MONITORING AND PROMoT7ON monitoring and prevention or treatment of anemia BE IMPROVED? in the mother and vitamin A deficiency in the child, monitoring the nutritional status of pregnant i) Though integraon into other healh care and nursing mothers, treating severe cases of acdvities malnutrition, and providing food supplements for moderate cases of malnutrition or a demonstration An assessment of health care services at of weaning foods (Sudan, Tanzania). Some also Solenzo in Burkina Faso (1989) reports that more include participation in agriculture-related than half of children had one of the risk factors activities to help to improve food availability in defined by the center: weight loss or weight cases where growth deficiency in children does below the third percentile; despite this, no not seem to be a result of problems of infection intervention, therapeutic treatment for associated (Mozambique, Tanznia) or activities designed infections, or nutrition treatment was proposed for to increase family income (or even access to them. They were simply given an appointment credits: Tanzania) or fish breeding (Zaire). This for the next consultation. Providing therapeutic type of measure, which tends to promote basic treatment for infections noted at the time of the solutions to problems of growth deficiency, is still weighing session should be systematic relatively rare. (accomplished in six sessions out of 14 in Zaire: Ceplanut, 1990), especially when they are ii) Through integraton into the community: accompanied by noticeable weight loss: the decentralion, home isis situation then clearly represents a risk of death. If the entire team at the center is not involved, the For most progms, integration into the infection is treated out of context, independent of conununity comes through using health care the broader issue of the child's growth vGerein, workers who come from the co.mmunity itself and Zaire). lntegrating growth monitoring into the creating village health care committees. The rest of health care center activities often runs into participation of the community is sometimes the problem of a limited number of health post sought to equip a building or provide minor managers overseeing both therapeutic and services. Overall, however, few programs have preventive activities at the same time. This is known how to use growth monitoring initiatives really a training problem. One solution is to effectively to involve the community in the health

118 Growthmonitoring ad promodon problems affecting them. The communitiesmust pervasive. Integrationof servicesprovided by the still be encouraged to participate (Gerein), and churches into nationalprograms does not seem to this should not be left to chancebut rather be part have createdany problems in Malawi (Cole-King, of staff training. This type of program should 1972) or Zaire. The three Zairian programs probably not be started until the community is analyzed by Gerein are typical of many of the aware of its importance and requests it. Only projects managedby NGOs: they benefit from a communityinvolvement, using growth monitoring more consistent source of financial support and as an indicator of what is really happeningwithin have a more flexible administative operational the community, can suggest appropriate system. They are very open to the community, interventions that are outside the realm of the innovation, and research. In terms of results, health care system (Kasongoin Zaire: work load there do not seem to be significant differences of the mothers). (Shorr, 1989; Yee, 1989).

To increase mothers' participation, the SCF i) SucCESU EXAMPLES project in Gambia (Shorr et al., 1989) held community-wide meetings to discuss malnutrition- A growth monitoring program involving five related problems, the inability of health care regions in Togo (approximately83,000 mothers) centers to satisfy the needs of all children, and in place since 1982 was evaluated in 1986 by thus the establishment of exact criteria before USAID, Pricor, and CRS, workdngwith national sending children for treatment (slowdown in health authorities (Badonte and Johnson-Welch, growth over the previous three months). A 1990; Govement of Togo, 1990). Appaently, women's conmmitteewas elected to help the nurse the program had not had yet an effect. The in rehabilitationprocedures, demonstrations,and mothers were not permitted consultationslastng preparationof appropriateweaning foods. more than one minute, and they were not really informed about the progression of their child's The establishmentof mobilizationcommittees growth. The staff was poorly trained and (FFH, Ghana: Shorr et al., 1989), working if supervised. Given this state of affairs, the possible in conjunction with the health care program leaders moved quickly to improve the committee, may ensure the cooperation of program to make it more responsiveto the needs important people and local political and health of the mothers, who participatedin the meetings. care leaders. If monitoringis indeed the basis of Among all of the types of problems identified, a nutrition managementsystem, the information four were studied by working groups to define gathered mustbe shared with them to sustaintheir solutions for them. To increase the number of support. home visits, "visitingmothers" were recruited to reinforcethe educationalmessages on health and According to the assessment in Zaire, few nutrition; weighing sessions were reorganized to differences seem to exist between the programs allow for longer individualmeetings; supervision managed by private charity groups and those run was strengthenedto allow staff to receive more by nurses worldng in the public health care assistance in technical terms for weighing and system; the nuns are highly motivated, but their consultations with the mothers; finally, home training is not better: the problem of retaining is monitoring by the center staff itself was

119 The health sector and nutritioninterventions in Africa

established specifically for children of weaning nutrition system - an ongoing ability to assess age. These measures were developed based on and change (Moneti and Yee, L989;ISS, 1989)- their feasibility and their cost-effectiveness. Program leaders decidedon a clear objective: The effect of these changes was evaluated to ensure that the population had a precise several months later: the participation of the understanding of what malnutrition is, its mothers was clearly encouragedby the "visiting prevalence in the village, and the measures mothers" and changes in d-teirbehavior resulting available to combat it. Growth monitoring, in a from the advice they received were evident; the simplified form (weighing once every three weighing sessions improved considerably,as the months) served as a tool to discuss the results at mothers took responsibilityfor distnbuting food a village level and underscore the possible causes while staff worked exclusively to weigh the of malnutritionand death. The overall program, children and provide advice for the mothers; the however,also includedhealth care initiatives(with supervisors became more involved in the entire a coverage rate of almost 100 percent), an process, and precise technical mamnals were improvementin hygienicconditions and the water produced. At the same time, the staff increased supply, and a higher level of financial and food their understanding of the community and their security. Of course, it is difficult to determine contact with the mothers by increasingthe number which factors actually played the most important of home visits. role in achieving the results obtained. This operationwas very expensive: $3.60 per child at Although the effects of various measures on the beginning, $5.30 for the expanded program, increasingchildren's growth in these regionswere and S8.05 for its annual operation. In addition, not measured, this program demonstrated how this program had a higher level of operations at different phases of the process mightbe improved, the outset than many other programs (65 percent albeit with the usual reservations regarding the of children already had a health care card in appropriatenessof the initiativesto achieving the 1980). objectives. Senegal, like other countries cooperatingwith No growth monitoring system works CRS in growth monitoring associated with food immediately at the outset; ideally, a periodic aid, was faced with the prospect of no longer evaluationshould be includedto make adjustments receiving this assistance. The reintegration of in the system; the Iringa program in Tanzania growth monitoring without food aid into institutionalized this step with some success peripheralhealth care services as a whole created through the three A concept (assessment,analysis, a number of problems: the need to sustain the and action). In fact, aside from a good initial mothers' motivationwithout the assistanceof food evaluation, this program included integration of distribution, to rethink the financing aspect, the various health care activities into a dynamic collection, and the analysis of data (formnerly process of commuunityparticipation, which are the conducted by CRS), to merge the mothers' basic ingredients of a good health care and committee and the health care committee, to coordinate this post's operations with those of

120 Growthmonitoring and promodon

other health posts also conducting growth exactly match the salary costs in Kasongo or monitoringprograms, etc. This was an occasion Katana in Zaire, for example (Gerein, 1988); to review the different ways of operating on a others like Boga, also in Zaire, or Gniby in local level, to listen to the mothers' requests, to Senegal (Diene, 1989) produce surpluses: this identify the existinggaps, and to try to define a has obvious imnplicationsfor any changes in national strategy based on the sum of these participatingconditions (age groups),for example. experiences. A pilot project was analyzedat the end of one year and readjusted (Diene, 1989; Given that only small projects with few Coly, 1989). This experiment, conductedunder participants and a good management structure other conditions, should be continued if it is seem to obtain results, costs will exceed their appropriateto extend. current level if the desire is to make them more effective, since this implies better staff training j) Cosn programs, more recruiting, and more sessions. Many programs also operate with equipment The average cost per child annually is (scales, growth charts) supplied by UNICEF or between $1 and $2 (salaries, health care records, the NGOs. An increase in coverage means staff transportation costs), according to Van providing more equipment, and thus incurring a Lerberghe (1989). As the UNICEF report in higher cost. Zaire (Ceplanut,1990) underscores, with the high inflationrate of some countries,the cost can vary The cost for families is not always assessed. significantly over the course of a few months In Senegal, the PPNS program includeda growth (between $.70 and $1.30 per child anmually monitoring component coupled with a between the beginningand the end of the study). supplementary food distribution program; H-sed on a total of $.70, investment costs participationcost $.67 per child at each monthly represent $.58 (scales, training and monitoring session. After food was no longer distributed,the vehicle), and operational costs represent $.11 family's contribution was reduced by half. (work sheets, gasoline, salaries of the staff Officials still support the idea of communities' weighingchildren and the supervisors). participatingin financing- and management- to ensure the autonomyand viability of thbistype McGuire and Austin (1978) cite $1.50 for of activity (Diene, 1989). In Zaire, in 1989 the Nigeria (accordingto Gwatiin et al., 1980), $.49 cost for the motherswas estimated at $.25 (one to in Indonesiaat the same time (but between $1.40 seven hours every two monthsby the mother), or and $4.10 currently), and $.98 in India. In one-quarterof that of healthcare operations. This Angola, accordingto Delahaye(1983), the cost of certainly seems high for such a poor result; the this type of program is relatively low when a question is whether this money could have been health care system is already in place and no better used, for example, to purchase additionalstaff are hired. Some systemsare in a supplementaryfood (Ceplanut, 1990). precarious financialsituation: the new registrants

121 The healthseator and nutritioninrerventions in Africa m Nutrition education2

Because nutrition education is included in many of these key questions for health care most health care services, we will consider it planners. Experimentsconducted in this area in separately. Africa will be examinedin light of this document.

Health care intervention programs call 1. Introduction nutrition education "information/education/ communication(JEC)" to highlight its principal characteristics. IEC cannot pretend to make any Nutrition educationhas been a routine part of immediatechanges in the socioeconomiccontext; many health care programs for some time; in activities in this area will thus be limited to 1978 (Austin), a survey of 201 programs in 66 promoting changes in behavior and practices countries mentioned that 91 percent of these within the family to encourage a redistnbution of programs already included it. But the real impact availableresources, particularlyfood, to the child of nutrition education is rarely analyzed; despite and mother. Depending on the circumstances, its almost universal use, nutrition education is this redistnbution leads to a better understanding generally viewed as relativelyineffective. Are the of the mntritionalneeds of the child and mother, a messages inappropriate, the method of change in attitudes (eliminatingtaboos or changing communicating them inadequate, or does attitudes about women and children), or a package education in fact do nodting to improve nutrition? of new and inewxensivetechniques to implement. Nutrition education has been shown to change New knowledge, attitudes, and practices should feeding practices and thus contributeto improving lead to an improvementin the nutritional status of nutrition, and as such, little debate exists the mother and child. concerning the last point. Opinions vary, however, on the first two points. In other words, when a nutrition education program is not 2. Findings effective, nutrition education itself should not be held responsible; one should analyze how the program operated and adjust it: "When a school a Results will not be forthcomingwithout a does not work, it is reformed, not closed, because significantinvestment in this effort. Awarenessof schools are the only way we have of learning; the this fact is still limited, although investment is same holds for nutrition education." essential to the implementationof well-designed programs that can produce results. Nutrition Effective ways of teaching nutridon education educationcannot operate as a low-priorityactivity exist. The issue is how much it will cost to obtain to which few resources are assigned, particularly the expected result and what means are necessary human resources. to do so. Is it reasonableto invest in educationas an isolated measure or should it be accompanied U Professionals (foreign or national) in the by other initiatives? A publication by ACCISCN areas of conununicationsand nutrition should be (Hornik, 1985) has already attempted to address called upon to design and supervise the program.

122 Nutrition education

This is a requirement for a successful program, health care workers, however, must be even if it involvesa greater cost. complementaryand not contradictoryin the least.

* For field and paramedical health care * Movingfrom the level of a pilot project to workers to play an educational role in the a large-scale project is not only a problem of nutrition area, they must be guaranteedadequate magnitude. New variablesare introducedwith the training and support. Training in educational increase in scale of a project. The supervision, activities should be part of health care workers' qualitycontrol, and coordinationofa large project basic and ongoing training. Training should be are more difficult. upgradedconstructively, supervised continuously, and accorded the same importance as other * Generaly, not much demandfor edon medical initiatives. The effect of educational exists; it is not easy to supervise because few activities should be evaluated, and the results methods exist to verify on an ongoing basis cornmunicatedto health care workers to motivate whether the educationalmessages are correctly them. transmitted, reach their target audience, and are complete!yunderstood. Their effectivenesswill * Coverage is the most significantproblem depend on the motivationof those responsiblefor after the relevance of the proposed initiatives,the them- In the long mn, this means broad politicl type of message, and the equipment used. support and a financial commitment, both of Interpersonal commnication is the most effective which are difficult to achieve insofar as few method, but it does not provide extensive results in terms of improved nutritionalstatus are coverage. Even the use of health care and available. communuityworkers, who are inexpensive and abundant, does not resolve this problem entirely. In fact, they are often overburdened,and training, 3. Recommendationson the steps to follow for retraining, and supervisingthem may be difficult- the design of a nutrition education project Mass communicationmay be used to supplement interpersonalcommunication initiatives. It allows greater coverage, and the educationalmessages The goal of an educational program is to may be aired repeatedly. It is not as effective as change behavior - that is, to replace harmful direct contact, however, nor as effective as practices with those that improve nutritional passing on complicatedmessages; finally, as the status; the nutritionalstatus of a given population mass media are not interactive, mass wDl improve oz'y if this change occurs. The communication cannot address individual challengefaced by nutritioneducationprograns is questions. As a result, it is generally used to to identify the interventions that will lead to inform, raise the awarenessof, and encouragethe modifications in a population's practices. To population to consult health workers. A good achieve this, certain steps should be followed in communications strategy will use many designingan educationnutrition program: complementarymeans of communication. The messages transmitted through various media and * First, the situationmust be analyzed. This analysis has several facets: one must li) do an

123 Thehakh sector and utrition intervnions in Africa inventoryof existing programs and their results; * Regularfollow-up and periodicevaluations (ii) study the target populationsin depth, their will be necessary to verify progress toward feeding practices, their problems and needs, the meeting the objectives and immediatelyreadjust socioeconomiccontext in which they live, and the program whenever problems arise. Many their sources of informationconcerning health and programshave been runningfor years withoutany nutrition; and (iii) evaluate the various means of knowledgeof their effectiveness. communicationused amongthese populations. In additionto these studies,an inventoryof available * Nutrition education initiativesare rarely human, financial, and logisticalresources should analyzed. They shouldbe analyzednot only while be drawn up. This prior analysis is necessaryto the project is running, as mentioned in the identifyrelevant and specificinterventions for the previous paragraph, but also upon completionof target populationsthat will be compatible with the project. This would allow one to determine their cultural practices. whetherthe project's objectiveshave been attained and to draw lessons that might be useful in the * Second, based on the results of the prior designof other projects. This analysis, however, analysis, precise and measurable objectives in will not be possibleunless it has been anticipated terms of changing attitudes or behavior and a from the beginning(necessity of baselinedata that communicationsstrategy that specifies the target must be collectedbefore the project begins) and groups and the means of communicationto be the objectivesare measurable. used to reach them should be defined. The contentof the educationalmessages and the choice U If the program is set up with the aid of of the communicationpathways to be used will be technicalassistants, ensuring that their knowledge determinedseparately for each target population. will be transferred quickly to the organization responsiblefor the program in the long run is a Once the conmunicationstrategy has been essential. Changingbehavior is a slow process developed, the first two interventionsshould be that requiresmessage repetition,and thus a lot of increasingthe awarenessof politicalleaders whose time. Without this transfer of knowledge, the support is necessaryto the program's successand program will not become autonomous and training staff responsible for carrying out the sustainable. This transfer process is called the program at all levels. institutionalizationof the program.

* Programcontent, equipment, and strategies should always be specifically defined (no 4. Background predetermninedmodels) based on the target populations and with their participation. All equipmentor messagesshould be pretestedon the Nutrition education, like education in other target populations to ensure that they are health-related areas, was traditionally provided understandableand compatiblewith local cultural essentiallyby staffat remotehealth care centers as practices. part of therapeuticinterventions or during growth monitoring sessions (Wone et al., 1981, for example). Its first limitation,therefore, was that

124 Nutriion education

it affected only those people who came to the itspzactices, needs, media-related customs, and its health centers. It was providedby staff rarely sourcesfor healthcare and nutritioninformation, trained in communicationtechniques, whose as do conmmercialmarketing specialists,- from knowledgewas limitedand sometimes out-of-ate, whom they borrowedsome of these techniques. and who had difficultyadapting their theoretical Basedon this analysis,one mayidentify available adviceto the needsof each individual,as theydid solutionsand developa solidintervention strategy. not necessarilyknow anythingabout their life- styles and individualproblems. These staff members, frequentlyoverburdened by routine Somehave comparedmultimedia instruction medicalcare providedby the center, had very to instruction provided directly, "person to little timeto devoteto this type of initiative.The person.' In fact, interpersonalcomnunication is systemaccumulated a lot of problems: limited one of the ends of a continuumextending from coverage of the target population, limited face-to-facecommunication to the mass media, knowledgein the communicationsand nutrition including small- and large-group instruction. area, difficulty in adapting the educational Each method has different characteristicsand messagesto the target population,and no way to needsto be used accordingto thesecharacteristics reinforcethe messages. Althoughthe example as one of the elements of the strategy. describedhere is overstated,health care centers Interpersonalcommunication is certainlythe most whose nutritioneducation programs have these effectivemeans of changingbehavior, especially gaps are frequentlyfound. if it is accompaniedby concretedemonstrations and sessionswhere the motherscan practicewhat Since the 1980s,new trends have influenced they have learned(learning by doing). To be the way in which health and nutritioneducation really effective,however, (i) educationalcontent programsare implemented.These tendencies take mustbe relevantand culturally appropriate - that intoconsideration the contributionsof behavioral is, it must be based, as for all other means of science, the epidemiologyof the problemone communication,on an in-depthanalysis of the wantsto resolve,the methodologyused formarket siuation and the targetpopulation, and (ii) every studies, and the techniquesof audiovisualand health worker shouldbe responsiblefor only a mass media professionals (Mahonne, 1982; small group of people at risk (18 in Jamaica).' UNESCO,1987). Thelast of theseoften receive Extendingcoverage in this type of system is a great deal of attentionbecause they are the only difficult. meansto reacha broadaudience and to broadcast the same messagesrepeatedly at a lower cost. Given the advantagesand disadvantagesof Theyalso representa wayto reachpeople without oliLerenttypes of communication,the tendencyis requiringthat theytravel anywhere. The women to use severalmeans of communication1 including can listen to the radio while engagedin their interpersonalcommunication, in one strategy. normalactivities. This has two advantages:it may be a way (i) to increasecoverage, since no single method wiUl The use of thesemedia has madehealth care reach the entire target population and (ii) to comnmunicationsprofessionals aware of the need reinforce the main message through various to do an in-depthanalysis of the targetpopulation,

