School feeding programs in developing countries: impacts onchildren's health and educational outcomes

Lamis H Jomaa, Elaine McDonnell, and Claudia Probart

School feeding programs (SFPs) are intended to alleviate short-term hunger,improve nutrition and cognition of children, and transfer income to families. Thepresent review explores the impact of SFPs on nutritional, health, and educationaloutcomes of school-aged children in developing countries. Peer-reviewed journalarticles and reviews published in the past 20 years were identified and screened forinclusion. Analysis of the articles revealed relatively consistent positive effects ofschool feeding in its different modalities on energy intake, micronutrient status,school enrollment, and attendance of the children participating in SFPs comparedto non-participants. However, the positive impact of school feeding on growth,cognition, and academic achievement of school-aged children receiving SFPscompared to non-school-fed children was less conclusive. This review identifiesresearch gaps and challenges that need to be addressed in the design andimplementation of SFPs and calls for theory-based impact evaluations to strengthenthe scientific evidence behind designing, funding, and implementing SFPs.© 2011 International Life Sciences Institute

INTRODUCTION

Hunger is an ongoing problem that affects more than 1.2billion people who do not have enough to eat in the worldtoday.1 The recent global economic crisis, fluctuations infood prices in 2006–2008, wars and political conflicts, anddevastating natural disasters have deprived millions ofpeople of access to adequate food. Most of the individualsaffected are in the most impoverished regions of theworld. Previous efforts by heads of states, internationalorganizations, and local agencies to address poverty andhunger-related issues resulted in the UN MillenniumDeclaration (2000). The goals outlined in the declarationwere established to eradicate poverty, alleviate hunger,reduce gender inequalities, improve health and longevity,overcome environmental degradation, and most impor-tantly, develop global partnerships to achieve the goals.2

Education and health were central components ofthe roadmap towards implementing the stated goals. The

first millennium development goal (MDG) emphasizedthe eradication of extreme poverty and hunger, whereasthe second and third MDGs focused on “achieving uni-versal primary education” and overcoming gender dis-parities in primary and secondary education. The UNdeclaration and its roadmap have set the platform forglobal trends and national efforts to meet the MDGswithin a reasonable timeframe (mostly by 2015) and theWorld Bank (WB) and its development partners, includ-ing the World Food Program (WFP), took more rapidsteps to meet these goals and launched the Education forAll (EFA) Fast Track Initiative (FTI) in 2002. The mainobjective of EFA FTI was to help low-income countriesmeet the MDGs, particularly the “education for all” goal.3

Despite the major efforts exerted, the progress towardsuniversal primary education (2nd MDG) has been slowand uneven. More than 121 million school-aged childrenare still out of school, and two-thirds of them are girlsliving in rural areas in the most vulnerable regions of the

Affiliations: LH Jomaa, E McDonnell, and C Probart are with the Department of Nutritional Sciences, College of Health and HumanDevelopment, The Pennsylvania State University, University Park, Pennsylvania, USA.

Correspondence: LH Jomaa, Department of Nutritional Sciences, College of Health and Human Development, The Pennsylvania StateUniversity, University Park, PA 16802, USA. E-mail: lamisjomaa@gmail.com.

Key words: children's health, cognitive functions, developing countries, dietary intake, education, growth, micronutrient status, school-agedchildren, school feeding programs

Special Article

doi:10.1111/j.1753-4887.2010.00369.xNutrition Reviews® Vol. 69(2):83–98 83

world. One of the major reasons for this lag in progresstoward universal primary education is the persistence ofpoverty, hunger, and malnutrition.4 Infants and childrenare among the most vulnerable population groups subjectto the adverse, and when very young, irreversible, short-and long-term cognitive, physical, and psychosocial con-sequences of hunger and undernourishment.5,6 There isalso increased evidence that childhood undernutritionimposes significant economic costs on individuals andnations, and that improving children’s diets and nutritioncan have positive effects on their academic performanceand behaviors at school as well as their long-term pro-ductivity as adults.7,8

School feeding programs (SFPs) have been continu-ously gaining popularity in developing countries, mostlyamong those affected severely by childhood hunger andmalnourishment. These programs aim to enhance theconcentration span and learning capacity of school chil-dren by providing meals in schools to reduce short-termhunger that may otherwise impair children’s perfor-mance.9 Currently, SFPs exist in 70 of the 108 low- andlower-middle income countries, and most of them havebeen initiated and funded by the WFP.10 Some of theseSFPs have evolved and been adopted nationally whileothers still rely on the assistance, funding, and/or exper-tise of the WFP and its development partners in varyingdegrees. The WFP and its development partners havebeen promoting school feeding in its different modalitiesfor years as effective interventions that help alleviatehunger and improve the cognitive and educational abili-ties of children. When children are provided with foodat school, not only do parents receive an incentive tosend their children to school, particularly girls, childrenare also encouraged to attend and complete a schoolday. Thus, SFPs can help developing countries and theirdevelopment partners meet a number of MDGs, includ-ing the eradication of hunger, achieving universalprimary education, and closing the gender gap by givingboys and girls equal opportunities for completion ofprimary schooling.11,12

According to Belgeron and Del Rosso’s conceptualframework for Food for Education (FFE),13 also knownas SFPs, FFE programs provide food transfer to childrenat school, income transfer to their families, and resourcetransfer to the schools operating these programs. In 2009,the World Bank and the WFP published a joint reviewon SFPs14 re-emphasizing the rationale and objectives ofthese programs. The three main objectives identified wereto provide safety nets for families to absorb social andeconomic shocks, improve the education and scholasticperformance of school-aged children, and enhance chil-dren’s nutrition and health status. To parallel the threemain pillars or objectives of SFPs (safety nets, nutrition,and education), a logical framework for SFPs was devel-

oped mapping the inputs, outputs, outcomes, and impactsfor each of the objectives and rationales of the pro-grams.15 Resources or inputs include micronutrientfortified meals, snacks, and take-home rations, as wellas anthelmentic treatments; outputs include the numbersof children fed, schools reached, and food rations/deworming tablets distributed. If nutrition is the objectiveof the program, the outcomes identified are the allevia-tion of hunger and improvement of the micronutrientstatus of school-aged children; the intended impactsare to improve the nutrition and health of beneficiarychildren and improve their learning capacities. The mainimpacts of SFPs with regard to meeting educationalobjectives are to improve learning, increase lifetime earn-ings of targeted children, and increase access to educationfor girls, orphans, and vulnerable children. In addition,educating families and future generations about familyplanning, HIV/AIDs prevention, and other health topicsare among the intended impacts of SFPs.

Although the benefits of school feeding are well-documented, controversy remains over the effectivenessof SFPs. According to Kristjansson et al.,16 “experts ata School Feeding/Food for Education Stakeholdersmeeting in 2000 concluded that there is little evidence fornutritional benefits of school feeding and that schoolfeeding only enhances learning when other improve-ments in school quality are made (World Bank, n.d)”. Thepresent review was conducted to summarize the nutritionliterature on the impact of school feeding, focusingprimarily on the relationship between school feedingand changes in the nutritional and health outcomesof targeted school-aged children and secondarily onmore established cognitive and educational outcomes.The review sheds light on gaps in the literature regardingschool feeding in developing countries and other chal-lenges faced by those implementing SFPs. Furthermore, ithighlights the need for theory-based impact evaluationstudies to strengthen the design and implementation ofexisting and future SFPs.

METHODOLOGY AND CRITERIA FOR SELECTION OFEMPIRICAL EVIDENCE INCLUDED IN THE REVIEW

Articles and manuscripts included in this review wereidentified using primarily PubMed and Web of Sciencedatabases, as well as the World Food Program headquar-ters online database and library. Research articlesaddressing this topic and published in peer-reviewedjournals in the past 20 years (1990–2009) were screenedand those that did not meet the inclusion criteria of ourreview were excluded (see Table 1). Variations of termswere used in the search process to find studies of differentdesigns and interventions conducted in developingcountries. Randomized controlled trials (RCTs), which

Nutrition Reviews® Vol. 69(2):83–9884

are experiments in which investigators allocate eligiblepeople randomly into treatment and control groups toreceive or not to receive one or more of the interventionsbeing compared, are regarded as the gold standard ofresearch. However, non-randomized trials are seen asimportant complimentary studies when randomizationor blinding is inappropriate, unethical, or sometimes notfeasible. Thus, our review included RCTs, intervention/control studies, crossover design studies, and effective-ness reports of existing SFPs. Table 2 includes asummarized description of the studies included in thereview, outlining their research designs, study partici-pants, duration of observations/interventions, outcomesmeasured, and the main findings.

