JOHN CHIMWEMWE PHUKA

Efficacy of Complementary Food Supplementationwith Lipid-Based Nutrient Supplements on

Growth of Malawian Children

ACADEMIC DISSERTATIONTo be presented, with the permission of

the Faculty of Medicine of the University of Tampere,for public discussion in the Jarmo Visakorpi Auditorium,

of the Arvo Building, Lääkärinkatu 1, Tampere, on September 25th, 2009, at 12 o’clock.

UNIVERSITY OF TAMPERE

Reviewed byProfessor Markku HeikinheimoUniversity of HelsinkiFinlandDocent Harri NiinikoskiUniversity of TurkuFinland

DistributionBookshop TAJUP.O. Box 61733014 University of TampereFinland

Tel. +358 3 3551 6055Fax +358 3 3551 7685 taju@uta.fiwww.uta.fi/tajuhttp://granum.uta.fi

Cover design byJuha Siro

Acta Universitatis Tamperensis 1451ISBN 978-951-44-7832-1 (print)ISSN-L 1455-1616ISSN 1455-1616

Acta Electronica Universitatis Tamperensis 884ISBN 978-951-44-7833-8 (pdf )ISSN 1456-954Xhttp://acta.uta.fi

Tampereen Yliopistopaino Oy – Juvenes PrintTampere 2009

ACADEMIC DISSERTATIONUniversity of Tampere, Medical SchoolFinland

Supervised byProfessor Per AshornUniversity of TampereFinlandProfessor Kenneth MaletaUniversity of MalawiMalawi

i

SUMMARY

In low income countries, 170 million under-five year-old children are stunted.

Prevalence of underweight and wasting are also high. In the past few years, several

analyses have documented that at least 35% of all childhood deaths are attributable to

underweight; that most of these deaths occur in moderate to severe underweight; that

undernutrition (wasting, stunting, underweight) are common between 6 and 24

months of age; that stunting is associated with poor cognition and development, and

lower physical capacity and income in adulthood; and that proper nutrient

supplementation can mitigate the effects of undernutrition.

Efforts to promote normal infant growth have so far produced modest growth

outcomes and infants in low income countries continue to experience growth faltering

during this critical period. The relatively new supplementary strategy food products

commonly known as the lipid-based nutrient supplements (LNS) have been successful

in treating severe forms of acute malnutrition in the community and are well accepted,

suggesting that similar products could be useful for promotion of growth and

prevention of undernutrition in infancy. Therefore, the present study tested the effects

of LNS as a complementary food on promotion of growth and development of infants

and young children, and its effect as a supplementary food on secondary prevention of

childhood undernutrition. Firstly, to test the effect of LNS as a supplementary food on

secondary prevention of severe acute undernutrition we conducted a supplementary

ii

feeding trial in which the infants received either daily doses of 50 g LNS called

fortified spread (FS50) or 71g iso-energetic corn-soy blend (CSB) called Likuni Phala

(LP). Secondly, to test the effect of provision of LNS as a complementary food

supplement for primary prevention of undernutrition and promotion of growth and

development we conducted a randomised controlled complementary feeding trial in

which the infants in 2 trial groups received LNS daily doses of either 50 g (FS50) or

25 g (FS25) and those in the control group received 71 g (LP) for 12 months.

In the supplementary feeding trial, mean weight-for-age increased by 0.22 (95% CI

0.07 to 0.37) and 0.28 (0.18 to 0.40) Z-score units in the LP and FS groups,

respectively. Comparable increase for mean weight-for-height was 0.39 (0.20 to 0.57)

and 0.52 (0.38 to 0.65) units. Recovery from underweight and wasting was 20% and

93% in the LP-group and 16% and 75% in the FS-group. Few individuals recovered

from stunting and mean length-for-age was not markedly changed.

In the complementary trial the mean weight and length gains in LP, FS50 and FS25

groups were 2.37, 2.47, and 2.37 kg (P = 0.66) and 12.7, 13.5, and 13.2 cm (P =

0.23), respectively. In the same groups, cumulative 12-month incidence of severe

stunting was 14.0%, 0.0% and 4.0% (P = 0.01), severe underweight 15.0%, 22.5%

and 16.9% (P = 0.71), and severe wasting 1.8%, 1.9% and 1.8% (P > 0.99). There

was a significant interaction between baseline length and intervention (P = 0.04);

among children with below-median length at enrolment, those given FS50 gained on

average 1.9 cm (0.3 to 3.5; P = 0.02) more than individuals receiving LP.

iii

In the supplementary feeding trial, the mean gain in haemoglobin concentration after

12 weeks supplementation was 3.8 g/L in the FS50-group compared to 1.5 g/L in the

LP-group. Compared to enrolment measurements in all the 3 trial groups, the mean

haemoglobin concentration declined marginally at the end of the intervention. The

smallest decline occurred in the LNS groups; <1 g/L, 4 g/L and 7 g/L in the FS50,

FS25 and LP groups, respectively.

At 18-months of age, the mean ± SD mental ages in the LP, FS50 and FS25 groups

were 17.9 ± 1.3, 17.9 ± 1.3 and 17.9 ± 1.2 months (P > 0.99), respectively. Length-

for-age Z-score (LAZ) gain during the intervention period and maternal education

were associated with developmental outcome at age of 18-months (P = 0.03 and P =

0.04; respectively).

The cumulative 36-month incidence of severe stunting was 19.6% in LP, 3.6% in

FS50 and 10.3% in FS25 groups (P = 0.03). Mean weight for age Z-score (WAZ)

change was -1.09, -0.76 and -1.22 (P = 0.04), mean LAZ change -0.47, -0.37, and -

0.71 (P = 0.10), and mean weight for height Z-score WHZ change -1.52, -1.18, and -

1.48 (P = 0.27). All differences were more marked among individuals with baseline

LAZ below median. Differences in length developed during the intervention, at ages

10-18 months, whereas weight differences continued to increase after the intervention.

iv

In conclusion, short-term supplementation with lipid-based nutrient supplements

(LNS) or corn-soy blends (CSB) improve weight gain similarly but neither of them

has short term effect on length. Long-term provision of LNS results in higher length

gain, weight gain and prevents incidence of stunting than provision of iso-energetic

CSB. Similarly, LNS may promote increased haemoglobin concentration more than

CSB. Provision of the multiple micronutrients through fortified LNS and CSB

supplements have similar effect on child development.

v

TABLE OF CONTENTS

SUMMARY............................................................................................................... iTABLE OF CONTENTS ...........................................................................................vLIST OF ORIGINAL ARTICLES............................................................................ viLIST OF FIGURES AND TABLES....................................................................... viiiABBREVIATIONS.................................................................................................. ix1 INTRODUCTION............................................................................................. 12 LITERATURE REVIEW .................................................................................. 4

2.1 Approach to the literature review ............................................................... 42.2 Biological and environmental factors of growth ......................................... 52.3 Assessment of growth and nutritional status..............................................152.4 Epidemiology of undernutrition and early growth faltering .......................222.5 Epidemiological strategies for promotion of growth and development.......262.6 Justification for the present study ..............................................................48

3 AIMS ...............................................................................................................524 METHODS ......................................................................................................54

4.1 Approach to the study ...............................................................................544.2 Study setting and subjects .........................................................................57

5 Study subjects...................................................................................................575.1 Nutritional interventions ...........................................................................585.2 Data collection ..........................................................................................605.3 Statistical approach ...................................................................................655.4 Ethical compliance....................................................................................67

6 RESULTS ........................................................................................................686.1 Supplementary feeding to moderately underweight infants (study I)..........686.2 Supplementation of complementary food to health infants (Study II).........726.3 Haematological changes (Study I and II)...................................................796.4 Compliance to the interventions supplements (Study I and II) ...................806.5 Morbidity and serious adverse events (Study I and II) ...............................80

7 DISCUSSION ..................................................................................................827.1 Strength and weaknesses for the study.......................................................827.2 Growth promotion and management of undernutrition ..............................857.3 Promotion of recovery from and prevention of anaemia ............................937.4 Developmental outcomes after complementary feeding.............................94

8 SCIENTIFIC CONCLUSIONS ........................................................................979 PUBLIC HEALTH IMPLICATION AND FUTURE RESEARCH NEEDS......9910 ACKNOWLEDGEMENTS ........................................................................10211 REFERENCES ...........................................................................................10512 APPENDICES............................................................................................142

12.1 Geographical and demographic profile of Malawi ...................................14212.2 Anthropometric measurements ................................................................14412.3 Mid-Upper Arm Circumference ..............................................................145

vi

LIST OF ORIGINAL ARTICLES

This thesis is based on the following original articles, referred to in the text of the

article by their roman numerals.

I. Phuka J, Thakwalakwa C, Maleta K, Cheung YB, Briend A, Manary M,

Ashorn P. Supplementary feeding with fortified spread among moderately

underweight 6-18-month-old rural Malawian children. Matern Child Nutr.

2009 Apr;5(2):159-70

II. Phuka JC, Thakwalakwa C, Maleta K, Cheung YB, Briend A, Manary MJ,

Ashorn P. Complementary feeding with fortified spread and incidence of

severe stunting in 6-to-18 month-old rural Malawians. Arch Pediatr Adolesc

Med. 2008 Jul;162(7):619-26. Erratum in Arch Pediatr Adolesc Med. 2008

Oct;162(10):942

III. Phuka JC, Thakwalakwa C, Maleta K, Cheung YB, Briend A, Manary MJ,

Ashorn P. Postintervention growth of Malawian children who received 12-

month dietary complementation with a lipid-based nutrient supplement or

maize-soy flour. Am J Clin Nutr. 2009 Jan;89(1):382-90.

vii

IV. Phuka JC, Thakwalakwa C, Maleta K, Cheung YB, Briend A, Manary M,

Ashorn P. Childhood development of 18 months old Malawians after a year of

complementary feeding with lipid-based nutrient supplements or corn-soy

blend. (Submitted)

viii

LIST OF FIGURES AND TABLES

FIGURES

Figure 1: The ICP growth model for boys. ................................................................. 7Figure 2: Length and length velocity for a normal Pakistani ...................................... 8Figure 3: Causes and consequences of undernutrition. ..............................................12Figure 4: Overall study design for the thesis .............................................................56Figure 5: Anthropometric population shift after supplementary feeding ....................72

TABLES

Table 1: Comparison of the two trials conducted for the study..................................55

ix

ABBREVIATIONS

ALA Alpha-linolenic acid

ARI Acute respiratory infection

ANOVA Analysis of variance

BDI-II Battelle Developmental inventory II

BMI Body mass index

BSID-II Bayley Scales if Infant Development II

CDC United States Centers for Disease Control

CI Confidence interval

CRF Case reporting form

CSB Corn-soy blend

CSM Corn-soy milk

DALY Disability-adjusted life year

DHA Docosahexaenoic acid

DNA Deoxyribonucleic acid

x

DSMB Data safety management board

DPT Vaccine against Diphtheria, Pertussis and Tetanus

EPI Expanded programme on immunization

FS Fortified spread

GDP Gross domestic product

GNI Gross national income

HAZ Height-for-age Z-score

HIV Human immunodeficiency syndrome

ICP The infancy, childhood and puberty growth model

ITN Insecticide-treated bed net

LA Linoleic acid

LAZ Length-for-age Z-score

LCPUFA Longchain polyunsaturated fatty acid

LMS Parameters need to generate percentile and Z-scores: median (M), the

generalized coefficient of variation (S), and the power in the Box-Cox

transformation (L)

LNS Lipid-based nutrient supplement

LP Likuni Phala

xi

MUAC Mid-upper-arm circumference

MGRS The World Health Organisation (WHO) Multicentre Growth Reference

Study

MSEL Mullen Scales of Early Learning

NCHS The United States National Center for Health Statistics

NRU Nutrition Rehabilitation Unit

NTD Neural tube defect

ORS Oral rehydration salts

ODA The net official development assistance

PCR Polymerase chain reaction

PDCAAS Protein digestibility-corrected amino acid score

PUFA Polyunsaturated fatty acid

RUTF Ready-to-use therapeutic food

SAE Serious adverse event

SD Standard deviation

SOP Standard operating procedures

TEM Technical error of measurement

xii

UNICEF United Nations Children’s Fund

USAID U.S. Agency for International Development

USDA U.S. Department of Agriculture

WAZ Weight-for-age Z-score

WHO World Health Organisation

WHZ Weight-for-height Z-score

WLZ Weight-for-length Z-score

1

1 INTRODUCTION

In absence of optimal environmental conditions, growth may either be down regulated

by internal programming early in foetal growth, or may be impaired. The impaired

growth may be due to inadequate substances for tissue accretion such as amino acids

and micronutrients, due to inadequate energy to support the accretion process, or due

to slower metabolism. The result of these adaptations or inadequacies may manifest as

growth faltering or as inadequate neuro-motor development. After short exposure to

undernutrition children commonly present with wasting. Subsequently after prolonged

undernutrition, wasting may progress to underweight or stunting (a statural growth

failure) and it is associated with poor child development.

Evidence from the WHO Global Database on Child Growth and Malnutrition based

on a collection of surveys worldwide indicate that prevalence of undernutrition based

on stunting, wasting and underweight is high in developing countries but on the

decline in all other parts of the world except in Africa (de Onis and Blössner 2003, de

Onis et al. 2004a, UNICEF 2007). Prevalence of stunting in developing countries in

under-five years old children was estimated to decline from 30% in 2000 to 16% in

2020, and for the same period in Africa from 35% to 31% (de Onis and Blössner

2003). Similarly, worldwide forecast for underweight from 1990 to 2015 were 30% to

19 %, but that is rising in Africa from 24% to 27% (de Onis et al. 2004a). Stunting

affects approximately 160 million children less than 5 years of age, with prevalence of

2

approximately 46% in southern Asia and 38% in sub-Saharan Africa (de Onis et al.

2004a, UNICEF 2007)

Maternal and childhood undernutrition remain a big cause of total global disease

burden. In 2002, it was estimated that childhood and maternal underweight were

together responsible for 9.5% of world disability-adjusted life years (DALY) (Ezzati

et al. 2002). Recent estimates suggest no decline; in 2008, maternal and childhood

undernutrition were estimated to be responsible for 11% of the total global disease

burden and 35% of child deaths (Black et al. 2008).

Stunting has been associated with poor child development, including poor sensori-

motor, cognitive-language and social-emotional development (Meeks Gardner 1999,

Cheung et al. 2001, Siegel et al. 2005, Olney et al. 2007). Childhood stunting and

poor preschool child development are also associated with poor school performance,

reduced physical performance and reduced income in adulthood, as well as lower

birth weight of the next generation (Mendez and Adair 1999, Hass et al. 1995, Daniel

and Adair 2004, Grantham-McGregor 2007, Victora 2008).

Global strategies to promote child growth and development and to prevent

undernutrition are at the top of global health priorities. Despite efforts dating to

several decades to provide complementary food supplements, results on growth are

modest and undernutrition remains a big cause of global disease burden (Marchione

3

2002, Black et al. 2008). Lipid-based nutrient supplements have been effective in

treating severe undernutrition in the communities and they are generally well accepted

by infants (Manary et al. 2004, Sandige et al. 2004, Ciliberto et al. 2005, Ndekha et al.

2005, Collins et al. 2006, Bhutta et al. 2008, Dewey and Adu-Afarwuah 2008). It has

been suggested that this approach may also be useful in promotion of child growth

and development and in prevention of undernutrition (Nestel 2003). Hence, the

present population-based randomised controlled study was conducted to test the

efficacy of the lipid-based nutrient supplements (LNS) as complementary food

supplements in Malawi.

4

2 LITERATURE REVIEW

2.1 Approach to the literature review

The purpose of this review is to give the background rationale for the current study.

Thus, the review presents: the knowledge on fundamentals of human growth; the

empirically guided methods of assessing human growth and nutritional status; the

timing of growth faltering in developing countries; the consequences of growth

faltering and its associated preceding factors; and the lessons learned from prior

efforts to promote human growth and development. Subsequently, the justification of

the current study, a feasibility assessment on nutrition strategies for promoting infant

growth and development, is presented. To give adequate width and depth of

background knowledge on promotion of growth and prevention of undernutrition,

both systematic and random search procedures were used. Published electronic

references were obtained by searching strategically relevant key words. The search

was mostly conducted through but not restricted to the PubMed of U.S National

Library of Medicine and the National Institutes of Health. Relevant references cited in

retrieved literature were identified and reviewed further. The main focus was on

primary results rather than secondary results, although findings from a few critical

reviews or books were included. As efforts to promote growth through

supplementation date back to at least 4 decades ago, accessible historical literature

dating that far was reviewed but more emphasis was given to recent data.

5

2.2 Biological and environmental factors of growth

2.2.1 Pattern of human growth

Human growth, a stature increasing process, involves somatic accretion,

differentiation and functional maturation of organs. Starting with conception, the

growth process continues until cessation in late teens or early twenties (Tanner 1963,

Karlberg 1989, Cameron 1992, Coley et al. 2006). Human postnatal growth is best

represented by a trace on a height growth curve or distance growth chart from birth to

adulthood (Tanner 1952). These charts are important tools for assessing the pattern of

growth over time. Traditionally, the series of measurements for construction of the

growth chart are conducted at intervals long enough to produce growth curves that

look smooth and continuous curve. More detailed measurements that are carried out at

shorter intervals, however, reveal that postnatal growth is not necessarily smooth, but

it rather occurs by saltation spaced by stasis (Lampl et al. 1992). Further evaluation

suggests that rather than one continuous curve starting from birth to adulthood, the

growth curve is a composition of a series of different overlapping growth curves. A

number of mathematical models demonstrating decomposed growth curves have been

suggested and have been rigorously evaluated before (Preece and Baines 1978, Stützle

et al. 1980, Karlberg 1989, van Dommelen et al. 2004).

The Infancy, Childhood, and Puberty Growth Model (ICP-Growth Model) constructed

by Karlberg is the most conceptually convincing mathematical model of human

6

growth and it has been accepted by other authors (van Dommelen et al. 2004,

Cameron 2007). This model superimposes curves from three independent growth

phases which are summed up to produce one postnatal growth curve (Karlberg et al.

1989) (Figure 1). The three independent phases in this growth-model curve correlates

better with regulatory growth mechanisms (Karlberg 1989) and it fits empirical data

better when compared to other models (van Dommelen et al. 2004). Each phase is

characterised by an initial rapid growth that later slows down before the next phase

commences (Karlberg 1989, Karlberg et al. 1994). The fastest growth occurs in the

first phase that is also characterised by rapid deceleration before it ceases at about 3-4

years of age (Karlberg et al. 1994) (Figure 2). The childhood growth phase starts at

about one year of age and slowly decelerates until it is interrupted by the pubertal

growth spurt at 13-15 years. However, rarely before the pubertal phases a small

interruption may occur at 6-8 years and this is known as the juvenile growth spurt

(Tanner and Cameron 1980).

7

Figure 1: The ICP growth model for boys.The mean functions are plotted for each of the three components as well as thecombined growth tC: Onset of Childhood component tE: Age at end of growth.(Karlberg 1989)

8

Figure 2: Length and length velocity for a normal PakistaniMonthly recorded values have been used in (a), every second month values in (b)(Karlberg 1994)

2.2.2 Regulation of human growth

The phenotypic variation of height is a summation of the genetic variation and the

random environmental effects (Silventoinen et al. 2008a). Intrinsic factors affecting

growth from infancy to adulthood include at least genetics, morphogens and

hormones (Owens 1991, Tabata and Takei 2004). Externally, the body obtains

substrates for building or maintaining the body (Owen 1991). Additionally, the

genetic control may interact with the environment independently; for example, the

9

genes may influence caloric intake, caloric expenditure and metabolism (Marti et al.

2004).

Regulation of heritability of human growth is to a large extent determined by genetics.

Evidence of the role of heritability on growth comes from strong height correlation

between siblings or between parents and their children. In adulthood, height

heritability ranges from 50 % in full sibling to above 90 % in monozygous twins

(Silventoinen et al. 2003a, 2008a, Czerwinski et al. 2007). Additional evidence of

strong genetic influence on growth traits such as height is a phenomenon known as

canalisation; this phenomenon refers to the tendency of growth traits to follow a

certain tendency or trajectory over time (Tanner 1963, Cameron 1992, Cameron

2007). Follow-up studies have shown that children failing to grow because of medical

conditions have attained expected heights after a correction of their medical condition

(Tanner 1963, Cameron 1992). The genetic control of stature has been shown to

increase with age (Dubois et al. 2007, Silventoinen et al. 2008a), suggesting that

environmental effect on growth is more important in early childhood. Apart from

height, other characteristics of growth showing strong heritability include muscle

strength, body weight, body mass index (BMI) and body composition (Schousboe et

al. 2004, van Dommelen et al. 2004, Dubois et al. 2007, Silventoinen et al. 2007,

2008b).

Several hormones participate in intrinsic regulation of human growth. Some hormones

like growth hormone are produced specifically to promote growth while for others

10

like insulin this is a secondary effect. Correlation of some hormones with different

phases of growth helps in understanding the timing of their hormonal influence on

growth. Each phase of the ICP growth model, for example, is regulated by at least one

hormone. However, the effects of some hormones are not limited to a single phase

(Karlberg 1989). Hormones regulating prenatal life are thought to continue to infancy

growth phase up to the end of the first postnatal year (Karlberg 1989). Animal studies

suggest that the prenatal growth is regulated by at least morphogens and hormones

(Fowden 1995). Morphogens are thought to be responsible for organ differentiation

and hormones are considered to be responsible for the general growth process. Insulin

like-growth factors 1 and 2, insulin and glucocorticoids are regarded as the most

important hormones regulating growth during this phase (Fowden 1995, 2003, Kaku

et al. 2007). Although growth hormone is present in foetal life, its effect has at least

not been shown to include regulation of prenatal growth (Owens 1991, Lee at al.

2003). After the first year of life, growth hormone in the presence of thyroid hormone

regulates growth until puberty starts at about 11-15 years (Karlberg 1989), when

additional growth effect comes from sexual hormones (Silventoinen 2003b).

2.2.3 Environmental factors influencing human growth

External effects from the environment compliment the intrinsic growth factors. The

environmental effects on growth include at least nutrition, geographical location,

physical environment, social economic status, morbidity, and education (Martorell

1995a, Silventoinen 2003b, Black et al. 2008). These environmental factors influence

both the intrauterine and the postnatal growth phases. In foetal life, uterine conditions

11

account for approximately half of the variation in birth weight, an indication of a

strong environmental influence on intrauterine growth (Adair and Pollitt 1985, Dubois

et al. 2007). Postnatally, the environment influence on growth is well known and its

largest impact is thought to be in the first few years (Martorell 1995a, Silventoinen

2003b, 2008a, Black et al. 2008). During the first 2 years of life, environmental

factors account to about half or more of the growth variation; thereafter, the

environmental influence on growth gradually decline with age to an extent that they

accounts for less than 10% growth variation towards adulthood (Demerath et al. 2007,

Dubois et al. 2007, Silventoinen et al. 2008a). Growth status is, thus, a good indicator

of a child’s general health, nutrition and socioeconomic status (Silventoinen 2003b).

Inadequate nutrition or undernutrition, a major risk factor of growth failure, may

occur due to disturbance at any point or level along the nutrient’s pathway from the

environment into the body (Figure 3). Along this pathway, three possible levels are

notable and will be described further below: the population, the household and the

individual levels (UNICEF 1998). Firstly, at the population level, basic causes operate

by reducing the quantity, the quality and the utilisation of potential resources; and

these basic causes include at least political, cultural, religious, economic and social

systems, including women’s status. Likewise, at the population level, inadequate or

inappropriate knowledge and discriminatory attitudes limit household access to the

actual resources. Secondly, at the household level, underlying causes operate through

insufficient access to food, inadequate maternal and child care practices, poor water

and sanitation, and poor health services. Thirdly, at the individual level, the

immediate causes of undernutrition operate through inadequate food intake and

12

morbidity. Thus, in low income settings, all these environmental factors act through

diseases or inadequate food intake or interaction of the diseases and inadequate intake

(Black et al. 2008, UNICEF 1998) to cause undernutrition and the subsequent growth

failure.

Figure 3: Causes and consequences of undernutrition.Framework of the relations between poverty, food insecurity, and other underlyingand immediate causes to maternal and child undernutrition and its short-term andlong-term consequences (Black et al. 2008)

Wide spread childhood infections in low income countries have been associated with

undernutrition (Black et al. 1984, Rowland et al. 1988, Adair and Guilkey 1997, Assis

et al. 2005, Olney et al. 2007, Walker et al. 2007, Hall et al. 2008). Infections can

cause undernutrition through reduction of nutrient intake and altering metabolism.

13

State of disease reduces in many occasions nutrient intake by reducing appetite,

reducing absorption, and loss of nutrients (UNICEF 1998). Reduced food intake

following loss of appetite is common in childhood illnesses including malaria,

diarrhoea, respiratory tract infections and helminth infestations. Because of altered

metabolism, nutrient demand also increases during childhood illnesses. The

pathophysiology of infections such as malaria and acute respiratory tract infections

(ARI) usually increases the energy demand of the body (Scrimshaw 1977).

Inadequate food intake negates nutrition; the process whereby living organisms take

in and transform extraneous solid and liquid substances necessary for growth,

maintenance of life, normal functioning of organs and production of energy. Short

term undernutrition often causes preclinical micronutrient related medical conditions

or wasting while prolonged exposure to undernutrition is often characterised by

stunting. Both wasting and stunting may results in underweight. In the low income

countries, micronutrient deficiencies commonly associated with undernutrition and

the associated growth failure include at least zinc, iron, iodine, selenium, and essential

fatty acids (Allen 1994).

Socio-economic status is another noticeable environmental factor that is associated

with growth. Some of the elements contributing to socio-economic status include

income, possessions, education, water, sanitation, family size and living environment.

Poor economic status is probably the most important socio-economic factor as poverty

is associated with several other factors: inadequate food, poor sanitation and hygiene,

14

infection, and maternal stress and depression (Fotso and Kaute-Defo 2005, Grantham-

McGregor et al. 2007). Education, especially that of the mothers is another socio-

economic factor strongly associated with growth and nutrition in childhood (Ivanans

1975, Liu et al. 1998, Pena et al. 2000, Griffiths et al. 2004, Larrea and Kawachi

2005, Walker et al. 2005, Hatt and Water 2006). Maternal education has been

consistently associated with positive growth and other health outcomes despite being

assessed differently in many studies (Fotso and Kaute-Defo 2005). Educated mothers

are more likely to use health promotion facilities and programmes, and may as well

earn more money enhancing their capacity to support children. Another factor

influencing height is secular trends; over the years in most but not limited to high

income countries, the mean population height has been increasing (Henneberg and

Berg 1990, Silventoinen 2003b). Reduced stress from disease burden and improved

nutritional environment are thought to contribute to secular trends (Silventoinen

2003b).

Taken together, human growth is genetically determined and biologically modified by

intrinsic as well as environmentally determined growth factors. The average

population growth potential of children is similar in both low and high income

countries (Bhandari et al. 2002, WHO 2006). Therefore, childhood growth failure at

population level in the low income countries is likely to be caused by environmental

factors rather than the genetically determined intrinsic growth regulators. Thus,

managing these environmental risk factors allow for an external opportunity for

promoting growth, prevention of growth failure and stimulation of catch-up growth.

15

2.3 Assessment of growth and nutritional status

2.3.1 Methods of assessing human growth and nutritional status

Growth is often assessed using anthropometric measurements; of which common

measurements include length/height, weight, head circumference and mid-upper arm

circumference (MUAC). In auxology, anthropometry usually refers to measurements

of external morphological (Ulijaszek and Deborah 1999) although anthropometry may

more generally refer to all measurement of man which ideally includes psychological,

physiological and anatomical traits. Collectively, these measurements are also

important for nutritional assessment, screening, surveillance and monitoring.

Nutritional assessment methods, however, also take many other forms appropriate for

characterising different stages of nutritional deficiency. These approaches include the

clinical assessment, nutrient balance, functional indicators of nutritional sufficiency

(physiological, biochemical and molecular techniques), anthropometric methods and

measures of optimal nutrient intake (WHO/FAO 2004). In longitudinal studies and

surveillance of undernutrition, anthropometric methods are widely used because they

are technically easier to use than other methods. For surveillance, anthropometric data

can easily be plotted on a distance growth chart from where failure to thrive can easily

be detected (Figures 1 and 2). Other methods like dietary assessments and

biochemical techniques are expensive and require substantial training.

16

2.3.2 Methods of assessing child development

Standardised tests for child development are psychometric measures designed to

inventory individual’s abilities and provide a comparison between the individual

performance on the test to that of the reference or standard. They are designed for

global assessment by providing an inventory of key developmental milestones;

however some also provide the option of assessing an individual’s ability within a

specific developmental domain. In general, standardised assessments comprise formal

tester-administered measures in which a qualified examiner administers the test

adhering to a stringent administration and scoring protocol. The objectivity of these

tests makes them ideal for large-scale use, such as routine follow-up, in which

multiple examiners may administer the test and which comparisons are made between

the results of individuals or groups. The most commonly used standardised

developmental tests at least includes the Mullen Scales of Early Learning (MSEL), the

Battelle Developmental inventory II (BDI-II), the Griffiths Mental Developmental

Scales- Revised (Griffiths Scales), and the Bayley Scales if Infant Development II

(BSID-II) (Johnson and Marlow 2006).

2.3.3 Standardisation of assessments methods

17

2.3.3.1 Standardisation of anthropometric measurements

To interpret growth measurements appropriately, anthropometric indices are useful.

The indices are calculated by combining at least two anthropometric measurements

(WHO 1995). The anthropometric indices are used to compare an individual child or a

group of children with a reference population, and they can be adjusted for age and

sex. In children, the most commonly used anthropometric indices are weight-for-

height, height-for-age, and weight-for-age (for wasting, stunting and underweight,

respectively) (WHO 1995). The common expressions of anthropometric indices in

children are Z-scores, percentiles, or percent of median. These reporting systems are

defined by WHO as follows:

Z-score- the deviation of the value for an individual from the median value of

the reference population divided by the standard deviation of the reference

population.

Percentile - the rank position of an individual on a given reference

distribution, stated in terms of what percentage of the group the individual

equals or exceeds.

Percent of median - the ratio of a measured value in the individual to the

median value of the reference population, expressed as a percentage.

Because of better utilisation characteristics, World Health Organisation (WHO)

recommends use of Z-score to standardise weight and height and for assessment of

malnutrition (WHO 1995). The Z-score system is preferable because it satisfies all of

18

the following four characteristics: adhering to reference distribution; working on

linear scale hence permitting summary statistics; using uniform criteria across indices;

and detecting changes at extremes of the distribution; the percentile and the percent of

median, each satisfies only two of those. A Z-score of less than -2 or greater than +2

is considered as evidence of malnutrition; the lower extremity is regarded as

undernutrition and the upper extremity is regarded as overnutrition. Of these,

undernutrition is the common problem for children in low income countries (de Onis

and Blössner 2003, de Onis et al. 2004a, UNICEF 2006, Black et al. 2008).

2.3.3.2 Standardisation of developmental scores

During construction, standardised developmental test are subject to rigorous empirical

analyses to assess the psychometric properties of the scale. Firstly, the reference

population is selected to be representative of the population for whom the test is

designed and the sample size has to large enough. Secondly, the developmental test

has to be reliable with high internal consistency, repeatability and inter observer

reliability. Lastly, the test has to be valid with high content validity, construct validity,

concurrent validity and predictive validity. Although some of the standardised tests

available were constructed on a wide ethnic background population none included

children from low income settings. As such when used in such population

standardised developmental tests need to be re-validated.

19

2.3.4 Reference population

2.3.4.1 Anthropometric reference population

Determination of nutritional status is made by comparing anthropometric

measurement to that of a growth reference or a growth standard constructed from

reference population. The US National Center for Health Statistics (NCHS)/ WHO

1977 international references and the CDC 2000 growth charts have been used until

recently (WHO 1983, Kuczmarski 2002). They both have received criticism on either

reference population or statistical methods employed in their construction.

The WHO/ NCHS growth references were criticised that they do not represent

children from low income settings, because the reference covering birth to three years

were constructed based on largely formula fed, affluent children of European ancestry

from a single community in the United States of America (Kuczmarski et al. 2002).

Breast fed children grow rather differently from largely bottle fed children (de Onis

and Habicht 1996, de Onis 1997, Dewey 1998, Victora et al. 1998, Garza and de Onis

1999). Additional criticisms have been on centile smoothing procedure and estimation

of extreme centiles (Dibley et al. 1987, Kuczmarski et al. 2002); and on poor

transition from length to height at 24 months because they were based on two

different data sets (Kuczmarski et al. 2002). At the time the NCHS/WHO growth

curves were constructed, the statistical methods available were too limited to correctly

model the pattern and variability of growth. Likewise, although the CDC 2000 growth

charts construction employed LMS methods that fit skewed data adequately and

20

generate fitted curves that follow closely empirical data, they were also based on

restricted population (Cole and Green 1992, Kuczmarski et al. 2002, WHO 2006).

It was with this background that the WHO Child Growth Standards were launched in

Geneva in on 27th April 2006 as a standard rather than a reference. Although used

interchangeably, a standard define how children should grow, therefore deviations

from the pattern it describes are evidence of abnormal growth; on the other hand, a

reference does not provide as sound a basis for such value judgements, therefore it is

only a tool for grouping and analysing data, and a common basis for comparing

population (WHO1995, 2006). The WHO Child Growth Standards were constructed

from widely diverse ethnic backgrounds and cultural settings (Brazil, Ghana, India,

Norway, Oman and USA) following the WHO Multicentre Growth Reference Study

(MGRS). The construction of these growth curves followed a methodical process

where initially the existing methods were examined, including type of distributions

and smoothing techniques; then flexible software allowing comparison of alternative

methods and generation of the curves was selected; and lastly the selected approach

was systematically applied to generate models that best fit the data (Borghi 2006,

WHO 2006). Like in the development of the CDC 2000 growth charts, LMS

procedures were employed as a simplification of the Box-Cox power exponential

(Rigby and Stasinopoulos 2004, Borghi 2006) and cubic splines were used for curve

smoothing (WHO 2006).

21

The MGRS is unique in that it was purposely designed to produce a standard by

selecting healthy children living under conditions likely to favour the achievement of

their full genetic growth potential (WHO 2006). Furthermore, the mothers of the

children selected for the construction of these standards engaged in fundamental

health promoting practices: exclusive breast feeding for at least 4 months;

introduction of complementary feeding by the age of 6 months with continued breast

feeding up to 12 months; and the mother not smoking before and after delivery (de

Onis et al. 2004b). Only single term birth children in absence of significant morbidity

were included (de Onis et al. 2004b).

2.3.4.2 Developmental scores norm-reference population

During standardisation, a test is administered to a large group of children for whom it

is designed; this group forms the normative sample or normal population. An

individual’s score obtained on the test is essentially a comparison to this normative

data and is used to determine how the individual is developing in relation to the norm.

Standardised tests thus yield norm-referenced or standardised scores. Standard score

typically have a mean score of 100 and standard deviation (SD) of 15. A Z-score of <-

2 SD is typically classified as delayed performance. Other norm-referenced score that

may be yielded are percentiles and age equivalents. Age equivalents are easy to

understand but should be used cautiously as developmental age below chronological

age may actually be within normal range of standardised scores.

22

2.4 Epidemiology of undernutrition and early growth faltering

2.4.1 Age distribution and determinants of early growth failure

Growth faltering in developing countries may start in utero or soon after birth, is

usually pronounced between 12 and 18 months and may continue until 40 months,

after which the decline often stops (Martorell et al. 1995b, Shrimpton et al. 2001,

Maleta et al. 2003a, Victora et al. 2008). Height deficit experienced at 2 or 3 years

usually persists with similar magnitude until adulthood (Coley et al. 2006, Victora et

al. 2008). Although catch up growth has been reported, it is often in relation to a

reference population (Tanner 1963, Cameron 1992, Adair 1999); differences in height

between stunted and non-stunted children per population usually remain the same

until adulthood. Catch-up growth is largely dependent on maturational delay that

prolongs growth period; however, this delay is not common in low income settings

(Martorell 1994). In developing countries, reasons for intrauterine growth restriction

include poor maternal nutrition and infections; restricted foetal growth especially

during the third trimester affects crucial period of brain development (Singh 2005,

Grantham-McGregor et al. 2007). Likewise, postnatal growth faltering during the first

three years of life is often associated with poor maternal and childhood nutrition and

infections (Black et al. 2008).

23

2.4.2 Micronutrient causes and determinants of growth faltering

In recent years, the significance of micronutrients as a cause of undernutrition has

been appreciated more while well known causes of protein energy malnutrition

remain an important concern (Allen et al. 2006). Because of the high prevalence

worldwide, iron, vitamin A and iodine are considered to be micronutrients of high

public health importance; in Africa for example the prevalence of anaemia is

estimated at 46%, that of inadequate iodine intake at 43% and that of vitamin A

deficiency at 49%. Similar high prevalence of micronutrient deficiencies are estimated

to occur in South East Asia (Allen et al. 2006). Other micronutrients likely to

contribute substantially to global burden are zinc, folate and vitamin D (Allen et al.

2006). Essential fatty acids are yet another group of micronutrients of high public

health important. In recent years, substantial data on their role on neuro-tissue

development have been reported. Whatever the cause of undernutrition may be, its

consequences on growth and development of the foetus or infants are severe and long

lasting; thus, underscoring the need for public health interventions to fight it.

2.4.3 Consequences of undernutrition and growth faltering

Prenatal growth restriction has been associated with poor developments in early

childhood and later in life. Low birth weight term infants in Brazil showed lower

developmental levels at ages of 6 and12 months than appropriate birth weight infants

24

(Grantham- McGregor et al. 1998); in Guatemala they had lower cognitive

developments at 2 and 3 years (Villar et al. 1984, Gorman and Pollitt 1992); in

Jamaica they had poorer problem solving ability at 7 months and lower development

at 15 and 24 months than normal birth weight infants (Gardner et al. 2003, Walker et

al. 2004). Effect of intrauterine growth restriction has been reported to continue to

adulthood (Strauss 2000) although not in all studies (Li et al. 2004).

Early childhood stunting has been associated with several poor developmental

outcomes from both cross-sectional studies and longitudinal studies. In early

childhood, stunting is associated with poor child development, including poor sensori-

motor, cognitive-language and social-emotional development (Meeks Gardner et al.

1999, Cheung et al. 2001, Siegel et al. 2005, Olney et al. 2007). History of stunting

and poor development in early childhood are associated with poor school performance

(Clarke et al. 1990, Mendez et al. 1999, Daniels et al. 2004, Grantham-McGregor et

al. 2007, Victora et al. 2008); schooling achievements and educational performance

are generally poor among currently stunted children across different ethnic

background and from different outcome measures (Bogin and MacVean 1983,

Jamison 1986, Moock and Leslie 1986, Steegmann et al. 1992, Hutchinson et al.