125 The health sector and nutriion interventionsin Africa

means, each with its own characteristics,for those the use of culturally acceptable materials was who are reachedby several meanssimultaneously. routinely tested: photographsof children before and after rehabilitation, proverbs, songs, agricultual demnonsirations,and radio programs. 5. Nutrition education experiments in Africa The problem of the energy density of food for the young child ca;ne up repeatedly, which led to research on the causes of and solutions to the a) IAnAL EXPERIMENls problem and the use of more appropriatenutrition advice. The results were convincing in several According to Schurch (1983), nutrition pilot tests in terms of increasingknowledge about education was characterized initially by a malnutrition, the nutritional value of food, and condescending approach, based on the feeding practices. In general, however, the results presumption that the mothers had no were often short-term; an intense but limited understanding of their problems and that it was exposure period cannot produce long-lasting necessary to repeat over and over the popular results. These programs did not emphasize that explanation for malnutrition at the time: an this new way of feedinga child should become the unbalanced protein intake in their child's food normal way to feed him, and insofaras the child's ration. Nutrition education sessions based on malnutrition was treated, the mother did not learning about three types of food groups always see the necessity of continuingwith what followed: the groups were those rich in protein, she consider-d a treatment rather than a normal energy, and vitamins and minerals, with the diet. emphasis clearly on the first, particularly for protein of animal ougin. The promulgation of Since this time, a comparison of education this protein myth did not produce the expected provided at a recuperationcenter within a hospital results, and it promoted a false idea in the minds with that of a center in a remote dispensary of numerous educators over the years. This (Hoorweg& McDowell, 1979) has indicated that fiasco, attributable to scientific ideas at the time, the results were better at a dispensarybecause the should make us reflect again today: how many sessions took place closer to the mothers' home failed nutrition education programs are the result and a social worker regularly visitedthe homes to of insufficientknowledge? reinforce the messages taught in the group sessions. Additional lessons drawn from these In view of these failures, a new educational experiments are that nutrkion education should: program was implemented,generally at nutrition (i) preciselyidentify the nature of the problem and rehabilitationcenters. It was based on closer staff possible solutionsin the local context; (ii) involve observation of the mothers' problems. The pilot the mothers and influential members of the projects implemented in Uganda (Schneideman, community, especially in the subsequent 1971; Stanfield, 1971) paved the way for close comnunication of the messages received; (iii) use monitoring of the family setting, collecting the traditionalchannels of communication;(iv) repeat views of members of the community, and the basic messages; (v) be integrated into all continually evaluating the effect of the advice health care- related educationalinitiatives; and (vi) given to the mothers. During these experiments, call on motivated individualscapable of initiatives

126 Nudritionedation to supervisethe wholeprogram (Stanfield, 1976). (which could occur at the time of weighing Beforethe 1980s,the joint effortsof healthcare sessions,for example)and groupmeeings. and socialscience professionals laid the basis of communitynutrition education in Africa along Finally.women living near the dispensaries linessimilar to thoseof the 'new education." who wererelatively well off werethe priorityof these programs in tenrs of their coverage; Apparently,these prudent lessonswere not coverageof the targetgroup in Ghana (8 percent) applied or were applied only partially. wasvery low, but it wasbetter in Lesotho(about Experimentsthat were preciselydescribed, or 30 percent). This seemsto indicatethat a good morerarely, analyzed, were compiled by Schurch level of coverage cannot be reached without and Wilquin(1982), Burgess(1982), and IsraTl resortingto communityhealth care workerswho and Tighe(1984). Theyseem amateur in termsof can helpto reinforcethe educationalmessages, a theirdesign or implementation,and neglect certn criticalfactor in makingthem ef.ective. stages that are critical to the success of a coMmuniCationprojea.4 Morocco (Gilmoreet al., 1980): A food supplementprogram for youngchildren was set b) PIOT PROSECIS: SOMESUCCESSES n MR up by CRS in 1975. In 1978, a nutritidon mAmnoivs educationcomponent, which had been ready for some time, was added to the program. The Ghana and Lesotho (Austinet al., 1981): assessment conducted in 1980 unexpectedly The CatholicRelief Services organized nutrition confirmedthat althoughthe distributionof food programsfor preschoolchildren that included, had been very effective, the addition of the among o,therfactors, food distributionassociated educational component really made this with growtn monitoringand nutritioneducation. interventionworthwhile, since the malnutrition Since these were multifactoralprograms, the rates (weLghtlageratio) of those who only specific role of each factor was difficult to receivedfood supplementsand those who had discern. An assessmentmade after six months' benefited from the supplements and the exposure to the program, however, showed a educationalcomponent were 33 percentand 11 significant correlation, after controlling for percent,respectively. There are goodreasons for different variables, between the mothers' the effectof educatonalmeasures in this context, knowledgelinked to education(understanding of and they are worth discussg here. First, the the growth curve, which could only have been implementationof the nutrition education acquired at the dispensary,for example)and- componetwas carefillyplanned: the Minist6re adequatenutritional status. Thisconclusion would des AffairesSociales (Mistry of SocialWelfare) tend to prove that the educationalfactor had an establishedan Institut de Nutrition (Nutrition effect; improvingthe mothers' undersing, Institute)in Marrakeshto train motivatedwomen however, was mostly linked to individualand to superviseoperations at the various regional personalizedcounseling rather tban to group levels; the trainers themselveshad undergonea instructionsessions. This is a very important three-yeartraining couTse at the InstitutNational observationbecause it indicatesthat a choicemust de Nutrition (NationalNutrition Institute) in be madebetween less frequentindividual contact Tunis. Five hundredwomen from rmalareas

127 T7hehealth sector and nutrition interventionsin Africa were then recruited as monitors at the village improved recipes; the goal of this alternativewas social developmentcenters. to allow broader coverage of the population foI less than the cost of the distribution of an These women had mastered their tasks industriallyproduced weaning food. Accordingto skillfully, which is not the case in mnanyother the assessment, the experiment had convinciag educationprograms. From the beginning,one of results in only onc district. the objectives was to orient the program toward the problems of mothers in poor areas by Beginningin 1969, the Yako district recruited proposing practical solutions to them. The 300 "pap monitors" in addition to customary availability of food supplements was a powerful medical personnel. The monitors attended motivating factor that led the women and their commnity educationsessions once a month with husbands to accept the difficult and long trips to a local education official; these monitors, the distribution centers, and it led to regular themselvesmothers with young children, returned attendance. The size of the rations distributedhad to their villages to do demonstrations for their in fact been calculated to provide a sufficient neighbors. They visited mothers with a child of quantity to young children in spite of the weaning age to give them personalized advice on inevitable dividing up that occurs among poor their child's diet and nutrition. In addition, they fmilies (three rations per family per month were responsible for screening malnourished providing a supplement of 500 calories and 16 children and monitoringchildren returning from grams of protein per daily ration, which exceeds rehabilitation centers. The monitors, compared all other known supplementationprograms). The with the communityhealth care workers, had no program contributed to changes in the mothers' other responsibilities: their only objective was to knowledge, attitudes, and weaning practices; a improve the diet of the young children in their synergistic effect also occurred in terms of the village. This approach, consisting of passing on evolution of the women's status. Without a a message that would be retransmitted in turn in doubt, coming to the center was to some extent an open discussion,proved to be effective because it emancipation for the women: they were more allowed the mothers to express themselvesand at open to the other services provided in the village the same time provided some supervision of the and gained some mastery over their environment. final version of the messagepassed on. Because The instructional initiatives did not stop at the of the success of this project, it was replicated in educational sessions but rather continued on an several other regionsin this country. ongoing basis infonmally,thanks to a remarable level of interaction with the mothers. This An assessmentwasconducted after eight years program was also limited, however, in its ability of operations in Yako (Zeitlin, 1981). It was to provide sufficient coverage of the target based on a small group of mothers (equivalent to population. that supervised by three outreach workers). Although several changes were made over the Burkina Faso (Zeitlin, 1981): Ways to course oi the project that niight have contributed improve the diet of the young child at home were to an improvementin nutritionalstatus, the project tested in one district in this country. The project seens to have had a real effect: overall, the Yako used illiterate conununity workers to distribute project probably decreased serious malnutrition

128 Nutritioneducatlon

(60 percentof the averageweightlage ratio) in dhe categoriesof motherswho shouldhave received project area to 0.2 percent, comparedwith an different types of instruction, based on the averagerate of 3 percentto 15 percentin the rest communicationmethod adaptedto each group. of the country. Ninety four percent of the Instructionwithin a healthcare frameworkis not mothers were able to recite one or more ableto providethis. The programnwas revised to nutritionally balanced pap recipes and includemore appropriate educational methods and implementedthe advice they had receivedon a tools, and a final summaryassessment showed regularbasis. Thisproject was compared with an improvedresults. Nevertheless,in a subsequent educationalsystem using mid-levelgovernment publication, Andrien (1986) wrote that this workers in another region: the latter had a initiativehad serious limitations;the causes of smallerimpact (78 percentof the motherscould malnutritionwere not analyzed,and the focuswas recite a recipe, but the recipes were often less limitedto dietaryrecommendations passed on in balanced)because the workerswere not fromthe conversationsthat gave the mothers themselves villages and used a more formal and less few opportunitiesto participate. The program participatorymethod of instruction. also depended on the dynamismof a highly motivatedleader, and withhis departure,it lost its In 1988,national leaders in chargeof health momentum. Finally, although consideredan educationin BurkdnaFaso tried to replicatethe exemplaryprogram for the country, it was not aforementionedproject in anotherprovince; as no reallyintegrated into an overallhealth care policy; documentationon the this project existed, its influenceremained very limited, and it was however,the governmentagency in charge of neverreplicated on a locallevel. nutritionwas not ableto repeatit. Instead,it had to "reinvent"a new method. This example c) LARGE-SCAEPROGRAMS CONDUCTED BYHEIHL? illustratesthe needto documenteach projectand CAE COMMNCATIONPROFESSIONALS not to relyon a fewmotivated individuals who are able to get pilot experimentsgoing successfully The previous pilot experimentsoriginated withoutintegrating them in a sustainedway into froma globalnutrition education movement that the operationsof the country'shealth care system. was attemptingto understandnot only what worked m this area but also why and how it CMted'Ivoire (Andrien,1983): Andrienhad worked (Gussow,1984). This resulted-in the the opportunityto evaluate a hospital-based definitionof a series of conccptsconcerning the traditional educational program provided to relationshipbetween knowledge,attitudes, and motherswhose children had been hospitalizedfor practices and the means to change them severe malnutrition. An initial formative fundamentaliythroughlarge-scalecampaigns using assessmentshowed that instructionabout various the mass media, of which the "Food is Life" food groupswas not well understoodby mostof programin 1975in Tanzania is an example(see the mothersand was not a necessarystep, given Hornik, 1985;p. 57). Designedto improvethe the program'sobjectives. This mighthave been population's knowledgein nutrition matters, due to a time lag in relation to the messages overcomefood taboos,and encourageimproved providedby other agencies. An analysisof the productionand storing methodsfor food, these target populationshowed that there were four effortsprobably reached an estimated1.5 million

129 7hehealth sector and nutriton interventionsin Africa people, thanks to broad political support and soon as the messagesceased, these new behaviors extensivepreparation; the effect, however, seems declined(Rasmuson et al., 1990). An assessment to have been short-term, which often characterizes of the nutritionalstatus of these childrenfollowing these "blitz" campaigns. the program, however. did not show any improvement;overall, the children lost weight, as Gambia (HealthCom., 1985): This USAID in a similar program in Nicaragua. This was program (messages concerning ORS and dietary probably caused by deteriorating economic practices during diarrhea episodes) was managed conditions during this period resulting from in direct consultationwith the Ministry of Health worseningdrought conditions. and Social Welfare. It lasted three years, beginning with a six-month-long preliminary Swaziland (Hornik et al., 1986): Radio program. It combined radio announcements broadcastswere also used in this case (five to six (600/year), printed materials (250,000), and times weekly, in addition to broadcasts or radio interpersonal commnunicationbetween community messagesto raise awarenessseveral times a day), and village workers. The messageswere changed distribution of pamphlets (260,000) and posters in accordance with the seasons and reinforcedby (7,500), and interpersonal conununication at a radio-sponsored lottery. A majority of the dispensariesduring consultationsfor diarrhea or target population had access to the radio or a through community agents who used the yellow community worker. Although one-third of the curtain method at the doorways to their homes to populationof womenwas illiterate, it was possible indicatetheir qualifications. In terms of nutrition, to provide explanations of the printed mattrials the messages emphasized the importance of (coded with colors that the population could continuing to feed the child during diarrhea recognize) through the two channels mentioned episodes and of giving him special foods above. In sum, three-quartersof the women had afterwards. a pamphlet, half of them heard the messageson the radio and had also heard of the lottery, and In all, 85 percent of mothers questionedwere the majorityof them had the opportunityto talk to exposed to at least one aspect of this campaign. the community workers. As such, the project One of every six people was in touch with a effectively used several information sources to communityworker, but 60 percent heard the radio communicatethe messages to the mothers. Th;is messages; even if personal contact seems more media combination was clearly effective in effective, broadcasting the messages by radio is reinforcing the messages: in terms of increasing essentialto reach the target population,at least in knowledge, it was a success for all issues related a country like Swaziland where a significant to rehydration. Follow-up demonstrated that a number of people have a radio, listen to radio learning-forgetting-relearning cycle occurs, programs,and where radio is a credibie source of making periodic repetition of the messages health care and nutrition information. necessary for about two years to make sure they are retained. Increased knowledge was Althoughsolid data on actualpractices and the accompaniedby changes in behavior, includingin ultimateeffect of the project on nutritionalstatus feeding practices during diarrhea: children were are not available, ORS practices seem to have put on a starvation diet less frequently; but as improved nationwide as a result of the radio

130 Nurition education campaign.The dispensariesdo not seemto have whether the mothershad run into any specific had an effectbecause they limited their effortsto problemss in the process of making the distributing packets, when available, without recommendedchanges. This crucial stage showed providingany explanation.As far as dietduring that the mothers are ready to change their diarrhealepisodes is concerned,convincing more practicesif whatis proposedis relevantand not mothers to feed their childrenat this time was too difficult to implement. In addition,group difficult,but a majorityrecognized the importance discussionsallowed the participantsto identifythe of feeding themnwith specific foods once the advantagesand disadvantagesof each of the diarrhealepisode was over. Thisdoes not seem changes. to be a resultof contradictorymessages but rather of the fact that the messagesconcerning feeding The next stage involved integrating the duringdiarrhea episodes were neither assimilated greatest number of agencies possible into nor understood. discussions concerning the project's implementation;awareness and participationfrom Once again, the relative success of this many areas are necessaryto institutionalizeand campaignwas the result of the enthusiasmof sustain a project. Among the merits of this severalwell-trained staff members. The idea of projectare the following: (i) it conducteda far- preventingdiarrheal illness with ORS had already reachingand irrdepthanalysis of the targetpublic, been propagatedin a variety of ways over the whichled to a good understandingof how and by courseof severalyears and wasnot an unknown, whom mothers are influenced at any given unlikethe initial situationin Gambia. But this moment, and (ii) it anticipated appropriate campaignwas managedoutside of the normal messagesfor each group. The projectresulted in institutionalframework, and whether it may be several recommendations aimed at modifying continuedand institutionalizedto produce longer- legislationrelated to food aid, promotingincome- term results is not clear. producing initiatives, involving street vendors, etc., according to the idea advanced by Berg Ghana (MOH,1989): A nationwideprogram (1976) and others but not often retained by made a careful inidal exanination of living educators to ;ate. The idea is to try to change the conditionsfor the youngchild in each regionto basicfactors that limit irnproving nutritional status underscorethe differencesand similaritiesbetween (Gussow, 1984). Unfortunately,an assessmentof the radically different environments in the the program's impact has not been conducted. country's northern and southern regions. While maintainingthe particularitiesof each region, the Swazilatti: A programnwas implemented projectemphasized the similarities,given that the under the aegis of the Health and Agriculture problem is basicallythe same throughoutthe Ministries,with technicalassistance funded by country: at six months, growth in terms of height UNICEFand USAID(National Nutrition Council, becomesprogressively delayed. A quantitative Swaziland,1988; Aphane, 1989). It developed and qualitative analysis was conductedfor each along the lines of the Ghana progrmn; once age group to develop coneru.e suggestionsfor the again, painstakingpreliminary studies slhowedthat mothers. Trial interventionswere conducted in mothers' feeding practices could be improved about 100 homes for three to four days to see from the birth of their children (rejection of

131 The heaUhsctor and nutritioninterventions in Africa colostrum, providing nmrsinginfants with water, differentprogram strategies, implementation,anr use of industrially produced milk) to the end of assessment. the weaning period (inadequate portions given to the child, food that has been too diluted, An assessment was conducted one year after insufficientnumber of meals, etc.). The solutions the program was initiated. It showed a significant were carefully discussed at group meetings and extension of coverage in the villages by the tested to verify their acceptability. Aside from community workers (participation in the growth these home-based initiatives, recommnendations monitoring sessions increased from 13 percent to were made to officials concerninghow to improve 44 percent in less than six months in some the situation nationwide: limiting importationand villages), and a modest improvement in the promotion of industrially produced milk and motlhers'knowledge (directly related to their level bottles, enviromnental clean-up measures, more of exposure to the program), but few concrete appropriate staff training, etc. As in the case of changes in their daily feeding practices for young Ghana, no assessment of the program's progress children. This project had a limitedbudget and a and impact was conducted. short initial phase. Staff training in communication techniques and supervision of Cameroon: This project was initiated by them was clearly inadequate, while the proposed Manoff International in conjunction with the techniques were rather sophisticated for staff of American Center for Education, CARE, and the this nature: different strategies depending on the government of Cameroon in two regions in the age groups, etc. Finally, inceiporating an country's most northern province. In fact, the ongoing evaluation component into the program project wassintegrated into a primary health care appeared necessary to adjust the activities better project that had been operated off and on since during the course of their implementation. In 1984 (Manoff, 1987 and 1989; Hollis et al., sum, although this project was based on the same 1989). This project is important because it used principles, it did not benefit from as careful the ability of public awareness measuresto change preparation and implementationas other projects. the habits and practices of mothers in areas where Only a new assessment can determine this access to the media was rare, and the surrounding program's degree of success in terms of its impact environment was rather barren, isolated, and and sustainability. without infrastructure. The populationwas almost entirely illiterate, and only 30 percent of Congo, 1980-1990: The CARE education households had a radio. The project was based project, known as NUTED, was initiated in 1980. essentially, therefore, on the efforts of community It represents a projuct conducted nationwide that health care workers already working on the CARE has lasted for an exceptionallylong time. It was project who visited the villages two or more times designed to train staff in various ministries in per month. During the first visit, the health nutrition and hygiene education, import or work!!r organized a weighing session with produce educational materials locally to support individual counseling provided to the mothers; the national initiatives, and create an educational second visit was dedicated to group educational resource center in Brazzaville. An iiiitial study sessions. The entire program was implementedin was conducted on nutrition problems in Congo four phases: basic analysis, formulation of with USAID (Huntington, 1986); it showed that