IMPACT OF SCHOOL FEEDING ON NUTRITION ANDHEALTH OF SCHOOL-AGED CHILDREN

Growth and body composition

Evidence of the impact of SFPs on children’s growthand body composition remains inconclusive due to themixed results reported from different studies. A fewstudies found a positive effect of school feeding on chil-dren’s growth and anthropometric indices, while othersshowed no effect. In 2006, Kristjansson et al.16 conductedthe first systematic review and meta-analysis to explorethe impact of school feeding on a number of physical andpsychosocial variables in school-aged children. A total of18 studies were included in the review, nine of whichwere conducted in developing/lower income countries(Kenya, Jamaica, India, Indonesia, and China), five ofwhich were RCTs, and four of which were controlledbefore and after (CBA) trials. The meta-analysis showed

an overall small, non-significant change in heightbetween children who did or did not receive mealsat school in RCTs,17–19 whereas a significant increasein height was observed in CBA trials.16,20–22 On average,school-fed children gained 1.43 cm in height more thancontrols. The meta-analysis of two RCTs within the samereview showed a small, yet significant effect of schoolfeeding on height-for-age (z-score change = 0.04 [95%CI:0.02–0.06]). However, authors found a stronger and moreconsistent effect of school feeding on weight gain fromthe three RCTs and three CBA trials that were analyzed; again that ranged between 0.25 to 0.75 kg a year.

Similar positive effects of school feeding werereported by Powell et al.18 Researchers reported positivenutritional outcomes from their RCT in which theyprovided 395 primary school-aged children (2nd to 5th

graders), from 16 rural Jamaican schools, breakfast mealsevery day for one school year (8 months). Children in thecontrol group (n = 396 children) were given one-quarterof an orange as a placebo. Researchers found that childrenin the breakfast group gained significantly more weight(b = 0.42), height (b = 0.25), and BMI (b = 0.16) thanchildren in the control group (P < 0.05). A slight substi-tution effect was observed among breakfast consumersover lunchtime, as they consumed 54 calories less onaverage than the control group. Substitution is a reduc-tion in home diet for students who are receiving food atschool. However, the decrease in energy intake did notoffset the total energy consumed from the school break-fast meal and a net increase in dietary (energy) intake wasobserved among breakfast consumers. Children receivingbreakfast gained, on average, an additional 0.25 cmduring the 8-month intervention, a gain which, whenextrapolated by researchers, was found to be equivalentto 0.4 cm increase in height per year or approximately 1

Table 1 Inclusion and exclusion criteria for studies and publicationsincluded in this review.Studycharacterstic

Inclusion criteria Exclusion criteria

Setting Developing countries Developed countriesSample Primary school-aged children Children < 5 years (infants,

toddlers, preschoolers)Children at secondary school

levelIntervention Effect of in-school meals,

fortified/supplementedmeals and/or snacks, andtake-home food rations(THRs)

Complimentary health/nutrition interventions (ex.De-worming)

Effect of micronutrientsupplements, not part of aschool feeding program orstudy

Publicationtimeframe

1990 and later Prior to 1990

Nutrition Reviews® Vol. 69(2):83–98 85

Tabl

e2

Sum

mar

yof

revi

ewed

pape

rson

the

impa

ctof

scho

olfe

edin

gon

nutr

itio

n,he

alth

,edu

cati

on,a

ndle

arni

ngof

scho

ol-a

ged

child

ren

inde

velo

ping

coun

trie

s(1

3st

udie

s).

Refe

renc

eSt

udy

desi

gnSe

ttin

g,pa

rtic

ipan

ts,i

nter

vent

ion,

and

dura

tion

ofst

udy

Out

com

esm

easu

red

Mai

nfin

ding

s

Kris

tjans

son

etal

.(20

07)16

Coch

rane

syst

emat

icre

view

Incl

uded

18st

udie

s,9

from

deve

lope

dan

d9

from

deve

lopi

ngco

untr

ies

Wei

ght/

heig

ht/m

icro

nutr

ient

stat

us/H

ble

vels

Mix

edre

sults

forc

hang

ein

heig

hts

betw

een

scho

ol-fe

dan

dno

n-sc

hool

-fed

child

ren

(RCT

san

dCB

Asfr

omlo

wer

and

high

erin

com

eco

untr

ies)

Scho

olen

rollm

enta

ndat

tend

ance

Scho

ol-fe

dch

ildre

nga

ined

anav

erag

eof

0.39

kgm

ore

than

cont

rols

(RCT

sin

low

erin

com

eco

untr

ies)

and

0.71

kgov

er11

.3m

onth

s(C

BAs

inlo

wer

inco

me

coun

trie

s)Sc

hool

achi

evem

enti

nm

ath,

read

ing,

and

spel

ling

Incr

ease

inm

uscl

em

ass

(ass

esse

din

only

1RC

T)

Atte

ntio

npr

oble

ms,

on-t

ask

beha

vior

Impr

ovem

ents

ince

rtai

nm

icro

nutr

ient

inta

kes

and

bioc

hem

ical

para

met

ers

(ex.

VitB

12in

take

and

bloo

dco

ncen

trat

ion

from

Neu

man

n’s

RCT

inKe

nya

and

impr

oved

Cafr

omst

udy

inBe

ijing

,Ch

ina)

Ahm

ed(2

004;

Bang

lade

sh)24

Prim

ary

data

colle

cted

from

scho

ol,h

ouse

hold

,an

dco

mm

unity

-leve

lsu

rvey

sin

Bang

lade

sh

SFP

prov

ides

mid

-mor

ning

snac

kco

nsis

ting

offo

rtifi

edw

heat

bisc

uits

toch

roni

cally

food

-inse

cure

area

sof

Bang

lade

sh

Scho

olen

rollm

ent,

atte

ndan

ce,a

nddr

opou

trat

e+

perf

orm

ance

onac

hiev

emen

ttes

ts

SFP

incr

ease

dsc

hool

enro

llmen

tby

14.2

%

Appr

ox.6

,000

prim

ary

scho

ols

(1.2

1m

illio

npr

imar

ysc

hool

-age

dch

ildre

n)Tw

o24

-hou

rrec

alls

wer

eco

llect

edpe

rho

useh

old

toes

timat

ein

divi

dual

inta

kean

dfo

odco

nsum

ptio

npa

tter

ns

Incr

ease

dsc

hool

atte

ndan

ceby

1.3

days

am

onth

Snac

kpr

ovid

es30

0kc

alan

d75

%of

RDAs

forv

itam

ins

and

min

eral

sRe

duce

ddr

opou

trat

eby

7.5%

Incr

ease

dac

hiev

emen

ttes

t*sc

ores

by15

.7%

poin

ts,m

ainl

yin

mat

hem

atic

sw

here

SFP

part

icip

ants

scor

ed28

.5%

high

erth

anno

npar

ticip

antc

ount

erpa

rts

Mot

hers

ofSF

Ppa

rtic

ipan

tsre

port

edin

crea

sed

conc

entr

atio

non

stud

ies

amon

gth

eirc

hild

ren

Sim

eon

(199

8;Ja

mai

ca)23

Revi

ewof

aca

se-c

ontr

olst

udy

Stud

yI:

115

child

ren

ingr

ade

7(1

2–13

y)in

3cl

assr

oom

s†at

aru

rals

choo

lin

Jam

aica

.O

necl

ass

(n=

44)w

asse

rved

anea

rlym

orni

ngsc

hool

mea

land

two

clas

ses

serv

edas

cont

rols

(1se

rved

adr

ink

aspl

aceb

oan

dth

eot

hern

othi

ng)n

=76

Stud

yI:

Atte

ndan

ce,w

eigh

t,an

dpe

rfor

man

ceon

spel

ling

and

arith

met

icte

sts

No

diffe

renc

ein

wei

ghtg

ain

betw

een

inte

rven

tion

grou

pan

dco

ntro

lgr

oup

Dur

atio

n=

2se

mes

ters

(1st

sem

este

rba

selin

e,2nd

sem

este

rint

erve

ntio

n)At

tend

ance

decr

ease

din

both

grou

ps,b

utpr

ovis

ion

ofbr

eakf

astl

ead

tolo

wer

drop

inat

tend

ance

Gre

ater

achi

evem

enti

nar

ithm

etic

test

s(b

reak

fast

grou

pve

rsus

cont

rol)

butn

odi

ffere

nce

insp

ellin

gSi

meo

n(1

998;

Jam

aica

)23St

udy

II:Cr

osso

ver

desi

gn(e

ach

child

test

edon

2m

orni

ngs,

1w

ithbr

eakf

ast,

2nd

with

out.

1w

eek

apar

tas

was

hout

perio

d)

Stud

yII:

90ch

ildre

n(9

–10

y)pe

rfor

man

ceun

derb

reak

fast

and

no-b

reak

fast

cond

ition

s(s

hort

-ter

mfo

odde

priv

atio

n)

Cogn

itive

batt

ery

test

s(7

test

s)us

edfo

rte

stin

gm

enta

larit

hmet

ic,v

erba

lflu

ency

,sho

rt-t

erm

mem

ory,

etc.

