Interventions for physical growth and psychological ...

DEPARTMENT OF CHILDAND ADOLESCENT HEALTHAND DEVELOPMENT

WORLD HEALTH ORGANIZATION

WHO/CHS/CAH/99.3ORIGINAL: ENGLISHDISTR.: GENERAL

A CRITICAL LINKInterventions for physical growthand psychological development

A REVIEW

CH

ILD

AN

D A

DO

LESC

EN

T H

EA

LTH

AN

D D

EV

ELO

PM

EN

T■

■■

■■

■■

■■

■■

■■

■■

■■

■■

■C

AH

DEPARTMENT OF CHILDAND ADOLESCENT HEALTHAND DEVELOPMENT

WORLD HEALTH ORGANIZATION

WHO/CHS/CAH/99.3ORIGINAL: ENGLISHDISTR.: GENERAL

CH

ILD

AN

D A

DO

LESC

EN

T H

EA

LTH

AN

D D

EV

ELO

PM

EN

T■

■■

■■

■■

■■

■■

■■

■■

■■

■■

■C

AH

A CRITICAL LINKInterventions for physical growthand psychological development

A REVIEW

© World Health Organization, 1999

This document is not a formal publication of the World Health Organization (WHO),and all rights are reserved by the Organization. The document may, however, be freely reviewed,abstracted, reproduced and translated, in part or in whole, but not for sale nor for usein conjunction with commercial purposes.

The views expressed in documents by named authors are solely the responsibility of those authors.

Editor: Mandy MikulencakCover photo: WHO/PAHO Carlos GaggeroDesigned by minimum graphicsPrinted in Switzerland

Acknowledgements vii

Executive Summary 1

Chapter 1. Introduction 4

Benefits of investing in early childhood care and development programmes 4

Benefits of investing in programmes to improve child physical growth 5

Benefits of investing in programmes that combine ECCD and nutrition interventions 5

Interaction at three critical points 5

The child 6

The child and family 6

Design and delivery of programmes 6

Questions to be addressed by the review 6

Chapter 2. Background and methodology for the review 8

Measuring impact: the selection of outcomes measures 8

Outcomes related to child development 8

Long-term measures of individual outcomes 9

Outcomes related to child growth 9

Theoretical background for the review 10

Steps or process by which interventions are developed 11

Dimensions of research rigor to consider in evaluating interventions 11

Degree of control: adequacy of delivery and evaluation design 12

Context and conditions of the intervention 13

Level of intervention: where in the causal pathway does the intervention occur? 13

Chapter 3. Interventions to support psychological development 15

Impact of ECCD interventions on psychological development 15

Child-focused interventions that provide psychosocial stimulationdirectly to the child 16

Parent-focused interventions to improve parenting skills and psychosocialstimulation in the home environment 17

Joint-focused interventions to improve parenting skills, and to providepsychosocial stimulation to children and supportive services to parents 18

Effectiveness of ECCD programmes 19

Impact of nutrition interventions on psychological development 20

Improving nutrition through supplementary feeding 20

Mechanisms for effects of malnutrition on development 20

iii

Contents

Timing and duration of supplementation 21

Activity levels and behavioural responsiveness 21

Issues in defining mechanisms 22

Rehabilitation of severely undernourished children 22

Correction of micronutrient deficiencies 23

Iodine 23

Iron 24

Zinc 25

Other micronutrients 25

Effects of breastfeeding on psychological development 25

Nutrition education and growth promotion 26

Chapter 4. Interventions to support physical growth 27

Impact of nutrition interventions on physical growth 27

Supplementary feeding of pregnant and lactating women 27

Supplementary feeding programmes for children less than 5 years 28

Community-based efficacy trials 28

Food supplementation programmes 29

Rehabilitation and feeding of undernourished children 29

Correction of micronutrient deficiencies 29

Iodine 30

Iron 30

Vitamin A 30

Zinc 31

Nutrition education to improve breastfeeding and complementary feeding 32

Nutrition education and physical growth 32

Nutrition education and feeding practices 33

Production of complementary foods for young children 33

Physical growth monitoring and promotion 34

Impact of other types of interventions on physical growth 35

Interventions to control disease 35

Improved water and sanitation 35

Control of disease through medical services and immunization 35

Interventions to increase maternal education 36

Interventions to increase household food security 36

Food subsidies 36

Food for work 36

Agricultural production 36

Credit programmes 37

Effectiveness of nutrition programmes 37

Chapter 5. Combined physical growth and psychological development interventions 39

Efficacy trials of combined physical growth and psychological developmentinterventions 39

Cali, Colombia 39

A CRITICAL LINK: INTERVENTIONS FOR PHYSICAL GROWTH AND PSYCHOLOGICAL DEVELOPMENT

iv

Bogota, Colombia 40

The Jamaica Study 41

A combined intervention for severely malnourished children 43

A combined intervention for children with non-organic failure-to-thrive(NOFTT) 43

Summary of evidence for combined effects from efficacy trials 44

Effectiveness evaluations of combined physical growth and psychologicaldevelopment programmes 44

Summary of additive effects from effectiveness and efficacy trials 45

Levels where interventions have an additive effect 45

The child 45

The child and family 46

Design and delivery of programmes 46

Implications for models of combined programmes 46

Programmes that provide centre-based pre-school education plus feeding 48

Programmes that provide food and stimulation to children, and educationon nutrition and responsive parenting for parents 49

Programmes that combine nutrition and education on responsive parenting 49

Other approaches 50

Programmatic issues 52

Potential programmatic advantages 52

Delivering interventions to those who are most likely to benefit 52

ECCD as an entry point and/or motivating factor for programmeparticipation 53

Co-ordination of messages, materials and approaches 53

Potential operational constraints to combined programmes 54

Chapter 6. Summary, conclusions and recommendations 56

Purpose of the review 56

Key questions addressed by the review 56

To what extent and through what means can psychological functioningbe improved for children living in disadvantaged environments? 56

To what extent and through what means can child physical growth beimproved in settings where chronic undernutrition is prevalent? 57

To what extent and through what means can nutrition and psychologicaldevelopment be improved simultaneously through combined health,nutrition, and psychosocial interventions? 58

Are there effective models for combined interventions, and are thesemodels feasible for implementation on a public health scale? 58

Examples of programme models that can incorporate nutrition, health andpsychosocial components 59

Conclusions 59

Recommendations 60

Annex. Seven Programme Models 62

1. Integrated Child Development Services (ICDS), India 62

2. Head Start, U.S.A 63

v

CONTENTS

3. PANDAI (Child Development & Mother’s Care) Project, Indonesia 64

4. PRONOEI, Peru 65

5. Programa de Alimentacao de Pre-escolar (PROAPE), Brazil 66

6. Hogares Comunitarios de Bienestar (Homes of Well-Being), Columbia 67

7. Integrated Programme for Child and Family Development (IFBECD),and Family Development Programme (FCP), Thailand 68

References 70

Figures Figure 1. Steps in the research Process 11

Figure 2. Dimensions of focus in early childhood care and development programmes 15

Tables Table 1. Cross-cutting dimensions to consider in the interpretation of impact evaluations 12

Table 2. Three combined physical growth and psychological development efficacy trials 40

Table 3. Efficacy studies of combined interventions 42

Table 4. Delivery mechanisms of combined nutrition and psychological developmentprogramme models in the review 45

Table 5. Evaluations of effectiveness of combined programmes 47

Table 6. Options for combined health, nutrition, and psychological developmentprogrammes for children from birth to age 3 years 51


vi

Acknowledgements

The authors of this review were Dr Gretel Pelto (Cornell University), Ms Katherine Dickin (Cornell Univer-sity), and Dr Patrice Engle (California Polytechnic State University, San Luis Obispo). Each has taken majorresponsibility for this project at its various stages.

The helpful suggestions and contributions to sections of the document by the following people are grate-fully acknowledged: Dr Jean-Pierre Habicht (Cornell University), Dr David Pelletier (Cornell University),Dr Maureen Black (University of Maryland, Baltimore), Dr Mercedes de Onis (WHO), Dr Ernesto Pollitt(University of California, Davis), Dr Lida Lhotska (UNICEF New York), and Dr Linda Richter (Universityof Natal, South Africa). Also appreciated are the efforts of Ms Maria Blaze-Kabitzer in checking the refer-ences.

Valuable assistance in reviewing the paper was provided by Dr Jane Lucas and Dr José Martines in theworking group on Family and Community Practices in the WHO Department of Child and AdolescentHealth and Development. The special contributions of Mandy Mikulencak’s editing and Sue Hobbs’ designare evident in the final document.

vii

Executive Summary

and psychosocial support for children’s develop-ment require similar skills and resources fromcaregivers. Research shows that parental feedingbehaviours can influence the quality and quan-tity of children’s dietary intake. Parental psycho-social support, such as providing learningmaterials or conversing with young children, alsosignificantly affect development. These two kindsof behaviours are closely related. For example, aparenting skill that could influence both feedingand psychosocial support is noticing attempts ofchildren to communicate, and responding in away that is appropriate to their needs and devel-opmental level. A parental resource that affectsboth feeding and psychosocial support is suffi-cient time and flexibility for the caregiver to pro-vide the support when it is needed.

In principle, combined growth and develop-ment interventions that help families practise “re-sponsive parenting” have the potential to promotebetter psychological development, as well asphysical growth. Combined growth and develop-ment interventions may also permit greater effi-ciency in the delivery of services to aid childrenand their families, such as through existing childhealth programmes.

Review findings

This review summarizes three types of interven-tions:

● Interventions (psychosocial or nutrition) tosupport psychological development.

● Nutrition interventions to support physicalgrowth.

● Combined interventions to improve bothgrowth and psychological development.

The following critical findings arose from thereview, and provide the basis for advocating morewidescale implementation and testing of com-bined programmes.

■ Psychosocial interventions significantlyimprove psychological development. For

Introduction

The future of human societies depends onchildren being able to achieve their optimal

physical growth and psychological development.Never before has there been so much knowledgeto assist families and societies in their desire toraise children to reach their potential.

This review documents the critical relationshipbetween nutritional status and psychologicaldevelopment, and demonstrates the potential ofcombining interventions that enhance earlychildhood development and those that improvechild health and nutrition into an integratedmodel of care.

In the past several decades, the relationshipsamong health, physical growth, psychologicaldevelopment and parental caregiving have becomeclearer. There is an extensive scientific basis forthe effectiveness of interventions to promotegrowth and to promote psychological develop-ment, even under poor socio-economic andenvironmental conditions. Moreover, when theseinterventions are implemented simultaneously,there is even greater impact. For example, chil-

dren who arebetter nourishedand less lethargicare more curiousand exploratory,and thereforemake better useof opportunitiesto learn how toproblem solveand manipulateobjects. Childrenare also moreable to explore

and learn when parents provide a stimulating en-vironment and are responsive to their children’sverbal and non-verbal cues.

The premise behind interventions to promotegrowth and development simultaneously is thatfeeding behaviours that increase nutrient intake

The premise behindinterventions to promote growthand development simultaneouslyis that feeding behaviours thatincrease nutrient intake and

psychosocial support forchildren’s development require

similar skills and resourcesfrom caregivers.

1

example, children attending pre-school centre-based programmes gain an average of about eightIQ points by the time they are ready to startschool. They are also less likely to repeat primaryschool grades or be placed in special educationclasses. Interventions that provide both directservices to children and support to parents,through parenting education and life skillsdevelopment, have been shown to be the mosteffective. These effects are also seen in large-scaleprogrammes such as the Integrated ChildDevelopment Services (ICDS) in India.

■ Nutrition interventions significantly im-prove psychological development in disadvan-taged populations. The review demonstrates thatincreased intake of nutrients and energy duringthe first two years of life, and prenatally throughsupplements to mothers, have significant positiveimpacts on cognitive and motor development.For example, interventions to prevent iodinedeficiency have dramatic effects on cognitivedevelopment, as well as on the prevention of thephysical stunting that accompanies iodine defi-ciency. Similarly, breastfeeding is associated withimproved cognitive development, and the rela-tionship may be causal.

■ Nutrition interventions significantly im-prove physical growth in poor and malnour-ished populations. Balanced protein and energysupplements during pregnancy improve birth-weight and reduce the incidence of intrauterinegrowth retardation. Food supplementation forinfants and young children has documentedimpacts on physical growth. Other types of effec-tive nutrition interventions include caregivereducation about diets and feeding practices foryoung children, breastfeeding promotion, andzinc supplementation in zinc deficient areas.Programmes that include education, food supple-mentation and/or micronutrient supplementationcan result in reductions in the prevalence ofmoderate and severe undernutrition.

■ Combined interventions to improve bothphysical growth and psychological develop-ment have even greater impact in disadvan-taged populations at risk of malnutrition.Research shows us that caregivers from poorfamilies in developed and developing countriescan acquire new nutrition knowledge, feedingskills, and responsive parenting techniques thatcan improve their children’s nutritional status andcognitive development simultaneously. In com-bined interventions, the psychosocial support

provided by increased stimulation appears to havegreater effects on psychological functioning thanon physical growth, whereas nutritional supple-mentation improves both growth and develop-ment. The combination of stimulation andsupplementation interventions appears to have agreater effect on cognitive development thaneither one alone. These combined nutrition andpsychosocial interventions had significant impactson both growth and development in every studythat tested this relationship. Even large-scaleprogrammes, such as the ICDS Project in Indiaor Head Start in the United States, have shownimprovements in nutritional status or cognitivedevelopment, and some have improved bothgrowth and development.

Having the greatest impact

The review also identified a number of conditions,such as children’s ages and programme charac-teristics, under which the impact on growth anddevelopment is most likely to be seen.

● Interventions during the earliest periods oflife—prenatally, during infancy and early child-hood—are likely to have the greatest impact.Interventions to support psychological devel-opment after this particularly vulnerable earlyperiod, however, are also effective.

● The children in greatest need due to povertyor parents’ lack of knowledge are generally theones who show the greatest response to growthand development interventions. At the sametime, families need a minimum level of re-sources. Identifying the families and childrenwho are mostly likely to benefit from such in-terventions should improve outcomes.

● Growth and development programmes thatutilize several types of interventions and morethan one delivery channel are more efficaciousthan those that are more restricted in scope.Types of interventions include nutritioneducation on diet and feeding practices,providing supplementary foods or micronutri-ent supplements, teaching parents aboutchild development milestones, demonstratingcognitive stimulation activities or other activi-ties to improve parenting skills. Types ofdelivery channels are individual home visits,group counselling, childcare centres and massmedia.

● Combined programmes could be more efficientat delivering services through reductions in


2

delivery costs, less duplication of services, andappropriate identification of those who aremost likely to benefit. There are also likely tobe savings for families as a result of easieraccess when services are combined and fami-lies are more likely to be motivated to seekthose services.

● Programme efficacy and effectiveness appearto be greater when parents are more involved.

Next steps

When discussing the future of combined inter-ventions, a number of challenges arise. Manypotential models for combined interventions topromote physical growth and the psychologicaldevelopment of infants and young children havenot yet been implemented. Others have been im-plemented but not systematically evaluated, whichis essential for the expansion of programmes.However, this review serves as a critical startingpoint in discussing recommendations for imme-diate action.

● We need to develop and test a model of com-bined interventions that could reach a largeproportion of children who are at risk of growthand development faltering. An example is aculturally-adaptable counselling package thatcombines nutrition counselling on comple-mentary feeding (with food supplementation,as necessary) with counselling on psycho-social care (e.g. warmth, attentive listening,proactive stimulation, and support for explor-ation and autonomy). The counselling should

involve the child’s primary caregiving system(e.g. the family) and build on its existingstrengths.

● We should develop and implement new activi-ties to promote appropriate feeding and respon-sive parenting in existing child healthprogrammes. Activities could be incorporatedinto well-child clinics, primary health care con-sultations for childhood diseases, prenatal care,and nutrition programmes, such as growthmonitoring, nutrition education, breastfeedingpromotion, and nutrition rehabilitation centres.

● We should expand and strengthen the health,nutrition, and breastfeeding components ofexisting early childhood care and development(ECCD) programmes. This can be done bothin childcare centres and with parents andcaregivers by providing counselling and train-ing on responsive parenting and appropriatefeeding.

● We need to design an expanded researchagenda to compare and evaluate the effective-ness of different content, programme venues,and delivery channels—for example, breast-feeding promotion through community-basedprimary health care, and delivery by commu-nity health workers, women’s groups, andschool teachers. Training materials for commu-nity workers, monitoring and evaluation tools,and other tools for cultural adaptation, plan-ning, and community participation should bedeveloped together with the counsellingstrategies.

EXECUTIVE SUMMARY

3


4

1

to health. These rights are outlined in the UnitedNations Convention on the Rights of the Child,now ratified by almost all member states of theUnited Nations, which states that children havethe right to “the enjoyment of the highest attain-able standard of health”. The Convention alsosupports States’ identifying the best interventionsfor child health, nutrition, and development.

Strengthening and expanding existing stra-tegies, and identifying and testing creative newapproaches can help ensure that all children haveaccess to the essential requirements for healthygrowth and development.

Benefits of investing in earlychildhood care and developmentprogrammes

Psychological development (mental, motor, socialand behavioural) occurs through maturation anddaily interaction with the environment. When thisenvironment is inadequate, children often do notachieve their full potential for cognitive, social,and behavioural development. Given the rapidrate of mental and motor development in infantsand young children (Condry, 1983), there hasbeen great interest in instituting early interven-tions to provide stimulating, responsive environ-ments that nurture psychological development,and prevent the cumulative deficits often seenamong disadvantaged children. Recent researchsuggests that ECCD programmes for childrenduring their first two or three years of life are morelikely to forestall deficits in learning and psycho-logical development than initiation of interven-tions in the pre-school or school-aged period(Ramey and Ramey, 1998).

ECCD programmes comprise a variety ofinterventions, including early education andsocialisation activities for children, education forparents, and social support for families. Theseprogrammes enhance cognitive development,motivation for learning, and readiness for school(Myers, 1992; Young, 1995), and improve

Survival, growth and development in the earli-est years of life are fundamental for the future

of every individual and for the future of the soci-eties into which those individuals are born. Neverbefore in human history have societies had somuch knowledge to assist families in raisinghealthy children.

However, these crucial formative years remaina time of peril and loss—disease and malnutri-tion not only claim the lives of millions of chil-dren throughout the world, but they also damage

their growth anddeve lopment ,diminish theirquality of life inthe present andc o m p r o m i s etheir future.

This reviewbrings togetherevidence show-ing that nutritionin t e r ven t i onsand early child-hood care andd e v e l o p m e n t

(ECCD) programmes have positive impacts on thephysical growth and cognitive, motor, and behav-ioural development of young children. It providesguidance for designing interventions that stimu-late psychological development and those thatimprove nutrition and child growth, and consid-ers various programme models for doing so.

Thus, the emphasis in the review process wasto identify and compare efficacy trials and pro-gramme evaluations of growth and developmentinterventions, and to interpret the evidence onthe factors contributing to their effectiveness. Thereview builds on the seminal work by Myers(1992) who described combined programmemodels almost 10 years ago.

The review was also undertaken with the viewthat it is important for society to provide chil-dren with the best services to ensure their rights

Chapter

Introduction

A remarkable degree ofconsensus is emerging on theessential requirements thatpositively influence a child’s

early growth and developmentas well as on the ways that

parents and others can provideour youngest childrenwith a healthy start.

Carnegie Task Force (1994)

5

parent-child interaction and family functioning(Olds and Kitzman, 1993; Benasich et al., 1992).

How society benefits from early childhood careand development programmes is difficult toquantify, as such calculations are by nature im-precise and subject to a number of assumptions.However, these calculations provide a good indi-cation of the value of investing in early childhood,such as:

● Increased human resource development (viabetter school achievement).

● Cost-savings and increased efficiency ofprimary schooling (lower rates of grade rep-etition and remedial education).

● Higher educational attainment.● Increased earning potential.● Reductions in juvenile delinquency and its

associated costs.● Increased commitment to marriage.● Increased social mobilisation and community

involvement, and reduced social and economicinequalities in developing countries (Zigler etal., 1992; Barnett, 1996; Schweinhart, Barnesand Weikert, 1993; Myers, 1992).

One early childhood care and developmentprogramme in the United States is a concrete ex-ample of the economic value of such projects. Acost-benefit assessment conducted when theproject participants reached young adulthoodconcluded that the benefits to society outweighedthe project costs more than five fold (Barnett,1985; Barnett, 1995). When participants turned27 years of age, an even greater benefit wasfound—an estimated $7 for each dollar invested(using constant dollars).

Benefits of investing in programmesto improve child physical growth

Physical growth is a very sensitive indicator ofnutritional and health status in infants and youngchildren. Undernutrition is associated with greaterrisks of death (Pelletier et al., 1994), severe infec-tion (Black et al., 1984), and delayed cognitiveand psychomotor development (Lasky et al.,1981; Pollitt et al., 1993). Most growth retarda-tion occurs very early in life. The two periods ofhighest vulnerability are during intrauterine de-velopment and between 6 and 24 months of agewhen the child is making the transition from ex-clusive reliance on breastmilk to consumption ofthe family diet (Brown and Begin, 1993). An in-fant born at low birthweight is at higher risk ofmortality and morbidity (McCormick, 1985) and

a variety of developmental problems (Aylward etal., 1989). Prevention of undernutrition in infantsand young children is critical because growthdeficits are generally not recouped, even withadequate feeding in later years (Martorell et al.,1992; Martorell, 1995).

Growth retardation in early childhood also hasimportant implications for human resource de-velopment through its effects on morbidity, schoolachievement, and work capacity in adulthood.From the perspective of public policy and expen-ditures for social services, the costs of many typesof nutrition interventions are low relative to thepotential benefits and relative to other health in-terventions (McGuire, 1996).

Benefits of investing in programmesthat combine ECCD and nutritioninterventions

Two basic principles outlined in the World HealthOrganization (WHO) Constitution provide astrong argument for designing programmes thatcombine ECCD and nutrition interventions:

● “The healthy development of the child is ofbasic importance.”

● “Health is a state of complete physical, mentaland social well being and not merely theabsence of disease or infirmity.”

A recent World Bank conference reiterated thisperspective: “… health and cognitive interven-tions need to be interactive and designed as inte-grated programmes of nutrition, health, andpsychosocial stimulation” (Young, 1997).

Although there is still much to be learned aboutthe interactions of physical growth, illness andpsychological development, and how they arerelated to family and social conditions, the inti-mate relationship of physical and psychologicalgrowth has been well established. Only recentlyhas the scientific evidence for these powerfulrelationships become available (see recent reviews,such as Engle, Menon and Haddad, 1997;Gorman, 1995; Martorell, 1997; Wachs andMcCabe, 1998; Walker et al., 1998).

Interaction at three critical points

Nutrition and psychological development inter-act at three different critical points: (1) at the levelof the child; (2) between the child and his or herfamily or caregivers; and (3) in the design anddelivery of programmes.

CHAPTER 1. INTRODUCTION


6

The child

The intimate relationship between physicalgrowth and psychological development is particu-larly evident in the first years of life. This helpsexplain why prenatal and early childhood nutri-tion interventions—without a psychosocialcomponent—can also have an impact on psycho-logical development (Martorell, 1997). Likewise,early psychosocial stimulation programmes to im-prove cognition (one aspect of psychologicaldevelopment) may also have effects on physicalgrowth (Super et al., 1990; Martorell, 1997). Themost significant fact, though, is that children whoreceive combined nutrition and stimulationprogrammes perform better than those who re-ceive either type of intervention alone. For bothnutrition and psychosocial interventions, earlierappropriate interventions appear to have greaterimpacts than later interventions (Pollitt et al.,1993; Husaini et al., 1991; Waber et al., 1981;Grantham-McGregor et al., 1991).

The child and the family

The growth and development of a child dependsnot only on the care, food and resources providedby a caregiver, but also on the characteristics andbehaviours of both the child and his or hercaregivers. For example, better-nourished childrentend to be more active and exploratory and moreable to elicit interaction from parents (Chavez etal., 1975), all of which contribute to learning.Similarly, a child with well-developed psycho-social skills, who is able to engage the caregiver,may then be more effective in demanding andobtaining food. Thus, early interaction betweenchildren and caregivers affects subsequent inter-actions (Engle and Ricuitti, 1995).

Increasing caregivers’ skills and resources mayenable them to provide not only improved nutri-tional care, but also improved psychosocial care.Caregivers who are responsive in a feeding situa-tion also tend to be more responsive and stimu-lating in play situations, and those who showdysfunctional behaviour while feeding their chil-dren are also dysfunctional in other interactions.As a result, implementing feeding and other carepractices in a responsive, stimulating way is likelyto result in social and cognitive, as well as nutri-tional, improvements (Black et al., 1994; Pucker-ing et al., 1995; Polan et al., 1991).

Design and delivery of programmes

The third level of interaction takes place at theprogramme planning and implementation stage.A programme that incorporates nutrition, health,and psychosocial stimulation may be able to pro-vide all three kinds of inputs more efficiently bycombining delivery mechanisms. There is a grow-ing emphasis on the integration of nutrition,health, and ECCD intervention strategies amongagencies that fund and implement internationaldevelopment programmes. For example, UNICEFhas developed a conceptual model for child sur-vival, growth, and development that places careat the centre (Jonsson, 1995; Engle, 1997; Engleet al., 1997; Zeitlin, 1996). Care is defined as “theprovision in the household and the communityof time, attention and support to meet the physi-cal, mental and social needs of the growing childand other household members” (Engle, 1992).The provision of care is a critical link betweenfood and health resources, and the child’s physi-cal growth and psychological development. TheWorld Bank also recommends the integration ofhealth, nutrition, and psychosocial stimulation inprogrammes for early childhood (Young, 1997).

The three levels of interaction—the child, thechild and the family, and programmes—will bediscussed throughout the paper.

Questions to be addressedby the review

A number of recent reviews have clarified whatwe know about the linkages of malnutrition andlack of stimulation with later growth and devel-opment (e.g. Martorell, 1997; Gorman, 1995;Young, 1997; Walker et al., 1998; Levitsky andStrupp, 1995; Strupp and Levitsky, 1995; Pollittet al., 1996). The potential benefits of investingin programmes to improve both physical growthand psychological development are substantial inlight of the role of undernutrition and disadvan-taged environments in perpetuating deficits ingrowth, cognitive function and other aspects ofpsychological development.

The issue is how best to promote healthygrowth and development. The review addressesfour key questions:

● To what extent and through what means canpsychological functioning be improved forchildren living in disadvantaged environments?

● To what extent and through what means canchild physical growth be improved in settingswhere chronic undernutrition is prevalent?

7

● To what extent and through what means cannutrition and psychological development beimproved simultaneously through combinedhealth, nutrition, and psychosocial interven-tions?

● Are there effective models for combined inter-ventions, and are these models feasible forimplementation on a public health scale?

The chapters that follow address these ques-tions based on data from many different sources,including experimental studies, community-basedefficacy studies, and evaluations of large-scaleprogrammes.

Chapter 2 discusses methodological issuesraised in the review. It addresses the outcomemeasures by which answers to the key questions,stated above, can be examined. It also orients thereader to basic issues that affect the interpreta-tion of studies.

Chapter 3 concerns interventions that areaimed at supporting and improving child psycho-logical development, including cognitive, motor,and social development.

Chapter 4 examines interventions aimed atsupporting physical growth.

Chapter 5 presents combined growth and de-velopment interventions, both experimental stud-ies that were designed to test the impact ofcombined growth and development interventions,and effectiveness trials of full-scale programmesthat have included both interventions. Thissection also examines the combined approachfrom a programmatic perspective, discussing thepossible advantages and disadvantages of variousmodels of combined approaches.

Chapter 6 summarizes the results of the re-view and discusses the findings in relation to thequestions raised above.

CHAPTER 1. INTRODUCTION


8

This review draws on several decades of re-search in nutrition, epidemiology, psychol-

ogy, anthropology, and other biological, medicaland social science fields. We utilized a number ofexcellent analytical reviews and meta-analyses ofspecific topics within the larger framework cov-ered by this review. We analysed primary sourcesonly for key studies for which greater detail wasneeded. For each topic, a computer-assisted searchidentified new work published after the date ofprevious reviews.

Measuring impact: the selection ofoutcome measures

Outcome measurement—of child growth and/orcognitive, motor, and psychosocial develop-ment—was a primary criterion when selectingstudies and programmes. The definition of anoutcome is discussed in general terms below. Theresults of intermediate outcomes related to childgrowth and development, when measured, werealso reviewed. Intermediate outcomes, such as

2

Development is usually categorized as:■ Cognitive (or mental) development, including

memory, problem-solving, and numerical under-standing (and, for some authors, language devel-opment).

■ Language development, the ability to communicatewith others, to comprehend speech and expressthoughts (receptive and expressive language).

■ Social-emotional development, including anunderstanding of the relationship of self to other,ability to regulate oneself and one’s emotions,development of social skills.

■ Temperament, a biologically-based tendency suchas the ease or difficulty with which a childapproaches routine and novel situations.

■ Fine and gross motor development, including child’sability to sit, walk, run, and handle small objects.

changes in parent knowledge, are importantbecause they help establish plausibility and iden-tify the pathways by which interventions have im-pact. Sequelae or long-term outcomes help estimatethe benefits to society of investing in growth anddevelopment interventions.

Outcomes related to child development

Psychological development refers to the emergenceof skills and competencies that help a child adaptand function in his or her environment. Theseskills and competencies become more complexas the child ages and matures.

Assessing children’s developmental status posesseveral problems because the process is complexand always affected by the cultural context. Theimpact of child development interventions is typi-cally assessed by standardized tests of cognitiveand motor development in infants and youngchildren. Also used are observational assessmentsof behaviours related to development, learning,social interaction, or subsequent performance inschool, and prevalence of mental deficiency. Thetests used to assess pre-school ability have mainlybeen designed in industrialized, Western coun-tries. Unless they are adapted for content andformat, and standardized locally, they are subjectto inaccurate interpretation. However, as notedby Pollitt (1998), the problems with externallydesigned tests may be less serious if testingevaluates programme progress (rather thanchildren) or compares groups within a similarcultural context.

Unique difficulties arise in assessing cognitiveabilities of infants. The most popular global as-sessments prior to 18 months of age, such as theBayley Scales of Infant Development, are not par-ticularly predictive of later functioning, whereastests from 3 years onward are quite predictive(Pollitt, 1998). Available assessment tools for in-fants that are predictive of later performance tendto be time intensive and difficult to administer,particularly under field conditions. The term

Chapter

Background and methodologyfor the review

9

psycho-motor development is used more often thancognitive development for infants because it isdifficult to distinguish cognitive from motor de-velopment in children younger than 12 months.

Other dimensions of psychological develop-ment, such as social-emotional development,receive less attention, particularly outside the in-dustrialized world. Yet, emotional adjustment isimportant for the child’s capacity to learn anddevelop, as well as for cognitive, motor, and lang-uage development. Emotions are linked to physi-cal conditions, including malnutrition (e.g. Lozoff,1998).

A few aspects of social-emotional functioninghave been evaluated outside of industrializedcountries. Several studies have examinedpredictors of aggressiveness such as exposure toviolence in the community (Liddell et al., 1994;Chikane, 1986). Quality of attachment—theunique and powerful affective relationship thatdevelops between child and caregiver—is asignificant predictor of many aspects of laterfunctioning (e.g. Valenzuela, 1997). Attachmenthas been assessed in many different cultures, butsome investigators have questioned whether theconstruct has cross-cultural validity. Relatively few

other measures ofsocial-emotionaldevelopment havebeen examined indifferent culturalcontexts.

Tools for evalu-ating the immedi-ate environment ofthe child have been

designed for use in research and programmeevaluation (e.g. Bradley and Caldwell, 1984).These instruments include measures of familyinteraction, including social-emotional character-istics, parents’ knowledge and practices relatedto childcare and feeding, and the quality of thehome environment.

Caregivers have many types of interactions withchildren, such as washing, feeding, and carrying.All these care behaviours have both an instrumen-tal (or task-related) and an affective (emotional)component. In addition to recording the instru-mental activities, these tools also measure affec-tive aspects, such as parental responsiveness,warmth, and encouragement of cognitive, lan-guage and motor development (Engle, Menon andHaddad, 1997; Engle and Riccuiti, 1995). Otheraspects that are typically measured with instru-ments such as the HOME Inventory (Bradley and

Caldwell, 1984) include avoidance of restrictionand punishment, organization of the environment,provision of a safe play environment for the child,and opportunities for variety in daily stimulation.An adapted HOME Inventory has been used in anumber of different cultural contexts (Bradley etal., 1989).

Long-term measures of individualoutcomes

The long-term effectiveness of early child devel-opment programmes has been assessed by meas-ures of functioning during later childhood,adolescence, and adulthood (Gomby et al., 1995).Variables measured include placement in specialneeds classes, school progress and achievement(grade levels passed and failed), functional or use-ful knowledge, earning potential, criminal record,and even marital status. Many researchers feel thatthese measures are more appropriate than IQ testscores, since they reflect important functionalconsequences of early experiences (Gomby et al.,1995). Although some of these measures will notbe appropriate in all settings, a similar approachto assessing functional competence can be utilizedin any environment (Pollitt et al., 1993).

Outcomes related to child growth

Child growth is an indicator of past and presentconditions including food intake, health status,and activity levels, and a predictor of futureimpairments in health and performance that mayresult from poor nutrition in childhood. Growthis an important indicator of the impact of nutri-tion interventions because it is a relatively short-term outcome that can be measured in astandardized wayand predicts a rangeof other functionalout-comes. Weightis likely to showchanges more rap-idly than height.

Child growth hasbeen measured twoways: (1) rates ofincrease and incre-ments over specifiedperiods of time, and(2) attained size at agiven point in time(e.g. weight-for-age, weight-for-height, andheight-for-age, based on standardized indices).

CHAPTER 2. BACKGROUND AND METHODOLOGY FOR THE REVIEW

Emotional adjustment isimportant for a child’s capacityto learn and develop, as wellas for cognitive, motor and

language development.

Child growthis an indicator of past and

present conditions includingfood intake, health status,and activity levels, and a

predictor of futureimpairments

in health and performancethat may result from poor

nutrition in childhood.