132 Nurntion education despitea prevalenceof moderatemalnutrition in country. The situationbegan to worsenin 1988, young children,food appearedto be sufficiently in part because of unfavorable economic available. The age at which food supplements conditionsand poor management,which led to were introducedvaried - eithertoo early or too several changes in foreign partners. A new late - and an increasinglymarked tendency to use assessment conducted by UNICEF in 1990 imported food supplementswas noted. The recommendedthat the projectbe discontinuedand situationseemed to be relatedto the family'slevel integrated into a nutrition project under the of instruction. An important program was tutelageof the Health Ministry (Ministtre de la implemented,therefo!e, that includedtraining for Sant6), whose principal educationalcomponent 300 field workers and their supervisors; had been replaced by sectoral educational development of educational and food initiatives. In particular,ORS messages,which demonstration materials, radio programs, a had beenpronulgated by anotheragenry without manual, posters, and academic programs for coordinationwith the NUTEDproject, could now school children in rural areas; and scales for be integratedinto it, thus avoidingredundancy maternal and infanthealth care centers. All of and contrcdictions(Dillon, 1990). these materials were tested beforehand. According to the first assessment team, the Becauseof a lackof appropriatebaseline data, program seemed to be comprehensiveand well it wasnot possibleto assessthe project's impact. planned, and its implementationfollowed the In operationalterms, one may concludethat it predeterminedschedule for the first few years. never reachedthe projectedperformance level in spite of its real successat popularizingimproved Not long after the project began, however, weaningtechniques between 1985 and 1988. As problemsappeared: stafftraining was considered the result of not maintaining real skills in too short, the objectivewas too narrow (basedon commnication techniques, the educational the introductionof comolementaryfood at the component linked to the media, which was right age), administrativeproblems received too initiallyimportant, was gradually phased out. The little attention,and the needto encouragepolitical implementationof a growth monitoringsystem leadershad largelybeen neglected. In addition, producedmixed results: 38 percent of mothers the foundation of ihe program was a base had a growth chart in 1987, but analysisshows assessment that had focused on malnutrition that the systemwas operatingwithout any real problemswithout analyzing their socioeconomic supervision and did not lead to appropriate and culturalaspects; changing mothers' behavior individualcounseling for the mothers. Thefactors without knowing what determined it seemed that contributedto these poor results were the difficult. Finally,the mediastrategy had not been verticality of the project, because of which clearlydefined. In viewof thesedeterminiations, essentialcolliboration from the Health Ministry adjustmentswere made in the programn,and its wasnot alwaysavailable; the absenceof precisely scopewas broadened: it promotedbreastfeeding, defined objectives(the only clearly formulated oral rehydrationtherapy. growth monitoring, and objective, "improvingthe nutritionalstatus of improved weaning conditions. The program Congo's young people," fell under a began to show some real success beginningin multidisciplinaryinitiative that was never really 1985, and it became popular throughout the implemented),and the lack of precise assessment

133 The health sector and uaridioninrve Cndons tn Afica criteria; the lack of supervision in a project that mnrsesto field nursing staff and paramedics. The was too centralized; and increasinglydiscouraged results show that in the PMI and hospitals, staff personnel. suggest that radio, television, or video be used for mntritioneducation. In fact, they do not want to The Sabel: A rapid overview of the work of take personal responsibilityfor a task they do not the African Nutrition EducationNetwork (R6seau believe will enhance their medical work, that takes d'education nutritionnelle Aficain or RENA) them awayfrom their consultations,and for which (Lejeune and Andrien, 1986) confirns that the they admit they are hardly qualified. Midwives French-speaking countries are not very advanced and nurses are rather discouraged in view of the in the area of education: projects are incomplete, few visible changes after so many "chats." use traditional education techniques, and are not Materials are the least underused at the remote adequately supervised or assessed. A recent centers; evaluation of the difficulties encountered in the education area by diarrheal disease prevention * Training staff in interpersonal programs in two countries in the Sabel provides a communicationand the use of the materials is good illustration of these problems (Pritech, inadequate: staff who are asked to take on these 1990). activides (midwives, nurse-aides, social workers, and less often, nurses) have generally little or no Four main problems have been identified: training in this area; this category of personnel are motivated and available, and they would like the * Lack of message coordination among the materials to be more diverse and better explained various projects, as in many other places, causes so that they feel more prepared; and confusion among the target public (Andrien and Lamotte, 1986): inasuburbofDakarinSenegal, * The media used are not always the messages broadcast by a university-based appropriate: Andrien demonstratedthat in Pikine, instituteto ensure educationalcoverage for several a suburb of Dakar (Andrien & Lamotte, 1986), neighborhoodswere different from those provided only 28 percent of residences had a radio and only by health care dispensaries in the same half of the targeted families did. One of the neighborhoods; characteristics of target populations affected by malnutrition is that they do not have access to * Undert-e of educational materials: as information dtrough the media. The attitude of preliminary assessments showedthat the principal doctors in Niger and Mali, given that the radio, source of information for more than 65 percent of television, or video must be used to educate the mothers is health care workers, diarrhea people, is tantamount to an abdication of prevention programs produced important responsibility(Pritech, 1990) on the part of health educationalmaterials; supervisoryvisits, however, care workers. showed that these materials were not used sufficiently. A study of the reasons for this underuse was ronducted in several treatnent centers that operated with a variety of health care personnel, ranging from hospital doctors and

134 Nutritioneducation

* Heath and nutrition education projects 6. Lessons learned from projectsimplemented should be designed by communication in Africa to date professionals based on a rigorous methodology. Specialists in the social sciences and communicationsmust work with nutritionists to This section contains a summary of the assist them in the design and pretesting of lessons leaamedfrom the projects describedabove. appropriate messages.' Overall, each of the projects seemed to have gaps. The following lessons emphasize that, for a RAISING THE AWARENESS OF project to be successful, each stage must be POLITICAL LEADERS CONCERNED: followedand certain key criteria must be satisfied. Making political leaders aware of the importance of the program so that they do not create any PROJECT DESIGN: Several factors must obstacles to its implementationand they support be considered when a communicationproject is it. This stage is often overlooked because leaders designed: are not directly involved in the project's execution. Their support is still needed. m Projects must be based on a prior analysis of the situation, providing answers to the TRAINING: As many projects have shown, following questions in the case of mntrition problems resulting from the project implementors' education: which behaviors affect the nutritional lack of training often arise. Training should be status of the target population; what are the improved for twotypes of staff workers: (i) those determint of these behaviors; what is the nature responsible for inerpersonal communication and of the resistance to changes in these behaviors; (ii) those in charge of the design, planning, and who are the varous target groups; what are the implementation of communication initiatives. most effective channels of communicationto reach These gaps became major problems in programs each group; how should messages be fomnulated in Congo (1980-1990), Cameron (1988), and for each chamnel of communication; does the freuently in the Sahel. planned initiative complementnational health care programs; how does one conciliate the promotion * Workers responsible for interpersonal of new behaviors with respect for local cultural communcation. The importance of health and practices? Any message that is read or heard is nutrition education should be reemphasized by not magically assimilated, leading to imediate defining precisely its content and status in all changes in behavior. The key is 9i)a prior health care taining programs and by maidng investigation of the problem at hand that helps to medical authorities aware tbat, even when they do identify acceptable solutions depending on the not participate, they should at least facilitate use target population(s); (ii) designing' messages that of the materials and ensure supervision. Basic must be pretested (in some cases, several times) trainig provided these staff should include through sample testing of this population; and (iii) instruction in communications and community continuous adjustment of the messages based on assessment. But until the basic training programs the public's reaction. are revised, and staff trained in them from the beginning is available, currently employed health

135 The health sector and nuaritioninterventions in Africa care workers should be retrained in interpersonal the characteristics of the target public (Is it communication and the use of IEC materials. literate? What are its credible sources of Cormmunityworkers should also be retrained in infonnation? What are its listening habits? What community organizing and interpersonal channels have the greatest likelihood of reaching communication. it?) and. available resources. A large-scale program in Zaire (Barnes-Kalunda, 1986) is an * Staff in charge of communication example of a bad choice: it was based on the use initiatives. Training for these staff members is of television, which was not widely accessible at often inadequate and fragmented. Many of them the time. Two other examples mentioned above need additional training, especially in are those of Senegal and Niger, where use of communications-relatedplanning. these media did not respond to the public's needs but rather represented an abdication of In addition, this retraining should be responsibility on the part of health care workers. accompanied by "constructive and positive' In fact, several media should be combined to supervision by field workers. This type of obtain a cumulative effect: the same message supervision plays a very important role because it broadcast by several sources will be reinforced. serves as a quality control measure for the initiatives, provides continuous training for staff, NEED TO TARGET THE MESSAGES: and provides them with moral support (Indonesia, Messages should be designed to ensure that the 1985). target public feels involved and identifies with the messages. The program in Burkina Faso also COM MUNITYACfTON: Extension demonstratedthat mothers were attracted by clear programs to villages and neighborhoods are messages that directly addressed their immediate needed; this ongoingdevelopment could be carried problems and whose expectedbenefit was visible: on by community health care workers or social a response to a problem that is widely felt is the workers, whoever is more motivated, provided key to motivation (McSweeney and Friedman, that resources are obtained for training and 1980). How far nutrition education messages are educational materials. The distribution and broadcast and their impact must also be monitored maintenanceof these materials should be managed to avoid any hannful effects. Pagezy and in an organized fashion. These workers also seem Subervie (1990) relate how messagesbroadcast by to be more effective when they are known by the a development program in a region with a low community rather than when they are outsiders level of agricultural productivity and monetary (see Burkina Faso, 1981). economic activity in the northem part of Zaire had a catastrophic effect in an adjacent region NEED TO EVALUATE THE AVAILABLE with a different environment.' MEANS OF COMMUNICATION: A close study should be done before any means of FOLLOW-UP AND ASSESSMENT: Fmm communication is chosen; this choice should not the moment the project is designed, follow-up be made under pressure from organizations measures and plans for ongoing and specializing in mass media or governments eager comprehensiveassessments should be developed. for new gadgets. The choice should be guided by As mentioned several times above, follow-up is

136 Nuritdon education very important in determining the necessity of INSOITONALIZATION: One of the modifying a project that is already underway. reasons a number of initially successful NGO, Ongoing assessment should document how a university, or foreign assistance programs have project was implemented, what worked, and what been unsuccessfulin the long run is that they have did not work. Such information will allow the been designed from the outset outside of official project to be replicated and improved at the same health care institutions, based on a very time (see the problem with a lack of "medicalized"treatment method that relies mostly documentation in the Burldna Faso project, on sophisticated health care structures. An 1980-1990). As far as the comprehensive initiative aimed at modifying behavior in the assessment is concerned, in imny cases baseline nutrition and health area cannot be designed studies must be conducted before the project without coordinationamong all the conmmnication begins, allowing for pre- and post- project strucure concerned (Andrien, 1986). Education comparisons. Indicators should be identified that programs should be comprehensive in scope and allow one to assess whether the project's should integrate all health care concepts related to objectives have been attained. Without an fanilies that are useful or immediately usable to assessment, as was seen in a large number of avoid any bias. The Ouando program in Benin projects descrbed above, determining whether a (Fakambi, 1990) emphasized nutrition in young project has had an impact is impossible, and children and because of this, the mothers were drawing useful lessons firom these projects is fairly knowledgeable in this area, although they difficult. clearly did not know enough about maternal nutrition, an important aspect in these regions. ROLE OF CAMPAIGNS: The use of Because too many programs are preoccupied with intensivecampaigns is common. These campaigns efficiency, they focus on specific issues; if help to raise the awareness of, inform, sensitize, initiatives are not pursued simltaneously to and mobilize the population (Tanzania, 1975; addrss the various bottlenecks, the effect on Gambia, 1985), but they will have only a short- changing pracces remains slight. In Congo, the term effect if they are not followedup by ongoing UNICEF assessment identified the lack of IEC initiatives that reinforce the messages integration of this type of program into other launched at the time of the campaign. programs run by the Ministry of Health as one of the reasons for its failure. A program's durability COORDINATION AMONG PROJECTS: also should not depend exclusively on a few A frequently encountered problem is the motivated individuals who will run the program multiplicity of similar educational messages or well but who will become "indispensable" to its initiatives that are not coordinated beforehand. operations. This was a problem in Cote d'Ivoire Several examples of this problem were mentioned (1980). Swaziland (1984-85), and Congo above, in particular in Congo, C8te d'Ivoire, and (1980-90). Senegal. This lack of cooperation is always inefficient, but it may also be disastrous if the messages are contradictory.

137 The health secor and nuntriioninterventions in Africa

mother (seven hours per week) but rather the 7. Costs number of cases of malnutritionavoided, the cost falls by 2.5 to 19 times, depending on the program components considered (this is a Adding education to the food distribution multifactoral program, including food program in Morocco (Gilmore, 1980) increased distribution). Improving cost effectiveness is the program's costs by $1-$3, making the total directly related to better coverage of the target cost $34.47 per person annually, compared with populationand substitutingless qualified for better $10 in Kenya and $32 in Colombia for programs qualified staff (Schflrch, 1983; p. 177). during the same time period. The contribution requested from, and accepted by, the mothers, One of the reasons cited for using the mass given the substantial amount of food supplied, media rather than personal instruction is the high allowed the centers to be financiallyindependent. cost of the latter for similar coverage. Programs Asking participants to contribute to an education like those at Yako, however, have shown to what program that does not provide them with a extent the populations were able to pay for some material return, however, appears problematic. of the costs, so that the government did not have to cover too much of the costs relative to the The cost of the educationalcomponent for the overall budget. Ghana and Lesotho programs was estimated at $2.80 and $2.48, respectively, per person As for more modem national programs, initial annually (Austin et al., 1981). research is estirnatedto cost between $30,000 and $50,000 (for a small country or region); the cost The cost for the government of the Yako of reproducing the materials, about $150,000; the program (Burkina Faso) was $0.53 per mother cost of project execution, between $40,000 and annually. This program was not costly because it $75,000, if materials used have already been was based essentially on community health care tested and are not excessive in quantity; and the workers. The cost of a similar program using cost of monitoring and evaluation, $50,000 per mid-level staff was higher with fewer results round. In most cases, using the media costs (U.S. $2.24). The cost of programs implemented nothing, and training and supervision of staff are at health care centers is clearly the lowest in included in integrated costs of normal operations absolute terms; but as the coverage provided is for health care services. Once the program has very low. the comparison of individuals reached been set up, which is the most expensive phase, is not valid. In the case of Yako, if one does not maintaining it should not exceed reasonable consider simply providing instruction to the budgetary limits for most countries (Griffiths, 1989).

138 III. CONCLUSIONS

This overviewof projectassessments suggests * Decisionmakers at every level shouldbe severalpriority recommendations for Africa: madeaware of nutritionproblems. The problem of breastfeedingis typical: numerous African a Health care services should use a decision makers ignore us when we tell them conmmunityapproach as early as possible. This about this problem because, as they say, all implies training staff in this area, defining African womenbreastfeed; this is true, but the community participation, and decentralizing womenoften breastfeedincorrectly and do not decision making, and hence, budgets, to the weanthe childproperly, as is demonstratedin the districtlevel. Withouta communityapproach, the chapterdealing with this topic. The health sector coverage,inpact, and viabilityof programswill ought to serve as an advocatefor nutrition in continueto be inadequate. relationto other sectors. This couldincrease the health sector's coverage and encourage it to * Nutritiontraining for healthcare workers undertakepreventive interventions. For example, and those in other sectorsshould be rethought, the onlyway an effectivesalt iodizationprogram more adapted to a communityapproach, less will be implementedis in collaborationwith the academicand morepractical, and sinmpleroverall. Ministryof Commerce. This is the case for basic taining as well as retraimingwhen workers are alreadyemployed. * Effectivedecision-linked monitoring and assessmentsystems should be put into place. * Health care servicesmust becomemore Many projectsare too smallor poorly assessed. efficientby concentratingon what they can do Ths shouldbe initiatedat the time the projectis best. This documenthas attemptedto showhow designed, when the objectivesare determined. these choicescan be made. Someinitiadves can Baselinedata shouldbe collectedto measurethe be undertakeneasily withouta lot of resources project's impactand cost. Too many midwives (salt iodization,prevention of anemia, among and nurses spend hours every week filling out others). No furitherdelay is justified in putting formsfor reportsthat are scarcelyread ratherthan such programs into place: the problemshave assistingin decisionmaking. been identified,solutions exist, and only the will to implementthem is lacking.

139 BIBLIOGRAPHY

Benefits already provided by the health sector

Dosseter, J.F.B., and H.C. Whittle. 1975. "Protein losing enteropathy and malabsorption in acute measles enteritis." British Medical Journal, 2:592-93.

Laditan, A.A.O., and P.J. Reeds. 1976. 'A study of the age of onset, diet and the importance of infection in the pattem of severe protein-energy malnutrition in Tbadan, Nigeria." British Journa of Nutnrion, 36:411-19.

Molla, A.M., S.A. Sarker, M. Khatoon, and M.M. Rahaman. 1983. 'Effects of acute diarhoea on absorption of macronutrients during disease and after recovery." in Clen, L.C., and N.S. Scrimshaw, eds 2T-rrhoea and Malnutrion: Interacions, Mechanisms and Interventions. New York: Plenum Press, pp. 143-54.

Morley, D. 1969. "Severe Measles in the Tropics." British Medical Journal, 1:297-300.

Tomkins, Andrew. 1981. "Tropical malabsorption: recent concepts in pathogenesis and nutritional significance." Clinical Science, 60:131-137.

Tomkins, Andrew, and Fiona Watson. 1989. Malnutrition and Infection, A Review. ACCISCN Nutrition Policy Discussion Paper No. 5. London. October.

140 Prevnionof vitaminA defidey

Prevention of vitamin A deficiency

Am6d6e-Manesme0, De Maeyer E. 1989. Le deftcit en vtamine A: Stratigiar diagnostiqueset hrapetutriques.INSERM - ORSTOM. Paris

BarclayAJG, Foster A, SommerA. 1987. "VitaminA supplemnentsand mortalityrelated to measles: a randomisedclinical trial." BdtishMedical Journal, 294:294-296.

BatesCJ. 1983. "VitaminA in pregnancyand lactation."Proc Nutr Soc 42:65-79.

CohenN. "VitaminA projectevaluation, March 1987- yuly1989. Helen Keller International/ Niger.

Costello(de L.) AM. 1991. "VitamineA suppl&mentation."Lancer, 338:568.

CoutsoudisA, BroughtonM, CoovadiaHM. 1991. "VitaminA supplementationreduces measles morbidity in youngAfrican children:a randomized,placebo-controlled, double-blind trial.' AmericanJournal of Clinical Nutition, 54:890-895.