Thes

ete

sts

wer

eus

edin

Nor

thAm

eric

anch

ildre

n

Sign

ifica

nttr

eatm

ent-

grou

pin

tera

ctio

nsin

verb

alflu

ency

and

men

tal

arith

met

icte

sts

(und

erno

uris

hed

grou

pha

dlo

wer

perf

orm

ance

whe

nbr

eakf

astw

asom

itted

,fas

ting

did

noth

ave

that

effec

ton

non-

stun

ted

grou

p)+

was

ted

child

ren

perf

orm

edlo

wer

onau

dito

rysh

ort-

term

mem

ory

vers

usno

n-w

aste

daf

term

issi

ngbr

eakf

ast)

Stud

yco

nduc

ted

ina

met

abol

icw

ard

Thre

egr

oups

(30

child

ren

inea

ch):

1)se

vere

lym

alno

uris

hed

grou

p(1

sttw

oye

ars

oflif

e)–

prev

ious

lyho

spita

lized

;2)

stun

ted

(HAZ

<- 2)

;3)n

on-s

tunt

ed(H

AZ>

- 1)G

roup

1fr

omho

spita

lrec

ords

and

grou

ps2

and

3fr

ompr

imar

ysc

hool

s

Nutrition Reviews® Vol. 69(2):83–9886

Pow

elle

tal.

(199

8;Ja

mai

ca)18

RCT

Brea

kfas

toffe

red

toun

dern

ouris

hed

(WAZ

�- 1

SDof

NCH

S)an

dad

equa

tely

nour

ishe

dch

ildre

nW

AZ>

- 1SD

(n=

395)

and

cont

rolg

roup

sw

ere

noto

ffere

dbr

eakf

ast(

n=

396)

Mea

sure

men

tsta

ken

atba

selin

ean

dat

end

ofin

terv

entio

nAd

equa

tely

nour

ishe

dch

ildre

nw

ere

ofhi

gher

SES,

had

high

erat

tend

ance

,bet

tern

utrit

iona

lsta

tus,

and

high

ersc

ores

onar

ithm

etic

,re

adin

g,an

dsp

ellin

gte

sts

atba

selin

eve

rsus

unde

rnou

rishe

dch

ildre

nRe

sults

afte

rint

erve

ntio

nw

ere

sim

ilar,

exce

ptfo

rdiff

eren

ces

inre

adin

gsc

ores

Prim

ary

scho

olch

ildre

n(g

rade

s2–

5)at

16ru

ralJ

amai

can

scho

ols;

mea

nag

e,9

y,W

eigh

tsan

dhe

ight

sBr

eakf

aste

ffect

:chi

ldre

nre

ceiv

ing

brea

kfas

thad

sign

ifica

ntly

bett

erat

tend

ance

,arit

hmet

icsc

ores

,and

incr

ease

dw

eigh

t,he

ight

,and

BMI

scor

esth

anco

ntro

lsD

urat

ion:

brea

kfas

tpro

vide

dev

ery

scho

olda

yfo

r1sc

hool

year

WRA

Tus

edto

asse

ssst

uden

tac

hiev

emen

t(re

adin

g,sp

ellin

g,an

dar

ithm

etic

)

Adeq

uate

lyno

uris

hed

grou

pga

ined

sign

ifica

ntly

mor

ehe

ight

and

BMI

vers

usco

ntro

lsbu

tnut

ritio

nals

tatu

sdi

dno

taffe

ctch

ange

sin

achi

evem

entt

ests

Scho

olat

tend

ance

(sch

oolr

egis

ters

)G

ende

rdiff

eren

ces

exis

t:fe

mal

esha

dhi

gher

gain

sin

wei

ght,

heig

ht,a

ndBM

I+im

prov

edre

adin

g/sp

ellin

g/ar

ithm

etic

test

sco

mpa

red

with

boys

Youn

gerc

hild

ren

inth

ebr

eakf

astg

roup

impr

oved

inar

ithm

etic

com

pare

dw

ithol

derc

hild

ren

Jaco

by(1

996;

Peru

)28Eff

ectiv

enes

str

ial

(ran

dom

ized

cont

rolle

d)St

udy

tom

easu

resh

ort-

term

impa

ctof

Peru

vian

scho

olbr

eakf

astp

rogr

amon

diet

,sch

oola

tten

danc

e,an

dco

gniti

onof

4than

d5th

grad

ers

24-h

reca

llsw

ere

used

toas

sess

the

amou

nts

ofnu

trie

nts,

E,pr

otei

n,an

dFe

cons

umed

perd

ayw

ithan

dw

ithou

tbre

akfa

stat

scho

ol

Sign

incr

ease

inE,

prot

ein,

and

iron

inta

kes

ofch

ildre

nre

ceiv

ing

scho

olbr

eakf

astv

ersu

sco

ntro

ls

10sc

hool

sfr

omou

tski

rts

ofan

Ande

anci

ty(H

uara

z)w

ere

rand

omly

bute

qual

lyas

sign

edto

trea

tmen

torc

ontr

olco

nditi

ons

(app

rox.

400

child

ren)

Atte

ndan

ce(t

each

er-r

ecor

ded

roll)

Impr

oved

rate

sof

atte

ndan

ce(b

reak

fast

effec

t)

Subs

ampl

eof

60ch

ildre

nse

lect

edto

test

effec

tofb

reak

fast

ondi

etar

yin

take

Psyc

ho-e

duca

tiona

lpar

amet

ers

(3te

sts

forc

ogni

tive

proc

esse

san

d3

test

sfo

rcom

plex

men

tala

bilit

ies;

read

ing/

voca

bula

ry/m

ath)

Impr

oved

perf

orm

ance

onvo

cabu

lary

test

,mai

nly

amon

gch

ildre

nof

grea

terw

eigh

tM

ealp

rovi

ded

30%

ofto

talE

requ

irem

ents

,60

%of

RDAs

form

iner

als

and

vita

min

s+

100%

RDA

fori

ron

No

sign

ifica

ntim

prov

emen

tsw

ere

obse

rved

onco

ding

,rea

ding

com

preh

ensi

on,o

rmat

hem

atic

ste

sts

amon

gth

eex

perim

enta

lgro

up

Afrid

i(20

10;

Indi

a)30

Rand

omiz

edpr

ogra

mev

alua

tion

A24

-hfo

odco

nsum

ptio

nsu

rvey

was

deve

lope

dto

asse

ssdi

etar

yin

take

ofch

ildre

n(n

=97

6,ag

es4–

14y)

enro

lled

inpr

ivat

epr

imar

ysc

hool

sin

the

cent

ral

regi

onof

Mad

hya

Prad

esh

ontw

oco

nsec

utiv

eda

ys(s

choo

land

non-

scho

olda

y).

Tota

ldai

lyco

nsum

ptio

nan

dca

lorie

inta

keon

scho

olan

dno

n-sc

hool

day

Indi

vidu

alnu

trie

nts

(pro

tein

,ca

rboh

ydra

te,c

alci

um,a

ndiro

n)on

scho

olan

dno

n-sc

hool

day

On

aver

age,

the

diffe

renc

ebe

twee

nRD

Aan

dac

tual

daily

inta

keof

child

ren

inpr

imar

ysc

hool

was

redu

ced

by10

0%fo

rpro

tein

inta

ke,

30%

forc

alor

iein

take

,and

10%

fori

ron

inta

ke,4

9–10

0%of

the

food

tran

sfer

sw

ere

refle

cted

inth

eto

tald

aily

inta

kes

ofch

ildre

n

Indi

vidu

alfo

odco

nsum

ptio

nda

taw

asus

edto

asse

ssin

take

ofes

sent

ialn

utrie

nts

Dai

lyac

tiviti

esre

call

data

(incl

ude

cook

ing,

hous

ehol

dcl

eani

ng,s

iblin

gca

re,l

ives

tock

care

,col

lect

ing

wat

eran

dfir

ewoo

d)Ja

coby

(199

8;Pe

ru)29

Ove

rvie

wpa

pera

ndev

alua

tion

ofa

SFP

Resu

ltsfr

omtw

ost

udie

sth

atex

plor

edth

eim

pact

ofsc

hool

brea

kfas

tpro

gram

ondi

et,n

utrit

iona

lsta

tus,

and

num

bero

fed

ucat

iona

lout

com

esof

scho

olch

ildre

nag

ed5–

10y

inAn

des

regi

onin

Peru

:1)

Jaco

byet

al.1

996

inH

uara

z;an

d2)