1997, Brooker et al. 1999, Beasley et al. 2000, Sharif et al. 2000, Glewwe et al. 2001,

Partnership for Child development 2001, Özmert et al. 2005,). Stunted children were

less likely to be enrolled to school in Tanzania (Beasley et al. 2000); more likely to

enrol late (Moock and Leslie 1986, Brooker et al. 1999), to attain lower achievement

levels or grades for their age (Bogin and MacVean 1983, Jamison 1986, Moock and

Leslie 1986, Clarke et al. 1990, Hutchinson et al. 1997, Brooker et al. 1999, Sharif et

25

al. 2000, Glewwe et al. 2001, Partnership for Child development 2001, Özmert et al.

2005), and often have poorer cognitive ability or achievement scores (Johnston et

al.1987, Sigman et al. 1989, Ivanovic 2004, Webb et al. 2005). On the contrary, only

few studies have reported no significant association between stunting and poor school

performance (Glewwe 1995, Cueto 2005).

Subsequently in adolescent and young adulthood, childhood stunting has been

associated with reduced physical performance (Haas et al. 1995) and lower human

capital (Victora et al. 2008). Short mothers are more likely to have small children who

are likely to grow into short adults themselves (Ramakrishnan et al. 1999, Victora et

al. 2008) and low maternal birth weight was associated with lower birth weight in the

next generation (Victora et al. 2008)

Childhood undernutrition has also been associated with morbidity later in life, but this

is an area of ongoing research. Studies on association between undernutrition and

many metabolic diseases (Type 2 diabetes, hypertension and cardiovascular disease)

show contradictory results and not many have been done in low income settings

(Victora et al. 2008). Foetal undernutrition combined with later quick catch up in

infancy appears to be associated with later adult sequelae like glucose intolerance and

obesity (Cameron and Demerath 2002). On the contrary, stunting did not predict

higher fat deposition in adolescence (Cameron et al. 2005). Results from some studies

have shown association between low nutrient intake, mental illnesses and immune

26

response but not with osteoporosis or cancer. Lung capacity reduction has been seen

in low birth weight infants (Victora et al. 2008).

Undernutrition has also been associated with childhood mortality, lethargy, apathy,

clinical manifestation of micronutrient deficiency disorders and morbidity from

infectious diseases including at least diarrhoea, pneumonia, malaria and measles

(Caulfield et al. 2004a, b). These disorders may cause reduced nutrient intake,

completing a viscous cycle. What is more shocking is the synergistic effect of mild to

moderate malnutrition with the infectious diseases on childhood mortality (Pelletier et

al. 1993, Caulfield et al. 2004a, b). In the past few decades 35-50% of world wide

under-five year olds’ deaths have been estimated to be directly or indirectly caused by

undernutrition (Pelletier et al. 1994, Black et al. 2008). Such serious mortality

consequences have been associated with at least stunting, wasting and underweight

(Pelletier et al. 1994).

2.5 Epidemiological strategies for promotion of growth and development

Several strategies have been used to promote child growth and development and to

prevent childhood undernutrition at different levels which have focused at birth,

infants or childhood outcomes. Their target has been to improve the quality and the

quantity of general nutrient intake or to reduce morbidity that interacts with nutrition.

Some of the notable interventions include health education, micronutrient

27

supplementation (iron-folate, multiple micronutrient, calcium and zinc vitamin A),

food fortification, supplementary and complementary feeding, infectious disease

control (deworming, intermittent treatment of malaria, insecticide-treated bed nets),

hand washing or hygiene interventions, promotion of breastfeeding, delayed cord

clamping and cash transfer.

2.5.1 Promotion of general nutrient intake to mothers

Maternal supplementation schemes may be aimed at either building nutrient reserves

in preparation for conception or promoting adequate placental transfer of nutrients to

the foetus during pregnancy or adequate nutrient transfer through breast milk during

lactation period. With necessary improvements, maternal supplementation may be one

of the most cost effective methods of tackling undernutrition. Interventions aimed at

improving general nutrient intake through maternal intervention so far have largely

been found to be effective in promotion of growth and prevention of undernutrition

(Bhutta et al. 2008). Maternal supplementary interventions thus far, have focused on

provision of micronutrients and general nutrient intake. Notable micronutrients that

have been provided to women in low income settings include folate, iron, zinc, iodine,

vitamin A and essential fatty acids; details of outcomes of supplementation with these

micronutrients will be discussed separately later on. Energy and protein

supplementary interventions in pregnancy have been provided together or

independently.

28

In some interventions energy and proteins were balanced and in others protein was

high. Prenatal provision of balanced energy and protein intake to mothers were found

to reduce small-for-gestation age babies (considered a proxy for intrauterine growth

restriction) and to modestly increase birth weight (Kramer 1993, Kramer and Kakuma

2003, Bhutta et al. 2008). Although supplementation in pregnancy was observed to

increase birth weight significantly in Gambia, Guatemala and Taiwan (McDonald et

al. 1981, Ceesay et al. 1997, Winkvist et al. 1998), other studies else where have

failed to show such clearly positive results (Adair and Pollitt 1985, Kramer and

Kakuma 2003). Despite small weight difference at birth, later on, especially within in

the first 60 postnatal months, energy supplementation during pregnancy was

associated with increasingly higher height and weight when compared to non-

supplementation (Kusin and Kardjat 1992). This observation may signify the effect of

a phenomenon called programming (Lucas et al. 1999). Similarly, postnatal

supplementation with energy to lactating mother has been shown to have positive

effects. It increases breast milk output volume and energy density, especially when

supplementation was done to undernourished mothers (Gonzalez-Cossio et al. 1998).

On the contrary, provision of protein alone or iso-caloric protein failed to show any

positive foetal or birth outcomes. Actually, in some cases data from intervention with

high protein supplements were suggestive of harmful birth outcomes (Kramer and

Kakuma 2003). These results were suggestive that provision of protein alone, without

energy in pregnancy may not have health benefit measurable at birth.

2.5.2 Promotion of general nutrient intake to infants

29

Intervention promoting nutrient intake in infancy aim at promoting exclusive

breastfeeding in the first 6 postnatal months which is the ideal food for healthy growth

and development at that age. Thereafter, they aim at promoting interventions that

promote timely introduction and provision of adequate complementary foods, in

conjunction with continued breastfeeding up to 2 years of age or more (Fifty-fourth

World Health Assembly 2001, WHO 2003). In undernourished children,

supplementary interventions aim at treatment, replacement of lost nutrients and

promotion of catch-up growth. Thus, they need to provide for even higher nutrient

intake.

2.5.2.1 Promotion of appropriate breastfeeding

In the first 6 months of life, the benefits of exclusive breastfeeding on prevention of

childhood morbidity and mortality have been well documented (WHO 2000, Arifeen

et al. 2001), thus justifying promotion exclusive breastfeeding up to 6 months of age

other than 4 months (Fifty-fourth World Health Assembly 2001). Results from

breastfeeding promotion interventions have been encouraging, leading to higher

uptake and prevalence, and longer duration of exclusive breastfeeding. These

interventions have been associated with reduction in morbidity and mortality in

infants and children (WHO 2000, Kramer and Kakuma 2002, Jones et al. 2003, Gau

2004, Britton et al. 2007, Bhutta et al. 2008).

Although 40 % of births in the world still occur at home, the large scale Baby

Friendly Hospital Initiative introduced by WHO and UNICEF in 1991 has been

30

effective in both low and high income countries; the initiative has been associated

with improved uptake of breastfeeding by mothers, with increased consumption of

breast milk by infants and with reduction in prevalence of childhood symptoms which

causes hospital consultation (UNICEF 2007, Walker 2007, Cardoso et al. 2008).

Although breast feeding has not been shown to improve weight or length gain more

than non breastfeeding infants, exclusively breastfed children have been shown to

weigh more than predominantly breastfed infants (Victora et al. 1998). Actually,

introduction of complementary feeding before 6 months in low income settings is

associated with stunting (Adair and Guilkey 1997). In long term, breastfeeding has

also been associated with significantly higher scores for cognitive development

(Fewtrell 2007).

2.5.2.2 Promotion of complementary feeding

Supplementary feeding and nutritional education interventions provided together or

separately to promote complementary feeding between 6 and 24 months are

suggestive of modest effective promotion of growth across different age and ethnic

groups, and different food security and socioeconomic backgrounds (Bhutta et al.

2008, Dewey and Adu-Afarwuah 2008). Higher growth impact, however, occurs in

food insecure regions (Dewey and Adu-Afarwuah 2008). Most intervention assessing

growth promotion efforts used height/length and weight gain as their outcomes, but a

few of them also evaluated morbidity and psychomotor development. In some of the

studies the largest impact on growth was seen in the younger populations (Schroeder

et al. 2002, Rivera et al. 2004). Because the improved growth observed in the

intervention groups were often modest, the growth faltering commonly seen in young

31

children between 6 and 24 months in developing countries is still persistent even with

these interventions (Dewey 2001, Shrimpton et al. 2001, Dewey and Adu-Afarwuah

2008). Thus, the impact of these interventions has mostly been to slow the growth

faltering but not necessarily stopping it. However, growth and developmental effects

from supplementary interventions have been found to be sustainable in many studies

and emerging data on long-term follow-up studies suggest that these interventions

also promote school performance, individual productivity and economic performance

in adulthood (Martorell 1995a, Li et al. 2003a, Stein et al. 2003, Hoddinott et al.

2008).

Caloric supplementary foods were mostly based on cereal and/or legumes fortified

with micronutrient, although fortified milk and lipid-based nutrient supplements

(LNS) were also used (Dewey and Adu-Afarwuah 2008). For many years there have

been concerns of suitability of energy and nutrient density in cereal/legume blend

used for the purpose of complementary feeding of 6-24 months old infants (Beaton

and Ghassemi 1982, Lutter 2000). Increased energy and nutrient density in porridge

made from cereal/legume flour can be achieved by increasing the thickness of gruel or

by adding oil. However, thick porridge is considered difficult to feed to infants by

most of the mothers. Cheap methods to increase energy density of porridge have not

been conclusive either; mixed results were found from 5 studies with efforts to

increase the energy density by adding amylase (Dewey and Adu-Afarwuah 2008). On

the other hand, increasing nutrient density was associated with reduced breast milk

intake (Islam et al. 2006); and edible oils are often too expensive to many poor

families in low income countries, thus unlikely to be effective without any

32

supplementation programmes. In short augmenting nutrient through gruel or porridge

intake without advanced technologies is rather difficult.

Except in milk based products or lipid-based nutrient supplements most of the other

complementary food provide plant protein from cereal and legumes, especially corn

and soy, which are considered to be of lower quality than animal protein. Historically,

cereal/soy blends were developed by the U.S. Department of Agriculture (USDA),

U.S. Agency for International Development (USAID) and National Institutes of

Health and formed the basis of food aid commodities. When introduced by USDA in

1966, they contained non-fat dry milk. As milk availability became difficult in the late

1980s, dairy was removed and thus all protein was derived from plant-based sources

(Marchione 2002). Despite being the provider of best legume protein and perhaps

appropriate for other age groups, unrefined proteins from soy are difficult for infants

and young to digest and contain anti-nutrient factors like phytic acid that limit

absorption of mineral micronutrients (Graham et al. 1971, Messina 1999, Hoppe et al.

2008). For many years and using different methods of assessment, including the

protein digestibility-corrected amino acid score (PDCAAS) which has been recently

adopted by WHO and FAO, unrefined corn and soy protein have been shown to be

poorer than that of milk as a source of amino acids (Graham et al. 1971, WHO 2007,

Hoppe et al. 2008).

Unlike growth, morbidity results were not universal in complementary interventions.

Results on morbidity are mixed with a few reporting reduced morbidity, others

33

reporting no effect and yet others reporting increased morbidity after the intervention

(Dewey and Adu-Afarwuah 2008). Results on developmental outcomes were not

conclusive. A few studies reported positive motor development while other showed

no difference (Dewey and Adu-Afarwuah 2008).

In recent years, community-based management of severe acute malnutrition with

lipid-based nutrient supplement (LNS) called ready-to-use therapeutic food (RUTF)

has been shown to be more effective than traditional milk-based management often

offered centrally through nutrition rehabilitation units (NRU). Compared to NRU

management, the community-based management has been effective in reducing case

fatality rates from malnutrition, promoting recovery from undernutrition, promoting

weight gain, reducing recurrences and inducing more compliance and acceptability

among caregiver and children (Manary et al. 2004, Sandige et al. 2004, Ciliberto et al.

2005, Ndekha et al. 2005, Collins et al. 2006, Bhutta et al. 2008, Dewey and Adu-

Afarwuah 2008).

These relatively new products typically contain milk protein, sugar, and a mixture of

micronutrients, embedded in a lipid base, need no cooking before use, and can be

stored for months even in warm conditions (Briend et al. 1999, Briend 2001). Because

they are lipid based, LNS can mask metallic taste of high mineral concentrations; it

can be used to deliver micronutrient daily dose ranging from 10 to 100 g. The energy

content of LNS is up-to 22 kJ/g (5.4 kcal/g); thus an intake of up to100 g/d can deliver

2200 kJ/d (540 kcal/d). With a dose of 20g/d which is estimated to cost about $0.06,

34

110 kcal can be consumed (Nestel et al. 2003). It is noteworthy that 110 kcal covers

for greater than 10% of daily total energy intake of a child aged 1-2 years.

The LNS used in therapeutic feeding is well accepted by undernourished children,

suggesting that similar product could be used as a complementary food supplement

(CFS) for prevention purposes and promotion of growth (Nestel et al. 2003). In

Malawi, data from dose finding trial suggested that home provision of LNS to

moderately underweight 6-17 months old infants improve their nutritional status and

results in markedly increased linear growth and improved weight gain (Kuusipalo

2004). In Ghana, these products were shown to promote growth and motor

development of infants (Adu-Afarwuah et al. 2007).

2.5.3 Promoting micronutrient intake to mothers and infants

There is growing evidence suggesting the nutritional importance of micronutrients.

Low nutrient intakes among infant in low income settings are attributable to low

intake and low micronutrient density of complementary foods (Kimmons et al. 2005).

Thus, some authors have concluded that dietary quality rather than quantity is the key

aspect of complementary food diets that needs to be improved (Lutter and Rivera

2003, Allen 2008). One way to improve quality of complementary food is by

provision of vitamin and mineral supplements.

35

Several micronutrient interventions have been implemented either as a single or a

multiple micronutrient supplementary intervention or as a single or a multiple food

fortification intervention. Because of cost effectiveness as evidenced by lower cost

per each disability-adjusted life year (DALY) saved and biological synergism seen

with some micronutrient combinations, the multiple micronutrient interventions

would be preferable to the single micronutrient interventions in low income settings

(Mejia and Arroyave 1982, Homedes 1996, Zimmermann et al. 2004, Horton 2006).

Likewise, a centralised fortification would be preferred to the supplementary

intervention or community-level fortification interventions (Baltussen et al. 2004,

Horton 2006). However, caution should be exercised where the food fortification with

a specific micronutrient may pose a health risk by supplementing to subgroups of the

population at risk of toxicity or negative interactions (Crane et al. 1995, Smith et al.

2008).

The supplementary interventions, however, may still be preferred where

supplementation only targets specific sub-groups of the population, like pregnant

women (Werler et al. 1999). Thus far, micronutrient fortification or supplementation

programs for folate, iron, zinc, iodine and vitamin A have been widely promoted in

low income settings (The World Health Report 2002, Allen et al. 2006); impact from

interventions with these micronutrients and those from vitamin D and essential fatty

acids intervention are outlined below.

36

2.5.3.1 Folate

Supplementary folate during peri-conception period has been shown to protect against

foetal neural tube defects (NTD) (Rush 1995). Even when food fortification is in

place supplementation has been recommended to attain adequate daily intake to

prevent NTD (Werler et al. 1999). Folate food fortification has been effective without

increasing neurologic damage attributed to vitamin B-12 deficiency masking (Mills et

al. 2003). In infants and children, however, folate deficiency is rare and breast milk

usually provide adequate intake (Allen 2003).

2.5.3.2 Iron

Iron supplementation and food fortification have been shown to increase haemoglobin

concentration in child bearing and pregnant women (Zimmermann et al. 2004, Van

Thuy et al. 2005, Bhutta et al. 2008). In childhood, iron supplementation and

fortification of complementary food have been associated with haematological,

growth and lower morbidity outcomes. Haematological outcomes include reducing

prevalence of anaemia, increasing haemoglobin concentration and ferritin

concentration in children (Baltussen et al. 2004, Dewey 2007, Bhutta et al. 2008).

Lower episodes of diarrhoea and ARI have been reported when iron supplementation

has been combined with milk and zinc supplementation (Bhutta et al. 2008). Evidence

also shows that iron promotes better cognitive, motor, social-emotional development

in supplemented than non-supplemented infants (Lazoff 2007).

37

2.5.3.3 Zinc

Supplementation with zinc has been shown to be effective in increasing zinc intake in

children (Sazawal et al. 1996). Among infants and children, several studies have

shown that zinc supplementation reduces occurrence of several infections, including

diarrhoea, dysentery, acute lower tract infection, and pneumonia (Sazawal et al. 1996,

Sazawal et al. 1998, Bhutta 1999). On the other hand impact of zinc fortification of

complementary food has not been encouraging. Provision of complementary food

supplements fortified with multiple micronutrients, including zinc, however, showed

little impact on plasma concentration (Dewey and Adu-Afarwuah 2008), suggesting

poor absorption.

2.5.3.4 Iodine

Centralised salt iodisation has been shown to be effective in increasing the proportion

of houses using iodised salt which has been shown to be associated with increased

consumption of iodine in women and children (Melse-Boonstra et al. 1998, Zhao and

van der Haar 2004). Health benefits of consuming iodised salt include reduction in

occurrence of goitre (Bhutta et al. 2008), increased birth weight (Manson et al. 2002)

and increased weight-for-age and length-for-age in children (Manson et al. 2002).

2.5.3.5 Vitamin D

Because vitamin D deficiency is mainly a problem above and below latitudes 40°N

and 40°S, where the intensity of ultraviolet radiation from sunlight, Vitamin D

38

deficiency has been considered a problem of industrialised countries, yet insufficient

dietary intake may be as common in low income countries (Zeghoud et al. 1994,

Thacher et al. 1999, Allen 2006). Supplementary and fortification interventions to

lactating mothers or infant and children have been shown to be effective in promoting

vitamin D intake in children (Allen 2006, Wagner et al. 2006). However, in low

income countries children with clinical features of vitamin D deficiency responded

more to calcium supplementation or combined calcium and vitamin D when compared

to those supplemented with vitamin D alone (Thacher et al. 1999).

2.5.3.6 Vitamin A

Centralized food fortification with vitamin A has been shown to contribute

significantly towards intake of the vitamin in the past 3 decades (Guillermo et al.

1981, Mejia and Arroyave 1982, Krause et al. 1998, Dewey and Adu-Afarwuah

2008). In a review of effect of vitamin A fortified complementary supplements, more

studies showed significantly higher retinol plasma concentration in infants from the

intervention groups than those from the control groups; however, in those that showed

no difference, investigators attributed the observation to wide spread vitamin

supplementation programmes (Dewey and Adu-Afarwuah 2008). After both

supplementary and fortification interventions, vitamin A has been shown to promote

linear growth, reduce morbidity and mortality, reduce the occurrence of Bitot’s spots,

diarrhoea, respiratory tract infections, measles and their associated mortality (Mejia

and Arroyave 1982, Muhilal et al. 1988, Pant et al. 1996, Underwood and Arthur

1996, Krause et al. 1998). Impact of supplementation with other micronutrients like

iron has been shown to improve when provided together with vitamin A, suggesting

39

synergistic properties of vitamin A to other micronutrient (Mejia and Guillermo

1982).

2.5.3.7 Essential fatty acids

There has been an increasing evidence indicating that polyunsaturated acids, the

essential fatty acids linoleic acid (LA, omega-6) and alpha-linolenic acid (ALA,

omega-3), are critical for development and maintenance of structural and functional

integrity of the central nervous system and the retina (Uauy et al. 2001, Singh 2005,

Innis 2007). These essential fatty acids LA and ALA are precursors of longchain

polyunsaturated acids (LCPUFA), including docosahexaenoic acid (DHA, 22:6 n-3);

the LCPUFA are one of building elements for membrane, and nervous tissue (Uauy et

al. 2001). Supplementary intervention with PUFA or LCPUFA to pregnant or

lactating mothers and directly to infants has been shown to increase intake of essential

fatty acids (Uauy et al. 2001, Helland et al. 2003, Helland et al. 2006). The impact of

the maternal supplementation with essential fatty acids has also been shown to

promote growth and development by increasing physical growth (Lauritzen et al.

2005a), visual acuity (Lauritzen et al. 2004b) and intelligence quotient (Helland et al.

2004).

2.5.3.8 Multiple micronutrient supplements

Sprinkles, a vitamin and mineral mix packaged in small sachets containing a daily

dose of micronutrients designed to mix with food, have also been tested. Sprinkles

40

containing iron, vitamin A, zinc and vitamin C have been shown to be effective in

treating anaemia and increasing haemoglobin concentration in Ghana (Zlotkin et al.

2003, Adu-Afarwuah 2007).

2.5.4 Prevention and treatment of infections associated with undernutrition

As significant contributors of undernutrition and growth faltering, infections have

been targeted as one of the strategies to promote growth. Malaria, diarrhoea, intestinal

helminth, and respiratory tract infections which affect a large proportion of under-

fives (Snow et al. 2005, Walker 2007) have been major focus for intervention to

promote growth in addition to vaccine preventable diseases.

2.5.4.1 Anti malarial interventions

A considerable amount of effort has been devoted anti-malarial interventions in the

past few decades. Anti-malarial interventions using insecticide treated bed net (ITN)

and regular intermittent presumptive treatment have been shown to be effective in

reducing maternal anaemia and increasing the birth weight; in children these

interventions have been associated with episodes of clinical malaria, less severe

anaemia and fewer hospital admissions (Lengeler 2004, Garner and Gülmezoglu

2006, Gamble et al. 2007, Msyamboza et al. 2007, Bhutta 2008, Meremikwu et al.

2008).

41

2.5.4.2 Deworming interventions

There have been benefits after treating pregnant women, school going children and

preschool children albeit lower worm loads occur in preschool children (Stoltzfus et

al. 2004). Deworming interventions in pregnancy have been shown to be effective in

reducing fall in haemoglobin concentration between first and third trimester and in

increasing birth weight (Torlesse and Hodges 2001, Bhutta et al. 2008). Although not

an intervention common in the first 12 months, deworming among preschool and

school children has been shown to be practical and to promote weight and length gain,

and reduce prevalence of anaemia (Stoltzfus et al. 2004, Hall et al. 2008, Alderman et

al. 2006, Kirwan et al. 2009, Taylor-Robinson et al. 2009).

2.5.4.3 Hygiene interventions

Hygiene interventions have been shown to be effective in reducing episodes of

diarrhoea. These include handwashing, water quality treatment, sanitation, and health

education (Fewtrell 2005). Improved sanitation resulted in less episode of diarrhoea in

Lesotho (Daniels et al. 1991)

2.5.4.4 Expanded programme on immunization

Expanded programme on immunization (EPI) introduced by WHO in 1974 has

overtime increasingly reported higher coverage, lower morbidity and lower mortality

from vaccine preventable diseases (Kim-Farley 1992, Cutts 1998, Arevshatian et al.

2007). High vaccination coverage for measles in low income countries have been

42

attained in low income countries resulting in reduced morbidity and mortality,

although sporadic epidemics have been reported lately (Arevshatian et al. 2007).

2.5.5 General interventions

Although rather different from supplementation and fortification interventions,

general approaches aiming at promoting dietary diversification, socioeconomic status

and female education are also promising.

2.5.5.1 Dietary diversification

To enhance dietary diversification, agriculture and small-animal production have been

promoted and, so far, these interventions have been associated with increase in

consumption of animal and economic status of families (Allen 2003b, Bhutta 2008).

2.5.5.2 Caretaker literacy and socioeconomic status

Intervention studies designed to assess causality effect of literacy and socioeconomic

status on growth promotion or prevention are rare. Evidence of association between

literacy and socioeconomic status mostly come from long-term observations and case

control studies. However, higher socioeconomic status and caretaker education

(especially maternal education) assessed differently and across different backgrounds

43

have consistently been associated with better health status and growth in childhood

(Fotso and Kaute-Defo 2005, Larrea and Kawachi 2005, Hatt and Water 2006).

Improvements in education, purchasing power and access to healthcare facilities

overtime have been associated with reduction in undernutrition in Brazil (Monteiro et

al. 2009). In rural China rapidly increasing village and household income was

associated with increased sanitation, hygiene behaviours and reduced undernutrition

(Cangjiang et al. 2009).

2.5.5.3 Health education and behaviour change

Often, nutritional interventions without appropriate behavioural change will have

somewhat small effect. To achieve substantial nutritional outcomes, households may

have to make substantial changes in their food practices during pregnancy, lactation

and weaning periods. Nutritional education to communities and healthcare workers

has been shown to reduce underweight (Zaman et al. 2008, Horton 2008, Roy et al.

2008, Monteiro et al. 2009); in pregnant women nutritional counselling has been

associated with an increase in haemoglobin concentration and a reduced prevalence of

anaemia (Garg and Kashyap 2006).

2.5.6 Concerns of growth promoting strategies

Long term implications of early childhood growth and somewhat negative outcomes

of some of the nutrition interventions have raised health concerns. Notably concerns

have been raised on breast feeding in HIV infected mother, later age metabolic

diseases after supplementary feeding, iron and folate supplementation in malaria

44

endemic areas, and vitamin A supplementation in pregnancy. Outlined below is a

discourse on these concerns.

2.5.6.1 Mother to child HIV transmission

Provision of breastfeeding counselling to HIV positive women is challenging for

healthcare workers with background knowledge of mother-to-child HIV transmission

through breast milk, especially with some studies suggesting better child survival

when infants are fed breast milk substitutes (Mulder 1996, Miotti et al. 1999, Nduati

et al. 2000). To the mothers in developing countries HIV brings the difficult choice of

balancing between the risk of transmitting HIV to their child through breastfeeding

and provision of substantial benefits of breastfeeding. However, overtime evidence

weigh more towards child benefits of breastfeeding against the risk of HIV

transmission in low income countries where women are left with few options. Most of

the trials assessing the benefit of breast feeding substitute provide no evidence of

higher negative outcomes among HIV exposed infants who were breastfed to formula

fed infants by 7 months of age (Iliff et al. 2005, Thior et al. 2006, Coovadia et al.

2007, Bhutta et al. 2008). Actually, formula feeding was associated with higher

mortality in Uganda (Kagaayi et al. 2008).

2.5.6.2 Metabolic syndrome

The hypothesis that intrauterine nutrition and growth can result in foetal adaptation

that programme future weight trajectory and project lifetime risk of metabolic

45

syndrome consequences has been extensively reviewed. Although several studies link

birth weight to adult cardiovascular conditions, postnatal upward shift of weight

centiles is considered more associated with risk factors of metabolic syndrome (Lucas

et al. 1999, Bhargava et al. 2004, Demerath et al. 2007, Victora et al. 2008). Because

of this concern, effect of growth promoting interventions on blood lipid, insulin

resistance and type 2 diabetes, blood pressure and cardiovascular diseases in adult life

are evaluated below. However, very few studies have reported metabolic syndromes

as an outcome of early childhood growth promoting intervention in low income

countries. While low birth weight is associated with high serum cholesterol values in

adults from developed countries, similar associations were not observed in low

income countries (Levitt 2000, Stein 2002, Victora et al. 2008). Supplementation in

pregnancy and infancy was also not associated with high serum cholesterol levels

(Stein et al. 2006).

Insulin resistance and secretion failure are risk factors of type 2 diabetes. Although

catch-up growth in infancy and childhood is associated with later insulin resistance

(Bhargava et al. 2004, Ong 2004), supplementation to both mothers and infants, and

long period of breastfeeding were associated with low fasting plasma glucose in

adulthood (Conlisk et al. 2004, Stuebe et al. 2005, Stein et al. 2006). These

observations suggest that nutritional interventions promote insulin tolerance and

prevent type 2 diabetes. Pooled analysis from low income countries adjusted for adult

covariates birth weight, weight for age and body mass index (BMI) at 2 years was

inversely associated with adult blood glucose concentration (Victora et al. 2008).

46

Some authors have reported association of rapid growth in infancy with increase in

adult blood pressure or hypertension although weight and height may relate inversely

with blood pressure (Cheung et al. 2000, Huxley et al. 2000). Results from low

income countries showed that weight and height at 2 years were associated with high

blood pressure. However, after adjustment for adult covariates inverse associated was

found in some cases (Victora et al. 2008).

In low income countries, obesity (high proportion of fat in the body) is not common.

However, because of its relation to other specific metabolic syndrome diseases it is a

concern in growth promotion interventions. Birth weight has a bimodal relationship

with obesity where higher adult fat proportion is associated with both the low and the

high birth weight extremities. Although height, weight and BMI at 2 years were

associated with adult BMI, independently linear growth in infancy, body mass index

and height at 2 years are associated adult lean mass (Li et al. 2003b, Victora et al.

2008).

Data on cardiovascular disease from low income show that birth weight is inversely

associated with prevalence of cardiovascular diseases, albeit few data are available

(Stein et al. 1996). Shorter child height and weight are associated with increased risk

of cardiovascular disease (Victora et al. 2008).

47

2.5.6.3 Iron and folate supplementation in malaria endemic areas

Because folate is essential for DNA synthesis and growth of malaria parasite, the

parasites acquire folate by de novo synthesis or salvaging preformed folate. Antifolate

antimalarial drugs work by inhibiting both the synthesis and the salvaging of the

folate from the host (Metz 2007). Hence, folic acid supplements given to pregnant

women and children may increase malaria parasites proliferation and may reduce

some anti-malarial drug’s effects (Metz 2007). Evidence from Africa suggest that

universal supplementation with folate in these areas increases the risk of mortality, the

severity of morbidity and treatment failure. (van Hensbroek et al. 1995, Dzinjalamala

et al. 2005, Carter et al. 2005, Mulenga et al. 2006, Sazawal et al. 2006). However,

targeted folic and iron supplementation combined with treatment of malaria and other

infections has been shown to be beneficial to children (Sazawal et al. 2006). There is

no evidence suggesting that supplementation in pregnancy may be detrimental either

to the mother or to the foetus (Metz 2007). Another concern of universal

supplementation with folate is masking of Vitamin B 12 deficiency, albeit rare (Smith

et al. 2008).

Evidence on detrimental effect of iron supplementation to young children where

malaria is endemic is rather contradictory (Stoltzfus et al. 2007). Some studies and

reviews reported an increase in the incidence of clinical malaria, increase in the

incidence of other infectious diseases and increase in mortality after universal

supplementation (Oppenheimer et al. 2001, Iannotti et al. 2006, Sazawal et al. 2006,

Prentice et al. 2007). On the other hand some studies reported no apparent harmful

effect or serious adverse event (Oppenheimer 2001, Gera and Sachdev 2002, Verhoef

48

et al. 2002, Desai et al. 2003, Iannotti et al. 2006, Sazawal et al. 2006, Lynch et al.

2007, Prentice et al. 2007, Stoltzfus et al. 2007). Like with folate supplementation,

subgroups venerable to anaemia and infections seem to benefit more from

supplementation with iron (Sazawal et al. 2006).

2.5.6.4 Teratogenicity of vitamin A

Because of teratogenic effects observed in animal studies, those observed after

introduction of oral vitamin A analogue isotretinoin for treatment of severe acne in

1983, and some suspicious birth defects reported after high dose exposure in

pregnancy, long term high dose vitamin A is better avoided (Rose et al. 1986, Kizer et

al. 1990, Rothman et al. 1995). However, low dose vitamin A supplementation may

be tolerable even in pregnancy without a risk of teratogenicity and it has been shown

to reduce the incidence of maternal mortality by 40% (West et al. 1999, Allen and

Haskell 2002, Ross 2002).

2.6 Justification for the present study

Human growth and development is a complex multi-dimensional phenomenon.

Hence, there are many causes of growth failure which can be categorised to 2 origins-

- genetic and environmental. Although the larger part of human growth seems to be

determined genetically, environmental factors, especially effective nutrition, are vital

for uninterrupted growth. Despite numerous growth promotion efforts, impact on

49

growth has only been modest thus far, suggesting adaptive genetic down regulation

growth to promote survival in low resources setting. Impact of down regulation of

growth seems to be transferable across several generations, further suggesting that

unwinding this growth faltering process may require interventions that go on to

several generations. However, results of interventions other than growth, especially on

developments, have been encouraging and they may have the effect of accelerating

the attainment of full growth potential in even fewer generations to come.

From the complexity of interacting determinant of growth no single approach seems

feasible to address growth faltering and its associated consequences. However,

adequately addressing complementary feeding for 6 to 24 months infants and nutrition

of pregnant and lactating women are some of the possibly effective ways. The

enormous amount of emerging knowledge on significance of micronutrients in

determining growth, complemented by recent reaffirmation on the need for animal

proteins and adequate energy is vital for development of new dietary strategies for

promotion of effective complementary feeding.

Approaches combining complementary efforts and rigorous morbidity management

seemed to have produced slightly better outcomes, suggesting the significance of

including management of morbidity in new interventions. Social factors, including

background economic status of the families, education levels of the caretakers and

indirect costs like time for food preparation also emerged important. However,

50

interventions for these are not limited to public health alone rather epidemiologists

need to collaborate with other relevant sectors for such wide approaches.

In the medical principal of “doing no harm”, safety of supplementary effort is a

limiting element to blanket strategies. Cheaper options like centralised fortification of

processed food may not be appropriate as some micronutrients may promote

morbidity to some selected sub-groups. Efficacy of centralised fortification is also

questionable among 6 to 24 months where wide range of complementary food intake

is likely to result in inadequate intake among those consuming small amounts. Again

it may be ethically questionable if such important growth promoting nutrients should

be left to chance purchase in populations living in dire poverty. However, the

potential negative effects of these do not outweigh the potential positive outcomes of

growth promoting efforts, rather, demands careful and evidence based interventions.

While the literature review also emphasises the importance of the age between 6 to 24

months as the window of opportunity to promote growth and to prevent development

of undernutrition or mitigate its effects, the findings also emphasises on the

importance of intervening early within this window. Other than through women’s

nutrition or sanitation and through infection control interventions, the most feasible

intervention to the infants themselves is early complementary feeding from the age of

6 months when breast milk is still very important part of the diet. Therefore, it

remains a public health crucial challenge to identify an effective complementary diet

that provides good quality protein, adequate energy and micronutrients with minimal

51

risk of displacing breast milk or introducing infections. From the available and

identified literature, lipid-based nutrient supplements have most of these properties. It

was therefore with this back ground that the present study and aims were formulated.

52

3 AIMS

The aim of the present study was to assess the effect of providing lipid-based

nutritional supplements (LNS) or corn-soy blend on infant’s or childhood growth and

development promotion, and on primary and secondary prevention of undernutrition.

The specific objectives were:

1. To compare growth and recovery from undernutrition among moderately

underweight children receiving micronutrient-fortified corn-soy blend (Likuni

Phala [LP]) or lipid-based nutrient supplements (LNS) for 12 weeks.

2. To compare growth and incidence of undernutrition in infants receiving long-

term dietary supplementation with micronutrient-fortified corn-soy blend

(Likuni phala [LP]) or lipid-based nutrient supplements (LNS).

3. To study term growth effect of 12-months supplementation with lipid-based

nutrient supplements (LNS) over a subsequent 2-year non intervention period

53

4. To compare the effect of a 12-month long complementary food

supplementation with lipid-based nutrient supplements (LNS) or corn-soy

blend on early childhood development at the age of 18 months

54

4 METHODS

4.1 Approach to the study

The aim of the study was addressed through two community-based randomised

intervention trials conducted in 6 to 18 month old infants. Although the two studies

were largely similar, to address specific aims for each study the enrolment criterion

and the total time of supplementation were different (Table 1).

55

Table 1: Comparison of the two trials conducted for the study

Criteria Supplementary feeding Complementary feeding

Location Lungwena Lungwena

Enrolment age 6-15 months 6 months

Trial groups 2 3

Intervention group 1 LNS 50 g/d LNS 50 g/d

Intervention group 2 None LNS 25 g/d

Control group LP 71 g/d LP 71g/ d

Test dose LNS 6 g LNS 6 g

Primary follow-up 12 weeks 12 months

Secondary follow-up None 24 months

Enrolment nutritional status WAZ < -2.0 WAZ - 2.0

Outcome 1 Anthropometry Anthropometry

Outcome 2 Haematology Haematology

Outcome 3 None Development outcomes

Data collectors Trained Trained

Trial registration NCT00131222 NCT00131209

The supplementary feeding trial was aimed at promoting recovery form moderate

underweight; its outcomes were mainly anthropometry and haematological outcomes.

In the other trial which aimed at long-term promotion of complementary feeding in

addition to anthropometry and haematological outcomes, developmental outcomes

and long-term anthropometric effects were also assessed. Thus, the complementary

feeding trial had 3 different sub-studies (Figure 4).

56

Trial ISupplemetary Feeding

Randomised to recieve eitherLP (n=88) or FS50 (n=88)

Trial IIComplementary Feeding

Randomised to recieve either LP (n=61) or FS50 (n=61) or FS25 (n=60)

Ai

ms

Trial

To analyse growthand recovery of

moderatelyunderweight infants

To analyse growthand incidence ofundernutrition

To assessdevelopmental

outcomes

To assess post-intervention effects

on growth

Sub-study I

AnthropometryHaematology

Sub-study II

AnthropometryHaematology

Sub-study III

Motor and neuro-development

Sub-study IV

Anthropometry

Figure 4: Overall study design for the thesis

57

4.2 Study setting and subjects

4.2.1 Study Area

The study was conducted in Lungwena area and its neighbouring areas in Mangochi

district in the Southern part of Malawi, a low income country in Africa (Appendix

12.1). The area is located on the south eastern shore of Lake Malawi was

approximately 20km long and 5km wide. The people were mostly from the Yao tribe

and predominantly spoke Chiyao language. Staple food was maize that was grown

once a year in rainy season between December and April. Lungwena Health Centre

provided free curative and primary health services, including family planning,

antenatal and delivery services, growth monitoring, vaccinations, distribution of ITN

and treatment of diseases. Those requiring further management were referred to

nearby hospitals.

5 Study subjects

The study participants were eligible infants from the study area or it neighbouring

villages. In the complementary feeding trail infants turning 6 months were recruited to

participate in a 12-month intervention study. On the other hand, the recruitment age

58

range in the supplementary feeding trial was wider; moderately underweight infants

aged 6 and less than 15 months were enrolled to the 12 weeks intervention.