Deiayer EM. 1986. "The WHO programme of preventionand control of Vitamin A deficiency, xerophtalmiaand nutritionalblindness.' Nurition and Health,4:105-112.

AmericanNutrition Review. 1991. 'Vitamin A administrationreduces mortality and morbidityfrom severe measles in populationsnonendemic for hypovitaminosis(Editorial).' Amecan NuMtion Review, 49:89-91.

Lancet. 1987. "VitaminA for measles(Editorial)." Lancet 1:1067-1068.

Lancet. 1990. 'Vitamin A and malnutrition/infecdoncomplex in developing countries.' Lance, 336:1349-1351.

FAO. 1980.Food balance sheets, 1975-1977 average. Rome.

FAOIOMS.1989. Besoinsen vitanine A, fer, addefolique et vitanine B12. Rapportd'une consltation conjointeFAO/OMS d'experts. Collection: Alimentation et mntritionN' 23. Rome.

FAOIOMS. 1992. Preventng Specfic MicronurriemDeficiencies. Theme Paper No. 6. Inteational Conferenceon Nutrition.

FAO. 1991. Le Programme WvitamineA- de la FAO. Troisitme rapport succinctde situation;les cinq premi&resann6es 1986-1990. Rome.

FosterA, Sonner A. 1986.'Childhood blindness from comneal ulceration in Africa:causes, prevention, and treatment."Bulletin WHO, 64:619-623.

141 Bibliography

Foster A, Kavishe F, Sonimer A, Taylor HR. 1986. 'A simple surveillance system for xerophtalmia and childhood corneal ulceration." BuUetinW1O. 64: 725-728.

Galan P, SamnbaC, Luzeau R. Amnd&-Manesme0. 1990. "Vitamin A deficiency in preschool Congolese chilaren during malarial attacks. Part 2: impact of parasitic disease on Vitamin A status." Int J Vt- Nutr Res, 60(3):224-228.

Helen Keller IternationallNiger. 1989. 'Vitamin A project." Operationalresearch data analysis report. Final Draft. Part two. Feb.

Helen Keller International/UNICEFIMinist&rede la SantEet de l'Action Sociale. 1989. Etude de prtaence de la carence en vizamineA chez Ies enfaws de 0-10 ans et twolAdvit des t4ze de Bitot chez ln efans de 6-10 ans. Ouagadougou,Burkina Faso. juin-juillet.

Horton, S. 1992. Unit Costs, Cost-Effectveness, and Fnwnacingof NutritionInterventions. PHR1Worldng Paper No. 952. August. The World Bank.

Hussey G.D., Klein M. A. 1990. "Randomized controlled trial of Vitamin A in children with severe measles." New England Journal of , 323:160-164.

Inua M, Duggan MB, West CE, et al. 1983. "Post measles comnealulceration in children, northern Nigeria: the role of Vitamin A, malnutritionand measles." Ann Trop Pediatr, 3:181-191.

Maindani M.. Ross D. 1989. VitaminA supplmentaon and child survival: magic budletorfalse hope? A review and selected, annotatedBIBLIOGRAPHIE. EPC PublicationNo 19, London Sch Hyg Trop Med.

Markowitz LE, Nzilambi N, Driskell WJ, et al. 1989. "Vitamin A levels and mortality among bospitalized measles patients, Kinshasa, Zaire.' J Trop Pedior, 35:109-112.

McLaren D.S. 1991. "VitaminA supplementationand mortality." Lane 338:1208.

Ministere de la Sant6IAEDIHKI.1992. Republiquedu Ni-ger:Evaluadon duprojetde conmuunicatonsociale en VitamineA. D6cenbre.

Muhilal, Murdiana A., Azis 1. et al. 1988. 'Vitanin A-fortified monosodium glutamate and Vitamin A status: a controlled field trial." Am J ain Nutr, 48:1271-76.

Pepping F. KavisheFP, Hackenitz EA, West CE. 1988. 'Prevalence of xerophtamniain relation to nutrition and general health in preschool-agechildren in three regions of Tanzania." Acta Paediatr Scarnd,77: 895-906.

Rahmathulla L., Underwood B.A., Thulasiraj R-D. et al. 1990. "Reduced mortality among children in southern India receiving a small weekly dose of Vitamin A." N EngI J Med, 323:929-35.

142 Preventionof iraminA deficieny RahmathullaL., UnderwoodB.A., ThulasirajR.D., MiltonR.C. 1991."Diarrhea, respiratory infections and growth are not affectedby a weeklylow-dose vitamin A supplement:a maskedcontrolled field trial in childrenin southernIndia." AmericanJournal of ClinicalNutrition, 54: 468477.

Save The Children, USA. 1990. Agencyfor internationaldevelopment fndly and communityprogram, VitaminA defciencyprogram support, Mbalachwrda and Mhota districts,Malawi. 9130/87- 9/29/90. Final evaluationreport.

Sight and Life. 1989. "Vitamin A workshopin Dakar, Senegal."Sight and Life Newsletter,4(2):6-11.

Sight and Life. 1990. "Regionalworkshop on VitaminA interventionsand child survival,Lusaka, Zambia. Sight and Life Newsletter,5(1):2-10.

Sight and Life. 1990.Annual Report. Basle,Switzerland.

SommerA, TarwotjoI, HussainiG, SusantoD. 1983. "Increasedmortality in mild VitaminA deficiency." Lancet, ii:585-588.

SommerA, Tarwoto I, Katz J. 1987. "Increasedrisk of xerophtalmiafollowing diarrhea and respiratory disease."Am J ClinNutr, 45:977-980.

SommerA., Tarwojo I., DjunaediE. et al. 1986. "Impactof VitaminA supplementationon childhood mortality:a randomisedcontrolled community trial.' Lancet, 1:1169-73.

SteirbifllerPG. 1983. "Nutritionalblindness in Africa."Soc Sc Med, 17(22):1715-1721.

Stoltzfus,Rebecca, and Jean-PierreHabicht. 1993. "Measuringthe Effects of VitaminA Supplementation, Letter to the Editor."American Journal of PublicHealth, 83(2).

Storms D. and Quinley 1. 1988. Afield guidefor addng WraminA interventionsto PVO child survival projects. Recommendationsfor Child Survivalproject managers. Report of a specialPVO Child Survival Task Force on VitaminA. Baltimore

Tielsch TM, West KP. 1990. "Costand efficiencyconsideations in communitybased trials of VitaminA in developingcountries." Stat Med 9:3543.

UnderwoodB. 1990. "VitaminA prophylaxisprograms in developingcountries: past experiencesand future prospects."Nutr Rev, 48:265-274.

VijayaraghavanK., RadhaiahG., PrakasamB.S., SannaK.V.R., ReddyV. 1990. "Effectof massive dose vitaminA on morbidityand mortalityin Indianchildren. " Lancet, 336:1342-45.

Villard L., Bates C.J. 1987. "Effectof vitanin A supplementationon plasma and breast milk vitamin A levels in poorly nourished Gambian women." Human Nutr - Clin Nutr, 41C:47-58.

143 Bibliography

VitaminA NewsNotes, HKL.1991. Printemps 4-5.

West KP, Sommer A. 1987. Delivery of oral doses of vitamin A to prevent vitanmn A defiiency and nutrtional blindness. A state-of-the-art review. ACC/SCN State-Of-The-ArtSeries, Nutrition Policy Di&wussionPaper No.2. FAO, Roine.

West KP. 1991."Dietry vitamin-Adeficiency: effects on growth,infection, and mortality.'Food Nurr Buil, 13:119-131.

West K.P., PeppingF., TemalilwaC.R. 1988. The compositionoffoods commonlyeaten in EastAfrica. CTA; ECSA; WageningenAgricultral University.Wageningen.

WestK.P., PokhrelR.P., KatzJ., et al. 1991."Efficacy of vitaminA in reducingpreschool in Nepal."Lancet, 338:67-71.

WHOIUNICEF/IVACGTask Force. 1988.Vitamin A supplements.A guideto theiruse in the treatmentand preventon of vitaminA deficerncyand xerophtaimia. WHO, Genbve.

144 Prewvntionof Iron defiency

Prevention of iron deficiency

ACCISCN.Report of joint ACCISCNworkshop on iron deficiencycontrol; Trinity College, Dublin, 648 June 1990.Genbve, Nations Unies, 1990.

Ajayi OA. Iron stores in pregantnigerians and their infantsat term. Eur I Clin Nutt 1988;42:23-28.

B6n6ficeE, Chevassus-AgnesS, MaireB, NdiayeAM. Enquttessur l'htat nutritionnelen zonetropicale s&he (Sahel 1976-1979):m6thodologie et r6sultats.In: Etat nutritionnelde la populationrurale du Sahel: rapportd'un groupede travail,Paris (France)28-29 avril 1980.Ottawa, Ont., CRDI, 1981; 160f:37-56.

BrunengoJF, Morier F, Tnkalri G. Toure IA, CUnacA.. Iron de i pregancy and breast-feedingin the Republicof Niger.Trans Roy SocTr Med Hf 188;P40-6•T:

Bundy,D.A.P. 1991. "Solutions:Strategies for worn control"in Proceedingsof a workshoppresented by the Imperial College of Science, Technologyand Medicineon "Parasiteworms and child development",Imperial College, London, March (p. 18).

ChippauxJP, SchneiderD, AploganA, DyckIL, Berger J. Effets de la supplementationen fer sur l'infectionpalustre. Bull Soc PathExot 1991;84:54-62.

DeMaeyerEM, Adiels-TegmanM. The prevalenceof anemiain the world. Wld Hlth Stats Q 1985; 38:302-316.

DeMaeyerEM, DallmanP, Gurnmj'JM, HallbergL, SoodSK, SrikantiaSG. Preventingand controlling iron deficiencyanaemia through pmary heldth care. A gude for health adminstaors and programne managers.Geneva, WHO, 1989.

ElegbeI, OjofeitimiEO, ElegbeIA. Traditionaltreatment of pregnancyanaemia. Trbpical Doctor 1984; 14:175-177.

FAO/OMS.Besoins en vitamineA, fer, acide folique et vitamineB12. Rapport d'unc consultation conjointeFAO/OMS d'experts. Rome,FAO 1989.

FAO/OMS. Preventing Spedfic MicronutrientDeficencies. Theme Paper No. 6. ICN, Rom, 1992.

Fleming AF, GhatouraGBS, HarrisonKA, Brigs ND, Dunn DT. The preventionof anaenia in pregnancyin primigravidaein the guinea savanna of Nigeria. Ann Trop Med Paraitol 1986; 80:211-233.

FlemingAF. Anemiematernelle dans le nord du Nig6ria:causes et solutions.Forun Mondialde la SantO 1987;8:365-370.

145 Biblography

Fleming AF. Tropicalobstetrics and .1. Anaemiain pregnancy in tropical Africa. Trans Roy Soc Trop Med Hyg 1989; 83: 441448.

Flores M. La femmneenceinte et l'antmie. M&reset Enfants, 1990; 19(3):6-7.

Galan P, Cherouvrier F, Mashako L, Nsibu C, Kapongo C, Preziosi P, Hercberg S. Iron bioavailability from Africanmeals with rice, cassava, or plantainforming the staple food. J Clin BiochemNutr 1991; 10:217-224.

Galan P, Cherouvrier F, Zohoun I, Zohoun T, Chauliac M Hercberg S. Iron absorption from typical Wect African meals containing contaminatingFe. Br J Nutr 1990;64: 541-546.

Galloway R. Determinantsof medical compliance.Washington, D.C. Consultancyreport to The World Bank; 1990.

Gillespie SR. Mason JB. Preventing anaemia. SCN News 1990;Numb. 6:1-7.

Greenwood BM. The effects of malaria chemoprophylaxisgiven by traditionalbirth attendants on the course and outcome of pregnancy. Trans Roy Soc Trop Med Hyg 1.989;83:589-594.

Guiro AT, Galan P, Cherouvrier F, Sall MG, Hereberg S. Iron absorption from african pearl millet and rice meals. Nutrition Research 1991; 1.1:885-893.

Hercberg S, Galan P, Dupin H. Iron deficiencyin Africa. Wld Rev Diet 1987; 54:201-36.

Hussein MA. Hassan HA, Abdel-GhaffarAA, Salem S. Effects of iron supplementson the occurrence of diarrhoea among children in rural Egypt. Food Nutr Bull 1988; 10:35-39.

Jackson RT, Jackson LC. Biologicaland behavioralcontributors to anemia during pregnancy in Liberia, West Africa. Hum Biol 1987; 59:585-597.

Jackson RT, Latham MC. Anemia of pregnancyin Liberia, West Africa: a therapeutictrial. Am J Clin Nutr 1982;35:710-714.

Jenkinson D. Single-dose intramuscular iron dextran in pregnancy for anaemia prevention in urban Zambia. I Trop Med Hyg 1984; 87:71-74.

Latham MC. StephensonLS. Kinoti SN, Zarr'anMS. Kurz KM. Improvementsin growth followingiron supplementationin young Kenyanschool children. Nutrition 1990; 6:159-165.

Levin HM. A benefit-costanalysis of nutritionalprograms for anemiareduction. Research Observer, The World Bank. 1986; 1:219-246.

146 Prevention of iron defciency

Maire B, Briend A, Carles C, Le Francois P, Chevassus-AgnesS, Ndiaye A. Etude des relations h6matologiques"n*re-cordon" dans unc maternite de la periph6rie de Daar. 29e Congres FEder Gyn6co et Obstetr Lgue Franc, Daar, 26-29 mai 1982.

Morrow 0. Iron supplementationduring pregnancy: why aren't womencomplying? A review of available informadon. Geneva, WHO, Safe Mofterhood programme;WHOIMCH/90.5; 1990.

Murray MJ, Murray AB, Murray MB, Murray Ca. The adverse effect of iron repletionon the course of certain infections.Br Med J 1978; 2:1113-1115.

Nyazema NZ. Towards better patient drug complianceand comprehension:a challenge to medical and pharmaceuticalservices in Zimbabwe.Soc Sci Med 1984; 18:551-554.

Ogunbode 0, Damole IO, Oluboyede OA, Path D. lIronsupplementation during pregnancy using three different iron regimes. Curr Ther Res 1980; 27:75-80.

OjofeitimiEO, Elegbe I. Babafemi J. Diet restrictionby pregeant women in Nigeria. Irt J Gynaecol Obstet 1982; 20:99-103.

Okafor LA, Diejomaoh FME, Oronsaye AU. Bone marrow status of anaemic pregnant women on supplementaliron and folic acid in a Nigerian community. 1985; 36:500-503.

OMS (WHO). Rapport sur la consultation r6gionaleafricaine sur la lutte contre l'an*mie pendant la grossesse. Brazzaville, Congo; 25-29 septembre 1989: OMS, Bureau regional pour l'Afrique, AFRJMCHt86;AFRINUIT/104; 1989.

Powers IU, BatesCJ, PrenticeAM, Lamb WH, Jepson M, BowmanH. The relative effectivenessof iron and iron with riboflavin in correcting a microcyticanaemia in men and children in rural Gambian. Hum Nutr: Clin Nutr 1983; 37C:413-25.

Preziosi P. Evaluationde l'efficaicited'une supplementationpar le fer dans la preventionde l'andmie en zone tropicale. 2tude controlee chez 347 enfants scolarises au Benin. DEA de Statistiqueet Sante, Universite de Paris-Sud, oct. 1990.

ScrimshawN. La carence en fer. Pour la Science (Scientific ''nerican) 1991, 170:90-96.

Smith AW, Hendrickse RG, Harrison C, Hayes RJ, GreenwoodBM. Iron-deficiencyanaemia and its response to oral iron: report of a study in rural Gambianchildren treated at home by their mothers. Ann Trop Paediatr 1989(a); 9:6-16.

Smith AW, HendrickseRG, Harrison C, Hayes RJ, GreenwoodBM. The effects on malaria of treatment of iron-deficiencyanaemia with oral iron in Gambianchildren. Ann Trop Paediatr 1989(b); 1:17-22.

StephensonL. The impact of helminth infectionson . London, Taylor & Francis, 1987.

147 Bibliogra'hy

SvanbergU, SandbergA-S. L'absorptionferrique am6liorte des produits de sevrage grice A la germinationet A la fermentation.In: Pour am6liorerl'alimentation des jeunes enfantsen Afrique orientaleet austraue:une teuhnolgie i la porte des m6nages.Compte rendu d'un atelier tenu i Nairobi,Kenya du 12 au 16 octobre1987. Ottawa, Ont., CRDI. 1989;[DRC-265 f: 415423.

Taylor P, Mutanbu SL. Compliancewith malaria chenoprophylaxis programmes. Acta Tropica1987; 44:423-431.

148 Preventionof iodine defciency

Preventionof iodinedeficiency

Abdel-WahabM.M., Elton M., OmerM.I.A., MukhtarE. KaballoA.M. Prophylaxisand treatmentof endemicgoitre in WestemSudan with intamuscular iodized oil. AnnTrop Paediatr 1984; 4:159-164.

ACC/SCN.A GlobalStrategy for the Preventionand Control of IDD:Proposal for a Ten YearProgramm of Supportto Countries.Report of the 13thSession of the ACCSub Committeeon Nutritionand its AdvisoryGroup on Nutrition.Washington D.C., 2-6March, Doc. No SCN87/MODA. January 1987, ACCISCN.c/o FAO,Rome.

AfricanTask Force on IodineDeficiency Disorders. Report of the fourthmeeting. Dar-es-Salaam, 28 Febr.-2 March 1990.

AlnwickDJ. Kenyaoulaws non-iodatedsalt. IDD Newsletter1988; 4(4):5.

BrownR., SeyfarthS., BrownR.C. A twelveyear follow-upof iodinedeficiency in Bulape,Zaire. Ann Soc beigeMed Trop 1989;69:251-55.

Cliff J, LundquistP. RoslingH, SorboB, Wide L. Thyroidfunction in a cassava-eaingpopulation affectedby epidemicspastic paraparesis. Acta endocrinol (Copenh) 1986; 113: 523-528.

DelangeF., AhluwaliaR. La toxicit6du maniocet la thyroide:recherches et qlestionsde santepublique; compterendu d'un coiloquetenu i Ottawa,Canada, du 31 mai au 2 juin 1982.Ottawa, Ont., CRDI, 1985.(IDRC-207f).

EkpechiOL. Iodinedeficiency disorders in Africa.In: The preventionand controlof iodinedeficiency disorders,B.S. Hetzel,J.T. Dunnand JLB.Stnbury eds. Amsterdam,Elsevier; 1987: 219-236.

EltomM., KarissonF.A., KamalA.M., Bostr6mH., DahlbergP-A. The effectivenessof oral iodized oil in the treatmentand prophylaxis of endemicgoiter. J ClinEndocrinol Metab 1985; 61(6):1112-17.

FAO/OMS(WHO). Preventing Speafic MicronurientDejliences. ThemePaper No. 6. International Conferenceon Nutrition,1992.