Prog

ram

eval

uatio

nco

nduc

ted

inM

atah

uasi

Effec

tofb

reak

fast

ondi

et,a

nem

ia,a

ndsc

hool

atte

ndan

cew

ere

asse

ssed

usin

gth

em

easu

res

indi

cate

din

Jaco

by19

96pa

per

Sign

ifica

ntdi

etar

yim

prov

emen

tsin

ener

gy,p

rote

in,a

ndes

sent

ial

nutr

ient

sin

take

Not

able

impr

ovem

enti

niro

nin

take

and

drop

inin

cide

nce

ofan

emia

from

66to

14%

in6

mon

ths

Nutrition Reviews® Vol. 69(2):83–98 87

Tabl

e2

Co

nti

nu

edRe

fere

nce

Stud

yde

sign

Sett

ing,

part

icip

ants

,int

erve

ntio

n,an

ddu

ratio

nof

stud

yO

utco

mes

mea

sure

dM

ain

findi

ngs

Van

Stui

jven

berg

(199

9;So

uth

Afric

a)32

RCT

Mic

ronu

trie

ntst

atus

of6–

11-y

ear-

old

child

ren

(gra

des

1–5)

ata

rura

lsch

ooli

nSo

uth

Afric

aas

sess

edaf

tert

hepr

ovis

ion

ofbi

scui

tsfo

rtifi

edw

ithFe

,be

ta-c

arot

ene,

and

iodi

ne+

aVi

tC-

fort

ified

cold

drin

kto

115

child

ren

(inte

rven

tion

grou

p)an

dco

mpa

red

toco

ntro

lgro

upof

child

ren

rece

ivin

gno

n-fo

rtifi

edbi

scui

ts(n

=11

3)

Wei

ghta

ndhe

ight

wer

em

easu

red

Mic

ronu

trie

nt(s

erum

ferr

itin,

seru

miro

n,iro

n-bi

ndin

gca

paci

ty,s

erum

retin

ol,u

rinar

yio

dine

),an

dan

thro

pom

etric

(wei

ght,

heig

ht,H

AZW

AZ)m

easu

rem

ents

Thyr

oid

size

was

exam

ined

Atba

selin

e:H

igh

leve

lsof

VitA

defic

ienc

yan

dgo

iteri

nta

rget

edpo

pula

tion

1/3

ofth

esa

mpl

ew

asin

fect

edw

ithat

leas

t1pa

rasi

teSm

all%

ofch

ildre

nw

ere

stun

ted

and

none

wer

eun

derw

eigh

t

Afte

rint

erve

ntio

n:N

osi

gnifi

cant

diffe

renc

ein

wei

ghto

rhei

ghtr

ecor

ded

over

12m

onth

sbe

twee

nin

terv

entio

nve

rsus

cont

rolg

roup

Cogn

itive

asse

ssm

entt

ests

wer

eco

nduc

ted

onch

ildre

nin

grad

es2–

4Re

duce

din

cide

nce

ofill

ness

es;f

ewer

scho

olda

ysm

isse

dby

inte

rven

tion

vers

usco

ntro

lgro

upD

urat

ion:

12m

onth

sM

orbi

dity

data

was

colle

cted

(ass

ess

reas

ons

fora

bsen

ce)

Impr

ovem

ents

inVi

tA,F

e,an

dio

dine

stat

usof

inte

rven

tion

grou

pan

dre

duce

dan

emia

,Vit

A,an

dio

dine

defic

ienc

ies

Cogn

itive

outc

omes

:Gre

ater

trea

tmen

teffe

cts

inch

ildre

nw

ithlo

wFe

and

Hb

and

with

goite

ratb

asel

ine

onon

eof

the

cogn

itive

test

sG

rille

nber

ger

etal

.(20

03;

Keny

a)25

RCT

554

stud

ents

inG

rade

1fr

om12

sele

cted

prim

ary

scho

ols

ina

rura

lm

alar

ia-e

ndem

icar

eain

Keny

aw

ere

rand

omly

assi

gned

toa

supp

lem

ent

grou

p(m

eat,

milk

,ore

nerg

ysu

pple

men

t)or

toa

cont

rolg

roup

Gro

wth

para

met

ers

(wei

ght,

heig

ht,

skin

fold

thic

knes

s,m

id-u

pper

arm

mus

cle,

and

fata

rea)

Child

ren

insu

pple

men

tatio

ngr

oups

gain

edap

prox

.0.4

kg(1

0%)m

ore

wei

ghtt

han

child

ren

inth

eco

ntro

lgro

upCh

ildre

nin

the

mea

t,m

ilkan

den

ergy

grou

psga

ined

0.33

,0.1

9,an

d0.

27cm

mor

eM

UAC,

resp

ectiv

ely,

than

child

ren

inth

eco

ntro

lgro

upCh

ildre

nin

3tr

eatm

entg

roup

s,m

ainl

ym

eat,

follo

wed

bym

ilkgr

oup,

gain

edm

ore

mus

cle

area

than

cont

rolg

roup

Child

ren

rece

ived

food

asa

mid

-mor

ning

snac

kin

scho

ols

over

anin

terv

entio

npe

riod

of23

mon

ths

Posi

tive

effec

tofm

ilksu

pple

men

ton

heig

htga

inw

asob

serv

edin

asu

bgro

up(s

tunt

edch

ildre

nw

ithH

AZ�

- 1.4)

Ove

rall

effec

tofs

uppl

emen

tatio

non

heig

ht,H

AZ,W

HZ,

and

mea

sure

sof

body

fatw

ere

insi

gnifi

cant

Neu

man

n(2

003;

Keny

a)17

RCT

Child

ren

ingr

ade

1(n

=55

4)fr

om12

scho

ols

ina

mal

aria

-end

emic

area

inKe

nya

wer

era

ndom

ized

into

4fe

edin

gin

terv

entio

ns:1

)mea

t,2)

milk

,3)e

nerg

y,4)

cont

rol

Base

line

mea

sure

sin

clud

ednu

triti

onal

stat

us,h

ome

food

inta

ke,

anth

ropo

met

ry,b

ioch

emic

alm

easu

res

ofm

icro

nutr

ient

stat

us,

mal

aria

,int

estin

alpa

rasi

tes,

heal

thst

atus

,and

cogn

itive

and

beha

vior

alou

tcom

es

Base

line

data

repo

rted

:In

adeq

uate

mic

ronu

trie

ntin

take

sth

roug

h24

-hou

rrec

alls

confi

rmed

with

high

bioc

hem

ical

defic

ienc

ies

ofVi

tA,B

12,z

inc,

iron,

and

ribofl

avin

Dur

atio

nof

stud

y:2

scho

olye

ars

(7te

rms;

each

term

3m

onth

s)~

30%

ofch

ildre

nw

ere

stun

ted

and

unde

rwei

ght

Hig

hpr

eval

ence

ofm

alar

ia(5

0%of

child

ren

have

enla

rged

sple

ens)

and

othe

rpar

asiti

cin

fect

ions

Neu

man

n(2

007;

Keny

a)26

RCT

Sam

eas

Neu

man

n20

03O

utco

mes

mea

sure

dat

base

line

and

long

itudi

nally

(atd

iffer

enti

nter

vals

thro

ugho

utth

e2-

year

inte

rven

tion)

and

pres

ente

dhe

re:w

eigh

t,M

UAC,

skin

fold

thic

knes

sm

easu

res,

and

obse

rvat

ions

fora

ctiv

ityle

vels

and

beha

vior

s(s

ocia

lint

erac

tions

incl

assr

oom

)

Mea

tand

ener

gy-s

uppl

emen

ted

child

ren

perf

orm

edsi

gnifi

cant

lybe

tter

onar

ithm

etic

test

sov

ertim

eco

mpa

red

tom

ilkan

dco

ntro

lgro

ups

Mea

tgro

upha

dth

egr

eate

st%

incr

ease

into

talt

ests

core

s(in

dica

toro

fsc

hool

perf

orm

ance

),m

ainl

yar

ithm

etic

,com

pare

dto

othe

rfee

ding

grou

psan

dco

ntro

lM

eatg

roup

show

edth

est

eepe

stra

teof

incr

ease

inm

id-u

pper

-arm

mus

cle

(lean

mas

s)an

dgr

eate

stin

crea

sein

%of

time

spen

ton

high

leve

lsof

phys

ical

activ

ity(fr

eepl

ayob

serv

ed)c

ompa

red

toal

loth

ergr

oups

Child

ren

inth

em

eat,

milk

,and

ener

gygr

oups

allg

aine

dw

eigh

tata

grea

terr

ate

than

cont

rolg

roup

and

youn

gers

tunt

edch

ildre

nin

the

milk

grou

psh

owed

agr

eate

rrat

eof

gain

inhe

ight

com

pare

dto

othe

rgr

oups

Nutrition Reviews® Vol. 69(2):83–9888

Siek

man

net

al.