5.1 Nutritional interventions

Control groups in both trials received 71g corn–soy blend called Likuni Phala (LP) in

Malawi which was cooked into thin porridge. In both trials there was an intervention

group that received iso-energetic lipid-based nutrient supplement (FS50-50 g/d) to the

LP. The complementary feeding trial had a second intervention group that received

half the dose of lipid-based nutrient supplement (FS25-25 g/d). Likuni phala was

purchased from a local producer (Rab Processors, Limbe, Malawi). Lipid-based

nutrient supplement was produced at a Malawian non-governmental organization,

Project Peanut Butter (Blantyre, Malawi) or Nutriset (Malaunay, France), from peanut

paste, milk powder, vegetable oil, sugar, and premade micronutrient mixture

(Nutriset, Malaunay, France). All supplements were fortified with micronutrients, but

the level of fortification varied between the products (Table 2)

Table 2. Energy and nutrient content of a daily ration of each food supplement used in

this trial

Intervention group

Variable LP FS50 FS 25

Weight (g) 71 50 25

Energy (kcal) 282 256 127

59

Intervention group

Variable LP FS50 FS 25

Protein (g) 10.3 7.0 3.5

Carbohydrates (g) NA 13.8 6.6

Fat (g) 3.1 16.9 8.5

Retinol (µg RE) 138 400 400

Folate (µg) 43 160 160

Niacin (mg) 3 6 6

Pantothenic acid (mg) NA 2 2

Riboflavin (mg) 0.3 0.5 0.5

Thiamin (mg) 0.1 0.5 0.5

Vitamin B6 (mg) 0.3 0.5 0.5

Vitamin B12 (µg) 0.9 0.9 0.9

Vitamin C (mg) 48 30 30

Vitamin D (µg) NA 5 5

Calcium (mg) 71 366 283

Copper (mg) NA 0.4 0.4

Iodine (µg) NA 135 135

Iron (mg) 5 8 8

Magnesium (mg) NA 60 60

Selenium (µg) NA 17 17

Zinc (mg) 3.6 8.4 8.4

60

5.2 Data collection

5.2.1 Enrolment

For both trials, inclusion criteria included acceptable age, residence in the study area,

and an informed consent from at least one authorised guardian. To assess recovery

from moderate underweight only those with WAZ < -2.0 were enrolled to the

supplementary feeding trial. Exclusion criteria were presence of oedema, history of

peanut allergy, severe illness warranting hospitalisation on the enrolment day,

concurrent participation in another clinical trial, or any symptoms of food intolerance

within 30 minutes after the ingestion of a test dose of lipid-based nutrient supplement,

one of the food supplements used in the trial.

The enrolment procedures were similar in both trials. To identify potential

participants, trained health surveillance assistants contacted all families who were

known to live in the area and have a baby of approximately right age. After initial

screening process in the village, infants were invited to an enrolment session at the

health centre, where further screening for eligibility took place and guardians were

given detailed information on the trial. Before enrolment, a guardian signed a written

consent form for trial participation.

61

For group allocation, blocked randomisation was used in both studies. The guardians

to eligible infants picked one from a set of identically appearing opaque envelopes,

each containing a paper indicating an identification number and randomly assigned

allocation to one of the intervention groups. The randomization lists and envelopes

were made by people not involved in the trial implementations and the codes were not

disclosed to the researchers or those assessing the outcomes until all data had been

entered into a database.

5.2.2 Follow-up

The intervention and follow-up periods were different in the 2 trials and the

complementary feeding trial and a long-term postintervention follow-up. However, in

both studies participant were visited at home every week when data on use

supplements provided was collected together with data on morbidity and other

possible adverse events. At each and every other third visit in the supplementary and

complementary trials additional food supplements were delivered to the homes. As a

way of checking food use within the family, empty food container were collected

back to the health centre every week. At regular intervals (every 6 weeks and 4

months in the supplementary and the complementary feeding trials, respectively), the

participants underwent a physical examination and anthropometric assessment at the

health centre. Haematological tests were limited to enrolment and final health visits

only except those for malaria which were done whenever a participant presented with

fever at the health centre.

62

5.2.3 Anthropometric measurements

Anthropometric measurements collected in both trials were weight, length/height,

head circumference, MUAC (Appendix 12.2). Weight was measured from lightly

naked infants with electronic infant weighing scale (SECA 834, Chasmors Ltd, UK)

and recorded to the nearest 10g. Length (at 24 months of age) and height (at > 24

months) were measured to the nearest 1 mm with a high quality length board

(Kiddimetre, Raven Equipment Ltd, Essex, UK) and a high quality a stadiometer

(Harpenden stadiometer; Child Growth Foundation, London, United Kingdom). Mid-

upper-arm and head circumferences were measured with nonstretchable plastic tapes

(Lasso-o tape, Harlow Printing Limited, South Shields, United Kingdom).

Anthropometric indices (WAZ, LAZ, and WLZ) were calculated with Epi-Info 3.3.2

software (Center for Disease Control and Prevention, Atlanta, GA), based on the CDC

2000 growth reference (Kuczmarski et al. 2000).

5.2.4 Haematological tests

A 2 ml venous blood sample was drawn and serum separated by centrifugation at the

beginning and end of follow-up. Haemoglobin concentration was measured from a

fresh blood drop with Hemo-Cue cuevettes and reader (HemoCue AB, Angelholm,

Sweden). In the complentary feeding trial only, serum ferritin concentration was

analysed from frozen sera with commercial test kits according to the manufacturer’s

instructions (Ramco Laboratories, Stafford, Texas, USA).

63

Thick and thin blood smears were made, stained with Giemsa, and screened

microscopically for malaria parasites from symptomatic participants at enrolment.

Screening for human immunodeficiency virus infection was done from fresh blood

with two antibody rapid tests and positive results were confirmed with DNA

amplification technology, according to the manufacturers’ instructions (Determine,

Abbot Laboratories, Abbot Park, USA, Uni-Gold, Trinity Biotech plc, Bray, Ireland,

Amplicor HIV-1 Monitor Test Version 1.5, Hoffmann-La Roche, Ltd, Basel,

Switzerland).

5.2.5 Developmental assessments

Development was assessed in the complementary feeding trial only. The assessments

was conducted using the Griffith’s developmental assessment tool for 0-2 year-olds

(Griffiths & Huntley 1996). This was adapted by removing some items considered

inappropriate in the rural Malawian setting. Local research assistants were trained and

certified in assessment and use of the Griffiths by a qualified trainer. The participant’s

raw score for the general scale and each subscale were expressed as the sum of all the

items scored (Cheung et al. 2008). The sub- and general raw scores were used to find

developmental quotients from the Griffith Manual (Griffith & Huntley 1996). The

mental age column in the Griffith Scale was used to read the mental age

corresponding to attained the subscale raw score, and general mental age was

calculated as the mean of all subscale mental ages.

64

5.2.6 Training and quality control

All data collection was implemented according to standard operating procedures

(SOP) developed before commencing the study. All research assistants for the trials

were trained on the use of these SOPs as well as the use of questionnaires and food

hygiene. The performance of field data collectors were daily monitored by an

experienced senior research assistant and the author.

All measurements were conducted by the author or one other colleague. Reliability on

anthropometric measurements was assessed for all measurements against an

experienced anthropometry measurement expert before the commencement of the

trial. Standardization measurements were collected from eight infants. Technical error

of measurement (TEM) for all people involved in collecting anthropometric data and

coefficient of reliability for length, MUAC and head circumference were calculated

from standardisation measurements collected from 8 infants not participating in the

study.

65

5.3 Statistical approach

5.3.1 Sample size calculation

Sample sizes were calculated from expected values for one of the primary outcome

(weight gain) and predicted equal standard deviation in comparison groups based on

an earlier preliminary dose-finding trial (Kuusipalo et al. 2006). The sample sizes

calculation was set to provide the trial with 80% power and 95% confidence. Finally,

the sample sizes were adjusted to allow for approximately 5 % attrition estimated

from previous studies (Kuusipalo et al. 2006).

5.3.2 Data management and analysis

Collected data were recorded on paper forms, transcribed to paper case report forms

(CRF) and double-entered into a tailor made Microsoft Access 2003 database. The

two entries were electronically compared and extreme or otherwise suspicious values

were confirmed or corrected.

Statistical analysis was carried out using Stata 9.0 (StataCorp, College Station, USA)

on intention-to-treat basis. Infants with no anthropometric data after enrolment were

66

only included in the comparison of baseline characteristics. The analyses on

anthropometric measures used data from young children who were available at

analyses time point, while infants and young children with at least one measurement

after enrolment were used to analyze incidence of stunting using survival analysis

method to deal with censoring.

For continuous and categorical outcomes, the intervention groups were compared with

analysis of variance (ANOVA) or t-test and Fisher’s exact test, respectively. Survival

analysis was used to determine cumulative probability of severe or moderate stunting

among different groups and the difference were tested by the log-rank test. An event

was considered to have happened at mid-point between the time the event was

detected and the previous measurement. Individuals with severe or moderate-to-

severe stunting already at enrolment were excluded from survival analyses concerning

the incidence of that outcome.

For comparison of continuous baseline and final anthropometric and haematological

group values paired t-test was used, where as categorical baseline and final variables

were compared using sign-test. For the studies of compliance using visits as units of

analysis, the Huber-White robust standard error was used to allow for correlated data

(multiple visits per child).

67

5.4 Ethical compliance

Both trials were performed according to the International Conference on

Harmonisation/Good Clinical Practice guidelines, and they adhered to the principles

of the Declaration of Helsinki. Before the onset of enrolment, both the trials protocols

were reviewed and approved by University of Malawi, College of Medicine Research

and Ethics Committee and the ethical committee of Pirkanmaa Hospital District in

Finland. The details of the protocols were published at the clinical trial registry of the

National Library of Medicine NCT00131222 and NCT00131209.

68

6 RESULTS

The main results of the 2 trials will be presented separately but those for the 3 sub-

studies in the complementary feeding trial will be presented rather integrated. The

results on supplementary feeding to underweight infants and children will be

presented first, followed by main result from the complementary feeding trial. Pooled

results on compliance to trial supplements, morbidity and serious adverse event (SAE)

will be presented together for the 2 trials.

6.1 Supplementary feeding to moderately underweight infants (study I)

6.1.1 Enrolment and Follow-up

Of the 1657 initially screened infants and children for the study on management of

moderate underweight, 367 were too old (aged > 14.99 months), 1010 were above

anthropometric cut-off for further evaluation (WAZ > - 1.80). Among the rest, 30

were not brought to the enrolment session, 72 were not underweight (WAZ > - 2.00),

and parents of 2 infants declined participation after receiving full information of the

trial. The remaining 176 infants and children were randomised into two intervention

groups. Two infants had their enrolment day rescheduled because they were too ill on

the first pre-planned enrolment day. There was no difference in the mean weight for

69

age Z-score between the enrolled children and those who were eligible but not

enrolled (P = 0.34).

Three participants died and another 3 were lost to follow-up before the end of the

intervention. The assumed causes of death were anaemia and respiratory tract

infection and respiratory tract infection. From the lost participants, one moved away

and two were unavailable for final measurements although they had participated in all

home visits and received and apparently eaten the intervention food. Comparison of

baseline WAZ between participants who completed the follow-up and those who died

or were lost did not suggest that they came from different populations (P = 0.15 and P

= 0.70; respectively). The losses to follow-up were also not significantly different

between the intervention groups (P = 0.61 for deaths and P = 0.68 for total loss to

follow-up; Fisher’s exact test). Sensitivity analysis to assess robustness of the results

after loss to follow-up produced similar standardised weight gain.

6.1.2 Background data and compliance to the interventions

At enrolment, participants in the LP group were on average five days younger, 100 g

lighter and 0.6 cm shorter than those in the FS group. The prevalence (number of

participants) with underweight (WAZ < -2) was 100% (86) in the LP and 99% (89) in

the FS group. Comparable figures for stunting (LAZ < -2) were 69% (59) vs. 64%

(33) and those for wasting (WLZ < -2) 16% (14) vs. 13% (12), in the LP and FS50

70

groups, respectively. Eight of the participants had a positive HIV-antibody test at

enrolment, but only five of them were truly HIV infected, as evidenced by a positive

polymerase chain reaction (PCR) test (Table 2 in original publication I).

6.1.3 Growth and recovery from undernutrition during supplementary feeding

After 12 weeks intervention, mean gain in weight was 50 g (95% CI -80 to 174 g)

higher in the FS50 group than the LP group. Comparable change for length was 0.2

cm (95% CI: -0.5 to 0.2 cm) lower in the FS group. Correspondingly, mean weight-

for-age Z-score (WAZ) and weight-for-length Z-score (WLZ) increased more and

mean length-for-age Z-score (LAZ) fell more in the FS50 group than the LP group,

but none of these differences reached statistical significance. Average changes in

mean mid upper arm circumference, head circumference, and blood haemoglobin

concentration were also quite similar in the two groups. All 95% CIs were consistent

with only relatively small between-group-differences to either direction.

Because of the limited outcome difference between the groups, we compared the

variation in weight gain to the predictions used for sample size calculation. The

observed standard deviation for weight gain was 0.38 kg in FS50 group and 0.46 kg in

LP group, compared to estimated SD of 0.39 kg in both groups. Whilst the observed

variation is slightly higher than expected, the 95% CI of the SD of the whole sample

and that of the two intervention groups included the expected value of 0.39 (Jacknife

method), suggesting that a marked between-group difference would have been

identified with the current sample and trial design if it existed in the population.

71

During the 12 week intervention, approximately 20% of the initially underweight

individuals, 85% of the initially wasted, and 10% of the initially stunted children

recovered from their condition, with little difference between the two intervention

groups. Correspondingly, prevalence of undernutrition at the end of follow-up (the

function of enrolment prevalence, incidence of new cases and recovery during the

intervention) was not significantly different across the groups. However, the

prevalence of underweight (WAZ < -2.00) was reduced significantly in both groups

(P < 0.001 for LP and 0.001 for FS50, sign test). A similarly reduced trend was

noticed for the prevalence wasting (P < 0.001 for LP and 0.07 for FS50, sign test),

and low mid-upper arm circumference (P = 0.02 for LP and < 0.001 for FS50, sign

test). The prevalence of stunting (P = 0.27 for LP and P > 0.99 for FS50, sign test)

and anaemia (P = 0.69 for LP and 0.04 for FS50, sign test) were not markedly

changed during the intervention.

At the end of intervention, the mean WAZ increased by 0.22 (95% CI: 0.07 to 0.37)

Z-score units in the LP group and by 0.28 (95% CI: 0.18 to 0.40) Z-score units in the

FS50 group (Figure 5a). Increase for mean weight-for-length was 0.39 (95% CI: 0.20

to 0.57) Z-score units and 0.52 (95% CI: 0.38 to 0.65) Z-score units for LP and FS50

groups, respectively (Figure 5b). No similar shift was observed for length for age Z-

score units (Figure 5c).

72

Figure 5: Anthropometric population shift after supplementary feeding

6.2 Supplementation of complementary food to health infants (Study II)

6.2.1 Enrolment and follow-up

Of the 303 initially screened infants, 65 were too old (aged > 6.99 months), 2 were

too young (aged < 5.5 months) and 2 were too ill on the day of screening or

enrolment. Of the remaining 234 infants, 49 were not brought to the enrolment session

(3 infants died, 16 infants moved away, parents of 7 infants were not interested and

Enrolment

Final

0

.2

.4

.6

0

.2

.4

.6

-8 -6 -4 -2 0

LP

FSDen

sity

Weight-for-age Z-score

0

.2

.4

.6

0

.2

.4

.6

-6 -4 -2 0 2

LP

FS

Weight-for-length Z-score

c

0

.2

.4

.6

0

.2

.4

.6

-6 -4 -2 0

LP

FS

Length-for-age Z-score

a b

73

we could not get any explanations from parents of 23 infants), and parents of 3

infants declined participation after receiving full information of the trial. The

remaining 182 infants were randomized into three intervention groups.

During the 12-month follow-up, 10 infants died and 4 were otherwise lost to follow-

up before age 18 months. The success rate of following up to age 18 months was not

significantly different between intervention groups (P = 0.47; Fisher’s exact test).

Only 6 participants had no anthropometric data at all after enrolment and there was no

difference in this between intervention group (P = 0.70; Fisher’s exact test). The

assumed causes for the 10 deaths were infections (diarrhoea, malaria, meningitis, and

respiratory tract infection) and injuries (drowning, poisoning). Of the 176 from whom

we collected anthropometric measurements 5 did not show-up for developmental

assessment. Thus 163 infants participated in the developmental assessment study.

Like for anthropometric measurements, no statistically significant differences were

observed in proportions of success to follow-up, deaths and dropouts across the trial

groups (P = 0.24, 0.77 and 0.13, respectively). Reasons for not showing up were not

established.

To assess sustainability of intervention effects after, we continued the follow-up for

another 24 months after the intervention. During the 24-month post-intervention

follow-up, there were six deaths and 13 further losses to follow-up, leaving 149

participants, who completed the 36 month assessment. Like in during the intervention

period, there was no statistically significant difference in the proportion of deaths and

74

dropouts at 36 months follow-up between the food intervention groups (P = 0.72 and

0.26, respectively).

6.2.2 Background data

At enrolment in the complementary feeding trial, mean anthropometric measurements

were comparable in the LP and FS50 groups, whereas infants in the FS25 group were

on average 250 and 380 g heavier than the FS50 and LP groups respectively. They

were also 0.3 and 0.7 cm longer than infants in the other 2 groups, respectively. No

participant was severely wasted at beginning; the prevalence (number of infants) of

severe underweight in the LP, FS50, and FS25, groups was 2% (1), 3% (2), and 0%

(0), and that of severe stunting 3% (2), 7% (4), and 2% (1), respectively. Nine of the

participants had a positive human immunodeficiency virus (HIV) antibody test at

enrolment, but only 1 of them was truly HIV infected, as evidenced by a PCR test.

6.2.3 Growth during complementary feeding and incidence of undernutrition

Of the 176 participants with anthropometric data, mean gains in weight and length

were 100 g (95% CI -143 to 343 g and 0.8 cm (95% CI: -0.1 to 1.7 cm) higher in the

FS50 than the LP group. Correspondingly, mean decreases in weight-for-age Z-score

(WAZ) or length-for-age Z-score (LAZ) were smaller among the FS50 than the FS25

75

or LP infants and there was also a smaller drop in blood haemoglobin concentration in

the FS50 group. None of the differences, however, reached statistical significance.

There was an interaction between intervention group (FS50 vs. LP) and baseline LAZ,

both using length gain (P = 0.04) or weight gain (P < 0.01) as an outcome. Among

subjects with baseline LAZ below median in this trial (-1.035), the mean gain in

length was 1.9 (95% CI: -0.3 to 3.5) cm or 0.40 (95% CI: -0.15 to 0.95) Z-score units

bigger in the FS50 than in the LP group. Comparable differences in weight gain were

404 (95% CI: -74 to 735) g or 0.43 (95% CI: 0.07 to 0.80) Z-score units. Among

subjects with baseline LAZ above median, the difference in length was -0.4 (95% CI:

-1.3 to 0.5) cm or -0.09 (95% CI: -0.58 to 0.39) Z-score units and that in weight was -

254 (95% CI: -604 to 96) g or -0.25 (95% CI -0.64 to 0.14) Z-score units. There was

no significant interaction between intervention and baseline WAZ (P = 0.14).

As secondary end-points, we looked at the proportion of subjects who developed

severe or moderate-to-severe underweight, stunting and wasting during the follow-up.

The proportion of subjects who developed other forms of undernutrition did not differ

markedly between the intervention groups, but severe stunting occurred significantly

less frequently in FS50 and FS25 than in LP group. After enrolment, no infant in

FS50 group, 4% in FS25, and 13% of LP infants developed severe stunting (P = 0.01;

Fisher’s exact test). The 95% CI for the difference between FS50 and LP was 4 to

21%.

76

Cumulative incidence of stunting during the 12-month period was also calculated

based on survival analysis methods that dealt with censoring differently from the

analysis confirmed that severe stunting developed less often and later in the FS50 and

FS25 than in the LP group (0%, 4% and 14% respectively; P = 0.01, log-rank test).

The cumulative incidence of moderate-to-severe stunting was similar in the three

groups, but infants in the FS50 group developed the condition on average somewhat

later. However, the latter result was not statistically significant (P = 0.66, log-rank

test).

6.2.4 Developmental outcomes during complementary feeding (Study III)

The background anthropometric characteristics at developmental assessment and

changes between enrolment and developmental assessment were mostly similar. The

age at developmental assessment was similar across the groups. The proportion of

boys and mean WAZ, LAZ and WLZ were lowest in LP group and highest in FS25.

Like in the main growth analysis, changes in anthropometric indices among

participants in developmental assessment during the intervention period did not differ

significantly between the groups (P = 0.80, 0.70 and 0.67 for WAZ, LAZ and WLZ,

respectively).

The Griffith scores were not different across the groups. The mean (SD) general raw

Griffiths scores were 227 (9.0), 226 (11) and 226 (10) for the LP, FS50, and FS25

77

groups, respectively (P = 0.90). Correspondingly, mean general quotients were 101

(7), 100 (9) and 100(8) (P = 0.88) and mean (SD) mental ages for each group were

17.9 (1.3), 17.9 (1.3) and 17.9 (1.2) months (P > 0.99). Similar comparisons of the

trial groups on all the developmental subscales showed no statistically significant

differences. There was no statistically significant differences between mean

chronological age and general mental age in each of the groups or when all the groups

were pooled together; pooled mean difference was 0.01 (95% CI: -0.19 to 0.21).

Exploratory analysis through multiple regression suggested that only the capacity of

mothers to write and the participant’s length-for-age Z-score at 18 months of age were

significantly associated with the general developmental score (P < 0.05).

6.2.5 Post intervention growth after complementary feeding (Study IV)

In the post-intervention follow-up, 4-7% of the participants in each group developed

severe stunting, bringing the cumulative 3-year incidence (number of cases) of severe

stunting to 20% (11) in LP, 4% (2) in FS50, and 10% (6) in FS25 groups (P = 0.03,

log-rank test). The point estimate for the difference between LP and FS50 was 16%

(95% CI: 5% to 26%) and number needed to treat to prevent one case of severe

stunting was 6 (95% CI: 4 to 20). Cumulative incidence of moderate-to-severe

stunting was similar in the three groups, although infants in the FS50 group tended to

develop the condition at on the average somewhat later age (P = 0.50, log-rank test).

78

Among the 149 participants who completed the follow-up, mean 3-year gains in

weight, height, mid upper arm and head circumferences were highest in the FS50

group and lowest in the FS25 groups. Average values for anthropometric indexes

weight-for-age (WAZ), height-for-age (HAZ) and weight-for-height (WHZ) fell in all

groups, but the reductions were smallest among the FS50 and largest in the FS25

groups. However, inter-group differences reached statistical significance among

unselected children only for WAZ-change. Compared to children in the FS50 group,

those getting LP had on average 0.33 (95% CI: -0.39 to 0.69) Z-score units and those

getting FS25 0.46 (95% CI: 0.10 to 0.82) Z-score units larger reduction in their WAZ

during the follow-up. Similarly, those getting LP had 0.10 (95% CI: -0.23 to 0.43) Z-

score units and those getting FS25 0.34 (95% CI: 0.02 to 0.67) Z-score units larger

reduction in their HAZ during the follow-up. Adjusting the analyses for baseline

weight resulted in similar findings to the unadjusted analyses (data not shown).

Because of our earlier observation showing that there was a strong interaction

between child’s length at 6 months of age and his/ her weight and length gains during

the intervention, we conducted stratified analyses on these outcomes also during the

post-intervention follow-up. Largest anthropometric gains were seen in the FS50

group and smallest in the FS25 group, both among children with baseline height

below the population median as well as those above it. However, the absolute

differences in weight and length were bigger and they more often reached statistical

significance in the initially shorter participants. Among these infants, those getting

FS50 had on average 0.61 (95% CI: -0.15 to 1.37) kg higher weight gain or 0.53 (95%

79

CI: 0.07 to 0.99) Z-score unit smaller reduction in WAZ than LP children and 0.91

(95% CI: 0.09 to 1.73) kg higher gain or 0.81 (95% CI: 0.31 to 1.30) Z-score unit

smaller reduction in WAZ than FS25 children. Differences in height gains showed a

similar pattern but did not reach statistical significance.

The differences in height-for-age (HAZ) between LP and FS groups started to develop

after 4 months (when children were 10 months old) and reached a peak 4-8 months

later (at 14-18 months of age). In contrast, differences in weight-for age or weight for

height increased gradually during the 12-month intervention and continued to grow

wider during the 2-year post-intervention follow-up. The patterns were similar among

participants with baseline height above or below median but absolute values were

more pronounced among those who were shorter at the beginning.

6.3 Haematological changes (Study I and II)

In the supplementary feeding trial, the mean gain in haemoglobin concentration after

12 weeks supplementation was 3.8 g/L in the FS50 group compared to 1.5 g/L in the

LP group; these shifted the group means to 94.5 and 94.8 g/L, respectively. Compared

to enrolment measurements, in all the 3 trial groups, the mean haemoglobin

concentration declined marginally after one year provision of complementary food

supplements the 3 groups. The smallest decline occurred in the LNS groups; <1 g/L, 4

g/L and 7 g/L in the FS50, FS25 and LP groups, respectively. Correspondingly after

one year of complementary intervention, all the three groups had groups had mean

80

haemoglobin concentration of 106 g/L, 109 g/L and 107 g/L respectively; at most 4g

below the cut-off for anaemia in children (110 g/L).

6.4 Compliance to the interventions supplements (Study I and II)

All mothers reported that their children readily ate the provided supplement and

diversion of any portion to someone else than the intended beneficiary was reported in

less than 1% food delivery interviews, for all the food groups in both trials. Within

trial comparisons showed no difference across the groups. From the home visits

during which trial products were checked, the percentage of visits found with

leftovers ranged from 3% to 10%. Between groups the proportion of leftovers in the

supplementary feeding trial were 4.9% and 6.5% in the LP and FS50 groups,

respectively (P = 0.18). The difference was significant in the complementary feeding

trial found to be 2.8%, 9.8%, and 5.6% in the LP, FS50 and FS25 groups, respectively

(P < 0.001).

6.5 Morbidity and serious adverse events (Study I and II)

All the food products (LNS and LP) were well tolerated by the participants in both

trials. None of the eligible participants who received the 6g test dose was allergic to

LNS. In addition to the 13 deaths, 6 participants were hospitalised and recorded as

81

having experienced a serious adverse event (SAE) during the follow-up. In both trials

the SAEs were not significantly different between the trial groups. Both the study

physician and the data monitoring boards for the trials considered all deaths and SAE

unlikely related to the trial interventions.

82

7 DISCUSSION

The present study was planned to compare the health benefits (promotion of growth

and development, primary prevention of undernutrition and secondary prevention of

severe acute undernutrition) associated with supplementary provision of either lipid-

based nutrient supplement or iso-energetic multiple micronutrient fortified corn-soy

blend. The complementary feeding trial tested the effect on primary prevention of

undernutrition and promotion of growth and development; on the other hand, the

supplementary feeding trial tested the effect on secondary prevention of

undernutrition. This chapter presents a discourse on the observed health benefits

including growth and developmental outcomes of this study.

7.1 Strength and weaknesses for the study

Several factors contributed to the strength of this study: the study design, population

based recruitment, high enrolment rate, high compliance to the study with minimal

attrition and minimisation of measurement bias. Group allocation in both trials was

random ensuring that the distribution of known and unknown confounding variables

was similar in each of the comparison trial groups. The randomisation was perceived

successful in both trials based on similarity of baseline group characteristics in each of

the trials and hoping that distribution of unknown confounders was also equally

successful. The recruitment to the trials was population based in which all guardians

83

to potential participants in the catchment area were directly invited to participate.

Turn up was high further minimising the risk of selection bias. The sample sizes were

calculated to provide both trials power of 80% and confidence of 95%; and to mitigate

effects of attrition, an additional 5% of the initially calculated sample size was added

to each.

After enrolment, adherence to the trials were high with few deaths and losses to

follow-up across the trial groups. Measurement bias was minimised by training of the

data collection team. Reliability indices for the two people who took anthropometric

measurements were high. All instruments and laboratory equipment were calibrated

regularly, ensuring minimal variability of outcomes.

The trial implementations were highly controlled. All food supplements were

delivered at home at appropriated interval for each study. Food use was verified by

collecting back empty containers; both nutritional supplements were well accepted

with greater 90 % consumption reported from all the trial groups. Disease symptoms

were monitored every week, and infants found ill referred for treatment at the nearest

health centre and where necessary the cost of treatment were met by the study.

Mothers were encouraged to continue breast feeding and all infants were still

breastfeeding by the end in the interventions.

84

However, despites the above strengths there are a few limitations on how far we can

make an inference from this study. Because in both trials for ethical reasons, there

were no non-supplemented control groups, thus conclusions drawn from the study are

mainly on comparison of health benefits after provision of lipid-based nutrient

supplements or corn-soy blend. Although used widely and for many years in food aid,

efficacy and effectiveness of corn-soy blends to vulnerable population has been

questioned (Marchione 2002, Hoppe et al. 2008); and furthermore data are scarce. In a

review on complementary food supplements it was found that in the past several

decades only 2 relatively new studies compared corn-soy blends to no food

supplements (Dewey and Adu-Afarwuah 2008). Initially supplied as corn-soy-milk

(CSM) because they contained non-fat milk, scarcity of milk in 1980s led to

adaptation to corn-soy blend (CSB) without field trials (Marchione 2002). Despite

this, we know that the CSB control group in our study provided a reasonable range of

protein, with the 2 crops (corn and soy) complementing each others amino acids

deficits. However, the protein quality only provided up to 65% PDCAAS. As plant

source they also carry anti-nutrients such as phytic acids which may limit absorption

of other minerals like zinc and iron. However, compared to no supplementation

results from efficacy studies conducted in Malawi, Ghana and South Africa suggest

modest growth promotion effects after supplementation with corn-soy blends

(Kuusipalo et al. 2006, Lartey et al. 1999, Oelofse at al. 2003); thus we expected the

corn-soy blend to have had some growth promotion effect.

85

Another potential weakness is lack of dietary data during the post-intervention period

in the complementary feeding trial. However, we assume that food consumption

patterns were distributed equally between the groups because of the randomisation.

It is with this background of the study’s strengths and weakness that conclusion from

the findings of the study are drawn. Despite the shortfall on how far the results of this

study can be extrapolated, the study design and implementation allow us to make firm

conclusions on the comparison between effect of LNS and CSB on promotion of

growth and development, and on prevention and treatment of moderate

undernutrition.

7.2 Growth promotion and management of undernutrition

7.2.1 Enrolment growth status

At enrolment age (6 months) in the complementary feeding trial as expected, mean

weight-for age and weight for length were consistently normal compared to reference

population (Shrimpton 2001). However, length-for-age was rather low with all groups

being mildly stunted. Corresponding enrolment indices for the supplementary feeding

trial were determined by inclusion and exclusion criteria. Thus as expected the mean

weight-for-age Z-score were < -2 units and weight-for-length Z-score > -2. All wasted

86

infants were referred for to the health centre for management of acute malnutrition in

accordance with the Ministry of Health (Malawi) policy on management of infant

malnutrition. Both groups in the supplementary feeding trial were, however, also

found to be stunted at enrolment, suggesting that the infants may have been

underweight because to stunting rather than wasting.

7.2.2 Promotion of growth

At the end of both interventions, compared to the LP group, mean gain in weight were

higher in the group of infants receiving iso-energetic LNS, albeit not statistically

significant. Similar gains for mean length were only observed in the complementary

feeding trial. There was interaction between baseline length-for-age Z-score versus

LNS for weight and length gains in the complementary feeding trial. Significantly

higher weight and length gains occurred in those below baseline median length-for-

age Z-score than in those above median; suggesting stronger effect of LNS in those

already undernourished at the start of trial. Twenty-four months after the

supplementary intervention stopped weight and length differences were at least the

same or larger as seen after 4 to 12 months and somewhat larger in the LNS groups

than in the LP group. These observations were similar either in absolute weight and

length values or standardised values and more marked in those with baseline length-

for-age Z-score less than median (Figure 3 in original publication III); and they

support the notion that the critical time to achieve maximum effect from promotion of

87

complementary feeding is as early as possible but of course not earlier than 6 months

of age (Bhandari 2001, Dewey 2001).

Compared to the LP group, the different upward direction taken by the length trend of

the FS50 group as early as 4 month into intervention period is consistent with the idea

that FS50 supplementation somehow affected the timing of initiating the childhood

growth phase or acceleration in the infants’ linear growth velocity in of the ICP

growth model (Karlberg et al. 1994). Such acceleration occurs at an average age of 9

months in industrialized countries, but often much later in low-income settings (Liu et

al. 2000, Xu et al. 2002). Each month of delay in childhood growth-spurt is associated

with on average approximately 0.5 cm deficits by 5 years in length gain (Liu et al.

2000, Xu et al. 2002).

Factors inducing the childhood growth-spurt in growing infants are unknown at

present, but dietary intake of cow’s milk may play a role in the process (Hoppe et al.

2006). Theoretically, it is thus possible, that the observed linear growth differences in

our trial were due to the cow milk protein fraction in FS50 (10 g in the 50 g daily

dose), an increased intake of which might have led to on average of 4-month earlier

induction of the childhood growth-spurt in the intervention compared to the control

group. This suggestion, whilst still quite hypothetical, receives some support from the

fact that children grew less in the group that received a lower dose of FS (FS25), with

same micronutrient content but only half the amount of milk. Furthermore, another

recent complementary feeding trial in Malawi, in which same age children were

88

supplemented with soy-containing modification of LNS (no milk powder) did not

document any length gain acceleration (Lin et al. 2008). Although corn and soy

complement each others limiting deficient amino acids (amino acid tryptophan for

corn and amino acids methionine and cysteine for soy), the PDCAAS of corn soy

blend is only up to 65%. Milk powder and peanut on the other hand have a PDCASS

of 95% each (WHO 2007, Hoppe et al. 2008). Other possible explanations to the

linear growth variation during intervention period include the different concentration

and bioavailability of other micronutrients, e.g. zinc or essential fatty acids, in the

FS50 and corn-soy blend (Brown et al. 2002 Adu-Afarwuah et al. 2007).

Contrary to the linear growth, weight differences between the groups grew larger

especially after the intervention. Whilst part of this may be attributed to the timing of

IC-spurt, other variables are likely to play a role, as well. Possible explanations

include the impact of FS50 supplementation on the children’s general health and

susceptibility to infections, or appetite. Alternatively, the guardians could have

continued continue providing nutritious snacks to children who had earlier received

snack-like FS50 but not to those, who had received LP. Whilst biologically plausible

explanations, however, we have no data to support any of these possibilities. Another

hypothetical explanation for this is programming; however, it has been suggested that

this phenomenon occurs in utero (Lucas et al. 1999)

Another notable point we would like to highlight from our results is the comparison of

outcomes among children receiving the higher (50 g/d) and the lower (25 g/d) dose of

89

FS. As the price for such (50 g) a dose (approx 0.2 USD / day) would be relatively

high for the rural families in Malawi, we wanted to see if half of the dose would

produce the same growth outcomes but more inexpensively. Unfortunately, this was

not the case, children in the lower-dose group had both higher incidence of severe

stunting and lower mean weight and height gains during the intervention and

especially thereafter. A comparable phenomenon has been earlier observed in a

Malawian dose-finding trial, where underweight infants were given different doses of

LNS for 12 weeks (Kuusipalo et al. 2006). Because a smaller (20 g/d) dose of a

similar supplement (Nutributter) yielded positive results on linear growth and motor

development among 6-12 month old infants in Ghana (Adu-Afarwuah et al. 2007), the

larger dose may thus be more appropriate in the Malawian setting. The apparent

discrepancy may be explained by the much higher degree of stunting in the Malawian

setting or the fact that the higher dose was not tested in the Ghanaian setting (Adu-

Afarwuah et al. 2007).

In the supplementary feeding trial, considering the length of intervention period and

timing of observations it is not surprising that no change was observed in length gain.

This does not necessarily contradict observations from the complementary feeding

trial but rather point toward unsynchronised timing of intervention and

anthropometric measurements. Length gain acceleration has often been shown to

follow weight gain increase with a certain lag period approximately 3 months in rural

Malawi (Maleta et al. 2003b). Mean gain in head circumference and mid-upper arm

circumference were almost comparable among the groups in each of the trials.

90

7.2.3 Promotion of recovery from undernutrition

In the supplementary feeding trial, gains for mean weight-for-age and weight-for-

length Z-score were higher in the LNS group than the LP groups, but not significantly

so. There was a somewhat decrease or no change in length-for-age Z-score.

Correspondingly, there was good proportion that recovered from wasting and

underweight but not as much on stunting. These observations suggest that either both

supplementation schemes were effective in promoting weight gain or it is an effect

from a confounding variable equally distributed in the trial groups such as seasonality;

or from regression to the mean phenomenon. A comparison to earlier studies in the

same area suggest that unsupplemented underweight children of this age would have a

rather lower weight gain velocity (Kuusipalo et al. 2006) and that seasonal effect on

weight gain would be markedly lower than the approximately 0.3 Z-score unit

increase observed in this trial (Maleta et al. 2003b). These increases in standardised

weight, however, are comparable to those observed after other supplementary feeding

intervention to similar target groups (Simondon et al. 1996, Lartey et al. 1999, Becket

et al. 2000, Bhandari et al. 2001, Oelofse et al. 2003, Adu-Afarwuah et al. 2007).

These results on recovery from undernutrition supports findings from other studies

and programmes suggesting that LNS promotes recovery from moderate wasting and

underweight (Patel et al. 2005, Defourny et al. 2007, Matilsky et al. 2009); thus acting

as a secondary prevention of severe acute undernutrition.

91

7.2.4 Prevention of undernutrition

Although the long-term LNS and LP supplementation failed to prevent plummeting of

mean weight-for-age, length-for-age and weight-for length Z-scores which are

common in infants aged 6 to 24 months in developing countries (Shrimpton 2001), the

decline was slowest in participants in the LNS group, and markedly so among those

with baseline length for age Z-score at enrolment (Figure 4, in original article III).

The corresponding incidence of stunting was lower in the group provided with LNS

than those provided with iso-energetic LP or half the dose of LNS. During the

intervention period no case of severe stunting occurred in the group receiving high

dose LNS, 4% in group receiving half the dose of LNS and 14% in the group

receiving LP. Postintervention assessment showed that incidence of severe stunting

continued to be significantly lower in the LNS groups than the LP groups, suggesting

that the observed results were sustainable (4 %, 10% and 20% for the FS50, FS25 and

LP groups, respectively). These observations were confirmed in a survival analysis

that dealt with censoring differently.

Despite the trends of weight-for-age and weight-for-length Z-score being similar to

that length-for-age Z-score, incidence of underweight and wasting were similar in all

the trial groups. As weight responds more to daily body composition, temporary body

composition changes may have influenced the incidence of underweight and wasting

while not having the same impact on much more stable measure like height and the

trends.