10

Attained size measures of height-for-age andweight-for-height reflect different nutritionalconditions. Weight-for-height usually reflectsshort-term variations in nutritional input, whereasheight-for-age reflects longer-term or chronic con-ditions (WHO, 1995). In many studies, height-for-age is associated with cognitive growth,whereas weight-for-height is rarely associated withcognition (UNICEF, 1998). Weight-for-age is alsoassociated with cognitive development, but lessclosely.

The effectiveness of interventions for pregnantwomen, designed to improve foetal growth, ismeasured by birthweight, differentiated into chil-dren born prematurely or at term. Children bornbelow 2500 grams at term (40 weeks) are calledsmall for gestational age, or SGA, and are likely tohave suffered inter-uterine growth retardation(IURG). Children born prematurely are also likelyto have low birthweight. The distinction betweenlow birthweight and small for gestational age maybe hard to ascertain if data on gestational age arenot available. In situations in which low birth-weight is caused by nutritional deficiencies in themother, birthweight and cognitive performancehave been linked, particularly among older chil-dren, but the data are not entirely consistent(Pollitt et al., 1996).

Assessing intermediate outcomes, such as dietor feeding practices, is also important when theycontribute to understanding how the interventionhad impact. A number of strategies for measur-ing feeding behaviour have been developed (e.g.Bentley et al., 1991; Engle et al., 1996). Selectedbiochemical indicators of nutritional status arerelevant for this review when nutrition interven-tions were designed to improve micronutrientstatus.

Theoretical backgroundfor the review

The four key questions addressed by this reviewstem from extensive basic research on physicalgrowth and psychological development. Thisbasic work provides the rationale for interven-tions to improve growth and development.Although this review was not intended to revisitthat evidence, the intervention studies includedoften refer to these theoretical underpinnings:

● Malnutrition causes both poor physical growthand developmental delays.

● Malnutrition—measured by growth faltering—is causally related to mortality in infancy and

early childhood, and interventions that reducethe incidence of malnutrition can be expectedto reduce mortality dramatically.

● There is evidence that in many communitieswith endemic malnutrition, both feedingpractices and the selection of foods for infantsand young children from the food sources thatare available within the community are notoptimal.

● In infancy and early childhood, when childrenare fully dependent on others for their nutri-ent intake, the proximate causes of malnutri-tion are: (1) receiving diets of poor quality andinadequate quantity, and (2) inappropriatefeeding practices and behaviours related tofood preparation, frequency, and interactions.There are, in turn, multiple socio-economic,cultural, and psychological determinants ofthese causes.

● Many households in conditions of povertypotentially have the resources to provide ad-equate diets and use good feeding practices thatsupport normal growth. Lacking are the knowl-edge and skills about how to do this withintheir local environmental and cultural context.

● Many other households in conditions of pov-erty are more severely constrained economi-cally and require assistance in the form ofsupplementary, nutritious foods and/or nutri-ent supplements for their infants and smallchildren, and for consumption by mothersduring pregnancy and lactation.

● Many children in developing countries anddisadvantaged populations in industrializedcountries experience delays in motor and cog-nitive development that negatively affect theirschool performance, their ability to maximisetheir educational opportunities, and theirsocial functioning later in life.

● An important factor in developmental delaysassociated with malnutrition is the evolutionof the interaction between the malnourishedchild and his or her caregivers, in which thechild becomes progressively more apatheticand less demanding, and caregivers provide lessstimulation and responsive interactions.

● Feeding is a central aspect of caregiving ininfancy and early childhood, and the teachingof feeding skills provides an opportunity toteach other caregiving skills, such as respon-sive parenting to stimulate motor and cogni-tive development.

11

In summary, interventions to address mal-nutrition (improving food and feeding practices)and to promote child development (responsiveparenting) are both important. There may beadditional benefits of combining activities andintegrating these interventions.

Steps or process by whichinterventions are developed

The development of an effective public healthintervention, such as oral rehydration therapy(ORT) for diarrhoea, involves a sequence ofresearch steps, each of which makes a differentcontribution to the formulation of the final inter-vention (Figure 1 from de Zoysa et al., 1998). Thesteps are dynamic and iterative. Not all steps arerequired for every intervention. However, anawareness of these steps helps to identify wherethe greatest research investments have been placedand which aspects have not received adequateattention.

The five steps most commonly found in thedevelopment of an intervention are:

1. Describe the problem.

2. Identify risk factors (such as biological andbehavioural risk factors).

3. Explore the context and identify the determi-nants of the risk factors.

4. Select or formulate possible interventions.

5. Test the interventions in carefully controlled,double-blind efficacy trials (de Zoysa et al.,1998).

Successful clinical efficacy trials of an interven-

tion do not mean it will also succeed as a publichealth intervention. Four additional steps are es-sential to determine the utility or effectiveness ofthe intervention under real-world conditions. Step6 is to formulate the intervention for implemen-tation in usual public health conditions, and Step7 involves testing it under controlled conditions,but in a public health context. These steps can belabelled efficacy trials of public health interven-tions (de Zoysa et al., 1998).

The last two steps assess the effectiveness ofthe public health intervention. Step 8 is the evalu-ation of how well the intervention functioned asa large-scale public health programme, rather thana small-scale experiment. Finally, Step 9 is thecontinual monitoring and evaluation of the indi-cators for programme improvement.

Unfortunately, many public health inter-ventions do not go through Steps 6 through 9.Often, interventions are scaled up without appro-priate concern for effectiveness and are later foundto be ineffective. In general, this review empha-sizes community-based interventions. Thus, thestudies analysed are primarily community-basedefficacy trials (Step 7) and effectiveness trials (Step8). However, some of the studies reported can beclassified as clinical trials to test an intervention(Step 5).

Dimensions of research rigor toconsider in evaluating interventions

There are three main dimensions in public healthinterventions that affect the rigor of the research,and therefore the analysis and interpretation ofdata. These dimensions also vary in relation tothe steps in the research process described above:


Figure 1. Steps in the research process

Adapted from de Zoysa et al., 1998.

1 2 3 4 5Describe Identify risk Explore the Select or Test

the problem factors context and formulate interventionsidentify the possible

determinants interventions

➞ ➞ ➞ ➞

6 7 8 9Formulate Assess the Assess the Monitor adequacy

public health efficacy of effectiveness and impact ofinterventions public health of public health large-scale public

interventions interventions health interventions

➞ ➞ ➞➞


12

1. The degree of control in the delivery of theintervention.

2. The context and conditions in which theintervention is implemented, and the degreeto which the intervention is appropriate to thatcontext.

3. The level of the intervention in terms of howdirectly the input is related to the biologicaloutcome. (For example, in vitamin A interven-tions, the most direct locus is represented bygiving capsules directly to children; an inter-mediate locus is represented by an interven-tion to change the availability of vitamin A inthe household such as home gardening; and amore distant locus is a food policy interven-tion that is intended to increase vitamin A inthe food supply.)

We refer to these three as dimensions to em-phasize the point that they are continua. Also, itshould be noted that the dimensions are cross-cutting, and each dimension influences theothers and has implications for the interpretationof results. Table 1 summarizes the main points toconsider for each dimension.

Degree of control: adequacy of deliveryand evaluation design

In identifying the degree of control in an evalua-tion, it is important to distinguish between effi-cacy and effectiveness. Efficacy refers to the impactof an intervention under ideal, controlled condi-tions. This is usually possible only within a

research study, and has been referred to here asStep 7 in the research and development process.Effectiveness describes the impact under usual con-ditions, such as during implementation of a pro-gramme (Step 8). Even under field researchconditions, the degree of control is not absoluteand results must be interpreted in light of thecontext and conditions under which the interven-tion was implemented. It is important also to havean assessment of whether or not the interventionactually took place, an adequacy assessment.

It is logical to first review the evidence of effi-cacy when assessing potential impact of any typeof intervention. If efficacy is established, then oneshould next consider evaluation studies of theeffectiveness of full-scale programmes that includeone or more of these proven approaches. Webegin each of the intervention review sections withan examination of efficacy studies and thenexamine the results of effectiveness evaluations,recognizing that the dividing line between modeland full-scale programmes, and between efficacyand effectiveness studies, is not always clear-cut.

In general, the degree of control over deliveryis highest in model interventions that are devel-oped to test efficacy. Probability designs, which arecharacterized by random assignment to treatmentand control groups, allow for statistical determi-nation of whether or not the intervention hadan effect. Such randomized trials are the preferreddesign for demonstrating efficacy. Plausibilitydesigns do not involve random assignment butachieve some degree of control by considering therelative impact either before and after or in groups

Table 1. Cross-cutting dimensions to consider in the interpretation of impact evaluations

Dimension Issues that affect interpretation

Degree of control ■ Adequacy or fidelity of implementation: the degree to which the intervention isdelivered as planned (e.g. verification that the supplement was actually consumed,home-visits conducted, etc.)

■ Evaluation design: the degree to which the design controls for bias and confounding,allowing causal inferences (e.g. probability, plausibility, or adequacy designs)

Context and conditions ■ Appropriateness of intervention to the culture and the community

■ Degree of risk if nothing is done and chance of benefit from intervention (affected bytargeting, rates of participation, availability of resources)

■ Appropriateness of the mode of delivery

■ Specification of other conditions under which an intervention is effective or effica-cious (i.e. effect modifiers)

Level of intervention ■ Interventions at the level of the individual, family, community, or society are affectedby different numbers and types of intervening and contextual factors that influenceimpact

13

only what type of intervention is applied that de-termines effectiveness, but also how and where theintervention is implemented, and to whom. Anintervention that is effective in one setting can-not necessarily be transplanted successfully toanother context without adaptation. Factorsexternal to the specific intervention, such asexisting infrastructure, policy, political and cul-tural context, may also influence effectiveness.

In summary, degree of control and the con-text, in combination, affect the degree to whichimpact can be attributed to a type of interven-tion. The strongest causal inference on impactcomes from designs that deliver probability orhigh plausibility results and interventions that areappropriate to the context where they are imple-mented. If either of these conditions is not met inan efficacy trial, it complicates the interpretation.For example, a supplementation trial of womenin New York City (Rush et al., 1980) was verywell-designed and implemented to investigatecausal relationships between dietary intake dur-ing pregnancy and birth outcome, but it was con-ducted in a population that was not actuallydeficient in the nutrients supplied. Thus, the re-sults did not replicate findings in malnourishedpopulations and contributed little to understand-ing of the role of supplementation during preg-nancy in malnourished women.

Level of intervention: where in the causalpathway does the intervention occur?

The third main dimension that affects the inter-pretation of impact is the level at which the inter-vention occurs in the causal pathway leading tothe expected outcomes. An intervention aimed ata cellular or a behavioural process may be imple-mented at the level of the individual, family,community, or society. The more distal theintervention is from the biological process orbehaviour it is intended to affect, the greater theopportunity for intervening and contextual fac-tors to affect its impact. As in the example above,if supplemental vitamin A capsules are givendirectly to children and no impact is seen, thiscould mean that the supplement was inactive, thatthe supplemented children did not have a defi-ciency of vitamin A, that the children lackedadequate retinol binding protein (due to proteinenergy malnutrition) or that vitamin A is notefficacious in improving the outcome of interest.If testing has eliminated the first three possibili-ties, stronger conclusions about the efficacy ofvitamin A supplementation per se could be made.


How well anintervention is actually

delivered is animportant considerationin the interpretation of

results from efficacystudies and

effectivenessevaluations of nutrition

and psychosocialinterventions.

with and without the intervention. Such designsare used to establish whether or not the interven-tion had an effect above and beyond that causedby other external influences. In reviewingevidence for efficacy, we looked for studies withinformation indicating a high degree of controlover implementation and that used probability orplausibility designs.

How well an intervention was actually deliv-ered is an important consideration in the inter-pretation of results from efficacy studies andeffectiveness evaluations. A lack of impact can-not be interpreted to indicate that the underlyingassumptions or intervention design were incor-rect without verifying that the implementation

occurred as planned. In-formation on the fidelityor adequacy of imple-mentation is consideredin the interpretation ofimpacts reported in stud-ies reviewed here.

The degree of experi-mental control is muchlower in full-scale pro-grammes than in efficacytrials. Thus, the effective-ness evaluations identi-fied by this reviewrepresent a lower degreeof control over imple-

mentation, and evaluations assess programmeeffectiveness in a real-world context, with all thefactors that limit or enhance implementation. Thefidelity or adequacy of implementation must beconsidered to understand the factors contribut-ing to the programme’s success or lack of impact.

Context and conditions of the intervention

Closely related to the issue of degree of control isthe appropriateness of the intervention for thecontext. Impact is influenced not only by thecharacteristics of the intervention, but also by thecharacteristics of the context where it is imple-mented, and how well they match. Thus, whenreviewing efficacy trials one should discusscontingent efficacy because the impact seen iscontingent on contextual issues, such as whetherthe intervention being tested is appropriate forthat population.

An intervention proven to be efficacious maybe ineffective if implemented in a programme thatis unable to achieve adequate coverage, utiliza-tion, or sustainability. In other words, it is not


14

On the other hand, if the intervention was a homegardening project, narrowing the field of possi-ble explanations would be quite different. Inter-vening factors could have occurred, for example,at the level of family or community participation,access to land or inputs, success of the foodproduction, consumption by the family, or con-sumption by the child.

Interventions on several levels will often berequired to address population needs in order tosupport the improvement of child growth anddevelopment. In general, this review concentrateson interventions at the individual and family levelbecause most of the model programmes designedto evaluate efficacy are implemented at this level.

15

This chapter reviews interventions aimed atimproving the psychological development of

disadvantaged or at-risk children and families,either through child centre-based education, orvia home visits or other activities to improveparenting skills and stimulation within the home.We included programmes for low birthweight(LBW) and non-organic failure-to-thrive (NOFTT)infants, but excluded interventions targetedspecifically to children with physical disabilitiesor organically-based mental disabilities. However,many of these programmes are designed to re-duce developmental delays caused by inadequatenutrition, health, and socio-economic conditions.

All outcomes were related to psychologicaldevelopment. Psychosocial interventions, includ-ing stimulation, are described first, followed bynutritional interventions. Next, the chapter dis-cusses efficacy trials of nutrition interventions thatdid not include any stimulation or psychosocialintervention, but that evaluated psychologicaloutcomes.

Impact of ECCD interventions onpsychological development

Model programmes in the United States providemuch of the evidence for the efficacy of early child-hood care and development (ECCD) interven-tions. Although ECCD programmes abound in

other parts of the world, there have been few effi-cacy trials outside the U.S. Examples of full-scaleearly education programmes that include a nutri-tion component (which is common for ECCDprogrammes, especially in developing countries)are described in Chapter 5.

The review examines whether and under whatcircumstances single-focused interventions areeffective, but does not cover the extensive bodyof literature that compares different theoreticalapproaches, such as specific teaching methods.Although this information is essential for inter-vention development, it is outside the scope ofthis review. This literature is reviewed elsewhere(Ottenbacher et al., 1987).

The most common model for early childhoodcare and development programmes is centre-basedpre-school education in which children arebrought from their homes to a central location,such as a school, a community building or otherpublic site, or in some cases, a private home. Otherforms of psychosocial intervention improve thechild’s psychological development within thecontext of the family and the home. Meisels (1992)suggests that centre-based and family-based inter-ventions can also be characterized as child-focused,parent-focused, or joint-focused as shown in Figure2. Child-focused approaches involve high inten-sity (frequent, or many hours or weeks), directinterventions with children, whereas parent-

3

Figure 2. Dimensions of focus in early childhood care and development programmes

(from Meisels, 1992)

High focus on parent

Parent-focused Joint-focusedLow focus High focuson child on child

Low intensity Child-focused

Low focus on parent

Chapter

Interventions to supportpsychological development


16

focused interventions involve direct contact withparents. Joint-focused programmes include highintensity components for both children andparents.

Behavioural changes for children and parentsare expected impacts of these programmes. Al-though this review is limited to psychologicaldevelopment outcomes, such as performance oncognitive and motor tests, parenting behaviourand parents’ life skills (e.g. education level) maybe equally significant and mediate or provide themeans by which interventions are effective, par-ticularly those with a focus on parents. Evidenceof the maternal benefits from early interventionshas been reviewed by Benasich et al. (1992).

Child-focused interventions thatprovide psychosocial stimulation directlyto the child

When psychological development activities inchildcare centres are implemented with high qual-ity, they consistently show a positive impact oncognitive function and IQ scores (Consortium for

Longitudinal Studies,1983; Haskins, 1989;Gomby et al., 1995;Hertzman and Wiens,1996). Barnett (1995)cites reviews of pro-grammes for disad-vantaged children,including those im-plemented on a largescale, that can resultin an average increaseof about eight IQpoints at programme

completion. In addition, four recent re-searchprojects in the United States have foundsimilar effects, especially the centre-basedprogrammes (Barnett, 1995). Positive outcomeshave been reported for programmes in otherdeveloped countries, including Ireland andAustralia (Boocock, 1995).

To address the question of longer-termeffects, Barnett (1995) reviewed 36 early child-hood programmes in the U.S. that followed chil-dren to at least age 8. Children who participate inearly child education programmes show strongerperformance on IQ tests until they enter school,after which there is often a decline. However, othertypes of important long-term impacts have beendocumented, such as lower rates of grade repeti-tion or placement of children in remedial or

special education classes (Barnett, 1995;Hertzman and Wiens, 1996; Consortium forLongitudinal Studies, 1983).

One of the most carefully evaluated U.S.projects is the Perry Pre-school Program inYpsilanti, Michigan, instituted in the early 1960s.As is characteristic of many children from impov-erished backgrounds, the young children in theproject initially appeared to have low cognitiveability, as judged by pre-school IQ tests. The in-tervention provided psychosocial stimulation andincreased children’s cognitive performance by thetime they entered school. From this head start,they experienced greater success and commitmentin school, which was reinforced by the percep-tions of parents and teachers. The interventioneventually led to less deviant behaviour, highereducational attainment and other indicators ofsocial development, even through 27 years of age(Schweinhart and Weikart, 1980; Schweinhart,Barnes and Weikart, 1993; Barnett, 1996).

A number of studies provide insight intoprogramme and participant characteristics thatincrease impact, such as intensity and durationof programmes (Gomby et al., 1995; Hertzmanand Wiens, 1996), and gender (Barnett, 1995).Although most pre-school programmes serve 3-to 5-year-olds, research on programme effective-ness and “neurological evidence that the environ-ment influences infant brain development” (p. 15,Gomby et al., 1995), show that services begunduring infancy have greater impact (Ramey andRamey, 1998). Early Head Start, a new initiativeof the Head Start Bureau in the U.S., reflects thisperspective by serving children from 0–3 years ofage (U.S. Department of Health and HumanServices, 1993).

Opinions vary concerning the extent to whichpre-school programmes have demonstrated long-term benefits (Gomby et al., 1995; Haskins, 1989)but there is enough evidence to suggest that sucheffects can be achieved with high-quality,intensive programming. The long-term results ofpre-school, centre-based interventions must be in-terpreted in light of the poor home and schoolenvironments that disadvantaged children con-tinue to experience after pre-school. Even partici-pants in the most successful programmes are stillat a considerable disadvantage compared tomiddle-class children. Results from programmesaround the world provide evidence that “pre-school experience helps low-income childrennarrow, but not close, the achievement gapseparating them from more advantaged children”(p. 94, Boocock, 1995).

Children who participate inearly childhood care and

development (ECCD)programmes show strongerperformance on IQ tests.

They are also less likely torepeat grades or be placed

in remedial or specialeducation classes.

17

Parent-focused interventions to improveparenting skills and psychosocialstimulation in the home environment

Many of the model interventions that are designedto improve parenting skills and aspects of thehome environment are implemented throughhome visits. Programmes using home visits varyin goals, level of intensity, type of families served,and type of staff.

In the U.S., programmes to improve parentingskills have shown inconsistent and more modesteffects on psychological outcomes than have child-focused programmes (Gomby et al., 1995). Pro-grammes that provide weekly or less frequenthome-visits are much less intensive than the child-focused programmes described above, so lowerimpact on the type of performance measured byIQ tests might be expected. However, home-visiting has been found to be an effective com-ponent of programmes targeting low birthweightand premature infants (Olds and Kitzman, 1993;Hertzman and Wiens, 1996), non-organic failureto thrive children (Black et al., 1995), and under-nourished children in developing countries(Grantham-McGregor et al., 1991; Waber et al.,1981). Olds and Kitzman (1993) reviewed fourrandomized trials of home visiting interventionsthat were designed to work with parents to im-prove the psychological development of pre-termand low birthweight newborns. All of the trialsfound consistent evidence of increased psycho-logical test performance.

In programmes directed to parents and chil-dren who are at social and economic risk, the

results are moremixed. A review of15 randomized tri-als with disadvan-taged families foundsix with significanteffects on psycho-logical developmentof children. Theprogrammes thathad an impact onpsychological devel-opment also dem-onstrated positiveeffects on parental

caregiving, such availability of toys and books,mother-child interaction and verbalization, paren-tal attitudes toward children, and behaviour man-agement. In addition, two trials that did not findsignificant cognitive effects did, however, improve

aspects of caregiving, such as responsiveness,teaching behaviour, and encouragement (Olds andKitzman, 1993). There is also some evidence thatbenefits may extend to siblings of the target child(Olds and Kitzman, 1993; Seitz and Apfel, 1994),probably as a result of improved parenting skillsand parent-child interaction.

The results of the review of these randomizedtrials show that programmatic aspects of deliv-ery, as well as participant characteristics, areextremely important. Many of these programmesbegan the intervention with pregnant women orpicked up children immediately after birth. Theduration of the intervention was usually about twoyears. The frequency of home-visits ranged fromweekly to monthly, with variation across pro-grammes, and also within programmes depend-ing on the age of child and the degree of need(St. Pierre et al., 1995).

The strongest impact was obtained in broad-based, comprehensive programmes that employedprofessionals as homevisitors, rather than innarrowly-focused pro-grammes that wereimplemented by para-professionals (Olds andKitzman, 1993). Intensity(or frequency) is alsoimportant. A carefully de-signed trial on effective-ness related to frequencyof home visits in Jamaicashowed significant im-provements with weeklyvisits, marginal effects with bi-weekly visits, andno effects for monthly visits. Frequency was moreimportant than professional versus para-professional status of the home visitor (Powell andGrantham-McGregor, 1989).

In sum, parent-focused interventions showmore impact on parenting behaviours, and child-focused interventions have stronger effects onchildren. Home visits are a promising strategy foraddressing the constellation of social problemsthat contribute to unfavourable home environ-ments. They have been most successful whendirected to those with greatest needs, such asyoung mothers and families of low birthweightbabies. Parent-focused interventions also havepotential benefits for siblings and other familymembers. The greatest impacts were seen in pro-grammes that augmented home visits with serv-ices such as centre-based early childhood careprogrammes (Gomby et al., 1993). This finding

CHAPTER 3. INTERVENTIONS TO SUPPORT PSYCHOLOGICAL DEVELOPMENT

Parent-focused interventionsto improve child

psychological developmentalso demonstrate positive

effects on parentalcaregiving, such as improvedmother-child interaction and

verbalization, parentalattitudes toward children,

and behaviour management.

Interventions thataugment home visitswith services such as

centre-based earlychildhood care

programmes show thegreatest impact to

psychologicaldevelopment.


18

led Weiss (1993) to conclude that home visitingis a necessary, but not sufficient, component ofprogrammes to improve child well-being.

Social and cultural contexts are significant. Aspreviously noted, most of the programmes withimpact evaluations were U.S.-based. Poverty inthe U.S. has changed over the past 30 years andis now more likely to be linked with lack of timefor children due to single-parenting and employ-ment of all parents, and social dysfunction suchas alcohol and drug use, domestic conflict, andstreet violence (Hernandez, 1995). Families indeveloping countries have a lack of knowledgeor skills, rather than social dysfunction. Becausethe proposed intervention provided knowledgeand skills rather than attention to other existingsocial problems such as drug abuse, activities weremore effective for families in developing coun-tries than for high-risk families in the U.S.

Joint-focused interventions to improveparenting skills, and to providepsychosocial stimulation to children andsupportive services to parents

Joint-focused programmes vary in the amount ofdirect attention given to parents and children.Some are fundamentally child-focused, in thatthey provide intensive centre-based services tochildren, but also provide supportive services toparents, possibly through parent meetings.

Some joint-focused programmes begin withhome visits and then add centre-based care as thechildren get older. The Infant Health and Devel-opment Programme (IHDP) is a multi-site,randomized trial in the United States that wasdesigned to test the efficacy of a comprehensiveearly intervention to reduce the developmentaland health problems of premature, low birth-weight infants. Activities were home-based in thefirst year of life. From 12 to 36 months thechildren attended an intensive centre-based pro-gramme five days a week, and regular group meet-ings were held for parents. At programmecompletion, the children in the intervention groupscored significantly higher on IQ tests, and theodds of scoring in the range of significant devel-opmental delay (below an IQ of 70) were 2.7 timeshigher for control children (IHDP, 1990).

Although it would be valuable to determinethe relative benefit of the parent and child com-ponents of joint interventions, it is difficult toseparate the effects of the different components,or to determine the advantage of intervening withboth children and parents. In studies in which

this has been possible, combined parent and childinterventions have been found to be more effec-tive than either one alone for children’s psycho-logical development. Project CARE in the U.S.found stronger effects on psychological develop-ment when the intervention combined intensiveeducational programmes for children with familysupport as compared with family support only(Miller and Bizzell,1983, cited in Hertz-man and Wiens,1996).

A study in Turkeyexamined the effectsof pre-school pro-grammes with educa-tional componentsversus custodialdaycare programmesversus staying athome. A training programme for mothers de-signed to help them foster cognitive development(using the Home Intervention Programme for Pre-school Youngsters or HIPPY curriculum) and tobecome sensitized to the child’s needs was assessedusing each of these care arrangements. Attend-ance at an educationally-oriented pre-schoolprogramme and, to a lesser extent, mothers’ par-ticipation in the home training programme, wereassociated with better scores on cognitive testsand school performance. The effects of the twocombined interventions (parent training andeducationally-based centre care) were the great-est (Kagitcibasi et al., 1988).

Greater degree of involvement with a pro-gramme has generally been associated with in-creased impact, but it is likely that more motivatedparents are both more likely to be involved, andmore likely to have better parenting skills any-way, thus requiring caution in the interpretationof findings. An evaluation of the impact of IHDPfound that high child involvement resulted in thegreatest impact on psychological development,regardless of the level of parental involvement.High parental involvement plus high child in-volvement resulted in the largest improvementin scores on a home environment evaluationprotocol (Liaw et al., 1995).

Other approaches to serving both children andparents are two generation programmes, such asEven Start and the Comprehensive Child Devel-opment Program. Two-generation programmes arebased on a relatively new approach in the U.S.that provides early childhood education to chil-dren and provides training on parenting skills, as

Interventions that are bothchild- and parent-focused

are more effective thaneither one alone for

children’s psychologicaldevelopment.

19

well as services such as education, literacy, or jobtraining to help parents improve their economicsituation (St. Pierre et al., 1995). These pro-grammes integrate services to parents and chil-dren with the aim of having a longer-term impacton both. However, since the programmes are re-cent, no longer-term evaluations are available.

Programmes that combine early education andfamily support have been reported to reduce therisk of antisocial behaviour and chronic delin-quency by improving parenting behaviours,enhancing mothers’ life and job opportunities, andimproving verbal abilities of children (Yoshikawa,1995). However, a review of six two generationprogrammes found more limited effects (St. Pierreet al., 1995). Of the four programmes that evalu-ated psychological development, only one showedsignificant effects. The lack of effect on psycho-logical development may reflect the fact that theeducation component for children was limited tohome visits prior to age 4, and then Head Startfor most of the programmes. Some programmessimply referred clients to services provided byother organizations, resulting in little consistencyin the quality and frequency of child educationactivities provided (Gomby et al., 1995; St. Pierreet al., 1995).

Five of the programmes documented positiveeffects on parenting behaviours, and three wereable to produce significant differences in the per-cent of mothers who obtained a high schooldiploma. However, these increases in educationalattainment were not reflected in income, and onlyone showed an impact on employment (St. Pierreet al., 1995). Two-generation programmes weresuccessful in increasing the participation of chil-dren and their parents in relevant social and edu-cational services. As with other programmes, thegreater the child or parental participation, thegreater the impact (St. Pierre et al., 1995).

St. Pierre et al., (1995) conclude that (1)effects on children can only be maximized throughhigh-intensity interventions with children; (2)effects on parents require quality services directedto parents, and (3) evidence is lacking on theefficacy of achieving effects on children solelythrough interventions with parents. They cautionthat high-quality, intensive services targeted toboth children and parents will be expensive, butbalance this against costs of remedial programmes,which are also expensive. Finally, they suggest thatwhile there has been considerable research on thecharacteristics of high-quality early childhood careand development programmes, further researchis needed to discover more innovative approaches

to interventions for parents. It should be notedthat the interventions reviewed by St. Pierre andcolleagues are almost exclusively in developedcountries, and their applicability in developingcountries remains to be determined.

Effectiveness of ECCD programmes

There are many effectiveness studies of ECCDprogrammes that examine quality and impactwhen these programmes are taken to full-scaleimplementation. In this review, we provide a fewexamples of programme effectiveness in differentcontexts.

As is true of efficacy trials, U.S. pre-school pro-grammes provide much of the current knowledgeon the effectiveness of full-scale early childhoodinterventions. Of these, Head Start has been themost broadly implemented and evaluated. Recentstudies compared the performance and progressof Head Start participants to non-participants withsimilar backgrounds. These studies have demon-strated important impacts of the Head Start pro-gramme on children (Lee et al., 1988; Currie andThomas, 1995). A review of various full-scalechild development programmes in the U.S. foundpositive effects on pre-school psychological testscores, although they were smaller and less con-sistent than those found for model programmes(Barnett, 1995). The long-term effects of these full-scale programmes on school progress were morefavourable, however, with eight of 10 studies find-ing statistically significant reductions in graderetention or special education placement (Barnett,1995).

In a review of research on cost-effectiveness,Barnett and Escobar (1989) concluded that avariety of early childhood education interventionsare effective, and theseinterventions for disad-vantaged children aresound public invest-ments. They note a lackof research on the linksbetween programmecharacteristics, the con-text of children’s lives,costs and outcomes.One example of theimportance of the context of children’s livescomes from Brooks-Gunn et al. (1993). Theyfound that characteristics of the neighbourhood,especially presence or lack of affluent neighbours,were associated with childhood IQ, teenage birthsand likelihood of leaving school early, even after


Early childhoodeducation interventions

for disadvantagedchildren are effectiveand a sound public

investment.


20

adjustment for family socio-economic status(SES).

Although significant efforts have been made toreview the information on programmes outsidethe U.S., these reviews have been hampered bythe lack of well-designed evaluations (Young,1996; Myers, 1992). Noting the difficulty in mak-ing definitive statements, Young (1996) concludesthat early childhood interventions are linked tosuch benefits as higher scores on tests of cogni-tive function, higher school enrollment and lowerrates of grade repetition and drop-outs.

Impact of nutrition interventions onpsychological development

This section examines the effects of nutrition in-terventions on psychological development. Theseinterventions were exclusively nutritional and didnot involve ECCD. (Combined ECCD and nutri-tion interventions are described in Chapter 5.)Unlike the previous section, much of the researchreported here was conducted in developingcountries.

Improving nutrition throughsupplementary feeding

The lack of knowledge about the importance ofdietary quality lead early investigators to focusprimarily on the effects of increased macro-nutrients (energy and protein) on children’spsychological development. In these supplemen-tary feeding interventions, the subjects (or in somestudies, their families) were provided with extrafood, with the intention of examining the effectsof greater dietary intake on physical growth andpsychological development. The results of theseinterventions suggest unequivocally that “there isevidence to support nutritional effects on behav-iour independent of social and environmentalfactors” (p. 2243S, Gorman, 1995; also Martorell,1997; Walker et al., 1998). Specifically:

● Supplementary feeding of women when theywere pregnant (Joos et al., 1983) and/orlactating has shown strong and consistent posi-tive effects on motor development of infants.Effects on children’s psychological develop-ment are more evident from 18 monthsonward than prior to 18 months (Gorman,1995; Pollitt and Oh, 1994; Grantham-McGregor et al., 1991; Husaini et al., 1991;Pollitt et al., 1993; Waber et al., 1981).

● Supplementary feeding of infants and youngchildren has resulted in significant increasesin “broad measures of cognitive development”(p. 2241S, Gorman, 1995). This has been con-sistently demonstrated in spite of variations inmeasurement instruments and in interventionmodels (Pollitt and Oh, 1994; Grantham-McGregor et al., 1991; Husaini et al., 1991;Pollitt et al., 1993; Waber et al., 1981; Engleet al., 1992).

● Longer-term impacts of early supplementationhave been reported in several studies. Pollittet al. (1993) found that participants who re-ceived a high-energy protein supplement(prenatally and in early childhood) performedbetter on numerical knowledge, functionalknowledge, vocabulary and reading achieve-ment in adolescence than those who receiveda low-energy, no protein supplement. Impactswere greatest among participants from the mostneedy families and among those given moreschooling opportunities. Chavez, Martinez, andSoberanes (1995), in a follow-up of a smallsample from Mexico, also reported long-termsignificant effects of early food supplementa-tion.