(200

3;Ke

nya)

36

RCT

Sam

eas

Gril

lenb

erge

rand

Neu

man

n(2

003)

Base

line

anth

ropo

met

ricda

ta(w

eigh

tan

dhe

ight

mea

sure

men

ts)

Plas

ma

VitB

12co

ncen

trat

ion

incr

ease

dsi

gnifi

cant

lyam

ong

mea

t-an

dm

ilk-s

uppl

emen

ted

grou

psle

adin

gto

a17

%de

crea

sein

VitB

12

defic

ienc

yam

ong

mea

tgro

upan

d27

%de

crea

sein

milk

grou

paf

ter1

year

ofsu

pple

men

tatio

nD

urat

ion:

1sc

hool

year

(9m

onth

s)Bi

oche

mic

alm

arke

rsw

ere

colle

cted

toas

sess

cert

ain

mic

ronu

trie

nts

(Vit

B 12,

VitA

,Fe,

Zn,C

u,fo

late

,Vit

B 2)a

ndin

flam

mat

ion

and

infe

ctio

nm

arke

rs(C

RP,s

plee

n,an

dst

ools

ampl

esto

asse

ssce

rtai

npa

rasi

tes

and

mal

aria

)

No

sign

ifica

ntch

ange

sbe

twee

ngr

oups

wer

eob

serv

edfo

roth

erm

icro

nutr

ient

s(Z

n,Fe

,Cu,

VitA

)Lo

wer

infe

ctio

npr

eval

ence

inth

ispo

pula

tion

afte

r1-y

eari

nter

vent

ion

Mur

phy

etal

.(2

003;

Keny

a)37

RCT

Sam

eas

Gril

lenb

erge

rand

Neu

man

n(2

003)

Ener

gy,p

rote

in,a

ndm

icro

nutr

ient

inta

kes

ofch

ildre

nw

ere

asse

ssed

atba

selin

ean

ddu

ring

2-ye

arfe

edin

gpe

riod

offe

edin

gst

udy

Tota

lEin

take

ofm

eatg

roup

grea

tert

han

milk

and

ener

gygr

oups

sepa

rate

lyIn

take

sof

VitB

12,V

itA,

ribofl

avin

,and

Cain

milk

grou

pgr

eate

rth

anco

ntro

lgro

upIn

take

sof

VitB

12,V

itA,

avai

labl

eFe

,and

Znin

mea

tgro

upgr

eate

rth

anco

ntro

lgro

upAr

esen

ault

etal

.(2

009;

Colo

mbi

a)31

Obs

erva

tiona

lin

vest

igat

ion

Mea

sure

the

impa

ctof

Bogo

ta’s

Secr

etar

yof

Educ

atio

nsn

ack

prog

ram

onhe

alth

and

nutr

ition

alst

atus

of3,

202

child

ren

(1,8

03in

scho

ols

that

rece

ived

snac

kpr

ogra

man

d1,

399

cont

rolc

hild

ren

insc

hool

sno

tco

vere

dby

snac

kpr

ogra

m)

Gro

wth

outc

omes

(HAZ

,BM

I-for

-age

z-sc

ores

)G

reat

erin

crea

ses

inVi

tB12

leve

lsam

ong

child

ren

atsc

hool

sre

ceiv

ing

the

snac

kco

mpa

red

toch

ildre

nat

scho

ols

notr

ecei

ving

snac

k(e

ffect

afte

r3

mon

ths)

Child

ren

recr

uite

d(5

–12

year

sol

d)M

icro

nutr

ient

stat

usou

tcom

es(p

lasm

afe

rriti

nan

dVi

tB12

and

eryt

hroc

yte

fola

tew

ere

mea

sure

d)

Snac

kw

asno

tass

ocia

ted

with

fola

teor

ferr

itin

leve

lsin

child

ren

HAZ

ofal

lchi

ldre

nde

crea

sed,

butc

hild

ren

rece

ivin

gsn

acks

had

smal

ler

decr

ease

inH

AZth

anch

ildre

nno

trec

eivi

ngth

esn

ack

Dur

atio

n:3

mon

ths

Mor

bidi

tyra

te(fe

ver,

coug

h,di

arrh

ea,

vom

iting

,no.

ofda

ysw

ithsy

mpt

oms,

no.o

fdoc

torv

isits

,no.

ofda

ysst

uden

tabs

entf

rom

scho

ol)

Chan

gein

BMI-f

or-a

gez-

scor

esw

asno

tsig

nific

antly

diffe

rent

betw

een

2gr

oups

ofch

ildre

nSi

gnifi

cant

lyfe

wer

mor

bidi

tysy

mpt

oms

(57%

few

erda

ysw

ithco

ugh

and

feve

rsym

ptom

s,30

%fe

wer

days

ofdi

arrh

ea,5

5%fe

wer

days

with

diar

rhea

and

vom

iting

)and

44%

few

erdo

ctor

visi

tsam

ong

child

ren

rece

ivin

gsn

acks

(�23

%fe

wer

days

ofsc

hool

abse

ntee

ism

)*

Conc

lusi

ve:c

onsi

sten

tfind

ings

that

supp

ortt

helin

kSF

Ppr

ovis

ions

and

inte

nded

outc

omes

.†

Inco

nclu

sive

:mix

edre

sults

from

the

liter

atur

e,w

eake

revi

denc

eon

link

betw

een

SFP

prov

isio

nsan

din

tend

edou

tcom

es.

‡U

nexp

lore

d:no

empi

rical

evid

ence

toex

plor

eth

ere

latio

nshi

pbe

twee

nSF

Ppr

ovis

ions

and

inte

nded

outc

omes

.Ab

brev

iatio

ns:B

MI,

body

mas

sin

dex;

Ca,c

alci

um;C

BA,c

ontr

olle

dbe

fore

and

afte

rtria

ls;C

u,co

pper

;E,e

nerg

y;Fe

,iro

n;H

AZ,h

eigh

t-fo

r-ag

ez-

scor

e;H

b,he

mog

lobi

n;kc

al,k

iloca

lorie

s;RC

T,ra

ndom

ized

cont

rolle

dtr

ial;

RDA,

reco

mm

ende

ddi

etar

yal

low

ance

;SFP

,sch

oolf

eedi

ngpr

ogra

m;V

itB 2

,rib

oflav

in;V

itB 1

2,co

bala

min

;WAZ

,wei

ght-

for-

age

z-sc

ore;

WH

Z,w

eigh

t-fo

r-he

ight

z-sc

ores

;WRA

T,w

ide

rang

eac

hiev

emen

ttes

t;Zn

,zi

nc.

Nutrition Reviews® Vol. 69(2):83–98 89

month of growth in the National Center for Health Sta-tistics (NCHS) reference population. The increase inweight among intervention group members was found tobe even greater than the increase in height, an equivalentof 2–3 months of weight gain in the reference population.Analyzing the effect of breakfast based on the nutritionalstatus of children, researchers observed more significantgains in height and BMI among adequately nourishedchildren compared to undernourished children, b = 0.77and b = 0.18 (P < 0.005), respectively.18 The significantlygreater increase in the height and BMI of adequatelynourished children compared to undernourished chil-dren was somewhat unexpected. Powell et al.18 arguedthat the “undernourished” group of children recruited inthis study was only moderately undernourished, whereaschildren in previous studies were more severely under-nourished. The undernourished children may havegreater dietary needs than their well-nourished counter-parts and these dietary requirements were either not metthrough the school breakfast meals or were offset byreduced dietary intakes at home. This study had somelimitations as it lacked data on the micronutrient status ofchildren and their health conditions and possible infec-tions that may interfere with the absorption of nutrientsfrom school meals. In addition, the sample of undernour-ished children may have been suffering from subclinicalinfections that could have impeded their absorption ofnutrients, such as iron and zinc, that are needed foradequate growth and development. It is also possible thatenergy and nutrient utilization were better in theadequately nourished group of children compared to themalnourished children.

Simeon’s review of a study conducted in a ruralschool in Jamaica did not show remarkable changes inthe weight gain of children following consumption of amorning meal each school day for two semesters.23 Atotal of 115 students in grade 7 (12–13 years old) wererecruited; one class (n = 44 students) received the schoolmeal (100 mL milk and either a slice of cake or a meat-filled pastry providing a total of 500 kcal on average)and two classes served as controls. Weight for age wasmeasured before the school feeding intervention started(1st semester) and at the beginning and end of the secondsemester (during which intervention took place). Stu-dents in both the intervention and the control groupsgained weight after the 2nd semester; however, there wasno significant difference in weight gain between bothgroups. The lack of impact of school feeding on theweight gain of these children may be attributed to severallimitations of this study. Data on the height of childrenat baseline was not collected due to time constraints;thus, researchers could not measure the change inheight after the intervention was completed and observewhether there were significant differences in height

change between the intervention and control groups. Inaddition, total dietary intake of participants was notassessed to learn if the school meal provided additionalcalories and nutrients to the child’s dietary intake or ifit was compensated for with diminished energy and foodconsumption outside of school hours.