92

After a rather high quality implementation of a supplementary trial we only managed

to mitigate the decline but not necessarily stop the pattern of declining anthropometric

indices commonly seen between 6 and 24 months in developing countries (Shrimpton

2001). The possible explanation for this could be inadequate energy and nutrient

intake, disease burden, or predetermined down regulation programming occurring

before the intervention period. All infants were breastfeeding up to 18 months

therefore they received an estimated of 413, 379 and 346 kcal/d for ages 6-8, 9-11 and

12-23 months (WHO 1998, Lutter 2003). With the LP and FS50 providing 282 and

256 kcal/d and the daily energy needs for the 3 age-ranges being 615, 686 and 894

kcal/d, it means the breast milk plus the LP or FS50 provided over 100%, 93% and

70% for the these age ranges (WHO 1998, Lutter 2003). These infants were also on

traditional household complementary foods, thus it unlikely that they were deficient

of energy during the intervention. Within the same group of participant there was no

evidence of breast milk intake displacement at for the first month after starting the

provision of the supplements (Galpin et al. 2006), although it has been reported

elsewhere (Bhandari et al. 2001). The distribution of days with fever or reported

diarrhoea or respiratory tract infections were equally distributed across the study

groups. The mean number of days with diarrhoea was 5 and the total mean days with

fever was 11 days per person per year. All morbidity cases were rigorously identified

through weekly home visits and mother were encouraged to go for medical care if

participants were found ill. It is therefore again unlikely that infections either directly

or through competing for nutrients explain the growth faltering. All the supplements

were fortified with multiple micronutrients suggesting that micronutrient deficiency is

also unlikely. The presence of anti-nutrients in corn soy blend, however, may explain

93

the relative difference between LP and FS50 but not the general decline in all groups.

Of course, all groups could have been receiving anti-nutrients from their traditional

complementary food stuff. Theoretically, the growth response to the supplementation

may have been constrained by predetermined prenatal programming or

intergenerational effect of maternal stunting (Dewey 2001). The explanation for the

persistent growth faltering despite the intervention is difficult to establish from this

study; thus, these results call further research to evaluate this question further.

7.3 Promotion of recovery from and prevention of anaemia

At enrolment, infant in the supplementary feeding trial were anaemic with the mean

group haemoglobin concentration of 93 g/L and 91 g/L for the LP and FS50 groups.

Those in the complementary feeding trial were 114 g/L, l06 g/L and 113 g/L for the

LP, FS50 and FS25 groups. Haemoglobin of less than 110 g/L defines anaemia in this

age group (Allen et al. 2006). In the supplementary feeding trial, the mean gain in

haemoglobin concentration was 3.8 g/L in the FS50 group compared to 1.5 g/L in the

LP group. Provision of micronutrient fortified supplements to anaemic population to

induce induces population recovery from anaemia. However, the provision period

may have not been long enough to reach the cut of point. All haemoglobin

concentration declined modestly after one year provision of complementary food

supplements the 3 groups. The smallest decline occurred in the LNS groups; <1 g/L, 4

g/L and 7 g/L in the FS50, FS25 and LP groups. All the groups had mean

haemoglobin concentration greater or equal to 106 g / L, reasonably close 110 g/L, the

cut-off for anaemia. This finding suggests that all supplementary schemes may have

94

had an effect on haemoglobin concentration. However, without non-supplemented

control group this interpretation needs to be considered cautiously as these finding

may be coming an equally distributed confounding factor.

7.4 Developmental outcomes after complementary feeding

Developmental assessment was conducted in the complementary feeding trial only. In

all its 3 trial groups the developmental findings at the end of intervention were equal.

These results are similar to those observed in other studies where food supplements

fortified with multiple micronutrients produced comparable developmental outcomes

(Black et al. 2004, Adu-Afarwuah et al. 2007). After a six-month supplementary

intervention from the age of 6 months, a trial in Ghana showed no significant

differences between micronutrient fortified LNS and micronutrient powder or

micronutrient tablets in promoting motor development (Adu-Afarwuah et al. 2007).

Another trial in Bangladesh showed no significant difference in psychomotor

development when a group receiving multiple micronutrients was compared to a

group receiving combined iron and zinc supplements (Black et al. 2004). However, all

micronutrient fortified supplements in Ghana, and multiple micronutrient or combined

iron and zinc supplements in Bangladesh promoted motor development more than

non-supplementation or placebo groups, respectively (Adu-Afarwuah et al. 2007,

Black et al. 2004). Several other studies have also shown that micronutrient

fortification of supplements (Moffat et al. 1994, Harahap et al. 2000, Olney et al.

2006,) or unfortified supplementation (Husaini et al. 1991) promotes neurobehavioral

development more than non supplementation in infants.

95

The 3 food supplements used in this study had different levels of essential fatty acids:

The corn soy blend, FS50 and FS 25 provided an additional 0.3, 0.8 and 0.4 g of alpha

linolenic acid (ALA) per day and 2.7, 7.3 and 3.6 g of linolenic acid (LA),

respectively. A differential impact on development was plausible as essential fatty

acids are needed for brain growth (Helland et al. 2003, Innis 2008 and 2009). It is

possible, however, that with the additional essential fatty acids provided by breast

milk, the overall difference of intake was not sufficient to have an influence on

development quotient. The recommended adequate intake of LA and ALA was 4.6

and 0.5 g/d (Food and Nutrition Board 2005), and breast milk was estimated to

provide at least 4.4 and 0.4g/d given the minimum intake was 830 g/d (Galpin et al.

2006).

Although with lack of a non-supplemented control group it is difficult to explain

what were the independent effects of the supplementation schemes, the fact that the

participants attained mean general mental ages comparable to their own chronological

age is, however, consistent with the possibility that all the schemes were effective.

This possibility is further supported by the notion that the observed mental ages were

similar to those from the reference population of the Griffiths Scales (Black et al.

2004) and those attained by British children of similar age in a different study (44).

Because the study area had a high prevalence of stunting, malaria is endemic and

income-level was low (UNICEF 2008, Maleta et al. 2003c), it was unlikely that the

participants from these living conditions would attain higher developmental outcomes

when compared to reference population from a high socioeconomic setting (Olney et

96

al. 2007, Siegel et al. 2005, Cheung et al. 2001, Grantham-McGregor et al. 2007,

Kariger et al. 2005). In a study mentioned earlier, the Ghana trial, using a somewhat

similar approach to ours, a difference in motor development was noticed in an

analysis that compared no intervention to one with lipid-based nutrient supplements

but not between two different supplement types (Adu-Afarwuah et al. 2007).

The adaptation method of the locomotor and personal social subscales for use in rural

Malawian community resulted in lower possible scores than would typically be in

those subscales. Therefore, the attained raw score were likely to be lower than could

normally be attained. No specific correction was performed on the raw score or the

corresponding mental ages in these subscales. Thus, the lower mental ages than

chronological ages attained in these specific subscales are likely or partly explained

by the adaptation of the Griffith tools. The lower mental ages in these subscales,

however, should not have affected our group comparison as the adaptation equally

affected all the groups.

97

8 SCIENTIFIC CONCLUSIONS

The present study was planned to compare the health benefits (promotion of growth

and development, primary prevention of undernutrition and secondary prevention of

severe acute undernutrition) associated with supplementary provision of either lipid-

based nutrient supplement or iso-energetic multiple micronutrient fortified corn-soy

blend. After the implementation the results show that

1. Supplementary provision with lipid-based nutrient supplements and corn-

soy blend to moderately undernourished 6 to 15 months old infants have

similar impact on weight and length gain in low income settings. The

results are also consistent with the notion that both interventions provided

for 12 weeks only promote recovery from wasting and underweight but not

stunting.

2. Compared to iso-energetic corn-soy blend, 12 months supplementary

provision with lipid-based nutrient supplements to otherwise healthy 6

months old infants induce earlier accelerated childhood growth phase of

the ICP growth model, with subsequent higher length and weight gains in

the next 36 months. Provision of LNS at half the energy and the milk dose

does not induce similar growth response when supplemented to initially

mildly stunted population.

98

3. Long-term provision with lipid-based nutrient supplements but not with

corn-soy blend protects against the incidence of severe stunting. However,

neither supplementation with LNS or CSB arrests the growth faltering

commonly seen between 6 and 24 months completely.

4. Provision of supplementary food fortified with multiple micronutrients,

including iron to mildly anaemic population of infants induces population

recovery; and when provided to non-anaemic infants for a long period they

have the potential to protect the population from gross manifestation of

anaemia.

5. Developmental status of groups of infants after one provision of

micronutrient fortified complementary food supplements is similar.

99

9 PUBLIC HEALTH IMPLICATION AND FUTURE RESEARCH NEEDS

This study shows that provision of infants with high dose lipid-based nutrient

supplements (LNS) achieves better growth rates than provision of low dose LNS or

corn-soy blends, especially among those already mildly stunted and prevent

development of stunting. Developmentally, their psychometric outcomes are not any

worse than reference population or any better than those in the CSB control group.

This may be one of possible solution to prevent development of undernutrition in low

in come countries, especially stunting which is associated with negative human capital

development. Failure to achieve completely normal growth rates during the

challenging period of infant growth with such a well implemented study, apparently

calls for more comprehensive approach such as maternal interventions as well.

The most effective public health strategy would be that which can be taken up by the

communities with minimal support from donor funds. The estimated ex-factory price

of LNS when made in Malawi is U.S. $ 4.00 / kg, corresponding to a daily cost of

U.S. $ 0.20 for a 50 g dose. Unsubsidized, this price is presumably unaffordable to

many families in rural Malawi or in other low income countries. Therefore, this

strategy is unlikely to be taken up by the communities without some kind of donor or

social support. Considering long-term effect of growth promotion intervention which

have been linked with better human capital it may still be cost effective approach

despite the high cost. Potential means for increased affordability include amendments

100

to the spread recipe (e.g. cheaper protein source) and social marketing.

Hypothetically, another way may be to vary dose based on prevalence of stunting at 6

months. In Ghana where prevalence of stunting at 6 months was lower, a 20 g daily

dose of LNS was found to be effective in promoting at least growth (Adu-Afarwuah

2007). Before implementing such a titre strategy aimed at lowering implementation

cost, this hypothesis need to be confirmed by clinical trials.

For populations where the focus is prevention of anaemia or promotion of

development only cheaper strategies of providing multiple micronutrients like the

Sprinkles need to be evaluated. Fortified corn-soy blend may be an equally good

option for promotion haemoglobin concentration and secondary prevention of severe

acute undernutrition.

Therefore, the findings of this study justify the need of several studies to find ways to

achieve full infant growth potential and to address the limited inference level of this

study.

1. An integrated intervention combining this approach with a trial arms that

promote reduction of prevalence of stunting at 6 months is justified. Growth

faltering seems to be programmed in utero or in early infancy. Such kind of

trial would target promoting pregnancy outcome or quality of breast milk

101

2. Interventions aimed at reducing cost of LNS are justified. Substituting milk

protein with soy protein extract is one option. Similarly, improving the quality

of corn soy blend by adding milk or using soy extracts is another option.

3. Trials using different doses of LNS doses titrated against baselines prevalence

of stunting at 6 months are justified.

4. To firmly make inferences on these finding trials with non-supplemented

control arms are warranted.

102

10 ACKNOWLEDGEMENTS

The study was carried out at the Department of International Health, Tampere

University Medical School and at the Department of Community Health, College of

Medicine, University of Malawi between 2004 and 2009.

Several people helped during my studies and it is impossible to individually

acknowledge all of them. I would the like to sincerely thank you all for your support.

The following, however, deserve special mention.

Professor Per Ashorn, the Professor of International Health Department at the

University of Tampere, for supervising and the guidance he has given me through out

this journey. His titrated supervision ensured my progress in challenging periods and

allowed me to take more self responsibility over this as work as my confidence grew.

I would also like to extend my appreciation for extending friendship to me and my

family and allowing me to make friend with his son Mikael.

Professor Kenneth Maleta, the Dean of postgraduate studies at the College of

Medicine, University of Malawi, for my recruitment, introduction into this fascinating

world of epidemiology and supervision. His academic support was crucial to my

103

commencement of this journey. His support to me and my family was ever amazing;

together with my daughter Chikumbutso we will never forget him, his son Kiran and

his wife Judith Maleta.

Dr Yin Bun Cheung for training me in statistics and all the wonderful guidance the

last 3 years of this work.

Chrissie Thakwalakwa, Andre Briend, Mark Manary, Lauren Galpin, Melissa

Gladstone for their collaboration in this study.

My colleague from Chrissie Thakwalakwa, Bernard Mbewe, Eunice Banda, Rose

Mzamu, Mercy Mpazo, Amina Atwabi, Misonzi Banda, Barry Kabai, Yusufu Saiti,

Lisa Jeremiya, Felix Baluwa, Richard Justine, Master Kamwendo, Lydia Mukala,

Osman Binoni, Ester Tsumba and Mariam Pongolani who made data collection in

very difficult remote conditions possible, I would like to thank you in a special way.

Special thanks goes to Lungwena Health Centre staff: Daniel Pondani, Pricilla

Pondani just to mention a few, without whose collaboration this work would have

been difficult.

104

Professor Jarmo Visakorpi, Professor Matti Salo for their independent observation

over my studies as members of my doctoral studies review board.

Late Dorothy Chilima, Tobias Chirwa, Steve Graham, Victor Mwapasa for their

participation in data safety boards of this study.

Thandizo, Chikumbutso and Chimwemwe Phuka, my wife and two beloved daughters

who survived long periods on my absence, thank you for the support, patience and

understanding.

Academy of Finland and Nestle Grant Foundation for financially supporting me and

my studies.

105

11 REFERENCES

Adair LS and Pollitt E (1985): Outcome of maternal nutritional supplementation: a

comprehensive review of the Bacon Chow study. Am J Clin Nutr 41:948-978.

Adair LS and Guilkey DK (1997): Age specific determinants of stunting in Filipino

children. J Nutr 127:314-320.

Adair LS (1999): Filipino children exhibit catch-up growth from 2 to 12 years J Nutr

129: 1140-1148.

Adu-Afarwuah S, Lartey A, Brown KH, Briend A, Zlotkin S and Dewey KG (2007):

Randomized comparison of 3 types of micronutrient supplements for home

fortification of complementary foods in Ghana: effects on growth and motor

development. Am J Clin Nutr 86:412-420.

Alderman H, Konde-Lule J, Sebuliba I, Bundy D and Hall A (2006): Effect on weight

gain of routinely giving albendazole to preschool children during child health days in

Uganda: cluster randomised controlled trial. BMJ 333:122.

Allen LH (1994): Nutritional influences on linear growth: a general review. Eur J Clin

Nutr 48:S75–89.

Allen LH (2003): B vitamins: proposed fortification levels for complementary foods

for young children. J Nutr 133:3000S-3007S.

106

Allen LH (2003 b): Intervention for micronutrient deficiency control in developing

countries: past, present and future. J Nutr 133:3875S-3978S.

Allen LH, de Benoist B, Dary O and Hurrell R (2006): Guidelines on food

fortification with micronutrients. WHO Press. Geneva.

Allen LH (2008): To what extent can food-based approaches improve micronutrient

status? Asia Pac J Clin Nutr 17(S1):103-105.

Arevshatian L. Clements CJ, Lwanga SK, Misore AO, Seward JF and Taylor P

(2007): An evaluation of infant immunization in Africa: is a transformation in

progress? Bull World Health Organ 85:449-457.

Arifeen S, Black RE, Antelman G, Baqui A, Caulfield L and Becker S (2001):

Exclusive breastfeeding reduces acute respiratory infection and diarrhea deaths among

infants in Dhaka slums. Pediatrics 108:e67.

Baltussen R, Knai C and Sharan M (2004): Iron fortification and iron supplementation

are cost-effective interventions to reduce iron deficiency in four subregions of the

world. J Nutr 134:2678-2684.

Beasley NMR, Hall A, Tomkins AM, Donnelly C, Ntimbwa P, Kivuga J, Kihamia

CM, Lorri W and Bundy DAP (2000): The health of enrolled and non enrolled

children of school age in Tanga, Tanzania. Acta Trop 76:223–229.

Becket C., Durnin JV, Aitchison TC and Pollitt E (2000): Effects of an energy and

micronutrient supplementation on anthropometry in undernourished children in

Indonesia. Eur J Clin Nutr 54(Suppl 2):S52-59.

107

Bhandari N, Bahl R, Taneja S, de Onis M and Bhan MK (2002): Growth performance

of affluent Indian children is similar to that in developed countries. Bull World Health

Organ 80:189-195.

Bhandari N, Bahl R, Nayyar B, Khokhar P. Rohde JE and Bhan MK (2001): Food

supplementation with encouragement to feed it to infants from 4 to 12 months of age

has a small impact on weight gain. J Nutr 131:1946-1951.

Bhargava SK, Sachdev HS, Fall CHD, Osmond C, Lakshmy R, Barker DJP, Biswas

SKD, Ramji S, Prabhakaran D and Reddy KS (2004): Relation of serial changes in

childhood body-mass index to impaired glucose tolerance in young adulthood. N Eng

J Med 350: 865-875.

Bhutta ZA, Black RE, Brown KH, Gardner JM, Gore S, Hidayat A, Khatun F,

Martorell R, Ninh NX, Penny ME, Rosado JL, Roy SK, Ruel M, Sazawal S and

Shankar A (1999): Prevention of diarrhoea and pneumonia by zinc supplementation in

children in developing countries: Pooled analysis of randomized controlled trials. J

Pediatr 135:689-697.

Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis M, Ezzati M, Mathers C and

Rivera J (2008): Maternal and child undernutrition: global and regional exposures and

health consequences. Lancet 317:243-260.

Black MM, Baqui AH, Zaman K, Persson LA, El Arifeen S, Le K, McNary SW,

Parveen M, Hamadani JD and Black RE (2004): Iron and Zinc supplementation

promote motor development and exploratory behavior among Bangladeshi infants.

Am J Clin Nutr 80:903-10.

108

Bogin B and MacVean RB (1983): The relationship of socioeconomic-status and sex

to body size, skeletal maturation, and cognitive status of Guatemala City

schoolchildren. Child Dev 54:115–128.

Borghi E, de Onis M, Garza C, Van den Broeck J, Frongillo EA, Grummer-Strawn L,

van Buuren S, Pan H, Molinari L, Martorell R, Onyango AW and Martines JC for the

WHO Multicentre Growth Reference Study Group (2006): Construction of the World

Health Organization child growth standards: selection of methods for attained growth

curves. Statistics in Medicine 25:247-265.

Britton C, McCormick FM, Renfrew MJ, Wade A and King SE (2007): Support for

breastfeeding mothers. Cochrane Database Sys Rev

1:CD0001141:10.100214651858.CD000141.

Brooker S, Hall A, Bundy DAP, et al. (1999): Short stature and the age of enrolment

in primary school: studies in two African countries. Soc Sci Med 48: 675–682.

Brown KH, Peerson JM, Rivera J and Allen LH (2002): Effect of supplemental zinc

on the growth and serum zinc concentration of pre-pubertal children: a meta-analysis

of randomized controlled trials. Am J Clin Nutr 2002;75:1062-71.

Cameron M (1992): The monitoring of growth and nutritional status in South Africa.

Am J Hum Biol 4:223-234.

Cameron N and Demerath WE (2002): Critical periods in human growth and their

relationship to disease of aging. Am J Phys Anthrop 119 (S35):159-184.

109

Cameron N, Wright MM, Griffiths PL, Norris SA and Pettifor JM (2005): Stunting at

2 years in relation to body composition at 9 years in African urban children. Obes Res

13:131-136.

Cangjiang Y, Sangthong R, Chongsuvivatwog V, McNeil E and Lin L (2009): Effect

of village income and household income on sanitation facilities, hygiene behaviours

and child undernutrition during rapid economic growth in rural cross border area,

Yunnan, China. J Epidemiol Community Health [Epub ahead of print].

Cardoso LO, Vicente AT, Damião JJ and Rito RVVF (2008): The impact of

implementation of the breastfeeding friendly primary care initiative on the prevalence

rates of breastfeeding and causes of consultations at basic healthcare center. J Pediatr

(Rio J) 84:147-153.

Carter JY, Loolpapit MP, Lema OE, Tome JL, Nagelkerke NJ and Watkins WM

(2005): Reduction of the efficacy of antifolate antimalarial therapy by folic acid

supplementation. Am J Trop Med Hyg 73:166-70.

Caulfield LE, de Onis M, Blössner M and Black RE (2004): Undernutrition as an

underlying cause of child death associated with diarrhoea, pneumonia, malaria, and

measles. Am J Clin Nutr 80:193-198.

Caulfield LE, Richard SA and Black RE (2004): Undernutrition as an underlying

cause of malaria morbidity and mortality in children less than five years old Am J

Trop Med Hyg Clin 71:55-63.

Ciliberto MA, Sandige H, Ndekha MJ, Ashorn P, Briend A, Ciliberto HM and

Manary MJ (2005): A comparison of home-based therapy with ready-to-use

110

therapeutic food with standard therapy in the treatment of malnourished Malawian

children: a controlled, clinical effectiveness trial. Am J Clin Nutr 81:864-870.

Cheung YB, Yip PSF and Karlberg JPE (2001): Fetal growth, early postnatal growth

and motor development in Pakistani infants. Int J Epidemiol 2001;30:518-26.

Cheung YB, Low L, Osmond C, Barker D and Karlberg J (2000): Fetal growth and

early postnatal growth are related to blood pressure in adults. Hypertension 36:795-

8000.

Cheung YB, Gladstone M, Maleta K, Duan X and Ashorn P (2008): Comparison of

four approaches to score development: a study of Malawian children. Trop Med Int

Health 13:987-93

Clarke N, Grantham-McGregor S and Powell C (1990): Health and nutrition

predictors of school failure in Kingston, Jamaica. Ecol Food Nutr 26: 1–11.

Coley AN, Milet J, Diallo A, Ndiaye T, Benefice E, Simondon F, Wade S and

Simondon KB (2006): Preschool stunting, adolescent migration catch-up growth, and

adult height in young Senegalese men and women of rural origin. J Nutr 136:2412-

2420.

Collins S, Dent N, Binns P, Bahwere P, Sadler K and Hallam A (2006): Management

of severe acute malnutrition in children. Lancet 368:1992-2000.

Coovadia HM Rollins NC, Bland RM, Little K, Coutsoudis A, Bennish M and Newell

ML (2007): Mother-to-child transmission if HIV-1 infection during exclusive

breastfeeding in the first 6 months of life: an intervention cohort study. Lancet

369:1107-1116.

111

Crane NT, Wilson DB, Cook DA, Lewis CJ, Yetley EA and Rader JI (1995):

Evaluating food fortification options: general principles revisited with folic acid. Am

J Public Health 85: 660-666.

Cueto S. (2005): Height, weight, and education achievement in rural Peru. Food Nutr

Bull 26 (Suppl 2): S251–260.

Cutts FT (1998): Advances and challenges fro the expanded programme on

immunization. Br Med Bull 54:445-461.

Czerwinski SA, Lee M, Choh AC, Wurzbacher K, Demerath EW, Towne B and

Siervogel RM (2007): Genetic factors in physical and development and their

relationship to subsequent health outcomes. Am J Hum Biol 19:684-691.

Daniels DL, Cousens SN, Makoae LN and Feachem RG (1991): A case-control study

of the impact of improved sanitation on diarrhoea morbidity in Lesotho. Bull World

Health Organ 68:455-463.

Daniels MC and Adair LS (2004): Growth in young Filipino children predicts

schooling trajectories through high school. J Nutr 134:1439-46.

de Onis M and Blössner M (2003): The World Health Organization Global Database

on Child Growth and Malnutrition: methodology and applications. Int J Epidemiol

32:518-26.

de Onis M, Blössner M, Borghi E, Morris R and Frongillo EA (2004a) Methodology

for estimating regional and global trends of child malnutrition. Int J Epidemiol

33:1260-1270.

112

de Onis M and Habicht JP (1996): Anthropometric reference data for international

use: recommendations from a World Health Organization Expert Committee. Am J

Clin Nutr 64:650–658.

de Onis M, Garza C, Victora CG, Onyango AW, Frongillo EA and Martines J for the

WHO Multicentre Growth Reference Study Group (2004b): The WHO Multicentre

Growth Reference Study: planning, study design and methodology. Food Nutr Bull

25(Suppl 2):S15–S26.

de Onis, Garza C and Habicht JP (1997): Time for a new growth reference.

Paediatrics 100(5): e8.

Defourny I., Seroux G., Abdelkader I. & Harczi G. (2007) Management of moderate

acute malnutrition with RUTF in Niger. Emergency Nutrition Network 31:2-4.

Demerath EW, Choh AC, Czerwinski SA, Lee M, Sun SS, Chumlea WMC, Duren D,

Sherwood RJ, Blangero J, Towne B and Siervogel RM (2007): Genetic and

environmental influences on infant weight and weight change: The Fels Longitudinal

Study. Am J Hum Biol 19:692-702.

Desai MR, Mei JV, Kariuki SK, Wannemuehler KA, Phillips-Howard PA, Nahlen

BL, Kager PA, Vulule JM and ter Kuile FO (2003): Randomized, controlled trial of

trial of daily iron supplementation and intermittent sulfadoxine-pyrimethamine for the

treatment of mild childhood anemia in Kenya. J Infect Dis 658-666.

Dewey KG (1998): Cross cultural patterns of growth and nutritional status of breast-

fed infants. Am J Clin Nutr 67: 10-17.

113

Dewey KG (2007): Increasing iron intake of children through complementary foods.

Food Nutr Bull 28:S595-S609.

Dewey KG and Adu-Afarwuah S (2008): Systematic review of the efficacy and

effectiveness of complementary feeding interventions in developing countries. Matern

Child Nutr 1:24-85.

Dewey KG (2001): The challenges of promoting optimal infants growth. J Nutr

131:1879-1880.

Dibley MJ, Goldsby JB, Staehling NW and Trowbridge FL (1987): Development of

normalized curves for the international growth reference: historical and technical

considerations. Am J Clin Nutr 46:736–748.

Dubois L, Girard M, Girard A, Tremblay R, Boivin M and Pérusse D (2007): Genetic

and environmental influences on body size in early childhood: A twin birth-cohort

study. Twin Res Hum Genet 10(3):479-485.

Dzinjalamala FK, Macheso A, Kublin JG, Taylor TE, Barnes KI, Molyneux ME,

Plowe CV and Smith PJ (2005): Blood folate concentrations and in vivo sulfadoxine-

pyrimethamine failure in Malawian children with uncomplicated plasmodium

falciparum malaria. Am J Trop Med Hyg 72:267-272.

Fewtrell MS (2004): The long –term benefits of having been breast-fed. Current

Paediatrics 14:97-103.

Food and Nutrition Board (FNB), Institute of Medicine, National Academies (2005):

Dietary reference intakes for energy, carbohydrate, fiber, fat, fatty acids, cholesterol,

114

protein, and amino acids (Macronutrients). [cited 2009 Mar 15]. Available from:

http://books.nap.edu/openbook.php?record_id=10490&page=1324

Fotso JC and Kaute-Defo B (2005): Socio-economic inequalities in early childhood

malnutrition and morbidity: modification of the household-level effects by the

community SES. Health & Place 11:205-225.

Fowden AL (1995): Endocrine regulation of fetal growth. Reprod Fertil Dev

7(3):351-363.

Fowden AL (2003): The insulin-like growth factors and feto-placental growth.

Placenta 24(8-9):803-812.

Gamble C, Ekwaru PJ, Garner P and ter Kuile FO (2007): Insecticide-treated nets for

the prevention of malaria in pregnancy: A systematic review of randomised controlled

trials. PLoS Med 4:e107.doi:10.137/journaLpmed.0040107.

Gardner JM, Walker SP, Powell CA and Grantham-McGregor S (2003): A

randomized controlled trial of a home-visiting intervention on cognition and behavior

in term low birth weight infants. J Pediatr 143:634–639.

Garg A and Kashyap S (2006): Effect of counselling on nutritional status during

pregnancy. Indian J Pediatr 73:687-692.

Garner P and Gulmezoglu AM (2006): Drugs for preventing malaria in pregnant

women. Cochrane Database of Syst Rev

4:CD000169.DOI:10.1002/14651858.CD000169.

115

Garza C and de Onis M for the WHO Multicentre Growth Reference Study Group

(2004): Rationale for developing a new international growth reference. Food and

Nutrition Bulletin 25(Suppl 2):S5–S14.

Garza C and de Onis M (1999): A new international growth reference for young

children. Am J Clin Nutr 70(suppl.):S169-S172.

Gau ML (2007): Evaluation of lactation intervention program to encourage

breastfeeding: a longitudinal study. Int J Nurs Stud 42:425-435.

Gera T and Sachdev HP (2002): Effect of iron supplementation on incidence of

infectious illness in children: systematic review. BMJ 325:1142.

Glewwe P, Jacoby HG and King EM (2001): Early childhood nutrition and academic

achievement: a longitudinal analysis. J Public Econ 81:345–368.

Glewwe P and Jacoby H (1995): An economic analysis of delayed primary school

enrollment in a low income country: the role of early childhood nutrition. Rev Econ

Stat 77:156–169.

Gonzalez-Cossio T, Habicht JP, Rasmussen KM and Delgado HL (1998) Impact of

food supplementation during lactation on infant breast-milk intake and on the

proportion of infants exclusively breast-fed. J Nutr 128:1692-1702.

Gorman KS and Pollitt E (1992): Relationship between weight and body

proportionality at birth, growth during the first year of life, and cognitive development

at 36, 48, and 60 months. Infant Behav Dev 15:279–296.

116

Graham GG, Morales E, Acevedo G, Placko RP, Cordano A (1971): Dietary protein

quality in infants and children. IV. A corn-soy-milk blend. Am J Clin Nutr 24:416-

422.

Grantham-McGregor S, Cheung YB, Cueto S, Glewwe P, Richter L and Strupp B

(2007): Developmental potential in the first 5 years for children in developing

countries. Lancet 369:60-70.

Grantham-McGregor SM, Lira PIC, Ashworth A, Morris SS and Assuncao AMS

(1998): The development of low birth weight term infants and effect of the

environment in Northeast Brazil. J Pediatr 132:661-666.

Griffiths P, Madise N, Whitworth A and Mathews Z (2004): A tale of two continents:

a multilevel comparison of determinants of child nutritional status from selected

African and Indian regions. Health & Place 10:183-199.

Griffiths R and Huntley M (1996): GMDS 0-2. Griffiths Mental Development Scales -

Revised: Birth to 2 years. Oxford: Hogrefe.

Guillermo A, Mejia LA and Aguilar JR (1981): The effect of vitamin A fortification

of sugar on serum vitamin A levels of preschool Guatemalan children: a longitudinal

evaluation. Am J Clin Nutr 34:41-49.

Haas JD, Martinez EJ, Murdoch S, Conlisk E, Rivera JA and Martorell R (1995):

Nutritional supplementation during the preschool years and physical work capacity in

adolescent and young adult Guatemalans. J Nutr 125:S1078-89.

Harahap H, Jahari AB, Husaini MA, Saco-Pollitt C and Pollitt E (2000): Effects of an

energy and micronutrient supplement in iron deficiency anemia, physical activity and

117

motor and mental development in undernourished children in Indonesia Eur J Clin

Nutr 54 Suppl 2:S114-S119.

Hatt LE and Waters HR (2006): Determinants of child morbidity in Latin America: A

pooled analysis of interactions between parental education and economic status. Soc

Sci Med 62:375-376.

Helland IB, Saugstad OD, Saarem K, Van Houwelingen AC, Nylander G and Drevon

CA (2006): Supplementation of n-3 fatty acids during pregnancy and lactation reduces

maternal plasma lipid levels and provides DHA to the infants. J Matern Fetal

Neonatal Med 62:316-322.

Helland IB, Smith L, Saarem K, Saugstad OD and Drevon CA (2003): Maternal

supplementation with very-long-chain n.3 fatty acids during pregnancy and lactation

augments children’s IQ at 4 years of age. Pediatrics 111:e39-44.

Henneberg M and Berg E R (1990): Test of socioeconomic causation of secular trend:

stature changes among favored and oppressed South Africans are parallel. Am J Phys

Anthrop 83:459–465.

Homedes N (1996): The disability-adjusted life year (DALY) definition, measurement

and potential use. Human Capital Working Paper: Report number 16128.

Hoppe C, Molgaard C and Michaelsen KF (2006): Cow’s milk and linear growth in

industrialized and developing countries. Ann Rev Nutr 26:131-173.

Horton S, Alderman H and Rivera JA (2008): Copenhagen consensus 2008 challenge

paper hunger and malnutrition. Copenhagen Consensus Center.

118

Husaini MA, Karyadi L, Husaini YK, Sandjaja, Karyadi D and Pollitt E (1991):

Developmental effects of short-term supplementary feeding in nutritionally-at-risk

Indonesian infants. Am J Clin Nutr 54:799-804.

Hutchinson SE, Powell CA, Walker SP, Chang SM and Grantham-McGregor SM

(1997): Nutrition, anaemia, geohelminth infection and school achievement in rural

Jamaican primary school children. Eur J Clin Nutr 51:729–735.

Huxley RR, Shiell AW and Law CM (2000): The role of size at birth and postnatal

catch-up growth in determining systolic blood pressure: a systematic review of

literature. J Hypertens 18:815-831.

Iannotti LL, Tielsch JM, Black MM and Black RE (2006): Iron supplementation in

childhood: health benefits and risks. Am J Clin Nutr 84:1261-76.

Iliff PJ, Piwoz EG, Tavengwa NV, Zunguza CD, Marinda ET, Nathoo KJ, Moulton

LH, Ward BJ, the ZVITAMBO study group and Humphrey JH (2005): Early

exclusive breastfeeding reduces the risk of postnatal HIV-1 transmission and

increased HIV-free survival. AIDS 19:699-708.

Innis SM (2007): Dietary (n-3) fatty acids and brain development. J Nutr 137:855-

859.

Innis SM (2008): Dietary omega fatty acids and developing brain. Brain Res 1237:35-

43.

Innis SM (2009): Omenga-3 fatty acids and neural development to 2 years of age: Do

we know enough fro dietary recommendations? J Pediatr Gastroenterol Nutr : 48:S16-

24.

119

Islam M , Peerson JM, Ahmed T, Dewey KG and Brown KH (2006): Effects of varied

energy density of complementary foods on breast-milk intakes and energy

consumption by healthy, breastfed Bangladeshi children. Am J Clin Nutr 83:851-858.

Ivanans I (1975): Effect of maternal education and ethnic background on infant

development. Arch Dis Child 5:454-457.

Ivanovic DM, Perez HT, Olivares MD, Diaz NS, Leyton BD and Ivanovic RM

(2004): Scholastic achievement: a multivariate analysis of nutritional, intellectual,

socioeconomic, sociocultural, familial, and demographic variables in Chilean school-

aged children. Nutrition 20:878–889.

Jamison DT (1986): Child malnutrition and school performance in China. J Dev Econ

20:299–309.

Johnson S and Marlow N (2006): Developmental screening or developmental testing?

Early Hum Dev 82:173-183.

Johnston FE, Low SM, de Baessa Y and MacVean RB (1987): Interaction of

nutritional and socioeconomic status as determinants of cognitive development in

disadvantaged urban Guatemalan children. Am J Phys Anthropol 73:501–506.

Jones G Steketee RW, Black RE, Bhutta ZA and Morris SS, Bellagio Child Survival

Study Group (2003): How many child deaths can we prevent this year? Lancet 362:

65-71.

Kagaayi J, Gray RH, Brahmbhatt H, Kigozi G, Nalugoda F, Wabwire-Mangen F,

Serwadda D, Sewankambo N, Ddungu V, Ssebagala D, Sekasanvu J, Kigozi G,

Makumbi F, Kiwanuka N, Lutalo T, Reynolds SJ and Wawer MJ (2008): Survival of

120

infants born to HIV-positive mothers, by feeding modality, in Rakai, Uganda. PLoS

ONE 3:e3877.

Kaku K, Osada H, Seki K and Sekiya S (2007): Insulin-like growth factors 2(IGF2)

and IGF2 receptor gene variants are associated with fetal growth. Acta Paediatr

96(3):363-367.

Kariger PK, Stoltzfus RJ, Olney D, Sazawal S, Black R, Tielsch JM, Frongillo EA,

Khalfan SS and Pollitt E (2005): Iron deficiency and physical growth predict

attainment of walking but not crawling in poorly nourished Zanzibari Infants. J Nutr

135:814-19.

Karlberg J (1989): A biologically-oriented mathematical model (ICP) for human

growth. Acta Paediatr Scand Suppl 350:70-94.

Karlberg J, Jalil F, Lam B, Low L and Yeung CY (1994): Linear growth retardation in

relation to the three phases of growth. Eur J Clin Nutr 48(Suppl 1):S25-S44.

Kimmons JE, Dewey KG, Hague E, Chakraborty J, Osendarp SJ and Brown KH

(2005): Low nutrient intakes among infants in rural Bangladesh are attributable to low

intake and micronutrient density of complementary foods. J. Nutr 135:444-451.

Kim-Farley R (1992): Global immunization. Annu Rev Public Health 13:223-237.

Kirwan P, Asaolu SO, Molloy SF, Abiona TC, Jackson AL and Holland CV (2009):

Patterns of soil-transmitted helminth infection and impact of four-monthly

albendazole treatments in preschool children from semi-urban communities in

Nigeria: a double-blind placebo-controlled randomised trial. BMC Infect Dis 9:20.

121

Kramer MS and Kakuma R (2003): Energy and protein intake in pregnancy (Review).

Cochrane Database Sys Rev 4:CD000032: 10.1002/14651858. CD000032.

Kramer MS and Kakuma R (2002): Optimal duration of exclusive breast feeding

(Review). Cochrane Database Sys Rev 1: CD003517:10.1002/14651858.CD003517.

Kramer MS (1993): Effects of energy and protein intakes on pregnancy outcome: an

overview of the research evidence from controlled clinical trials. Am J Clin Nutr

58:627-635.

Krause VM, Delisle H and Solomons NW (1998): Fortified foods contribute one half

of recommended vitamin A intake in poor urban Guatemalan toddlers. J Nutr

128:860-864.

Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z, Wei

R, Curtin LR, Roche AF and Johnson CL (2002): 2000 CDC growth charts for the

United States: Methods and development. National Center for Health Statistics. Vital

Health Stat 11. (246):1-190.

Kuusipalo H, Maleta K, Briend A, Manary M and Ashorn P (2006): Growth and

change in blood haemoglobin concentration among underweight Malawian infants

receiving fortified spreads for 12 weeks: A preliminary trial. J Pediatr Gastroenterol

Nutr 43:525-532.

Lampl M, Veldhuis JD and Johnson ML (1992): Saltation and stasis: A model of

human growth. Science 258(5083):801-803.

Larrea C and Kawachi I (2005): Does economic inequality affect child malnutrition?

The case of Ecuador. Soc Sci Med 60:165-178.

122

Lartey A, Manu A, Brown KH, Peerson JM and Dewey KG (1999): A randomized,

community-based trial of the effects of improved, centrally processed complementary

foods on growth and micronutrient status of Ghanaian infants from 6 to 12 mo of age.

Am J Clin Nutr 70:391-404.

Lauritzen L, Hoppe C, Straarup EM and Michaelsen KF (2005): Maternal fish oil

supplementation in lactation and growth during the first 2.5 years of life. Pediatr Res

58:235-242.