Mechanisms for effects of malnutritionon development

How improvements in nutrition operate to affectpsychological development is not well understood(Gorman, 1995; Meeks-Gardner et al., 1995). Theinformation provided by supplementation trialsdoes not provide definitive answers to questionsof possible mechanisms (Gorman, 1995). How-ever, three decades of research on linkages be-tween malnutrition and behaviour has led tosignificant changes in the theoretical models ofhow nutrition could affect brain function. In the1960s, “it was feared that malnutrition enduredduring certain sensitive periods in early develop-ment and would produce irreversible braindamage, possibly resulting in mental retardationand an impairment in brain function” (Levitskyand Strupp, 1995). In their recent review ofnon-human animal studies, Levitsky and Struppconclude:

We now know that most of the alternations inthe growth of various brain structures eventu-ally recover (to some extent)…. However,recent neuro-pharmacological research hasrevealed long lasting, if not permanent, changesin brain neural receptor function resulting from

21

an early episode of malnutrition. …The kindsof behaviours and cognitive functions impairedby malnutrition may be more related to emo-tional responses to stressful events than tocognitive deficits per se, the age range ofvulnerability to these long-term effects ofmalnutrition may be much greater than we hadsuspected, and the minimal amount of malnu-trition (hunger) necessary to produce theselong-term alternations is unknown.” (p. 2212S)

The most consistently observed effects ofgestational or lactational malnutrition in animalsare changes in motivation, increased emotionalreactivity and less ability to learn new things fromthe environment (Strupp and Levitsky, 1995).These emotional and motivational effects havebeen reported in many animal studies. Animals,like humans, tend to be stressed by novel situa-tions, aversive reinforcement such as electricshock, or loss of a reward (Smart, 1998). Malnu-trition seems to lower the animal’s threshold ofarousal, resulting in more intense reactions tounpleasant stimulation or a more frequent expres-sion of emotionality or anxiety in situations thatevoke stress. Strupp and Levitsky (1995) suggestthat in addition to direct effects of malnutrition,deficits in cognitive function may occur becausean animal that experiences a protracted period ofmalnutrition learns to interact less fully with theenvironment, or experiences the poorer caringenvironment provided by a depressed andmalnourished mother.

Timing and duration of supplementation

In her review of the effects of nutrition supple-mentation on psychological development,Gorman (1995) concluded that gestation and thefirst two years of life are the most importantperiods for supplementation, and that longerduration of supplementation is also associatedwith better outcomes. Similarly, Pollitt (1996) con-cluded that greater impacts on cognitive outcomeswere seen with earlier and longer supplementa-tion. His generalizations are based on compara-tive analyses of supplementation trials inGuatemala and Colombia that initiated supple-mentation at different times (ranging from pre-natal to 72 months) and for different durations.However, very early supplementation has beenshown to have negative results, possibly becauseit interferes with breastfeeding. In the Guatema-lan study, supplementation was very limiteduntil children were in their second year, due to a

cultural pattern of extended exclusive breast-feeding. Therefore, it probably did not replacebreastfeeding (Pollitt et al., 1993). In a longitudi-nal study in Bogota, Colombia, supplementationduring the first six months was not associated withcognitive outcomes, and subjects who begansupplementation after 6 months of age actuallyperformed better than those who began earlier(Waber et al., 1981).

Gorman (1995) and Pollitt (1996) draw atten-tion to the positive effects of nutrition interven-tions after early childhood. The interpretation ofresults must take into account the chronologicalpattern of malnutrition in different populations.In some parts of the world, children are at risk ofmalnutrition mainly in infancy and early child-hood, whereas in other places a combination ofdietary characteristics and practices, and diseasepatterns result in a much longer period ofvulnerability.

There is some flexibility in the capacity ofchildren to respond to the timing of nutrition sup-plementation; there is not a single point beyondwhich improved nutrient intake will have noeffect. However, this generalisation does notnecessarily hold for micronutrients, for whichcritical periods are more clearly defined. Theseissues are discussed later in the chapter.

Activity levels and behaviouralresponsiveness

Pollitt et al. (1993) have proposed that theassociation between nutritional status and psy-chological development may be mediated bymotor maturation, activity level, and exploratorybehaviour. This is consistent with the data show-ing significant effects of supplementation onmotor development in infants.

In a longitudinal intervention study in Mexico,supplemented infants were shown to be moreactive and to spend much more time playing andless time being carried or restricted in a crib(Chavez et al., 1975). As they grew older, theyspent more time talking and less time crying thanun-supplemented children did. These behaviourpatterns, in turn, led to changes in parental be-haviour such that supplemented children receivedmore care and attention from both parents, werespoken to and listened to more, and received moreinstructions, praise and rewards. Supplementedchildren also scored higher on tests of psycho-logical development (Chwang et al., 1995).

In Indonesia, a short-term (14 weeks) trial ofsupplementary feeding of infants found signifi-



22

cant impacts on motor (but not mental) develop-ment scores (Husaini et al., 1991).

Analysis showed that effects of supplementa-tion on motor development were independent of,rather than mediated by, effects on weight. Theauthors hypothesise that the additional energyindependently resulted in both the “accretion ofadipose tissues and neural activation that broad-ened children’s behavioural repertoire” (Husainiet al., 1991).

In the Jamaican supplementation trial withchildren aged 9-24 months (described in Chap-ter␣ 5), the stunted children were less active thannon-stunted children at the time of enrollment,and activity levels were correlated with loco-motor development. The difference in activitylevel disappeared by the 6-month follow-up,regardless of whether or not the children weresupplemented. Nutritional supplementation hada positive effect on psychological development,but there was no evidence that this effect wasmediated through increased levels of gross motoractivity (Meeks-Gardner et al., 1995). Grantham-McGregor (1995) has suggested that activity levelper se is not the critical factor, but rather thequality of the child’s exploration, which relatesto the environment in which the activity takesplace.

Issues in defining mechanisms

A number of methodological and theoreticalissues must be considered when interpreting theeffects of improved macronutrient nutrition onpsychological development. The validity andrelevance of the measures of psychological devel-opment must be considered. Several authors havesuggested that the developmental scales used,particularly those related to psychological devel-opment in infancy, may not measure the indica-tors that are most sensitive to nutritional insultor most predictive of later achievement (Gorman,1995; Husaini et al., 1991).

Most of the studies were not able to randomizecontrolled trials. The studies cited here can becharacterized as high plausibility designs (de Zoysaet al., 1998). Treatments were assigned randomly,but often by community rather than by individual.With this design, it may not be possible to con-trol for all potential differences among commu-nities even with statistical controls. Re-analysesof data from the trials in Guatemala, for example,indicated that despite similarity of the villagesregarding children’s growth and dietary patterns,there were some differences in social and eco-

nomic characteristics (Engle et al., 1992; Pollittet al., 1993).

In many of these studies, researchers had diffi-culty determining that there was actually a netincrease in nutrient intake, even when the inter-vention was implemented under relatively con-trolled field conditions. Therefore, it is difficultto determine whether there is a dose-responserelationship between nutrient intake and psycho-logical test scores (Gorman, 1995). Finally, inorder to isolate the effects of energy and proteinsupplementation, many of these studies providedmicronutrient supplementation to both treatmentand control groups. It is not known howmicronutrients may have complemented theeffects of the food supplements on children’sdevelopment.

An important theoretical issue is the degree ofundernutrition and deprivation in the population.Most of these studies were implemented inpopulations with chronic undernutrition and ahigh prevalence of stunting among children. Thegreater the malnutrition, the greater the potentialfor response (Pollitt et al., 1993).

Another important issue concerns the availabil-ity of resources to complement the provision offood from an external source. For example, inBogota, the strongest effects of the interventionwere seen in children from homes with greaterresources, such ashigher levels of ma-ternal education(Waber et al., 1981).Gorman (1995) sug-gests that both need(e.g. poor socio-economic status) andresources (e.g. mater-nal education) inter-act to improve theoutcome of nutritioninterventions.

Despite the methodological and theoreticalproblems that affect interpretation, these studiesprovide consistent evidence that supplementaryfeeding interventions are efficacious in signifi-cantly improving psychological development inyoung children, and that that these effects canpersist over time.

Rehabilitation of severelyundernourished children

Severely malnourished children score poorly ontests of psychological development and exhibit

Supplementary feedinginterventions can

significantly improvepsychological development

in young children, andthese effects can persist

over time.

23

behavioural abnormalities that could affect learn-ing (e.g. apathy and reduced exploration of theenvironment) (Grantham-McGregor, 1995). How-ever, few studies have assessed the effects ofnutrition rehabilitation following an episode ofsevere malnutrition on psychological develop-ment. It is very difficult to establish a causal rela-tionship because the effects of the acute episodeof undernutrition cannot be separated from thechildren’s poor home environments, which oftenprovide inadequate stimulation and care, as wellas inadequate diets. In theory, this question wouldbe addressed by randomized assignment todietary or stimulation interventions. However,such studies obviously cannot be conducted onethical grounds.

Grantham-McGregor (1995) reviewed theavailable evidence and reported that developmen-tal test scores of severely malnourished childrenare usually extremely low, but scores tend to im-prove shortly after these children recover fromthe acute malnutrition. Thereafter, however, per-formance remains low compared to that ofmatched controls or siblings. No specific type ofdeficits have been demonstrated to be associatedwith acute malnutrition, and she postulated thatthe acute episode probably adds little to theeffects of the underlying chronic undernutritionand deprived environment.

Correction of micronutrient deficiencies

Diets that are low in energy and protein are nearlyalways deficient in at least some micronutrients,and poor dietary quality is also a significant con-tributor to malnutrition of children in the devel-oping world. This section reviews studies thathave examined how correcting micronutrientdeficiencies affects psychological development.

Iodine

Iodine deficiency during pregnancy results in thebirth of babies with severe retardation of physicalgrowth and psychological development (cretin-ism). These severe effects of intrauterine iodinedeficiency (IDD) are irreversible (Levin et al.,1993). There is also considerable evidence thateven milder forms of the deficiency, in utero, areassociated with lower scores on psychologicaltests. Children and adults who are iodine defi-cient tend to be cognitively limited compared tothose who are iodine replete. Based on a meta-analysis of 18 studies, Bleichrodt and colleaguesfound that the mean cognitive scores for iodine-

deficient groups of children and/or adults wasabout 13 IQ points lower than those of non-deficient groups (Bleichrodt and Born, 1994;Bleichrodt et al., 1996).

The efficacy of prenatal iodine supplementa-tion in correcting and preventing IDD is wellaccepted and internationally recognized (ICN,1992*1). In Ecuador (Fierro-Benitez et al., 1989*),children whose mothers had received iodized oilprior to the secondtrimester of preg-nancy demonstratedbetter school per-formance than con-trols, although bothgroups had impairedperformance. Con-trolled randomizedtrials in Papua New Guinea demonstrated thatinjecting pregnant women with iodized oil canprevent cretinism in their children (if womenreceive injections prior to pregnancy), and reducestillbirth, infant mortality and defects in motorperformance of apparently normal children(Conolly et al., 1979*). Whereas the effects ofiodized oil injections were shown to persist forthree to four years or longer, oral doses of 1mlare considered to provide protection against IDDfor one year, and doses of 2␣ ml provide coveragefor three years.

Iodine deficiency during childhood also seemsto affect cognitive function, although to a lesserdegree than in utero. In Bolivia, school childrenwere randomly assigned to receive oral iodizedoil or a mineral oil placebo (Bautista et al., 1982*).Improved iodine nutrition and psychological per-formance were seen in both groups, but there wereno differences between groups. This finding wasinterpreted to reflect the contamination of thevillage environment with iodine, probably due tourinary excretion by treated children. Shrestha(1994) conducted an iodine supplementation trialamong schoolchildren in Malawi and found thatsupplementation resulted in significant improve-ments in psychological development.

These results suggest that in iodine deficientareas, supplementation during pregnancy is criti-cally important. Evidence for the effectiveness ofiodine supplementation during childhood onpsychological development is mixed, but supple-mentation of school children, particularly girlswho will be mothers, is recommended.


1 All references marked with an asterisk are includedas cited in Hetzel, 1989.

Iodine supplementationduring pregnancy is critical,

particularly in iodine-deficient areas.


24

Iron

Iron deficiency anaemia (IDA) in infants andyoung children is associated with significantlylower scores on psychological tests (Pollitt, 1993a;Lozoff et al., 1991; Lozoff, 1998). Pollitt notedthat deficits of 0.5 to 1.5 standard deviation unitsin scores on infant development scales or IQ testsof children have been found quite consistentlyacross studies and age groups. Moreover, theseeffects of iron deficiency anaemia during infancyare associated with lower developmental testscores at 5 years of age (Lozoff et al., 1991) andhave been shown to have effects during schoolage in studies in France, Israel, and the U.S. (sum-marized in Lozoff, 1998), even controlling forother differences between children and families.Studies of less severe iron depletion have foundpsychological impairment less consistently

(Idjradinata andPollitt, 1993; Soe-wondo et al., 1989).

The question ofwhether the conse-quences of irondeficiency anaemiacan be reversed orprevented is criticalgiven the consistent

findings of long-term effects of IDA. The data re-garding the reversibility of psychological deficitsamong young children with IDA does not pro-vide a clear answer (Lozoff, 1998). Some studieshave found significant improvements on somedevelopmental test scores after supplementation(Oski and Honig, 1978; Idjradinata and Pollitt,1993; Soewondo et al., 1989; Aukett et al., 1986).Other studies have not (Lozoff et al., 1991; Lozoffet al., 1982; Walter et al., 1989). In five more care-fully designed studies, one has shown a strongpositive effect (Idjradinata and Pollitt, 1993), twostudies have shown effects only among childrenwho have clearly improved in iron status aftertreatment, and two showed no effect aftertreatment regardless of hematological status(summarized in Lozoff, 1998). On the other hand,most studies have suggested that iron supplemen-tation among school-aged children leads toimproved school performance (Watkins andPollitt, 1998).

The second key question is whether preven-tion through iron supplementation can be effec-tive in eliminating cognitive delays by limiting irondeficiency anaemia. Moffatt et al. (1994) randomlyassigned iron-fortified or non-fortified formula to

a group of low-income infants in Canada andfound that the fortified formula prevented thedecline in psychological development scores seenamong control children. The scores of thechildren in the control group improved at 15months such that differences between the groupsdisappeared. No effects on mental scales wereseen. Results of an investigation in Chile shouldprovide additional answers to this question(Lozoff, 1998).

In iron intervention studies, it is important tocontrol for compliance because the recipients ofsupplementation suffer more side effects and aretherefore less compliant than the control group.For example, Aukett et al. (1986) found a muchhigher impact of supplementation when compli-ance was statistically controlled.

How iron deficiency affects psychological de-velopment is under debate. There may be directeffects on brain development. Lozoff (1998) andRoncagliolo et al. (1996) have proposed that irondeficiency during the first year of life, when thebrain grows most rapidly, may result in perma-nent harm to psychological functioning. Theysuggest that iron deficiency may result in inad-equate myelination of the central nervous system.Children with IDA have also been found to befearful, wary, hesitant, restless, and less happy inseveral studies, which may also affect test perform-ance and learning (Lozoff, 1998). It is still notclear whether IDA affects specific cognitiveprocesses (related to different types of learning,problem complexity, information load, etc.), gen-eralized intelligence, or attention and motivationalfactors that affect test performance (Lozoff, 1989;Soewondo et al., 1989; Pollitt, 1993).

Iron supplementation interventions that suc-cessfully improve the iron status of infants andyoung children with iron-deficiency anaemia canimprove performance on tests of psychologicaldevelopment, although not always to the level ofchildren who have not suffered anaemia. From apublic health perspective, this is importantbecause a substantial proportion of children indeveloping countries experience iron deficiencyto some extent. Lozoff (1998) concludes, “eventhough questions about causality and reversibil-ity with treatment remain, it appears that iron-deficiency anaemia identifies children at risk forlong-lasting developmental disadvantage relativeto peers.” (p. 179)

Iron deficiency anaemia ininfants and young children

is associated withsignificantly lower scores

on psychological tests.

25

Zinc

A review of the few studies available on the im-pact of zinc supplementation (Golub et al., 1995)found little evidence of an effect on psychologi-cal development. However, Friel et al. (1993)reported an impact on motor development scoresof very low birthweight (<1500 g) babies supple-mented with zinc. Zinc supplementation forperiods of one to seven months has resulted inincreased activity levels among children aged12–13 months in India (Sazawal et al., 1996) andinfants aged 6-9 through 11–14 months in Gua-temala (Bentley et al., 1997). The relationshipsamong zinc status, activity patterns and psycho-logical development merit further investigation.

Other micronutrients

Evaluations of direct effects of other micronutri-ent interventions on psychological developmenthave not been reported. It is reasonable to assumethat the well described disability and morbiditythat results from some micronutrient deficiencieswould have effects on learning ability and accessto education (Levin et al., 1993).

Effects of breastfeeding on psychologicaldevelopment

A number of recent studies have found signifi-cant associations between breastfeeding and psy-chological development of children (de Andracaet al., 1998). However, identifying the mecha-nisms for these associations is difficult. First,breastfed and non-breastfed infants differ on manycharacteristics that may be associated with thechoice of whether to breastfeed, such as degreeof prematurity or perceived weight (Doan andPopkin, 1996). Second, the formula given to non-breastfed infants varies considerably by study,depending on the kind of formula available whenthe study was performed. Third, mothers whochoose to breastfeed differ on educational as wellas personality dimensions from those who choosenot to breastfeed, and these subtle differences areprobably associated with parenting styles thataffect psychological development of children.Finally, the act of breastfeeding itself may havesignificant impacts on children’s development.Because breastfeeding mothers tend to differ fromnon-breastfeeding mothers on hard-to-measurevariables, and because breastfeeding cannot berandomly assigned, it is doubtful that a conclu-sive test can be designed to show breastfeeding

improves psychological development (de Andracaet al., 1998).

Data from seven studies of the effects ofbreastfeeding on psychological development dur-ing infancy and at later ages were summarized byde Andraca et al. (1998). These studies indicatesignificant differences in the test scores of thechildren ranging from 4 to 10 IQ points, but thesedifferences diminish, and in some cases disappear,when differences in families’ education and eco-nomic level are statistically controlled (e.g. Roganand Gladen, 1993, Morrow-Tlucak et al., 1988;Bauer et al., 1991). Given the limitations of re-search designs to test the hypothesis conclusively,the results are open to alternative interpretations(Uauy and de Andraca, 1995).

Some mechanisms proposed to explainpsycho-neurological differences between breastfedand artificially-fed children relate to the presenceof components in breastmilk such as long-chainlipids that are important for structural develop-ment of the nervous system (Lanting et al., 1994).Another proposedmechanism is im-proved mother-childcommunication andinteraction, whichlead to enhancedpsychological devel-opment (de Andracaet al., 1998).

In general, benefi-cial effects of feedinghuman milk to chil-dren appear to be more evident for lowbirthweight pre-term infants than term infants.Lucas et al. (1992) examined the effects ofbreastmilk on pre-term infants, independent ofbreastfeeding. Infants (< 1,850 g) fed breast milkby tube in early life had higher cognitive scores at18 months and at 7–8 years of age than thosewho did not receive breast milk. Three groupswere defined: infants whose mothers chose toexpress breastmilk for feeding, infants whosemothers chose not to give breastmilk, and a thirdsmall group of infants whose mothers wanted toprovide breastmilk but were unable to do so. Itshould be noted that the formula was not forti-fied to the level that is currently recommended.This study can be characterized as a plausibilitystudy that attempted to control for other possibleexplanations statistically when random assign-ment was not possible. Despite efforts to controlfor socioeconomic status and mother’s education,the authors admit that their procedures might not


The act of breastfeedingitself increases the quality

of the mother-childinteraction, and impactspositively on the child’s

development.


26

fully control for differences in parenting andgenetic endowment. However, they argue that theevidence for a role of breastmilk in psychologicaldevelopment is strengthened by the demonstra-tion of a dose-response relationship between theproportion of diet supplied by breastmilk and theoutcomes. Further, children whose motherswanted to express breastmilk but were not suc-cessful did not differ on test scores from childrenwhose mothers chose not to provide breastmilk.Therefore, the authors suggest that breastmilk,rather than individual differences between moth-ers based on breastfeeding choice, should explainthese results.

A definitive test of the psychological effects ofbreastfeeding has not been possible given the dif-ficulties in constructing a probability design notedabove. However, studies have shown that infantsare highly sensitive even in the first week of lifeto stimuli related to breastfeeding. For example,they can differentiate the smell of their ownmothers from other women, particularly the smellof breastmilk (Porter, 1989; Varendi, Porter andWinberg, 1994). They respond to visual contactwith the mother, which is more frequent duringbreastfeeding than bottle feeding (de Andraca etal., 1998). The hormonal system of the motheralso responds to the stimulus of sucking and child

presence (Jellife and Jellife, 1978). Thus, it is likelythat the act of breastfeeding itself increases thequality of the mother-child interaction.

Nutrition education and growth promotion

To the best of our knowledge, there have not beenany efficacy trials of nutrition education or growthpromotion programmes in which psychologicaldevelopment was examined as an outcome. Tothe extent that such interventions are successfulin improving diets of undernourished children,one might expect an effect on psychological de-velopment.

Interventions that provide information andmotivation to improve child feeding practicescould have an effect on child developmentthrough increasing responsive, positive inter-actions between the caregiver and the child (Engleand Ricciuti, 1995). A few programmes have be-gun to integrate parenting education or activitiesto improve caring behaviour into nutrition edu-cation, and an example is discussed in Chapter 4.Research trials of the impact of nutrition supple-mentation implemented in combination with earlychildhood development activities are reviewed inChapter 5.

27

ducted to show the effect of supplementary feed-ing of pregnant women on intrauterine growth.A recent review of seven trials shows that supple-mentation was associated with increases inmaternal weight gain and mean birthweight, anda decrease in the number of small-for-gestationalage (SGA) babies. A recent community-based trialfound that supplements delivered to pregnantwomen in Gambia through a primary health caresystem resulted in a significant increase in meanbirthweight. This effect was particularly markedin hungry season, when the rate of LBW wasreduced by 33 percent (Ceesay et al., 1997).

Susser (1991) concluded that the expectedmean change in birthweight associated with foodsupplementation is 300–400 grams in a faminesituation, 50–90 grams in undernourishedpopulations in developing countries, and about40 grams in socially-deprived populations indeveloped countries. Pinstrup-Anderson et al.(1993) estimated that increased energy intake canimpact birthweight by 8–34 grams per 100␣ 000kcal ingested. The latest review of the evidencefrom randomized, controlled trials concludes thatthe mean change in birthweight from supplemen-tation is about 100 grams (de Onis et al., 1998).

How supplementary feeding affects birth out-come is unclear. Susser (1991) found that mater-nal diet had a stronger effect on birthweight thanmaternal weight gain had. While some reviewers(Kramer, 1993; Gülmezoglu et al., 1997) cite thelack of a consistent relationship between theimpact of supplementary feeding and thedegree of pre-pregnancy undernutrition, others(Pinstrup-Anderson et al., 1993) note that agreater benefit is seen among women who aremore malnourished.

One explanation for the lack of consistent con-clusions is that, even among undernourishedwomen, the extra intake from the supplementsaffects the mother and foetus differently. Supple-mentation of moderately malnourished womentends to increase birthweight without muchimpact on maternal weight, apparently because

Impact of nutrition interventionson physical growth

This chapter examines the positive impacts thatnutrition interventions have on measures of

physical growth, such as birthweight, weight-for-age, weight-for-height, height-for-age, or incre-ments in weight or height. Physical growth as anoutcome is important because of the evidencelinking poor growth to subsequent morbidity,

mortality, and lowerperformance in workcapacity and schoolachievement (Mar-torell, 1995). Ouremphasis here onphysical growth doesnot imply that growthper se is an essentialgoal of interventionsto improve childwell-being, but thatthe various measuresof physical growth are

valid indicators of past and present nutritionaldeprivation, and of future outcomes of functionalsignificance, such as school achievement oremployment.

Supplementary feeding of pregnant andlactating women

Energy intake has a well-established direct causalimpact on intrauterine growth (Kramer, 1987) andmaternal undernutrition is one cause of lowbirthweight (LBW). A clear example of this comesfrom a study of the Dutch famine during whichsharp drops in energy intake resulted in decreasesin maternal weight and subsequent reductions inbirthweight (Stein and Susser, 1975; Susser,1991).

As noted in Chapter 3, food supplementationhas been the main method used to demonstratethe effects of improved energy and protein intakeon physical growth. Several studies have been con-

4

Various measures ofphysical growth are valid

indicators of past andpresent nutritional

deprivation, and of futureoutcomes of functional

significance, such as schoolachievement oremployment.

Chapter

Interventions to supportphysical growth


28

energy is channelled to the foetus. Severely mal-nourished women, on the other hand, may notbe able to afford channelling the nutrition to thefoetus. As a result, supplementation has a greaterimpact on maternal weight gain and less onbirthweight (Olson, 1994; Winkvist et al., 1994;Winkvist et al., 1998). Because previous studieshave not examined this maternal versus infantimpact, the effects of supplementation are not yetwell described.

Most of the available studies used probabilityor high plausibility designs and were able toachieve a high degree of control over implemen-tation. Usually there are other factors that cannotbe fully controlled in field situations. The resultsof maternal supplementation studies are contin-gent on maternal nutritional status prior to sup-plementation. Take-home supplements are oftenshared with other family members, and on-sitefeeding may result in a reduced home intake.Thus, the net increase in intake is likely to beconsiderably less than the amount of supplementthat was provided.

The effects on child growth of maternal sup-plementation during lactation have not been stud-ied as extensively as has supplementation duringpregnancy. There is evidence that supplementa-tion of malnourished mothers can result inincreased breastmilk production, as measured byinfants’ intake (Gonzales-Cossio et al., 1991), andsuch effects are presumed to bode well for infantgrowth. However, an impact of supplementationof lactating women on infant growth has not yetbeen clearly demonstrated.

Supplementary feeding programmes forchildren less than 5 years

Community-based efficacy trials

Community-based efficacy trials with a highdegree of control over the intervention haveshown significant effects of food supplementationon physical growth. In Indonesia, 90 days of sup-

plementary feeding ofinfants aged 6–20months producedsignificant improve-ments in children’sw e i g h t - f o r - a g e(Husaini et al., 1991).A supplementationtrial in Jamaica found

significant effects on increments in weight andlength (Walker et al., 1991). This study is dis-

cussed further in the chapter on the efficacy ofcombined nutrition and ECCD interventions.

Evidence of both short and long-term effectsof supplementation on physical growth comesfrom follow-up studies of the supplementationtrials in Guatemala. Rivera et al. (1995) reportedthat children from villages in which a highprotein-energy supplement (atole plus milk) wasprovided to children from birth to 3 years of agewere taller and heavier both in childhood and ado-lescence than children from villages that receiveda low-energy, no protein supplement (fresco). Thegrowth advantage was statistically significant foradolescent girls but not for boys. Differences be-tween treatment groups at adolescence weregreater than differences at age 3 for weight, butless than differences at age 3 for height. Whenheight at age 3 was included in the model, differ-ences in height at adolescence disappeared, dem-onstrating that the growth advantage in later yearswas accounted for by growth gains during theperiod of supplementation in early childhood.

Supplementation appears to reduce the harm-ful effects of infection on physical growth. Analy-ses of data from a supplementation trial in Bogota,Colombia, found a negative impact on lengthamong unsupplemented children, but there wasno impact of diarrhoea on growth among the sup-plemented children (Lutter et al., 1989). The im-pact of supplementation was small and notstatistically significant among children with thelowest rate of episodes of diarrhoea. For childrenwith the highest rate of diarrhoea, supplementa-tion resulted in nearly a 5cm increment in heightcompared to the control group. Similar findingsfrom Guatemala and Peru support the view thatinadequate nutritional intake and diarrhoeal dis-ease work together in negatively affecting childgrowth (Lutter et al., 1992). In Jamaica, Walkeret al. (1992) found that reductions in lineargrowth associated with fever and lower respira-tory tract infections occurred only among theunsupplemented children.

Even in a carefully designed supplementationtrial for physical growth, full control over the in-tervention is not possible. The lack of control overthe net increase in consumption is a key factorlimiting the impact of food supplementation ongrowth in field-based efficacy trials. Due to sub-stitution for the usual diet (for on-site feeding)and sharing or other uses of the supplement (fortake-home supplementation), the actual increasein intake is always less than the amount of sup-plement provided. Other factors limit the abilityto demonstrate an impact of supplementary feed-

Supplemental feeding ofyoung children significantly

improves weight- andheight-for-age.

29

ing. These include the prevalence of undernutri-tion among the participants, the children’s age,the incidence of illness, and the potential for theadditional energy intake to be used in increasedactivity rather than for growth.

Food supplementation programmes

Although provision of supplementary foods hasbeen a component of numerous nutrition pro-grammes, relatively few of these have beendesigned in a way that permits an assessment ofthe direct impact on physical growth. Reviews ofthe extensive literature on the subject have notalways distinguished between evaluations ofefficacy and effectiveness (Beaton and Ghassemi,1982).

Habicht and Butz (1979) reviewed early large-scale supplementation trials to identify indicatorsfor evaluating nutrition programmes. Of the ninetrials reviewed, four showed significant impactson physical growth, and the remaining five didnot include information to verify that the supple-ment actually reached the target children. On theother hand, not all of those showing an impactincluded adequate controls for other potentiallycausal factors. This illustrates the difficulty infinding well-controlled efficacy trials of foodsupplementation, and the need for cautious in-terpretation of existing data. However, the authorsconcluded that there was some evidence that sup-plementation improved physical growth inpopulations with poor growth rates.

A recent review (Pinstrup-Anderson et al.,1993) restricted its scope to three studies inIndia, Colombia, and Guatemala that allowedestimations of the effect of food supplementationon physical growth. Comparison across studieswas difficult due to variation in the age of chil-dren, degree of deficiency, and duration of sup-plementation. However, the effect on height wasestimated to range from 0.8 cm to 5.0 cm per100␣ 000 kcal ingested, and the impact on weightfrom 40 to 800 grams per 100␣ 000 kcal ingested.The authors concluded that supplementation sig-nificantly improved growth in weight and height.

Rehabilitation and feeding ofundernourished children

Typically, community-based supplementary feed-ing programmes recruit children into a trial with-out regard to their initial height and weight. Thus,the samples include both well-nourished andmalnourished children. On the other hand, stud-

ies of rehabilitation programmes provide evidenceof the impact of supplementary feeding on chil-dren identified as malnourished. Much of whathas been reported about nutritional rehabilitationis based on programmes in hospitals or specialfeeding centres for malnourished children andtheir mothers. Early evaluations, including thestudies reviewed by Beaton and Ghassemi (1982),often suffered from a lack of an appropriate con-trol group. This design flaw makes it impossibleto distinguish the effect of treatment from changesthat would have occurred anyway as most meas-ures tend to move closer to the average over time(Kirkwood, 1988).

More recently, Rivera et al (1991) have dem-onstrated the efficacy of providing supplementaryfood to promote recovery from undernutrition.Guatemalan children aged 6–24 months who weremoderately wasted2 and consumed more than10 percent of the daily recommended intake ofenergy from supplements recovered over a threemonth period, and much of this recovery was at-tributable to the supplement. Children receivinga low-energy supplement were significantly lesslikely to recover.

Rehabilitation programmes address only acutecases and do not prevent the incidence of mild tomoderate malnutrition, as these children are notselected in screening procedures. However, if therehabilitation programmes include nutrition edu-cation and other support for improved caregiving,they may prevent reoccurrence and may also pro-tect other siblings.

Correction of micronutrient deficiencies

Improvement of physical growth is not the pri-mary objective of most interventions to improvemicronutrient status, but there is evidence thatdeficiencies of iodine, iron, zinc, and possiblyvitamin A can contribute to poor growth. Dem-onstration of the efficacy of a micronutrient in-tervention should be based on (1) evidence thatimprovement of status does improve physicalgrowth, and (2) evidence that interventions docorrect the deficiency.3 Most micronutrient inter-

CHAPTER 4. INTERVENTIONS TO SUPPORT PHYSICAL DEVELOPMENT

2 Moderately wasted at beginning of 3 month intervalif <␣ 90% weight-for-length, according to NCHS andWHO standards, recovered at end of interval if >␣ 90%weight-for-length.

3 Micronutrient deficiencies may also be corrected orprevented by encouraging greater dietary diversity;however, such efforts will not be covered here dueto the lack of efficacy trials.


30

ventions represent both preventive and curativetherapies (Levin et al., 1993) and impact may bedemonstrated by recovery of deficient individu-als or by reduced prevalence of poor micronutri-ent status in the population.

Iodine

Iodine deficiency diseases include a spectrum ofdisorders of which dwarfism, associated withmental retardation, is one of the most readily rec-ognized. The effects of mild iodine deficiency areless well understood, but potentially affect abroader population. Although retarded physicalgrowth is recognized as an outcome of iodinedeficiency, the focus of most studies has been onthe significant impacts on psychological develop-ment (as discussed earlier). In conditions wheredistribution and utilization of iodine-fortified saltis practical, the effectiveness of programmes thatpromote fortification and consumption of iodizedsalt is well established. Elsewhere, iodine regula-tion through injection or oral dosing has also beenshown to be effective (Hetzel et al., 1987).

Iron

Maternal iron-deficiency anaemia (IDA) andpregnancy outcome

There is some evidence that iron-deficiency anae-mia increases the risk of pre-term delivery andlow birthweight, but not small-for-gestational agebirths (Scholl et al., 1992). We were not able tofind efficacy trials, however, that demonstratedthat iron supplementation during pregnancyprevents low birthweight. Two reviews of the lit-erature on the factors that affect foetal growthidentified only two trials of routine iron supple-mentation. Neither trial demonstrated an impacton birth size or gestational age, despite the effectof supplementation in increasing maternal serumferritin and haemoglobin levels (Gülmezoglu etal., 1997; de Onis et al., 1998).

Childhood iron-deficiency anaemia (IDA)and physical growth

Iron deficiency anaemia in children is associatedwith mild growth retardation, which may berelated to the role of iron in metabolism, immuno-competence (Chwang et al., 1988), or aneffect of IDA on appetite (Levin et al., 1993).Several studies have found a positive effect of ironsupplementation on growth of schoolchildren

(Lawless et al., 1991; Chwang et al., 1988), butonly one published study with pre-school-agedchildren was identified. Aukett et al. (1986) sup-plemented anaemic children age 17–19 monthsfor two months and found a significantly higherrate of weight gainamong treated chil-dren, as comparedwith controls.

Numerous studiesof the impact of ironsupplementation onIDA in infants andpre-schoolers havefound that two tofour months of super-vised supplementation virtually eliminates IDA inthe treatment groups (Pollitt, 1993a; Lozoff et al.,1991; Soewondo et al., 1989). Shorter periods ofsupplementation were less effective. There is alsoevidence of the efficacy of food fortification. Inpast decades, when IDA was a more commonproblem among the paediatric population in theU.S., numerous studies demonstrated the effec-tiveness of iron-fortified formula in correcting orpreventing iron deficiency among infants(Andelman and Sered, 1966; Gorten and Cross,1964; Marsh et al., 1959; Ross Laboratories,1970). Similar effects of iron-fortified formula andcereals have been confirmed more recently in aChilean study (Walter et al., 1993; Pizarro et al.,1991).