A large study that was conducted in Bangladeshby Ahmed Akhter and the International Food PolicyResearch Institute provided evidence regarding the posi-tive impact of SFPs on dietary intakes and educationaloutcomes of school-aged children. The SFP was imple-mented by theWFP and the Government of Bangladesh inapproximately 6,000 primary schools located in highlyfood-insecure rural areas and four slum areas in Dhakacity, Bangladesh. Ahmed24 reported that the BMI of chil-dren receiving a midmorning snack at school increased by4.3% compared to children in control schools who werenot receiving these snacks. The midmorning snack con-sisted of a packet of fortified wheat biscuits providing atotal of 300 calories and 75% of the recommended dailyallowances of vitamins and minerals for school-agedchildren; it was given to students for each day of schoolattendance. The average energy intakes of participantswere 11% and 19% higher in rural and urban slum areas,respectively, than the energy intakes of sex- and age-matched students in control schools who did not partici-pate in the SFP. Energy intake consumed from biscuitswas 97% additional to the child’s normal diet; thus, the SFPimproved net food consumption of the participating chil-dren, and the extra energy from the biscuits was not com-pensated for by a decrease in food consumption at home.Using household food consumption surveys, researchersobserved that the majority of students receiving SF sharedfortified biscuits with other members of their families on aregular or intermittent basis. These spillover effects of SFbiscuits on food consumption of the participants’ siblings(ages 2–5 years) led to a 7% increase in total caloriesconsumed, on average, by these preschoolers.

In terms of body composition, studies have alsoshown promising results. Grillenberger et al.25,26 con-ducted an RCT in 12 primary schools from a ruralmalaria-endemic area in Kenya (Embu District). Schoolswere assigned to one of three food supplement groups: 1)energy, 2) milk, or 3) meat supplement. In larger schoolswith more than one grade 1, classes were randomlyassigned to one of the three isocaloric supplementgroups. A total of 544 students participated in the studyand, depending on the group to which their school orclassroom was assigned, they were further randomlyassigned to a food supplement group (treatment) or nosupplement group (control). The number of children ineach group was relatively equal, approximately 140 chil-dren per group. Children in the treatment groups (receiv-ing food supplements) were fed Githeri, a local vegetable

Nutrition Reviews® Vol. 69(2):83–9890

stew, with meat, milk, or extra oil added, dependingon the treatment group to which they were assigned, i.e.,meat, milk, or energy. The control groups did not receivefood. Researchers observed that children in each of thesupplementation groups gained approximately 0.4 kg(10%) more weight than children in the control groups.In addition, children in the three treatment groupsshowed increases in mean upper-arm area circumference(MUAC), a measure that is usually used as an indicatorof protein-energy malnutrition. They also gained moremid-upper arm muscle circumference (MUAMC), whichis a fine indicator for the total body muscle mass, than thecontrol groups; this was mainly observed in membersof the meat group, whose gain reached up to 90%. Thesignificant gain in MUAMC among children in the meatgroup highlights the importance of providing childrenwith a good source of high-quality protein that is alsorich in multiple micronutrients and can increase the bio-availability of iron and zinc. This is important, especiallyamong children living in rural communities where thestaple plant-based foods are low in iron and zinc contentand people are susceptible to micronutrient deficiencies.The overall effects of supplementation on height, heightfor age, weight for height z-scores, and measures of bodyfat were insignificant in this study; however, researchersobserved a positive effect of milk supplements on heightgain in a subgroup of children (stunted children withheight for age z scores � -1.4).

In another study conducted in South Africa,27

the fortification of soup powders with iron and vitaminC, when combined with deworming within a SFP, led tosignificant improvements in height, height-for-age, andweight-for-height z scores of primary school childrenaged 6–8 years; the improvements were mainly amongthose with low baseline iron stores. Furthermore, com-bined food fortification and deworming led to improve-ments in weight and weight for age z scores of childrenwho had adequate iron stores at baseline. One importantfinding of this study is that the combined positive effectsof iron fortification and deworming on children’s anthro-pometric status surpassed the individual effects of eachtreatment alone.

Dietary and micronutrient status

The nutritional benefits of school breakfast programswere further documented in studies conducted inPeru.28,29 In 1993, the government of Peru launched aschool breakfast program in one of the poorest provincesof the country, the Andes region. An evaluation of theschool breakfast program implemented in the outskirtsof two Peruvian cities (Matahuasi and Huaraz) showedsignificant improvements in the energy, protein, andmicronutrient intakes of a subsample of children receiv-

ing the school breakfast. A total of 120 children wererecruited from the 3rd and 4th grades and were dividedequally between the experimental and control groups.After implementation of the school breakfast program,significant differences were seen between the experimen-tal group and the control group with respect to totalenergy intake (2,182 versus 1,731 kcal/day), proteinintake (56.1 versus 43.6 g/day), and iron intake (21.6versus 12.5 mg), P < 0.001.28 The energy intake of chil-dren receiving the school breakfast increased by 15.2%,their protein intake increased by 16.1%, and their dietaryiron intake increased by 60%. One of the strengths of thisstudy is that the observed increase in the dietary intakeof children within the experimental group was notcompensated for by lower food consumption at home.Instead, children consuming a breakfast meal at schoolhad significantly higher overall dietary intakes comparedto their control counterparts. In addition, the study con-trolled for a number of variables, including children’sstature (height-for-age) and nutritional status (weight-for-age) in their regression analyses in an attempt tomeasure the independent effect of school breakfast con-sumption on dietary intake and a number of educationaland cognitive markers. Using data from a nationally man-dated school meal program in India, Afridi30 conductedan empirical analysis of 24-hour food consumptionrecalls for children (age range, 4–14 years; n = 976) onschool and non-school days to estimate the extent towhich children benefit from the targeted school mealprogram. Afridi’s study showed that the meal programprovided children with a significant proportion of theirdaily intake of five nutrients (energy/calories, proteins,carbohydrates, calcium, and iron). Furthermore, 49–100%of the food transfers were reflected in the total dailyintakes of children, indicating that the program suc-ceeded in improving dietary intake for the five essentialnutrients for which the diets of children in India werefound to be highly deficient.

In Bogota, Colombia, Arsenault et al.31 conducteda longitudinal observation of 3,202 children aged 5–12years in public schools to examine whether a state-launched snack program (initiated in 2004 in all publicprimary schools) had any impact on the nutritional statusof children. The snack consisted of a beverage, a cerealand/or a protein component, was well as a “sweet” com-ponent that included peanuts and Petit Suisse cheese,as well as fruits on most days. The snack was designed toprovide children with a certain percentage of recom-mended daily intakes (30% of energy, 50% of iron, and40% of calcium). Only 3 months after the program waspilot-tested, researchers observed greater increases inserum vitamin B12 levels among children receiving thesnacks compared to the controls. However, significantchanges were not observed in the hemoglobin, serum

Nutrition Reviews® Vol. 69(2):83–98 91

ferritin, or folate levels of children. This can be explainedby the low prevalence of iron deficiency at baseline andthe fact that children in both the schools receiving thesnack and the controls were receiving iron supplementsprior to this program’s initiation.

Food fortification with micronutrients or the supple-mentation of children with multiple nutrients are amongthe strategies used in certain SFPs to reduce multiplemicronutrient deficiencies and improve the nutritionalstatus and cognitive and learning capacities of school-aged children. Van Stuijvenberg et al.32 were interested indetermining the effects of providing schoolchildren withmineral-fortified biscuits on their micronutrient status.The researchers conducted their study in a rural school inSouth Africa located in an area with a high prevalence ofmicronutrient deficiencies. They recruited 115 children(6–11 years old) to serve as the treatment group and 113children served as controls. The treatment group receivedbiscuits fortified with iron, beta-carotene, and iodine (at50% of the RDA) whereas the control group consumednon-fortified biscuits. Significant improvements wereobserved in the vitamin A, iron, and iodine status ofchildren who received the fortified biscuits. In addition,anemia, vitamin A, and iodine deficiencies were reducedby 13%, 28%, and 67%, respectively, in this sample.However, improvements seen in the vitamin A and ironstatus of children were not sustained after consumption ofthe fortified biscuits was interrupted during the school’ssummer break. The researchers concluded that since thebiscuits provided only 50% of the RDA for vitamin A inthe form of carotene, this may have been only sufficient tomaintain the day-to-day vitamin A levels but not sufficientfor replenishing very low or depleted vitamin A stores.Also, the dietary intake of vitamin A from meals con-sumed at home during the summer months, were assessedto be approximately 10% of the RDA per day; thus, duringthe period in which the fortified biscuit were not offered tostudents, retinol values reverted to the baseline level. Allthese findings indicate that the diets of these childrencontained very low amounts of micronutrients and thechildren’s body stores were depleted rapidly during thesummer months. According to the same authors, resolv-ing iron deficiency through food fortification can be verychallenging due to the low bioavailability and high reac-tivity of the iron compounds used in food fortification.33

However, the provision of vitamin C-fortified drinks (richin ascorbic acid) along with fortified food has been shownto improve iron status and is thus encouraged in SFPstargeting children with high levels of iron deficiency andiron-deficiency anemia, when feasible.34 Only iodine didnot return to the levels measured prior to the interven-tion, and that was probably because iodization of salt wasmandated in South Africa 6 months after the interventionwas in place.