Lauritzen L, Jørgensen MH, Mikkelsen TB, Skovgaard M, Olsen SF, Hoy CE,

Straarup EM and Michaelsen KF (2004): Maternal fish oil supplementation in

lactation: effect on visual acuity and n-3 fatty acid content of infant erythrocytes.

Lipids 39:195-206.

Lazoff B (2007): Iron deficiency and child development. Food Nutr Bull 28:S560-

S571.

Lee PA, Kendig JW and Kerrigan JR (2003): Persistent short stature, other potential

outcomes and the effect of growth hormone treatment in children who are born small

for gestational age. Pediatrics 112: 150-162.

Lengeler C (2004): Insecticide-treated bed net and curtains for preventing malaria

(Review). Cochrane Database of Syst Rev

2:CD000363.DOI.1002/14651858.CD000363.

Levitt NS, Lambert EV, Woods D, Hales CK, Andrew R and Seckl JR (2000):

Impaired glucose tolerance and elevated blood pressure in low birth weight, nonobese,

123

young South African Adults: Early programming of cortisol axis. J Clin Endocrinol

Metab 85:4611-4618.

Li H, DiGirolamo AM, Barnhart HX, Stein AD and Martorell R (2004): Relative

importance of birth size and postnatal growth for women’s educational achievement.

Early Hum Dev 76:1–16.

Li H, Barnhart HX, Stein AD and Martorell R (2003a): Effects of early childhood

supplementation n educational achievement of women. Pediatrics 112:1156-1162.

Li H, Stain AD, Barnhart HX, Ramakrishnan U and Martorell R (2003b):

Associations between prenatal and postnatal growth and adult body size and

composition. Am J Clin Nutr 77:1498-1505.

Lin CA, Manary MA, Maleta K, Briend A and Ashorn P (2008): An Energy-dense

complementary food is associated with a modest increase in weight gain when

compared with a fortified porridge in Malawian Children Aged 6-18 months. J Nutr

138:593-598.

Lindsay HA and Haskell M (2002): Estimating the potential for vitamin A toxicity in

women and young children. J Nutr 132:2907S-2919S.

Liu X Y, Jalil F and Karlberg J (1998): Risk factors for impaired length growth in

early life viewed in terms of the infancy–childhood–puberty (ICP) growth model.

Acta Paediatr 87:237–243.

Liu Y, Albertsson-Wikland K and Karlberg J (2000): Long-Term Consequences of

early linear growth retardation (stunting) in Swedish children. Pediatr Res

2000;47:475-480.

124

Lucas A, Fewtrell MS, Cole TJ (1999): Fetal origin of adult disease-the hypothesis

revisited. BMJ 319:245-249.

Lutter CK and Rivera JA (2003): Nutritional status of infants and young children and

characteristic of their diets. J Nutr 133:2941S-2949S.

Lutter CK (2000): Processed complementary foods: Summary of nutritional

characteristics, methods of production and distribution, and costs. Food Nutr Bull

21:95-100.

Lynch S, Stoltzfus R and Rabat R (2007): Critical review of strategies to prevent and

control iron deficiency in children. Food Nutr Bull 28:S610-S620.

Maleta K, Virtanen S, Espo M, Kulmala T and Ashorn P (2003a): Timing of growth

faltering in rural Malawi. Arch Dis Child 88:574-578.

Maleta K, Virtanen SM, Espo M, Kulmala T and Ashorn P (2003b): Seasonality of

growth and the relationship between weight and height gain in children under three

years of age in rural Malawi. Acta Paediatr 92:491-497.

Maleta K, Virtanen S, Espo M, Kulmala T and Ashorn P (2003c): Child malnutrition

and its predictors in Rural Malawi. Paediatr Perinat Epidemiol 17:384-390.

Manary MJ, Ndekha MJ, Ashorn P, Maleta K and Briend A (2004): Home based

therapy for sever malnutrition with ready-use food. Arch Dis Child 89:557-561.

Manson JB, Deitchler M, Gilman A, Gillenwater K, Shuaib M, Hotchkiss D, Mason

K, Mock N and Sethuraman K (2002): Iodine fortification is related to increased

weight-for age and birth weight in children in Asia. Food Nutr Bull 23:292-308.

125

Marchione TJ (2002): Foods provided through U.S. Government Emergency Food

Aid Programs: Policies and customs governing their formulation, selection and

distribution. J Nutr 132:2104S-2111S.

Marti A, Moreno-Aliaga MJ, Hebebrand J and Martínez JA (2004): Genes, lifestyles

and obesity. Int J Obes Relat Metab Disord 28(Suppl 3):S29-36.

Martorell R, Khan LK and Schroeder DG (1994): Reversibility of stunting:

epidemiological findings in children from developing countries. Eur J Clin Nutr 48

(Suppl. 1):S45-57.

Martorell R (1995a): Results and implications of the INCAP follow-up study. J Nutr

125:1127S-1138S.

Martorell R, Schroeder DG, Rivera JA and Kaplowitz HJ (1995b): Pattern of linear

growth in rural Guatemala adolescents and children. J Nutr 125:S1060-1067.

McDonald EC, Pollitt E, Mueller W, Hsueh AM and Sherwin R (1981): The Bacon

Chow study: maternal nutrition supplementation and birth weight of offspring. Am J

Clin Nutr 34:2133-2144.

Meeks Gardner JM, Grantham-McGregor S, Himes J and Chang S (1999): Behaviour

and development of stunted and nonstunted Jamaican children. J Child Psychol

Psychiatry 40:817-27.

Mejia LA and Arroyave G (1982): The effect of vitamin A fortification of sugar on

iron metabolism in preschool children in Guatemala. Am J Clin Nutr 36:87-93.

126

Melse-Boonstra A, Rozendaal M, Rexwinkel H, Gerichhausen MJW, van den Briel T,

Bulux J, Solomons NW and West CE (1998): Determination of discretionary salt

intake in rural Guatemala and Benin to determine the iodine fortification of salt

required to control iodine deficiency disorders: studies using lithium-labeled salt. Am

J Clin Nutr 68:636-641.

Mendez MA and Adair LS (1999): Severity and timing of stunting in the first two

years of life affect performance on cognitive tests in late childhood. J Nutr 125:1555-

62.

Meremikwu MM, Donegan S and Esu E (2008): Chemoprophylaxis and intermittent

treatment for preventing malaria in children. Cochrane Database Syst Rev

2:CDoo3756.DOI:10.1002/146551858.CD003756.

Messina MJ (1999): Legumes and soybeans: overview of their nutritional profiles and

health effects. Am J Clin Nutr 70:430S-450S.

Metz J (2007): Folic acid metabolism and malaria. Food Nutr Bull 28:S540-549.

Mills JL, Kohorn IV, Conley MR, Zeller JA, Cox C, Williamson RE and Dufour DR

(2003): Low vitamin B-12 concentrations inpatient without anemia: the effect of folic

acid fortification of grain. Am J Clin Nutr 77:1474-1477.

Matilsky DK, Maleta K, Castleman T, and Manary MJ (2009): Supplementary

feeding with fortified spreads results in higher recovery than with a corn-soy blend in

moderately wasted children. J Nutr 139:773-778.

127

Miotti PG, Taha TET, Kumwenda NI, Broadhead R, Mtimavyale LAR, Van der

Hoeven L, Chiphangwi JD, Liomba G and Biggar RJ (1999): HIV transmission

through breastfeeding: A study from Malawi. JAMA 282:744-749.

Moffat MEK, Longstaffe S, Besant J and Dureski C (1994): Prevention of iron

deficiency and psychomotor decline in high-risk infants through use of iron-fortified

infants formula: a randomized clinical trial. J Pediatr 125:527-534.

Monteiro CA, Benicio MH, Konno SC, da Silva AC, de Lima AL and Conde WL

(2009): Causes for the decline under-nutrition in Brazil, 1996-2007. Rev Saude

Publica 43:35-43.

Moock PR and Leslie J (1986): Childhood malnutrition and schooling in the Terai

region of Nepal. J Dev Econ 20:33–52.

Msyamboza K, Senga E, Tetteh-Ashong E, Kazembe E and Brabin BJ (2007):

Estimation of effectiveness of interventions for malaria control in pregnancy using the

screening method. Int J Epidemiol 36:406-411.

Muhilal, Permeisih D, Idjradinata YR, Muherdiyantiningsih and Karyadi D (1988):

Vitamin A-fortified monosodium glutamate and health, growth, and survival of

children: a controlled field trial. Am J Clin Nutr 48:1271-1276.

Mulder DW, Nunn A, Kamali A and Kengeya-Kayondo JF (1996): Post-natal

incidence of HIV-1 infection among children in a rural Ugandan population: no

evidence other than mother to child. Trop Med Int Health 1:81-85.

128

Mulenga M, Malunga P, Bennett S, Thuma P, Shulman C, Fielding K and Greenwood

B (2006): Folic acid treatment of Zambian children with moderate to severe malaria

anemia. Am J Trop Med Hyg 74:986-990.

Ndekha MJ, Manary MJ, Ashorn P and Briend A (2005) Home-based therapy with

ready-to-use food is of benefit to malnourished, HIV-infected Malawian children.

Acta Paediatr 94:222-225.

Nduati R, John G, Mbori-Ngacha D, Richardson B, Overbaugh J, Mwatha A, Ndinya-

Achola J, Bwayo J, Onyango FE, Hughes J and Kreiss J (2000): Effect of

breastfeeding and formula feeding on transmission of HIV-1: A randomized clinical

trial. JAMA 283:1167-1174.

Nestel P, Briend A, de Benoist B, Decker E, Ferguson E, Fontaine O, Micardi A,

Nalubola R (2003): J Pediatr Gastroenterol Nutr 36:316-328.

Oelofse A, Van Raaij JM, Benade AJ, Dhansay MA, Tolboom JJ and Hautvast JG

(2003) The effect of a micronutrient-fortified complementary food on micronutrient

status, growth and development of 6 to 12-month-old disadvantaged urban South

African infants. Int J Food Sci Nutr 54:399-407.

Olney DK, Pollitt E, Kariger PK, Khalfan SS, Ali NS, Tielsch JM, Sazawal S, Black

R, Mast D, Allen LH and Stoltzfus RJ (2007): Young Zanzibar children with iron

deficiency, iron deficiency anemia, stunting, or malaria have lower motor activity

scores and spend less time in locomotion. J Nutr 137:2756-2762.

Olney DK, Pollitt E, Kariger PK, Khalfan SS, Ali NS, Tielsch JM, Sazawal S, Black

R. Allen LH and Stoltzfus RJ (2006): Combined iron and folic acid supplementation

129

with or without zinc reduces time to walking unassisted among Zanzibar infants 5 to

11 mo old. J Nutr 136:2427-2434.

Ong KK (2006): Size at birth, postnatal growth and risk of obesity. Horm Res

65(S):65-69.

Ong KK and Dunger DB (2004): Birth weight, infant growth and insulin resistance.

Eur J Endocrinol 151:U131-139.

Oppenheimer SJ (2001): Iron and its relation to immunity and infectious disease. J

Nutr 131:616S-635S.

Owens JA (1991): Endocrine and substrate control of fetal growth: Placental and

maternal influences and insulin-like growth factors. Reprod Fertil Dev 3(5):501-517.

Özmert EN, Yurdakök K, Soysal S, Kulak-Kayikci ME, Belgin E, Özmert E, Laleli Y

and Saracbasi O (2005): Relationship between physical, environmental and

sociodemographic factors and school performance in primary schoolchildren. J Trop

Pediatr 51:25–32.

Pant CR, Pokharel GP, Curtale F, Pokhrel RP, Grosse RN, Lepkowski J, Muhilal,

Bannister M, Gorstein J, Pak-Gorstein S, Atmarita and Tilden RL (1996): Impact of

nutrition education and mega-dose vitamin A supplementation on the health of

children in Nepal Bull World Health Organ 74:533-545.

Partnership for Child Development (2001): An association between chronic

undernutrition and educational test scores in Vietnamese children. Eur J Clin Nutr

55:801–804.

130

Patel MP, Sandige HL, Ndekha MJ, Briend A, Ashorn P and Manary MJ (2005):

Supplemental feeding with ready-to-use therapeutic food in Malawian children at risk

of malnutrition. J Health Popul Nutr 23:351-357.

Pelletier DL, Frongillo EA and Habicht JP (1993): Epidemiological evidence for

potentiating effects of malnutrition on child mortality. Am J Public health 83:1130-

1133.

Pelletier DL (1994): The relationship between child anthropometric and mortality in

developing countries: Implications for policy, programs and future research. J Nutr

124:2047S-2981S.

Pena R, Wall S and Persson LA (2000): The effect of poverty, social inequality and

maternal education on infant mortality in Nicaragua, 1988-1993. Am J Public Health

90:64-69.

Preece MA and Baines MJ (1978): A new family of mathematical models describing

the human growth curve. Ann Hum Biol 5:1-24.

Prentice AM, Ghattas Doherty C and Cox SE (2007): Iron metabolism and malaria.

Food Nutr Bull 28:S524-S539.

Ramakrishnan U, Martorell R, Schroeder DG and Flores R (1999) Role id

intergenerational effects in linear growth. J Nutr 129(Suppl): 544S-549S.

Rigby RA and Stasinopoulos DM (2004): Smooth centile curves for skew and kurtotic

data modelled using the Box-Cox power exponential distribution. Statistics in

Medicine 23:3053–3076.

131

Rigby RA and Stasinopoulos DM (2005): Generalized additive models for location,

scale and shape. Journal of the Royal Statistical Society - Series C - Applied Statistics

54:507–544.

Rivera JA, Sotres-Alvarez D, Habicht JP, Shamah T and Villalpando S (2004): Impact

of the Mexican program for education, health, and nutrition (Progresa) on rates of

growth anemia in infants and young children. JAMA 291:2563-2570.

Rosa FW, Wilk AL and Kelsey FO (1986): Teratogen update: Vitamin A congeners.

Teratology 33:355-364.

Ross DA (2002): Recommendations for vitamin A supplementation. J Nutr

132:2902S-2906S.

Roy SK, Jolly SP, Shafique S, Fuchs GJ, Mahmud Z, Chakraborty B and Roy S

(2008): Prevention of malnutrition among young children in rural Bangladesh by a

food-health-care educational intervention: A randomized, controlled trial. Food Nutr

Bull 29:67-75.

Rothman KJ, Moore LL, Singer MR, Nguyen UDT, Mannino S and Milunsky A

(1995): Teratogenicity of high vitamin A intake. N Engl J Med 333:1369-73.

Rush D (1994): Periconceptional folate and neural tubes defect. Am J Clin Nutr

59:511S-516S.

Sandige H, Ndekha MJ, Briend A, Ashorn P and Manary MJ (2004): Home-based

treatment of malnourished Malawian children with locally produced or imported

ready-to-use food. J Pediatr Gastroenterol Nutr 39:141-146.

132

Sazawal S, Black RE, Ramsan M, Chwaya HM, Stolzfus RJ, Dutta A, Dhingra U,

Kabole I, Deb S, Othman MK and Kabole FM (2006): Effects of routine prophylactic

supplementation with iron and folic acid on admission to hospital and mortality in

preschool children in a high malaria transmission setting: community-based,

randomized, placebo-controlled trial. Lancet 367:133-143.

Schousboe K, Visscher PM, Erbas B, Kyvik KO, Hopper JL, Henriksen JE, Heitmann

BL and Sørensen TIA (2004): Twin study of genetic and environmental influences on

adult body size, shape and composition. Int J Obes 28(1):39-48.

Scrimshaw NS (1977): Effect of infection on nutrient requirements. Am J Clin Nutr

30:1536-1544.

Shariff ZM, Bond JT and Johnson NE (2000): Nutrition and educational achievement

of urban primary schoolchildren in Malaysia. Asia Pacific J Clin Nutr 9:264–273.

Shrimpton R, Victora CG, de Onis M, Lima RC, Blössner M and Clugston G (2001):

Worldwide timing of growth faltering: implications for nutritional interventions.

Pediatrics 107: E75.

Siegel EM, Stoltzfus RJ, Kariger PK, Katz J, Khatry SK, LeClerq SC, Pollitt E and

Tielsch JM (2005): Growth indices, anemia and diet independently predict motor

milestone acquisition of infants in South Central Nepal. J Nutr 135:2840-2844.

Sigman M, Neumann C, Jansen AA and Bwibo N (1989): Cognitive abilities of

Kenyan children in relation to nutrition, family characteristics, and education. Child

Dev 60:1463–1474.

133

Silventoinen K, Pietiläinen KH, Tynelius P, Sørensen TIA, Kaprio J and Rasmussen F

(2008a): Genetic regulation of growth from birth to 18 years of age: The Swedish

young male twins study. Am J Hum Biol 20(3):292-298.

Silventoinen K, Sammalisto S, Perola M, Boomsma DI, Cornes BK, Davis C, Dunkel

L, de Lange M, Harris JR, Hjelmborg JVB, Luciano M, Martin NG, Mortensen J,

Nisticò L, Pedersen NL, Skytthe A, Spector TD, Stazi MA, Willemsen G and Kaprio J

(2003a): Heritability of adult body height: A comparative study of twin cohorts in

eight countries. Twin Res 6(5):399-408.

Silventoinen K (2003b): Determinants of variation in adult body weight. J Biosoc Sci

35(2):263-285.

Silventoinen K, Magnusson PKE, Tynelius P, Kaprio J and Rasmussen F (2008b):

Heritability of body size and muscle strength in young adulthood: A study of one

million Swedish men. Genet Epidemiol 32(4):341-349.

Silventoinen K, Pietiläinen KH, Tynelius P, Sørensen TIA, Kaprio J and Rasmussen F

(2007): Genetic and environmental factors in relative weight from birth to age 18: The

Swedish young male twins study. Int J Obes 31(4):615-621.

Simondon KB, Gartner A, Berger J, Cornu A, Massamba JP, Miguel JLS, Ly C,

Missotte I, Simondon F, Traissac P, Delpeuch F and Maire B (1996): Effect of early,

short-term supplementation on weight and linear growth of 4-7-mo-old infants in

developing countries: A four-country randomized trial. Am J Clin Nutr 64:537-545.

Singh M (2005): Essential fatty acids, DHA and human brain. Indian J Pediatr

72:239-242.

134

Smith AD, Kim YI and Refsum H (2008): Is folic acid good for everyone? Am J Clin

Nutr 87:517-533.

Snow RW, Guerra CA, Noor AM, Myint HY and Hay SI (2005): The global

distribution of clinical episodes of plasmodium falciparum malaria. Nature 434:214-

217.

Steegmann AT, Datar FA and Steegmann RM (1992): Physical size, school

performance and visual-motor maturity in the Philippines. Am J Hum Biol 4:247–252.

Stein CE, Fall CHD, Kumaran K, Osmond C, Cox V and Barker DJP (1996): Fetal

growth and coronary heart disease in South India. Lancet 348:1269-1273.

Stein AD, Wang M, Ramirez-Zea M, Flores R, Grajeda R, Melgar P, Ramakrishnan U

and Martorell R (2006): Exposure to nutrition supplementation intervention in early

childhood and risk factors for cardiovascular in Adulthood: Evidence from

Guatemala. Am J Epidemiol 164:1160-1170.

Stein AD, Conlisk A, Torun B, Schroeder DG, Grajeda R and Martorell R (2002):

Cardiovascular disease risk factors of are related to adult adiposity but no birth weight

in young Guatemalan adults. J Nutr 132:2208-2214.

Stein AD , Barnhart HX, Hickey M, Ramakrishnan U, Schroeder DG and Martorell R

(2003): Prospective study of protein-energy supplementation early in Life and growth

in the subsequent generation in Guatemala. Am J Clin Nutr 78:162-167.

Stoltzfus RJ, Chway HM, Montresor A, Tielsch JM, Jape JK, Albonico M and Savioli

L (2004): Low dose daily supplementation improves iron status and appetite but not

135

anemia, whereas quarterly anthelminthic treatment improves growth, appetite and

anemia in Zanzibar preschool children. J Nutr 134:348-356.

Strauss RS. (2000): Adult functional outcome of those born small for gestational age:

twenty-six-year follow-up of the 1970 British Birth Cohort. JAMA 283:625–632.

Stuebe AM, Rich-Edwards JW, Willet WC, Manson JE and Michels KB (2005):

Duration of lactation and incidence of type 2 diabetes. JAMA 294:2601-2610.

Stützle W, Gasser Th, Molinari L, Largo RH, Prader A and Huber PJ (1980): Shape-

invariant modelling of human growth. Ann Hum Bio 7:507-528.

Tanner JM (1963): The regulation of human growth. Child Dev 34:817-847.

Tanner JM (1952): The assessment of growth and development in children. Arch Dis

Chid 27(131):10-33.

Tanner JM and Cameron N (1980): Investigation of the mid growth spurt in height,

weight and limb circumferences in single-year velocity data from the London 1966-67

growth survey. Ann Hum Biol 7:565-577.

Taylor-Robinson DC, Jones AP and Garner P (2007): Deworming drugs for treating

soil-transmitted intestinal worms in children: effects on growth and school

performance. Cochrane Database of Syst Rev

4:CD000371.DOI:10.1002/14651858.CD000371.

Thacher TD, Fischer PR, Pettifor JM, Lawson JO, Isichei CO, Reading JC and Chan

GM (1999): A comparison of calcium, vitamin D, or both for nutritional rickets in

Nigerian children. N Engl J Med 341:563-568.

136

The World Report 2002 (2002): Reducing risks, promoting healthy life. World Health

Organisation, Geneva.

Thior I, Lockman S, Smeaton LM, Shapiro RL, Wester C, Heymann SJ, Gilbert PB,

Stevens L, Peter T, Kim S, van Widenfelt E, Moffat C, Ndase P, Arimi P,

Kebaabetswe, Mazonde P, Makhema J, McIntosh K, Novitsky V, lee TH, Marlink R

and Essex M for the Mashi Study Team (2006): Breastfeeding plus infant zidovudine

prophylaxis for 6 months vs formula feeding plus zidovudine for 1 month to reduce

mother-to-child HIV transmission in Botswana: a randomised controlled trial. JAMA

296:794-805.

Torlesse H and Hodges M (2001): Albendazole therapy and reduced decline in

haemoglobin concentration during pregnancy (Sierra Leone). Trans R Soc Trop Med

Hyg 95:195-201.

Uauy R, Hoffman DR, Peirano P, Birch DG and Birch EE (2001): Essential fatty

acids in visual and brain development. Lipids 36:885-895.

Ulijaszek SJ and Kerr DA (1999): Anthropometric measurements error and the

assessment of nutritional status, Br J Nutr 82(3):165-177.

UNICEF (2007): State of the world children 2008. New York

UNICEF (1998): The state of the world children 1998. New York. Oxford University

Press

Underwood BA and Arthur P (1996): The contribution of vitamin A to public health.

FASEB J 10:1040-1048.

137

van Dommelen P, de Gunst MCM, van der Vaart AW and Boomsma DI (2007):

Genetic study of the height and weight process during infancy. Twin Res 7(6):607-

616.

van Hensbroek MB, Morris-Jones S, Meisner S, Jaffar S, Bayo L, Dackour R,

Christine P, Greenwood BM (1995): Iron, folic acid, combined with effective

antimalarial therapy promotes haematological recovery in African after acute

falciparum malaria. Trans R Soc Trop Med Hyg 89:672-676.

Van Thuy P, Berger J, Nakanishi Y, Khan NC, Lynch S and Dixon P (2005): The use

of NaFeEDTA-fortified fish source is an effective tool for controlling iron deficiency

in women of childbearing age in rural Vietnam. J Nutr 135:2596-2601.

Verhoef H, West CE, Nzyuko SM, de Vogel S, van der Valk R, Wanga MA, Kuijsten

A, Veenemans J and Kok FJ (2002): Intermittent administration of iron and

sulfadoxine-pyrimethamine to control anaemia in Kenyan children: a randomised

controlled trial. Lancet 360:908-914.

Victora CG, Morris SS, Barros FC, de Onis M and Yip R (1998): The NCHS

reference and the growth of breast and bottle fed infants. J Nutr 128:1134-1138.

Victora CG, Linda A, Caroline F, Hallam PC, Martorell R, Richter L and Sachdev HS

for the Maternal Undernutrition Study Group (2008): Maternal and child

undernutrition: consequences for adult health and human capital. Lancet 371:340-357.

Villar J, Smeriglio V, Martorell R, Brown CH and Klein RE (1984): Heterogeneous

growth and mental development of intrauterine growth-retarded infants during the

first 3 years of life. Pediatrics 74:783–791.

138

Wagner CL, Hulsey TC, Fanning D, Ebeling M and Hollis BW (2006): High-dose

vitamin D3 supplementation in a cohort of breastfeeding mothers and their infants: a

6-month follow-up pilot study. Breastfeed Med 1:59-70.

Walker M (2007): International breastfeeding initiatives and their relevance to the

current state of breastfeeding in the United States. J Midwifery Womens health

52:549-555.

Walker SP, Chang SM, Powell CA and Grantham-McGregor SM (2005): Effects of

early childhood psychosocial stimulation and nutritional supplementation on

cognition and education in growth stunted Jamaican children: prospective cohort

study. Lancet 366:1804-1807.

Walker SP, Chang SM, Powell CA and Grantham-McGregor SM (2004):

Psychosocial intervention improves the development of term low birth-weight infants.

J Nutr 134:1417–1423.

Webb KE, Horton NJ and Katz DL (2005): Parental IQ and cognitive development of

malnourished Indonesian children. Euro J Clin Nutr 59:618–620.

Werler MM, Louik C and Mitchell AA (1999): Achieving a public health

recommendation for preventing neural tube defects with folic acid. Am J Public

Health 89:1637-1640.

West KP, Katz J, Khatry SK, LeClerq SC, Pradhan EK, Shrestha SR, Connor PB, Dali

SM, Christian P, Pokhrel RP and Sommer A on behalf of NNIPS-2 Group (1999):

Double blind, cluster randomised trial of low dose supplementation with vitamin A or

carotene on mortality related to pregnancy in Nepal. BMJ 318:570-575.

139

WHO (1983): Measuring change in Nutritional Status. Geneva: World Health

Organization.

WHO and FAO (2004): Vitamin and mineral requirements in human nutrition: report

of a joint FAO/WHO expert consultation, Bangkok, Thailand.

WHO (2007): Protein and amino acid requirements in human nutrition: report of a

joint FAO/WHO/UNU expert consultation. WHO technical report series; 935

WHO (1998): Complementary feeding of young children in developing countries: a

review of current scientific knowledge. Geneva, WHO/NUT/98.1

WHO Collaborative Study Team on the Role of Breastfeeding on the Prevention of

Infant Mortality (2000): Effect of breastfeeding on infant and child mortality due to

infectious diseases in less developed countries: a pooled analyses. Lancet 355:451-

455.

WHO Expert Committee of Physical Status (1995): The use and interpretation of

anthropometry physical status. WHO technical report series; 854.

WHO Multicentre Growth Reference Study Group (2006): Assessment of differences

in linear growth among populations in the WHO Multicentre Growth Reference

Study. Acta Paediatr 450(suppl.):56-65.

WHO (2006): WHO Child Growth Standards Technical report: length/height-for-age,

weight-for-age, weight-for-length, weight-for-height and body mass index-for-age:

methods and development. Geneva.

WHO (2003): Global strategy for infant and young child feeding. Geneva.

140

Winkvist A, Habicht JP and Rasmussen KM (1998): Linking maternal and infant

benefits of a nutritional supplement during pregnancy and lactation. Am J Clin Nutr

68:656-661.

Xu X, Wang W, Guo Z and Karlberg J (2002): Longitudinal Growth during infancy

and childhood in children from Shanghai: predictors and consequences of the age at

onset of the childhood phase of growth. Pediatr Res 51:377-385.

Zaman S, Ashraf RN and Martines J (2008): Training in complementary feeding

counselling of healthcare worker and its influence on maternal behaviours and child

growth: A cluster-randomized controlled trial in Lahore, Pakistan. J Health Popul

Nutr 26:210-222.

Zeghoud F, Ben-Mekhbi H, Djeghri N and Garabédian M (1994): Vitamin D

prophylaxis during infancy: comparison of the long-term effects of three intermittent

doses (15, 5, or 2.5 mg) on 25-hydroxyvitamin D concentrations. Am J Clin Nutr

60:393-396.

Zhao J and van der Haar F (2004): Progress in salt iodisation and improved iodine

nutrition in China 1995-99. Food Nutr Bull 29:9-16.

Zimmermann MB, Wegmueller R, Zeder C, Chaouki N, Biebinger R, Hurrell RF and

Windhab E (2004) Triple fortification of salt with microcapsules of iodine, iron, and

vitamin A. Am J Clin Nutr 80:1283-1290.

Zlotkin S, Antwi KY, Schauer C and Yeung G (2003): Use of microencapsulation (II)

fumarate sprinkles to prevent recurrence of anaemia in infants and young children at

high risk. Bull World Health Organ 81:108-115.

141

ORIGINAL PUBLICATIONS

142

12 APPENDICES

12.1 Geographical and demographic profile of Malawi

Malawi has a total area of has 94,276 km2 and is located mostly on the western side of

Lake Malawi and both eastern and western sides of Shire river that drain water form

Lake Malawi. The lake itself is the south most lake within the Great African Rift

Valley. Malawi has a total population of 12.8 million with annual growth rate 2.1%.

The population of under fives was 2,340, 000. The crude birth and death rates were 43

and 21 per 1000 population, respectively. The life expectancy at birth was at 40 and

the total fertility rate is at 5.9. The population living in urban areas was 17% and

mean growth rate of urban population was 4.6%. Maternal mortality ratio was

reported at 980 per 100,000. The latest adjusted maternal mortality ratio for under-

reporting and misclassification was at 1800 per 100,000 live births in 2000 and its

corresponding lifetime risk of maternal death was 1 in 7.

12.1.1 Health indicators

Up to 73% of the population was drinking water from improved sources and 61%

used sanitation facilities. Immunisation coverage was high; DPT 1 and DPT3 were 99

and 93 % respectively, polio was 94%, measles was 82%, hepB3 and Hib3 were at

143

93%. Twenty percent of under-fives slept under bed net and similar age group

sleeping under treated bed nets were 15%. Among the under fives with fever

receiving anti malarial drugs was 28% and those with diarrhoea receiving Oral

Rehydration Salts (ORS) was 51%. The estimated adult HIV prevalence rate was

14.1%.

12.1.2 Nutritional indicators

In Malawi 16% infants were born with low birth weight. Prevalence of exclusive

breast feeding for less than 6 months was 56%, but 89% were breast fed combined

with complementary feeding between 6 and 9 months and 73% of children aged 20 to

23 months were still breast fed. Among under-fives, 22% and 5% were moderate-to-

severe and severe underweight, respectively. Correspondingly, 5% and 48% were

moderate-to-severe wasting and stunting, respectively. Vitamin A supplementation

coverage was at among 6-59 mo old was at 57% and 49% of households consumed

iodised salt.

12.1.3 Economic indicators

Malawi is one of the low income countries with it economy relying on agriculture

exports. At the time of the study gross national income (GNI) was at 160 US$ per

144

capita. Gross domestic product (GDP) per capita mean growth rate was at 1.0%. The

mean annual rate of inflation was at 29% and 42% of the population lived on less than

1 US$ per day. Of the total annual budget 7% was spent on health, 12% on education

and 5% on defence. The net official development assistance (ODA) inflow was at 476

million US$ that contributed to 23% of the GNI. The countries debt service as

percentage of exports of goods and services was at 6%.

12.2 Anthropometric measurements

12.2.1 Height

Height (stature) is uni-dimensional measurement that indicates the length of long bone

of the lower limb and bones in the vertebra column. Length is measured in supine

position up to 24 months using a fixed board then standing after height is measured in

standing position using a stadiometer. During growth period height is dependent on

age and sex. For several reasons, including malformations, deformations, and

congenital disorders, an individual may become dysmorphic resulting in abnormal

stature. In the absence of dysmorphism extreme variations of normality, prenatal

problems, malnutrition, psychological deprivation, chronic diseases, treatments and

endocrines disorders may be responsible for short stature

145

12.2.2 Weight

Weight, a measure of body mass, is dependent on height and body composition (total

body water, muscle mass, total lean body mass, fat) in addition to age and sex.

Because of its dependency on body composition which may vary within a short period

of time, weight changes may be observed within as short period as a day. Weight is

measured using scales of preferred accuracy with minimum clothing or appropriately

adjusted for if clothes are on.

12.2.3 Head Circumference

Head circumference is a proxy to measuring the brain size. Out of normal range head

circumference measurements help in diagnosing brain abnormalities like

hydrocephalous or microcephaly. Head circumference is measured using non

stretching tape measures.

12.3 Mid-Upper Arm Circumference

Mid-upper arm circumference is a proxy of measuring the amount of fat layer in the

arm. It is a simpler way of assessing wasting or weight for height. Arm circumference

146

grows very fast in the first year, thereafter it remains the same up to 60 months;

making MUAC an important to assess wasting. MUAC is measured using non

stretching tape measures.

Supplementary feeding with fortified spread amongmoderately underweight 6–18-month-old ruralMalawian children

John Phuka*†, Chrissie Thakwalakwa*, Kenneth Maleta*, Yin Bun Cheung‡,André Briend§, Mark Manary¶ and Per Ashorn†***College of Medicine, University of Malawi, P/Bag 360 Blantyre, Malawi, †Department of International Health, University of Tampere Medical School,Tampere, Finland, ‡London School of Hygiene and Tropical Medicine, London, UK, §World Health Organization, Department of Child Health andDevelopment, Geneva, Switzerland and IRD, Département Sociétés et Santé, Paris, France, ¶Washington University School of Medicine, St. Louis, Missouri,USA, and **Department of Paediatrics, Tampere University Hospital, Tampere, Finland

Abstract

We aimed to analyse growth and recovery from undernutrition among moderately underweightambulatory children receiving micronutrient-fortified maize–soy flour (Likuni Phala, LP) orready-to-use fortified spread (FS) supplementary diet. One hundred and seventy-six 6–18-month-old individuals were randomized to receive 500 g LP or 350 g FS weekly for 12 weeks.Baseline and end of intervention measurements were used to calculate anthropometric gainsand recovery from underweight, wasting and stunting. Mean weight-for-age increased by 0.22(95% CI 0.07–0.37) and 0.28 (0.18–0.40) Z-score units in the LP and FS groups respectively.Comparable increase for mean weight-for-length was 0.39 (0.20–0.57) and 0.52 (0.38–0.65)Z-score units. Recovery from underweight and wasting was 20% and 93% in LP group and 16%and 75% in FS group. Few individuals recovered from stunting and mean length-for-age was notmarkedly changed. There were no statistically significant differences between the outcomes inthe two intervention groups. In a poor food-security setting, underweight infants and childrenreceiving supplementary feeding for 12 weeks with ready-to-use FS or maize–soy flour porridgeshow similar recovery from moderate wasting and underweight. Neither intervention, if limitedto a 12-week duration, appears to have significant impact on the process of linear growth orstunting.

Keywords: fortified spread, infants and young children, randomized controlled trial, supplemen-tary feeding, moderately underweight, undernutrition.

Correspondence: John Phuka, College of Medicine, University of Malawi, PO Box 431, Mangochi, Malawi. E-mail: johnphuka@

malawi.net

DOI: 10.1111/j.1740-8709.2008.00162.x

Original Article

1© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Introduction

In Malawi, like most low-income countries, childhoodmalnutrition is very common. Between 22% and 48%of Malawian children are undernourished by the ageof 18 months (National Statistical Office & ORCMacro 2001). The peak incidence is between 6 and 18months of age and moderate underweight is the mostcommon presentation (National Statistical Office &ORC Macro 2001; Maleta et al. 2003a). Apart fromcausing acute morbidity and adverse long-termsequelae, malnutrition is estimated to contribute toapproximately half of the worldwide deaths in childrenunder 5 years of age (Pelletier et al. 1995; Caulfieldet al. 2004).

The epidemiology of undernutrition in Malawinecessitates emphasis on prevention or early home-based management of children who have developedsigns of malnutrition but who do not yet require hos-pitalization. There are, however, no easy options forsuch approaches as infection control has not provenwidely successful and not many indigenous foodsare rich in all the nutrients or available throughoutthe year (ACC/SCN 2001). Furthermore, widespreadpoverty in many communities reduces the possibilityto purchase commercially available complementaryfoods. Thus, there is a need to develop and test newand inexpensive food supplements that could beeasily used at the community level.

Cereal and legume mixtures that resemble theindigenous diet and are prepared in a mannersimilar to staple food have often been recommendedfor supplementary feeding of moderately undernour-ished children (Maleta et al. 2004). One such item isporridge made of vitamin and mineral-enrichedmaize–soy flour [Likuni Phala (LP)], but data onits actual impact to the children’s growth andnutritional status are scarce. Another option aremicronutrient-fortified, high-energy spreads (ready-to-use therapeutic foods, RUTF) that have provenvery beneficial for severely malnourished childrenboth in famine and non-famine situations (Briendet al. 1999; Briend 2001; Collins 2001; Diop el et al.2003; Manary et al. 2004; Sandige et al. 2004;Ciliberto et al. 2005, 2006; Ndekha et al. 2005).Similar products have recently been given also to

less severely wasted children (Patel et al. 2005;Defourny et al. 2007), and recent data from a pre-liminary dose-finding trial in Malawi suggested thatthey might improve growth also among moderatelyunderweight young individuals (Kuusipalo et al.2006). Thus far, however, the effects of fortifiedspreads (FS) and maize–soy flour have not beencompared in this target group with a randomizedcontrolled trial.

To more comprehensively analyse the efficacy of FSand maize–soy flour in promoting growth and recov-ery from malnutrition among moderately under-weight 6–18-month-old children, we conducted arandomized controlled, single-blind trial where mod-erately underweight children were provided for 12weeks with weekly supplementary rations of eithermaize–soy flour or FS. The present communicationreports mean growth, recovery from malnutritionand change in blood haemoglobin concentrationamong 6–18-month-old children getting the twosupplements.

Methods

Study area and timing

The study was conducted in Lungwena, Mangochidistrict of Malawi, South-Eastern Africa. InLungwena, exclusive breastfeeding for 6 months isalmost non-existent and infant diet is typicallycomplemented with thin maize porridge (‘phala’,10% dry weight of maize flour) as early as from2 to 6 months of age. Underweight and stuntingare very common: in a recent prospective cohortstudy of 813 newborns, underweight [weight-for-ageZ-score (WAZ) < -2] and stunting [length-for-ageZ-score (LAZ) < -2] prevalence was 10% and50% by 6 months of age and 40% and almost 80%by 18 months of age respectively (Maleta et al. 2003a).The climate in the area has one rainy seasonbetween December and March during which thestaple food maize is grown. Enrolment to the trial wasperformed during the growing season when foodlevels were at the lowest, i.e. in March 2005. The12-week follow-up of the last participant ended inJuly 2005.