Vitamin A

Vitamin A deficiency in experimental animalsresults in cessation of bone growth, weight loss,and loss of appetite (various references cited inLevin et al., 1993; West et al., 1988). Recent in-terest in the effects of vitamin A deficiency onmorbidity and mortality have led to numerouscontrolled intervention trials of vitamin A sup-plementation of pre-schoolers, some of whichhave measured physical growth as an outcome.

A field trial of vitamin A fortification of mono-sodium glutamate (MSG) in five programme andfive control villages in Indonesia found no effecton weight gain but a strong and consistent trendof increased linear growth among programmechildren (Muhilal et al., 1988). The mean incre-ment in height between baseline and final exam(after 11 months) was significantly greater amongprogramme versus control children ages 1 and 2(about 1 cm), with a consistent trend among chil-dren aged 3, 4 and 5 years.

Iron supplementation ininfants and pre-schoolers

for 2-4 months can virtuallyeliminate iron deficiency

anaemia.

31

In contrast, a randomized (no placebo) com-munity trial of vitamin A supplementation of chil-dren in Aceh, Indonesia, found an effect on weightgain among boys but no significant effect onlinear growth. Mean annual weight increment wasgreater among boys in programme villages thanin control villages and the differences were larg-est and statistically significant among boys aged4 and 5 years, with similar effects on arm circum-ference. There were no consistent statisticallysignificant differences in ponderal growth amonggirls, nor an impact on linear growth amongeither boys or girls. The authors note that thesegender-specific results are internally consistentwith findings of a greater impact of vitamin Asupplementation on mortality among boys (Westet al., 1988).

Three randomized double-blind (placebo-controlled) vitamin A supplementation trials havefound no consistent effects on physical growth,despite achieving improved serum retinol levels.In India, mean annual growth increments forheight, weight, and arm circumference did notdiffer between vitamin A-treated and controlgroups of pre-schoolers (Ramakrishnan et al.,1995). Trials in Ghana designed to assess effecton morbidity and mortality also included heightand weight measures. Effects on linear andponderal growth were inconsistent, and the onlystatistically significant finding was a mean weightgain of 3 grams per month less among thesupplemented children aged > 36 months, not

considered to be offunctional signifi-cance (Kirkwood etal., 1996). Similarly,no significant effectson growth were seenin a vitamin A supple-mentation trial withinfants and pre-schoolers in China(Lie et al., 1993).

Although both direct and indirect (via morbid-ity) mechanisms can be proposed to explain howthe vitamin A status of children could affect physi-cal growth, evidence to date does not show a con-sistent impact. It should be noted that thesesupplementation trials generally have excludedseverely malnourished or vitamin A deficient chil-dren.

The studies cited above, as well as other vita-min A trials, have demonstrated the efficacy ofsupplementation in improving vitamin A status,even when no impact on growth was seen. There

are many important benefits to justify widespreadimplementation of interventions to improve vita-min A status including reductions in prevalenceof xerophthalmia, decreased child mortality, andreductions in risk of severe illness. Whereas it isnot possible at this time to advocate vitamin Ainterventions as a means to improve growth, thisdoes not imply that such programmes are not aworthwhile investment to improve child healthand survival.

Zinc

A recent meta-analysis of the results of interven-tion trials of zinc supplementation in children(Brown et al., 1998) found small but highlysignificant effects on height (0.22 standarddeviation, or SD) and weight (0.26 SD) withsupplementation.The 25 studiesincluded in theanalysis werecommunity andclinical trials withadequate interven-tion designs, indeveloped anddeveloping coun-tries, including studies of low birthweight infants,pre-schoolers and school children, but excludingadolescents.

An effect of zinc supplementation on heightwas found only for interventions with groups ofchildren whose initial mean Z-score for height wasless than -2.0, ie., moderately malnourished. Inthese subsets of stunted children the average im-provement in height was considerable (0.49 SD).The effect on weight was found to be greatest ingroups with low initial plasma zinc concentra-tions. The authors concluded that “there is nowsufficient information to indicate that programmesto enhance zinc status should be considered as apotential intervention to improve children’sgrowth in those settings with high rates of stunt-ing and/or low plasma zinc concentrations”(Brown et al., 1988). UNICEF (1993) also con-cluded that zinc supplementation can improvegrowth among children with low zinc intakes orstatus.

The effects of zinc supplementation in preg-nancy have also been examined. There are fourrandomized trials, involving a total of 1␣ 400women. The available data provide no convinc-ing case for routine zinc supplementation duringpregnancy (de Onis et al., 1998). More trials are


Interventions to improvevitamin A status reduce the

prevalence ofxerophthalmia, decrease

child mortality, and reducethe risk of severe illness.

Zinc supplementation canresult in small, but highly

significant, effects on heightand weight.


32

needed in selected communities at high risk ofzinc deficiency, as well as in populations indeveloping countries where poor foetal growth isprevalent.

Supplementation is efficacious for increasingzinc intakes among vulnerable groups, althoughthe best form (a number of zinc salts are avail-able) and frequency are yet to be determined.Little information is available on the efficacy ofzinc fortification or the most appropriate foodvehicles. There are important issues that must beresolved related to the bioavailability of zinc andof other micronutrients whose bioavailability maybe affected by zinc (UNICEF, 1993).

In summary, the evidence on how well physi-cal growth is impacted by correcting micro-

nutrient deficienciesis mixed. The preven-tion of cretinic dwarf-ism through iodinesupp lementa t i onprenatally is incon-trovertible. Evidenceof an effect on growthis strong for zinc, insettings with highrates of stunting and/

or low plasma zinc concentrations. It is also likelythat iron supplementation may have an effect,possibly through improving appetite. Allen(1994), however, reviewed the literature onnutritional influences on linear growth andconcluded that no single nutrient supplement(including energy, protein, or various micro-nutrients) had a major, consistent effect on lineargrowth. She identifies several issues related todesign and sample selection that may help toexplain the lack of clear results, but notes thatsupplementation should have had an effect if aparticular nutrient was growth limiting. Sheconcludes with the suggestion that poor growthmay be a consequence of the multiple deficien-cies that result when children consume diets ofpoor quality.

Nutrition education to improvebreastfeeding and complementary feeding

Since the early 1960s, there has been a majorevolution in thinking about nutrition educationinterventions for young children. Earlier nutri-tion education emphasized information aboutappropriate foods, and communication aboutfeeding practices was generally limited tohygienic food preparation techniques. More

recently, nutrition education has also included agreater focus on feeding practices because of theincreasing evidence about the importance of feed-ing practices for nutrient intake.

Nutrition education and physical growth

Evidence from nutrition education trials, includ-ing the promotion of breastfeeding and improvedcomplementary feeding practices, shows that edu-cation interventions can affect physical growth.For example, a study in Bangladesh evaluated hownutrition education provided by community vol-unteers affected diet and growth of infants (Brownet al., 1992). Over a five-month period, childrenin the intervention villages gained an average of460 grams (0.46 SD) more weight for age thanchildren in the control villages. A significantlygreater percentage of the children in the controlgroup became severely malnourished, demon-strating a preventive effect of the nutrition edu-cation. The energy and protein adequacy of foodsgiven to children improved among interventionchildren and declined among controls. Althoughthe unit of randomization was the village, the unitof analysis was the individual. This design fea-ture of the study makes it more difficult toattribute benefits to the effect of the intervention.However, the study provides important evidencethat nutrition education can positively affectphysical growth, even in the context of chronicpoverty, undernutrition, and infection.

Evaluations of nutrition education interven-tions often measure their impact on caregivers’knowledge and practices related to child feeding,but provide little information on impact on physi-cal growth. However, Ashworth and Feachem(1985) reviewed data from twelve developingcountries and found promising indications thatwell-designed and implemented nutritioneducation can improveyoung children’s nutri-tional status even inpoor communities.They suggest that, inspite of flaws in theevaluation designs, theevidence shows thatnutrition education canbe effective, especiallywhen the approach todelivering appropriatemessages is throughinterpersonal communication by local workers,with reinforcement through mass media.

Poor growth may be aconsequence of the multiple

deficiencies that resultwhen children consume

diets of poor quality.

Highly effective nutritioneducation usesinterpersonal

communication by localworkers that is

subsequently reinforcedthrough mass media.

33

The conclusions reached by Ashworth andFeachem are supported in a recent review byCaulfield et al., (1998). The authors examined theresults of five efficacy trials, and 16 effectivenessevaluations of nutrition education programmesfor children from age 0–3 years, most of whichoccurred after 1985. Two questions were ad-dressed: (1) Under highly controlled researchconditions, what improvements in dietary intakesand growth of infants have been made? and (2)What improvements in dietary intakes and growthof infants have been achieved with programmesin developing countries? They concluded “despitevariability in the results, the majority of researchand programmatic efforts improved growth ratesby 0.10 to 0.50 SDs” (Caulfield et al., 1998). Thefive efficacy trials (Guatemala, Colombia, Jamaica,Indonesia, and Bangladesh) were able to increaseenergy intakes from non-breastmilk foods by70–300 kcal/d. The total improvement in intakewas lower because breastmilk declined as nutri-ent intake increased (Caulfield et al., 1998). Atthe end of the interventions, improvements innutritional status ranged from .25 to .46 SD forweight-for-age and .04 to .35 SD for height-for-age. Similar improvements were seen in theeffectiveness trials, both in intake and in changesin height and weight.

Nutrition education and feeding practices

Breastfeeding is the most widely and intenselypromoted feeding practice. Interventions to pro-mote breastfeeding have been evaluated prima-rily in terms of their effectiveness in improvingbreastfeeding behaviour rather than on physicalgrowth (Pinstrup-Anderson et al., 1993; Huffmanand Steel, 1995; Feachem and Koblinsky, 1984).Zeitlin (1996) cites descriptive studies in whichspecific breastfeeding behaviours (such as fre-quency of breastfeeding, halting other activitiesin order to breastfeed, and coaxing and active in-volvement with the child during a breastfeed) arepositively correlated with measures of infantgrowth. The benefits of exclusive breastfeedingin early infancy on reducing mortality and mor-bidity have been clearly demonstrated (Huffmanand Steel, 1995). An impact of breastfeeding pro-motion on infant growth is therefore likely.

In their review of nutrition education pro-grammes, Caulfield et al. (1998) identify some ofthe key factors that appear to contribute to pro-gramme success. First, successful programmes arecomprehensive and address the changing needsof the infant and young child. Second, they build

on current local practices. Third, they describenot only what but also how to feed infants.Although the design of the studies does not per-mit one to distinguish the effects of behaviouralchanges in practices from changes in foods, anumber of successful projects identified practicesthat may have served as barriers to adequate in-take. Messages developed to address these prac-tices focused on parental aspirations for childrenas a motivator, feeding frequency, using a sepa-rate feeding bowl, supervising feeding, feedingpatiently andpersistently, andcontinuing to feedeven when thechild appeared tobe full.

Successful nu-trition educationp r o g r a m m e sinvolve a multi-faceted approachincluding breast-feeding promotion, education about complemen-tary foods, and recommendations on how to feedchildren. In sum, it appears that nutrition educa-tion can be effective in improving nutrient intakesand physical growth of young children, and afocus on feeding practices as well as foods seemsto be an important component of these pro-grammes.

Production of complementary foods foryoung children

The concept of complementary foods, weaning foodsor specially formulated foods refers to a class offoods for infants and young children that are notpart of the regular family diet. These foods maybe prepared at home, in the local community orcommercially manufactured (Mitzner et al., 1984).The use of specially formulated mixtures in therehabilitation of malnourished children has beendocumented through clinical efficacy trials(Scrimshaw, 1980). However, less systematic at-tention has been given to assessment of the con-ditions under which different types of formulatedfoods are appropriate. For example, commerciallyproduced complementary foods have been criti-cized as a viable solution for complementary feed-ing in poor countries because they cannot bepurchased by the neediest families (Orr, 1977;Wise, 1980; Heimendinger et al., 1981). On theother hand, Scrimshaw (1980) notes thatIncaparina, one of the better-known formulated


Successful nutrition educationprogrammes address the

changing needs of the infantand young child, build on

current local practices, anddescribe not only what but also

how to feed infants.


34

foods, was intended to “provide a beverage withthe nutritional equivalent of milk in a culturallyacceptable form at as low a cost as possible forthe benefit of a sector of the population withmodest purchasing power, leaving to other pro-grammes the problem of reaching that part of thepopulation unable to purchase weaning foods.”

The evaluation of commercially producedcomplementary foods should be assessed in twosteps: first, in terms of its success in making theproduct available, and second, by indicators ofdecreased prevalence of undernutrition amongproposed users. We were not able to identify stud-ies of such foods that included both of these steps.

A number of the successful nutrition educa-tion interventions summarized by Caulfield et al.(1998) developed new recipes based on locallyavailable foods, using a technique of formativeresearch and recipe trials with the potential pro-gramme beneficiaries (Dickin et al., 1997). Theseprojects identified foods within the local area thatwould provide a more nutrient dense food andwere affordable. According to project reports,mothers were usually willing to try these newfoods as long as their children responded posi-tively to them. The factors that were identified asconstraints to continued use were time availabil-ity to prepare the food and affordability (Caulfieldet al., 1998).

Physical growth monitoring and promotion

Growth monitoring and promotion (GMP) refersto nutrition interventions that not only measureand chart the weight of children, but use theinformation on physical growth to counsel the par-ent in order to motivate actions that improvegrowth. Growth monitoring without the counsel-ling component could not be expected to have adirect impact on growth. The total package ofgrowth monitoring and promotion should beassessed. Unfortunately, most GMP programmeshave focused on the weighing and charting func-tions; and, not surprisingly, evaluations havefound little impact on physical growth. Growthmonitoring is a tool for decision-making relatedto interventions and can be expected to affect childgrowth only to the extent that it improves com-munication and actions to improve physicalgrowth (Griffiths et al., 1996).

Ruel (1995) laid out the theoretical framework,multiple purposes, and evidence of impact ofGMP. She found that GMP can be effective as aneducational tool, particularly in community-basedrather than clinic-based settings, but that there

was little evidence that screening through growthmonitoring was cost-effective or that growthmonitoring was appropriate for assessing theprevalence of undernutrition or for programmeevaluation. As a public health intervention, growthmonitoring was generally implemented withoutfirst being subjected to testing through efficacytrials (Step 7 in the research sequence identifiedby de Zoysa et al., 1998, and presented in

Chapter 2). Consequently, Ruel’s conclusions arebased mainly on assessments of programmeeffectiveness. It is difficult, therefore, to determinewhether the problems are in the basic conceptsor in the ways these concepts have been imple-mented.

Studies to test the first mechanism have shownthat growth monitoring facilitates awarenessamong health workers of the importance of physi-cal growth as an indicator of child well-being. Italso facilitates awareness among mothers (Ruel etal., 1990) when appropriate growth charts areused and growth monitoring is accompanied byappropriate nutrition education. In Lesotho, themothers with the lowest level of initial knowl-edge benefited most from adding nutrition edu-cation to growth monitoring (Ruel and Habicht,1992). Given the magnitude of investment ingrowth monitoring programmes, it is surprisingthat so little research has been devoted to howgrowth monitoring affects parents’ knowledge.

Only one study has examined the issue of GMPimproving the efficiency of programmes viatargeting and providing a focal point for inter-ventions (George et al., 1993). Unfortunately, boththe GMP and control groups improved, and therole of GMP as an intervention to improve serv-ices in usual circumstances was not able to betested.

As an educational tool, growth monitoring andpromotion (GMP) activities can:

■ Raise mothers’ and health workers’ awareness andknowledge of the importance of physical growthand practices that promote physical growth—hence motivating behaviour change.

■ Improve the efficiency of programmes viatargeting and by providing a focal point forinterventions.

35

Impact of other types of interventionson physical growth

In addition to the interventions that primarilyfocus on nutrition and physical growth, there areother types of interventions that may have animpact on child physical growth and/or psycho-logical development. Often the effects on physi-cal growth and psychological development are lessdirect and may not be viewed as priority goals ofthese interventions. As a result, individual childoutcomes may not be measured, although posi-tive effects are assumed to follow from improve-ments in food security, health care, sanitation, orincome of caregivers.

Because the effects of these interventions areless direct, their discussion here is much briefer.Detailed reviews have been prepared concerninga range of interventions and approaches relatedto child nutrition, physical growth and psycho-logical development (ACC/SCN, 1991b; Pinstrup-Anderson et al., 1993; Pinstrup-Anderson et al.,1995; Kennedy and Alderman, 1987; Myers,1992; Young, 1996). What follows is a samplingof available information on the impact of avariety of interventions that are intersectoral oroutside the usual domains of nutrition or childdevelopment sectors and that have been evalu-ated for their effect on physical growth and/orpsychological development outcomes. Evalua-tions of efficacy and programme effectiveness arecombined in this discussion. Some interventionsare directed to families, while others occur at thecommunity level.

Interventions to control disease

Improved water and sanitation

Improved water supplies are expected to affectphysical growth through a decreased incidenceof diarrhoeal disease. However, few studies haveactually measured this impact. Of those that have,two studies found no impact of improved watersupply on child nutritional status. In Malawi, childgrowth in families that used piped water was notbetter than in families using traditional sources,according to a study that evaluated child anthro-pometry one year before and one year after theintroduction of a piped water supply (Lindskoget al., 1987). In Bangladesh, a water andsani-tation intervention (hand pumps, latrines,hygiene education) reduced the incidence ofdiarrhoea among children under age 5 but hadno impact on physical growth (Hasan et al., 1989).Indicators of water and latrine use were not

significantly related to child nutritional status.On the other hand, Esrey et al (1988) com-

pared children in families that exclusively usedimproved water supplies to those in families thatused mixed sources, in a plausibility study con-ducted in villages in Lesotho with access to im-proved water sources. Differences in water use hadlittle impact on the physical growth of infants,likely due to breastfeeding. Among children age13 to 60 months, the exclusive improved waterusers gained on average 235 grams more in weightand 0.4 cm more in height than the mixed usersover a six-month period. The authors estimatedan expected improvement in physical growth overthe first five years of life of 4.4 cm and 2.3 kg,and showed that improved water supplies mustbe coupled with health education to encourageexclusive use of clean water for drinking and cook-ing, and maintenance strategies to ensure con-tinuous functioning of the water supplies.

An extensive review of hygiene interventionsthrough the early 1980s concluded that mostevaluations found little effect on diarrhoea or re-lated outcomes (Feachem and Koblinsky, 1984).However, more recently, a community-based in-tervention to improve hygiene practices in ruralBangladesh was successful in reducing rates ofdiarrhoea and prevalence of severe underweightin the intervention village (Ahmed et al., 1993).A community intervention trial in Kenya, using asimple method of solar disinfection of drinkingwater, demonstrated a significant reduction indiarrhoea morbidity (Conroy et al., 1996).

Control of disease through medical servicesand immunization

An intervention trial in Narangwal, India, com-pared the impact of nutrition services and medi-cal care (immunization and treatment of illness)singly and in combination, on child growth,morbidity and mortality (Kielmann et al., 1978).Interventions were assigned at the communitylevel, apparently not randomly, so results mustbe interpreted in light of potential selection bias.

Children’s weights at 17 months and abovewere significantly greater among children receiv-ing nutritional services (with or without medicalcare) than among children receiving only medi-cal care. However, medical care alone also had asignificant impact on change in weight relative tocontrols. Findings on height were similar, withthe effect of the nutrition intervention significantlygreater than the control, and an intermediateeffect of medical care. These findings indicate that



36

medical efforts to control infection can have apositive influence on physical growth.

Huffman and Steel (1995) reviewed evidencerelated to the impact of child survival interven-tions on nutrition. There is some evidence thatdietary management of disease interventions canimprove breastfeeding and feeding practices dur-ing and after illness, but no full-scale programmeevaluations measuring an impact on physicalgrowth were identified. Despite the relationshipof diarrhoeal and acute respiratory infections topoor growth, no studies were found that assessedinterventions to prevent or treat these illnesses interms of their impact on child nutritional statusor growth.

Three studies have examined immunization formeasles and nutrition or anthropometric status.Children in Gambia, Zaire, and Haiti, wereassessed before and after immunization, and im-munized children were compared with childrenwho had not received immunization on sched-ule. Although measles immunization mightappear likely as an intervention to improve physi-cal growth, given the magnitude of weight lossassociated with measles, the proportion ofchildren who contracted measles was too low topermit the demonstration of a significant differ-ence in weight. Therefore, it is not surprising thatno effect was found (Huffman and Steel, 1995).

Interventions to increasematernal education

There are numerous studies that show a signifi-cant association between women’s education andchild health and nutritional status. This effect maybe greater under certain conditions. One epide-miological study showed that maternal educationhad a strong positive association with child growthwhen household economic resources were low,but not precarious. When household foodresources were precarious, however, mothers’schooling was no longer associated with betterphysical growth (Reed et al. 1996). However, thereare no prospective studies that examine theeffects of education for girls on health and nutri-tion status of their children.

Interventions to increase householdfood security

As noted above, interventions to change the foodsupply to families may show smaller effects ongrowth than direct food supplementation becausethe causal links to intake and growth are more

distant. However, there is some evidence thatpositive effects of family or community-levelinterventions can be found. A few illustrationsfollow.

Food subsidies

Positive impacts on weight-for-age and energy andprotein intakes of pre-school children have beenreported for food subsidies in the Philippines(Garcia and Pinstrup-Anderson, 1987), and thereis evidence that a subsidized food ration improvedweight-for-age of children in Kerala State, India(Kumar, 1979, as cited in Pinstrup-Anderson etal., 1993). Kennedy and Alderman (1987) men-tion several examples of positive effects of subsi-dies on increasing food and calorie consumptionof households, but found little information onimpact on individuals.

Food for work

Kennedy and Alderman (1987) reviewed evidenceon the impact of food-for-work programmes,noting that the primary objectives have often beenthe creation of employment and the developmentof rural infrastructure, rather than improvementin nutrition. The review found limited data onimpacts on household food consumption andcited only one study that found a positive impacton child anthropometric status. Lack of impactmay be related to the short duration of most food-for-work programmes and the fact that the foodmay be sold by participants rather than consumed,often at a price lower than the usual wage for thelabour. Kennedy and Alderman conclude thatwhile such programmes may be effective inaddressing seasonal shortages of work and food,they do not represent a long-term solution, andthere is no evidence to prove that payment in foodrather than cash is more effective in addressingnutrition problems.

Agricultural production

The nutritional impact of agriculture-based inter-ventions is discussed in reviews by Pinstrup-Anderson et al. (1993) and Kennedy andAlderman (1987). In general, studies that evalu-ated child nutritional status show that the effectstend to be limited. The results of programmes toincrease cash cropping are mixed; some have re-ported a positive impact on nutritional status ofchildren, while others are have reported negativeor neutral effects (Kennedy and Garcia, 1993). A

37

major factor in the differential outcomes is thecontext of the intervention. When the profits fromcash cropping go the owners or managers of plan-tations, one would expect little benefit to thelabourers. In contrast, when the profits go to thefamilies who work the land, the results are morepositive, but may still be limited if funds that hadpreviously gone to women are redirected to themen in the families (Engle, 1993).

There are few systematic evaluations of theeffect of home-gardening interventions on foodconsumption by, or nutritional status of, childrenin low-income households (Kennedy and Alder-man, citing unpublished review by Brownrigg,1987). Home gardening, combined with a diver-sified nutrition education campaign, producedsignificant differences in vitamin A status of youngchildren (Smitasiri and Dhanamitta, 1996). Somereports claim an impact on dietary variety andmicronutrient intake, but none was identified inwhich an impact on child physical growth isreported.

Credit programmes

Micro-enterprise and credit programmes,especially those that target women, are often as-sumed to have a positive effect on children’s healthand nutrition through increasing maternal in-come. The only study found that measuredanthropometry as an outcome used an econo-metric plausibility design to analyse data fromthree credit programmes in Bangladesh, includ-ing the Grameen Bank (Pitt and Khandker, 1996).Despite finding positive effects on indicatorsrelated to income, assets, and consumption, noimpact was seen on the anthropometric status ofboys or girls (under 10 years of age). The authorsnote that anthropometric measurements weremade on a sub-sample such that the design wasless able to identify differences. Credit to womentended to have a greater impact on householdbehaviour than did credit to men, except in rela-tion to contraceptive use and fertility. There wasa strong positive effect of Grameen Bank credit towomen on girl’s schooling.

Evidence also suggests a positive impact ofprogrammes that integrate credit with health andnutrition education but no reports were found ofevaluations with adequate designs to demonstratethis relationship.

Effectiveness of nutritionprogrammes

Since several extensive reviews of evaluations ofnutrition programmes are available (Gwatkin etal., 1980; USAID, 1989; ACC/SCN, 1991; ACC/SCN, 1996), only the key findings will be sum-marized here. Effectiveness in improving childnutrition has been demonstrated in a largenumber of full-scale government and non-governmental organization (NGO) nutritionprogrammes, including the Tamil Nadu IntegratedNutrition Programme in India, the IringaNutrition Programme in Tanzania, the NationalFamily Improvement Programme (UPGK) inIndonesia, the Nutrition and Primary Health CareProgramme in Thailand, and the Applied Nutri-tion Education Programme in the DominicanRepublic (ACC/SCN, 1991; USAID, 1989;Pelletier and Shrimpton, 1994). While some suc-cessful nutrition programmes are relatively smallin scale, others are large state-wide or nationalprogrammes that have endured for many years.Micronutrient supplementation and/or fortifica-tion programmes, such as those in India, Bolivia,China, Indonesia, and Nepal, have also beenfound to be effective (Hetzel, 1989; USAID, 1989).

Although fewevaluations dem-onstrated effectsthat were clearlyattributable tospecific nutritionactivities, anACC/SCN report(1996) recog-nized that mostfull-scale nutri-tion programmes can reduce the prevalence ofmoderate and severe undernutrition in largepopulations by at least 1–2 percentage points peryear.

The components of full-scale nutrition pro-grammes are relatively consistent across countries,and usually include some combination of nutri-tion education, health services, and food supple-mentation or home gardening to improve thehousehold food supply (ACC/SCN, 1991). Thebasic nutrition intervention strategies are nowrelatively well delineated, and while no single setof interventions is best in all situations, the knownapproaches can be adapted to various contexts.Comparisons across programmes suggest that theprocesses of planning and implementation andthe quality of personnel are more critical to effec-


The processes of planningand implementation and the

quality of personnel are morecritical to programme

effectiveness than the specificcontent or type of intervention.


38

tiveness than the specific content or type of inter-vention (USAID, 1989; ACC/SCN, 1991). Thisconclusion should be tempered by the recogni-tion that there is an insufficient number of well-managed programmes to permit comparisons ofthe relative effectiveness of different types ofinterventions.

Key characteristics related to programmeeffectiveness include:

● Awareness of the importance and magnitudeof nutrition problems.

● Use of information for advocacy and planning.

● Community mobilization and participation.

● Training and support of staff, including com-munity-based nutrition workers.

● Needs assessment.

● Programmatic goals (involving all stake-holders) to guide planning and implemen-tation.

● Effective management and supervision (USAID,1989; ACC/SCN, 1991; Jonsson, 1995, as citedin ACC/SCN, 1996).

The importance of genuine community own-ership and involvement in decision-making,planning, and implementation is consideredparamount by most reviewers (ACC/SCN, 1996;Jonsson, 1995; ACC/SCN, 1991; USAID, 1989).As a result of increasing experience with andanalysis of its importance, there is now a greaterunderstanding of strategies for increasing the roleof the community. It is also recognized that bothcommunity-based and centrally-implementedactivities have different and complementarymerits, and a combination of approaches is likelyto be most effective (ACC/SCN, 1996).

More specific issues related to personnel,supervision, and management include:

● Clear definition of tasks.

● Manageable workloads for front-line staff andsupervisors.

● Reasonable worker-to-client and supervisor-to-worker ratios.

● A mix of existing staff, new staff, and locally-recruited community workers.

● A management information system to provideinformation at all levels for monitoring, evalu-ation, and decision-making (ACC/SCN, 1991;Pelletier and Shrimpton, 1994).

In addition to the characteristics already noted,flexibility and attention to availability of localresources (human and material) and existingadministrative structures have been identified ascritical to sustainability and replicability (USAID,1989; ACC/SCN, 1991).

Different contextual factors affect the relativeimportance of particular programme characteris-tics. For example, targeting is more importantwhere prevalence of undernutrition is low. Thesecontextual factors also affect the feasibility of im-plementing preferred approaches. For example,cultural and political considerations affect thesuccess of encouraging greater roles for commu-nity members, especially women. Thus, the fit ofa style of implementation with the context is alsocritical. In a review of nutrition programmes inSouth Asia, Jonsson(1995) identified con-textual issues related tosuccess, includingpolitical commitment,the presence of com-munity organizationsand charismatic com-munity leaders, lowgender discriminationas indicated by womenbeing literate, empow-ered and involved indecision-making, andcultural norms forfavourable child carepractices.

There is now sufficient evidence to concludethat large-scale nutrition programmes can beeffective in improving child physical growth andnutritional status. Many nutrition interventionsthat were first demonstrated to be efficacious inpilot projects or research trials are now full-scalenutrition programmes. Although there has beenlittle systematic attention to identifying thedeterminants of programme success (Schilling,1990; Pelto and Tuomainen, 1996), it appears thatthe factors that contribute to success reflect ap-proaches to decision-making, implementation,and management that are relevant to many typesof development programmes, rather than beingspecific to nutrition.

Programme successdepends on contextualissues such as political

commitment, thepresence of community

organizations andcharismatic community

leaders, low genderdiscrimination, andcultural norms for

favourable child carepractices.

39

Several different strategies for combining inter-ventions for physical growth and psychologi-

cal development have been instituted in the pastfew decades. This chapter examines interventionmodels that have been systematically evaluated,as well as other programme models that have yetto be scientifically assessed. We begin with a dis-cussion of the only three efficacy trials of com-bined interventions in developing countries thatwe have been able to identify. Two other relevantinterventions—one for severely malnourishedchildren and one for children in the U.S. withnon-organic failure-to-thrive (NOFTT)—are alsodescribed. We then examine the evidence fromeffectiveness evaluations. The chapter concludeswith a discussion of programmatic and policyissues affecting the feasibility of combined pro-grammes.

Efficacy trials of combined physicalgrowth and psychologicaldevelopment interventions

The scientific basis for combining nutritional andpsychosocial interventions is that commonmechanisms underlie both physical growth andpsychological development. Second, physicalgrowth influences the course of psychologicaldevelopment, and variations in psychologicaldevelopment can affect growth. Improvements inboth children’s nutritional intake and psycho-social stimulation should have a greater effect onpsychological development then either improve-ment alone.

As discussed in Chapter 1, there are threelevels where combined programmes interact andcan have greater, or additive, impact. The first isat the level of the child. A better-nourished childshould be more able to benefit from a stimulatingenvironment. A better-nourished child should alsobe more able to elicit more responsive caregivingthrough increased energy, verbal interaction, orhappiness. Second, additional effects can occurbetween the child and family. For example, improv-

ing practices related to nutrition and feedingshould also have an impact on practices relatedto psychosocial care, and vice versa. Both changesin parental practice involve increased attentionand ability or willingness to respond to children’sneeds. Third, in the design and delivery of pro-grammes, combined interventions should facili-tate both greater efficiency and better quality ofservices, and increased use of all services byparents (Erkel et al., 1994; Scott et al., 1998).

Three intervention trials have assessed theimpact of the combination of nutritional andearly childhood psychosocial interventions inpopulations with a high prevalence of undernu-trition. All provided nutritional supplementation,usually containing energy, protein and somemicronutrients. Psychosocial activities were im-plemented either through home-visits or centre-based programmes. A combined group wasincluded in each. The trials tested whether theeffect of the combined group was greater thaneither of the other two groups alone. Outcomesrelated to both physical growth and psycho-logical development were assessed. Table 2 showsthe characteristics of the projects.

Cali, Colombia

A study in Cali, Colombia, compared three typesof groups: food supplementation alone followedby a year of pre-school and health care; a com-bined programme of nutritional supplementation,health interventions, and pre-school; and amiddle-class control group. The study wasdesigned to test whether age of participation andduration of participation in pre-schools affectedimpact. For the nutrition intervention, childrenwere enrolled at ages 3, 4, or 5, followed by ayear of pre-school and feeding (about 17 pergroup). For the combined intervention, groupsof children were enrolled at ages 3, 4, 5, and 6years (about 50 per group). All of the childrenhad some pre-school experience in their final year,

5Chapter

Combined physical growthand psychological development

interventions


40

a minimum of 1 six-month period up a maximumof 4 six-month periods.

Positive effects were seen on a variety of meas-ures of psychological development, and differ-

ences persisted atfollow-up, one yearafter the interventionended. The longer theduration of the inter-vention, the greaterthe gains, and thesecognitive gains weremore evident atyounger ages (McKayet al., 1978). Unfortu-nately, it is not possi-

ble to separate the effects of age of entry into theprogramme and duration of exposure since ageand duration were confounded in the design.Effects of the intervention on physical growthshowed a similar pattern. Children exposedearlier and for a longer duration showed the great-est changes in linear growth and weight duringthe pre-school years (Pollitt and Perez-Escamilla,1995). The combined supplementation andstimulation had the greatest effects on psycho-logical development. Supplementation alone,before initiating the pre-school, had no effect onpsychological development.