Hetzel B.S. The preventionand control of iodinedeficiency disorders. ACCISCN State-Of-The-Art Series;Nutrition Policy Discussion Paper NO 3; clo FAO,Rome 1988.

HetzelBS. Thestory of iodinedeficiency. An interationalchallenge in nutrition.Oxford, Oxford Univ. Press, 1989.

IDD Newsletter.IDD in Eastem,Central and SouthemAfrica. IDD Newsletter 1987; 3(3):3.

IDD Newsletter.Zaire plansIDD control. IDD Newsletter 1988; 4(3):7-8.

149 Bibliography

Ingenbleek Y, Beckers C. Triiodothyronineand TSH hormones in PCM in -infants. Lancet 1975; 2: 8454847.

Ingenbleek Y, De Vissher M. Hormonal and nutritional status: critical conditions for endemic goitre epidemiology?Metabolism 1979; 28:9-19.

Madinger E. Ethiopia: advocacyseninar on the preventionand control of IDD. IDD Newsletter 1988; 4(4):4.

Mannar MGV. Control of IDD in Tanzania, Malawi, and Kenya through salt iodation. IDD Newsletter 1988; 4(1):7.

Mannar MGV. The status of IDD control in ten Eas"-,Central, and South African countries. IDD Newsletter 1989; 5(1):1-5.

OMS (WHO). Regional strategy and plan of action for the control of iodine deficiencydisorders (IDD) in the African Region. OMS, Office r6gional pour lIAfrique, Brazzaville, Congo, 1986, document AFRINUTI96.

OMS (WHO)/UNICEFIICCIDD.Rapport de la 36me reunion du Groupe Sp6cialsur les Troubles de la Carence en Iode. Abidjan, 13-17 nov. 1989.

Osman A.K., Basu T.K., DickersonJ.W.T. A goitrogenicagent from millet (pennisetumtyphoides) in Darfur Province, Western Sudan. AmnNutr Metab 1983; 27:14-18.

Phillips D., Osmond C. Iodine supplementationwith oral or intramuscular iodized oil. A two year follow-upof a comparativetrial. Int J Epidemiol1989; 18(4):907-10.

Phillips D.l., Lusty T.D., Osmond C., Church D. Iodine supplementation:comparison of oral or intramusculariodized oil with oral potassiumiodide. A controllrd trial in Zalre. Int J Epidemiol1988; 17(1):142-7.

Thilly C.H. Goitre et cretinisme end6miques,r6le 6tiologiquede la consomniationde manioc et strategie d'6radication. Bull Acad Roy Med Belg 1981; 136(horss6rie):389-412.

Thilly C.H., Delange F., Ramioul L., Lagasse R., Luvivila K., Ermans A.M. Strategy of goitre and cretinism control in central Africa. It J Epidemiol1977; 6:43-54.

Van Amelsvoort V. Rural water supply and the recent appearance of endemic goitre. Trop Geog Med 1971; 23:304-305.

Van der Haar F., Kavishe F.P. Iodine deficiencydisorders in the Region Eastem, Central and Southem Africa: proceedings of a symposium held in Gaborone (Botswana), 20 Nov. 1986. Wageningen. NINI/ICFSN, Report 5.

150 Pmendon of iodinedefilency

Wichter W., Mvungi M., K6nig A., Pickardt C.R., Scriba P.C.S. Prevalence of goitre and hypothyroidismin SouthemTanznia: effectof iodisedoil on thyroidhormone deficiency. J Epidemiol CommunityHealth 1986; 40:86-91.

WHOIUNICEF/ICCIDD.Regional Seminar on Control of Iodine Deficiency Disorders in Africa. Working Papers. 23-25 March 1987, Yaounde,Cameroon.

151 libfography

Promoting breasteeding

Academyfor EducationalDevelopment. 1991. "Questionset riponsessur l'alimenationdu nourisson: un grouped'experts rEexamine le sujet.",Washington, D.C.

Akre, James (ed.). 1989. "InfantFeeding: The PhysiologicalBasis.' Bulletin,67(Supplement). The ScientificJournal of WHO.Geneva, Switzerland.

APHA. 1989. "CIaringhouseon InfantFeeding and MaternalNutrition." Legislationand Policiesto SupportMaternal and ChildNutrition. Report No. 6

Amningshjalpen.1990. Personal Communication.

Bamisaiyc, A., and Oyediran, M.A. 1983. "Breastfeedingamong female employees at a major health institution in Lagos, Nigeria." Social Science and Medidne, 17(23):1867-1871.

Baumnslag.N. 1988.Breas(feeding: The Passportto Life. UNICEF,New York.

Cuniingham,A.S., D.B. Jelliffe,and E.F. PatriceJelliffe. 1992. "Breastfeeding, Growth and Illness:An Annotated Bibliography."Healh in the 1980s. UNICEF.

DaVanzo,J. d al. 1985. "Quantitativestudies of mortalitydecline in the developingworld." World Bank Staff Working Papers No. 683. Washington, D.C.

De Oliveira, D.N. 1990. Director of , Taguatinga Hospital, personal communication.

Dimond, Hilary J., and Ann Ashworth. 1987. "InfantFeeding Practices in Kenya, Mexico, and Malaysia: The Rarity of the ExclusivelyBreastfed Infant." Human Nutrition:Applied Nutrition, 41A:51-64.

Green, Cyntia P. 1989. "MediaPromotion of Breastfeeding: A Decade's Experience." Nutnition ConmnunicationProject of the USAID.

Hardy, E. 1990. "Presentation at Inter-Agency Meeting on Breastfeedingand Waged Employment." UNICEF,Brazilia.

Hargot,E. 1989. "CaseStudies of the MelkaOba-Yetnora Cooperatives in Ethiopia,"a paper prepared for the ConsultativeGroup on Early ChildhoodCare and Development,New York, May 1989. Mimeo. Cited in Myers, Robert. 1992. 'The Twelve Who Survive, StrengtheningProgrammes of Early Childhood Developmentin the Third World." Routledgein cooperationwith UNESCO for the ConsultativeGroup on Early Childhood Care and Development.London and New York.

Hartman, P. 1990. "Presentationat ILCA Conference," Phoenix, Arizona.

152 PromotingbreWfeedixg

Hitchcock, N. and J. Coy. i988. "Infant Feeding Practicesin Western Australia and Tasmana: A Joint Survey, 1984-1985." The Medical Journal of Australia, 148(February1):1114-117.

Huffinan, S.L., and A. Steel. 1992. "BreastfeedingPromotion in Belize." Center to Prevent Childhood MalnutritionPolicy Series 3(1):1-14.

Jelliffe, D.B. and E.F. Patrice Jelliffe. 1988. TProgrammesto Promote Breastfeeding."OUP Delhi.

Katcher, A., andM. Lanese. 1989. "BreastfeedingbyEmployedMothers:A ReasonableAccommodation in the Workplace. Pediatrics,75(4):644-647.

Labbok, M., and M. McDonald. 1990. "Proceedingsof the Interagency Workshop on Health Care Practices Related to Breastfeeding/InternationalJournal of Gynaecologyand ", 31 (Suppl. 1).

Lawrence, Ruth. 1991. "Presentationto the US Congress on the Marketing of Infant Formula."

Mata, Leonardo et al. 1983. "Effect of Hospital an HLeldInterventions on Breastfeeding,Health and Survival of Infants." Paper presented at the Western Hemisphere Nutrition Congress VII, Miami, Florida, August 7-12, 1983.

Minei, Alfred Potou. 1992. "A DiscrepancyAnalysis of the Situation: Evaluationof the Implementation of the InternationalCode of Marketing of Breastnilk Substitutesin papua New Guinea in Country PresentationsTraining Course on Implemnentingthe Intemational Code of Marketing of Breastmilk Substitutes," Penang, Malaysia 17-28 Febuary 1992.

Miter, Swasti. 1986. "Common Fate CommonBond." London: Pluto Press.

Ojofeitimi,E.O., I. Elegbe, U.T. Etuknwa. 1982. "Knowledgeand BreastfeedingPractices among Nurses and Teachers in Ile-Ife, Nigeria." PediatricNursing, Nov-Dec, 8(6):400402.

Phillips, Margaret A., Richard Feacham, and Anne Mills. 1987. "Options for Di,--hoea Control: The Cost and Cost-Effectivenessof Selected Interventions for the Prevention of Diarrhoea." EPC PublicationNo. 13. London School of Hygiene and .

Rea, M. Ferreira. 1990. "The BrazilianNational Breastfeeding Program: a SuccessStory." Internadonal Journal of GynecologicalObstetrics, 31(Suppl. 1):79-82.

Tomkins, Andrew, and Fiona Watson. 1989. "Malnutritionand Infection: A Review." ACC/SCN State of the Art Series. Nutrition Policy DiscussionPaper No. 5.

Traore, Baba, Mamdon Traore, and Oona Campbell. 1990. "Etude des Effets de l'Allaitement sur la Morbidit6Diarrh6ique, I'Etat Nutritionnelet la Mortalit6des Enfantsau Mali. " in Hill, Alan G. (ced.). 1990. "Determinantsof Health and Mortality in Africa." Demographicand Health Surveys No. 10. Institute for Research Development.London Schoolof Hygiene and Tropical Medicine.

153 Bibliography

UNICEF. undated. "Adoptezl'initiative des amis des b6b6s." 15 page pamphlet. New York.

Van Esterick, Penny. 1992. "Women,Work, and Breastfeeding."Cornell InternationalMonograph Series No. 23.

Victora, C.G. et al. 1989. "Infant Feeding and Deaths due to Diarrhea." American Journal of Epidemiology, 129:1032-1041.

Welistart, and Perinasia. 1988. "Proceedingsof the InvitationalAsian RegionalLactation Management Workshop and Related Events". June 29-July 1988. Bali, Indonesia.

WHO/UNICEF. 1981. "TheWHOIUNICEF Intemational Code of Marketingof Breastnilk Substitutes." WHO Geneva.

WHOIUNICEF. 1990a. "Breastfeedingin the 1990s:Review and Implicationsfor a Global Strategy." Technical Meeting. 25-28 Juin 1990, Geneva.

WHOIUNICEF. 1990b. "Innocenti Declaration on the Protection, Promotion, and Support of Breastfeeding."adopted at WHO/UNICEFPolicymakers' rneeting on "Breastfeedingin the 1990s:A Global Initiative."

WHO. 1991. "Indicatorsfor Assessing Breast-FeedingPractices." Report of an informal meeting, June 11-12, Division of Diarrhoealand Acute RespiratoryDisease Control, Geneva.

WHO. 1992. "Indicatorsfor AssessingBreastfeeding Practices: Report of an Informal Meeting." 11-12 June 1992. WHO/CDD, Geneva.

Zinkin, P. 1983. MozambiqueMOH. PersonalCommunication

154 Dietarytreatment of infections

Dietary management of infections

Abosede OA. Traditionalinterventions for diarrhoeas in Nigeria: Views of communityhealth personnel. Proceedings of Nigeria/JapanJoint Conference,pp. 143-6, 1987.

ACC/SCN. Second report on the world nutrition situation. Volume I: Global and regional results. ACCISCN; Geneva, 1992.

Akre J (ed). Infant feeding: The physiologicbasis. Bull WHO 1989; 67(suppl):1-95.

Ashworth A, Draper A. The potential of traditional technologiesfor increasing the energy density of weaning foods: A critical review of existing knowledge with particular reference to nmaltingand fermentation.Geneva: World Health Organization, 1992.

Behrens R, Nichter M, Herman E, Brown KH. Integrating ORT and feeding: Dietary managementof diarrhoea at the household level. In: Cereal based oral rehydration therapy for diarrhoea. Elliot K, ,AttawellK, Wilson R, et al (eds). Report of the International Symposiumon Cereal-Based Oral RehydrationTherapy, November 12-14, 1989. Geneva: Aga Khan Foundation, 1990.

Bentley ME, Dettwyler KA, Caulfield LE. Child anorexia and its managementin developingcountry settings: Review and recommendations. Paper prepared for the workshop and expert meeting on guidelines for nutrition and feeding of children 0-5 years in the Latin American region. Porlanar, Venezuela, March 1419, 1993.

Bentley ME, Dickin KL, Mebrahatu S et at. Developmentof a nutritionallyadequate and culturally appropriate weaning food in Kwara State, Nigeria: An interdisciplinaryapproach. Soc Sez Med 1991;33(10):1103-1I.

Brown KH, Dickin KL, BentleyME et al. Consurmptionof weaning foods from fermented cereals in Kwara State, Nigeria. In: Improvingyoung child feeding in Eastern and SouthernAfrica: Household- level food technology.Alnwick D., Moses S, SchmidtOG (eds). Proceedingsof a workshopheld in Nairobi, Kenya, October 12-16, 1987. Canada:IDRC, 1988.

CARE/Cameroon,Education Development Center, Inc., The Manoff Group, Inc. Improvingyoung child feeding practices in Cameroon:Projea overview. Washington,D.C.: Manoff Group, Inc. 1989.

CoulibalyM. Progranmnenational de lutte contre les maladiesdiarrheiques de la Mauritanie. Etude de la prise en charge de cas de diarrhee dars les structures de sante. PRITECH trip report, May, 1989.

Dettwyler KA. Breastfeeding,weaning, and other infant feeding practices in Mali and their effects on growth and development. Ann Arbor: UniversityMicrofilms International, 1985.

Dettwyler KA. Interaction of anorexia and cultural beliefs in infant malnutritionin Mali. Am J Human Bio! 19g9;1:683-95. 155 Bibliography

Dickin KL, Brown KI, Fagbule D et at. Effect of diarrhea on dietary intake by infants and young children in Kwara State, Nigeria. Euro J Clin Nutr 1991;44:307-317.

Diene S.M., 'bnproving Child Feeding Practices During and After Diarrhea and Malnutrition in the Fatick Region of Senegal". report to Pritech, August 1993 (unpublished).

Duran-BordierA. Pre-trainingneeds assessmentreport: BurkinaFaso. NutritionConununications Project trip report. March-April, 1992.

Esrey SA, Guptill KS, Rosales F. Evaluation of pilot interventionsin Peru and Nigeria. Report of the Dietary Managementof Diarrhea (MDM) Project: 1985-89.Undated mamnscript.

Gillespie Stuart et Mason John. Nutrition-RelevantActions: Some Experiences from the Eighties and Lessons for the Nineties. UN's ACC-SCNNutriton Policy DiscussionPaper No. 10. October 1991.

Griffiths M. Improvingyoung child feeding practices: The Weaning Project Final Report. Washington D.C.: The Manoff Group, Inc. 1992.

Guptill KS. Evaluation of an infant weaning food intervention in Kwara State, Nigeria. Doctoral dissertation.The Johns Hopkins University, 1990.

Guptill KS, Esrey SA, Oni GA, Brown KH. Evaluadon of a face-to-faceweaning food intervention in Kwara State, Nigeria: Knowledge,trial, and adoptionof a home-preparedweaning food. Soc Sci Med 1993;36(5):665-72.

Herman E and Keith N. A sunmary of PRUIECH-sponsoredresearch to improve child feeding during diarrhea. Report submittedto the Nigerian Ministry of Health and Social Affairs. June 29, 1993.

Hung MM. Evaluationde la valeur nutritive de trois preparationstraditionnelles (koko, fura et tuwo) et de recettes pr6conis6espar le projet PRITECH. Niamey, Mai 1992.

Keith N. Fura and koko during and after diarrhea episodes: A study to determine the feasibility of enrichingtraditional weaning foods. PRITECHtrip report. November26-December 17, 1990.

Keith N. Young childJeeding,weaning and diarrhea illness: Practicein a Hausavitlage and educational implicarions.Part I: The edmographicstudy. Virginia: PRITECH, 1991a.

Keith N. Fura and koko during and after diarrhea episodes: A study to determine the feasibility of enrichingtraditional weaning foods. PRITECHtrip report on Phase II, February 22- April 20, 1991.

Konde-LuleJK. Elasu S. MusongeDL. Knowledge,attitudes, and practices and their policy implications in childhoolddiarrhoea in Uganda. J DiarrhoealDis Res 1992;10(l):25-30.

Ministry of Health. Improvingyoung childfeeding practices in Ghana. Volume 2: Household interview and observations.Accra: Nutrition Division, 1989a.

156 Dietarytreatmen of infections

Ministry of Health. Improvngyoung childfeeding practices in Ghana.Volume 3: Householdintervendons trials. Accra: Nutrihon Di'visibn, 1989b.

Ministry of Health. Improvng young child feeding practices in Ghana. Volume 4: Focus group discussions.Accra: Nutrition Division, 1989c.

Ministry of Healt. Improvingyoung childfeeding practice in Ghana. Volume5: SummaryReport. Accra: Nutrition Division, 1989d.

Mosha AC, SvanbergU. The acceptanceand intake of bulk-reducedweaning foods: The Luganga village study. Foodand Nutr Bull 1990;12(1):69-74.

National Nutrition Council. Focusgroup discussions: Projects for promotionof improvedyoung child feeding. UNICEF; Swaziland, 1987.

NationalNutrition Council. Household observations and trias: Projectsforpromotionof improvedyoung childfeeding.UNICEF: Swaziland, 1988.

National Nutrition Council. The Manoff Group, Inc. Improvingyoung child feeding pracices in Swaziland:Project overview. Washington D.C.: The Manoff Group, Inc., 1992.

OdebiyiAl, Food taboos in maternaland child health: The views of traditionalhealers in lfe-Ife, Nigeria. Soc Sd Med 1989;28(9):985-96.

Oni GA, Brown KH, BentleyME, Dickin KL, KayodeB, Alade J. Feeding practices and prevalence of hand-feedingof infants and young children in Kwara State, Nigeria. Ecol FoodNutr 1990;25:1-11.

PRITECH/Sahel and ORANAISenegal. Les Maladies Dian*reiquesdans le Sahel : Donnees pidimiologiqueset premiersresulars de programme de lute. Dakar: PRTECH, 1989 (Revised 1993).

Rasmuson M, Booth EM. The role of formative evaluation in the Mass Media and Health Practices Project. In HEALTHCOM Commication for Child Survival Field Notes. Washington D.C.: Academy for EducationalDevelopment, 1985, pp.84-91.

RasmusonM. Seidel R, GabbidonH. Dietary managementof diarrhea: The GambianExperience. JNuMr Educ 199O:22 -fl71 RohdeF, Thiam A, FishmanC, Kante D. Mid-pointassessment of the NutritionCommunications Project. Washington D.C.: Acadeny for EducationalDevelopment, 1993.

Roisin A, Zerbo PJ, Corbin C. Etude de la prise en charge des cas de diarrlte dans deux groupes de quatre provinces du Burkina Faso. PRITECHtrip report. November-December,1990.

Samba K, GittelsohnJ. Improvingchild feeding practicesin The Gambia: A report of research conducted by the Gambia Food and NutritionAssociation. Banjul, March 21, 1991.

157 Bibliography

Samba Ndure K. Improvingchild feeding practices in T'he Gambia: A summary report of weaning food and educationmessage trials conductedby the GambiaFoodl and NutritionAssociation between April to October, 1992. Banjul, June 24, 1993.