J. Phuka et al.2

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Eligibility criteria, enrolment and randomizationof the trial participants

Inclusion criteria for the trial included age of at least6 months but less than 15 months, low WAZ(WAZ < -2.0), assumed residence in the studyarea throughout the follow-up period and signedinformed consent from at least one authorizedguardian. Exclusion criteria were severe wasting,weight-for-length Z-score (WLZ < -3.0), presence ofoedema, history of peanut allergy, severe illness war-ranting hospitalization on the enrolment day, concur-rent participation in another clinical trial or anysymptoms of food intolerance within 30 min afterthe ingestion of a 6-g test dose of FS, one of the foodsupplements used in the trial (given to all potentialparticipants to exclude the possibility of peanutallergy).

For enrolment, trained health surveillance assis-tants from a local health centre contacted families,who were registered to live in the study area andknown to have a baby of approximately right age.Infants and children, whose ages were appropriateand verified from an under-five-clinic card, wereweighed in their homes and those who were moder-ately underweight or near the cut-off point for it(WAZ < -1.8) were invited for further screening atthe health centre.At the enrolment session, guardianswere given detailed information on the trial contentsand the infants and children were fully assessed foreligibility.

For group allocation, consenting guardians of eli-gible participants were shown and asked to pick onefrom a set of 10 identical opaque envelopes contain-ing information on the group allocation of the partici-pant. Because one set had to be finished before usingthe next, for each block the first guardian chose onefrom a total of 10, the second chose one from a total of9 and the 10th picked the last envelope. The blockedrandomization list and envelopes were created bypeople not directly involved in trial implementationand the code was not broken until all data had beencollected and entered into a database. The actual ran-domization was done with a tailor-made computerprogram, using random number and rank functionsof a Microsoft Excel spreadsheet.

Interventions and follow-up

Trial participants received one of two interventions:An average of 71 g day-1 of micronutrient-fortifiedmaize–soy flour (LP) or an average 50 g of micronu-trient FS. The supplements were home-delivered tothe participants at weekly intervals (either 500 g LPor 350 g FS at each food delivery). LP was packed inbags of 500 g and was purchased from a local pro-ducer (Rab Processors, Limbe, Malawi). Fortifiedspread was produced and packed in 50 g daily dosepackets by Nutriset (Malaunay, France). Ingredientsof LP were maize flour, soy flour and micronutrientsand those of FS were peanut butter, milk, vegetableoil, sugar and micronutrients (Table 1).

FS could be eaten as such, whereas the maize-soyflour (LP) required cooking into porridge beforeconsumption. The guardians were provided withspoons and advised to daily offer their infants orchildren either a packet of FS (FS group) or por-ridge containing 12 spoonfuls of maize–soy flour

Table 1. Energy and nutrient content of the daily ration of the foodsupplement used in the trial

Variable LP group FS group Recommendednutrient intake*

Weight 71 g 50 gEnergy (kcal) 282 256Protein (g) 10.3 7.0Carbohydrates (g) n/a 13.8Fat (g) 3.1 16.9Retinol (mg RE) 138 400 400Folate (mg) 43 160 160Niacin (mg) 3 6 6Panthothenic ac.(mg) n/a 2 2Riboflavin (mg) 0.3 0.5 0.5Thiamin (mg) 0.1 0.5 0.5Vitamin B6 (mg) 0.3 0.5 0.5Vitamin B12 (mg) 0.9 0.9 0.9Vitamin C (mg) 48 30 30Vitamin D (mg) n/a 5 5Calcium (mg) 71 366 500Copper (mg) n/a 0.4 0.4Iodine (mg) n/a 135 75Iron (mg) 5 8 5Magnesium (mg) n/a 60 60Selenium (mg) n/a 17 17Zinc (mg) 3.6 8.4 8.4

LP, Likuni Phala (maize–soy flour); FS, fortified spread; n/a, informa-tion not available; * FAO & WHO 2002.

Supplementary feeding with fortified spread 3

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

(LP group). All mothers were encouraged to con-tinue breastfeeding on demand and to feed theirchildren only as much of the food supplement as thechild wanted to consume at a time (i.e. using respon-sive feeding practices and recognizing signs ofsatiety).

The participants were visited weekly at theirhomes to collect information on supplement use andpossible adverse events. Empty food containers werecollected every 3 weeks. At 6 and 12 weeks afterenrolment, the participants were invited to a localhealth centre where they underwent physical exami-nation, anthropometric assessment and laboratorytests.

Measurement of outcome variables

The primary outcome of the trial was weight gainduring the 12-week follow-up. Secondary outcomescomprised length gain, mean change in anthropomet-ric indices WAZ, LAZ and WLZ, recovery from mod-erate underweight, stunting or wasting, change in midupper arm circumference (MUAC) and change inblood haemoglobin concentration.

Weight was measured from naked children with anelectronic children weighing scale (SECA 834, Chas-mors Ltd, London, UK) and recorded to the nearest10 g. Length was measured to the nearest 1 mm witha high quality length board (Kiddimetre, RavenEquipment Ltd, Essex, UK). MUAC and head cir-cumference were measured to the nearest 1 mm withnon-stretchable plastic tapes (Lasso-o tape, HarlowPrinting Limited, South Shields, Tyne and Wear, UK).All measurements were done in triplicate by oneauthor (JP), whose measurement reliability wasassessed at the start of the study and who was blindedof the participant study allocation from enrolment tothe end of follow-up. All instruments were calibratedevery day before the measurements. Anthropometricindices (WAZ, LAZ and WLZ) were calculated withEpi-Info 3.3.2 software (CDC, Atlanta, USA), basedon the CDC 2000 growth reference (Kuczmarski et al.2002), which was the latest available reference mate-rial at the time of enrolment to the trial.

A 2 ml venous blood sample was drawn and serumwas separated by centrifugation at the beginning and

end of follow-up. Haemoglobin concentration wasmeasured from a fresh blood drop with Hemo-Cue®cuvettes and reader (HemoCue AB, Angelholm,Sweden).

All participants were tested for malaria parasitesand human immunodeficiency virus (HIV) infectionat the beginning of the trial. Thick and thin bloodsmears were made, stained with Giemsa and screenedmicroscopically for malaria parasites from all partici-pants. Screening for HIV-infection was done fromfresh blood with two antibody rapid tests, accordingto the manufacturers’ instructions (Determine,Abbott Laboratories, Abbot Park, IL, USA andUni-Gold, Trinity Biotech plc, Bray, Ireland). Driedfilter paper blood samples from participants with apositive result in either of the rapid tests were furthertested with DNA amplification technology, accordingto the manufacturers’ instructions (Amplicor HIV-1Monitor Test Version 1.5, Hoffmann-La Roche, Ltd,Basel, Switzerland).

Training and quality control

All data collection was implemented according tostandard operating procedures (SOP) developedbefore commencing the study. The four researchassistants for the trial were trained on the use of theseSOPs as well as the use of questionnaires and foodhygiene. The performance of data collectors weredaily monitored by one author (JP) and a seniorresearch assistant (MB).

The reliability of anthropometric measurementswas assessed for one authors (JP) and one researchassistant (MB) before the commencement of the trial.In standardization measurements from eight infants,the technical error of measurement for JP was0.41 cm, 0.13 cm and 0.35 cm for length, MUAC andhead circumference respectively (WHO, MulticentreGrowth Reference Study Group 2006). Comparedwith the expert (MB), the observer (JP) overesti-mated MUAC by 0.35 cm and underestimated headcircumference and length on average by 0.35 cm andby 0.37 cm respectively. Coefficient of reliability was0.98, 0.89 and 0.96 for length, MUAC and headcircumference respectively (Marks et al. 1989).

J. Phuka et al.4

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Sample size calculation

Sample size was calculated from expected values forthe primary outcome (weight gain) based on anearlier preliminary dose-finding trial (Kuusipalo et al.2006). Assuming a standard deviation (SD) of 0.39 kgfor changes in weight in both groups during interven-tion period and a mean gain difference of 0.17 kgbetween the intervention group (FS) and the controlgroup (LP), a sample size of 84 infants per group wasrequired to provide the trial with 80% power and95% confidence. To allow for approximately 5%attrition, the target enrolment was 88 infants pergroup.

Data management and analysis

Raw data were initially recorded on paper formsand their coherence was checked daily by a seniorresearch assistant. Summary data were transcribed topaper case report forms and then double-entered intoa tailor-made Microsoft Access 2003 database. Thetwo entries were electronically compared; all extremeand otherwise susceptible values were confirmed orcorrected.

Statistical analysis was carried out using Stata 9.2(StataCorp, College Station, TX, USA). For the sta-tistical analysis, the study participants were classifiedinto two groups according to the trial allocation. Par-ticipants with no anthropometric data at the finalmeasurement were excluded from outcome analysesbut included in the comparison of baseline character-istics, and all those with at least one follow-up mea-surement were included in the sensitivity analysis.

For continuous and categorical outcomes, the twointervention groups were compared with t-test andFisher’s exact test respectively. For comparison ofcontinuous baseline and final anthropometric andhaematological group values, paired t-test was used,whereas categorical baseline and final variables werecompared using sign test. For the studies of compli-ance using visits as units of analysis, the Huber–Whiterobust standard error was used to allow for correlateddata (multiple visits per child). Confidence intervalfor SD was calculated using the Jacknife method.

Recovery from underweight was defined asWAZ � -2.0 at enrolment and WAZ > -2.0 at the end

follow-up. Recovery from wasting (WLZ � -2.0 atenrolment) and stunting (LAZ � -2.0 at enrolment)were defined accordingly as WLZ > -2.0 andLAZ > -2.0 at completion of the trial respectively.

Ethics, study registration and participant safety

The trial was performed according to InternationalConference of Harmonization–Good Clinical Prac-tice (ICH-GCP) guidelines and it adhered to theprinciples of Helsinki declaration and regulatoryguidelines in Malawi. Before the onset of enrolment,the trial protocol was reviewed and approved by theCollege of Medicine Research and Ethics Committee(University of Malawi) and the Ethical Committee ofPirkanmaa Hospital District (Finland). Key details ofthe protocol were published at the clinical trial regis-try of the National Library of Medicine, Bethesda,MD, USA (http://www.clinicaltrials.gov, trial identifi-cation is NCT00131222).

A data safety and monitoring board (DSMB) con-tinuously monitored the safety of the trial. All sus-pected serious adverse events (SAE) were reportedto the DSMB for assessment. SAE was defined as anyuntoward medical occurrence that either resulted indeath or was life-threatening, or required inpatienthospitalization or prolongation of existing hospital-ization, or results in persistent or significant disability/incapacity or other serious medical condition.

Results

Of the 1657 initially screened infants and children,367 were too old (>14.99 months), 1010 wereabove anthropometric cut-off for further evaluation(WAZ > -1.80) and two were too ill on the day ofscreening. Among the rest, 30 were not brought tothe enrolment session, 72 were not underweight(WAZ > -2.00) and 2 declined participation afterreceiving full information of the trial. The remaining176 infants and children were randomized into twointervention groups (Fig. 1). There was no differencein the mean WAZ between the enrolled children andthose who were eligible but not enrolled (P = 0.34).None of the participants showed any adverse reactionto the 6-g test dose of FS.

Supplementary feeding with fortified spread 5

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Table 2 shows the baseline characteristics andanthropometric measurements of the participants byintervention group. At enrolment, participants in theLP group were on average 5 days younger, 100 glighter and 0.6 cm shorter than those in the FS group(Table 2). The prevalence (number of participants)with underweight (WAZ < -2) was 100% (86) in theLP and 99% (89) in the FS group. Comparablefigures for stunting (LAZ < -2) were 69% (59) vs.64% (33) and those for wasting (WLZ < -2) were16% (14) vs. 13% (12), in the LP and FS groupsrespectively. Eight of the participants had a positiveHIV-antibody test at enrolment, but only five of themwere truly HIV-infected, as evidenced by a positivePolymerase Chain Reaction (PCR) test.

During the 12-week follow-up, three participantsdied and three were lost to follow-up (Fig. 1). Theassumed causes of death were anaemia and respira-tory tract infection in the LP group and respiratorytract infection in the FS group. From the lost partici-

pants, one moved away and two were unavailable forfinal measurements although they had participated inall home visits and received and apparently eaten theintervention food. Comparison of baseline WAZbetween participants who completed the follow-upand those who died or were lost did not suggest thatthey come from different populations (P = 0.15 andP = 0.70 respectively). The losses to follow-up werealso not significantly different between the interven-tion groups (P = 0.61 for deaths and P = 0.68 for totalloss to follow-up; Fisher’s exact test).

All mothers reported that their children readily atethe provided supplement and diversion of any portionto someone else than the intended beneficiary wasreported only at 2/2065 (0.10%) food delivery inter-views, both in LP and FS groups. From the weeklyhome visits during which trial products were checked,the percentage of visits with leftovers found were4.9% and 6.5% in the LP and the FS groups respec-tively (P = 0.18).

1657 screened

280 invited for enrolment session

86 randomized to LP group

84 with at least 1 follow-up visit

367 age > 14.99 months1010 WAZ > –1.80

30 no show up 2 refused to participate

72 WAZ > –2.00

90 randomized to FS group

89 with at least 1 follow-up visit

1 death2 deaths

86 finished follow-up

3 dropouts

84 finished follow-up

Fig. 1. Flow of participants. FS, fortified spread; LP, LikuniPhala (maize–soy flour); WAZ, weight-for-age Z-score.

J. Phuka et al.6

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Table 3 shows the main outcome data for continu-ous variables. Among the 170 participants followedup for 12 weeks, mean gain in weight was 50 g [95%confidence interval (CI) -80–174 g] higher in the FSgroup than the LP group. Comparable change forlength was 1.5 mm (95% CI -5.1–2.2 mm) lower inthe FS group. Correspondingly, mean WAZ and WLZincreased more and mean LAZ fell more in the FSgroup than the LP group. None of these differencesreached statistical significance. Average changes inmean MUAC, head circumference and blood haemo-globin concentration were also quite similar in thetwo groups (Table 3). As shown, all 95% CIs wereconsistent with only relatively small between-groupdifferences to either direction.

Because of the limited outcome difference betweenthe groups, we compared the variation in weight gainto the predictions used for sample size calculation.Theobserved standard deviation for weight gain was0.38 kg in FS group and 0.46 kg in LP group,comparedwith estimated SD of 0.39 kg in both groups.While theobserved variation is slightly different in either direc-tion than expected, the 95% CI of the SDs of the wholesample and that of the two intervention groups

included the expected value of 0.39 (Jacknife method),suggesting that a marked between-group differencewould have been identified with the current sampleand trial design if it existed in the population.

Table 4 documents recovery from various forms ofmoderate malnutrition. During the 12-week interven-tion, approximately 20% of the initially underweightindividuals, 85% of the initially wasted, and 10% ofthe initially stunted children recovered from theircondition, with little difference between the two inter-vention groups.

Table 5 demonstrates the prevalence of malnutri-tion at the end of follow-up, i.e. the function of enrol-ment prevalence, incidence of new cases and recoveryduring the intervention. Again, no statistically signifi-cant group-level differences were observed (Table 5).

In a comparison between enrolment status (Table 2)and end of intervention (Tables 3,5), the mean WAZincreased by 0.22 (95% CI 0.07–0.37) Z-score units inthe LP group and by 0.28 (95% CI 0.18–0.40) Z-scoreunits in the FS group. Increase for mean weight-for-length was 0.39 (95% CI 0.20–0.57) and 0.52 (95% CI0.38–0.65) Z-score units for LP and FS groups respec-tively. The prevalence of underweight (WAZ < -2.00)was reduced significantly in both groups (P < 0.001 forLP and P = 0.001 for FS, sign test).A similarly reducedtrend was noticed for the prevalence wasting (P <0.001 for LP and P = 0.07 for FS, sign test), and lowMUAC (P = 0.02 for LP and P < 0.001 for FS,sign test).The prevalence of stunting (P = 0.27 for LP andP > 0.99 for FS, sign test) and anaemia (P = 0.69 for LPand P = 0.04 for FS, sign test) were not markedlychanged during the intervention.

Sensitivity analysis to assess robustness of theresults after loss to follow-up produced similar stan-dardized weight gain as those reported above: 0.30Z-score (95% CI 0.19–0.41) and 0.22 Z-score (95% CI0.19–0.41) for FS and LP groups respectively.

There was no significant interaction between inter-vention and baseline WAZ, WLZ or LAZ, eitherusing weight gain (P = 0.26 for WAZ, P = 0.93 forWLZ and P = 0.25 for LAZ) or length gain (P = 0.47for WAZ, P = 0.19 for WLZ and P = 0.10 for LAZ) asan outcome.

Morbidity was similar in both groups (Table 6). Allinterventions were well tolerated by the participants.

Table 2. Background characteristics of the participants at enrolment

Variable LP group FS group

Number of participants 86 90Mean (SD) age (months) 11.58 (2.77) 11.73 (2.46)Proportion of boys 49/86 (57.0%) 44/90 (49.0%)Mean (SD) weight (kg) 7.02 (1.05) 7.12 (0.82)Mean (SD) length (cm) 67.3 (4.5) 67.9 (3.8)Mean (SD) mid-upper arm

circumference (cm)12.9 (1.0) 13.0 (0.87)

Mean (SD) head circumference(cm)

44.3 (1.9) 44.4 (1.6)

Mean (SD) weight-for-age Z-score -3.09 (0.85) -2.95 (0.71)Mean (SD) length-for-age Z-score -2.45 (0.93) -2.26 (0.88)Mean (SD) weight-for-length

Z-score-1.20 (0.83) -1.21 (0.69)

Mean (SD) blood haemoglobinconcentration (g L-1)

93 (15) 91 (17)

Proportion with PCR confirmedHIV infection

4/82 (4.9%) 1/82 (1.2%)

Proportion with peripheral bloodmalaria parasitaemia

10/86 (11.6%) 10/89 (11.2%)

LP, Likuni Phala (maize–soy flour); FS, fortified spread; SD, standarddeviation; PCR, Polymerase Chain Reaction; HIV, human immuno-deficiency virus.

Supplementary feeding with fortified spread 7

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Table 4. Recovery from moderate malnutrition among participants in maize–soy flour and FS groups

Variable LP group FS group Absolute riskdifference (95% CI)

Relative risk(95% CI)

Incidence (%) Incidence (%)

Weight-for-age Z-score � -2.0 17/84 (20.2) 14/88 (15.9) -4.3 (-15.8 to 7.2) 0.8 (0.4 to 1.5)Weight-for-length Z-score � -2.0 13/14 (92.9) 9/12 (75.0) 17.9 (-45.8 to 10.1) 0.8 (0.6 to 1.2)Length-for-age Z-score � -2.0 5/57 (8.8) 7/57 (12.3) 3.5 (-7.7 to 14.8) 1.4 (0.5 to 4.2)

LP, Likuni Phala (maize–soy flour); FS, fortified spread; CI, confidence interval.

Table 5. Prevalence of malnutrition with various criteria at final measurement among participants in maize–soy flour and FS groups

Variable LP group (n = 84)* FS group (n = 86)* Difference ofproportions(95% CI)Number (%) Number (%)

Weight-for-age Z-score < -2 68 (81.0) 74 (86.1) 5.1 (-6.1 to 16.2)Weight-for-length Z-score < -2 4 (4.8) 5 (5.8) 1.0 (-5.7 to 7.8)Length-for-age Z-score < -2 62 (73) 54 (62.8) -11.0 (-24.9 to 28.7)Mid-upper arm circumference < 12.5 cm 17 (20.2) 7 (8.1) -12.1 (-22.5 to 17.4)Blood haemoglobin < 100 g L-1 48 (57.8) 51 (60.0) 2.2 (-12.7 to 17.0)

LP, Likuni Phala (maize–soy flour); FS, fortified spread; CI, confidence interval; *n for blood haemoglobin was 83 and 85 for LP and FSrespectively.

Table 6. Occurrence of morbidity among participants in maize–soy flour and FS groups

Variable LP group (n = 1020) FS group (n = 1063) Difference ofproportions(95% CI)Number (%) Number (%)

Home visits fever was reported 36 (3.6) 54 (5.1) 1.6 (-0.5 to 3.6)Home visits diarrhoea was reported 74 (7.3) 79 (7.4) 0.2 (-2.7 to 3.1)Home visits ARI was reported 93 (9.1) 95 (8.9) -0.2 (-2.7 to 2.2)Home visits other illnesses were reported 78 (7.7) 63 (5.9) -1.7 (-4.3 to 0.9)

LP, Likuni Phala (maize–soy flour); FS, fortified spread; CI, confidence interval; ARI, acute respiratory tract infection.

Table 3. Anthropometric and haematological outcomes among underweight infants receiving a 12-week dietary supplementation with either FS ormaize–soy flour

Variable LP group FS group Difference in meanchanges (95% CI)

Mean (SD) Mean (SD)

Change in weight (kg) 0.84 (0.46) 0.89 (0.38) 0.05 (-0.80 to 0.17)Change in length (cm) 2.65 (1.1) 2.50 (1.3) -0.14 (-0.51 to 0.22)Change in mid-upper arm circumference (cm) 0.3 (0.8) 0.4 (0.8) 0.1 (-0.1 to 0.3)Change in head circumference (cm) 0.5 (0.6) 0.4 (0.5) -0.1 (-0.3 to 0.1)Change in weight-for-age Z-score 0.22 (0.69) 0.29 (0.51) 0.07 (-0.11 to 0.26)Change in length-for-age Z-score -0.08 (0.41) -0.13 (0.42) -0.05 (-0.17 to 0.08)Change in weight-for-length Z-score 0.39 (0.85) 0.52 (0.63) 0.13 (-0.09 to 0.36)Change in blood haemoglobin concentration (g L-1) 1.5 (16.1) 3.8 (16.9) 2.3 (-2.7 to 7.3)

LP, Likuni Phala (maize–soy flour); FS, fortified spread; CI, confidence interval; SD, standard deviation.

J. Phuka et al.8

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

Besides the three deaths, three participants were hos-pitalized and recorded as having experienced an SAEduring the follow-up. The LP and FS groups sharedfour and two of these SAEs (P = 0.44; Fisher’s exacttest). Both the study physician and the DSMB consid-ered all deaths and SAE unlikely related to the trialinterventions.

Discussion

The present trial was carried out to test the growth-promoting effects of a micronutrient-fortified, energydense ready-to-use spread (FS), when used as asupplementary food among 6–18-month-old under-weight infants and children in rural Malawi. Enrol-ment rate for the trial was high, group allocation wasrandom, trial groups were rather similar at enrolment,follow-up was identical for all groups, compliancewith the intervention appeared good, loss tofollow-up was infrequent and balance between thegroups and people measuring the outcomes wereblinded to the group allocation. Additionally, becausecalculation of anthropometric indices (WAZ, LAZand WLZ) standardizes for gender, baseline differ-ences of gender proportions between the trial groupsdid not affect the interpretation of the main results.Hence, the observed results are likely to be unbiasedand thus representative of the population from whichthe sample was drawn.

In our sample, participants receiving FS for 12weeks gained on average slightly more weight but lesslength than those receiving an iso-energetic portionof maize–soy flour (LP). Mean changes in MUAC andblood haemoglobin concentration were also higher inthe FS group but head circumference increased morein the LP group. However, all differences between thegroups were marginal and the 95% CIs for pointestimates excluded large difference in population.Although the higher end of the 95% CI for the dif-ference in mean weight gain just overlapped with thepredefined cut-off for clinical significance, the datado not therefore support our initial hypothesis thatweekly home provision of FS to 6–15 months under-weight children in rural Malawi would be noticeablybetter in alleviating their being underweight orpromoting catch-up growth over a 12-week period

than similar supplementation scheme with maize–soyflour. Rather, our results are consistent with the twointerventions having a similar or only slightly differ-ent impact on these outcomes.

Children in both intervention groups showed goodrecovery from wasting, some recovery from beingunderweight and no change in their stunting status.For ethical reasons, however, we did not include anunsupplemented control group in our trial design.This omission, while allowing a comparison betweenthe two interventions, limits our possibility to makeconclusions on their effect as compared with youngchildren receiving no intervention. It is thus possiblethat the observed recovery from wasting or beingunderweight, and the increase in mean weight-for-ageor weight-for-length result from a confounder like aseasonal effect on growth or regression to the meanphenomenon. A comparison to earlier studies in thesame area suggest, however, that unsupplementedunderweight children of this age would have asomewhat lower weight and length gain velocity(Kuusipalo et al. 2006) and that seasonal effect onweight gain would be markedly lower than theapproximately 0.3 Z-score unit increase observed inthis trial (Maleta et al. 2003b). Rather, the increase inrelative weight is comparable to observations withother supplementary feeding interventions in similartarget groups (Simondon et al. 1996; Lartey et al. 1999;Becket et al. 2000; Bhandari et al. 2001; Oelofse et al.2003; Adu-Afarwuah et al. 2007).

Micronutrient FS were initially designed to beused as a rehabilitation food for malnourished chil-dren (Briend et al. 1999). In a succession of studies,therapeutic versions of the spreads (RUTF) wereshown to induce recovery, first in a hospital setting,then as outpatients in efficacy trials and finally inprogrammatic settings (Collins 2001; Diop el et al.2003; Manary et al. 2004; Sandige et al. 2004;Ciliberto et al. 2005, 2006; Ndekha et al. 2005).Recently, there has been a major interest to expandthe use of spreads to treatment of less severe con-ditions or primary prevention malnutrition. Theresults from our study support the findings of Patelet al. (2005) in Malawi and those of Medicins SansFrontieres in Niger (Defourny et al. 2007), suggestingthat FS can also efficiently promote recovery from

Supplementary feeding with fortified spread 9

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

moderate wasting, thus acting as a secondary pre-vention of severe malnutrition.

In contrast to recovery from wasting and associatedincrease in weight-for-age, linear growth and theprocess of stunting seems to be very little affected by a12-week supplementary feeding with FS or LP, as evi-denced by a decreasing LAZ and a very low recoveryrate from stunting in both groups in our trial.This is notsurprising, as length gain acceleration has often beenshown to follow weight gain increase with a certainlag period, e.g. by 3 months in rural Malawi (Walker& Golden 1988; Costello 1989; Heikens et al. 1989;Maleta et al. 2003b). Hence, only a longer interventionmight have the potential for the primary prevention ofstunting or rehabilitation of moderately underweightchildren in an area, where low weight is usually afunction of inadequate linear growth.First results fromsuch an approach, i.e. 6–12-month supplementation ofcomplementary feeding with 20–50 g FS have beenpromising, documenting increased length gain andprevention of severe stunting among rural infants inGhana and Malawi (Adu-Afarwuah et al. 2007; Phukaet al. 2008). Possible nutrients in FS that could accountfor improved linear growth in longer interventionsinclude, e.g. essential fatty acids, zinc or componentsof cow milk (Brown et al. 2002; Hoppe et al. 2006;Adu-Afarwuah et al. 2007).

Taken together, our results suggest that 12 week-long supplementary feeding with FS and LP porridgehave similar impacts on the weight and length gain ofunderweight infants in a low-income setting likeMalawi. Although the results are consistent with theidea that both interventions promote recovery fromwasting or moderate underweight, the lack of no-foodcontrol group hinders any firm conclusions on theindependent effect of either intervention.This, and thepossible effect of seasonality on the growth outcomes,should be addressed in further controlled trials.

Acknowledgements

We are grateful to the people of Lungwena, the staffat the Lungwena Training Health Centre and ourresearch assistants for their positive attitude, supportand help in all stages of the study,and to Laszlo Csonkafor designing the collection tools and data entry pro-

grams.André Briend is currently a staff member of theWorld Health Organization (WHO). André Briendalone is responsible for the views expressed in thispublication and they do not necessarily represent thedecisions or stated policy of the WHO.

Source of funding

The trial was funded by grants from the Academy ofFinland (grants 200720 and 109796), Foundation forPaediatric Research in Finland and Medical ResearchFund of Tampere University Hospital. The micronu-trient mixture used in the production of FS was pro-vided free of charge by Nutriset, Inc. (Malaunay,France). John Phuka and Chrissie Thakwalakwa arereceiving personal stipends from Nestlé Foundation.

Conflict of interest statement

André Briend was a consultant to Nutriset untilDecember 2003 and the company has also financiallysupported the planning of another research projectby the same study team through Per Ashorn and theUniversity of Tampere after the completion of thistrial. Other authors declare no conflict of interest.Thefunders of trial had no role in the implementation,analysis or reporting.

References

ACC/SCN (2001) What works? In: A Review of the Effi-cacy and Effectiveness of Nutrition Interventions (edsL.H. Allen & S.R. Gillespie), pp 69–87. ACC/SCN:Geneva in collaboration with the Asian DevelopmentBank: Manila.

Adu-Afarwuah S., Lartey A., Brown K.H., Briend A.,Zlotkin S. & Dewey K.G. (2007) Randomized compari-son of 3 types of micronutrient supplements for homefortification of complementary foods in Ghana: effectson growth and motor development. American Journal ofClinical Nutrition 86, 412–420.

Becket C., Durnin J.V., Aitchison T.C. & Pollitt E. (2000)Effects of an energy and micronutrient supplementationon anthropometry in undernourished children in Indo-nesia. European Journal of Clinical Nutrition 54 (Suppl.2), S52–59.

Bhandari N., Bahl R., Nayyar B., Khokhar P., Rohde J.E. &Bhan M.K. (2001) Food supplementation with encour-agement to feed it to infants from 4 to 12 months of

J. Phuka et al.10

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

age has a small impact on weight gain. Journal of Nutri-tion 131, 1946–1951.

Briend A. (2001) Highly nutrient-dense spreads: a newapproach to delivering multiple micronutrients to high-risk groups. British Journal of Nutrition 85, S175–179.

Briend A., Lacsala R., Prudhon C., Mounier B., Grellety Y.& Golden M.H. (1999) Ready to use therapeutic foodfor treatment of marasmus. Lancet 353, 1767–1768.

Brown K.H., Peerson J.M., Rivera J. & Allen L.H. (2002)Effect of supplemental zinc on the growth and serumzinc concentrations of prepubertal children: a meta-analysis of randomized controlled trials. AmericanJournal of Clinical Nutrition 75, 1062–1071.

Caulfield L.E., de Onis M., Blossner M. & Black R.E.(2004) Undernutrition as an underlying cause of childdeaths associated with diarrhea, pneumonia, malaria,and measles. American Journal of Clinical Nutrition 80,193–198.

Ciliberto M.A., Sandige H., Ndekha M.J., Ashorn P.,Briend A., Ciliberto H.M. et al. (2005) A comparison ofhome-based therapy with ready-to-use therapeutic foodwith standard therapy in the treatment of malnourishedMalawian children: a controlled, clinical effectivenesstrial. American Journal of Clinical Nutrition 81, 864–870.

Ciliberto M.A., Manary M.J., Ndekha M.J., Briend A. &Ashorn P. (2006) Home-based therapy for oedematousmalnutrition with ready-to-use therapeutic food. ActaPaediatrica 95, 1012–1015.

Collins S. (2001) Changing the way we address severe mal-nutrition during famine. Lancet 358, 498–501.

Costello A.M. (1989) Growth velocity and stunting in ruralNepal. Archives of Disease in Childhood 64, 1478–1482.

Defourny I., Seroux G., Abdelkader I. & Harczi G. (2007)Management of moderate acute malnutrition withRUTF in Niger. Emergency Nutrition Network 31, 2–4.

Diop el H.I., Dossou N.I., Ndour M.M., Briend A. & WadeS. (2003) Comparison of the efficacy of a solid ready-to-use food and a liquid, milk-based diet for the rehabilita-tion of severely malnourished children: a randomizedtrial. American Journal of Clinical Nutrition 78, 302–307.

FAO & WHO Joint Expert Consultation Report (2002)Vitamin and Mineral Requirements in Human Nutrition.World Health Organization: Geneva, Switzerland.

Heikens G.T., Schofield W.N., Dawson S. & McGregor S.(1989) The Kingston project. I. Growth of malnourishedchildren during rehabilitation in the community, given ahigh energy supplement. European Journal of ClinicalNutrition 43, 145–160.

Hoppe C., Molgaard C. & Michaelsen K.F. (2006) Cow’smilk and linear growth in industrialized and developingcountries. Annual Review of Nutrition 26, 131–173.

Kuczmarski R.J., Ogden C.L., Guo S.S., Grummer-StrawnL.M., Flegel K.M., Mei Z. et al. (2002) 2000 CDC growthcharts for the United States: methods and development.

National Center for Health Statistics. Vital & Health Sta-tistics 11, 246.

Kuusipalo H., Maleta K., Briend A., Manary M. & AshornP. (2006) Growth and change in blood haemoglobin con-centration among underweight Malawian infants receiv-ing fortified spreads for 12 weeks: a preliminary trial.Journal of Pediatric Gastroenterology and Nutrition 43,525–532.

Lartey A., Manu A., Brown K.H., Peerson J.M. & DeweyK.G. (1999) A randomized, community-based trial of theeffects of improved, centrally processed complementaryfoods on growth and micronutrient status of Ghanaianinfants from 6 to 12 mo of age. American Journal ofClinical Nutrition 70, 391–404.

Maleta K., Virtanen S.M., Espo M., Kulmala T. & AshornP. (2003a) Childhood malnutrition and its predictors inrural Malawi. Paediatric and Perinatal Epidemiology 17,384–390.

Maleta K., Virtanen S.M., Espo M., Kulmala T. & AshornP. (2003b) Seasonality of growth and the relationshipbetween weight and height gain in children under threeyears of age in rural Malawi. Acta Paediatrica 92,491–497.

Maleta K., Kuittinen J., Duggan M.B., Briend A., ManaryM., Wales J. et al. (2004) Supplementary feeding ofunderweight, stunted Malawian children with a ready-to-use food. Journal of Pediatric Gastroenterology andNutrition 38, 152–158.

Manary M.J., Ndekha M., Ashorn P., Maleta K. & BriendA. (2004) Home-based therapy for severe malnutritionwith ready-to-use food. Archives of Disease in Child-hood 89, 557–561.

Marks G.C., Habicht J.P. & Mueller W.H. (1989) Reliabil-ity, dependability, and precision of anthropometric mea-surements. The second national health and nutritionexamination survey 1976–1980. American Journal of Epi-demiology 130, 578–587.

National Statistical Office (Malawi) & ORC Macro. (2001)Malawi Demographic and Health Survey 2000. NationalStatistical Office and ORC Macro: Zomba, Malawi andCalverton, Maryland, USA.

Ndekha M.J., Manary M.J., Ashorn P. & Briend A. (2005)Home-based therapy with ready-to-use therapeutic foodis of benefit to malnourished, HIV-infected Malawianchildren. Acta Paediatrica 94, 222–225.

Oelofse A., Van Raaij J.M., Benade A.J., Dhansay M.A.,Tolboom J.J. & Hautvast J.G. (2003) The effect of amicronutrient-fortified complementary food on micronu-trient status, growth and development of 6 to 12-month-old disadvantaged urban South African infants.International Journal of Food Sciences and Nutrition 54,399–407.

Patel M.P., Sandige H.L., Ndekha M.J., Briend A., AshornP. & Manary M.J. (2005) Supplemental feeding with

Supplementary feeding with fortified spread 11

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

ready-to-use therapeutic food in Malawian children atrisk of malnutrition. Journal of Health, Population, andNutrition 23, 351–357.

Pelletier D.L., Frongillo E.A., Schroeder D.G. & HabichtJ.P. (1995) The effects of malnutrition on child mortalityin developing countries. Bulletin of the World HealthOrganization 73, 443–448.

Phuka J.C., Maleta K., Thakwalakwa C., Cheung Y.B.,Briend A., Manary, M.J. et al. (2008) Complementaryfeeding with fortified spread and incidence of severestunting in 6–18 month old rural Malawians. Archives ofPediatrics & Adolescent Medicine 162, 619–626.

Sandige H., Ndekha M.J., Briend A., Ashorn P. & ManaryM.J. (2004) Home-based treatment of malnourishedMalawian children with locally produced or imported

ready-to-use food. Journal of Paediatric Gastroenterol-ogy and Nutrition 39, 141–146.

Simondon K.B., Gartner A., Berger J., Cornu A., Massa-mba J.P., Miguel J.L.S. et al. (1996) Effect of early, short-term supplementation on weight and linear growth of4–7-mo-old infants in developing countries: a four-country randomized trial. American Journal of ClinicalNutrition 64, 537–545.

Walker S.P. & Golden M.H.N. (1988) Growth in length ofchildren recovering from severe malnutrition. EuropeanJournal of Clinical Nutrition 42, 395–404.

WHO, Multicentre Growth Reference Study Group (2006)Reliability of anthropometric measurements in theWHO Multicentre Growth Reference Study. Acta Paedi-atrica 450 (Suppl.), 38–46.

J. Phuka et al.12

© 2009 The Authors. Journal compilation © 2009 Blackwell Publishing Ltd Maternal and Child Nutrition (2009)

ARTICLE

Complementary Feeding With Fortified Spreadand Incidence of Severe Stuntingin 6- to 18-Month-Old Rural MalawiansJohn C. Phuka, MBBS; Kenneth Maleta, PhD; Chrissie Thakwalakwa, BEd;Yin Bun Cheung, PhD; Andre Briend, PhD; Mark J. Manary, MD; Per Ashorn, PhD

Objective: To compare growth and incidence of mal-nutrition in infants receiving long-term dietary supple-mentation with ready-to-use fortified spread (FS) or mi-cronutrient-fortified maize–soy flour (likuni phala [LP]).

Design: Randomized, controlled, single-blind trial.

Setting: Rural Malawi.

Participants: A total of 182 six-month-old infants.

Intervention: Participants were randomized to re-ceive 1 year of daily supplementation with 71 g of LP (282kcal), 50 g of FS (FS50) (256 kcal), or 25 g of FS (FS25)(127 kcal).

Outcome Measures: Weight and length gains and theincidences of severe stunting, underweight, and wasting.

Results: Mean weight and length gains in the LP, FS50,and FS25 groups were 2.37, 2.47, and 2.37 kg (P=.66) and12.7, 13.5, and 13.2 cm (P=.23), respectively. In the same

groups, the cumulative 12-month incidence of severe stunt-ing was 13.3%, 0.0%, and 3.5% (P=.01), of severe under-weight was 15.0%, 22.5%, and 16.9% (P=.71), and of se-vere wasting was 1.8%, 1.9%, and 1.8% (P�.99). Comparedwith LP-supplemented infants, those given FS50 gained amean of 100 g more weight and 0.8 cm more length. Therewas a significant interaction between baseline length andintervention (P=.04); in children with below-median lengthat enrollment, those given FS50 gained a mean of 1.9 cmmore than individuals receiving LP.

Conclusion: One-year-long complementary feeding withFS does not have a significantly larger effect than LP onmean weight gain in all infants, but it is likely to boostlinear growth in the most disadvantaged individuals and,hence, decrease the incidence of severe stunting.

Trial Registration: clinicaltrials.govIdentifier:NCT00131209.