This study provides promising results in terms offood fortification and its impact on children’s micronu-trient status. The fortification of food proved successful inalleviating highly prevalent micronutrient deficiencies,such as iron and vitamin A, which can lead to seriousconsequences for the health and cognition of children.However, the amount and duration of fortification wereinsufficient to maintain these changes. Thus, the challengeremains for researchers and policy planners to designSFPs that not only provide the targeted children withadequate types and amounts of micronutrients, depend-ing on their deficiencies, but also ensure that the durationof fortification or micronutrient supplementation issufficient to replenish depleted body stores.

In another intervention-control study conductedin South Africa by Kruger et al.,27 the fortification of souppowder with iron and vitamin C, when combined withthe use of anthelmintic treatment (deworming) as part ofa school feeding scheme, resulted in significant, positivechanges in the hemoglobin, mean corpuscular volume(MCV), and serum ferritin levels of children receivingfortified soups compared to controls. These positiveeffects were observed mainly among children with lowbaseline iron stores. Furthermore, fortification of biscuitswith bovine-hemoglobin as part of a nationwide SFP inChile led to significant improvements in the iron statusof children (aged 6–11 years) who received fortifiedbiscuits compared to children consuming non-fortifiedbiscuits.35 These positive findings were observed despitethe low rates of iron-deficiency anemia found amongthe recruited samples of school children. In addition,increases in hemoglobin concentrations and iron stores(serum ferritin) were observed among boys across all agesand among girls, mainly after menarche. These findingshighlight the importance of providing micronutrients toschool-aged children, particularly during growth phaseswith high micronutrient demands.

The use of food supplements to overcome multiplemicronutrient deficiencies among school-aged childrenin developing countries is another area of interest to someresearchers and program planners. Neumann et al.17

assessed the micronutrient status of children before inter-vention using 24-hour recalls and biochemical measures.They found that children had low micronutrient intakesof vitamins A, B12, B2, zinc, and iron. Furthermore, atbaseline, approximately 30% of children were found tobe stunted and underweight. The same research grouplater analyzed the effect of providing three different foodsupplements (meat, milk, and energy) on the micronutri-ent status of these primary school children. In their 1-yearintervention, researchers succeeded in reducing theprevalence of vitamin B12 deficiency by 17% for the meatsupplement group and by 27% for the milk group. Thefact that a similar effect was not observed in the energy

Nutrition Reviews® Vol. 69(2):83–9892

group can be explained by the higher level of vitamin B12

in animal-source foods such as meats and dairy prod-ucts.36 The prevalence of infection among children par-ticipating in this study decreased after 1 year of receivingschool food supplementations (lower prevalence ofelevated C-reactive protein levels and enlarged spleenswith a decrease in plasma ferritin and copper levels).However, the researchers did not observe any significantimprovements in the levels of other micronutrients, suchas iron, zinc, copper, or vitamin A and could not findany diet-related explanation for the lack of significanteffects of meat and milk supplementation on these micro-nutrients. Researchers attributed the lack of significantincreases in vitamin A, iron, zinc, and copper levels to theconfounding effect of malaria and the high rate of infec-tion among this population. In a follow-up study, Murphyet al.37 reported changes in the children’s micronutrientintakes after comparing three 24-hour recalls collectedbefore school feeding started with three 24-hour recallscollected after feeding was initiated. The researchersobserved that the intakes of vitamin B12, vitamin A, ribo-flavin, and calcium of children in the milk group weregreater than those of children in the control group. Also,children in the meat group had higher intakes of vitaminB12, vitamin A, and available iron and zinc than thecontrol group. The total energy intake of children inthe meat group was also higher than that found in eitherthe milk or the energy groups.Although the energy intakeof children at home decreased upon consuming oneof the fortified snacks, the decrease in energy intake waslowest in the meat group. Thus, consumption of the meatsnacks in this study improved the dietary quality (micro-nutrient status) and dietary quantity (total energy intake)of participant children, whereas the milk snacks had apositive impact solely on dietary quality.

IMPACT OF SCHOOL FEEDING ON EDUCATIONAND LEARNING

In addition to the promising nutritional outcomes,studies highlight the importance and benefits of offeringschool-aged children school meals and/or snacks toimprove certain cognitive functions and scholasticachievement, especially among disadvantaged malnour-ished children. Thus, school feeding is once again of par-ticular significance in developing countries with thehighest percentages of malnourished children, low schoolenrollment, and high dropout rates. Almost all types ofSFPs reported in the literature demonstrate a positiveimpact on school enrollment and attendance.23,24,28,32,38,39

The impact of school feeding on academic achievementshows consistent positive effects on arithmetic tests,but lower effects on reading, writing, and spellingtests.16,18,23,24,26 School feeding helps improve school

progress by reducing the dropout rate. This applies toboth school meals and take-home rations, with greaterbenefits to girls, in particular, when both modalities ofschool feedings are offered together at school.38

EVALUATING THE EVIDENCE

Findings from the nutrition literature may be mixedand equivocal at times, thus weakening the evidence onhow effective SFPs are for improving various markers ofnutrition status, growth, and health. Mixed findings maybe attributed to a multitude of factors, mainly differencesin the objectives and methodologies used in SFPs. Differ-ences lie in the design of household and school surveys,the quality and quantity of food served to children, avail-able school resources, durations of the interventions(ranging from 1-month to 2- or 3-year interventions),and the modalities of school feeding (school meals, forti-fied snacks, take-home rations) in various settings andstudies. In addition, findings from various studies andnational programs are difficult to compare as outcomevariables and indicators vary, and the age groups, degreeof malnourishment, and severity of worm infections andillnesses in targeted school-aged children differ from oneprogram to another. Since study designs, sample popula-tions, and outcome variables vary considerably, meta-analyses of studies become more difficult to perform.Despite the somewhat mixed results, the present reviewdemonstrates that SFPs have promising, positive impactson the nutrition and health status of school-aged chil-dren. Providing food to children in the form of schoolmeals, snacks, or take-home rations can help alleviatehunger, address the nutrition needs of children, improvechildren’s micronutrient status (if food is fortified withessential micronutrients or if micronutrient supplementsare provided), and reduce the susceptibility of childrento infectious diseases and illnesses (as summarized inTable 3). However, the school feeding studies and pro-grams in developing countries published to date lack anin-depth investigation of whether children are receivingculturally and developmentally appropriate nutritionand health education lessons to complement the nutri-tious foods and snacks being offered during a regularschool day. According to Powell et al.,18 in order for theachievement levels of children to improve and for chil-dren in developing countries to fully benefit from theschool feeding and supplemental services offered at theirschools, integrated interventions that include nutrition,health, and educational components are needed. Success-ful school nutrition and feeding programs in developedcountries have learned the importance of integratingnutrition education into these programs.40,41

A meta-analysis that explores school feeding and itsimpact on the nutrition and cognition of school-aged

Nutrition Reviews® Vol. 69(2):83–98 93

children showed mixed, yet promising, results on changesin the nutritional status of school-fed children comparedto non-school-fed children in developing countries.16

Numerous studies in the nutrition literature, which werediscussed in this review, show that school feeding canenhance children’s diets by increasing the total energyintake of children consuming meals and/or snacks atschool. Although a substitution effect does occur, studieshave shown that the increase in total dietary intake fromschool feeding offsets any diminished intake at home. Thesubstitution effect varies depending on the modalityof school feeding (e.g., snacks/biscuits result in lowersubstitution effects than meals), as well as the timingand composition of meals (e.g., breakfast or mid-morning meals result in lower substitution effects thanlunches).16,42–44

A growing body of literature supports micronutrientfortification (fortifying commonly eaten foods) andsupplementation (providing nutrients through micronu-trient pills or suspensions) as part of the SFPs in commu-nities with high levels of micronutrient deficiencies.Studies highlight the importance of school deworming(anthelmintic treatments) in regions where high rates ofworm infections prevail. As a result, the WFP has beenstrongly recommending that multiple nutrient fortifica-tion and meal supplementation, as well as deworming, be

complementary services to school feeding and includedin the “essential package.”45 In addition, research showsthat the effect of school feeding on the dietary intake andgrowth of school-aged children is greater among thosethat are malnourished compared to adequately nourishedchildren.25,27 Furthermore, greater benefits are observedamong younger school-aged children.16 These findingshave very important implications regarding the designand targeting criteria that are set for SFPs. Thus, programdesigners and planners are encouraged to target childrenfrom highly food-insecure areas with high infection ratesin their populations in order to provide food to those whoare most susceptible to hunger and malnutrition and toovercome the burdens of illness and diseases that mayotherwise impair the children’s growth and nutrition.