Sene M. Etudes sur les pratiques alimentaireschez les enfants de 0-5 ans pendant la diarrhee au Senegal : Phase I. PRITECHtrip report, mai 1991 (mrvisedJanuary, 1993).

Sene M, Diop M, Ndiaye A, Samb A, Leye I. Etude sur les pratiques alimentaireschez les enfants de 0-5 ans pendant la diarrh6e au Sdnigal : Phase II. PRITECHtrip report, f6vrier 1992.

Spain P. CDD in Kenya: Policy and research on home treatment. Virginia: PRITECH, 1991.

SserunjogiL, Tomkins A. The use of fermentedand germinatedcereals and tubers for improvedfeeding of infants and children in Uganda. Trans Roy Soc TropMed and Hyg 1990;84:443-6.

StephensonL. Impactof HelminthInfections on Hunan Nutrition.London: Taylor & Francis, 1987.

TomnkinsAM. Nutrient cost of protracted diarrhoea in young Gambian children. GUT 1983;24:A459.

Tomkins AM, Alnwick D, Watson F. The potential role of fermentationand germination in improving young child feeding in Eastern and SouthernAfrica. A Joint UNICEF/LSHTMproject funded by the Overseas DevelopmentAdministration, 1990.

Toure C. Pratiquesalimentaires et maladiesdiarrh6iques des enfants de 0 a 5 ans au Mali. PRITECHtrip report. June, 1991.

Tulane University, Care/Cameroon, Education DevelopmentCenter, Inc., The Manoff Group, Inc. Evalutationof tire CnrneroonWeaning Project. WashingtonD.C.: The Manoff Group, Inc. 1989.

World Bank. Commentamdliorer la contributiondu secteur de la sante dans la lutte contre la malnutrition : Revue d'evaluations de projets en Afrique. Division des Ressources Humaines, DNpartement Technique, Region Afrique. Draft, janvier 1993.

WHO. Diarrhea treatment chart. Geneva: World Health Organization, 1992.

Youssouf G. Etude de la prise en charge des cas de diarrh6e dans les strLcruresde sante au Niger. Ministry of Health. Republicof Niger, May, 1992.

158 Dietary supplements

Food supplementation

Adair LS, and Pollitt E. Outcome of maternal nutritionalsupplementation: A comprehensivereview of the Bacon Chow study. American .ournalof Clinical Nutrition,vol. 41, pp. 948-978, 1985.

AndersonMA, AustinJE, Wray JD, ZeitlinMF. Supplementaryfeeding. Study 1, in: AustinJE, Zeitlin MF, Nutrition interventionin developingcountries, an overview. Cambridge(USA), Oelgeschiager, Gunn & Hain Publishers,Inc., 1981.

Beaton GH, OhassemiH. Supplementaryfeeding programs for young children in developingcountries. Am. J. Clin. Nutr., 1982, 35(4):864-916.

Bremer-FoxJ, Lang P, Pines J. Assessmentof the PL480 Outreach and EnhancementGrant Programs. US-AID, Washington, PD AAY 553, 1987.

BremnsS., Berg, A. Eating Down During Pregnancy: Nutrition, Obstetric and Cultural Considerations in the Third World. Paper prepared for the 1988 and 1989 meeting of the Advisory Group on Nutrition of the ACC Sub-Committeeon Nutrition.Washington, D.C.: 1989.

Burkhardt R., Field J.0. et al. Supplementaryfeeding in rural Egypt: a summary profile of the health system in action. M.I.T. and Cairo, University Health Care Delivery Systems Project. Monograph No.5, M.I.T. TechnologicalPlanning Program, M.I.T., Mass. USA, 1980.

CallowayDR, Murphy SP, Beaton GH. Food intake and humanfunction: a cross-project perspectiveof the collaborative research support progamme in Egypt, Kenya and Mexico. Final report, Univ. Califomia, Berkeley, Aug.1988.

Chanetsa J. Food and nutrition programmes : sectoral and problem-orierntedapproaches, child supplementary feeding. (Zimbabwe). In: van der Haar F (ed.), Executinagfood and nutrition programmnesin East, Central and SouthernAfrica: experience and practice. Workshop, 22 Aug.- 2 Sept. 1983, Harare (Zimbabwe).Report 2, Wageningen,NINI, ICFSN, 1983: 149-58.

Chaves A, Martinez C. The effect of maternal supplementationon infant development. Arch Latinoam Nutr, 1979;29(suppl1):143-53.

Demeke T, Wolde-GabrielZ. The impact of supplementaryfeeding programmeson the nutritional status of beneficiariesin Addis Ababa. Food and Nutr Bull 1985; 7:38-50.

Diene M. Le programme de r6hydratationpar voie orale et les activit6sde nutrition dans les regions de Fatick et Kaolack. Rapport de visite; Dakar, SANAS/ DHPS 1989.

Echenberg DF, Stubbs HA, King JM.. Evaluation du PL-480 title R et du PPNS; (programnv de protection nutritionnelleet sanitaire - groupes vulnerables);Sn*Egal, TS-AID 1984.

159 Bibliography

Esrey SA, Habicht JP, Butz W. A methodology to review public health interventions: results from nutrition supplementationand water and sanitationprojects. Cornell Intermat.Nutr. Monograph Series, Nber 15, 1985.

Fleuret A. Consumption Effects of Food Aid. Household Utilization of Title II Foods in Rural Kenya. Washington, US-AID, PN-AAW 219, 1985.

Ghassemi H. Supplementary feeding revisited. Consultancy Report, Food Policy and Nutrition Division, FAO, Rome, September 1989. [ not to be quoted]

Gilmore JW, Adelman CC, Meyer AJ, Thorne MC. Morocco: food aid and nutrition education. AID project impact evaluation report N°8, 1980.

Hartog (den) A.P. Food and nutrition programmes : sectorial and problem-oriented approaches: local and donated foods in supplementary feeding programmes. In: van der Haar F (ed.), Executing food and nutrition programmes in East, Central and Southern Africa: experience and practice. Workshop, 22 Aug.- 2 Sept. 1983, Harare (Z;imbabwe).Report 2, Wageningen, NINI, ICFSN, 1983: 167-78.

Hautvast J G A J and Maletnlema T N, (eds). Practical Considerations for Childfceding in East, Central and Southern African Countries. Report of a Workshop held in Arusha; (Tanzania) 16-19 Dec. 1981. Report 1. NINI/ICFSN, Wageningen.

Herrera MG, Mora JO, de Paredes B., Wagner M. Maternal weighttheight and the effect of food supplementation during pregnancy and lactation. In Aebi Hl, Whitehead RG, eds. Maternal mtrition during pregnancy and lactation. Bem: Hans Huber, 1980:252-63.

Hoorweg J. NiemneijerR. Intervention in child nutrition. Monographs from the African Studies Centre, Leiden. London, KPI, 1989.

Horton, S. Unit Costs, Cost-Effectiveness,and Financing of Nutrition Interventions, PHR Working Paper No. 952, August 1992, The World Bank.

Jones E. Berkowitz SA, Roussel R. Evaluation of title II. Food for peace in Ghana. Development Associates, Inc. Washington, US-AID, PD-AAI-761, 1981.

Kennedy E, Knudsen 0. A review of supplementary feeding progranunes and reconmnendationson their design. In: Biswas M. and Pinstrup-AndersenP. (eds.). Nutrition and Development. Oxford Univ. Press, 1985: 77-96.

Kent L. Review of recipient selection criteria, beneficiary profile, and nutritionalimpact of catholic relief service's use of title II commodities in Madagascar. US-AID, Washington, PD-AAY-933, 1988.

Kielmann A. Taylor CE. Parker RL. The Narangwal nutrition study: a summraryreview. Am J Clin Nutr 1978; 31:2040-2052.

160 Dietary supplements

King JM, Seaton KW. Evaluation of P.L. 480 Title II Food AID Program in Camneroon.Systematics -General Corporation. Washington, US-AID, PD-AAL-683, 1982.

King IM, Smith Lloyd E, Prins A, Rasmuson M. Evaluation report, Catholic Relief Services, Djibouti food and nutritionprogram. InternationalScience and TechnologyInstitute, Inc. Washington, US-AID, PD-AAR-800, 1984.

Kusin, J.A., Kardjati, S., Houtkooper, J.M., Renqvist, U.H. "Energy supplementationduring pregnancy and postnatal growth", Lancet, vol. 340, pp. 623-26, sept. 1992.

Lechtig A, Habicht JP, Delgado H, Klein RE, Yarbrough C, Martorell R. Effect of food supplementation during pregnancy on birth weight. Pediatrics, vol. 56, pp. 508-520, 1975.

Lechtig A. Klein RE. Maternal food supplementationand infant health: results of a study in rural areas of Guatemala. In Aebi H, Whitehead RO, eds. Maternal nutrition during pregnancy and lactation. Bern: Hans Huber, 1980:285-313.

Lotfi M., Mason J. Direct intervention programes to imnproveinfant and child nutrition. Draft; ACC/SCN, c/o OMS (WHO), Geneve; 1989.

Lutter CK, Mora JO, Habicht JP, Rasmussen-KM,Robson DS, Herrera MG. Age specific responsivness of weight and length to nutritional supplementation.Am J Clin Nutr 1990; 51: 359-364. -

MO-H1-, Zimbabwe. The children's supplementary feeding programme in Zimbabwe. 1982.

Marchione TJ. Food supplementationprograms and breastfeeding practices. It. J. Gynecol. Obstet., 1990; 31(Suppl.1):125-30.

Mason E, Tagwireyi J. From relief feeding to food production: case study of childs supplementary feeding programme and supplementary food production programme in Zimbabwe 1980-1989. "Nutrition for non nutritionists". World Bank, 794/3 1989.

McClelland DG, Gray Cowan L, Markey LS, Stengel CA. Evaluation report: Mauritania PL 480 Title 11Section 206 Program. US-AID, Washington, PD-AAD-968, 1984.

Morn JO, King JM, Teller CH. The effectiveness of maternal and child health (MCH) supplementary feeding programs. An analysis of performance in the 1980S and potential role in the 1990S. Washington, U.S. Agency for International Development, 1990.

Nyiribibi M. Suivi de la croissance des enfants d'ige prescolaire; experience du Rwanda. Atelier sur la surveillance alimentaire et nutritionnelle das les pays francophones d'Afrique Centrale et de l'Est; Kinshasa, 4-15 juin 1990.

Prentice A.M. et al. Increased birthweight after prenatal dietary supplementationof rural African women. American Journal of Clinical Nutrition, vol. 46, pp. 912-25, 1987.

161 Bibliogmphy

Quinn V, Cohen M, Mason J, Kgosidintsi BN. Crisis-proofing the economy: the response of Botswana to economic recession and droughlt. 1986; 2-27.

Rivera JA, Habicht JP, Robson DS. Effect of supplementary feeding in recovery from mild to moderate wasting in preschool children. Am J Clin Nutr 1991; 54:62-68.

Rose D. and Martorell R. "The impact of protein-enargysupplementation interventions on child morbidity and mortality." Paper presented at the seminar 'Child Health Priorities for the 1990s." June 20-22, 1991 at the Johns Hopkins University, School of Hygiene and Public Health.

Seepnappa M, Ljunqvist B. Food security and nutritional status of children in the Iringa nutrition programme in Tanzania. Third Africa Nutrition Congress, Harare, 5-8 Sept. 1988.

Stefanak MA, Jarjoura D. Weight gain in supervised and take-home feeding programmes in Chad. J. Trop. Pediatr., 1989; 35:214-7.

Stephens B, Forman L, Gaylord L. Outreach grant project assessment - Togo and Benin. Intemational Science and Technology Institute, Inc. Washington, US-AID, PN-AAR 340, 1984.

Targeted food aid: criteria for commodity choice. Report of a workshop held in Wageningen (The Netherlands) 23-26 September, 1986. Wageningen, NINIICFSN, Report 4, 1986.

US-AID. Kenya food for peace, P.L. 480 Title II Evaluation Final report. Practical Concepts Inc. Washington, PN-AAJ-799, 1980.

US-AID. Upper Volta food for peace/title II evaluation. Final report. Washington, PD-AAJ-909, 1981.

US-AID. PL 480 title II evaluations, 1980-1985: the lessons of experience. Washington, AID/FVA/PPE, June 1985.

US-AID. A review of developmental aspects of the cathwel program in Burkina Faso with options for future program directions. Washington, PD-AAU-585, 1986.

US-AID. Lesotho: phaseout of CRS MCH program- implicmtionsfor the future. Draft report. Washington, PD-AAY-562, 1988.

Van der Haar, F. Executing food and nutrition programmes in East, Central and Southern Africa: experience and practice. Proceedings of a workshop held in Harare (Zimbabwe), 22 August-2 September - 1983. Wageningen, Netherlands International Institute; International Course in Food Science and Nutrition. NINIIICFSN; 2. 1983.

WHOIUNICEF. Joint support for the improvementof nutrition in the United Republic of Tanzania, Iringa 1983-1988. Evaluation report, Dar-es-Salaarn, 1988.

162 Dietary supplemes

Zeitlin M, Bogyo S, Gold R, Kidd R, Nldabungu B. Interin evaluation report. April 1987. AID-ORT food for peace. Maternal-childhealth (Title I1) component, area nutrition improvement project. Zaire. Washington, US-AID, PD-AAZ-136, 1987.

163 Bibliography

Nutritional rehabilitation

Andrien M. Evaluation du programme d'6ducation nutritionmelledu Centre Hospitalier R6gional de Bouake. Service d'Evaluation du Minisd&rede l'Enseignement Prinaire et de IPEducationT6levisuelle de Cote d'lvoire; Laboratoire de P6dagogie experimentale de Li&ge,Belgique, mars 1983.

Anonvnie. Nutrition rehabilitation centers. Nutr Rev 1968; 26:230-232.

Asokumar B., Enahoro FO. Recovery rate of malnourished children in Zaria, Nigeria. J Trop Pediatr 1991; 37:77-82.

Assimadi K, Kessie K, T.ntaganK, Degboe A, Begue P. Le traitement hospitalier de la malnutrition proteino-calorique. Etudes Mtdicales 1984; N04:213-216.

Atakouma DY. Malnutrition proteino-energ6tiqueinfantile au C.H.U. de Lorn-Tokoin; (Togo) et prise en charge ambulatoire des enfants malnourris. M.Sc. Thesis, Institut de Medecine Tropicale Prince Ldopold, IMTA/MScBT/C2, Anvers, 1990; 1:1-57.

Bac M. Progress towards health for all in the Gelukspan health ward - 1985. M.D. Thesis, Medical University of Southern Africa; 1986.

Beau JP. Fontaine 0, Garenne M. Managementof malnourished children with acute diarrhoea and sugar intolerance. J Trop Pediatr 1990; 36:86-89.

Beau, JP. Le centre de rehydratation et de rer,utrition de Pikine (Senegal). L'histoire de Mamadou. ENDA, Dakar; 1990.

Beaudry-Darisme M. Nutrition rehabilitation centers. An evaluation of their performance. Env Child Health 1973; Sept.: 29E-332.

Beghin I, Van Lerberghe W. La courbe poids-cible": un outil pratique pour le suivi des enfants malnourris pris en charge. Communicationaux 4tncs Journ6es Internationalesdu GERM, SPA, 1989.

Beghin ID, Viteri FE. Nutritional rehabilitation centers: an evaluation of their performance. Environ Child Health 1973; 19:404-416.

Bengoa JM. Les programmnesde nutrition envisages sous l'angle de la sante publique, p.18 59; in: Nutrition et al. imentation tropicales, recucil des conferences du deuxibne cours de nutrition pour l'Afrique Noire, Marseille, oct.-dtc. 1955. Rome, FAO, 1957.

Bengoa JM. Nutrition rehabilitation centres. J Trop Pediatr 1967; 13:169-176.

164 Nuaridonal rabilitadion

Bengoa JM. Nutritional rehabilitation, chap. 23, pp.321-334; in: Nutrition in Preventive Medicine, G.H.Beaton and J.M.Bengoa eds.,Gentve, OMS, 1976.

Birem-Etchebes S., Gonzales JP. Essai de rEsolutionde la malnutritionprotido-calorique du jeune enfant dans une province de la R6publiqueCentrafricaine. Bull Liais Doe OCEAC 1987; 80:133-140.

Bouvier M. Centre de rehabilitation nutritionnelle: Hopital Protestant de Dabou; Cote d'lvoire. Avril 1985.

Bowie MD, Moodie AD, Mann MD, HansenJDL. A prospective 15-year follow-up study of kwashiorkor patients. Part I. Physical growth and development. S A Med J 1980; 58:671-676.

Briers PJ, Hoorweg J, Stanfield JP. The long term effects of protein energy malnutrition in early childhood on bone age, bone cortical thickness and height. Acta Paediatr Scand 1975; 64:853-858.

Brown JE, Brown RC. An evaluation of nutrition centre effectiveness by measurement of younger siblings. Trans Roy Soc Trop Med Hyg 1979; 73(1)70-73.

Brown RC, Brown JE, Teeter RA. Evaluation of a nutrition center program in rral Africa. J Trop Pediatr 1980; 26(l):37-41.

Cameron N, Jones PRM, Moodie A, Mitchell J, Bowie MD, Mann MD, Hansen JDL. Timing and magnitude of adolescent growth in eight and weight in Cape coloured children after kwashiorkor. J Pediatr 1986; 109:548-555.

Case study: Mwunii, a people centered approach to community health has positive impact on health status in rural Tanzania. World Neighbors, 1987; 9189:1-23.

Cook R. Is hospital the place for the treatment of noshed children? Environ Child Health 1971; Marcdh:15-25.

Dagnelie PC. Mobile nutrition rehabilitation of malnourished children. organization, cost and results of rehabilitation at home, with regular visits, in small fixed groups, to a nutrition unit in Enib6, Kenya. Wageningen, Agricultural University, 1979.

Degboe AF, Kampatibe N, Doh A, Tatagan K, Begue P, Assimadi KI Les formes cliniques de la malnutrition prottino-calorique (M.P.C.) au Togo. Etudes Medicales, 1984; N0 4:199-210.

Ferre P. Evaluation du centre de rEhabilitationnutritionnelle de Dabou (C8te d'Ivoire). DESS, Univ. de Montpellier U, 1991.

Ferre P. Prise en charge de la malnutrition i l'Hopital Protestant de Dabou (C6te d'Ivoire). Thtse, Facult6 de M6decine de Montpellier, 1991.

Fisher MM, Killeross MC. Simonsson M, Elgie KA. Malnutrition and reasoning ability in Zambian school children. Trans Roy Soc Trop Med Hyg 1972; 66:471 478.

165 Bibliography

Fontaine 0, Beau iP, Briend A. Un centre exp6rimental de rehydratation et de r6habilitation nutritionnelle:rEsultats preliminaires. Rev Int Pediatr 1984; N° 142:21-23.