Differences in psychological performance werestill evident when children reached age 9,

Table 2. Three combined physical growth and psychological development interventions

Location of Interventions

Cali Bogota Jamaica

Age of participants 3, 4, 5 or 6 years of age Prenatally 9–24 months of ageat enrolment (third trimester)

Nutrition Daily energy and protein Milk and protein mix for Milk-based formula, homecomponent requirement (75%), provided infants, plus vitamin and delivered, 750 kcal and

at centre, plus vitamin and mineral supplements; 623 kcal 20g protein/day (expectedmineral supplements and 20 g protein for all sharing); cornmeal for rest

children >␣ 1 year of family

Child psychological Centre-based pre-school, Home visits twice weekly; Home visits weekly todevelopment 6 hours/day, 5 days/week; activities to stimulate demonstrate toys andcomponent activities to stimulate psychological development, activities, with teaching of

psychological development with teaching of parents and parents and direct to childdirect to child

Other intervention Medical care Medical care Medical care

Duration of Minimum of 1 up to maximum Until 3 years of age 24 monthsintervention of 4 six-month intervention

periods

although the magnitude of effect was smaller(Sinisterra et al., 1979). However, no differencesin physical growth remained (Pollitt and Perez-Escamilla, 1995).

Bogota, Colombia

A study in Bogota was expressly designed to ex-amine the independent and combined effects offood supplementation and psychosocial stimula-tion (called maternal education) on physicalgrowth and psychological development. Fourgroups relevant to this discussion were defined:supplementation only, stimulation only, supple-mentation and stimulation, and no intervention.Treatment continued from the prenatal perioduntil age 3 years.

The study began with a nutrition component,and a home-based stimulation intervention wasadded later. Mothers were recruited in the lasttrimester of pregnancy. Children were randomlyassigned to intervention or control groups, basedon the city block in which they resided. Thenutrition component consisted of a package offoods delivered to the home as well as micro-nutrient supplements. The stimulation interven-tion consisted of home visits twice weekly for threeyears to provide stimulation to children andeducation to the mothers (Waber et al., 1981).

A total of 187 children were available for analy-sis at age 3 years. Psychological development was

Combined programmeshave greater impact on

psychological developmentwhen children enter them

at an earlier age andparticipate for longer

periods.

41

measured with the Griffiths test, a widely recog-nized measure of cognitive ability. Supplementa-tion had a clear effect on five subscales of theGriffiths test and the overall scale, and appearedto have a stronger effect than the stimulationintervention. There was no added benefit of thecombined supplementation-stimulation interven-tions on psychological development. The twodifferent intervention modalities apparentlyaffected different aspects of psychological devel-opment (Waber et al., 1981).

Scores declined with age, following the pat-tern that is typical for children raised in impover-ished environments (see Chapter 3). Homevisiting with stimulation activities appeared tohave a stronger effect on psychological develop-ment in the initial period (Waber et al., 1981).Behavioural changes in both mothers and chil-dren were an outcome of the home stimulationvisits. For example, the mothers who received

home visits became moreresponsive to their infants.Supplementation affectedbehaviour as well. Infantswho received food supple-mentation were more active(Super et al., 1981).

Supplementation had aneffect on physical growth,but stimulation did not.However, the combinationof supplementation andstimulation had a greater

effect on physical growth than supplementationalone (Super, Herrera and Mora, 1990). An evenlarger effect was found three years later, with noadditional intervention. In a follow-up study con-ducted when children reached age 6, children whohad received both stimulation and supplementa-tion were significantly taller than the other groups,suggesting that the interventions may have re-sulted in enduring changes in caregiving behav-iours (Super, Herrera, and Mora, 1990).

As in many studies, children who were the mostmalnourished were most likely to benefit fromthe intervention. For the children who were mostmalnourished at birth, the combined supplemen-tation-stimulation intervention significantlyreduced the probability of low weight-for-age,while the nutrition supplementation alonesignificantly reduced the probability of stunting.For less malnourished children, supplementationpositively affected physical growth, but there wereno additional effects of stimulation on their physi-cal growth (Super et al., 1990).

The Jamaica Study

A third study also examined the possible effectsof combined nutritional and psychosocial stimu-lation interventions on physical growth andpsychological development. The study wasconducted with 129 stunted (low height-for-age)9- to 24-month-old children in Jamaica. Childrenidentified as stunted were randomly assigned tocontrol, supplementation, stimulation, or com-bined intervention groups (Grantham-McGregoret al., 1991; Walker et al., 1991). There was alsoa non-stunted control group. The supplementedfamilies were provided with 1 kg of milk-basedformula per week, plus other food for family mem-bers, and the stimulation consisted of weeklyhome-visits with structured play sessions for chil-dren and caregivers.

Supplementation, with or without stimulation,had positive effects on weight and length incre-ments in the first six months that children werereceiving the intervention. Gains were sustainedin the second six months. Children who wereyounger and thinner benefited most from the sup-plementation.

All three treatment groups showed consistentimprovement each time they were tested acrossthe two-year period of the study. Children receiv-ing stimulation showed significant improvementson all subscales and total development quotient(DQ, a measure of the rate of the child’s develop-ment), with the largest impact of stimulationoccurring at younger ages. The effects of supple-mentation were mainly on locomotor skills in thefirst year, and then became apparent on overallDQ (Grantham-McGregor et al., 1991). Thestimulation and nutrition interventions independ-ently benefited children’s developmental quotients(DQ), preventing much of the decline seen instunted controls.

The combined intervention resulted in thegreatest benefit to DQ (Grantham-McGregor etal., 1991). The children who received both stimu-lation and nutrition interventions performed thebest, and approached the level of the non-stuntedgroup (although still below the level of middle-class Jamaicans). The children who receivedeither supplementation or stimulation performedbetter than the controls but worse than the com-bined group. In contrast, the scores of the chil-dren in the control group declined initially, afterwhich they began to improve somewhat. Theauthors concluded that the effects of the combinedtreatments were additive, but not synergistic.

At a follow-up four years after the study ended,

CHAPTER 5. COMBINED PHYSICAL GROWTH AND PSYCHOLOGICAL DEVELOPMENT INTERVENTIONS

The combination ofsupplementationand stimulation

interventions has agreater effect on

physical growth thansupplementation

alone.


42

Tab

le 3. T

hree efficacy stu

dies of com

bin

ed in

tervention

s

a References are in Table 4.

Project andsam

ple

Cali,

Colom

biaa

333 low-incom

eurban fam

ilies;age 3 to 6 years;follow

ed toage 10 years.

Jamaica

a

129 stuntedchildren, 32 non-stunted controls;age 9–24 m

onths.to 21–36 m

onths;follow

ed toage 7 years.

Bogota,

Colom

biaa

187 poorurban fam

ilies;age prenatalto 36 m

onths;follow

ed toage 6 years.

Design

8 groups:4 w

ith food andcentre attend-ance, eachstarting asubsequent year(ages 3,4,5,6);3 w

ith food onlyat 3, 4, 5, w

ith1 year centreattendance foreach; 1 m

iddleclass group.

5 groups:supplem

enta-tion, stim

ulation,both, control,and non-stuntedcontrol;intervention fortw

o years.

4 groups:stim

ulation,supplem

enta-tion, both, andneither;intervention for3 years.

Nutrition

intervention

Centrefeeding

Foodsdeliveredat hom

e

Foodsdeliveredat hom

e

Psycho-socialintervention

Centreprogram

me

Weekly

home

visits

Twice-w

eeklyhom

e visits forchild stim

ulation

Was there an

effect onphysical grow

th?

Yes, for childrenexposed for longestand starting earliest(3 years); largestincrease in heightand w

eight duringpre-school period.N

o effects3 years later.

Yes, forsupplem

entation.N

o, forstim

ulationalone.

Yes, for supple-m

entation. Greatest

effect for combined

supplementation

and stimulation.

Was there an

effect onpsychologicaldevelopm

ent?

Yes, for number

of years in pre-school; greatereffects w

ith earlierentry and duration.Effects still seen atage 9. N

o effects offood w

ithout pre-school on psycho-logical developm

ent.

Yes, for stimulation

groups;yes, for supple-m

entation;greatest effectw

ith both.

Yes, for stimulation

and for bothstim

ulation andsupplem

entation.Supplem

entationaffected m

otor andoverall score;stim

ulation affectedlanguage.

Adequacy of

intervention

Verygood

Verygood

Verygood

Effects ofcom

binedprogram

mes

Combined supplem

enta-tion and stim

ulation hadgreatest effecton cognition. N

o effectsof supplem

entation onpsychological develop-m

ent without pre-school.

Age of treatment and

duration of treatment are

confounded in the design.

Additive effects were

seen on cognition.O

nly supplementation

affected physical growth.

No additive effects seen

on cognition. Additiveeffects on grow

th seen atages 3 and 6. Stim

ulationplus supplem

entation hadlarger effect on grow

ththan supplem

entationalone.

43

small global benefits in psychological develop-ment remained for the subjects who had receivedstimulation. However, the control group as wellas the supplemented groups had increased inheight relative to standards, suggesting a generalimprovement in nutritional status of children inJamaica during this period (Grantham-McGregoret al., 1997).

A combined intervention for severelymalnourished children

Although not a community-based efficacy study,the results of another controlled study in Jamaicaare relevant to the discussion (Grantham-McGregor et al., 1987). Children hospitalized forsevere undernutrition were enrolled in the study.This was not a combined intervention in the usualsense, but provided stimulating play experiencesfor children while in the hospital receiving nutri-tional and medical treatment, and afterwards.After hospitalization, the intervention continuedwith home visits for three years, during whichtime parents were provided with instruction andsupport on how to interact with their children inways to stimulate psychological development, butreceived no additional nutritional support. It wasconsidered unethical to assign treatmentrandomly within a hospital ward, so a non-intervention group was recruited from childrenhospitalized in the previous year and followedduring that time. Thus, the two malnourishedgroups were recruited during different periods. Asecond control group was composed of adequatelynourished children who were hospitalized forother, serious illness.

Psychological development indicators of thetreated children improved greatly (Grantham-McGregor et al., 1987). After two years, the chil-dren in the intervention group caught up tochildren from higher socioeconomic status homeswho were not malnourished (but had beenhospitalized for other illness). Performance thendeclined until a plateau was reached at about age7. The non-intervention group of malnourishedchildren did not improve, and the gap betweentheir performance and that of well-nourished con-trols did not reduce. There were few impacts ofthe home-visiting intervention on mothers’ child-rearing behaviour (Grantham-McGregor et al.,1994).

In a follow-up evaluation conducted 14 yearsafter the children had been hospitalized, signifi-cant differences in performance were still found(Grantham-McGregor et al., 1994). The interven-

tion group scored at a level midway between thewell-nourished controls and the non-interventiongroup. The children who had participated in thestimulation programme performed significantlybetter than the non-intervention children on anIQ test (the Wechsler Intelligence Scale for Chil-dren, WISC) and differences approached signifi-cance on the Wide Range Achievement Test.Moreover, differences in IQ and achievement testscores between the intervention group and thewell-nourished controls were not significant. Onthe other hand, there were no differences betweenthe two previously malnourished groups in bodymass index or height-for-age at the 14-yearfollow-up, which indicates that the damage tophysical growth incurred during the period ofsevere malnutrition was not recoverable with thehome visiting intervention. As would be expected,the well-nourished controls were significantlytaller. The authors conclude that the stimulationprogramme resulted in marked improvements inpsychological development, which were sustainedinto adolescence.

A combined intervention for children withnon-organic failure-to-thrive (NOFTT)

In an efficacy trial of a home-based intervention,Black et al. (1995) randomly assigned NOFTTchildren from low-income, inner city neighbour-hoods in Baltimore, Maryland (U.S.A.), to inter-vention and control groups. Over the 12-monthtrial, both groups received multi-disciplinaryclinic services and nutrition counselling. The treat-ment group also received home visits, which in-cluded general family support, such as helpingparents access welfare services or providingsupport or advice on relationship issues. As thesechildren were already falling far behind in physi-cal growth, it was expected that they would con-tinue to decline developmentally. However, thechildren in the intervention group, who begantreatment prior to 12 months, showed less de-cline in psychological development in infancy andless decline in language development than con-trols. Physical growth improved for both groups,independent of the home-visiting intervention.There were no significant effects of the interven-tion on parent-child interaction, but the home en-vironments of children in the home-visiting groupwere more child-centred, according to a standard-ized measurement tool.



44

Summary of evidence for combined effectsfrom efficacy trials

Table 3 summarizes the results from the threeefficacy trials. All three studies found that apsychosocial intervention significantly affectedpsychological development, and that nutritionintervention significantly affected physicalgrowth. Two of the three studies found that thecombination of the nutritional and psychosocialinterventions had a greater effect on psychologi-cal development than either intervention alone.The third study reported that supplementationand stimulation affected different domains ofpsychological development. Surprisingly, the thirdstudy reported that the combination of inter-ventions had a greater effect on physical growththan either one alone.

In almost all cases, effects were greater for chil-dren who were the most malnourished. Effectsalso tended to be greater for younger children;

both the psy-chosocial in-tervention andthe nutritionalinterventionwere moreeffective whenthe childrenwere underage 3 years.Several of thestudies also

found that the impact of the psychosocialintervention on parenting behaviour might havelong-term implications for these children and forothers in the same family.

Effectiveness evaluations of combinedphysical growth and psychologicaldevelopment programmes

Effectiveness evaluations examine the implemen-tation and impact of interventions under usualconditions, such as in full-scale or nationalprogrammes. Effectiveness evaluations providevaluable information about the impact of inter-ventions in specific contexts and help pinpointproblems. These assess not only programme out-comes, but also how well the intervention wasdelivered, sometimes called process evaluations.In contrast to efficacy trials, these evaluations aredesigned to maximise inferences about causal re-lationships between interventions and outcomes.

Our original intent was to review a range of

programmes that illustrated differences in design,implementation, effectiveness and context. Forexample, programme components may differ interms of the provision of food supplementationversus nutrition education, or in terms of educa-tional experiences for children through parentingeducation or centre-based activities. There are alsovariations in the ages of children to which serv-ices are directed, in staff characteristics such asprofessionals or community members, in linkagesto health services and other programmes, and inthe relative emphasis on nutrition or psycho-logical development. Initially, the criteria forselection were:

■ Explicit inclusion of activities aimed at pro-moting both physical growth (or other aspectsof nutritional status) and psychological devel-opment.

■ Full-scale implementation, defined as a pro-gramme that covers a substantial population,has been implemented over a sustained periodof time, and is not a pilot or research project.

■ Availability of evaluation data on physicalgrowth and/or psychological development out-comes.

■ Availability of cost and cost-effectiveness data.

Programmes designed to promote both physi-cal growth and psychological development haveincreased with the increasing world-wide inter-est in early childhood psychological development(Young, 1996). However, the third criterionseverely limited our choices of programmes toreview. We were able to identify only seven pro-grammes that evaluated impact on physicalgrowth and psychological development. Therewere so little data on the cost-effectiveness of pro-grammes in developing countries that thiscriterion had to be dropped.

The specific form of the combined interven-tions differs depending on the delivery mecha-nism. Table 4 shows the programmes that wereidentified, classified by their delivery mechanismsfor the nutrition and psychosocial components.To facilitate comparison with the efficacy studies,described earlier, these projects are also includedin this and subsequent tables.

Five of the evaluated combined programmesare delivered in childcare centres and are designedprimarily for children ages 3–5. In three of theseprogrammes, the food supplementation compo-nent was quite limited. Two of the sevenprogrammes include home visits, although one is

The effects of combinedprogrammes to improve physical

growth and psychologicaldevelopment are often greatest for

children under 3 years old, andthose who are the most

malnourished.

45

primarily a centre-based approach. Only onefocuses on parent education for child psycho-logical development, and the evaluation data isfrom a small-scale pilot project.

The programme descriptions shown in theAnnex are brief summaries that describe objec-tives, components, personnel, training, super-vision and management, implementing orco-ordinating agencies, coverage, limited data oncost, and effectiveness measured by impact onphysical growth, psychological development, andother relevant indicators. As these are shortsummaries, the reader is referred to the primarysources of information for details. In large-scaleprogrammes, implementation often varies overtime and across sites because adaptations are madeto local conditions and needs, and approaches arerefined based on what is effective. Although thisflexibility and evolution probably contribute toprogramme effectiveness, evaluation is difficultwhen making comparisons across time andlocation.

Summary of additive effects fromeffectiveness and efficacy trials

Table 5 summarizes the results from the effective-ness evaluations. In each case, benefits and dis-advantages of the combined intervention methodsare noted. Unlike efficacy trials, it is usually notpossible to separate out the effects of nutritional

and psychosocial interventions on physical growthand psychological development. However, theseevaluations allow us to determine what kinds ofimpacts are seen.

In the seven programmes summarized, sixreported significant effects on various measuresof psychological development. On the other hand,only two reported clear effects on physical growth,and two found what could be interpreted aseffects on physical growth. There are several rea-sons for failure to find effects on physical growth:children were older than 3 years when enrolled,at which point nutritional inputs are unlikely tohave an effect; children were not at risk of mal-nutrition; or the programmes were primarilyfocused on early childhood stimulation.

Levels where interventions have anadditive effect

The results from the seven effectiveness evalua-tions and the three efficacy studies provide someinformation on the three levels where interven-tions have an additive effect: the child, the childand the family, and the design and delivery ofprogrammes.

The child

Of the three efficacy studies, two found evidencefor additive effects of the combined interventions


Table 4. Delivery mechanisms of combined nutrition and psychological developmentprogramme models in the review

Child-focused Family- or Parent-focused

Psychosocial Formal childcare centre Informal, home daycare Home visiting fromcomponent community-based health or

nutrition workers

NutritioncomponentSupplementary food Jamaica, Bogotadelivered to house

Supplementary food ICDS, Head Start, Cali PRONEI, PROAPE, HWBin feeding site

Nutrition education PANDAI, IFBECD

Efficacy Trials: Cali, Colombia (McKay et al., 1978); Bogota, Colombia (Waber et al., 1981);Jamaica (Grantham-McGregor et al., 1991)

Effectiveness Trials: ICDS India (ICDS, 1995); Head Start, USA (Currie and Thomas, 1995)HWB [Homes of Well-Being, Colombia] ( ICBF, 1997); PANDAI, Indonesia (Satoto, 1996)PRONOEI, Peru (Myers et al., 1985); IFBECD, Thailand (Herscovitch, 1997)PROAPE, Brazil (Myers, 1992)


46

on psychological development, and the thirdfound evidence for an additive effect on physicalgrowth. These findings suggest that combiningthe two interventions will have a greater effectthan either alone in a population at risk of mal-nutrition and poverty.

The child and the family

The evaluations show that caregiving was posi-tively influenced by the interventions. The posi-tive effects of stimulation on physical growth inthe Bogota study were attributed to changingfamily practices, such as how food was usedwithin the family. Programme evaluations gener-ally reported that the more involved the parents,the greater the impact on children, but self-selection may contribute to these effects. Anumber of programmes found effects on thequality of the home environment (e.g. Black etal., 1995) or on parental perceptions and attitudes(PRONOEI, PANDAI), but the analyses did notlink these changes with child outcomes.

Design and delivery of programmes

An important benefit of combined programmesis at the level of the delivery of the intervention,or its use by families. For example, qualitative datafrom programme evaluations noted an increasedparental interest, positive effects on the commu-nities, and a greater knowledge by staff who workwith families of malnourished children. Pro-gramme administrators also commented on howcombined programmes benefited parents (e.g.PANDAI). On the other hand, problems of in-creased workload, failure to focus equally on thenutrition and psychosocial components, and alack of understanding of the psychosocial com-ponent by general health staff were also noted.As is the case with most programmes, the mostdifficult problem is the training and supervisionof workers so that the planned programme isactually implemented. Thus, combining interven-tions requires commitment and patience.

Although classified as combined programmes,most of these interventions present the nutritionalor health component and the psychological com-ponent separately. The links between nutritionand psychological development are not specified,and are often not even mentioned in the pro-grammes. There is usually a nutrition componentthat provides food or recommends certain foodsand the timing of feeding, and a psychosocial com-ponent or pre-school intervention that describes

developmental stages, activities that children needfor learning, and strategies for stimulation. Thepsychological aspects of feeding and nutrition, orthe health and nutritional aspects of psycho-logical development, are infrequently mentioned.

However, this review clearly demonstrates link-ages. Nutrition and psychosocial interventionshave an additive effect for the child’s psychologi-cal develop-ment. Theskills requiredfor feeding andpsychosocialcare are simi-lar. In the firsttwo years ofthe child’s life,both feedingand psycho-social stimula-tion areimproved bywarmth, re-sponsivenessto children’s attempts to communicate, and aware-ness of developmental changes by the parents. Acombined approach would focus on responsiveparenting and its relationship to feeding andpsychosocial support. These interventions wouldbe more consistent with the parent’s holisticperception of the child than the current approachof dividing the child into discrete domains ofhealth, nutrition, and psychological development.For example, parents tend to use children’semotional condition as an indication of health. Acombined programme should also be moreaccessible to parents, easier to put into practice,and more effective.

Implications for models of combinedprogrammes

Evidence from programmes discussed in the lastsection demonstrates that it is feasible to com-bine interventions to improve physical growth andchild psychological development in large-scaleprogrammes. However, there are implementationproblems that need to be resolved.

These combined approaches are still relativelynew, and many existing programmes give greateremphasis to one component over another, ratherthan being truly integrated. Further systematicevaluations are needed to provide the knowledgenecessary for informed planning and expansionof programmes. This is critical for the design and

Because the skills required forfeeding and psychosocial care are

similar, nutrition andpsychosocial interventions have

an additive effect for a child’spsychological development. A

combined approach would focuson responsive parenting and its

relationship to feeding andpsychosocial support.

47


Wor

kers

rep

orte

d be

ing

over

taxe

d; s

pent

so

muc

htim

e on

food

pre

para

tion

that

ther

e w

as in

suffi

cien

ttim

e fo

r ho

me

visi

ts o

r EC

CD p

rogr

amm

e. H

owev

er,

wor

kers

kne

w w

hich

fam

ilies

had

mal

nour

ishe

dch

ildre

n, p

erm

ittin

g be

tter

targ

etin

g.

Prog

ram

me

impr

oved

imm

uniz

atio

n st

atus

in H

ead

Star

t chi

ldre

n.

Incr

ease

d pa

rent

inte

rest

and

par

ticip

atio

n in

GM

Ppr

ogra

mm

e.

Posi

tive

effe

cts

on c

omm

unity

dev

elop

men

t rep

orte

d,si

nce

scho

ol c

onst

ruct

ion,

mat

eria

ls, f

ood

prep

ara-

tion

and

teac

her

supp

ort w

ere

part

of a

com

mun

itypr

ojec

t.

Evid

ence

of p

rogr

amm

e co

nver

genc

e si

nce

spon

sore

d by

bot

h m

inis

trie

s of

hea

lth a

nd o

fed

ucat

ion.

Chan

ges

in p

aren

tal p

erce

ptio

ns re

port

ed. P

ilot

proj

ect j

udge

d su

cces

sful

, chi

ld d

evel

opm

ent

wor

kers

gai

ned

cred

ibili

ty, a

nd d

eman

d w

as c

reat

ed .

Nat

iona

l sca

le-u

p w

as m

uch

less

suc

cess

ful b

ecau

seof

too

man

y go

als

and

rapi

d im

plem

enta

tion.

Bette

r qu

ality

of d

ay c

are

prog

ram

mes

was

asso

ciat

ed w

ith b

ette

r gl

obal

wel

l-bei

ng o

f chi

ldre

n(h

ealth

, nut

ritio

n, a

nd p

sych

olog

ical

dev

elop

men

tco

mbi

ned)

.

Lack

of i

nvol

vem

ent o

f par

ents

in d

ay c

are

prog

ram

me

and

expe

ctat

ions

of p

rogr

amm

e w

ere

too

high

.

CDS

(3–6

yea

rs)

Indi

a, v

ario

us s

mal

lst

udie

s in

diff

eren

tar

eas

Head

Sta

rt(3

–5 y

ears

) USA

,na

tiona

l sam

ple

PAN

DAI

(0–3

yea

rs)

Indo

nesi

a

PRO

NO

EI(3

–6 y

ears

) Per

u,ru

ral a

nd u

rban

area

s

PRO

APE

(3–6

yea

rs)

Braz

il, n

atio

nal

IFBE

CD(0

–6 y

ears

)Th

aila

nd, i

nitia

lst

udy

in s

elec

ted

dist

ricts

HWB

(Hom

es o

fW

ell-B

eing

)(2

–6 y

ears

)Co

lom

bia,

nat

iona

l

Com

pare

d IC

DS

and

non-

ICD

S vi

llage

s in

one

stud

y; c

ompa

red

part

icip

ants

and

non

-pa

rtic

ipan

ts in

oth

ers.

Com

pare

d si

blin

gs w

hoat

tend

ed to

thos

e w

hodi

d no

t; re

tros

pect

ive.

Com

pare

d pa

rtic

ipan

tsbe

fore

and

afte

r pi

lot

stud

y.

Mat

ched

PRO

NO

EI a

ndco

ntro

l com

mu-

nitie

sse

lect

ed in

3 r

egio

ns.

Chi

ldre

n fo

llow

edth

roug

h fir

st g

rade

.

Com

pare

d pa

rtic

ipan

tsto

non

-par

ticip

ants

.

Com

pare

d co

mm

uni-

ties

sele

cted

in tw

ore

gion

s.

Sam

pled

onl

y ch

ildre

nin

cen

tres

, and

com

pare

d ou

tcom

esw

ith le

ngth

of e

xpos

ure

to th

e ce

ntre

, and

with

qual

ity o

f cen

tre.

Cent

re fe

edin

g

Cent

re fe

edin

g

Gro

wth

mon

itorin

g an

dpr

omot

ion

Cent

re fe

edin

g(m

othe

rspr

epar

eddo

nate

d fo

od)

Cent

re fe

edin

g

MoH

clin

ics,

cent

re fe

edin

g

Cent

re fe

edin

g

Tab

le 5

. E

valu

atio

ns

of E

ffec

tive

nes

s of

Com

bin

ed P

rogr

amm

es

Was

the

re a

nW

as t

here

an

effe

ctA

dequ

acy

Proj

ect

and

Nut

ritio

nPs

ycho

soci

alef

fect

on

phys

ical

on p

sych

olog

ical

ofEf

fect

s of

com

bine

dsa

mpl

eD

esig

nin

terv

entio

nin

terv

entio

ngr

owth

?de

velo

pmen

t?in

terv

entio

npr

ogra

mm

es

Info

rmal

ECC

Dce

ntre

Form

al c

entr

e

Hom

e vi

sits

and

com

mun

ityse

ssio

ns

Info

rmal

cen

tre

Info

rmal

cen

tre

Hom

e vi

sits

by

child

dev

elop

-m

ent v

olun

teer

s

Fam

ily h

ome

dayc

are

cent

res

run

byco

mm

unity

volu

ntee

rs

Yes,

in o

ver

30sm

all s

tudi

es.

No,

but

chi

ldre

nw

ere

not a

t ris

k of

mal

nutr

ition

.

No

effe

ct o

n gr

owth

,bu

t fee

ding

prac

tices

impr

oved

.

No

effe

ct o

n gr

owth

.

Mar

gina

l

Yes,

but

bot

h pr

o-gr

amm

e an

d co

ntro

lgr

oups

impr

oved

;m

ost i

mpr

ovem

ent

in p

oore

st v

illag

es.

Yes,

in p

oore

stgr

oups

, but

use

ddu

ratio

n of

exp

osur

eto

pro

gram

me

asin

depe

nden

tva

riabl

e.

Yes,

low

er s

choo

l dro

pout

and

IQ in

vill

age

com

pa-

rison

s. Y

es, i

n w

ithin

-vill

age

com

paris

ons,

but

thes

est

udie

s co

uld

not c

ontr

olfo

r se

lf-se

lect

ion.

Yes,

with

a g

reat

er b

enef

itw

hen

mot

hers

had

hig

her

IQ s

core

s.

Yes

Yes,

in 1

out

of 3

reg

ions

.N

o ef

fect

s on

sch

ool

repe

titio

n, b

ut e

ffect

sse

en o

n ea

rlier

age

of

scho

ol e

ntry

.

Yes,

in r

ates

of f

irst g

rade

repe

titio

n.

Yes,

incr

ease

d %

of

aver

age

and

brig

htch

ildre

n in

pro

gram

me

com

mun

ities

onl

y.

Not

as

a fu

nctio

n of

dura

tion

of e

xpos

ure

topr

ogra

mm

e.

Varia

ble

Not

rep

orte

d

Pilo

t pro

ject

Poor

qua

lity

in m

any

case

s

Not

repo

rted

Muc

hst

rong

er in

one

area

(with

an

NG

O)

than

the

othe

r

Varia

ble,

and

in m

any

case

s po

or

Refe

renc

es:

ICD

S In

dia

(ICD

S, 1

995)

; Hea

d St

art,

USA

(Cur

rie a

nd T

hom

as, 1

995)

; PAN

DAI

, Ind

ones

ia (S

atot

o, 1

996)

; PRO

NO

EI, P

eru

(Mye

rs e

t al.,

198

5);

PRO

APE,

Bra

zil (

Mye

rs, 1

992)

; IFB

ECD

, Tha

iland

(Her

scov

itch,

199

7); H

WB,

Col

ombi

a (IC

BF, 1

997)


48

implementation ofcombined interventionsthat are feasible, cost-effective, and appropri-ate to the context andcommunities for whichthey are intended. Theobjective of this section,then, is to review themodels or approaches to

programme design, and to discuss factors thataffect their suitability in different contexts.

Programmes that provide centre-basedpre-school education plus feeding

A centre-based approach to providing pre-schooleducation plus feeding is illustrated by the exam-ples of ICDS, Head Start, PRONOEI, Homes ofWell-Being, and PROAPE. The emphasis in theseprogrammes is on pre-school education—thenutrition-related goals are secondary. This modelhas strong advantages in terms of psychologicaldevelopment and school readiness of children,given the evidence that well-implementedpre-school education programmes improve theseoutcomes. Centre-based programmes also offerparents the advantage of childcare, although manydo not provide care for the full day. Various com-munity-based projects, such as the nutrition pro-gramme in Iringa, Tanzania, have developed groupchildcare and child feeding programmes at therequest of participating families, an indication thatthis approach is appropriate to local needs(Jonsson et al., 1993).

The agencies and local committees that areformed to oversee implementation may sharemanagement responsibilities. The staff of centresare typically recruited locally. To achieve signifi-cant developmental gains in children, these work-ers need to be well trained and supervised in theimplementation of appropriate activities. Systemsfor remuneration of workers vary from govern-ment or agency stipends to community supportto use of community volunteers, or a combina-tion of these. High staff turnover can be a prob-lem, probably due to low pay and lack ofadvancement opportunities.

It is important to be clear about the role of thefeeding component of centre-based programmes.If the children are ages 3 and 4, and are stunted,but not acutely undernourished, then substantialimpacts on nutritional status cannot be expected.However, there are other important rationales forproviding meals and snacks in pre-school settings.

First, feeding may be required simply because ofthe amount of time children spend in the centre.Second, the provision of nutritious meals andsnacks improves the quality of the children’s diet.Third, mothers’ participation in food preparationintroduces and reinforces nutrition education forcaregivers. Finally, learning activities for childrencan be built around food and feeding. On the otherhand, costs for food tend to make up a large pro-portion of recurrent costs of programmes thatprovide substantial supplementary feeding. Thisfact may constrain plans to add a feeding compo-nent to an existing early childhood care and de-velopment (ECCD) programme. The staff timeneeded for food preparation is another constraint.One approach to overcome this problem is to askparents to contribute time for this activity, as isthe case with PRONOEI. The use of in-kind foodcontributions from families may also help todefray costs, but may reduce the amount andquality of food available or discourage attendanceof children from the poorest families.

Systematic evaluationsare needed to provide

the knowledgenecessary for informedplanning and expansion

of programmes.

Conditions that support centre-based pre-schooleducation plus feeding programmes*

■ Concentration within a delimited area of suffi-cient numbers of children so that participation inthe activities will be convenient and feasible.

■ Parental interest in group activities for childrenand goals (e.g. school readiness), such thatfamilies are motivated to bring children regularly.

■ Perceived need (on the part of potentially support-ive institutions) for a focal point around which toorganize and combine other types of servicedelivery for families.

■ Suitable local programme staff, and adequatenumbers of trained personnel to supervise thesites.

■ An appropriate curriculum for centre-basededucation, or resources to develop one, and themeans to train teachers.

■ An appropriate, safe, and convenient site for thepre-school activities or the resources to build andfurnish a site.

■ Low-cost nutritious foods (local or donated)appropriate for child feeding, and cooking andserving facilities.

* The authors suggest that these are supportive conditionsbecause they are neither necessary nor sufficient, but can beregarded as issues to consider in the selection of approachesfor combined programmes.

49

Programmes that provide food andstimulation to children, and education onnutrition and responsive parentingfor parents

Two examples of this approach are the studies inJamaica and Bogota, Colombia. Both interventionsimproved physical growth and psychological de-velopment of participants. The effectiveness offull-scale programmes based on this model is notknown as these models have only been used inefficacy trials. However, there are full-scale pro-grammes that include a home-visiting component.The example of the ICDS Anganwadi programmeis relevant here in that children received some foodwhile attending the group activities, and addi-tional supplementary food was targeted specifi-cally to undernourished children and mothers.In the ICDS programme, home visits are notroutine for all families, but are specifically targetedto needy families. Another example of targetedhome visits is the Iringa nutrition programme inTanzania, in which growth-monitoring data areused to select families for home visits to delivernutrition education.

A home-based approach is preferred in manycircumstances. In some cultures, it is not consid-ered appropriate to take infants and young chil-dren out of the home, except under unusualcircumstances, such as caregiver illness or theabrupt weaning of a child when the mother ispregnant. Also, including very young children incentre-based programmes requires much higherstaff-to-child ratios. Home visits are a useful strat-egy for very early intervention and a way to reachfamilies who may be in great need but less likelyto attend centre-based interventions. The focuson parents and improving caregiving may havesustainable benefits and affect other children inthe family.

Programmes that combine nutrition andeducation on responsive parenting

Models that combine nutrition and education onresponsive parenting are a variant on the previ-ous one, but do not include the provision of foodor direct stimulation to children. Project PANDAIin Indonesia is an example of a parenting educa-tion programme that combines nutrition educa-tion with demonstration or training in childcareand psychological development. ICDS in Indiaalso includes these components. The limitingfactor for this review, however, was the lack ofevaluation data on such programmes.