Arch Pediatr Adolesc Med. 2008;162(7):619-626

T HE HIGH INCIDENCE AND SE-rious consequences ofchildhood undernutritionin sub-Saharan Africa andsome parts of southern Asia

necessitate emphasis on early preven-tion, but there are no easy options for it.1-4

Infection control has not proved to bewidely successful, few indigenous foods arerich in all-important nutrients or avail-able throughout the year, and poverty lim-its possibilities for purchasing commer-cially available nutritious foods.5 InMalawi, thin porridge made of micronu-trient-enriched maize and soy flour is of-ten promoted as the main complemen-tary food for infants and young children.However, it has low calorie density, andit resembles the staple food in the area,which may result in displacement of ha-bitual foods from the beneficiary’s diet ordiversion of the complementary food toother family members.6

Recently, Briend and his collabora-tors7,8 developed the concept of highly nu-trient- and calorie-dense spreads, which aresimple to produce, need no cooking be-fore use, and can be stored for months evenin warm conditions. The best-known for-mulation of such spreads is called ready-to-use therapeutic food.9,10 Several clinicaltrials6,11-15 in Malawi have shown that ready-to-use therapeutic food is safe and effec-tive in the rehabilitation of severely mal-nourished children and that a modification

of it interferes with habitual diet less thana porridge supplement and has a positiveeffect on 3- to 4-year-old children who areless severely underweight and experienceless stunting. Recent data from a prelimi-nary dose-finding trial16 suggested thathome provision of fortified spread (FS) to

For editorial commentsee page 692

Author Affiliations are listed atthe end of this article.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM619

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

moderately underweight 6- to 17-month-old infants im-proves their nutritional status and results in markedly in-creased linear growth and improved weight gain.

Because of their efficacy in promoting growth in se-verely or moderately malnourished infants and their easeof use, simple production, and relatively low price, FSsmight logically have a potential as complementary foodsin the sub-Saharan setting. To compare the efficacy ofFS with that of a traditionally used complementary foodin promoting health and preventing the development ofundernutrition, we conducted a randomized, con-trolled, single-blind trial in which initially healthy in-fants were provided with daily rations of either micro-nutrient-fortified maize–soy flour (likuni phala [LP]) orFS for 1 years. The present study analyzes the hypoth-esis that infants receiving FS would grow better and beless likely to develop undernutrition than those receiv-ing LP. Morbidity and developmental outcomes will bereported in separate studies.

METHODS

STUDY AREA AND TIMING

This study was conducted between October 11, 2004, and De-cember 19, 2005, in Lungwena, a rural Malawian communitywith a high prevalence of early childhood stunting and under-weight.2,17 The staple food, maize, was grown during the singlerainy season between December and March. Exclusive breast-feeding for infants was almost nonexistent, and the infant dietwas typically complemented with thin maize porridge alreadyfrom 2 to 6 months of age.

ELIGIBILITY CRITERIA, ENROLLMENT,AND RANDOMIZATION

The inclusion criteria included age 5.50 to 6.99 months, resi-dence in the study area, and informed consent from at least 1 au-thorized guardian. The exclusion criteria were low weight forlength (�−2.0), presence of edema, history of peanut allergy, se-vere illness warranting hospitalization on the enrollment day, con-current participation in another clinical trial, and any symptomsof food intolerance within 30 minutes after ingesting a 6-g testdose of FS, 1 of the food supplements used in the trial.

For enrollment, trained health surveillance assistants con-tacted all the families known to live in the area and known tohave an infant of approximately the right age. Infants were in-vited to an enrollment session, where they were screened foreligibility, and guardians were given detailed information onthe trial contents. Before enrollment, a guardian signed a writ-ten consent form for trial participation.

For group allocation, guardians chose 1 envelope from a setof identical-appearing opaque envelopes, each containing a pieceof paper indicating an identification number and randomly as-signed allocation to 1 of the 3 interventions. The randomiza-tion list and envelopes were made by individuals not involvedin trial implementation, and the code was not disclosed to theresearchers or to those assessing the outcomes until all data hadbeen entered into a database.

INTERVENTIONS AND FOLLOW-UP

There were 3 intervention schemes. Infants in the control groupwere provided with a mean of 71 g/d of LP. Participants in theother 2 groups received a mean of either 50 or 25 g/d of mi-cronutrient-fortified spread (FS50 or FS25, respectively). Thesupplements were home delivered at 3-week intervals (at eachfood delivery, three 500-g bags of LP, four 262-g jars of FS50,or two 262-g jars of FS25 were given).

Likuni phala was purchased from a local producer (Rab Pro-cessors, Blantyre, Malawi). Fortified spread was produced at aMalawian nongovernmental organization, Project Peanut But-ter (Blantyre), from peanut paste, milk powder, vegetable oil,sugar, and premade micronutrient mixture (Nutriset Inc, Ma-launay, France). All the supplements were fortified with mi-cronutrients, but the level of fortification varied between theproducts. The difference between the FS50 and FS25 supple-mentation was in the amount of food base given (50 vs 25 g/d).The micronutrient content, however, was adjusted so that chil-dren in both FS groups received similar daily micronutrientdoses. Table 1 provides the calories and nutrient contents ofa daily ration of each supplementation scheme.

Both FS50 and FS25 could be eaten as such, whereas theLP required cooking into porridge before consumption. Guard-ians were provided with spoons and were advised to daily of-fer their infants porridge containing 12 spoonfuls of LP, 8 spoon-fuls of FS50, or 4 spoonfuls of FS25, divided into 2 to 3 dailydoses. All mothers were encouraged to continue breastfeedingon demand and to feed their infants only as much of the foodsupplement as the infants wanted to consume at a time.

Participants were visited weekly at their homes to collect in-formation on supplement use and possible adverse events. Emptyfood containers were collected every 3 weeks. At 17, 34, and 52weeks after enrollment, the participants underwent a physical ex-amination, an anthropometric assessment, and laboratory tests.

MEASUREMENT OF OUTCOME VARIABLES

The primary outcome was weight gain during 12-month follow-up. Secondary outcomes included length gain; mean change in

Table 1. Caloric and Nutrient Contents of a Daily Rationof Each Food Supplement Used in This Trial

Variable

Intervention Group

LP FS50 FS25

Weight, g 71 50 25Caloric intake, kcal 282 256 127Protein, g 10.3 7.0 3.5Carbohydrates, g NA 13.8 6.6Fat, g 3.1 16.9 8.5Retinol, µg RE 138 400 400Folate, µg 43 160 160Niacin, mg 3 6 6Pantothenic acid, mg NA 2 2Riboflavin, mg 0.3 0.5 0.5Thiamin, mg 0.1 0.5 0.5Vitamin B6, mg 0.3 0.5 0.5Vitamin B12, µg 0.9 0.9 0.9Vitamin C, mg 48 30 30Vitamin D, µg NA 5 5Calcium, mg 71 366 283Copper, mg NA 0.4 0.4Iodine, µg NA 135 135Iron, mg 5 8 8Magnesium, mg NA 60 60Selenium, µg NA 17 17Zinc, mg 3.6 8.4 8.4

Abbreviations: FS25, fortified spread, 25 g/d; FS50, fortified spread,50 g/d; LP, likuni phala; NA, not available; RE, retinol equivalents.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM620

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

the anthropometric indexes of weight-for-age (WAZ), length-for-age (LAZ), and weight-for-length (WLZ) z scores; the inci-dence of severe (WAZ, LAZ, and WLZ �−3) or moderate to se-vere (WAZ, LAZ, and WLZ �−2) underweight, stunting, orwasting; change in head or middle upper arm circumference; andchange in blood hemoglobin and serum ferritin concentrations.

Unclothed infants were weighed using an electronic infantweighing scale (SECA 834; Chasmors Ltd, London, England),and weights were recorded to the nearest 10 g. Length was mea-sured to the nearest 1 mm using a high-quality length board(Kiddimetre; Raven Equipment Ltd, Essex, England). Middleupper arm circumference and head circumference were mea-sured using nonstretchable plastic tape measures (Lasso-o Tape;Harlow Printing Ltd, South Shields, Tyne & Wear, England).Anthropometric indexes (WAZ, LAZ, and WLZ) were calcu-lated using Epi Info 3.3.2 software (Centers for Disease Con-trol and Prevention, Atlanta, Georgia), based on the Centersfor Disease Control and Prevention 2000 growth reference.18

A 2-mL venous blood sample was collected, and serum wasseparated by means of centrifugation at the beginning and endof follow-up. Hemoglobin concentration was measured froma fresh blood drop using cuvettes and a reader (HemoCue AB,Angelholm, Sweden). Serum ferritin concentration was ana-lyzed from frozen serum samples using commercial test kitsaccording to the manufacturer’s instructions (Ramco Labora-tories, Stafford, Texas).

Thick and thin blood smears were stained with Giemsa andscreened microscopically for malaria parasites from sympto-matic participants at enrollment. Screening for human immu-nodeficiency virus infection was performed from fresh bloodusing 2 antibody rapid tests, and positive results were con-firmed using DNA amplification technology according to themanufacturers’ instructions (Determine; Abbott Laboratories,Abbott Park, Illinois; Uni-Gold; Trinity Biotech Plc, Bray, Ire-land; and Amplicor HIV-1 Monitor Test Version 1.5; Hoff-mann–La Roche Ltd, Basel, Switzerland).

DATA MANAGEMENT AND ANALYSIS

Collected data were recorded on paper forms, transcribed topaper case report forms, and double entered into a tailor-made database (Microsoft Access 2003; Microsoft Corp, Red-mond, Washington). The 2 entries were electronically com-pared, and extreme or otherwise suspicious values wereconfirmed or corrected.

Statistical analysis was performed using Stata 9.0 (Stata Corp,College Station, Texas) on an intention-to-treat basis. Infants withno anthropometric data after enrollment were excluded from out-come analyses but were included in the comparison of baselinecharacteristics. The main analyses on anthropometric measuresused data from all 176 analyzable children using the last valuescarried forward (and back-transform WAZ and LAZ at 18 monthsto kilograms and centimeters for metric presentation) for the 8children who dropped out early or using the survival analysismethod to deal with censoring. Sensitivity analysis limited to the168 children with complete data gave similar findings (details notshown).

For continuous and categorical outcomes, the 3 interven-tion groups were compared using analysis of variance and theFisher exact test, respectively. Survival analysis was used to de-termine the cumulative probability of severe or moderate mal-nutrition in different groups, and the differences were testedby means of the log-rank test. An event was considered to havehappened at the midpoint between the time the event was de-tected and the previous measurement. Individuals with a par-ticular form of malnutrition already at enrollment were ex-cluded from survival analyses concerning the incidence of that

outcome. For the studies of compliance using visits as the unitof analysis, the Huber-White robust standard error was usedto allow for correlated data (multiple visits per child).

ETHICS, STUDY REGISTRATION,AND PARTICIPANT SAFETY

This trial was performed according to International Confer-ence on Harmonisation/Good Clinical Practice guidelines, andit adhered to the principles of the Declaration of Helsinki andregulatory guidelines in Malawi. Before the onset of enroll-ment, the trial protocol was reviewed and approved by the Uni-versity of Malawi College of Medicine research and ethics com-mittee and the ethical committee of Pirkanmaa Hospital District(Finland). Key details of the protocol were published at the clini-cal trial registry of the National Library of Medicine.

A data safety and monitoring board continuously moni-tored the incidence of suspected serious adverse events, de-fined as any untoward medical occurrence that either resultedin death or was life threatening or required inpatient hospital-ization or prolongation of existing hospitalization or resultedin persistent or significant disability or incapacity or other se-rious medical conditions.

RESULTS

Of the 303 initially screened infants, 65 were too old (aged�6.99 months), 2 were too young (aged �5.5 months),and 2 were too ill on the day of screening or enrollment.Of the remaining 234 infants, 49 were not brought to theenrollment session (3 died, 16 moved away, 7 parents werenot interested, and 23 parents gave no explanation) and3 parents declined participation after receiving full infor-mation about the trial. The remaining 182 infants were ran-domized into 3 intervention groups (Figure 1). None ofthe eligible participants who received the 6-g test dose wereallergic to FS.

Table2 provides the baseline characteristics of the par-ticipants by intervention group. At enrollment, the meananthropometric measurements were comparable in the LPand FS50 groups, whereas infants in the FS25 group werea mean of 250 and 380 g heavier than those in the FS50and LP groups, respectively (Table 2). Infants in the FS25group were also 0.3 and 0.7 cm longer than infants in theother 2 groups, respectively. No participant experiencedsevere wasting at the beginning. The prevalence of severeunderweight in the LP, FS50, and FS25 groups was 1.6%(n=1), 3.3% (n=2), and 0.0%, respectively; and that of se-vere stunting was 3.3% (n=2), 6.6% (n=4), and 1.7%(n=1), respectively. Nine participants had a positive hu-man immunodeficiency virus antibody test result at en-rollment, but only 1 of them was truly human immuno-deficiency virus infected, as evidenced by a positivepolymerase chain reaction test result.

During the 12-month follow-up, 10 infants died, 2 werenot located at the end of follow-up, and 2 were unavail-able for follow-up before the age of 18 months (Figure 1).The success rate of following up to age 18 months was notsignificantly different among intervention groups (P=.47,Fisher exact test). Only 6 participants had no anthropo-metric data at all after enrollment, and there was no dif-ference in this among intervention groups (P=.70, Fisherexact test). The assumed causes of the 10 deaths were di-

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM621

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

arrhea, malaria, and drowning in the LP group; diarrhea,meningitis, malaria, and poisoning in the FS50 group; andrespiratory tract infection and malaria in the FS25 group.

All mothers reported that their infants readily ate theprovided supplement, and the diversion of any portionto someone other than the intended beneficiary was re-ported at only 4 of 8864 food delivery interviews (0.05%),3 in the LP group and 1 in the FS50 group. From the3-week home visits during which leftover trial products

were checked, the percentages of visits with leftoversfound were 2.8%, 9.8%, and 5.6% in the LP, FS50, andFS25 groups, respectively (P� .001).

Of the 176 participants with anthropometric data, meangains in weight and length were 100 g (95% confidenceinterval [CI], −143 to 343 g) and 0.8 cm (95% CI, −0.1to 1.7 cm), respectively, higher in the FS50 group thanin the LP group. Correspondingly, mean decreases in WAZand LAZ were smaller in FS50 infants than in FS25 or

With ≥ 1 follow-up visit58With ≥ 1 follow-up visit58 With ≥ 1 follow-up visit60

Infants received FS2560Infants received LP61 Infants received FS5061

Infants screened303

Infants invited forenrollment session

241

Randomized182

Finished follow-up57Finished follow-up57 Finished follow-up54

Death before 12 mo1Death before 12 mo1 Deaths and 2 dropoutsbefore 8 mo

4

Death and 1 dropoutbefore 4 mo

1Deaths before 4 mo3 Dropout before 4 mo1

Were aged > 6.99 mo62

Did not show up49Were ill at enrollment2Parents refused to participate

3

Were aged > 6.99 mo3Were aged > 5.50 mo2

Figure 1. Flow of participants in the study. FS25 indicates fortified spread, 25 g/d; FS50, fortified spread, 50 g/d; and LP, likuni phala, 72 g/d.

Table 2. Baseline Characteristics of the 182 Participants at Enrollment

Variable

Intervention Group

LP(n=61)

FS50(n=61)

FS25(n=60)

Male sex, No./total No. (%) 24/61 (39.3) 33/61 (54.1) 34/60 (56.7)PCR-confirmed HIV infection, No./total No. (%) 0/55 0/55 1/55 (1.8)HIV antibodies, No./total No. (%) 3/55 (5.5) 2/55 (3.6) 4/55 (7.3)Clinical malaria, No./total No. (%)a 1/61 (1.6) 1/61 (1.6) 2/60 (3.3)Children aged �5 y per participant household, mean (SD) 2 (0.9) 2 (0.8) 2 (0.9)Age, mean (SD), mo 5.91 (0.41) 5.93 (0.44) 5.89 (0.36)Weight, mean (SD), kg 6.92 (0.93) 7.05 (0.90) 7.30 (0.92)Length, mean (SD), cm 62.8 (2.1) 63.2 (2.6) 63.5 (2.4)Middle upper arm circumference, mean (SD), cm 13.4 (0.9) 13.5 (1.1) 13.9 (1.1)Head circumference, mean (SD), cm 43.0 (1.5) 43.1 (1.7) 43.2 (1.3)Weight-for-age z score, mean (SD) −0.65 (1.07) −0.62 (1.04) −0.33 (0.94)Length-for-age z score, mean (SD) −1.20 (0.82) −1.20 (1.01) −1.00 (0.77)Weight-for-length z score, mean (SD) 0.48 (1.08) 0.55 (0.90) 0.75 (0.86)Blood hemoglobin concentration, mean (SD), g/dL 11.4 (1.6) 10.6 (1.7) 11.3 (1.5)Serum ferritin concentration, mean (SD), ng/mL 56.9 (73.6) 45.7 (37.4) 67.2 (74.4)

Abbreviations: FS25, fortified spread, 25 g/d; FS50, fortified spread, 50 g/d; HIV, human immunodeficiency virus; LP, likuni phala; PCR, polymerase chain reaction.SI conversion factors: To convert ferritin to picomoles per liter, multiply by 2.247; hemoglobin to grams per liter, multiply by 10.aReported fever and observed peripheral blood malaria parasitemia.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM622

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

LP infants, and there was also a smaller decrease in bloodhemoglobin concentration in the FS50 group. None ofthe differences, however, reached statistical signifi-cance (Table 3).

There was an interaction between intervention group(FS50 vs LP) and baseline LAZ using length gain (P=.04)or weight gain (P=.002) as an outcome. In infants with base-line LAZ below the median in this trial (−1.04), the meangain in length was 1.9 cm (95% CI, 0.3-3.5 cm) or 0.40 zscore (95% CI, −0.15 to 0.95 z score) bigger in the FS50group than in the LP group. Comparable differences inweight gain were 404 g (95% CI, 74-735 g) or 0.43 z score(95% CI, 0.07-0.80 z score). In infants with baseline LAZabove the median, the difference in length was −0.4 cm (95%CI, −1.3 to 0.5 cm) or −0.09 z score (95% CI, −0.58 to 0.39z score); the difference in weight was −254 g (95% CI, −604to 96 g) or −0.25 z score (95% CI, −0.64 to 0.14 z score).There was no significant interaction between interventionand baseline WAZ (P=.14).

As secondary end points, we looked at the propor-tion of infants who developed severe or moderate to se-vere underweight, stunting, and wasting during follow-up. The proportion of infants who developed other formsof malnutrition did not differ markedly among the in-tervention groups, but severe stunting occurred signifi-cantly less frequently in the FS50 and FS25 groups thanin the LP group (Table 4). After enrollment, no infants

in the FS50 group, 3.5% in the FS25 group, and 12.5%in the LP group developed severe stunting (P=.008, Fisherexact test). The 95% CI for the difference between theFS50 and LP groups was 3.8% to 21.2%.

The cumulative incidence of stunting during the 12-month period was also calculated based on survival analy-sis methods that dealt with censoring differently from theanalysis given in Table 4. This approach confirmed thatsevere stunting developed less often and later in the FS50and FS25 groups than in the LP group (0.0%, 3.5%, and13.3%, respectively; P=.01, log-rank test) (Figure 2A).The cumulative incidence of moderate to severe stunt-ing was similar in the 3 groups, but infants in the FS50group developed the condition on average somewhat later.However, the latter result was not significant (P=.66, log-rank test) (Figure 2B). Calculated from the absolute riskreduction, the number of infants needed to be suppliedfor 1 year with FS50 to prevent 1 case of severe stuntingwas 8 (95% CI, 5-26).

The proportion of those developing severe and mod-erate to severe underweight and severe and moderate tosevere wasting did not differ significantly between thetrial groups (Table 4). The cumulative incidence of se-vere underweight was 15.0%, 22.5%, and 16.9% for theLP, FS50, and FS25 groups, respectively (P=.71); the cu-mulative incidence of severe wasting was 1.8%, 1.9%, and1.8% for the same groups, respectively (P� .99).

Table 3. Outcome Changes in Infants Receiving Different Doses of FS and LP During Up to 12 Months of Follow-up

Variable

Intervention Group

P ValueaLP FS50 FS25

Change in weight, mean (SD), kg 2.37 (0.60) 2.47 (0.77) 2.37 (0.61) .66Change in length, mean (SD), cm 12.7 (1.7) 13.5 (2.9) 13.2 (2.9) .23Change in middle upper arm circumference, mean (SD), cm 1.1 (0.9) 1.0 (1.1) 1.0 (0.8) .64Change in head circumference, mean (SD), cm 3.7 (0.5) 3.7 (0.8) 3.7 (0.6) .96Change in weight-for-age z score, mean (SD) −1.29 (0.63) −1.18 (0.90) −1.32 (0.65) .53Change in length-for-age z score, mean (SD) −0.74 (0.95) −0.59 (1.22) −0.64 (0.86) .71Change in weight-for-length z score, mean (SD) −0.98 (0.83) −1.05 (0.86) −1.13 (0.75) .62Change in blood hemoglobin concentration, mean (SD), g/dL −0.68 (2.06) −0.04 (2.21) −0.38 (1.82) .28Change in serum ferritin concentration, mean (SD), ng/mL 16.2 (116.4) 11.5 (61.2) -0.7 (172.4) .83

Abbreviations: FS, fortified spread; FS25, FS, 25 g/d; FS50, FS, 50 g/d; LP, likuni phala.SI conversion factors: To convert ferritin to picomoles per liter, multiply by 2.247; hemoglobin to grams per liter, multiply by 10.aObtained by analysis of variance.

Table 4. Proportion of Participants Developing Various Forms of Undernutrition During Trial Follow-up

Variable

Intervention Group

P ValueaLP FS50 FS25

Ever developed severe stunting (LAZ �−3), No./total No. (%) 7/56 (12.5) 0/56 2/58 (3.5) .008Ever developed moderate to severe stunting (LAZ �−2), No./total No. (%) 11/49 (22.5) 10/51 (19.6) 15/55 (27.3) .64Ever developed severe underweight (WAZ �−3), No./total No. (%) 8/57 (14.0) 11/58 (19.0) 9/58 (15.5) .81Ever developed moderate to severe underweight (WAZ �−2), No./total No. (%) 23/53 (43.4) 20/56 (35.7) 20/57 (35.1) .62Ever developed severe wasting (WHZ �−3), No./total No. (%) 1/58 (1.7) 1/60 (1.7) 1/58 (1.7) �.99Ever developed moderate to severe wasting (WHZ �−2), No./total No. (%) 5/58 (8.6) 7/60 (11.7) 3/58 (5.2) .47

Abbreviations: FS25, fortified spread, 25 g/d; FS50, fortified spread, 50 g/d; LAZ, length-for-age z score; LP, likuni phala; WAZ, weight-for-age z score;WHZ, weight-for-height z score.

aObtained by Fisher exact test.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM623

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

All the interventions were well tolerated by the partici-pants. In addition to the 10 deaths (4, 4, and 2 in the LP,FS50, and FS25 groups, respectively) (Figure 1), 3 otherparticipants (1 in each intervention group) were hospital-ized and recorded as having experienced a serious adverseevent during follow-up. The LP, FS50, and FS25 groupsshared 5, 5, and 3 of these serious adverse events, respec-tively (P=.82, Fisher exact test). Three of the serious ad-verse events were considered to be unrelated, and 10 wereprobably unrelated to the trial interventions.

COMMENT

The present trial was performed to test the growth-promotingeffectsof2micronutrient-fortified,calorie-dense,ready-to-usespreads(FS50andFS25)whenusedascomple-mentary foods for 6- to 18-month-old infants in ruralMalawi. The enrollment rate for the trial was high, groupallocation was random, follow-up was identical for allgroups, loss to follow-up and unavailability for follow-upwere infrequent, and the people measuring the outcomeswere masked to group allocation. Hence, the observed re-sults are likely to be unbiased and, thus, representative ofthe population from which the sample was drawn.

In this sample, infants receiving FS50 or FS25 for 1 yeargained on average slightly more weight and length thanthose receiving an approximately isoenergetic portion ofLP. However, the mean differences between the FS50 andLP groups were modest (100 g and 0.8 cm), and statisticalhypothesis testing could not exclude the possibility of ran-dom findings. Therefore, the data do not support the ini-tial hypothesis that 1 year of complementary provision ofFS to all infants older than 6 months in rural Malawi wouldbe noticeably better than LP in promoting their mean weightor length gain by 18 months of age. Instead, the results seemto be mostly in line with the modest findings from other

dietary intervention trials19-24 for infants in similar set-tings, documenting a 0- to 0.6-SD higher mean weight (SD,0-400 g) or length (SD, 0-1 cm) gains in the interventiongroups than in unsupplemented controls.

In contrast to the mean gains, the present study dem-onstrated a marked and statistically significant differ-ence in the incidence of severe stunting between the FS50and LP groups. Also, a stratified analysis suggested aninteraction between initial height and intervention group,as demonstrated by bigger between-group differences inlength and weight gain in infants with at least mild stunt-ing at enrollment. Caution must be exercised when in-terpreting these results because the incidence of severestunting was only one of several secondary outcomes ofthe trial and the interaction finding came from a post hocexploratory analysis. The results are, however, biologi-cally plausible, and there was a trend toward a dose re-sponse. Therefore, we believe that the results in thissample appropriately represent the larger population. Fur-thermore, the results are consistent with those of an ear-lier trial from West Africa25 in which supplementationof children with FS who experienced stunting inducedmarked catch-up growth among them. These data sug-gest that a dietary intervention with FS can boost lineargrowth and reduce the incidence of the severest formsof stunting, especially when directed to initially disad-vantaged infants. If the whole unselected infant groupneeds to be targeted, the intervention should probablyalso address the other major risk factors for growth fail-ure (infections and inappropriate child care).26

In the present sample, 8 infants needed supplemen-tation for 1 year with FS50 to prevent 1 case of severestunting. It is difficult to assess the public health impactresulting from such an intervention and the potential re-duction in severe stunting. Stunting is associated withvarious adverse sequelae, such as developmental delay,

0.20

0.05

0

0.15

0.10

1230 96Time on Trial, mo

Cum

ulat

ive

Inci

denc

e

LPFS25FS50

A

0.30

0.35

0.05

0

0.15

0.20

0.25

0.10

1230 96Time on Trial, mo

Cum

ulat

ive

Inci

denc

e

B

Figure 2. Cumulative incidence functions of severe stunting (A) and moderate to severe stunting (B) in children in the likuni phala (LP), fortified spread, 25 g/d(FS25), and fortified spread, 50 g/d (FS50) groups.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM624

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

lower work capacity, worse economic status as an adult,and delivery problems.27 However, it is not known whetherthese outcomes are particularly associated with any lin-ear growth failure or only its most severe forms. Such as-sociations and the possibility of preventing various healthconsequences with an FS intervention need to be ad-dressed in later trials. Such studies are further justifiedby recent evidence from a Ghanaian trial, suggesting thatinfant motor development can be markedly enhanced ina low-income setting by a 6-month-long provision of Nu-tributter, another lipid-based nutrient supplement.24

Although we collected information on compliance, itcame from parental recollections, which are often unre-liable in dietary interventions. Hence, we cannot rule outthe possibility of food sharing and its potential effect onthe results, especially because the 2 supplement types dif-fered in many ways from each other. In fact, earlier stud-ies from the same research site suggest that LP and FSare shared within families, but this occurs more often withLP.6,28 In this respect, this trial, thus, looked at effective-ness (ie, infant outcomes after the provision of differentfood supplements to the family), rather than assessinggrowth under ideal conditions in which participants trulyate all the supplements that were intended for them.

Another major limitation of the trial is the lack of anonsupplemented control group; therefore, we can makesolid conclusions only on the relative value of FS and LPsupplementation but not on their independent effect. Ina previous study17 conducted 10 years earlier in the samearea, the differences in the 50th percentile weight andlength between 6 and 18 months of age were 2.3 kg and11.9 cm, respectively, in unsupplemented infants (ie, 0.1kg and 0.8 cm less than mean gains for the LP controlgroup in the present trial). However, the different an-thropometric methods used in the earlier trial and its his-torical nature limit its use for conclusive comparisons.

The ex-factory price of FS, when made in Malawi, isapproximately US $4 per kilogram, corresponding to adaily cost of US $0.20 for a 50-g dose. Unsubsidized, thisprice is presumably unaffordable to many families in ru-ral Malawi. The present results with FS25, and those froma recent Ghanaian trial24 using a daily dose of 20 g, sug-gest, however, that a lower and markedly cheaper dosemay be as efficient in growth and development promo-tion as the FS50 dose used in the present trial. Other po-tential means for increased affordability include amend-ments to the FS recipe (eg, a cheaper protein source) andsocial marketing, both of which are being tested in thesub-Saharan African setting.

Taken together, these results suggest that 1-year-long complementary feeding with FS does not have amarkedly larger effect than LP supplementation on meanweight gain in all infants, but it seems to boost lineargrowth in the most disadvantaged individuals and, hence,decrease the incidence of severe stunting in a popula-tion in which it is otherwise common. A larger and longertrial, however, is needed to confirm the finding and tolook at the effect on outcomes other than growth, to ana-lyze whether LP supplementation has some health im-pact, to see whether the growth effect persists beyond theage of 18 months, and to guide policy decisions on theuse of spreads or corn–soy flour in malnutrition preven-

tion in Malawi and similar settings. Further studies onthe possible mechanisms of the growth-promoting effectof FS are also warranted.

Accepted for Publication: November 7, 2007.Author Affiliations: Community Health Department, Col-lege of Medicine, University of Malawi, Blantyre(Drs Phuka and Maleta and Ms Thakwalakwa); Depart-ment of International Health, University of Tampere Medi-cal School, Tampere, Finland (Drs Phuka and Ashorn);Department of Epidemiology and Public Health, Lon-don School of Hygiene and Tropical Medicine, London,England (Dr Cheung); Department of Child Health andDevelopment, World Health Organization, Geneva, Swit-zerland (Dr Briend); Departement Societes et Sante, IRD,Paris, France (Dr Briend); Department of Pediatrics, Wash-ington University School of Medicine, St Louis, Mis-souri (Dr Manary); and Department of Paediatrics, Tam-pere University Hospital, Tampere (Dr Ashorn).Correspondence: John C. Phuka, MBBS, College of Medi-cine, University of Malawi, PO Box 431, Mangochi, Malawi(johnphuka@malawi.net).Author Contributions: Dr Phuka had full access to allof the data in the study and takes responsibility for theintegrity of the data and the accuracy of the data analy-sis. Study concept and design: Phuka, Maleta, Thakwalakwa,Briend, Manary, and Ashorn. Acquisition of data: Phuka,Maleta, and Thakwalakwa. Analysis and interpretation ofdata: Phuka, Maleta, Thakwalakwa, Cheung, Manary, andAshorn. Drafting of the manuscript: Phuka, Maleta,Thakwalakwa, Briend, and Ashorn. Critical revision of themanuscript for important intellectual content: Phuka, Maleta,Thakwalakwa, Cheung, Manary, and Ashorn. Statisticalanalysis: Phuka, Cheung, Briend, and Ashorn. Obtainedfunding: Ashorn. Administrative, technical, or material sup-port: Phuka, Maleta, Thakwalakwa, Manary, and Ashorn.Study supervision: Maleta, Manary, and Ashorn.Financial Disclosure: Dr Briend was a consultant to Nu-triset Inc until December 2003, and the company has alsofinancially supported the planning of another researchproject by the same study team through Dr Ashorn andthe University of Tampere after the completion of this trial.Funding/Support: This study was supported by grants200720 and 109796 from the Academy of Finland, theFoundation for Paediatric Research in Finland, andthe Medical Research Fund of Tampere UniversityHospital.Role of the Sponsor: The funding bodies had no role inthe design and conduct of the study; in the collection,analysis, and interpretation of the data; or in the prepa-ration, review, or approval of the manuscript.Disclaimer: The authors alone are responsible for theviews expressed in this publication, and they do not nec-essarily represent the decisions or the stated policy of theWorld Health Organization.Additional Contributions: The people of Lungwena, thestaff at the Lungwena Training Health Centre, and theresearch assistants provided positive attitudes, support,and help in all stages of the study; Laszlo Csonka de-signed the collection tools and data entry programs; andNutriset Inc provided the micronutrient mixture used inthe production of FS free of charge.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM625

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

REFERENCES

1. National Statistical Office (Malawi) and OCR Macro. Malawi Demographic andHealth Survey 2000. Zomba, Malawi: National Statistical Office and OCR Macro;2001.

2. Maleta K, Virtanen S, Espo M, Kulmala T, Ashorn P. Child malnutrition and itspredictors in rural Malawi. Paediatr Perinat Epidemiol. 2003;17(4):384-390.

3. Pelletier DL, Frongillo EA, Schroeder DG, Habicht JP. The effects of malnutritionon child mortality in developing countries. Bull World Health Organ. 1995;73(4):443-448.

4. Caulfield LE, de Onis M, Blossner M, Black RE. Undernutrition as an underlyingcause of child deaths associated with diarrhea, pneumonia, malaria, and measles.Am J Clin Nutr. 2004;80(1):193-198.

5. United Nations Administrative Committee on Coordination/Sub-Committee onNutrition (ACC/SCN); Allen LH, Gillespie SR. What Works? A Review of theEfficacy and Effectiveness of Nutrition Interventions. Geneva, Switzerland: ACC/SCN in collaboration with the Asian Development Bank (Manila, the Philip-pines); 2001.

6. Maleta K, Kuittinen J, Duggan MB, et al. Supplementary feeding of underweight,stunted Malawian children with a ready-to-use food. J Pediatr Gastroenterol Nutr.2004;38(2):152-158.

7. Briend A, Lacsala R, Prudhon C, Mounier B, Grellety Y, Golden MH. Ready-to-use therapeutic food for treatment of marasmus. Lancet. 1999;353(9166):1767-1768.

8. Briend A. Highly nutrient-dense spreads: a new approach to delivering multiplemicronutrients to high-risk groups. Br J Nutr. 2001;85(suppl 2):S175-S179.

9. Collins S. Changing the way we address severe malnutrition during famine. Lancet.2001;358(9280):498-501.

10. Diop el HI, Dossou NI, Ndour MM, Briend A, Wade S. Comparison of the efficacyof a solid ready-to-use food and a liquid milk-based diet for the rehabilitation ofseverely malnourished children: a randomized trial. Am J Clin Nutr. 2003;78(2):302-307.

11. Manary MJ, Ndekha M, Ashorn P, Maleta K, Briend A. Home-based therapy for se-vere malnutrition with ready-to-use food. Arch Dis Child. 2004;89(6):557-561.

12. Sandige H, Ndekha MJ, Briend A, Ashorn P, Manary MJ. Home-based treatmentof malnourished Malawian children with locally produced or imported ready-to-use food. J Pediatr Gastroenterol Nutr. 2004;39(2):141-146.

13. Ciliberto MA, Sandige H, Ndekha MJ, et al. Comparison of home-based therapywith ready-to-use therapeutic food with standard therapy in the treatment of mal-nourished Malawian children: a controlled, clinical effectiveness trial. Am J ClinNutr. 2005;81(4):864-870.

14. Ndekha MJ, Manary MJ, Ashorn P, Briend A. Home-based therapy with ready-to-use therapeutic food is of benefit to malnourished, HIV infected Malawianchildren. Acta Paediatr. 2005;94(2):222-225.

15. Ciliberto MA, Manary MJ, Ndekha MJ, Briend A, Ashorn P. Home-based therapyfor edematous malnutrition with ready-to-use therapeutic food. Acta Paediatr.2006;95(8):1012-1015.

16. Kuusipalo H, Maleta K, Briend A, Manary M, Ashorn P. Growth and change inblood haemoglobin concentration among underweight Malawian infants receiv-ing fortified spreads for 12 weeks: a preliminary trial. J Pediatr Gastroenterol Nutr.2006;43(4):525-532.

17. Maleta K, Virtanen S, Espo M, Kulmala T, Ashorn P. Timing of growth falteringin rural Malawi. Arch Dis Child. 2003;88(7):574-578.

18. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the UnitedStates: methods and development. Vital Health Stat 11. 2002;(246):1-190.

19. Simondon KB, Gartner A, Berger J, et al. Effect of early, short-term supplemen-tation on weight and linear growth of 4-7-mo-old infants in developing coun-tries: a four-country randomized trial. Am J Clin Nutr. 1996;64(4):537-545.

20. Lartey A, Manu A, Brown KH, Peerson JM, Dewey KG. A randomized, community-based trial of the effects of improved, centrally processed complementary foodson growth and micronutrient status of Ghanaian infants from 6 to 12 mo of age.Am J Clin Nutr. 1999;70(3):391-404.

21. Beckett C, Durnin JV, Aitchison TC, Pollitt E. Effects of an energy and micronu-trient supplementation on anthropometry in undernourished children in Indonesia.Eur J Clin Nutr. 2000;54(suppl 2):S52-S59.

22. Bhandari N, Bahl R, Nayyar B, Khokhar P, Rohde JE, Bhan MK. Food supple-mentation with encouragement to feed it to infants from 4 to 12 months of agehas a small impact on weight gain. J Nutr. 2001;131(7):1946-1951.

23. Oelofse A, Van Raaij JM, Benade AJ, Dhansay MA, Tolboom JJ, Hautvast JG.The effect of a micronutrient-fortified complementary food on micronutrient sta-tus, growth and development of 6- to 12-month-old disadvantaged urban SouthAfrican infants. Int J Food Sci Nutr. 2003;54(5):399-407.

24. Adu-Afarwuah S, Lartey A, Brown KH, Briend A, Zlotkin Z, Dewey KG. Random-ized comparison of 3 types of micronutrient supplements for home fortificationof complementary foods in Ghana: effects on growth and motor development.Am J Clin Nutr. 2007;86(2):412-420.

25. Lopriore C, Guidoum Y, Briend A, Branca F. Spread fortified with vitamins andminerals induces catch-up growth and eradicates severe anemia in stunted refu-gee children age 3-6 y. Am J Clin Nutr. 2004;80(4):973-981.

26. Martorell R, Kettel KL, Schroeder DG. Reversibility of stunting: epidemiologicalfindings in children from developing countries. Eur J Clin Nutr. 1994;48(suppl1):S45-S57.

27. Grantham-McGregor S, Cheung YB, Cueto S, Glewwe P, Richter L, Strupp B.Developmental potential in the first 5 years for children in developing countries.Lancet. 2007;369(9555):60-70.

28. Flax VL, Ashorn U, Phuka J, Maleta K, Manary MJ, Ashorn P. Feeding patternsof underweight children in rural Malawi given supplementary fortified spread athome. Matern Child Nutr. 2008;4(1):65-73.

Announcement

Sign Up for Alerts—It’s Free! Archives of Pediatrics &Adolescent Medicine offers the ability to automatically re-ceive the table of contents of Archives via e-mail when itis published online. This also allows you to link to in-dividual articles and view the abstract. It makes keep-ing up-to-date even easier! Go to http://pubs.ama-assn.org/misc/alerts.dtl to sign up for this free service.