Fontaine O., Beau J.P. Ort and nutritional rehabilitation in malnourished infants with diarrhoea. Joint third scientific working group on drug developmentand managementof acute diarrhoeas. Diarrhoeal Diseases Control Programme, WHO; working paper n 16; Geneva, 1-3 Oct. 1984.

Fontaine O., Beau J.P., NDiaye AM. Rehydratationpar voie orale et rdhabilitationnutritiomelle chez des enfants s6verementmalnutris. Enfant Milieu Trop 1987; 33:177-180.

Gatler JR, Ramsey FC, Salt P, Archer E. Long-term effects of early kwashiorkor compared with marasmus. III. Fine motor skills. I Ped Gastr Nutr 1987; 6:855459.

Gallin M, Pecoul B. Etude de l'impact des centres de recup6ration et d'6ducation nmtritiomelles ambulatoires; (CRENA) au Niger. Medecins sans fronti6res; Epicentre, 1988.

Glatthaar II, Fehrsen GS, Irwig LM, Reinach SG. Protein-energy malnutrition: the role of nutrition education in rehabilitation. Hum Nutr: Clin Nutr 1986; 40C:271-285.

Gressard A. Le centre nutritionnel de Ruhengeri (Rwanda) et l'alimentation infantile au Rwanda. Med Trop 1974; 34(3):427-444.

Hazoume FA, ToukourouR. Ricup6rationdes enfants atteints de marasme nutritionnelet de kwashiorkor marastique. Afr Med 1982; 21(205):695-697.

Hennart P, Beghin D, BossuytM. Long-termfollow-up of severe protein-energymalnutrition in Eastern Zaire. I Trop Pediatr 1987; 33:10-12.

Hone NM. A survey of the managementof severe childhood malnutritionin Zambian hospitals. J Trop Pediatr 1986; 32:249-250.

Hoorweg J. Niemeijer R. The effects of nutrition rehabilitationat three family life training centres in central province, Kenya. Leiden, The Netherlands, African Studies Center, 1982.

Hoorweg J. Impact evaluation of child nutrition programmes. Food Pol 1988; 13:199-208.

Jansen AAM,Verkley MTB. Ambulatory home nutrition rehabilitation in rural Kenya. J Trop Pediatr 1986; 32:258-262.

Janssen F. Bouton JM, Vuye A, Vis HL. Evaluation of clinical and biological parameters in marastic kwashiorkor children treated by parenteral nutrition. J Parent Enter Nutr 1983; 7(1):26-36.

Korte R. Patel PM. Operational aspects of mobile and stationnary young child clinics in Lushoto, Tanzania. J Trop Pediatr 1974; 20, Monograph no 33.

166 Nutritional rehabilitation Kraut H, Kreysler J, Lal K, MndemeK, Moshi H, OltersdorfU, Plesser T, Schach E, Bock E. Rehabilitationof undernourishedchildren in Tanzaniausing locallyavailable food. Ecol Food Nutr 1978;6:231-242.

LaboyaN, De GraeveG, Tellier V. El trataniento anbulatoriode los ninosdesnutritdos en Kepolowe; (Zaire - Shaba). Semin.Intemac. Rehab. IntegralNutric. 1990;La Paz, IBBA,libro de resumenes: p.9 1 .

LarchetM. Cosnard G, Joullie M. Le centre de rthabilitationinfantile et d'tducation nutritionnellede Ouahigouya(Haute-Volta). Med Trop 1977;37(3):257-264.

LemaireB, NikobitungwaR, BuheturaS. R6sultatsde la r6cup6rationambulatoire d'enfants malnourris dans la plainede l'Imbo, Burundi.Section de Nutrition,Ministtre de la SantEPublique, Bujumbura, Burundi, 1990.

LesbordesJL, CoulaudB, GeorgesAJ. Malnutritionet virus de PinununodEficiencehumaine; (VEI) i Bangui(Centrafrique). Med Trop 1990; 50(2):161-165.

LifandaK. Consultancyreport on Save-The-Childrenhealth and nutritionprogramne. BurkinaFaso field office, 1988.

Mauppm-PierreCRJ. Reflexionsmnthodologiques k retrer d'une action de rehabilitationnutritionnelle dans un secteurpopulaire de Dakar (Senegal).A propos de 596 observations.Thtse de medecine, Universitede Lile n; 1986.

OjofeitimiEO, TeniolaSO. Evaluationof a nutritionrehabilitation centre in ile-Ife, Oyo State, Nigeria. Wld Rev Nutr Diet 1980; 35:87-95.

OkehialamTC. Growthof nigerianchildren withi marasmus after hospitaltreatment. J Ntnl Med Ass 1983;75:7540.

OMS(WHO). La malnutritionprotdino-Energdtique sdvere: traitement et conduitetherapeutique. Geneve, OMS, 1982.

PecoulB, SoutifC, HounkpeviM, DucosM. Efficacyof therapeuticfeeding center evaluatedduring the hospitalizationand a follow-upperiod, Tahoua,Niger 1987-1988.Paris, M6decinsSans Frontieres, 1988.

Rey JL, Roy C, Denis F, Leonard0, HoudierR, DarracqR, Soro B. L'infectionVIH dans un centre de recuperationnutritionnelle i Abidjan(CGte d'lvoire). Rev P6diatr1990; XXVI(8):271-274.

Satgt P, Mattei JF, Dan V. Avenirsomatique des enfants atteintsde kwashiorkor.Ann P6diat 1970; 17:368-381.

167 Bibliography

Schneidman I, Bennett FJ, and Rutishauser HE. The nutrition rehabilitation unit at Mulago hospital, Kanpala. Development and evaluation, 1965-1967. J Trop Pediatr Environ Child Hlth 1971; 15:25-32.

Smith IF. Rehabilitationof protein-energymetabolism and hormonal status in Nigerian children. Nutr Rep Int 1981; 24(3):603-613.

Smith IF, Taiwo 0, Golden MHN. Plant protein rehabilitation diets and iron supplementation of the protein-energy malnourished child. Eur J Clin Nutr 1989; 43:763-768.

Stanfield JP. The Luteete family health centre: nutrition rehabilitation in a comprehensive rural development strategy. J Trop Pediatr Environ Child HIth 1971; 17:67-82.

Stoch MB, Smythe PM. Fifteen-year developmentalstudy on subsequentphysical growth and intellectual functioning. Arch Dis Child 1976; 51:327-336.

Tellier V. Rapport d'une visite de travail i Kapolowe et a Lubumbashi. Anvers, Inst Med Trop, Dept de Nutrition, 1990.

Teyssier J, Lallement AM, Imbert P, Diaine C, Terrissol M. Etude de la morbidite et de la mortalit6d'un service de pediatrie &Dakar. Med Trop 1986; 46(1):51-61.

Tolboom JIM, Ralitapole-MarupingAP, Kabir H, Molatselli P. Severe protein energy malnutrition in Lesotho, death and survival in hospital, clinical findings. Trop Geogr Med 1986; 38:351-358.

Van Roosmalen-WiebengaMW, Kusin JA, De With C. Nutritional rehabilitation in hospital. A waste of time and money? Evaluation of nutrition rehabilitation in a rural district hospital in Southwest Tanzania. I. Short term results. J Trop Pediat 1986, 32:240-243.

Van Roosmalen-WiebengaMW, Kusin JA, De With C. Nutritional rehabilitation in hospital. A waste of time and money? Evaluation of nutrition rehabilitation in a rural district hospital in Southwest Tanzania. 11. Long term results. J Trop Pediat 1987, 33:24-28.

Verbeek K. Une etude evaluative du centre de rEhabilitationnutritionnelle de l'hopital protestant de Dabou, Cote d'lvoire. Rapport, Univ. Agr. Wageningen, 1986.

Verkley MTB. Jansen AAJ. A mixed ambulatory-homenutrition rehabilitationprogranune in a rural area in Kenya. East Afr Med J 1983; 60(1):15-21.

Whittaker DE, Le Roux I, Disler P. The cost effectiveness of the Philani nutrition day centre in Crossroads squatter camp, Cape Town. S Afr Med J 1985; 68:174-176.

Whyte SR, Yambo M. Vestgaard J, Michaelsen KF, Vagnby B. Evaluation of the family life training programme. Ministry of culture and social services, Kenya. DANIDA, file n°104. El, Ken.791Eval., 1989.

168 Gmwth monitoringand promotion Growth monitoring and promotion

Alnwick D, Moses S, SchmidtOG, eds. Improving young child feeding in Eastem and Southern Africa. Household-levelfood technology. Proceedings of a workshop held in Nairobi, Kenya, 12-16 October 1987. IDRC, 1988.

APHA. Growth monitoring. InternationalHealth Programmes series. Washington, D.C., AmericanPublic health Association, 1981.

Armar-Klemesu MA, Wheeler EF, Brakohiapa LA, Yamamoto S. Infant feeding and growth in ral Ghana: is the use of the traditional fermented porridge a case for early supplementation ? J Trop Pediatr 1991; 37: 111-115.

Bac M. Progress towards health for all in the Gelukspan health ward. Thesis; Medical University of Southern Africa, 1985.

Badonte ST, Johnson-WelchC. L'amelioration de la surveillancede la croissance. Meres et Enfants 1990; 9(3):4-5.

Beaton G, Kelly A, Kevany J, Martorell R, Mason J. Appropriate uses of anthropometric indices in children. ACC/SCN State of the Art Series; Nutrition Policy Discussion Paper No 7. Geneva, United Nations, 1990.

Berg A. Can nutrition efforts make a difference ? in: Malnourished people, a policy view. Poverty and Basic Needs Series, June 1981. Appendix C:73-92.

Black R.E. Would control of childhDod infectious diseases reduce malnutrition ? Acta Paediatr Scand 1991; Suppl 374:133-140.

Briend A, Dyckwewicz C, Graven K, Mazunder RN, Wojtyniak B, Bennish M. Usefulness of nutritional indices and classifications in predicting death of malnourishedchildren. Br Med J 1986; 293:373-375.

Briend A, Bari A. Critical assessmentof growth monitoring for identifying high risk children in prmary health care programmes. Br Med J 1989; 298:1607-1611.

Briend A. Is diarrhoea a major cause of malnutrition among the under-fives in developing countries ? A review of available evidence. Eur J Clin Nutr 1990; 44:611-628.

Carlson BA, Wardlaw TM. A global, regional and country assessment of child malnutrition. Staff working papers No 7; New York, UNICEF, 1990.

Ceplanut. Proc6dures d'evaluation rapide. Programme de suivi et de promotion de la croissance. R6publiquedu Zaire, Centre de planificationNutritionnelle, UNICEF Zaire/ Bureau de l'Evaluation, NY. 1990.

169 Bibliography

Chen LC, Chowdhury AKMA, Huffman S. Anthropometric assessment of energy-protein malnutrition and subsequent risk of mortalityamong preschool aged children. Am J Clin Nutr 1980; 33:1836-1845.

Cole TJ, Paul AA, Eccles M, Whitehead RG. The use of a multiple growth standard to highlight the effects of diet and infection on growth, pp. 91-100; in: 88. Perspectives in the science of growth and development, J.M. Tanner ed., London, Smith-Gordon 1989.

Cole-King SM. Under-fives clinic in Malawi. The development of a national programme. Environ Child Hlth 1975; 21(4):183-191.

Coly M. Enquete sur la surveillance nutritionnelle dans les communautes rurales de Gniby et Boulel (Departement de Kaffrine). Institut de Sante et D6veloppement, Univ. Cheikh Anta Diop, Dakar, Senegal, 1989.

Comu A, Delpeuch F, Simondon F, Tchibindat F, Faucon LD, MassambaJP, Goma I, Olivola D, et al. Enqu&tenationale sur l'etat nutritiomneldes enfants d'ige pr6scolaire au Congo. Coll. Etudes et Theses; Paris, Edit. de l'ORSTOM, 1990.

CRS/Caritas. Growth monitoring and nutrition education. Washington, U.S. AID, Bureau for Science and Technology, Office of Nutrition, February 1988.

Delahaye P. Introduction des fiches de croissance en Angola. Quelques aspects de programmation. Carnets de l'Enfance 1983; 61/62:271-286.

Dickin KL, Brown KH, Fagbule D, Adedoyin M, GittelsohnJ, Esrey SA, Oni GA. Effect of diarrhoea on dietary intake by infants and young children in rural villages of Kwara State, Nigeria. Eur i Clin Nutr 1990; 44:307-317.

Diene M. Le progranmmede rehydratation par voie orale et les activites de nutrition dans les regions de Kaolack et Fatick. Rapport de visite; Dakar, SANAStDHPS 1989.

Editorial. Growth Monitoring: intermediate technologyor expensive luxury ? Lancet, 1985;2:1337-1338.

Editorial. Diarrhoea and malnutrition. Lancet 1991; 338:921-922.

El Samani EFZ, Willett WC, Ware JH. Association of malnutrition and diarrhoea in children aged under five years. Am J Epidemiol 1988; 128:93-105.

Garenne M, Maire B, Fontaine 0, Dieng K, Briend A. Risques de deces associes i diff6rens Etats nutritionnels chez l'enfant d'age pr6scolaire. Dakar: Rapport ORSTOM-ORANA; 1987.

Genece E, Rohde J. Growth monitoring as an entry point for primary health care. Ind J Ped, 1988; 55 (suppl. 1):S78-S83.

170 Growth monitoring And promontion

Gerein NM. An evaluationof growth monitoringin Zaire. Ph.D. Thesis, Fac. Med., Univ. London, 1988.

GoldenMHN. The roleof individualnutrient deficiencies in growthretardation of childrenas exemplified by zinc and protein, pp.143-160;in: Linear growth retardationin less developedcountries, J.C. Waterlowed., Nestl6Nutrition Workshop Series Vol. 14, 1988.New York, RavenPress and Vevey, Nestl6.

Gopalan C., Chattejee M. Use of growth charts for promotingchild nutrition. A review of global experience.Nutrition Foundation of India. SpecialPublication Series N°2, Delhi, 1985.

Governementof Togo, CRS/Togoand LogicalTechnical Services. Improving mothers' participationin growthmonitoring and promotion.Final ProjectReport. Bethesda,Pricore; 1990.

Grantham-McGregorS. PowellC, WalkerS. Nutritionalsupplements, stunting, and childdevelopment. Lancet 1989;2:809-810.

Greaves P, Hendrata L. Breastfeedingand growth monitoring. Int J Gynecol Obstet 1990; 31(Suppl.1): 121-123.

Griffiths M. Growthmonitoring: making it a tool for education.In: Growth monitoring:information packet.Washington, DC. AmericanPublic Health Association, Clearinghouse on InfantFeeding and MaternalNutrition, 1987a.

GriffithsM. Le graphiqueA bulles. Mtres et Enfants1987b; 6(1):7.

GriffithsM, Berg A. The bubblechart: an updateon its development.Food Nutr Bull 1988; 10:71-74.

GwatldinDR, WilcoxJ, Wray J. Can health and nutritioninterventions make a difference,monograph no. 13, WashingtonD.C., OverseasDevelopment Council, 1980.

Hendrata L, Rohde J. Ten pitfalls of growth monitoring and promotion. hId J Pediatr 1988; 55(Suppl):S9-S15.

Henry FJ, BriendA, CooperE. Targetingnutritional interventions: is there a role for growthmonitoring? Health Pol Plan 1989; 4:295-300.

Huttly SRA, Victora CG, Barros FC, Texeira AMB, VaughanJP. The tining of nutritionalstatus determinations:implications for interventionsand growthmonitoring. Eur J ClinNutr 1991;45:85-95.

1SS (IstitutoSuperiore di Sanita).Joint WHO-UNICEFNutrition Support Progranme (JNSP) Global EvaluationFinal Report. Vol. 1 & 2. Rome,7 december1989.

KasongoProject Team. Anthropometncassessment of youngchildren's nutritional status as an indicator of subsequentrisk of dying. J Trop Pediatr 1983;29:69-75.

171 Bibliograhy

Kasongo Project Team. Growth decelerations among under-5-year old dhildren in Kasongo (Zaire). H. Relationships with subsequent risk of dying and operational consequences. Bull WHO 1986; 64: 703-709.

Kielman AA, Mc Cord C. Weight-for-age as an index of risk of death in children.Lancet 1978; 1:1247-1250.

Lifanda K. Consultancy report on Save the Children health and nutritionprogramme. Burkina Faso field office, May 1988.

Lotfi M. Growth monitoring: a brief literature review of current knowledge. Food Nutr Bull, 1988; 10 (4):3-9.

Lotfi M, Mason J. Direct intervention programnnesto improve infant and child nutrition. ACCISCN; Draft 4 Oct. 1989. Gen6ve, Nations Unies.

Lovel H, De Graaf J, Gordo G. Commentles meres mesurent la croissance de leurs enfants. Dimensions wommunautairesde l'expension des fiches de croissance au Ghana. Carnets de l'Enfance 1984; 65/68:303-322.

Litter CK, MoraJO, HabichtiP, RasmussenKM. Robson DS, Herrera MG. Age-specificresponsiveness of weight and length to nutritional supplementation.Am J Clin Nutr 1990; 51:359-364.

Maire B, Garenne M, Briend A, Dieng K, Fontaine 0. Weight change and risk of dying in West-African children. Communic.at the 14th InternationalCongress of Nutrition. Aug. 20-25, 1989, seoul, Korea.

Maire B. Prevalence de la malnutrition du jeune enfant: du probabiliste au rtel. These; Universit6 de Montpellier H, 1990.

Mc Guire JS, Austin JE. Beyond survival: children's growth for national development. Assignment children 1987: 141.

Mensah PPA, Tomkins AM, Draser BS, Harrison TJ. Effect of fermentationof Ghanaian maize dough on the survival and proliferation of four strains of shigella flexneri. Trans Roy Soc Trop Med Hyg 1988; 82:635-636.

Moneti F, Yee V. Mobilisationen faveur dc la nutritikn; resultats de l'exp6rience d'Iringa. Mtres et Enfants 1989; 8(2):1-3.

Morley D, Parent M. Le chemin de la sante. Commentsuivre la croissance de l'enfant. Bureau d'Etudes et de Recherchespour la Promotion de la Sante, Kangu-Mayombe,Republique du Zaire; 1983.

Morley D. A health and weightchart for use in developingcountries. Trop Geog Med, 1968; 20:101-107.

172 Growthmonftoring and promotion Morley D. The spread of comprehensivecare throughunder-fives' clinics. Tras Roy Soc Trop Med -Hyg, 1973; 67 (27: 155-170. " ; .

Moy RJD, Choto R, BoothIW, Mc NeishAS. Diarrhoeaand catch-upgrowth. Lancet 1990; i:600.

M.S.A.S./Unicef/Orstom.Snainaire-Atelier sur les bouilliesde sevage en AfriqueCentrale. Minist&re de la Sante et de l'ActionSociale, Bcazzaville, 21-24 mai 1991.

NabarroD, ChinnockP. Growthmonitoring - inappropriatepromotion of an appropriatetechnology. Soc Sci Med, 1988; 26 (9):941-948.