In Project PANDAI, the parental educationcomponent was added to a growth monitoringand nutrition education programme that was al-ready being implemented at the community level.One advantage of an approach that emphasizesparental education is the feasibility of combiningit with existing nutrition education and growthmonitoring programmes just by adding new edu-cational components. The costs for developmentor adaptation of a curriculum and materials, fortraining staff, and for the additional staff timeneeded for implementation, monitoring, andsupervision should not be underestimated. Add-ing new information and activities to an on-going programme may be a more feasibleapproach to integration than creating centre-basedpre-schools or significantly expanding the paren-tal education of suchprogrammes. In fact,the PANDAI projectsuggests that broad-ening the scope of aprogramme may bemotivating for par-ents, staff, and com-munities.

Conditions that support home-based programmesthat provide food and stimulation to children, andeducation on nutrition and responsive parenting forparents

■ Intent to target infants and toddlers rather than3–5 year olds.

■ Intent to reach poor families who may not useservices due to constraints on time, income, orperceived need.

■ A dispersed population or cultural factors that donot favour children’s participation in centre-basedactivities.

■ Young children often found at home with aprimary care-giver; low rates of female employ-ment outside of the home.

■ Existence of a home-visiting health deliverysystem that could be expanded to include nutri-tion and/or ECCD components.

■ Lack of suitable spaces for groups of children tobe cared for.

■ Significant incidence of family food insecurity thatnecessitates food supplementation to improvechild nutrition.

Educational components onpsychosocial stimulationcan be added to existingnutrition education and

growth monitoringprogrammes.



50

Other approaches

The three models presented above are examplesof possible combined approaches based on pro-grammes that we have reviewed. Several othermodels were not reviewed because of lack of in-formation on their implementation and evalua-tion. They may, however, represent appropriateand effective approaches. There are many non-evaluated projects whose aim is to increase theresources or support systems for families throughactivities such as mothers’ groups, credit andmicro-enterprise development, or systems to linkhealth, educational, and community services.

A child development programme in Chile, forexample, which has introduced formal pre-schooland parental education in urban areas, imple-ments a different model in rural areas using moth-ers’ groups and a special curriculum to reachchildren who could not attend regular pre-school(Young, 1996). Centre-based programmes, suchas the network of pre-schools in Kenya and theInitial Education Project in Mexico, are startingto include nutrition components and reach outto parents and communities through activitiessuch as cooking demonstrations and nutritionmessages. Child-to-child activities can be imple-mented through local schools. Incorporatingteaching on health, nutrition and childcare intothe school curriculum can reach an importantaudience of school children who often care for

Conditions that support programmes that combinenutrition and responsive parenting education forparents

■ Nutrition education would be appropriate giventhe existing nutritional situation (i.e. undernutri-tion is related to non-optimal feeding practices inconditions where household food security is poorbut not limiting).

■ Parental behaviours related to child psychologicaldevelopment and parent-child interaction could beimproved.

■ Parental interest in investing time and effort inchild psychological development activities, butconditions are not conducive for a centre-basedapproach.

■ Children at highest risk are younger than pre-school age.

■ Existence of a community-based nutrition, health,or child psychological development programmewith the potential to take on new activities.

their younger siblings and may soon becomeparents themselves. The Little Teachers pro-gramme in Botswana is an example of such amodel.

A number of organizations that support wom-en’s employment have also developed childcareprogrammes, many of which are targeted to chil-dren from birth to age 3. The effect of these pro-grammes on children’s nutrition was summarizedby Mehra, Kurz and Paolisso (1992). Programmesreviewed included India’s Mobile Crèches,Senegal’s seasonal day care centres, and Ghana’sAccra Market Women’s Association. These pro-grammes tend to be closely connected to the wom-en’s workplace, and are under the governance ofthe women. According to project reports, therewere significant improvements in nutritionalstatus in over half of the projects. Children in daycare homes or pre-schools also had lower rates ofmortality and morbidity than their non-schooledmates. The authors conclude that the food itself,as well as the cleanliness and safety of the childcare locations, caused these significant effects onchildren’s health.

The psychological and educational interven-tions of the programmes reviewed by Mehra andcolleagues were not specifically evaluated. How-ever, the authorsobserved that thiscomponent of theinstitutional pro-grammes was notnearly as strong as thehealth and nutritioncomponent. The ratioof caregiver to childwas about 1:15 inboth institutional andhome day care. Thisratio contrasts withthe recommendationin the U.S.A. of 1:3 forchildren under age 3 year. Kits for educationalinstruction were sometimes available but were notalways used because workers lacked knowledgeon their use, or feared that the children wouldharm the materials.

The lessons that Mehra and colleagues drewfrom their review are that:

● Easy access to centres is a key determinant inthe use of day care programmes.

● There is a need for quality control and train-ing of caregivers in child psychological devel-opment, nutrition, health and hygiene.

Centre-based programmes,such as pre-schools, can

add nutrition componentsand reach out to parents

and communities throughactivities such as cooking

demonstrations anddissemination of nutrition

messages.

51


Table 6. Options for combined health, nutrition, and psychological developmentprogrammes for children from birth to age 3 years

Type of Delivery Delivery mechanismprogramme Specific mechanism for psychological(primary focus) programme for nutrition development Examples

Health Primary health care Nutrition information Health centre staff IFBECD in Thailandservices and preventative care and counselling provide information on (Herscovitch, 1997);

development; mile- Health Cards in Malaysiastones on health cards (Shah, 1995)

Primary health care Screening for Screening forscreening nutritional delay developmental delays

Maternity care, safe Nutrition information Parent support groups Brazelton trainingmotherhood promotion and counselling provide information on (Widmeyer and Field,

development in hospital 1981)or during prenatal care (experimental only)

Nutrition Growth promotion, Community based Community based PANDAI in Indonesiaprogrammes nutrition education weighing and counselling assessment of child (Satoto, 1996)

psychologicaldevelopment

Food supplementation Food delivered at home Home visiting for Grantham-McGregor, 1991;stimulation Waber et al., 1981

(experimental only)

Breastfeeding support Information in mother Information on psycho-groups, mother support support groups logical development,groups for child health support groupsand survival

Child care Child care facilities Food supplied during Curriculum provided by Homes for Well-Being,for working for working mothers attendance at centre trained community day Colombia (ICBF, 1997),mothers care provider; parent WHO/DFH Child Care

education or parent Programmes (1995)involvement programmes

Early child Community child Food provided at ECCD centre-based ICDS (1995), PRONOEIpsychological centre programmes centre programmes (Myers et al., 1985)developmentprogrammes Parent support groups Nutrition education either Parenting education either Philippines Parent

in groups or in home visits in groups or in home Effectiveness Programme;with community workers visits with community Honduras CCF Guide

workers Mothers Programme; ChileParents and ChildrenProgramme

Community PVO/NGO community Nutrition education based Informal pre-school, home Golf project in Philippines;development development Programmes on local knowledge, visits, parent-derived Laos ECFD Project

participative methods curriculum using (Arnold, 1998)participative techniques

Micro-credit Savings or income- Nutrition education Women’s groups discuss Save the Children infor women generating projects for within women’s groups children’s psychosocial Bangladesh (Arnold, 1998)

women needs, establish childcare centres

School-based School curricula and Nutrition education Psychosocial education PROAPE in Brazilteachers, or child-to-child (Myers, 1992)programmes

High-risk Rehabilitation programmes Nutrition rehabilitation Stimulation provided by Grantham-McGregor et al.,children for severe malnutrition in health centres home visits (1987); Black et al., (1995)

Hospital or TBA care for Nutrition, breast- Home visits for Combined Health andlow birth weight infants; feeding education psychological Development ProgrammeKangaroo methods development in U.S.A. (IHDP, 1990)

Other child risks: AIDS, Richter et al.(1990, 1994)chronic illness, etc.

General Mass media, videos Information Information Kotchabhakdi et al. (1987);information Black and Teti, (1997)


52

● Such programmes can significantly improvenutrition.

Other arenas for promoting combined physi-cal growth and psychological development areprogrammes with agriculture, sanitation, health,and environmental activities. Combined pro-grammes to improve nutritional, health, andpsychological outcomes are now being plannedat the national level in countries such as Ugandaand the Philippines. Designs include a strongemphasis on community initiatives and buildingon existing programmes and institutions. Momen-tum is mounting for projects that combine inter-ventions aimed at the whole development of thechild, in the context of family and community. Asummary of programme approaches is shown inTable 6.

Programmatic issues

So far this section of the review has addressed thebenefits of combining physical growth andpsychological development interventions frombiological and psychosocial perspectives. Thereis also a practical rationale for integration. A com-bined programme should achieve cost efficienciesover two separate programmes. Multiple use offacilities and transportation, for example, as wellas training and co-ordination of personnel to covera broader range of needs, should contribute tomore efficient resource utilization. Parental timecosts to use combined services should also belower, resulting in greater coverage.

Initially, we intended to examine whether thetheoretical rationale for cost efficiencies was borneout in practice. Ideally, one would assess the ad-vantages of combined programmes, and of oneprogrammatic model relative to another, by com-paring data on cost-effectiveness. The standardmethodology for doing so would compare the costof reducing child malnutrition by a given amount(e.g. a decrease in the prevalence of 10 percent-age points) across programmes with differentinputs.

The standard methodology is not adequate inthis case, however, because of our dual interestin nutritional outcomes and psychological devel-opment outcomes. These two outcomes are meas-ured in different units (e.g. kilograms versus IQpoints) and are valued differently by society,thereby precluding a simple method for combin-ing them. The method most commonly appliedin this situation is to convert both outcomes to acommon unit of benefit (monetary units) in the

form of cost-benefit analysis (CBA). The CBA notonly resolves the problem of mixed units of meas-urements; it also provides an estimate of the long-term returns to society for investment in suchprogrammes. Such analyses have not yet beencarried out for combined nutrition and ECCDinterventions.

Marginal costs associated specifically with anutrition or ECCD component are of particularinterest when considering the advisability of com-bining interventions. We could not obtain the datafrom combined interventions needed to estimatemarginal costs, but we consider this to be a highpriority for future research. Much of the marginalcost of an added intervention will be related tostaff time for training, implementing, super-vising and monitoring the new activities. Studiesof time spent on different activities in pilot pro-grammes, such as those conducted as part of theNarangwal Nutrition Study in India (Kielmannet al., 1978) would provide valuable insights intorelative and marginal costs of programme com-ponents. Costs reported in terms of time require-ments are more directly comparable thanmonetary values and are useful for programmeplanning in other contexts. Thus, there would beno need to convert time requirements to currencyvalues. Methodologies for cost analysis are beyondthe scope of this discussion, but it is worth not-ing that even without complete information onprogramme costs, estimates of inputs such as timeuse would be valuable for planning purposes.

Potential programmatic advantages

Delivering interventions to those who aremost likely to benefit

Underlying the argument for combined pro-grammes is the recognition that the children whoare most vulnerable to growth-faltering are alsoat highest risk of disruptions in psychologicaldevelopment, and that efficient targeting of pro-grammes requires reaching these children andtheir care-givers. Parallel periods of rapid physi-cal growth and psychological developmentduring the first two to three years of life, and thebroad effects of poverty on psychological devel-opment, mean that the same children in the samefamilies are the most likely to be at risk of pooroutcomes, and most likely to benefit from serv-ices or improved care-giving behaviours.

Reviewers of both nutrition and child psycho-logical development programmes frequently notethat potential impacts have not been achieved

53

because the youngest children are usually notreached (Beaton and Ghassemi, 1982; Gomby etal., 1995). The same could be said of failure toreach the poorest. It makes sense to provide com-

bined services oncethe effort has beenmade to establishcontact and motivateparticipation amongthose most in needor those families withinfants and youngchildren. It also re-duces the time coststo the family.

Combined pro-grammes could alsosimplify screening

procedures. For example, the association of poorphysical growth with poor developmental out-comes indicates that monitoring of physicalgrowth could be used to identify children in needof ECCD programmes, and measurement ofphysical growth is generally more feasible thanassessment of cognitive functioning. Zeskind andRamey (1979) found that low ponderal index (PI,height and weight) at birth predicted poor per-formance on infant mental scales at three monthsand again at 18 months for control children,whereas the low PI babies who received an in-structional day care programme achieved normalscores at 18 months. This suggests that foetal mal-nourishment is an indicator of infants likely tobenefit from intervention.

ECCD as an entry point and/or motivatingfactor for programme participation

A key principle underlying successful behaviourchange programmes is that they must address thefelt needs of the participants, and include the vitalconcerns that can motivate people to take action.Parents are generally very interested in promot-ing the psychological development of their chil-dren, particularly as societies become moretechnologically developed and school successgains importance (Richter, 1994). In contrast,chronic undernutrition is often not recognized,in part because often the norms for child size arefar below those that exist in better-nourishedpopulations.

Evidence suggests that caregivers may be moremotivated to participate in combined programmesthat provide information and activities related tochild psychological development. For example, a

child development chart designed as an assess-ment and education tool to be used in conjunc-tion with growth assessment in Indonesia provedso popular with mothers that attendance at growthmonitoring and promotion sessions increased(Zeitlin, 1996). Comments from mothers whoparticipated in home trials of improved feedingpractices in Tanzania and Ghana showed that thepositive impact of the new practices include theirperceptions about the effects on child behaviour.Mothers said that their children were more activeand playful, cried less, and enjoyed eating more(Dickin et al., 1997). Similar comments weremade in the evaluation of the HEARTH pro-gramme in Haiti. As a result of two weeks ofsupplementary feeding, parents reported that theirchildren were happier and more energetic(Basics, 1997). Caulfield (1998), in her survey ofnutrition programmes, reported that one of themessages linked to programme success was in-volving parents’ aspirations for their children. Theconnection between adequate feeding and subse-quent behaviour could motivate parents as muchor more than the effects on physical growth andhealth that are usually assumed to be motivationalfor parents.

Co-ordination of messages, materialsand approaches

Nutrition education programme materials,personnel, and contacts with families create ex-cellent opportunities for discussing appropriateways of interacting with and responding to chil-dren. Materials that deal simultaneously with childfeeding, care, and educational practices have beendeveloped by various programmes, and person-nel have been trained to use them and to workwith care-givers to facilitate the overall psycho-logical development of the child. However, manyof these are small-scale programmes and notevaluated.

Planners and personnel who are aware of bothnutritional and developmental needs are alsobetter equipped to avoid contradictory messages.An example from Sri Lanka illustrates the type ofconflicting recommendations that could beavoided. A child psychological developmentproject began to promote self-feeding of fingerfoods. The aim was to stimulate motor develop-ment and encourage more proactive feeding in apopulation that is characterized by high levels ofgrowth faltering in toddlers. Unfortunately, themessage conflicted with the recommendationsfrom nutrition programmes in the same area,


Children who are mostvulnerable to growth-

faltering are also at highestrisk of disruptions in

psychological development.The dual focus of combined

programmes can addressboth needs.


54

which implied that children should be fed with aspoon only. A combined programme, however,would take into consideration the impact of eachrecommendation on multiple aspects of the well-being of a child.

Potential operational constraints tocombined programmes

There may be disadvantages to combined pro-gramme models related to either operationalconstraints or cultural expectations in particularsettings.

At the level of the family, there are constraintson the amount of time and other resources avail-able to mothers and family members. Programmesmust not over-burden families with additionaltasks that make childcare more time-consuming.Programmes aimed at behaviour change mustrecognize the need to work in small steps andnot provide too much information or expect toomany additional activities to be adopted at once.

Programmes must ensure that they do not makeparents feel at fault for deprivations in the homeor community environment. Similarly, pro-grammes must be aware of the potential forcreating feelings of inadequacy about care-takingbehaviours in the home. Both Richter and Myersalso cite the dangers of disempowering parentsthrough ECCD activities that remove culturally-mandated responsibilities from the parent to thechildcare centre, or replace the role of the motherduring home visits (Richter, 1994; Myers, 1992).

Parents should recognize that the child’s ownbehaviour influences parental behaviour. Thefindings concerning the behavioural effects of asupplementary feeding intervention in Mexicoillustrate this issue (Chavez and Martinez, 1975).Parental behaviours changed in response to themore demanding and engaging behaviour of thesupplemented children. (This was perhaps dueto the larger body size and faster psychologicaldevelopment of the supplemented children, whichled to them being perceived as and responded toas older children, rather than babies.) The authorsalso noted that the greater activity levels requiredmore precautions to protect the safety of thesechildren, and that the supplemented childrenwere more mischievous, disobedient and evenaggressive.

Chavez and Martinez do not describe any nega-tive shift in parents’ attitude toward the children.To the contrary, they noted that the supplementedchildren were perceived as smart and good-looking, generating parental pride. In another

cultural setting, however, such behaviour mightbe considered inappropriate and lead to less will-ingness to participate in a programme or newactivity.

Cultural differences in child rearing may causedifficulties in programming. For example, pro-gramme credibility could be jeopardized if onecomponent is inappropriate to the context or pro-duces either no results or unacceptable outcomes.Harkness and Super (1996) noted that parents’interpretations of child behaviours vary acrosscultures, as does their understanding of children’spsychological development. “These culturally or-ganized understandings relate in systematic waysto action—including, for example, styles of talk-ing to children, methods of discipline, or seekingadvice from experts.” Such cultural belief patternswould also influence motivation of parents toparticipate in a child psychological developmentprogramme and the type of behavioural anddevelopmental outcomes expected from such aprogramme. For ex-ample, parents in-volved in PRONOEIreported that theirchildren learned tosing and dance in thepre-school, but theyhoped the childrenwould also learn toread (Myers et al.,1985).

The answer to thisconcern is to involvethe community inthe design and im-plementation of theprogramme to ensure that objectives are valuedby participants, activities are culturally-appropri-ate, and participation is convenient and withinhousehold time and resource limitations.

Combined programmes can provide a focalpoint for community action. Above all, it isessential that communities have an early and con-tinuing, meaningful role in the process of identi-fying needs and planning solutions. This not onlyensures appropriate and useful programmes fortheir clientele, but also community ownership anddecision-making, which are the foundation ofeffective, sustainable implementation.

At the programme level, problems to beavoided include overloading of the supervisoryand management systems and the personnel, thefacilities, and the people who are responsible forimplementing activities in the community. Exist-

The communityshould be involved in the

design and implementationof programmes to ensure

that objectives are valued byparticipants,

activities are culturally-appropriate, and

participation is convenientand within household timeand resource limitations.

55

ing personnel may not have the expertise for bothnutrition and child psychological developmentinterventions. It is obviously important not tosacrifice quality in the interests of combining nu-trition and child psychological development, andplanners need to develop clear strategies for train-ing, supervision, and monitoring. Tools, curricu-lum guides, and assessment measures are neededto assist communities in learning about psycho-logical development, assessing community needs,and developing or adapting a relevant curricu-lum.

Another potential source of difficulty is co-

ordination across departments or agencies. Myers(1992) concludes that rather than trying to forceorganizational integration, the emphasis shouldbe on convergence and ensuring that various serv-ices arrive at the same place. Strategies includeintegrating content of materials, planning collabo-ratively, focussing on key components, creatinginter-organizational activities (training pro-grammes, materials, or experimental projects),and building political will. Also, an emphasis onbottom-up planning and implementation will behelpful, given that integration at this level is lessproblematic.



56

Purpose of the review

This review explores the critical relationshipbetween nutritional status and psychological

development, and demonstrates the potential ofcombining interventions that enhance early child-hood development and those that improve childhealth and nutrition into an integrated model ofcare. The premises that underlie the review arederived from a strong research base and include:

● Malnutrition causes both poor physical growthand psychological developmental delays.

● Malnutrition—measured by growth faltering—is causally related to mortality in infancy andearly childhood, and interventions that reducethe incidence of malnutrition can significantlyreduce mortality.

● In infancy and early childhood, when childrenare fully dependent on others for their nutri-ent intake, the proximate causes of malnutri-tion are: (1) receiving diets of poor quality andinadequate quantity, and (2) inappropriatefeeding practices and behaviours, related tofood preparation, frequency and interactions.There are, in turn, multiple socio-economic,cultural and psychological determinants ofthese proximate causes.

● There is evidence that in many communitieswith endemic malnutrition, both feeding prac-tices and the selection of foods for infants andyoung children from the food sources that arepotentially available within the community arenot optimal.

● Many children in developing countries anddisadvantaged populations in industrializedcountries experience delays in psychologicaldevelopment that negatively affect their schoolperformance, their ability to maximise educa-tional opportunities, and their social function-ing later in life.

Key questions addressed by thereview

To encourage more widespread development,evaluation and expansion of combined interven-tions, the review addressed four key questions:

● To what extent and through what means canpsychological functioning be improved forchildren living in disadvantaged environments?

● To what extent and through what means canchild physical growth be improved in settingswhere chronic undernutrition is prevalent?

● To what extent and through what means cannutrition and psychological development beimproved simultaneously through combinedhealth, nutrition, and psychosocial interven-tions?

● Are there effective models for combined inter-ventions, and are these models feasible forimplementation on a public health scale?

To what extent and through what meanscan psychological functioning be improvedfor children living in disadvantagedenvironments?

Nutrition and education interventionssignificantly improve psychologicaldevelopment in disadvantaged populations.

● Model pre-school programmes have demon-strated significant impacts on psychologicaldevelopment. For example, children attendingpre-school centre-based programmes gain anaverage of about eight IQ points by the timethey are ready to start school. They are alsoless likely to repeat primary school grades orbe placed in special education classes. There isalso evidence that these programmes lead toimproved school performance in adolescence,reduced anti-social behaviour, and improvedemployment opportunities in adulthood.

6Chapter

Summary, conclusions andrecommendations

57

● Psychosocial interventions for low birth-weight infants have successfully prevented orameliorated developmental delays, a high riskfor these infants.

● Food supplementation of infants and youngchildren has significant impact on tests of psy-chological development. In infancy, the effectsof nutritional interventions are most apparenton motor skills, but for older children, effectsare seen on a variety of tests.

● Micronutrient interventions to prevent iodinedeficiency have dramatic effects on psycho-logical development, as well as the physicalstunting associated with this deficiency. Thereis also solid evidence that correcting iron defi-ciency anaemia improves learning.

● Breastfeeding is associated with moderatelyhigher levels of psychological development ina number of studies.

Significant improvements in psychologicaldevelopment are achieved through many differ-ent mechanisms:

● Child-focused interventions that providestimulation directly to children in communitychild care centres;

● Parent-focused interventions, implementedthrough home visits, which improve parentingskills, including skills in providing stimulationto children;

● Nutrition interventions for infants and youngchildren that are delivered through health serv-ices, nutrition services, social services, com-munity organizations, and other mechanisms;

● Micronutrient interventions for infants andchildren that are provided through food forti-fication or supplements;

● Food supplementation and/or micronutrientinterventions for pregnant women providedthrough health services, nutrition services,social services, community organizations, andother mechanisms;

● Joint-focused (direct services to child, indirectto parent) and combined (health and psycho-social) interventions that are delivered throughhealth services, nutrition services, social serv-ices, community organizations, and othermechanisms.

In efficacy trials, these interventions resultedin statistically significant improvements in meas-ures of psychological development and functional

performance. With respect to early childhood careand development (ECCD) interventions, child-centred actions have been more successful inshowing impacts on children than parent-focusedinterventions (although both are efficacious). Thisgeneralisation is based on evaluations of pro-grammes in developed countries (mainly theUnited States). In less developed countries wheremalnutrition plays a much greater role in devel-opmental delays, parent-focused programmes,with home visiting, have been quite effective.

To what extent and through what meanscan child physical growth be improved insettings where chronic undernutrition isprevalent?

Nutrition interventions significantly improvephysical growth in poor and malnourishedpopulations.

● Protein and energy supplementation in preg-nancy can reduce the incidence of intrauterinegrowth retardation and improve birthweight,with a mean increase of 100 grams.

● Food supplementation of infants and youngchildren has significant impact on physicalgrowth, ranging from 0.8 to 5.0 centimetreson height and 40 to 800 grams on weight. Foodsupplementation also prevents growth falter-ing in children who experience high rates ofdiarrhoeal disease, permitting them to achieveheights and weights that are comparable topeers who do not suffer frequent episodes ofdiarrhoea.

Other types of nutrition interventions alsohave documented impacts on physical growth,although it is more difficult to estimate their mag-nitude from presently available evidence:

● Interventions that change diets and feedingpractices improve physical growth. Well-designed and implemented nutrition educationprogrammes can prevent serious malnutritionin communities where children are at high risk,without needing to raise family incomes(Caulfield et al., 1998).

● Interventions to promote breastfeeding im-prove breastfeeding behaviours, and these, inturn, are linked to improved physical growth.

● Micronutrient interventions do not appear tosignificantly improve physical growth, with theimportant exception of iodine and possibly ofzinc. Well-controlled intervention trials with

CHAPTER 6. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS


58

zinc supplementation indicate the potential forsmall, but significant improvements in heightand weight in settings with high rates of stunt-ing.

● Intersectoral programmes to promote improve-ments in water and sanitation, agricultural pro-duction and food security can have positiveimpacts on physical growth if the interventionis appropriate to the social and economic con-text.

● Immunization activities, unless they are accom-panied by improved health services, generallyhave not shown an impact on physical growth.

Evaluations of large-scale nutrition pro-grammes have documented the feasibility andeffectiveness of interventions to improve physi-cal growth. However, for interventions to fulfiltheir potential, they must be well implemented,appropriate to the context, and directed to thosein need. There is general agreement among pro-fessionals in public health nutrition that the qualityof implementation and management is an impor-tant determinant of programme effectiveness.

To what extent and through what meanscan nutrition and psychologicaldevelopment be improved simultaneouslythrough combined health, nutrition, andpsychosocial interventions?

Combined interventions to improve bothphysical growth and psychologicaldevelopment have even greater impact indisadvantaged populations at risk ofmalnutrition.

● The stimulation, or psychosocial, componenthave significant effects on psychological devel-opment and language, particularly for youngerchildren. Generally, physical growth is notaffected.

● The nutrition component affects both physi-cal growth and psychological development.Nutrition supplementation initially producesstronger effects on motor than mental devel-opment among infants, but affects a wide rangeof skills when children are over age 2 years.

● Combined nutrition and psychosocial interven-tions have a greater impact on psychologicalfunctioning than either intervention alone.

● Caregivers with limited formal education indeveloped and developing countries can

acquire knowledge of nutrition, and feedingand parenting skills.

Combined interventions are delivered throughchild care centres, with feeding at the centre, orhome visits providing parent education for childpsychological development and food supple-ments. We point out that there are only a few well-designed community-based efficacy studies thathave investigated the effects of combined inter-ventions and more research is needed.

Are there effective models for combinedinterventions, and are these modelsfeasible for implementation on a publichealth scale?

Full-scale programmes that include bothnutrition and psychosocial components havebeen implemented throughout the world, andcontinue to increase.

● Interventions combining pre-school (stimula-tion and educational activities) with a minornutrition component significantly affect psy-chological development, but have limitedeffects on physical growth. Most of the formallyevaluated programmes use this model.

● The impact on physical growth is larger inprogrammes that include home visits andsupplementary foods for younger aged under-nourished children.

● Other promising models include using analready established growth-monitoring pro-gramme to teach parents about child psycho-logical development and to promote newparenting skills, and disseminating feedingrecommendations with information aboutpsychological development via local media.

● Combined interventions are likely to be moreefficient than separate interventions becausethey are intended for the same population andmake use of the same facilities, transportation,and client contacts. From an economic stand-point, the marginal costs are expected to below, relative to impact.

● From the perspective of the family, a combinedapproach increases access to services. It mayalso increase overall effectiveness because fami-lies who need early intervention often have avariety of risk factors (e.g. lack of maternal edu-cation, low birthweight, poverty), several ofwhich may need to be addressed.

59

Examples of programme models thatcan incorporate nutrition, health andpsychosocial components

● Incorporation of child psychological develop-ment into primary health care through the useof developmental milestones on health cardsand the inclusion of simple messages forparents on how to facilitate psychologicaldevelopment.

● Promotion and support of home-based groupchild care, combined with supplementary feed-ing for children of working mothers, sometimeswith a micro-credit programme.

● A child-to-child strategy in which older sib-lings learn skills to help improve the psycho-logical development, health, and nutrition ofpre-schoolers.

● Community development projects that usehome visiting and pre-school programmes asan entry point for other interventions such asincome improvements.

● Interventions with high-risk children, such aslow birthweight infants, that combine bothpsychosocial and nutritional care.

● Parent education courses and mothers’ groups,including breastfeeding support groups.

● Mass media programmes (radio, TV, videos)that target both physical growth and psycho-logical development.

Conclusions

Although the number of combined programmesthat have been evaluated is limited, such pro-grammes are effective, particularly if they areappropriate to the context. There are several con-ditions under which the greatest impact on growthand development are most likely to be seen.

● Interventions during the earliest periods oflife—prenatally, during infancy and early child-hood—are likely to have the greatest impact.Interventions to support psychological devel-opment after this particularly vulnerable earlyperiod, however, are also effective.

● The children in greatest need due to povertyor parents’ lack of knowledge are generally theones who show the greatest response to growthand development interventions. Identifying thefamilies and children who are mostly likely tobenefit from such interventions should im-prove outcomes. There is also evidence that

certain positive characteristics or resources offamilies (such as maternal education) contrib-ute to a greater response.

● Growth and development programmes thatutilize several types of interventions and morethan one delivery channel are more efficaciousthan those that are more restricted in scope.Types of interventions include nutrition edu-cation on diet and feeding practices, provid-ing supplementary foods or micronutrientsupplements, teaching parents about childdevelopment milestones, demonstrating cog-nitive stimulation activities or other activitiesto improve parenting skills. Types of deliverychannels are individual home visits, groupcounselling, childcare centres and massmedia.

● Greater effects are usually seen with interven-tions of longer duration and higher intensity.However, positive effects of short-term inter-ventions, particularly with micronutrientsupplementation, have also been demon-strated.

● Combined programmes could be more efficientat delivering services through reductions indelivery costs, less duplication of services, andappropriate identification of those who aremost likely to benefit. There are also likely tobe savings for families as a result of easieraccess when services are combined and fami-lies are more likely to be motivated to seekthose services.

● Programme efficacy and effectiveness appearto be greater when parents are more involved.

Issues affecting programme success

Despite the success of interventions, children whoare nutritionally or socio-economically disadvan-taged never fully catch up to the level of well-nourished, more privileged children. Over time,children in disadvantaged environments becomeprogressively more disadvantaged with respect tooutcome measures on psychological develop-mental tests, and these test scores reflect the pro-gressive social and educational disadvantages thatthey suffer. In many cases, the effect of interven-tions is to prevent or slow down the progressivedeterioration that is a common fate of children indeprived environments.

Supplementary feeding or increased stimula-tion during relatively short periods of children’slives (such as periods of malnutrition) will not



60

obviate all the difficulties of growing up in a dis-advantaged environment. The positive effects ofECCD and nutrition interventions cannot be ex-pected to persist indefinitely in the face of thedaily effects of such an environment. There is aneed for continuing programmes of support thatare tailored to the changing needs of children andadolescents.

Recommendations

We need to develop and test a model ofcombined interventions that could reach alarge proportion of children who are atrisk of growth and development faltering.

Appropriate feeding (food and practices) andresponsive parenting (e.g. attentive listening,proactive stimulation, and appropriate responses)need to be promoted where children are suffer-ing from malnutrition and developmental delaysdue to poverty. The review points to the value ofinterventions that combine activities to supportboth.

A necessary first step is to further develop andtest a model that combines nutrition and psycho-social interventions. The model should be likelyto be effective and feasible within the context of apublic health programme that is delivered throughthe health services or community-based pro-grammes. It should focus on children from age0-3 years, their most vulnerable years. It shouldseek to improve the health and developmentbroadly among children, rather than be limitedto children in particular risk groups or with iden-tified disabilities. The intervention should iden-tify ways to deliver and strengthen the counsellingof caregivers in order to improve their knowledgeand skills about appropriate foods and feedingpractices, including breastfeeding, and help themprovide more responsive and stimulating care toenhance the psychological development of theirchildren. The counselling should also be specifi-cally related to the child’s changing nutritionalrequirements and readiness to learn new skills ashe or she grows and develops. Whenever neces-sary, the intervention should also provide sup-plementary foods and/or nutrient supplements.

To operationalize this type of intervention, thefollowing steps are suggested:

● Develop a generic counselling package, withguidelines on how to adapt it to local culturaland environment conditions.

● Define the most appropriate approach to pre-senting this information and building skills

based on principles of modelling, feedback,and culturally appropriate communication.

● Define special groups of children or conditionsand identify particular information for thosegroups (e.g. malnourished children).

● Identify mechanisms and channels for the de-livery of the package (e.g. growth promotionprogrammes, health service contacts, commu-nity health worker programmes, breastfeedingpromotion or reproductive health programmes,or other community organizations).

● Develop guidelines for identifying the familieswho need additional supports in order to im-plement the recommended behaviours (e.g.supplementary foods, micronutrient supple-ments, and other types of social support).

● Develop tools for training, monitoring, andsupervision of the counselling and supportpackage.

● Develop evaluation tools, including methods

Rationale for combined nutrition and developmentinterventions■ Many households in conditions of poverty

potentially have the resources to provide adequatediets and use good feeding practices of the typethat support normal growth. What they lack areknowledge and skills about how to do this withintheir local environmental and cultural context.

■ Many other households in conditions of povertyare more severely constrained economically. Inaddition to knowledge and skills, they requireassistance in the form of supplementary, nutri-tious foods and/or nutrient supplements for theirinfants and small children, and for consumptionby mothers during pregnancy and lactation.

■ An important factor in developmental delaysassociated with malnutrition is the evolution ofthe behavioural transactions between the malnour-ished child and his or her caregivers, in which thechild becomes progressively more apathetic andless demanding and caregivers provide lessstimulation and responsive interactions.

■ Feeding is a central aspect of caregiving ininfancy and early childhood, and the teaching offeeding skills provides an opportunity to teachother caregiving skills, including responsiveparenting to provide stimulation for psychologicaldevelopment.

61

for assessing psychological development thatare culturally appropriate.