(REPRINTED) ARCH PEDIATR ADOLESC MED/ VOL 162 (NO. 7), JULY 2008 WWW.ARCHPEDIATRICS.COM626

©2008 American Medical Association. All rights reserved. on July 8, 2008 www.archpediatrics.comDownloaded from

Postintervention growth of Malawian children who received 12-modietary complementation with a lipid-based nutrient supplement ormaize-soy flour1–4

John C Phuka, Kenneth Maleta, Chrissie Thakwalakwa, Yin Bun Cheung, Andre Briend, Mark J Manary, and Per Ashorn

ABSTRACTBackground: Therapeutic feeding with micronutrient-fortifiedlipid-based nutrient supplements (LNSs) has proven useful in therehabilitation of severely malnourished children. We recently re-ported that complementary feeding of 6–18-mo-old infants withan LNS known as FS50 was associated with improved linear growthand a reduction in the incidence of severe stunting during the sup-plementation period.Objective: Our objective was to assess whether a reduction instunting seen with 12-mo LNS supplementation was sustained overa subsequent 2-y nonintervention period.Design: One hundred eighty-two 6-mo-old healthy rural Malawianinfants were randomly assigned to receive daily supplementation for12 mo with 71 g of maize-soy flour [likuni phala (LP); controlgroup, 282 kcal] or either 50 g of FS50 (264 kcal; main interventiongroup), or 25 g of FS25 (130 kcal). Main outcome measures wereincidence of severe stunting and mean z score changes in weight-for-age, length-for-age, and weight-for-length during a 36-mo follow-up period.Results: The cumulative 36-mo incidence of severe stunting was19.6% in LP, 3.6% in FS50, and 10.3% in FS25 groups (P ¼ 0.03).Mean weight-for-age changes were 21.09, 20.76, and 21.22 (P ¼0.04); mean length-for-age changes were 20.47, 20.37, and 20.71(P¼ 0.10); and mean weight-for-length changes were 21.52, 21.18,and 21.48 (P ¼ 0.27). All differences were more marked amongindividuals with baseline length-for-age below the median. Differ-ences in length developed during the intervention at age 10–18 mo,whereas weight differences continued to increase after the inter-vention.Conclusions: Twelve-month-long complementary feeding with 50g/d FS50 is likely to have a positive and sustained impact on theincidence of severe stunting in rural Malawi. Half-dose interventionmay not have the same effect. This trial was registered at clinical-trials.gov as NCT00131209. Am J Clin Nutr 2009;89:382–90.

INTRODUCTION

Stunting, or statural growth failure, affects ’170 millionchildren ,5 y of age, with a prevalence of 40% in southern Asiaand 50% in sub-Saharan Africa (1–5). The condition is associ-ated with many long-term consequences, such as poor cognitiveor school performance, delayed motor development, impairedphysical performance, reduced income in adulthood, obstetricemergencies, and lower birth weight in offspring (4, 6–15).

Because of these adverse sequelae and the persistence ofstunting after the age of 2 y (16, 17), prevention strategies area global health priority. To date, however, few interventions haveproven successful in markedly promoting sustainable lineargrowth of children ,2 y old in low-resource settings (18).

In recent years, the treatment of children with severe acutemalnutrition has drastically changed through the use of lipid-based nutrient supplements (LNSs) (19–25). These productstypically contain milk protein, sugar, and a mixture of micro-nutrients, embedded in a lipid base (26). Because of their pos-itive impact on severely malnourished individuals, ease ofproduction and use, excellent storage properties, and flexibility ofthe fortified micronutrient composition, modified LNSs have alsobeen considered a potential means for the prevention of stunting.Actual research data on this possibility are still scarce, but 3 recentclinical trials from sub-Saharan Africa suggest that provision ofsmall daily doses of complementary LNS to 6–18-mo-old infantscan promote their linear growth and yield other health benefits (27–29). To date, however, no longer term results have been publishedfrom trials using LNSs to prevent stunting.

1 From the College of Medicine, University of Malawi, Blantyre, Malawi

(JCP, KM, and CT); the Department of International Health, University of

Tampere Medical School, Finland (JCP, YBC, and PA); the Clinical Trials

and Epidemiology Research Unit, Singapore (YBC), the Department of Child

Health and Development, World Health Organization, Geneva, Switzerland,

and IRD, Departement Societes et Sante, Paris, France (AB) 5Washington

University School of Medicine, St Louis, MO (MJM); and the Department

of Paediatrics, Tampere University Hospital, Tampere, Finland (PA).2 AB is a staff member of the World Health Organization. The author

alone is responsible for the views expressed in this publication, which do

not necessarily represent the decisions or the stated policy of the World

Health Organization. The funders of this trial had no role in its implementa-

tion, analysis, or reporting.3 Supported by the Academy of Finland (grants 200720 and 109796), the

Foundation for Paediatric Research in Finland, and the Medical Research

Fund of Tampere University Hospital. The micronutrient mixture used in the

production of fortified spread was provided free of charge by Nutriset Inc

(Malaunay, France). JCP and CT received personal stipends from the Nestle

Foundation.4 Reprints not available. Address correspondence to J Phuka, College of

Medicine, University of Malawi, PO Box 431, Mangochi, Malawi. E-mail:

johnphuka@malawi.net.

Received June 3, 2008. Accepted for publication October 27, 2008.

First published online December 3, 2008; doi: 10.3945/ajcn.2008.26483.

382 Am J Clin Nutr 2009;89:382–90. Printed in USA. � 2009 American Society for Nutrition

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

We recently reported findings from a randomized trial inMalawi, which showed a 0% incidence of severe stunting[length-for-age z score (LAZ) ,23] between 6 and 18 mo ofage among infants receiving daily dietary supplementationwith 50 g LNS for 1 y, compared with 13% among thosereceiving an iso-energetic dose of maize-soy flour and 4%among those receiving 25 g/d LNS (29). In the current article,we extend the follow-up of the same children to the subsequent2-y period with no further dietary intervention. The newanalysis focused on the incidence of stunting as well as meanlength and weight gain between 6 and 42 mo of age.

SUBJECTS AND METHODS

Study area and timing

Thestudywasconductedbetween11October2004and8January2008 in Lungwena, a rural Malawian community with a highprevalence of early childhood stunting and underweight children(1, 30). The staple food, maize, was grown during a single rainyseason between December and March. Exclusive breastfeeding forbabies was almost nonexistent and infant diet was typicallycomplemented with thin maize porridge from 2 to 6 mo of age.

Eligibility criteria, enrollment, and randomization of thetrial participants

Inclusion criteria included age of 5.50–6.99 mo, residence inthe study area, and an informed consent from at least one au-thorized guardian. Exclusion criteria were low weight-for-lengthz score (WLZ; ,22.0), presence of edema, history of peanutallergy, severe illness warranting hospitalization on the enroll-ment day, concurrent participation in another clinical trial, orany symptoms of food intolerance �30 min after the ingestionof a 6-g test dose of the peanut-based LNS used as a trial in-tervention.

For enrollment, trained health surveillance assistants contactedall families who were known to live in the area with a childbetween 5 and 7 mo of age. Infants were invited to an enrollmentsession where they were screened for eligibility, and guardianswere given detailed information on the trial contents.

For group allocation, guardians picked one from a set ofidentically appearing opaque envelopes, each containing a paperindicating an identification number and a randomly assignedallocation to 1 of the 3 interventions. The randomization listand envelopes were made by persons not involved in trial im-plementation, and those assessing the outcomes were blindedthroughout the study.

Interventions and follow-up

There were 3 intervention schemes given for 12 mo from theage of 6 mo, after which the participants were monitored foranother 24 mo without additional intervention. Infants in thecontrol group were provided with an average of 71 g/d ofmicronutrient-fortified maize and soy flour, locally called likuniphala (LP). Participants in the other 2 groups received, on av-erage, either 50 or 25 g/d of an LNS known as fortified spread(FS; FS50 or FS25). The supplements were home delivered at 3weekly intervals (at each food delivery, three 500-g bags of LP,four 262-g jars of FS50, or two 262-g jars of FS25 were pro-vided).

LP was purchased from a local producer (Rab Processors,Limbe, Malawi). LNS was produced at a Malawian nongov-ernmental organization (Project Peanut Butter, Blantyre, Malawi)from peanut paste, milk powder, vegetable oil, sugar, and pre-made micronutrient mixture (Nutriset, Malaunay, France). Allsupplements were fortified with micronutrients, but the level offortification varied between the products. The difference betweenthe FS50 and FS25 supplementation was the amount of food basegiven (50 compared with 25 g/d). The micronutrient content,however, was adjusted so that children in both FS50 and FS25groups received similar daily micronutrient doses. The energyand nutrient contents of a daily ration of each supplementationscheme are shown in Table 1.

FS50 and FS25 could be eaten raw, whereas the maize-soyflour (LP) required cooking into porridge before consumption.The guardians were provided with spoons and advised to dailyoffer their infants porridge containing 12 spoonfuls of LP, 8spoonfuls of FS50, or 4 spoonfuls of FS25, which were dividedinto 2–3 daily doses. All mothers were encouraged to continuebreastfeeding on demand and to feed their infants only as muchof the food supplement as the infant wanted to consume ata feeding.

During the intervention, the participants were visited weekly attheir homes to collect information on supplement use and pos-sible adverse events during the intervention period. Empty foodcontainers were collected every 3 wk. At 4, 8, 12, 18, and 36 moafter enrollment, the participants visited the research office ata nearby health center and underwent a physical examination andanthropometric assessment.

TABLE 1

Energy and nutrient content of a daily ration of each food supplement used in

this trial1

Intervention group

Variable LP FS50 FS25

Weight (g) 71 50 25

Energy (kcal) 282 256 127

Protein (g) 10.3 7.0 3.5

Carbohydrate (g) NA 13.8 6.6

Fat (g) 3.1 16.9 8.5

Retinol (lg RE) 138 400 400

Folate (lg) 43 160 160

Niacin (mg) 3 6 6

Pantothenic acid (mg) NA 2 2

Riboflavin (mg) 0.3 0.5 0.5

Thiamine (mg) 0.1 0.5 0.5

Vitamin B-6 (mg) 0.3 0.5 0.5

Vitamin B-12 (lg) 0.9 0.9 0.9

Vitamin C (mg) 48 30 30

Vitamin D (lg) NA 5 5

Calcium (mg) 71 366 283

Copper (mg) NA 0.4 0.4

Iodine (lg) NA 135 135

Iron (mg) 5 8 8

Magnesium (mg) NA 60 60

Selenium (lg) NA 17 17

Zinc (mg) 3.6 8.4 8.4

1 FS25, fortified spread 25 g/d; FS50, fortified spread 50 g/d; LP, likuni

phala; NA, not available; RE, retinol equivalents.

SUSTAINED GROWTH EFFECT OF COMPLEMENTARY FEEDING 383

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

Measurement of outcome variables

The primary outcome for the initial trial was weight gain.Secondary outcomes included the following: mean changes inlength, head circumference, and mid-upper-arm circumference;the anthropometric indexes weight-for-age z score (WAZ), length-and height-for-age z score (LAZ/HAZ), and WLZ; and the in-cidence of underweight children, wasting, and stunting. In thissecondary phase, we focused the analyses on the incidence ofsevere stunting (LAZ/HAZ , 23 z score units) or moderate-to-severe stunting (LAZ/HAZ , 22 z score units) and changes inthe mean anthropometric indexes.

Weight was measured with naked infants by using an elec-tronic infant weighing scale (SECA 834; Chasmors Ltd, London,United Kingdom) and recorded to the nearest 10 g. Length (at�24 mo of age) and height (at .24 mo) were measured to thenearest 1 mm with a high-quality length board (Kiddimetre;Raven Equipment Ltd, Essex, United Kingdom) and a high-quality stadiometer (Harpenden stadiometer; Child GrowthFoundation, London, United Kingdom). Mid-upper-arm andhead circumferences were measured with nonstretchable plastictapes (Lasso-o tape; Harlow Printing Ltd, South Shields, UnitedKingdom). Anthropometric indexes (WAZ, LAZ, and WLZ) werecalculated with Epi-Info 3.3.2 software (Centers for DiseaseControl and Prevention, Atlanta, GA) with the use of the the CDC2000 growth reference charts (31).

Sample size calculation

Sample size was calculated from expected weight gain at theend of the initial 12-mo intervention period. Assuming a SD of1.40 kg in all intervention groups and a difference of 0.75 kgbetween the means in the main intervention group (FS50) and thecontrol group (LP), a sample size of 55 infants/group was cal-culated to provide the trial with 80% power and 95% confidence.To allow for ’10% attrition, the target enrollment was 60 infants/group.

Data management and analysis

Data were recorded on paper forms, transcribed to paper casereport forms, and entered twice into a custom Microsoft Access2003 database program (Microsoft Corp, Redmond, WA). Thedouble entries were electronically compared, and extreme valueswere confirmed or corrected.

Statistical analysiswasperformedbyusingStata9.0(StataCorp,College Station, TX) on an intention-to-treat basis. Infants withno anthropometric data after enrollment were included only inthe comparison of baseline characteristics. The analyses of an-thropometric measures used data from young children who wereavailableat42moofage;data frominfantsandyoungchildrenwithat least one measurement after enrollment were analyzed for in-cidence of stunting with the use of the survival analysis method tocontrol for censoring.

For continuous and categorical outcomes, the 3 interventiongroups were compared by using analysis of variance (ANOVA)and Fisher’s exact test, respectively. Post hoc pairwise com-parisons for statistically significant intergroup differences wereanalyzed using the Tukey wholly significant difference test.Survival analysis was used to determine the cumulative proba-bility of severe or moderate stunting among different groups,

and the differences were tested by the log-rank test. An eventwas considered to have occurred at midpoint when it occurredbetween the time the event was detected and the previous mea-surement. Individuals with severe or moderate-to-severe stuntingat enrollment were excluded from survival analyses on the in-cidence of that outcome. Effect sizes were calculated by dividingthe difference in mean z scores between FS50 and either FS25 orLP groups by the pooled SDs of the 2 means.

Ethics, study registration, and participant safety

The trial was performed according to International Conferenceof Harmonisation–Good Clinical Practice guidelines (ICH-GCP),and it adhered to the principles of the Declaration of Helsinki andthe regulatory guidelines in Malawi. Before the onset of enroll-ment, the trial protocol was reviewed and approved by the Col-lege of Medicine Research and Ethics Committee (University ofMalawi) and the Ethical Committee of Pirkanmaa HospitalDistrict (Finland). Key details of the protocol were published atthe clinical trial registry of the National Library of Medicine,Bethesda, MD. Before enrollment and again between the initial12-mo intervention and the subsequent 24-mo follow-up, par-ticipant guardians were briefed about the study, and they signeda written consent form for trial participation.

A data safety and monitoring board continuously monitoredthe incidence of suspected serious adverse events during theintervention period, defined as any untoward medical occurrencethat resulted in death, was life threatening, required inpatienthospitalization or prolongation of existing hospitalization, orresulted in a persistent or significant disability or incapacity orother serious medical condition.

RESULTS

Of the 303 initially screened infants, 182 were enrolled in thecomplementary feeding trial. Of these, 10 died and 4 were lostto follow-up during the 12-mo food-supplementation period;thus, 168 participated in the postintervention follow-up. Duringthe 24-mo postintervention follow-up, there were 6 deaths and13 further losses to follow-up, which left 149 participants whocompleted the 36-mo assessment (Figure 1). There was nostatistically significant difference in the proportion of deaths andnumber of dropouts between the food intervention groups (P ¼0.72 and 0.26, respectively).

Selected baseline characteristics of the participants by in-tervention group are shown in Table 2. At enrollment, meananthropometric measurements were comparable in the LP andFS50 groups, whereas infants in the FS25 group were, on av-erage, slightly heavier and taller (Table 2). The prevalences(number of infants) of severe stunting and severe underweight atenrollment in the LP, FS50, and FS25, groups, respectively, wereas follows: severe stunting: 3.3% (n ¼ 2), 6.6% (n ¼ 4), and1.7% (n ¼ 1); severe underweight: 1.6% (n ¼ 1), 3.3% (n ¼ 2),and 0.0% (n ¼ 0).

Figure 2 describes a time-to-event analysis of stunting indifferent intervention groups. By the end of the 12-mo in-tervention period, 13.3% (n ¼ 7), 0.0% (n ¼ 0), and 3.5%(n ¼ 2) of the participants in the LP, FS50, and FS25 groups,respectively, had developed severe stunting (29). In the post-intervention follow-up, 4–7% of the participants in each group

384 PHUKA ET AL

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

developed the condition, which increased the cumulative 3-yincidence (number of cases) of severe stunting to 19.6% (n ¼11) in LP, 3.6% (n ¼ 2) in FS50, and 10.3% (n ¼ 6) in FS25groups (P ¼ 0.03, log-rank test; Figure 2A). The point estimate(95% CI) for the difference between the LP and FS50 groupswas 16% (5–26%), and the number needed to treat or preventone case of severe stunting was 6 (4–20). The cumulative in-cidence of moderate-to-severe stunting was more similar in the 3groups, although infants in the FS50 group tended to develop thecondition on average somewhat later (P ¼ 0.50, log-rank test;Figure 2B).

Among the 149 participants who completed the follow-up,mean3-y gains in weight, height, and mid-upper-arm and head cir-

cumferences were highest in the FS50 group and lowest in the FS25group. Average values for the anthropometric indexes WAZ, HAZ,and weight-for-height z score (WHZ) decreased in all groups, butthe reductions were smallest among the FS50 group and largest inthe FS25 group. However, intergroup differences reached statis-tical significance among unselected children only for WAZ change(Table 3). During the follow-up period, those who received LP orFS25, on average, had larger reductions in their WAZ comparedwith FS50 children: 0.33 (95% CI: 20.39, 0.69) 0.46 (95% CI:0.10, 0.82) z score units for LP and FS25 groups, respectively.Similarly, those who received LP or FS25 had larger reductions intheir HAZ during the follow-up: 0.10 (95% CI: 20.23, 0.43) and0.34 (95% CI: 0.02, 0.67) z score units for LP and FS25 groups,

FIGURE 1. Study flow of participants. LP, likuni phala; FS50, fortified spread 50 g/d; FS25, fortified spread 25 g/d.

TABLE 2

Baseline characteristics of the 182 participants at enrollment1

Intervention group

Variable LP (n ¼ 61) FS50 (n ¼ 61) FS25 (n ¼ 60)

Male sex 24/61 (39.3)2 33/61 (54.1) 34/60 (56.7)

PCR-confirmed HIV infection 0/55 (0.0) 0/55 (0.0) 1/55 (1.8)

Age (mo) 5.91 6 0.413 5.93 6 0.44 5.89 6 0.36

Weight (kg) 6.92 6 0.93 7.05 6 0.90 7.30 6 0.92

Length (cm) 62.8 6 2.1 63.2 6 2.6 63.5 6 2.4

Mid-upper-arm circumference (cm) 13.4 6 0.9 13.5 6 1.1 13.9 6 1.1

Head circumference (cm) 43.0 6 1.5 43.1 6 1.7 43.2 6 1.3

Weight-for-age z score 20.65 6 1.07 20.62 6 1.04 20.33 6 0.94

Length-for-age z score 21.20 6 0.82 21.20 6 1.01 21.00 6 0.77

Weight-for-length z score 0.48 6 1.08 0.55 6 0.90 0.75 6 0.86

1 FS25, fortified spread 25 g/d; FS50, fortified spread 50 g/d; LP, likuni phala; PCR, polymerase chain reaction.2 n/total n; percentage in parentheses (all such values).3 Mean 6 SD (all such values).

SUSTAINED GROWTH EFFECT OF COMPLEMENTARY FEEDING 385

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

respectively. Adjusting the analyses for baseline weight resulted infindings similar to the unadjusted analyses (data not shown).

Because of our earlier observation that there was a strong in-teraction between a child’s length at 6 mo of age and his or herweight and length gains during the intervention (P¼0.002 and P¼0.04, respectively) (29), we also conducted stratified analyses onthese outcomes during the postintervention follow-up. As shownin Table 3, the largest and smallest anthropometric gains were seenin the FS50 and FS25 groups, respectively, both among childrenwith baseline height below the population median and those above

it. However, the absolute differences in weight and length werebigger, and they more often reached statistical significance in theinitially shorter participants (Table 3). Among these infants, thosewho received FS50 had, on average, a 0.61 kg (95% CI: 20.15,1.37 kg) higher weight gain or a 0.53 (95% CI: 0.07, 0.99) z scoreunit smaller reduction in WAZ than did LP children and a 0.91 kg(95% CI: 0.09, 1.73 kg) higher gain or a 0.81 (95% CI: 0.31, 1.30) zscore unit smaller reduction in WAZ than did FS25 children.Differences in height gains showed a similar pattern but did notreach statistical significance (Table 3).

FIGURE 2. Cumulative incidence functions of (A) severe stunting [length- and height-for-age z score (LAZ/HAZ) , 23 z score units] and (B) moderate-to-severe stunting (LAZ/HAZ , 23 z score units) among children in the LP, FS25, and FS50 groups. Throughout the follow-up, all surviving participantswere used as the denominator for the incidence. Supplementation was provided for the first 12 mo only. LP, likuni phala; FS50, fortified spread 50 g/d; FS25,fortified spread 25 g/d.

TABLE 3

Outcome changes during 36 mo of follow-up in young children who received different doses of fortified spread (FS) and likuni phala (LP)1

All participants Baseline LAZ , median Baseline LAZ � median

Variable

LP group

(n ¼ 50)

FS50 group

(n ¼ 46)

FS25 group

(n ¼ 53) P value2

LP group

(n ¼ 29)

FS50 group

(n ¼ 22)

FS25 group

(n ¼ 21) P value2

LP group

(n ¼ 21)

FS50 group

(n ¼ 24)

FS25 group

(n ¼ 32) P value2

Change in weight (kg) 5.61 6 1.123 6.00 6 1.31 5.58 6 1.08 0.14 5.54 6 1.05 6.15 6 1.33 5.24 6 0.98 0.03 5.70 6 1.24 5.87 6 1.30 5.80 6 1.09 0.90

Change in height (cm) 28.4 6 2.7 28.6 6 2.8 27.9 6 3.1 0.40 28.3 6 2.9 28.8 6 3.2 27.7 6 3.4 0.49 28.6 6 2.4 28.4 6 2.4 28.0 6 2.9 0.67

Change in mid-

upper-arm

circumference (cm)

1.3 6 1.1 1.4 6 1.3 1.0 6 1.1 0.20 1.5 6 1.1 1.8 6 1.1 1.2 6 1.0 0.25 1.1 6 1.1 1.1 6 1.3 0.9 6 1.1 0.71

Change in head

circumference (cm)

5.6 6 0.7 5.8 6 0.9 5.6 6 0.8 0.56 5.6 6 0.7 5.9 6 0.8 5.7 6 0.9 0.35 5.5 6 0.7 5.6 6 0.9 5.6 6 0.8 0.97

Change in weight-

for-age z score

21.09 6 0.90 20.76 6 0.9921.22 6 0.84 0.04 21.02 6 0.7120.49 6 0.8821.29 6 0.88 0.01 21.19 6 1.1321.01 6 1.0221.17 6 0.82 0.79

Change in height-

for-age z score

20.47 6 0.76 20.37 6 0.8320.71 6 0.87 0.10 20.28 6 0.7820.07 6 1.0020.52 6 0.97 0.28 20.72 6 0.6820.64 6 0.5320.84 6 0.79 0.56

Change in weight-

for-length z score

21.52 6 1.20 21.18 6 1.2821.48 6 0.92 0.27 21.67 6 0.8621.20 6 0.9421.83 6 0.96 0.06 21.30 6 1.5521.15 6 1.5621.25 6 0.83 0.92

1 FS25, 25 g/d; FS50, 50 g/d; LAZ, length-for-age z score.2 Obtained by ANOVA.3 Mean 6 SD (all such values).

386 PHUKA ET AL

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

To demonstrate the timing of growth failure in the 3 interventiongroups, we plotted the changes of children’s mean anthropometricindexes as a function of the trial duration (Figure 3). As shown, thedifference in HAZ between LP and FS groups started to developafter 4 mo (when children were 10 mo old) and reached a peak 4–8mo later (at 14–18 mo of age). In contrast, differences in WAZ orWHZ increased gradually during the 12-mo intervention andcontinued to grow wider during the 2-y postintervention follow-up(Figure 3). The patterns were similar among participants withbaseline height above or below the median, but absolute valueswere more pronounced among those who were shorter at the be-ginning (data for individuals with higher baseline HAZ notshown). The corresponding attained weight or height by age andintervention group, both among all children and only those withbaseline LAZ below the median, is shown in Figure 4.

DISCUSSION

We previously reported findings from a randomized controlledtrial that reported a lower incidence of severe linear growthfailure (stunting) among infants whose diet was supplemented for1 y with a specific formula (FS50) of micronutrient-fortified LNSrather than an iso-energetic ration of maize and soy flour (LP)

(29). In the present study, we analyzed the duration of the effectby extending the growth follow-up for an additional 2 y withoutproviding further food supplements. After the intervention, se-vere stunting continued to occur slightly more often in thecontrol group, so that over the entire 3-y follow-up, its incidencewas almost 20% in the LP group but only 4% in the FS50 group.Hypothesis testing suggested that the observed difference wasunlikely to be caused by chance alone. Selection and im-plementation bias was avoided by procedures such as population-based enrollment, use of a control group, randomized group al-location and blinding of outcome assessors, strict control ofdropouts, and the use of standardized outcome variables. Theobserved results are thus consistent with the idea that 1-y sup-plementation of infant diet with 50 g/d FS50 is associated in ruralMalawi with a reduced incidence of severe stunting during a pe-riod that covers the intervention and subsequent 2 y.

Similar to our earlier results (29), the growth difference be-tween the intervention and the control group was largest whenmeasured with dichotomous and more severe outcomes or amongthose individuals who were somehow already disadvantaged(mildly stunted) at enrollment. Differences in mean length orheight developed mostly during the intervention (at 10–18 mo ofage), whereas weight differences became more pronounced later,

FIGURE 3. Development of mean cumulative anthropometric changes in all participants and in those with baseline length-for-age z score (LAZ) belowmedian. WAZ, weight-for-age z score; HAZ, height-for-age z score; WHZ, weight-for-height z score; LP, likuni phala; FS50, fortified spread 50 g/d; FS25,fortified spread 25 g/d.

SUSTAINED GROWTH EFFECT OF COMPLEMENTARY FEEDING 387

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

in the second and third year of life. Although no conclusive ex-planation can be given to this phenomenon, the results are con-sistent with the idea that FS50 supplementation somehow affectedthe timing of acceleration in the infants’ linear growth veloc-ity (32). Such acceleration, known as the infancy-to-childhoodgrowth spurt (IC spurt), occurs at an average age of 9 mo in in-dustrialized countries but often much later in low-income settings(17, 33). Each month of delay in IC spurt is associated, on average,with ’0.5-cm deficit in length gain by 5 y (17, 33).

Factors that induce the IC spurt in infants are unknown atpresent, but dietary intake of cow milk may play a role in theprocess (34). Theoretically, it is thus possible that the observedlinear growth differences in our trial were due to the cow milkprotein fraction in FS50 (10 g in the 50-g/d dose), an increasedintake of which might have led, on average, to a 4 mo earlierinduction of IC spurt in the intervention compared with thecontrol group. This suggestion, while still hypothetical, receivessome support from the fact that children grew less in the groupthat received a lower dose of FS (FS25), which had the samemicronutrient content but only half the amount of milk. Fur-thermore, another recent complementary feeding trial in Malawi,in which the same-age children were supplemented with a soy-

containing modification of LNS (no milk powder), did notdocument any length-gain acceleration (35). Other possibleexplanations of the linear growth variation during the interventionperiod include the different concentration and bioavailability ofother nutrients, eg, zinc or essential fatty acids, in the FS50 andmaize-soy rations (27, 36).

Contrary to the findings on linear growth, weight differencesbetween the groups grew larger, especially after the intervention.Whereas part of this finding may be attributed to the timing ofthe IC spurt, other variables are likely to play a role as well.Possible explanations include the impact of FS50 supplemen-tation on the children’s general health and susceptibility to in-fections or on appetite. Alternatively, the guardians may havecontinued to provide nutritious snacks to children who hadearlier received FS50 but not to those who had received LP.Although these are biologically plausible explanations, we haveno data to support any of these possibilities.

A third important point from our results is the comparison ofoutcomes among children who received the higher (50 g/d) and thelower (25g/d)dosesofFS.Thehigherdosewasdesigned toprovideapproximately the recommended amount of energy from sup-plement foods for 6-mo-old infants without affecting the breast

FIGURE 4. Attained weight and length in all participants and those with baseline length-for-age z score (LAZ) below median during the follow-up. LP,likuni phala; FS50, fortified spread 50 g/d; FS25, fortified spread 25 g/d.

388 PHUKA ET AL

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

milk intake of the recipients (37, 38). Because the price for sucha dose (’$0.20/d) would be relatively high for the rural families inMalawi, we wanted to see if half of the dose would produce thesame growth outcomes but more inexpensively. Unfortunately,this was not the case; children in the lower-dose group had botha higher incidence of severe stunting and lower mean weight andheight gains during the intervention and especially thereafter. Acomparable phenomenon was earlier observed in a Malawiandose-finding trial, in which underweight infants were given dif-ferent doses of LNS for 12 wk (39). Although a smaller (20 g/d)dose of a similar supplement (Nutributter; Nutriset, Malaunay,France) yielded positive results on linear growth and motor de-velopment among 6–12-mo-old infants in Ghana (27), the largerdose may thus be more appropriate in the Malawian setting. Theapparent discrepancy may be explained by the much higher degreeof stunting in the Malawian setting or the fact that the higher dosewas not tested in the Ghanaian setting (27, 29).

The 2 main limitations of our trial were the lack of com-prehensive data on the participants’ dietary intakes during or afterthe intervention and the lack of a nonsupplemented control group.Breast milk intakes were comparable in all trial groups before theintervention and 4 wk after its onset, but subsequent dietaryintakes are unknown (40). Furthermore, because of the lack of anunsupplemented control group, we can make solid conclusionsonly on the relative value of the FS50 and maize-soy flour (LP)supplementations but not on their independent effect. In thiscomparison, the effect size on linear growth was comparablewith other complementary-feeding trials that made comparisonswith unsupplemented control children (18). However, in oursetting, even the control intervention (LP supplementation) mayhave had some impact on the growth outcomes, as suggested bybetter outcomes in the control group than in the half-dose LNSgroup. If compared with no-food control children, the growth-promoting effect of LNS might thus prove even more favorablethan what was observed in this sample. Further trials shouldinvestigate this possibility.

We are grateful to the people of Lungwena, the staff at the Lungwena

Training Health Centre, and our research assistants for their positive attitude,

support, and help in all stages of the study, and to Laszlo Csonka for designing

the collection tools and data entry programs.

The authors’ responsibilities were as follows—JCP, KM, CT, AB, MJM,

and PA: designed the trial; JCP and CT: responsible for data collection; YBC:

responsible for analytical strategy; and YBC, KM, and PA: supervised the

analysis. All authors commented on the analysis and participated in writing

of the manuscript. JCP had full access to all the data in the study and is

responsible for its integrity and the accuracy of the data analysis. AB was

a consultant to Nutriset until December 2003, and the company also sup-

ported the planning of another research project by the same study team

through PA and the University of Tampere after the completion of this trial.

The other authors declared no conflict of interest.

REFERENCES1. Maleta K, Virtanen S, Espo M, Kulmala T, Ashorn P. Child malnutrition and

its predictors in rural Malawi. Paediatr Perinat Epidemiol 2003;17:384–90.2. de Onis M, Blossner M. The World Health Organization Global Data-

base on Child Growth and Malnutrition: methodology and applications.Int J Epidemiol 2003;32:518–26.

3. National Statistical Office (Malawi) and ORC Macro. Malawi de-mographic and health survey 2004: Zomba, Malawi and Calverton.Bethesda, MD: National Statistical Office and ORC Macro, 2005.

4. Grantham-McGregor S, Cheung YB, Cueto S, Glewwe P, Richter L,Strupp B. Developmental potential in the first 5 years for children indeveloping countries. Lancet 2007;369:60–70.

5. UNICEF. The state of the world’s children 2007. New York, NY:UNICEF House, 2006.

6. Mendez MA, Adair LS. Severity and timing of stunting in the first twoyears of life affect performance on cognitive tests in late childhood.J Nutr 1999;129:1555–62.

7. Daniels MC, Adair LS. Growth in young Filipino children predictsschooling trajectories through high school. J Nutr 2004;134:1439–46.

8. Benefice E, Fouere T, Malina RM. Early nutritional history and motorperformance of Senegalese children 4–6 years of age. Ann Hum Biol1999;26:443–55.

9. Cheung YB, Yip PSF, Karlberg JPE. Fetal growth, early postnatalgrowth and motor development in Pakistani infants. Int J Epidemiol2001;30:66–72.

10. Siegel EH, Stoltzfus RJ, Kariger PK, et al. Growth indices, anemia anddiet independently predict motor milestone acquisition of infants inSouth Central Nepal. J Nutr 2005;135:2840–4.

11. Olney DK, Pollit E, Kariger PK, et al. Young Zanzibar children with irondeficiency, iron deficiency anemia, stunting, or malaria have lower motoractivity scores and spend less time in locomotion. J Nutr 2007;137:2756–62.

12. Gardner JM, Grantham-McGregor S, Himes J, Chang S. Behaviour anddevelopment of stunted and nonstunted Jamaican children. J ChildPsychol Psychiatry 1999;40:819–27.

13. Haas JD, Martinez EJ, Murdoch S, Conlisk E, Rivera JA, Martorell R.Nutritional supplementation during the preschool years and physicalwork capacity in adolescent and young adult Guatemalans. J Nutr 1995;125:1078S–89S.

14. Rush D. Nutritional and maternal mortality in developing world. AmJ Clin Nutr 2000;72(suppl):212S–40S.

15. Victora CG, Adair L, Fall C, et al. Maternal and child undernutrition:consequences for adult and human capital. Lancet 2008;371:340–57.

16. Martorell R, Khan LK, Schroeder DG. Reversibility of stunting: epi-demiological findings in children from developing countries. Eur J ClinNutr 1994;48(suppl):S45–57.

17. Liu Y, Albertsson-Wikland K, Karlberg J. Long-term consequences ofearly linear growth retardation (stunting) in Swedish children. PediatrRes 2000;47:475–80.

18. Dewey KG, Adu-Afarwuah S. Systemic review of the efficacy andeffectiveness of complementary feeding interventions in developingcountries. Matern Child Nutr 2008;4(suppl 1):24–85.

19. Briend A. Highly nutrient-dense spreads: a new approach to deliveringmultiple micronutrients to high-risk groups. Br J Nutr 2001;85(suppl 2):S175–9.

20. Collins S. Changing the way we address severe malnutrition duringfamine. Lancet 2001;358:498–501.

21. Diop el HI, Dossou NI, Ndour MM, et al. Comparison of the efficacy ofa solid ready to use food and a liquid milk-based diet for the rehabil-itation of severely malnourished children: a randomized trial. Am J ClinNutr 2003;78:302–7.

22. Manary MJ, Ndekha M, Ashorn P, Maleta K, Briend A. Home-basedtherapy for severe malnutrition with ready-to-use food. Arch Dis Child2004;89:557–61.

23. Sandige H, Ndekha MJ, Briend A, Ashorn P, Manary MJ. Home-basedtreatment of malnourished Malawian children with locally produced orimported ready-to-use food. J Pediatr Gastroenterol Nutr 2004;39:141–6.

24. Ciliberto MA, Sandige H, Ndekha MJ, et al. A comparison of home-basedtherapy with ready-to-use therapeutic food with standard therapy in thetreatment of malnourished Malawian children: a controlled, clinical ef-fectiveness trial. Am J Clin Nutr 2005;81:864–70.

25. Emergency Nutrition Network. Management of moderate acute malnu-trition with RUTF in Niger. Issue 31, September . 2007. Available from:http://www.ennonline.net/fex/31/fex31.pdf (cited 19 November 2008).

26. Briend A, Lacsala R, Prudhon C, Mounier B, Grellety Y, Golden MH.Ready-to-use therapeutic food for treatment of marasmus. Lancet 1999;353:1767–8.

27. Adu-Afarwuah S, Lartey A, Brown KH, Briend A, Zlotkin S, DeweyKG. Randomized comparison of 3 types of micronutrient supplementsfor home fortification of complementary foods in Ghana: effects ongrowth and motor development. Am J Clin Nutr 2007;86:412–20.

28. Adu-Afarwuah S, Lartey A, Brown KH, Zlotkin S, Briend A, DeweyKG. Home fortification of complementary foods with micronutrient

SUSTAINED GROWTH EFFECT OF COMPLEMENTARY FEEDING 389

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from

supplements is well accepted and has positive effects on infant ironstatus in Ghana. Am J Clin Nutr 2008;87:929–38.

29. Phuka JC, Maleta K, Thakwalakwa C, et al. Complementary feeding withfortified spread and incidence of severe stunting in 6- to 18-month-oldrural Malawians. Arch Pediatr Adolesc Med 2008;162:619–26.

30. Maleta K, Virtanen S, Espo M, Kulmala T, Ashorn P. Timing of growthfaltering in rural Malawi. Arch Dis Child 2003;88:574–8.

31. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC growth charts forthe United States: methods and development. National Center for HealthStatistics. Vital Health Stat 11 2002;246:1–190.

32. Karlberg J, Jalil F, Lam B, Low L, Yeung CY. Linear growth retardation inrelation to the three phases of growth. Eur J Clin Nutr 1994;48(suppl):S25–44.

33. Xu X, Wang W, Guo Z, Karlberg J. Longitudinal growth during infancy andchildhood in children from Shanghai: predictors and consequences of theage at onset of the childhood phase of growth. Pediatr Res 2002;51:377–85.

34. Hoppe C, Molgaard C, Michaelsen KF. Cow’s milk and linear growthin industrialized and developing countries. Annu Rev Nutr 2006;26:131–73.

35. Lin CA, Manary MA, Maleta K, Briend A, Ashorn P. An energy-densecomplementary food is associated with a modest increase in weight gain

when compared with a fortified porridge in Malawian children aged6–18 months. J Nutr 2008;138:593–8.

36. Brown KH, Peerson JM, Rivera J, Allen LH. Effect of supplemental zincon the growth and serum zinc concentration of prepubertal children:a meta-analysis of randomized controlled trials. Am J Clin Nutr 2002;75:1062–71.

37. Lutter CK, Dewey KD. Proposed nutrient composition for fortifiedcomplementary foods. J Nutr 2003;133:3011S–20S.

38. World Health Organization. Complementary feeding, guiding principlesfor complementary feeding of the breastfed child PAHO/WHO. Wash-ington, DC: Division of Health Promotion and Protection/Food andNutrition Program, 2003.

39. Kuusipalo H, Maleta K, Briend A, Manary M, Ashorn P. Growth andchange in blood haemoglobin concentration among underweight Mala-wian infants receiving fortified spreads for 12 weeks: a preliminary trial.J Pediatr Gastroenterol Nutr 2006;43:525–32.

40. Galpin L, Thakwalakwa C, Phuka J. Breast milk intake is not reducedmore by introduction of energy dense complementary food than bytypical infant porridge. J Nutr 2007;137:1828–33.

390 PHUKA ET AL

at Tam

pere University Library D

epartment of H

ealth Sciences on January 10, 2009

ww

w.ajcn.org

Dow

nloaded from