Niger Ministere de la Sant6AEDIH. 1992. REpubliquedu Niger: Evaluon du projet de communicationsociale en vitamineA. D6cembre.

NyiribibiM. Suivi de la croissancedes enfantsd'Age pr6scolaire: expErience du Rwanda. Ateliersur la surveillancealimentaire et nutritionnelle,pays francophonesde l'Afrique Centrale et de 1'Est. Kinshasa,N'sele, 4-15 Juin 1990.

OMS rWHO).La fichede croissance;son utilisation pour les soins amxnourrissons et aux enfants.OMS, Genave, 1986.

OMS (WHO). Soins de sante maternelleet infantile: Fiche intemationalede croissance.Directives h l'usage des agentsde soins de santEprimaires. OMS, Genbve, 1978.

Pagezy H, SubervieAM. R6flexionssur une interventionnutritionnelle en Afrique centrale(Zaire). Laboratoired'Ecologie Humaine, Aix en Provence,1990.

RohdeJ.E. Surveillancede la croissance.Sante du Monde,1984; Oct.,8-10.

RowlandMGM; Cole TJ, WhiteheadRG. A quantitativestudy into the role of infectionin determining nutritionalstatus in Gambianvillage children. Br I Nutr 1977;37:441-450.

RuelMT, PelletierDL, HabichtJP, MasonJB, ChobokoaneCS, MarupingAP. Comparisonof mother's understandingof two child growthcharts in Lesotho.Bull WHO 1990;68(4)-483-491.

Ruel MT, Pelletier DL, HabichtJP, Mason JB, ChobolkaneCS, MarupingAP. Comparisonof two growthcharts in Lesotho:health workers' abilityto uaderstandand use themfor action.Am I Public Health 1991;81:610 616.

SauerbornR, NougtaraA, SorghoG, BidigaJ, DiesfeldHI. Assessmentof MCH servicesin the district of Solenzo,Burkina Faso. m. Effectivenessof MCH servicesin detectingof and caring for mothers and childrenat risk. J Trop Pediatr 1989;35:l-17.

Serdula MK, Aphane JM, Kunene PF, Gania DM, StaehlingN, Peck R, Seward J, SullivanB, TrowbridgeFL. Acuteand chronicundenutriton in Swazland.J Trop Pediatr 1987a;33:35-42.

173 Bibliography

Serdula MK,Herman D, WilliamsonDF, Binkin NJ, Aphane JM, Trowbridge F. Validity of clinic-based nutritional surveillance for prevalence estimationof undernutrition. Bull WHO 1987b; 65:529-533.

SimondonKB, SimondonF, Cornu A, DelpeuchF. The utility of infancy weight curves for the prediction of linear growth retardation in preschool children. Acta Paediatr. Scand., 1991; 80(1):1-6.

Smcdman L, Sterky G, Mellander L, Wall S. Anthropometry and subsequent mortality in groups of children aged 6-59 months in Guinea-Bissau.Am I Clin Nutr 1987; 46:369-373.

Steveny J. Standardizedinterpretation of under-5s weight curves. Trop Doct 1982; 12:133-135.

Tomkins A. Nutritionalstatus and severity of diarrhoea amongpreschool childrenin rural Nigeria. Lancet 1981; 1:860-862.

Tomkins AM. Nutritionalstatus and risk of morbidityamong young Gambian clildren allowing for social and environmentalfactors. Trans R Soc Trop Med Hyg 1989; 83:282-287.

Tremlett G, Lovel H, Morley D. Guidelines of the design of national weight-for-age growth charts. AssignmentChildren 61162, 143-175.

UNICEF. State of the world's children 1984. Oxford: Oxford UniversityPress, 1983.

UNICEF. Assessment of growth monitoring and promotion (GMP) activities in Indonesia. New York, Evaluation Office, UNICEF, 1990.

UNICEF. Evaluationof Growth Monitoringand PromotionProgramMes, Worlkhop Report, Nairobi 7-9 may 1992, Evaluation Office, UNICEF, july 1992.

Shorr [I, Ucellani V, Ferrera D. . Promising PVO strategies in growth monitoring and promotion. A report on a workshop sponsored by the Nutrition CosmmunicationProject of the Acaatmy for EducationalDevelopm^ it in coordinationwith the Office of Nlutritionof A.I.D. Washington, D.C., June 15-17 1989.

Van Lerberghe, W. Growth, infection and mortvity: is growth monitoring an efficient screening instrument? pp.101-109, in: Auxology 88. Perspectives in the science of growth and development, J.M. Tanner ed., London, Smith-Gordon, 1989.

Van Loon H. Epidemiology of malnutrition in developing countries. Thesis, Katholieke Universiteit Leuven, 1987.

VelemaJP, AlihonouEM, Gaudaho T, Hounye FH. Childhood mortalityamonfg users and non-users of primary health care inia rural West African communiy. Int J Epidemiol 1991; 20:474-479.

Whitehead RG, Paul AA. Growth charts and the assessment of infant feeding practices in the western world ans in developing countries. Early Hum Dev 1984; 9:187-207.

174 Growth monistorfngand promotion

WHO. Guidelines for training conmmunityhealth workers in nutrition. 2nd edition; Geneva, WHO 1986.

Wilcox JR, Teller CH, Horwitz A. Making primaqy health care nutrition work: issues raised in a review of the record. Iternational Nutrition Plarmers Forum Conference on Nutrition in P.H.C., Cairo, Egypt, Jan. 16-19 1984.

Yamnbi0. Nutritionalstatus and the risk of death: a prospectivestudy of children six to thirty months old in Iringa Region, Tanzania. PhD Thesis, Cornell University, USA, 1988.

Yee V, Zerfas A. Issues in growth monitoring and promotion. LTS/InternationalNutrition Unit. In: Growth monitoring: information packet. Wasington, DC. American Public health Association, Clearinghouseon Infant Feeding and Maternal Nutrition, 1987.

Yee V. Suivi de la croissance. L'experience des organisationsb6n6voles privEcs. M&reset Enfants 1989; 9(1):6-7.

Zerfas A.l. Anthropometricfield methods: general.pp. 339-364. In: Nutrition and Growth, New York, PlemnmPress, 1979.

175 Bibliography

Nutrition education

Andrien M. Evaluation du programme d'dducation nutritionnelle du Centre Hospitalier Regional de Bouak6. Service d'Evaluationdu Ministere de l'EnseignementPrimaire et de I'Education Televisuelle dz Cote d'Ivoire; Laboratoire de PEdagogieexp6rimentale de Liege, Belgique, mars 1983.

Andrien M. Limites de l'education nutritiomelle conventionnelle:un exemnpleen milieu urbain d'Afrique de l'Ouest. Hygie 1986; V(4):21-25.

Andrien M, Lamotte P. L'education nutritionnelledes meres d'enfantc en bas ige: etude des besoins A Pikine (Senegal). Liege, Laboraloirede Pedagogie Exptrimentale, CEERES, Coll. Sante Education No 3; 1986.

Aphane J. New research methods to identifyproblems and possiblesolutions. In: Proceedingsof the 14th InternationalCongress of Nutrition. Seoul, Korea, August 20-25, 1989: 596-600.

AustinJE, et al. Annotated Directory of NutritionPrograms in Developing Countries. Cambridge, MA, Harvard Institute for InternationalDevelopment; 1978.

AustinJE, BeldihgTK, Brooks R, Cash R, Fisher J, Morrow R, Pielemeyer N, Pyle D, Wray JD, Zeitlin MF. Case study of the cost and effectivenessof preschoolclinics in Ghana and Lesotho. Chap. Four, in: Integrated Nutrition and Primary health care programs. Vol.5, Study VII, Nutrition Intervention in Developing Countries. Cambridge, MA, Oelschlager.Gunn and Main, 1981.

Barnes-Kalunda,S. Evaluationof the mass media componentof AID's nutrition planningproject in Zaire. Consultant report; Newton, MA, InternationalNutrition Communication Service, 1981.

Berg A. Fear of trying. J Am diet Assoc 1976; 7:279-298.

Burgess A. Evaluationof nutrition interventions An annotatedbibliography and review of methodologies and results. 2nd edition; FAO Food and NutritionPaper, no 24. Rome, FAO 1982.

Child to child. London, Instituteof Education, 1990.

Dillon JC. Evaluation du projet NUTED; rapport de mission. Brazzaville, Congo, UNICEF 1990.

Fakambi LK. Factors affecting the nutritional status of mothers: the food and nutrition program of the Ouando horticulture and nutrition center in the People's Republic of Benin. Matemal nutrition and health care program, Research Report Series no 1 l. Washington, InternationalCenter for Research on Women, 1990.

Gilmore JW, Adelnian CC, Meyer AJ. Thorne MC. Morocco: food aid and nutrition education. AID project impact evaluation report N48, 1980.

176 Nutrition education

Griffiths M. The role of education and training in improving child feeding practices,pp.605-607; in: Proceedings of the 14th Intemational Congress of Nutrition, Vol 1, ed. by K.W.Young et al, Seoul, Korea, 1989.

Gussow J.D., Contento 1. Nutrition education in a changing world. A conceptualizationand selective review. Wld Rev Nutr Diet 1984; 44:1-56.

Healithcom,Communication for Child Survival. The mass media and health practices: evaluations in Honduras and the Gambia: Summary report of the major findings. A report to US AID prepared by Applied CommunicationTechnology, Academy for Educational Development, Sept.1985.

Healthcom, Commuunicationfor Child Survival. Lessons from five countries: Honduras, The Gambia, Swaziland, Ecuador, Peru. Academy for EducationalDevelopment, Stanford University, Agency for InternationalDevelopment, October 1985.

Hollis C, Seumo E, Mal Bappa A, Griffiths M. Improving young child feeding practices in Cameroon : project overview. The Weaning Project. MOH Cameroon, CARE/Cameroon, Education Development Center inc., The Manoff Group. Oct.1989.

Hoorweg J, McDowell I. Evaluation of nutrition education in Africa. Community research in Uganda, 1971-1972. Afrika Studie Centrum, Leiden. The Hague, Mouton Publ. 1979.

Hoorweg J, Niemeijer R. Intervention in child nutrition. Monographs from the african studies centre, Leiden. London, KPI, 1989.

Hornik R.C. Nutrition Education: a State-Of-The-Art Review. ACCISCN State-Of-The-Art Series; NutritionPolicy DiscussionPaper No.1. ACC/SCN, c/o FAO, Rome, 1985.

Hornick R, Sankar P, Huntington D, Matsebula G, Mndzebele A, Magongo B. Comnmnication for diarrheal disease control: Swaziland programil evaluation 1984-85. Annenberg School of Communications, Philadelphia, Univ. Pennsylvania 1986.

Huntington D. The CARE/Congonutrition educationproject. In: Hollins C, ed., Using Communications to Solve Nutrition Problems: a Compendium. Newton, Mass, Education Development Center, International Nutrition CommunicationService, June 1986: 120-128.

Israel R.C., Tighe J.P.N. Nutrition education: the state of the art review and analysis of the literature. Nutrition Education Series 1984, issue 7; Division of Scielnce,Technical and Vocational Educaton, UNESCO, Paris France.

Lejeune A., Andrien M. Communication et Nutrition. Naissance d'un r6seau pour l'6ducation nutritionnelleen Afrique (RENA). ACCT, Paris 1986.

1

-' ~~177 Bibliogr hy

Manoff RK. The importane of a social marketingapproach, pp.95-98; in: Nutrition Education:a state-of-the-anreview, Hornick RC, ed. ACCISCNState of the art Series,Nutrition policy discussion paper No 1, clo FAO, Rome1985.

Manoff IntemationalInc. Weaningpractices assow-ment for the extremenorth provinceCameroon. A summary of Findings.The WeaningProject, Report by CARE/Cameroon,MOH/Cameroon and EducationDevelopment Center, July 1987.

ManoffGroup Inc. Theweaning project. Evaluation of the Cameroonweaning project. Tulane University and CAREtCameroon.Oct 1989.

McSweeneyBG, Freedman M. Lack of time as an obstacleto women's education. The case of Upper-Volta.Comp Educ Rev 1980;24(2):S-124/S-139.

MOHGhana; Nutrition Division. Inprovingyoung child feeding practices in Ghana;5 vol; Washington, US-AID,RESDO, 1989.

MorleyD. Le programnede l'Enfantpour l'enfant.Carnets de l'Enfance, 1979;47/48:185-195.

NationalNutrition Council Swaziland I UNICEF.Project for promotionof improvedyoung child feeding. 4 vol;1988.

PagezyH., SubervieAM. Rtfexionsur une interventionnutritiounelle en AfdiqueCentrale (Zaire). Le cas de la bouilliede sevrage chcz les Ntombadu Zaire (aspectsanthropologiques). CNRS, Lab. d'EcologieHumaine, Univ. d'Aix en Provence,1990.

Pritech. Probl6matiquede l'utilisationdu mat6riel6ducatif dans les structuresde sante au Mali et au Niger. Programmenational de lutte contreles maladiesdiarrhtiques. Pritech, 1990.

RasmusonM, SeidelR, GabiddonH. Dietarymanagement of diarrhea:the Gambianexeperience. J. Nutr. Educ., 1990;22(1):15-23.

SchneidemanI, BennettFJ, RutishauserIHE. The nutrition rehabilitationunit at Mulago Hospital, Kampala;development and evaluation,1965-1967. J Trop Pediatr 1971;17 Monograph13:25-34.

Schurch B, Wilquin L. L'education nutritionnelledans les conunautes du Tiers-Monde. Urn bibliographieannotEe. Fondation Nestde pour l'etudedes probltmesde l'alimentationdans le monde. FondationNestlE, Lausanne, Suisse; 1982.

SchuirchB. Evaluationof Nutrition Educationin Third World Communities.Nestle Foundation publicationSeries No 3, Hans Huber Publishers;Bern, Suisse; 1983.

Stanfield J.P.. Nutrition educationin the conuextof early childhood malnutritionin low-resource communities.Proc Nutr Soc 1976;35:131-138.

178 Man tion education

UNESCO. L'6ducation nutritionnelleau cours des ann6es '80. Dossier d'information du BIE No 9. Gen&ve,Bureau Internationald'Education; 1987.

Wane I, De Lauture H, Penot C, Perier-ScheerM, Cohen M et al. Premitre dvaluationd'une acion d'educationnutritionnelle dans la mndinade Dakar. Dakar Med 1981;26(4):420426.

Zeitlin M.F., Formacion C.S. Nutrition Education. Study II, volume 2, NuxtritionIntervention in DevelopingCountries, Oclgeschlager, Gunn and Hain, Cambridge;Mass., 1980.

Zeitlin M.F. Upper Volta case study of home-basedcommunity-level weaning food development.Chap. 4, in: NutritionIntervention in DevelopingCountries; volume 3, studyIV: FormulatedFoods, ed. by J.Heimendinger,M.F. Zeitlin and I.E. Austin. Oelgesehlager,Gunn and Hain; Cambridge,Mass., 1981.

179 NOTES

1. This reviewfocuses on programsaimed at preschool-agechildren and pregnantwomen: it does not address emergencyfoodaid, schoolfeedingprograms,foodforification.foodfor workT dprograms,orhospital and nutrition rehabilitationcenter food supplementprograns. These aspects,however, are sometimesintegrated into a few programs.

2. Nutritioneducation within rhefoarnaframeworkof a schoolwill not be addressedhere; although it is certainly an imponrtantarea, no assessmentsare curntly avaiable thkatwould illustrate how this typeof instructionoperates, its cost, and its impact. Educationalaspects of specificprograms such as thosefor iodinedeficiency or vitaminA will not be mentioned:the greatestamount of attentionwill be paid to instructionprovided to motherswithin the frameworkof preventingprotein-energy malnurtrition in young children.

3. Manysmall-scale programs in Colombia(the Promotora h Candlaria Programserving 920famiies), Thailand, and Jamaica(serving a populationof 60,000people) (Hornik,1985, pp. 42-43) managedto improvethe nutrition situationin this way. BroadlyeVending coveragethrough such a system, however,is difficult.

4. For example,in Zaire in 1980, in a largenutrition education project that usedmass media (radio,relevision), an initialsurvey was indeed conductedof the knowledgeand behaviorof mothersand the listeningaudiencss of radioprograms, but the programwas then designedbehind closed doors at the centralgovernment level, without any prior testingof the messagesor assessmentof theirpotential impact. Infact, a closer anabsis of the inital surveyrevealed that livingconditions for themothers (littleaccess to food, lack of fuel, lack of time) were more likely to be the issue than a lack of knowledge(I'andberg). n. Theparticipation of the targetpublic informulatingthe messagesturned out to be a good way to comeup with relevantmessages and to ide4ify acceptablebehaviorsjfor the population.

6. Thesuccess of the programin Indonesia(Manoff, 1985), basedon iheuse of the mediaand communityhealth care workerskwwn as kaders, is basicaly attributableto the quaity of superviion and its appropriateness, allowingoperanonal difficulties related to the targetedgoals to be continuallycorrectd Theprogram objectives shouldbe preciselydefined to guide in the developmentof a strazegyand to allowthe assessmentand follow-up to be conducted.

7 The message recommendedrhe adoption of a corn-flourbased pap, called 'energyfl in the course of broadcastingthe message,the necessarysuppkments were gradually lft out. In thispopulation,the trnsition used to be made directlyto thefamily diet based on cassava,supplemented by green leafy vegetables,fish, and wild gane. After several years, corn cultivationwas broadly extendedinto the region, contribudngto alcohol's manufacture;mothers increased their work load with longcomn pounding sessions, which resulted in children'snot receivingpap on a regularbasis; weaningfoods werefrequently modified without n eviden benfit to the children,

180 The WorldBank Africa Technical Department would like to knowif this publicationis of interestto you, and whetheryou will use its informationin your work. Pleasetake a momentto completethis questionnaireand returnit to our department.It willguide future publication decisions. Thankyoul

Is this publicationrelevant to your work? o very relevant o somewhatrelevant ni not relevant,but of personalinterest O neitherrelevant nor of personalinterest

Can you see a use for this publicationin your work (checkany that apply) a policyor programplanning o referencefor articles,speeches, or discussions o workshopor seminartopic O sharewith colleagues O library o personal use o other

Is the way the informationis presentedin this publicationuseful to you? o yes o no - I wuuldprefer to see this information

Wouldyou like additionalcopies of this publicationto give to othersin your field? o yes o how many? o no

Pleaseindicate the main focusof your day-to-daywork:

Whattopics relating to your work wouldyou like to see coveredin other puiblications?

Othercomments:

(Mailing instructionsare on the badc of this sheet) TO: Tonia Marek The World Bank 1818 H Street, N.W. Washington, D-C. 20433

Please fold this form, staole it and Your name return it via air mail, or place it in an envelopeand send it to Tonia Marek Your organization at the World Bank. You may also fax it to her at (202)473-8216 or send a Your address message via intemet to [email protected].

Thank you for your cooperation Country

Fax

EMai i I