● Identify mechanisms for strengthening familyresources for providing responsive care andstimulation, including greater control of care-givers over resources, time availability, andinfrastructure systems, such as improvedwater and sanitation.

● Identify mechanisms and channels for support-ing and facilitating family behaviour changewith respect to appropriate feeding and respon-sive parenting through activities directed to thecommunity (e.g. communication and commu-nity development activities).

We should develop and implement newactivities to promote appropriate feedingand responsive parenting in existing childhealth programmes.

We should expand and strengthen the health,nutrition, and breastfeeding components ofexisting early childhood care and development(ECCD) programmes.

Health and nutrition service infrastructures andcommunity-based outreach programmes providea potential basis for new and expanded packagesof nutrition, health promotion and child devel-opment activities. This requires bringing togetherexpertise in child psychological development,health services, and nutrition to develop a pack-age of integrated activities.

Activities could be incorporated into well-childclinics, primary health care consultations forchildhood diseases, prenatal care, and nutritionprogrammes, such as growth monitoring, nutri-tion education, breastfeeding promotion, andnutrition rehabilitation centres. Health staff couldwork with community agencies, non-governmen-tal organizations (NGOs), and other groups toincorporate appropriate feeding and parentingeducation into on-going community programmes.Informational materials about appropriate feed-ing and responsive parenting could be utilizedand disseminated by the educational sector.

Actions outside the health sector

The need to address the determinants of malnu-trition in children has been effectively articulatedin a number of international arenas, including theInternational Conference on Nutrition (FAOand WHO, 1992) and the World Food Summit(1996). We note that there are also actions that

can be taken by the health sector in support ofnon-health sector activities. The health sector hasan important role to play in creating and dissemi-nating informational materials about the closerelationship of appropriate feeding and respon-sive parenting for the child’s psychological devel-opment. It can also motivate and support theeducational sector and the media to provideinformation and support to families on appropri-ate feeding and responsive caregiving.

ECCD and day care programmes afford vitalopportunities to address problems of malnutri-tion and promote physical growth. However, itappears that the nutrition component of theseprogrammes is often weak in comparison withactivities aimed at the promotion of psychologi-cal and social health. We recommend that highpriority be given to efforts to identify ways to ex-pand and strengthen this component, particularlyin relation to family and caregiver behaviours.

We need to design an expanded researchagenda to compare and evaluate theeffectiveness of different content,programme venues, and delivery channels.

The research community must be mobilized.Research efforts are required not only to addresscritical gaps in knowledge, but also to evaluateproposed programmes and their components, toidentify means of overcoming barriers to imple-mentation, and to develop new and more efficientmethods for institutionalising and sustainingprogrammes.

Research and development should focus oncombined (integrated or convergent) nutrition,health, and psychosocial interventions for chil-dren from 0-3 years of age. The interventionsdeveloped should involve the family, which is thechild’s primary caregiving system. Support for theinterventions (e.g. technical background, train-ing materials, tools for evaluation and monitor-ing) should also be developed.

As it is difficult to draw conclusions for pro-gramme implementation from individual fieldtrials, a research agenda to evaluate differentapproaches and to clarify operational issuesshould be developed. The intervention researchagenda should include comparisons of differentchannels or types of delivery mechanisms (e.g.health workers, parent support groups, and homevisiting). The research must be conducted indeveloping countries in collaboration with estab-lished in-country investigators.



62

1. Integrated Child DevelopmentServices (ICDS), India

Objectives

● Improve nutritional and health status of chil-dren under age 6.

● Lay the foundation for proper child physicalgrowth and psychological development.

● Reduce mortality, morbidity, malnutrition andschool dropout rates.

● Achieve effective coordination of policy andimplementation.

● Enhance the mother’s ability to look after thenormal health and nutritional needs of thechild.

Programme participants and components

● 3- to 6-year-olds.● Non-formal pre-school education provided

courtyard centres.● Supplementary feeding for children, malnour-

ished children, pregnant and nursing women.● Health and nutrition education, and growth

promotion.● Health services and links with primary health

care: immunization, vitamin A supplementa-tion, referral, treatment of minor illness.

Personnel

● 300␣ 000 Anganwadi workers (AWW, or localwomen selected by government) and equalnumber of helpers.

● AWWs expected to have matriculated (10 yearsof education) and live in the community butcriteria are relaxed (5–8 years of education) ifnecessary.

● AWWs paid a minimal stipend.

Training

● Formerly three-month pre-service training inone of over 300 training centres. Now replacedwith 3 phases totalling 3 months of institu-tional and 4 months of community-basedtraining.

● Short refresher training courses.

Supervision and management

Each child development programme officer isresponsible for implementation and managementof about 100 centres, covered by 5 supervisors,who in turn supervise 20 AWWs.

Responsible agency

Department of Women and Child Developmentin the Ministry of Human Resource Development.

Implementation

1975 to present. 3␣ 907 centres operational in1995.

Coverage

17.8 million children (under 6 years) and3.8 million expectant and nursing mothers fromdisadvantaged groups and communities. Thelargest programme of its kind in the world.Almost 70 percent of country’s community de-velopment blocks.

Cost

$10 to $22 per child per year (or 1/15 minimumwage).

Effectiveness

Evaluation design

● Numerous small studies and a national evalu-ation.

● Adequacy designs, with little control for self-selection biases in some studies.

Impact on physical growth

● National evaluation found lower levels ofundernutrition in ICDS areas (about 2 percent-age points lower Grade III and IV; statisticalsignificance and basis of comparison groupsunknown).

Annex

Seven Programme Models

63

● Evaluation of USAID-assisted sites in Panch-mahals and Chandrapur found reductions ininfant and toddler mortality rates and improve-ments in weight-for-age, despite occurrence ofdrought. Changes were seen over a 10-yearperiod, with no comparison groups.

Impact on psychological development

● National Institute of Nutrition evaluation com-pared beneficiaries, defined as children whoattended regularly (3 times/week for 6 months)to non-participants in same communities.Beneficiaries scored higher on cognitive teststhan non-participants, but still lower than ex-pected for their chronological age, especiallyamong the older children. No relationshipswere found between weight-for-age and testscores.

● Chaturvedi and colleagues (1987) compared6- to 8-year-old children in ICDS and controlcommunities, and found significant differencesin percent of children in school, age of schoolentry, and for females, Raven’s ProgressiveMatrices Test scores.

Other outcomes

● Reductions in infant mortality rates and in-crease in immunisation coverage in ICDS vs.non-ICDS areas.

● Higher primary school enrolment, especiallyamong girls.

Factors influencing effectiveness

● AWWs’ responsibilities include knowing andvisiting homes of malnourished children,making them accountable for reaching needyfamilies.

● AWWs reported to be over-burdened.

Sources: ICDS, 1995; Myers, 1992; Tandon, 1989;Consultative Group on Early Childhood Care and De-velopment (CGECCD), 1993; ACC/SCN, 1991; Vazirand Kashinath, 1995; USAID, no date; Chaturvedi etal., 1987.

2. Head Start, U.S.A.

Objectives

● Provide comprehensive child developmentservices for low-income children and theirfamilies.

● Meet pre-school children’s educational, health,nutritional, and psychological needs.


● 3- to 5-year-olds.● Pre-school education (usually part-day for the

duration of school year). At least 90 percentmust be from families below poverty line,10 percent disabled. Most programmes havewaiting lists and selection criteria vary, butgenerally reflect need.

● Multi-cultural, bilingual curriculum.● Meals/snacks provide at least one-third of daily

nutritional needs.● Health care services: physical exam, assess-

ments of growth and immunisation status,vision, hearing, speech, anaemia and otherprevalent health problems, with referral to freemedical care.

Personnel

One-third of employees are parents of current orformer Head Start students.

Training

Not specified.


Grants are awarded to local public agencies,private non-profit organizations and schoolsystems to operate programmes at the commu-nity level.

Responsible agency

U.S. Department of Health and Human Services(DHHS).

Implementation

1964 to present. In 1994, there were 1␣ 405 grant-ees and 40␣ 295 Head Start classrooms. Pro-grammes vary since locally operated. Recentinitiatives include Head Start transition projectsthat continue support services through thirdgrade, family service centres to provide literacy,treatment of substance abuse and employmentprogrammes for families, and home-based pro-grammes for children under 3 years of age.

Coverage

In 1995, 622␣ 000 children or about 30 percentof eligible 3- to 5-year-olds.

Cost

$4␣ 000 per child per year (Currie and Thomas,1996).

ANNEX: SEVEN PROGRAMME MODELS


64

Effectiveness

Evaluation design

● Most studies suffer from selection biases andlack of appropriate comparison groups.

● A recent analysis of two national data sets hada high plausibility design comparing siblingswho did and did not participate (to control forunobservable differences in families and tocontrol for observable child characteristics) andcomparing the effect of other pre-schoolsrelative to no pre-school (Currie and Thomas,1995; Currie and Thomas, 1996).


● No impact seen on height-for-age. (None wasexpected since not a period or populationvulnerable to active stunting).


● Positive effects noted on the Peabody PictureVocabulary Test (PPVT). Effects persisted intoand translated into improved school attainmentfor white and Hispanic (especially of Mexicanorigin) children but not African-Americanchildren.

● By various measures, Head Start closed one-fourth to two-thirds of the gap betweenchildren attending the programme and theirmore advantaged peers.

Other outcomes

● Currie and Thomas (1995) extrapolate thatenrolment increased future expected wage by4 percent and chance of completing school by5 percent.

● Any pre-school experience, including HeadStart, was associated with improved immuni-sation status.


● Database did not allow analysis of programmecharacteristics.

● Some evidence that children of mothers withhigher test scores benefited more from HeadStart.

● Ethnic and language characteristics of homeinfluenced impact.

Sources: Currie and Thomas, 1995; Currie and Tho-mas, 1996; Lee et al., 1988; The Future of Children,1995.

3. PANDAI (Child Development &Mother’s Care) Project, Indonesia

Objectives

● Improve child growth and developmentthrough changing child care-taking behaviourand improving home and environmentalstimulation.

● Increase parental sensitivity to children’s needsand capabilities.


● 0- to 5-years-old, but materials developedspecifically for children up to 36 months.

● Curriculum based on 36 developmentallysequenced items or milestones that form thebasis of the Child Development Card (KKA),given to parents (similar to growth monitor-ing card).

● Home-visits to improve parent-child inter-actions, to teach parents to interact with childin ways to stimulate learning, and to teach par-ents to monitor development of child withKKA.

Personnel

Kaders, women from community who are alreadyinvolved as volunteers in the local health centreactivities, conducted home-visits and assessments.

Training

Kaders given additional training on appropriatechild care-taking behaviours, their importance forchild development, and use of Child DevelopmentChart to assess child’s progress.


Kaders supervised by local health centre staff, whoin turn are supervised by a medical doctor andpsychologist.

Responsible agency

Ministry of Health and Diponegoro University,Semarang, Indonesia.

Implementation

Initial pilot test with 150 children and illustratedmaterials based on the Portage curriculum. Nowexpanded to wider area (but less intensive) as acomponent of nation-wide Child DevelopmentProgramme (BKB). Added to on-going nutritioneducation and growth promotion programmeimplemented in communities.

65

Coverage

Not specified.

Cost

Not specified.

Effectiveness

Evaluation design

● Pre- and post-test evaluation of impact of 18-month pilot intervention with 150 children,compared to 150 control children from com-munities without the intervention.

● Full-scale programme has not yet been evalu-ated.


● No significant impact seen on weight-for-age.● Some evidence of improvements seen in child

feeding practices and reported dietary intake.


● Post-test scores for mental and motor devel-opment significantly improved relative to pre-test among intervention group only.

● Greatest effects were for motor developmentof younger children.

● Significant improvements seen in parental be-haviour in intervention group, measured byHOME test, with improved scores for verbaland emotional responsiveness, play materials,and parental involvement.

Other outcomes

● Qualitative assessments indicated that moth-ers and fathers were motivated by the inter-vention and requested more information.

● Some spill-over effects on neighbours werenoted.


● The intensity of this pilot intervention and thefact that kaders chose as participants only 4 or5 families that they knew from their on-goingcommunity health activities had an effect.

● Cannot be considered as an evaluation of a full-scale programme. However, it did generateenough interest that other parts of the countryasked to become involved and expand the pro-gramme.

Sources: Unpublished materials provided by Dr Satoto,Diponegoro University, Semarang, Indonesia, many ofwhich were jointly prepared with Dr Nancy Colletta,University of Maryland, Baltimore, MD.

4. PRONOEI, Peru

Objectives

Not specified.


● 3- to 5-year-olds.● Parents and community members involved in

parent committees.● Non-formal pre-school held 4–5 mornings per

week, using Piagetian curriculum.● Snack or noon meal.● Some income-generating projects as spin-off.

Personnel

● Animators (men and women selected fromcommunity) run the pre-school and are paidminimal incentive by Ministry of Education,which is supposed to be supplemented by com-munity.

● Mothers take turns preparing the snack or mealfor the children.

Training

Little training provided, although guidance is tobe given by coordinators. Increasingly, animatorshave experience as an assistant before becomingthe animator.


Teacher coordinators responsible for training,supervising and advising animators. Coordinatorsvisit 6–24 communities 5 times/year. Parent com-mittees also involved in management.

Responsible agency

Ministry of Education.

Implementation

Initiated on small scale in Puno in 1967, expandedto Department of Puno 5 years later, subsequentlyexpanded to 4 Departments. Site often donatedby community members, and some built by com-munity with materials provide by donors.

Coverage

About 2␣ 000 PRONOEI in 4 regions receivingUSAID assistance in 1984, which represents aboutone-third of all PRONOEI. Enrolment from 20–32 (average 29) but not all attend regularly.

Cost

$28 per child per year excluding community con-



66

tribution; $40 including in-kind contributions(1985). (Also estimated as 1/14 of minimum wageor 1/40 of GNP per capita.)

Effectiveness

Evaluation design

● Testing of mental, motor and social develop-ment with locally developed scales.

● Comparison of PRONOEI and non-PRONOEIcommunities in 3 Departments. In two depart-ments, the non-participant mothers werebetter educated.

● Pre-/post-test comparisons available on a smallsample.


● Effects on anthropometric indicators werenot seen, except that the gap between boys’and girls’ nutritional status was less in somePRONOEI groups.

● Nutritional supplementation has not been aconsistent focus of the programme, amountsprovided are not substantial, and children arenot in period of active growth faltering, so thelack of effect is not surprising.


● In Puno, where the programme is well estab-lished and control group was comparable,PRONOEI children scored significantly higherthan controls on all three subscales.

● No differences seen in the other two depart-ments except for higher scores for PRONOEIchildren on social sub-scale in one department.

● Pre-test/post-test comparisons suggest greaterimprovement over 5 months among PRONOEIchildren, among those who begin with very lowlevel of ability, and especially social skills.

Other outcomes

● Differences in school progress and performancein first three years of primary school were notfound, but this may reflect poor quality of theprimary schools themselves.

● Some effects on community such as attentionto education, nutrition and health topics atcommunity meetings.


● Low attendance (due to illness, distance,parental apathy, lack of food supplementation).

● Lack of trained coordinators, and support toanimators.

● High turnover of animators, lack of paymentby community.

● Location in remote communities.

Sources: Myers et al., 1985; Myers, 1992.

5. Programa de Alimentacao dePre-escolar (PROAPE), Brazil

Objectives

Not specified.


● 3- to 6-year-olds.● Snack, including milk.● Supervised psycho-motor activities.● Health component: check-ups, immunisation,

dental care, hygiene, and visual examinations.

Personnel

● Combination of trained personnel (usuallypara-professionals from the community) andparents.

● Staffing varied in different areas but generallyincluded paid staff.

Training

Not specified.


Not specified.

Responsible agency

Ministries of Health and Education.

Implementation

Pilot began in 1977. Expanded to 10 other statesin 1981. Held in community sites with groups ofabout 100 children. Not community-controlled.Not currently operating—this is an example of aprogramme shown to result in cost savings (seebelow) by government evaluations that wasnevertheless discontinued for other reasons.

Coverage

10 states, with reportedly large numbers of chil-dren.

Cost

$28 per child per year.

67

Effectiveness

Evaluation design

● Compared participants and non-participantsin various locations.

● Other aspects of design and comparison groupsnot known (reports cited by Myers not avail-able).


● Marginal impact seen as expected givenfactors cited for Head Start and PRONOEI.


● Combined repetition and dropout rate of 39percent for PROAPE children vs. 52 percentfor controls in first grade, and 27 percent vs.44 percent in second grade. Total cost for sec-ond grade graduate 11 percent less for PROAPEparticipants (including cost of PROAPE).

● In another evaluation, 73 percent PROAPEchildren passed first grade vs. 53 percent ofchildren with no pre-school experience. Costper first grade graduate was 17 percent lowerfor PROAPE participants.

Other outcomes

Not specified.


Not specified.

Source: Myers, 1992.

6. Hogares Comunitarios de Bienestar(Homes of Well-Being), Colombia

Objectives

● Improve psychosocial, moral, and physicaldevelopment of children under 7 years(directed to poorest sectors).

● Provide stimulation and support for social-isation.

● Improve nutrition and living conditions.● Strengthen the responsibility of parents in

combination with the community for care ofchildren.


● 2- to 5-year-olds (can have up to 2 childrenless than 2 years of age per centre).

● Centres in 1␣ 042 municipalities; 882␣ 000

children; 54.3 percent of target population ofpoorest families.

● Community mothers (CM) hold pre-schools forup to 15 children from 0–7 five days a week.

● CM provide love and protection, educationalactivities, and food (68 percent of require-ments).

● Association of Parents (up to 25 homes) is re-sponsible for the programme and receivesfunding from the state.

● Children were weighed and measured onaverage three times a year.

Personnel

Community mothers.

Training

Community mothers receive training prior tostarting.

Supervision and Management

Centres supervised by Association of Parents, anda committee for oversight. Association supervisedand trained by a Centre for the Zone.

Responsible agency

Colombian Institute for Family Welfare.

Implementation

Not specified.

Effectiveness

Evaluation design

● Nation-wide probability sampling of commu-nity homes: 4␣ 762 day care homes, 798 com-munity associations, 69␣ 000 children with asub-sample of 23␣ 810 children, and a slightlysmaller number of parents.

● Two types of analysis: Description of day carehomes, community associations, children andparents; and multivariate analysis of multipleinfluences on children. No control groups.

● Analysis examined the relation of quality ofservice and length of exposure to child nutri-tional status, health, and psychosocial devel-opment. A second analysis compared qualityof programme to overall child outcomes.


● In general, no evidence for improvement innutritional status.

● Rates for malnutrition lower in children from



68

24–47 months living in particular areas withvery poor parents, and in day care centres inwhich the dietary recommendations of ICBFwere met.


● Qualitative development scale (Escala deValorizaci Cualitativa del Desarollo Infantil) toclassified children in terms of 12 processes intothree categories: at risk (three or more pro-cesses at risk), normal, and advanced.

● Higher percent of at-risk children found atyounger ages: 8.6 percent at risk at age 3versus 0.8 percent for age 6. However, therewas no relation between time of exposure andpercent at risk.

Other outcomes

● A significant association (r=0.17) betweenquality of the programme and degree of well-being based on a global indicator of health,nutrition, and psychosocial development wasfound.

● Quality was defined by health conditions ofday care home, practice of evaluating psycho-social development, knowledge of the com-munity mother (CM) about child care, foodprepared for the children, attitudes of CMtoward children, keeping records on children,and the characteristics of the CM’s family.

● Families had low levels of involvement in theprogramme, contrary to the stated goals of theproject.


● In 86 percent of day care homes, the menuwas correct, but in the visits, only 29 percentof community mothers prepared the lunch foodadequately.

● Some community mothers never trained (14percent), and conditions of houses were poor.

● About 28 percent of community mothers neverplanned any pedagogical activities.

● The longer the community mother had beenrunning the centre, the lower the percent ofchildren at risk of developmental delay.

Sources: ICBF, 1997; Instituto Colombina de BienestarFamiliar, Bogota, Colombia, 1997.

7. Integrated Programme for Childand Family Development(IFBECD), and Family Develop-ment Programme (FCP), Thailand

Objectives

● Enhance public health and nutrition servicewith other aspects of child development usinghealth “pink books” revised to include ECCDmessages.


● 0- to 6-year-olds.● Home-based, centre-based and mobile ECCD

programmes, enhanced by volunteers.● Non-formal education including information

on parenting and family life education.● Primary education incorporating child-to-child

messages for 5th and 6th graders includingtopics such as cleaning, making toys, playing,reading, food and nutrition, etc.

● Toy library.● Agricultural extension to increase food secu-

rity.

Personnel

Trained child development worker in thecommunity (also often the health worker) andwomen’s groups in the community.

Training

Not specified.


Not specified.

Responsible agency

Initially a partnership of the Government of Thai-land, UNICEF, NGOs, and a university under arotating committee secretariat, with partners fromseveral parts of government (e.g. Ministries ofHealth and Agriculture). Now run by NationalInstitute for Child and Family Development, withmany new partners as project expanded to includeyouth and adolescence. Local community organi-zations selected volunteers for project.

Implementation

Pilot began in 1990. Expanded to 16 provincesin 1992, and expanded to all 75 provinces in 1995as FCP following government commitment.Current phase to terminate in 1998.

69

Coverage

All provinces and districts except Bangkok (FCP).More than 23,739 children under age 6 years,3␣ 192 families, and some older children andyouth.

Cost

Not specified.

Effectiveness

Evaluation design

● Pre/post comparison and treatment and con-trol communities in two provinces for firstphase of the project (IFBECG).

● Two different sites or provinces were evalu-ated—one high intensity, and one low inten-sity.


● High intensity: Both groups increased innutritional status.

● Low intensity: No effects on growth overall,except that greatest changes seen in poorestvillages.


● High intensity: Programme villages showed anincreased proportion of average and brightchildren, but both groups improved.

● High intensity: Parents’ perceptions of chil-dren’s developmental level improved more inprogramme villages.

● Low intensity: IQ scores improved modestlyin both groups. In programme sites, fewer chil-dren had a decline in IQ scores over time.

● Low intensity: Home scores improved more inprogramme sites.

Other outcomes

Not specified.


● Prepared materials that could be used by allworkers.

● Difficult to have workers focus on psycho-social care rather than physical care andgrowth.

● Organization and coordination between vari-ous groups, including funding for purposes ofcoordination, formed a spirit of cooperation.

● Strong village organization builds on in imple-menting activities.

● Political commitment shown by government.● Focus lost when project moved from em-

phasis on 0- to 6-year-olds to the larger FCPapproach which targeted 0- to 18-year-olds.

● Workers were unclear of roles, there was in-sufficient time to do all of the projects, andonly 32 percent of workers felt that they un-derstood the purpose of the project. Manyworkers felt that this was simply an additionaljob added to their already overloaded job, andlittle occurred.

Source: Herscovitch, 1997.



70

References

Barnett SW. Long-term effects of early childhood pro-grams on cognitive school outcomes. The future ofchildren, 1995, 5:25-50.

Barnett SW. Lives in the balance: age 27 benefit-cost analy-sis of the High/Scope Perry Preschool Program.Ypsilanti, MI, High/Scope Press (No. 1056), 1996.

Barnett SW, Escobar CM. Research on the cost-effec-tiveness of early educational intervention: implica-tions for research and policy. American journal ofcommunity psychology, 1989, 17:677-704.

Bauer G. et al. Breastfeeding and cognitive develop-ment of three-year-old children. Psychological re-ports, 1991, 68 (June):1218.

Beaton GH, Ghassemi H. Supplementary feeding pro-grams for young children in developing countries.American journal of clinical nutrition, 1982, 35:863-916.

Benasich AA, Brooks-Gunn J, Clewell BC. How domothers benefit from early intervention programs?Journal of applied developmental psychology, 1992,13:311-362.

Bentley ME, Caulfield LE, Ram M, Santizo, MC,Hurtado E, Rivera JA, Ruel MT, Brown KH. Zincsupplementation affects the activity patterns ofrural Guatemalan infants. Journal of nutrition, 1997,127:1333-1338.

Bentley ME, Stallings RY, Fukumoto M, Elder JA.Maternal feeding behavior and child acceptance offood during diarrhea, convalescence, and health inthe central Sierra of Peru. American journal of publichealth, 1991, 81:43-47.

Black MM, Dubowitz H, Hutcheson JJ, Berenson-Howard J, Starr RH. A randomized clinical trial ofhome intervention for children with non-organicfailure to thrive. Pediatrics, 1995, 95:807-814.

Black MM, Hutcheson JJ, Dubowitz H, Berenson-Howard J. Parenting style and developmentalstatus among children with non-organic failure tothrive. Journal of pediatric psychology, 1994, 19:689-707.

ACC/SCN. Managing successful nutrition programmes.Geneva, ACC/SCN, 1991 (ACC/SCN State-of-the-Art Series, Nutrition Policy Discussion PaperNo. 8).

ACC/SCN. Nutrition-relevant actions. Geneva, ACC/SCN, 1991b (ACC/SCN State-of-the-Art Series,Nutrition Policy Discussion Paper No. 10).

ACC/SCN. How nutrition improves. Geneva, ACC/SCN,1996 (ACC/SCN State-of-the-Art Series, NutritionPolicy Discussion Paper No. 15).

Ahmed NU, Zeitlin MF, Beiser AS, Super CM. A lon-gitudinal study of the impact of behavioural changeinterventions on cleanliness, diarrhoeal morbidityand growth of children in rural Bangladesh. Socialscience and medicine, 1993, 37:159-71.

Allen, LH. Nutritional influences on linear growth: ageneral review. European journal of clinical nutrition,1994, 48(S1):S75-S89.

Andelman MB, Sered BR. Utilization of dietary iron byterm infants. American journal of the diseases of chil-dren, 1966, 111:45-55.

Arnold C. Early childhood development programmes: someexamples from South Asia focusing on parenting/caregiving and the 0–3s. Paper presented at theAnnual Meeting of the Consortium for ChildhoodCare and Development, Paris, April 20-25, 1998.

Ashworth A, Feachem RG. Interventions for the con-trol of diarrhoeal diseases among young children:weaning education. Bulletin of the World HealthOrganization, 1985, 63:1115-1127.

Aukett MA, Parks YA, Scott PH, Wharton BA. Treat-ment with iron increases weight gain and psycho-motor development. Archives of disease in childhood,1986, 61:849-857.

Aylward GP, Pfeiffer SI, Wright A, Verhulst SJ. Out-come studies of low birth weight infants publishedin the last decade; a meta-analysis. Journal ofpediatrics, 1989, 115:515-520.

Barnett SW. Benefit-cost analysis of the Perry PreschoolProgram and its policy implications. Educationalevaluation and policy analysis, 1985, 7:333-342.

71

Black MM, Teti LO. Promoting mealtime communica-tion between adolescent mothers and their infantsthrough videotape. Pediatrics, 1997, 3:432-436.

Black RE, Brown KH, Becker S. Malnutrition is adetermining factor in diarrheal duration, but notincidence, among young children in a longitudinalstudy in rural Bangladesh. American journal of clini-cal nutrition, 1984, 39:87-94.

Bleichrodt N, Born MPh. A meta-analysis of researchon iodine and its relationship to cognitive develop-ment. In: Stanbury JB, ed. The damaged brain ofiodine deficiency. New York, Cognizant Communi-cation Corporation, 1994:195-200.

Bleichrodt N, Shrestha RM, West CE, Hautvast JG, vande Vijver FJR, Born MPh. The benefits of adequateiodine intake. Nutrition reviews, 1996, 54:S72-S78.

Boocock SS. Early childhood programs in othernations: goals and outcomes. The future of children,1995, 5:94-114.

Bradley RH, Caldwell BM. 174 children: a study of therelationship between home environment and cog-nitive development during the first five years. In:Gottfried AW, ed. Home environment and early cog-nitive development. Orlando, FL, Academic Press,1984.

Bradley RH, Caldwell BM, Rock SL, Barnard C, GrayMA, Hammond S, Mitchell L, Siegel L, Ramey CT,Gottfried AW, Johnson DL. Home environment andcognitive development in the first three years of life:a collaborative study involving six sites and threeethnic groups in North America. Developmental psy-chology, 1989, 25:217-235.

Brooks-Gunn J, Duncan GJ, Klabanov PK, Sealand N.Do neighborhoods influence child and adolescentdevelopment? American journal of sociology, 1993,99:353-395.

Brown KH, Begin F. Malnutrition among weanlings ofdeveloping countries: still a problem begging forsolutions. Journal of pediatric gastroenterology andnutrition, 1993, 17:132-138.

Brown KH, Peerson JM, Allen LH. Effect of zinc sup-plementation on children’s growth: a meta-analysisof intervention trials. In: Sandstron B, Walter P, eds.Trace elements, growth and development: role of traceelements for health promotion and disease prevention.Basel, Karger, 1988, 54:76-83.

Brown LV, Zeitlin MF, Peterson KE, Chowdhury AMR,Robers BL, Weld LH, Gershoff, SN. Evaluation ofthe impact of weaning food messages on infant feed-ing practices and child growth in rural Bangladesh.American journal of clinical nutrition, 1992, 56:994-1003.

Carnegie Task Force on Meeting the Needs of YoungChildren. Starting points: meeting the needs of ouryoungest children. New York, Carnegie Corporationof New York, 1994.

Caulfield LE, Huffman S, Piwoz E. Interventions toimprove complementary food intakes of 6-12 monthold infants in developing countries: impact ongrowth, prevalence of malnutrition, and potentialcontribution to child survival. 1998. Unpublishedmimeo prepared for the Academy of EducationalDevelopment Linkages Program, Washington, DC.

Ceesay SM, Prentice AM, Cole TJ et al. Effects onbirthweight and perinatal mortality of maternaldietary supplementation in a primary health caresetting in rural Gambia. British medical journal, 1997,315:786-790.

Chaturvedi S, Srivastava BC, Singh JV, Prasad M.Impact of six years exposure to ICDS scheme onpsychosocial development. Indian paediatrics, 1987,24:153-160.

Chavez A, Martinez H, Yachine T. Nutrition, behavioraldevelopment and mother-child interaction in youngrural children. Federation proceedings, 1975,34:1574-1582.

Chavez A, Martinez C, Soberanes B. The effect of mal-nutrition on human development: a 24-year studyof well-nourished and malnourished children liv-ing in a poor Mexican village. In Scrimshaw NS,ed. Community-based longitudinal nutrition and healthstudies: classical examples from Guatemala, Haiti, andMexico. Boston, International Nutrition Foundationfor Developing Countries, 1995.

Chikane R. The effects of township unrest on children.In Burman S, Reynolds P, eds. Growing up in adivided society: the context of childhood in SouthAfrica. Johannesburg, South Africa, Raven Press,1986:35-53.

Chwang L, Soemantri AG, Pollitt E. Iron supplemen-tation and physical growth of rural Indonesian chil-dren. American journal of clinical nutrition, 1988,47:496-501.

Condry S. History and background of preschool inter-vention programs and the Consortium for Longitu-dinal Studies. In: The Consortium for LongitudinalStudies, as the twig is bent...lasting effects of preschoolprograms. Hillsdale, NJ, Lawrence Earlbaum Asso-ciates, 1983, 1-30.

Conroy RM, Elmore-Meegan M, Joyce T, McGuigan KG,Barnes J. Solar disinfection of drinking water anddiarrhoea in Maasai children: a controlled field trial.Lancet, 1996, 348:1695-1697.

REFERENCES


72

Consortium for Longitudinal Studies. As the twig isbent...lasting effects of preschool programs. Hillsdale,NJ, Lawrence Earlbaum Associates, 1983.

Consultative Group on Early Childhood Care andDevelopment (CGECCD). Meeting basic needsthrough programmes of early childhood care anddevelopment. The education for all forum: the secondmeeting. New Delhi, India, 1993.

Currie J, Thomas D. Does Head Start make a differ-ence? The American economic review, 1995, 85:341-346.

Currie J, Thomas D. Does Head Start help Hispanicchildren? RAND labor and population program, 1996(Working paper series, No. 96-17).

de Andraca I, Peirano P, Uauy, R. Nutrition and care inthe preterm and neonatal periods and later devel-opment: human milk is best for optimal mentaldevelopment. In Nutrition, health, and child develop-ment: research advances and policy recommendations.Pan American Health Organization, Tropicalmetabolism Research Unit of the University of theWest Indies, and The World Bank. Scientific Publi-cation 566. 1998:43-68.

de Onis M, Villar J, Gülmezoglu M. Nutritional inter-ventions to prevent intrauterine growth retardation:evidence from randomized controlled trials. Euro-pean journal of clinical nutrition, 1998, 52(S1):S83-S93.

de Zoysa I, Habicht J-P, Pelto GH, Martines J. Researchsteps in the development and evaluation of publichealth interventions. Bulletin of the World HealthOrganization, 1998, 76:127-133.

Dickin K, Griffiths M, Piwoz E. Designing by dialogue:program planners’ guide to consultative research forimproving young child feeding. Washington, DC,Academy for Educational Development, 1997.

Doan RM, Popkin BM. Women’s work and infant carein the Philippines. Social science and medicine, 1996,36:297-304.

Engle PL. Care and child nutrition. Report for Interna-tional Congress on Nutrition. New York, UNICEF,1992.

Engle PL. Influences of mothers’ and fathers’ incomeon children´s nutritional status in Guatemala.Social science and medicine, 1993, 37:1303-1312.

Engle PL. The care initiative: assessment, analysis, andaction to improve care for nutrition. New York,UNICEF, 1997.

Engle PL, Carmichael SL, Gorman K, Pollitt E. Demo-graphic and socio-economic changes in the Guate-malan Oriente families from 1967 to 1987. Foodand nutrition bulletin, 1992, l4:237-245.

Engle PL, Menon P, Haddad L. Care and nutrition:concepts and measurement. Washington, DC, Inter-national Food Policy Research Institute, 1997.

Engle PL, Ricciuti HN. Psychosocial aspects of care andnutrition. Food and nutrition bulletin, 1995, 16:356-77.

Engle PL, Zeitlin M. Active feeding behavior compen-sates for low child demand among Nicaraguan one-year-olds. Journal of nutrition, 1996, 126:1808-1816.

Engle PL, Zeitlin M, Medrano Y, Garcia LH. Growthconsequences of low income Nicaraguan mothers’theories about feeding one year olds. In: Super C,Harkness S, eds. Parental cultural belief systems. NewYork, Guilford Press, 1996:428-446.

Erkel EA, Morgan EP, Staples MA, Assey VH, Michel Y.Case management and preventive services amonginfants from low-income families. Public health nurs-ing, 1994, 11:352-360.

Esrey SA, Habicht J-P, Latham MC, Sisler DG, CasellaG. Drinking water source, diarrheal morbidity andchild growth in villages with both traditional andimproved water supplies in rural Lesotho, South-ern Africa. American journal of public health, 1988,78:1451-1455.

Feachem RG, Koblinsky MA. Interventions for the con-trol of diarrhoeal diseases among young children:promotion of breastfeeding. Bulletin of the WorldHealth Organization, 1984, 62:271-291.

Food and Agricultural Organization of the UnitedNations and World Health Organization. Majorissues for nutrition strategies. International Confer-ence on Nutrition. ICN/92/INF/7. Rome andGeneva, FAO and WHO, 1992.

Friel JK, Andrews WL, Matthew JD, Long DR, CornelAM, Cox M, McKim E, Zerbe GO. Zinc supplemen-tation in very-low-birth-weight infants. Journal ofpediatric gastroenterology and nutrition, 1993, 17:97-104.

Garcia M, Pinstrup-Anderson P. The pilot food price sub-sidy scheme in the Philippines: its impact on income,food consumption, and nutritional status. ResearchReport No. 61. Washington, DC, International FoodPolicy Research Institute, 1987.

George SM, Latham MC, Abel R, Ethirajan N, FrongilloEA. Evaluation of effectiveness of good growthmonitoring in south Indian villages. Lancet, 1993,342:348-352.

Golub MS, Keen CL, Gershwin ME, Hendrick AG. Zincdeficiency and behavior. Journal of nutrition, 1995,125(8S):2263S-2271S.

73

Gomby DS, Larner MB, Stevenson CS, Lewit EM,Behrman RE. Long-term outcomes of early child-hood programs: analysis and recommendations. Thefuture of children, 1995, 5:6-24.

Gomby DS, Larson CS, Lewit EM, Behrman RE. Homevisiting: analysis and recommendations. The futureof children, 1993, 3:6-22.

Gonzales-Cossio T, Habicht J-P, Delgado H, RasmussenKM. Food supplementation during lactation in-creases infant milk intake and the proportion ofexclusive breastfeeding. Federation of Americansocieties of experimental biology journal, 1991,5:A917.

Gorman KS. Malnutrition and cognitive development:evidence from experimental /quasi-experimentalstudies among the mild-to-moderately malnour-ished. Journal of nutrition, 1995, 125(8S):2239S-2244S.

Gorten MK, Cross ER. Iron metabolism in prematureinfants: II. Prevention of iron deficiency. Journal ofpediatrics, 1964, 64:509-520.

Grantham-McGregor SM. A review of the effect ofsevere malnutrition on mental development. Jour-nal of nutrition, 1995, 125:2233S-2238S.

Grantham-McGregor SM, Powell C, Walker S, ChangS, Fletcher P. The long-term follow-up of severelymalnourished children who participated in anintervention program. Child development, 1994,65:428-439.

Grantham-McGregor SM, Powell CA, Walker SP,Himes JH. Nutritional supplementation, psycho-social stimulation, and mental development ofstunted children: the Jamaican Study. Lancet, 1991,338:1-5.

Grantham-McGregor SM, Schoefield W, Powell C.Development of severely malnourished childrenwho received psychosocial stimulation: a six-yearfollow-up. Pediatrics, 1987, 79:247-254.

Grantham-McGregor SM, Walker SP, Chang SM, PowellCA. Effects of early childhood supplementation withand without stimulation on later development instunted Jamaican children. American journal of clini-cal nutrition, 1997:66, 247-253.

Griffiths M, Dickin K, Favin M. Promoting the growth ofchildren: what works. Washington, DC, The WorldBank Human Development Department, 1996.

Gülmezoglu M, de Onis M, Villar J. Effectiveness ofinterventions to prevent or treat impaired fetalgrowth. Obstetrical and gynecological survey, 1997,52:139-149.

Gwatkin DR, Wilcox JR, Wray JD. Can health and nu-trition interventions make a difference? Washington,DC, Overseas Development Council monograph,No. 13, 1980.

Habicht J-P. Discussion: mortality, malnutrition andsynergies: determinants and interventions. In: HillK, ed. Child health priorities for the 1990s. Baltimore,MD, The Johns Hopkins University Institute forInternational Programs, 1992.

Habicht J-P, Butz WP. Measurement of health and nu-trition effects of large-scale nutrition interventionprojects. In: Klein RE et al, eds. Evaluating theimpact of nutrition and health programs. Plenum Pub-lishing Corporation, 1979:133-182.

Harkness S, Super CM, eds. Parents’ cultural belief sys-tems: their origins, expressions, and consequences. NewYork, The Guilford Press, 1996.

Hasan KZ, Briend A, Aziz KMA, Hoque BA, PatwaryMY, Huttly SRA. Lack of impact of a water andsanitation intervention on the nutritional status ofchildren in rural Bangladesh. European journal ofclinical nutrition, 1989, 43:837-843.

Haskins R. Beyond metaphor: the efficacy of early child-hood education. American psychologist, 1989,44:274-282.

Heimendinger J et al. Nutrition intervention in develop-ing countries, study IV: formulated foods. Cambridge,MA, Oelgeschlager, Gunn and Hain, 1981.

Hernandez DP. Changing demographics: past andfuture demands for early childhood programs. Thefuture of children, 1995, 5:145-160.

Herscovitch L. Moving child and family programs to scalein Thailand, Program Review. Bangkok, Thailand,UNICEF, 1997.

Hertzman C, Wiens M. Child development and long-term outcomes: a population health perspective andsummary of successful interventions. Social scienceand medicine, 1996, 43:1083-1095.

Hetzel BS. The story of iodine deficiency: an internationalchallenge in nutrition. Oxford, Oxford UniversityPress, 1989.

Hetzel BS, Dunn JT, Stanbury JB, eds. The preventionand control of iodine deficiency disorders. Oxford,Elsevier, 1987.

Huffman SL, Steel A. Do child survival interventionsreduce malnutrition? The dark side of child sur-vival. In: Pinstrup-Anderson P, Pelletier D, Alder-man H, eds. Child growth and nutrition in developingcountries: priorities for action. Ithaca, NY, CornellUniversity Press, 1995:139-152.

REFERENCES


74

Husaini MA, Karyadi L, Husaini YK, Sandjaja KaryadiD, Pollit E. Developmental effects of short-termsupplementary feeding in nutritionally-at-riskIndonesian infants. American journal of clinicalnutrition, 1991, 54:799-804.

ICBF (Instituto Colombiano de Bienestar Familiar).Primera encuesta, Sistema de evaluación de impacto,Hogares comunitarios de bienestar 0-6 años (Firstsystematic evaluation of impact of the Houses ofWell-Being). Bogotá, Colombia, ICBF, 1997.

ICDS (Integrated Child Development Services) ProjectReport. New Delhi, Department of Women andChild Development, Ministry of Human ResourceDevelopment, Government of India, 1995.

Idjradinata P, Pollitt E. Reversal of developmentaldelays in iron-deficient anaemic infants treated withiron. Lancet, 1993, 341:1-4.

Infant Health and Development Program (IHDP).Enhancing the outcomes of low-birth-weight,premature infants: a multi-site, randomized trial.Journal of the American Medical Association, 1990,263:3035-3042.

Jellife DB, Jellife FP. Human milk in the modern world.Oxford, Oxford University Press, 1978.

Jonsson U. Towards an improved strategy for nutri-tion surveillance. Food and nutrition bulletin, 1995,16:102.

Jonsson U, Ljungqvist B, Yambi O. Mobilization fornutrition in Tanzania. In: Rohde J, Chatterjee M,Morley D, eds. Reaching health for all. Delhi, Ox-ford University Press, 1993.

Joos SK, Pollitt E, Mueller WH, Albright DL. TheBacon Chow study: maternal nutritional supple-mentation and infant behavioral development. Childdevelopment, 1983, 54:669-676.

Kagitcibasi C, Sunar D, Bekman S. Comprehensive pre-school education project: final report. Ottawa, Canada,International Development Research Centre, 1988.

Kennedy E, Garcia M, eds. Effects of selected policies andprograms on women’s health and nutritional status,Washington, DC, International Food Policy Re-search Institute, 1993.

Kennedy ET, Alderman HH. Comparative analyses ofnutritional effectiveness of food subsidies and other food-related interventions. Washington, DC, InternationalFood Policy Research Institute, 1987.

Kielmann AA, Taylor CE, Parker RL. The Narangwalnutrition study: a summary review. American jour-nal of clinical nutrition, 1978, 31:2040-2057.

Kirkwood BR. Essentials of medical statistics. Oxford,Blackwell Scientific Publications, 1988.

Kirkwood BR, Ross DA, Arthur P, Morris SS, DollimoreN, Binka FN, Shier RP, Gyapong JO, Addy HA,Smith PG. Effect of vitamin A supplementation onthe growth of young children in northern Ghana.American journal of clinical nutrition, 1996, 63:773-781.

Kotchabhakdi NJ, Winichagoon P, Smitasiri S,Dhanamitta S, Valyasevi, A. The integration of psy-chosocial components of early childhood developmentin a nutrition education programme of Northeast Thai-land. Paper prepared for the Third Inter-Agencymeeting of the Consultative Group on Early Child-hood Care and Development, Washington, DC,January 1987.

Kramer MS. Determinants of low birth weight: meth-odological assessment and meta-analysis. Bulletin ofthe World Health Organization, 1987, 65:663-737.

Kramer MS. Effects of energy and protein intakes onpregnancy outcome: an overview of the researchevidence from controlled clinical trials. Americanjournal of clinical nutrition, 1993, 58:627-635.

Lanting CI, FidlerV, Huisman M, Touwen BCL, BoersmaER. Neurological differences between 9-year-oldchildren fed breast-milk or formula-milk as babies.Lancet, 1994, 344:1319-1322.

Lasky RE, Klein RR, Yarborough C, Engle PL, LechtigA, Martorell R. The relationship between physicalgrowth and infant behavioral development in ruralGuatemala. Child development, 1981, 52:219-226.

Lawless JW, Latham MC, Stephenson LS, Kinoti SN,Pertet A, Adams E. Iron supplementation and physi-cal growth in Kenyan children. Federation of Ameri-can societies of experimental biology journal, 1991,5:A720.

Lee VE, Brooks-Gunn J, Schnur E. Does Head Startwork? A 1-year follow-up comparison of disadvan-taged children attending Head Start, no preschool,and other preschool programs. Developmental psy-chology, 1988, 24:210-222.

Levin HM, Pollit E, Galloway R, McGurie J. Micronu-trient deficiency disorders. In: Jamison DT, MosleyWH, Measham AR, Bobadilla JL, eds. Disease con-trol priorities in developing countries. Oxford, WorldBank/Oxford University Press, 1993:421-451.

Levitsky DA, Strupp BJ. Malnutrition and the brain:changing concepts, changing concerns. Journal ofnutrition, 1995, 125(8S):2212S-2220S.

Liaw F-R, Meisels SJ, Brooks-Gunn J. The effects ofexperience of early intervention on low birth weight,premature children: the Infant Health DevelopmentProgram. Early childhood research quarterly, 1995,10:405-431.

75

Liddell C, Kvalsvig J, Qotyana P, Shabala A. Commu-nity violence and young South African children’sinvolvement in aggression. International journal onbehavioral development, 1994, 17:613-628.

Lie C, Ying C, Wang EL, Brun T, Geissler C. Impact oflarge-dose vitamin A supplementation on childhooddiarrhoea, respiratory disease and growth. Europeanjournal of clinical nutrition, 1993, 47:88-96.

Lindskog U, Lindskog P, Gebre-Medhim M. Childhealth and household water supply: a longitudinalstudy of growth and its environmental determinantsin rural Malawi. Human nutrition: clinical nutrition,1987, 41:409-423.

Lozoff B. Comments. American journal of clinical nutri-tion, 1989, 50:673-674.

Lozoff B. Explanatory mechanisms for poorer devel-opment in iron-deficient anemic infants. In: Nutri-tion, health, and child development: research advancesand policy recommendations. Pan American HealthOrganization, Tropical Metabolism Research Unitof the University of the West Indies, and The WorldBank. Scientific Publication 566. 1998:162-178.

Lozoff B, Jimenez E, Wolf AW. Long-term developmen-tal outcomes of infants with iron deficiency. NewEngland journal of medicine, 1991, 325:687-694.

Lozoff B, Brittenham GM, Viteri FE, Wolf AS, UrrutiaJJ. The effects of short-term oral iron therapy ondevelopment deficits in iron deficient anemic in-fants. Journal of pediatrics, 1982, 100:351-357.

Lucas A, Morley R, Cole TJ, Lister G, Leeson-Payne C.Breastmilk and subsequent intelligence quotient inchildren born preterm. Lancet, 1992, 339:261-264.

Lutter C. Recommended length of exclusive breastfeeding,age of introduction of complementary foods and theweanling dilemma. Geneva, WHO Program for theControl of Diarrhoeal Diseases, 1992 (unpublishedreport).

Lutter CK, Habicht J-P, Rivera JA, Martorell R. The re-lationship between energy intakes and diarrhoealdisease in their effects on child growth: biologicalmodel, evidence, and implications for public healthpolicy. Food and nutrition bulletin, 1992, 14:36-42.

Lutter CK, Mora JO, Habicht J-P, Rasmussen KM,Robson DS, Sellers SG, Super CM, Herrera MG.Nutritional supplementation: effects on child stunt-ing because of diarrhea. American journal of clinicalnutrition, 1989, 50:1-8.

Marsh A, Long H, Stierwait RN. Comparativehematological response to iron fortification of a milkformula for infants. Pediatrics, 1959, 24:404-412.

Martorell R. Promoting healthy growth: rationale andbenefits. In: Pinstrup-Anderson P, Pelletier D,Alderman H, eds. Child growth and nutrition indeveloping countries. Ithaca, NY, Cornell UniversityPress, 1995.

Martorell R. Undernutrition during pregnancy andearly childhood: consequences for cognitive andbehavioral development. In: Young ME, ed. Earlychild development: investing in our children’s future.Amsterdam, Elsevier, 1997:39-83.

Martorell R, Rivera J, Kaplowitz H, Pollitt E. Long-termconsequences of growth retardation during earlychildhood. In: Hernandez M , Argente J, eds.Human growth: basic and clinical aspects. Amsterdam,Elsevier Science Publishers B.V., 1992:143-149.

McCormick MC. The contribution of low birth weightto infant mortality and childhood morbidity. NewEngland journal of medicine, 1985, 312:82-90.

McGuire J. The payoff from improving nutrition. Wash-ington, DC, The World Bank, 1996.

McKay H, Sinisterra L, McKay A, Gomez H, Lioreda P.Improving cognitive ability in chronically deprivedchildren. Science, 1978, 200:270-278.

Meeks-Gardner JM, Grantham-McGregor SM , ChangSM , Himes JH , Powell CA. Activity and behavioraldevelopment in stunted and non-stunted childrenand response to nutritional supplementation. Childdevelopment, 1995, 66:1785-1797.

Mehra R, Kurz K, Paolisso M. Child care options forworking mothers in developing countries. Washington,DC, ICRW, 1992.

Meisels SJ. Early intervention: a matter of context. Zeroto three, 1992, 3:1-6.

Mitzner K, Scrimshaw N, Morgan R, eds. Improving thenutrition food and nutrition program. Cambridge, MA,Massachusetts Institute of Technology, 1984.

Moffatt MEK, Longstaffe S, Besant J, Dureski C. Pre-vention of iron deficiency and psychomotor declinein high-risk infants through use of iron-fortifiedinfant formula: a randomized clinical trial. Journalof pediatrics, 1994, 125:527-534.

Morrow-Tlucak M, Haude RH, Ernhart CB. Breast-feeding and cognitive development in the first twoyears of life. Social science and medicine, 1988,26:635-639.

Muhilal, Permeisih D, Idjradinata YR, Muherdiyanti-ningsih Karyadi D. Vitamin A-fortified monosodiumglutamate and health, growth, and survival ofchildren: a controlled field trial. American journal ofclinical nutrition, 1988, 48:1271-1276.

REFERENCES


76

Myers R. et al. Pre-school education as a catalyst forcommunity development. Final Report to USAID/Lima, 1985.

Myers R. The twelve who survive: strengthening pro-grammes of early childhood development in the thirdworld. London, Routledge, 1992.

Olds DL, Kitzman H. Review of research on homevisiting for pregnant women and parents of youngchildren. The future of children, 1993, 3:53-92.

Olson RK. Developing indicators that predict benefit fromprenatal energy supplementation [Dissertation].Ithaca, NY, Cornell University, 1994.

Orr E. The contribution of new food mixtures to therelief of malnutrition. A second look. Food nutri-tion, 1977, 3:2-10.

Oski FA, Honig AS. The effects of therapy on thedevelopmental scores of iron-deficient infants.Journal of pediatrics, 1978, 92:21-25.

Ottenbacher KJ, Muller L, Brandt D, Heintzelman A.The effectiveness of tactile stimulation as a form ofearly intervention: a quantitative evaluation. Jour-nal of developmental and behavioral pediatrics, 1987,8:68-76.

Pelletier DL. The potentiating effects of malnutritionon child mortality: epidemiological evidence andpolicy implications. Nutrition reviews, 1994, 52:409-415.

Pelletier DL , Shrimpton R. The role of information inthe planning, management and evaluation of com-munity nutrition programmes. Health policy andplanning, 1994, 9:171-184.

Pelto GH, Tuomainen H. Evaluating social and psy-chological responses to vitamin A interventions. In:Garza C, Haas JD, Habicht J-P, Pelletier DL, eds.Beyond nutritional recommendations: implementingscience for healthier populations. Ithaca, NY, CornellUniversity Division of Nutritional Sciences,1996:145-162.

Pinstrup-Anderson P, Burger S, Habicht J-P, PetersonK. Protein-energy malnutrition. In: Jamison DT,Mosley WH , Measham AR, Bobadilla JL, eds. Dis-ease control priorities in developing countries. Oxford,World Bank/Oxford University Press, 1993: 391-420.

Pinstrup-Anderson P, Pelletier D, Alderman H, eds.Child growth and nutrition in developing countries.Ithaca, NY, Cornell University Press, 1995.

Pitt MM, Khandker SR. Household and intrahouseholdimpact of the Gramreen Bank and similar targetedcredit programs in Bangladesh. In: World Bank dis-cussion papers. Washington, DC, The World Bank,1996.

Pizarro F, Yip R, Dallman PR, Olivares M, HertrampfE, Walter T. Iron status with different infant feed-ing regimens: relevance to screening and preven-tion of iron deficiency. Journal of pediatrics, 1991,118:687-692.

Polan HJ, Leon A, Kaplan MD, Kessler DB, Stern DN,Ward MJ. Disturbance of affect expression in fail-ure to thrive. Journal of the academy of child andadolescent psychiatry, 1991, 30:897-903.

Pollitt E. Iron deficiency and cognitive function.Annual reviews of nutrition, 1993a, 13:521-537.

Pollitt E. Timing and vulnerability in research on mal-nutrition and cognition. Nutrition reviews, 1996,54:S49-S55.

Pollitt E. Forecasting the developmental impact of earlychildhood development programs. Paper presented atthe Conference on Early Childhood Development,Centre for International Health at Wye College,Kent, UK, April 4-7, 1998.

Pollitt E, Gersovitz MS, Garguilo M. Educational ben-efits of the United States school feeding program: acritical review of the literature. American journal ofpublic health, 1978, 68:477-481.

Pollitt E, Golub M, Gorman K, Grantham-McGregorS, Levitsky D, Schurch B, Strupp B, Wachs T. Areconceptualization of the effects of undernutritionon children’s biological, psychosocial, and be-havioral development. Social policy report, Societyfor research in child development, 1996, 10:1-21.

Pollitt E, Gorman KS, Engle PL, Martorell R, Rivera J.Early supplementary feeding and cognition: effectsover two decades. Society for research in child devel-opment, 1993, 58:1-99 (Monograph Series No. 235).

Pollitt E, Oh SY. Early supplemental feeding and healthpolicy. Food and nutrition bulletin, 1994, 15:208-214.

Pollitt E, Perez-Escamilla R. Growth improvements inchildren above 3 years of age: the Cali study. Jour-nal of nutrition, 1995, 125(4):885-93.

Porter M. Attractiveness of lactating females’ breastodors to neonates. Child development, 1989, 60:803-810.

Powell C, Grantham-McGregor SM. Home visitingof varying frequency and child development.Pediatrics, 1989, 84:157-164.

Puckering C, Pckles A, Skuse D, Heptinstall E,Dowdney L, Zur-Szpiro S. Mother-child interactionand the cognitive and behavioural development offour-year-old children with poor growth. Journal ofchild psychology and psychiatry, 1995, 36:573-595.

77

Ramakrishnan U, Latham MC, Abel R. Vitamin A sup-plementation does not improve growth of preschoolchildren: a randomized, double-blind field trial insouth India. Journal of nutrition, 1995, 125:202-211.

Ramey CT, Ramey SL. Early intervention and earlyexperience. American psychologist, 1998, 53:109-120.

Reed BA, Habicht J-P, Niameogo C. The effects of ma-ternal education on child nutritional status dependon socio-environmental conditions. Internationaljournal of epidemiology, 1996, 25:585-592.

Richter L. Many kinds of deprivation: young childrenand their families in South Africa. In: Eldering L,Leseman P, eds. Early intervention and culture: prepa-ration for literacy: the interface between theory andpractice. Paris, UNESCO, 1994.

Richter L, Bac M, Hay IT. Psycho-biological aspects ofhealth care of young children. South African familypractice, 1990, 11:490-497.

Rivera JA, Habicht J-P, Robson DS. Effect of supple-mentary feeding on recovery from mild to moder-ate wasting in preschool children. American journalof clinical nutrition, 1991, 54:62-68.

Rivera JA, Martorell R, Ruel MT, Habicht J-P, Haas JD.Nutritional supplementation during the preschoolyears influences body size and composition ofGuatemalan adolescents. Journal of nutrition, 1995,125(4S):1068S-1077S.

Rogan WJ, Gladen BC. Breastfeeding and cognitivedevelopment. Early human development, 1993,31:181-193.

Roncagliolo M, Garrido M, Williamson A, Lozoff B,Periano P. Delayed maturation of auditory brain-stem responses in iron-deficient anemic infants.Pediatric research, 1996, 39:20A.

Ross Laboratories. Iron nutrition in infancy: report of TheSixty-Second Ross Conference on Pediatric Research.Columbus, OH, Ross Laboratories, 1970.

Ruel MT. Growth monitoring as an educational tool,an integration strategy, and a source of information:a review of experience. In: Pinstrup-Anderson P,Pelletier D, Alderman H, eds. Child growth andnutrition. Ithaca, NY, Cornell University Press,1995:78-96.

Ruel MT, Habicht J-P. Growth charts only marginallyimproved maternal learning from nutrition educa-tion and growth monitoring in Lesotho. Journal ofnutrition, 1992, 122:1772-1780.

Ruel MT, Habicht J-P, Olson C. Impact of a clinic-basedgrowth monitoring programme on maternal nutri-tion knowledge in Lesotho. International journal ofepidemiology, 1992, 21:59-65.

Ruel MT, Pelletier D, Habicht J-P, Mason JB,Chobokoane CS, Maruping AP. A comparison ofmothers’ understanding of two growth charts inLesotho. Bulletin of the World Health Organization,1990, 68:483-491.

Rush D, Stein Z, Susser M. Diet in pregnancy: arandomized controlled trial of prenatal nutritionalsupplementation. National Foundation-March ofDimes Birth Defects original article series, Vol. XVI.New York, Alan Liss, 1980.

Satoto. Care and child feeding, growth, and development.Paper prepared for the Indonesian Conference onComplementary Feeding, Surabaya, Indonesia,January 1996.

Sazawal S, Bentley M, Black RE, Dhingra P, George S,Bhan MK. Effect of zinc supplementation on ob-served activity in low socioeconomic Indian pre-school children. Pediatrics, 1996, 98:1132-1137.

Schilling PR. Supplementary feeding programs; a criti-cal analysis. Revista saude publicacao, 1990, 24:412-419.

Scholl TO, Hediger ML, Fischer RL, Shearer JW. Anemiavs. iron deficiency: increased risk of preterm deliv-ery in a prospective study. American journal of clini-cal nutrition, 1992, 55:985-988.

Schweinhart LJ, Barnes HV, Weikart DP. Significantbenefits: the High/Scope Perry Preschool Study throughage 27. Ypsilanti, MI, High/Scope Press, 1993.

Schweinhart LJ, Weikart DP. Young children grow up: theeffects of the Perry Preschool Program on youths throughage 15. Ypsilanti, MI, The High/Scope Press, 1980.(Monograph of the High/Scope Educational Re-search Foundation, No. 7).

Scott KG, Nonkin RI, Holomon HA. Epidemiology ofchild developmental problems: the extent of the prob-lems of poor development in children from deprivedbackgrounds. Paper presented at the Conference onEarly Childhood Development, Centre for Interna-tional Health, at Wye College, Kent, UK, April 4-7,1998.

Scrimshaw NS. A look at the Incaparina experience inGuatemala: the background and history of Incapa-rina. Food and nutrition bulletin, 1980, 2:1-2

Seitz V, Apfel NH. Parent-focused intervention: diffu-sion effects on siblings. Child development, 1994,65:677-683.

Shah PM. The assessment of the use of home-basedchild’s health, growth, and development record inMalaysia. 30 October- 3 November, 1995. Report.

REFERENCES


78

Shrestha RM. Effect of iodine and iron supplementationon physical, psychomotor and mental development inprimary school children in Malawi [Dissertation].Wageningen, Wageningen Agricultural University,1994.

Sinisterra L, McKay H, McKay A, Gomez H, Korgi J.Response of malnourished children to multidisci-plinary intervention. In. Brozek J, ed., Behavioraleffects of energy and protein deficits proceedings. In-ternational Nutrition Conference. NIH PublicationNo. 79-1906. Washington, DC, Government Print-ing Office, 1979.

Smart JL. Malnutrition, learning, and behavior: recentadvances in laboratory animal research. In: Nutri-tion, health, and child development: research advancesand policy recommendations. Pan American HealthOrganization, Tropical Metabolism Research Unitof the University of the West Indies, and The WorldBank. Scientific Publication 566. 1998:1-13.

Smitasiri S, Dhanamitta S. Nutri-action analysis as aresearch strategy to improve nutrition information,education and communication interventions in Asia.Biomedical and environmental science, 1996, 9(2-3):290-5.

Soewondo S, Husaini M, Pollitt E. Effects of iron defi-ciency on attention and learning processes in pre-school children: Bandung, Indonesia. Americanjournal of clinical nutrition, 1989, 50:667-673.

St. Pierre RG, Layzer JI, Barnes HV. Two-generationprograms: design, cost, and short-term effectiveness.The future of children, 1995, 5:76-93.

Stein Z, Susser M. The Dutch Famine, 1944-45, andthe reproductive process. II. Interrelations ofcaloric rations and six indices at birth. Pediatricresearch, 1975, 9:76-83.

Strupp BJ, Levitsky DA. Enduring cognitive effects ofearly malnutrition: a theoretical reappraisal. Jour-nal of nutrition, 1995, 125:2221S-2232S.

Super CM, Clement J, Vuori L, Christiansen N, MoraJO, Herrera MG. Infant and caretaker behavior asmediators of nutritional and social interventions inthe barrios of Bogota. In: Field TM, Sostek AM,Vietze P, Leiderman PH, eds. Culture and early inter-actions. Hillsdale, NJ, L. Erlbaum Associates,1981:171-188.

Super CM, Herrera MA, Mora JO. Long term effects offood supplementation and psychosocial interven-tion on the physical growth of Colombian infantsat risk of malnutrition. Child development, 1990,61:29-49.

Susser M. Maternal weight gain, infant birth weight,and diet: causal sequences. American journal of clini-cal nutrition, 1991, 53:1384-1396.

Tandon BN. Nutritional interventions through primaryhealth care: impact of the ICDS projects in India.Bulletin of the World Health Organization, 1989,67:77-80.

Uauy R, de Andraca I. Human milk and breast feedingfor optimal mental development. The journal of nu-trition, 1995, 125(8S):2278S-2280S.

UNICEF. Consensus statement on zinc nutrition and pub-lic health in developing countries. New York, UNICEF,1993.

UNICEF. The state of the world’s children 1998. New York,UNICEF, and Oxford, Oxford University Press,1998.

USAID. Crucial elements of successful communitynutrition programs. In: Report of the Fifth Interna-tional Conference of the International Nutrition Plan-ners Forum, Seoul, Korea, August 1989. Washington,DC, USAID, 1989.

US Department of Health and Human Services. Creat-ing a 21st century head start. Final Report of theAdvisory Committee on Head Start Quality andExpansion. Washington, DC, US Department ofHealth and Human Services, 1993.

Valenzuela M. Maternal sensitivity in a developingsociety: the context of urban poverty and infantchronic undernutrition. Developmental psychology,1997, 33, 5, 845-855.

Varendi H, Porter RH, Winberg J. Does the newbornbaby find the nipple by smell? Lancet, 1994, 344:989-990.

Vazir S, Kashinath K. ICDS and psychosocial develop-ment. Nutrition news, 1995, 16:1-5.

Waber DP, Vuori-Christiansen L, Ortiz N, Clement JR,Christiansen NE, Mora JO, Reed RB, Herrera MG.Nutritional supplementation, maternal education,and cognitive development of infants at risk ofmalnutrition. American journal of clinical nutrition,1981, 34:807-813.

Wachs TD, McCabe G. The role of the environment inhuman nutritional research and intervention. InNutrition, health and child development: research ad-vances and policy recommendations. Pan AmericanHealth Organization, Tropical Metabolism ResearchUnit of the University of the West Indies, and TheWorld Bank. Scientific Publication 566. 1998, 14-31.

79

Walker SP, Grantham-McGregor SM, Powell CA, HimesJH, Simeon DT. Morbidity and the growth of stuntedand non-stunted children, and the effect of supple-mentation. American journal of clinical nutrition,1992, 56:504-510.

Walker SP, Powell CA, Grantham-McGregor SM. Earlychildhood supplementation and cognitive develop-ment, during and after intervention. In: Nutrition,health, and child development. Pan American HealthOrganization, Tropical Metabolism Research Unitof the University of the West Indies, and The WorldBank. Scientific Publication 566, 1998:69-81.

Walker SP, Powell CA, Grantham-McGregor SM, HimesJH, Chang SM. Nutritional supplementation, psy-chosocial stimulation, and growth of stunted chil-dren: the Jamaican study. American journal of clinicalnutrition, 1991, 54:642-648.

Walter T, Dallman PR, Pizarro F, Velozo L, Pena G.Effectiveness of iron-fortified infant cereal in pre-vention of iron deficiency anemia. Pediatrics, 1993,91:976-982.

Walter T, de Andraca I, Chadud P, Perales CG. Irondeficiency anemia: adverse effects on infant psycho-motor development. Pediatrics, 1989, 84:7-17.

Watkins WE, Pollitt E. Iron deficiency and cognitionamong school-aged children. In: Nutrition, health,and child development: research advances and policyrecommendations. Pan American Health Organiza-tion, Tropical Metabolism Research Unit of theUniversity of the West Indies, and The World Bank.Scientific Publication 566, 1998:179-197.

Weiss HB. Home visits: necessary but not sufficient.The future of children, 1993, 3:113-128.

West KP, Djunaedi E, Pandji A, Kusdiono, Tarwotjo I,Sommer A, and the Aceh Study Group. Vitamin Asupplementation and growth: a randomized com-munity trial. American journal of clinical nutrition,1988, 48:1257-1264.

Widmayer SM, Field TM. Effects of Brazelton demon-strations for mothers on the development of preterminfants. Pediatrics, 1981, 67:711-714.

Winkvist A, Jalil F, Habicht J-P, Rasmussen KM. Mater-nal energy depletion is buffered among malnour-ished women in Punjab, Pakistan. Journal ofnutrition, 1994, 124:2376-2385.

Winkvist A, Habicht J-P, Rasmussen KM. Linkingmaternal and infant benefits of a nutritional sup-plement during pregnancy and lactation. Americanjournal of clinical nutrition, 1998, 68:656-661.

Wise RP. The case of Incaparina in Guatemala. Foodand nutrition bulletin, 1980, 2:3-8.

Wollinka O, Keeley E, Burkhalter BR, Bashir N, eds.Hearth nutrition model: applications in Haiti, Vitnam,and Bangladesh. Washington, DC, BASICS andWorld Relief, 1997.

World Health Organization Expert Committee on: physi-cal status: the use and interpretation of anthropometry.Geneva, WHO, 1995. (TRS No. 854.)

Yoshikawa H. Long-term effects of early childhoodprograms on social outcomes and delinquency. Thefuture of children, 1995, 5:51-75.

Young ME. Investing in young children. In: World Bankdiscussion papers, No. 275. Washington, DC, TheWorld Bank, 1995.

Young ME. Early child development: investing in thefuture. Washington, DC, The World Bank, 1996.

Young ME. Introduction. In: Young ME, ed. Early childdevelopment: investing in our children’s future. Amster-dam, Elsevier, 1997:1:13.

Zeitlin MF. Child care and nutrition: the findings frompositive deviance research. Ithaca, NY, Cornell Inter-national, 1996 (Nutrition Monograph SeriesNo. 27).

Zeskind PS, Ramey CT. Fetal malnutrition: an experi-mental study of its consequences on infant devel-opment in two caregiving environments. Childdevelopment, 1979, 49:1155-1162.

Zigler E, Taussig C, Black K. Early childhood inter-vention: a promising preventative for juveniledelinquency. American psychologist, 1992, 47:997-100.

REFERENCES

Interventions for physical growth and psychological ...

Documents

Transcript of Interventions for physical growth and psychological ...