Evidence on the positive impact of school feedingon education outcomes seems to be even stronger thanthat observed and reported with nutritional outcomes.The present review of the literature shows that SFPs havea positive impact on children’s school enrollment andattendance.23,24,28,32,38,39 The impact of school feeding onschool-aged children’s academic achievement is consis-tently positive for arithmetic tests, yet inconclusive forreading, writing, and spelling tests.16,18,23,24,26 Althoughthere is reasonable evidence for school feeding and itspositive effects on short-term cognitive functions and

Table 3 Summary of the scientific evidence on the impact of school feeding activities and provisions on intendednutrition, health, and educational outcomes in school-aged children and their households.School feeding provisions and activities Positive intended outcomesIn-school meals (breakfast, lunch, or mid-day), snacks, and

take-home rationsEnergy and micronutrient content of school meals, snacks,

and/or take-home rations

School-aged children (participants):Energy intake*Nutritional status*School enrollment*School attendance*Growth (weight and height)†

Cognition (memory, complex mental abilities) andclassroom behavior (attention and participation)†

Educational achievement (arithmetic and literacy tests)†

Household:Energy intake of siblings and other family members

(in-school meals and take-home rations)†

DewormingFrequency and dosage of anthelmentic treatment

Decreased morbidities and illnesses*Improved micronutrient/nutritional status of school-aged

children – if deworming is coupled with micronutrientfortification of school meals/snacks†

Health and nutrition education curriculaAge, developmentally and culturally appropriate nutrition

and health education lesson plans

School-aged children:Nutrition and health knowledge, attitudes, and behaviors‡

Household:(c) Nutrition and health knowledge and attitudes of

household members, and allocation of food and healthresources‡

* Conclusive: consistent findings that support the link between SFP provisions and intended outcomes.† Inconclusive: mixed results from the literature, weaker evidence on link between SFP provisions and intended outcomes.‡ Unexplored: no empirical evidence to explore the relationship between SFP provisions and intended outcomes.

Nutrition Reviews® Vol. 69(2):83–9894

various dimensions of children’s school performance,randomized controlled trials and well-designed evalua-tion studies are needed to demonstrate the long-termimpact of these programs on children’s school comple-tion and productivity as adults. Table 3 includes asummary of the scientific evidence that justifies the linkbetween school feeding provisions and activities andintended positive outcomes on the nutrition, health, andacademic performance of school-aged children. The evi-dence was considered conclusive if consistent findingswere reported across studies reviewed in this paper; theywere considered inconclusive if findings were mixed, andunexplored if there is a lack of scientific measurementof the link between the SFP provision or activity and theintended outcome.

RESEARCH AND EVALUATION GAPS

The school feeding literature is rich with well-designedtrials and country reports; however, rigorous theory-based impact evaluations are required in this area. Experi-mental and quasi-experimental studies on the nutritionalbenefits of school feeding may target and evaluate specificoutcomes and biomarkers of children’s nutrition andhealth status; however, these studies mostly fall short ofidentifying the causal chain or theory behind the imple-mented programs. Furthermore, many of these studieslack a detailed description of the political, social, andeconomic context of a SFP, thereby limiting the possibilityof replicating the program in similar contexts or thegeneralizability of the findings.

The International Initiative for Impact Evaluation(3ie)46 highlights the importance of theory-basedimpact evaluations of programs in developing countriesdesigned to improve the livelihood of populations.Numerous principles are needed for a theory-basedimpact evaluation including the existence of a programtheory (causal chain), understanding the context of theprogram, anticipating heterogeneity, conducting rigorousfactual analysis, and using mixed methods to validatethe outcomes and impact of a program.47 In the case ofschool feeding, there are insufficient rigorous theory-based evaluations of the nutritional and health impact ofthese programs on school-aged children in developingcountries. The theory and context of SFPs that are fundedand managed by the WFP or transitioning to becomenationally owned are well-defined. However, the design ofthese programs falls short from fulfilling other principlesneeded for theory-based impact evaluations, includingthe use or reporting of mixed methods (qualitative andquantitative data), and the availability of factual andcounterfactual analyses (experimental versus controlor pre- and post-intervention results) in some of theseprograms. On the other hand, experimental and quasi-

experimental studies assessing school feeding, such asthose reviewed in this paper, meet the scientific caliberneeded in impact evaluations yet often lack a clear defi-nition of the logical framework of the SFP being imple-mented; moreover, their inconclusive findings, at times,limit them from validating the links between resources,outputs, and outcomes within a causal chain of a schoolfeeding framework.

Furthermore, several questions were raised from theliterature on school feeding and should be addressedwhen future programs and trials are designed to test theeffectiveness of SFPs. Some of these research gaps werepointed out throughout this review, while others are sum-marized as follows. 1) Studies on the SFPs vary in designand some lack scientific rigor. Thus, efforts need to beexerted to provide guidance to researchers on what stan-dardized methods, designs, and outcomes would be bestimplemented and measured to improve the quality ofevidence on school feeding. 2) The benefits of schoolfeeding on nutrition, health, and education of school-aged children are identified in the literature; however,the scale of these benefits requires further exploration.3) Independent and combined effects of SFPs withsupplemental health and educational interventions,including school-based health and nutrition education,on the nutritional knowledge and eating behavior oftargeted children and their households are areas ofneeded study. 4) Studies that explore or control for thesubstitution effect of SFPs based on various factorsincluding modality, timing, and composition of snacks ormeals, age of child, and other household characteristicsare still limited, yet necessary. 5) Long-term studies andprogram evaluations are needed to track the impact ofschool feeding on nutritional/health status, educationalattainment, and productivity of children as they reachadulthood. 6) Research on the cost-drivers, cost-effectiveness, and benefit-to-cost ratio of various feedingprograms, including cross-sectional and randomizedtrials, is needed to better advise governments on how tobest allocate their limited resources

CHALLENGES AND FUTURE DIRECTIONS

School feeding programs face numerous and continuouschallenges, some of which are context-specific; howeverothers are more universal and apply to all SFPs imple-mented in developing countries. The sustainability ofthese programs, procurement of food in light of foodprice fluctuations and environmental and agriculturalchanges, as well as questions of a program’s cost-effectiveness are common challenges faced by SFP plan-ners and designers. Despite their importance, these issuesare beyond the scope of this review and were discussed in

Nutrition Reviews® Vol. 69(2):83–98 95

depth in the most recent joint WB and WFP publicationon school feeding.48,49

Based on the most recent “transition process” theo-ries and modules developed by the WFP and the WB withthe collaboration of national governments, the trend forSFPs is to become government-driven and owned, morecost-contained and cost-efficient, and linked with widerschool health and nutrition services such as water sanita-tion, supplementation, and deworming, among otherservices. The aim is for these SFPs to become embeddedinto national policies and frameworks. Questions remainabout the feasibility and effectiveness of nationallyowned, funded, and managed SFPs. Thus, a working data-base of success stories, country and program evaluationreports that can be publicly accessible and updated notonly by the international agencies and funders but alsoby national governments can help advance research inthis area. This would be an invaluable database and areference document that could be used by various stake-holders involved in designing, funding, implementing,and evaluating SFPs.

Local procurement of food is yet another area thatseems to be the focus of international non-governmentaland governmental organizations and that will be expand-ing to provide full coverage to the growing number ofSFPs, especially in response to the financial and foodcrises. In addition, home-grown school feeding (HGSF) isa new framework that is suggested to link school feedingwith local agricultural production.50 HGSF seems to bethe next revolution in school feeding. Therefore, moreinterest and funding might be channeled to expand onthis new framework. Furthermore, SFPs that offer morenutritious food fortified or supplemented with requiredmicronutrients will also be sought in order to increase theeffectiveness of SFPs and their impact on the nutrition,cognition, and health of children.

Researchers design rigorous randomized controlledtrials to examine the efficacy of SFPs and to establishcause and effect relationships, when possible, betweenschool feeding in its different modalities and positivenutrition and educational outcomes. However, ethicalissues arise as researchers deny or delay the provision offoods and complimentary services, such as anthelmentictreatment and micronutrient supplements, to childrenin control groups that may otherwise benefit from thetreatment.

CONCLUSION

Global efforts were exerted to set the MDGs, yet we arestill far from achieving those highly aspired goals. Mil-lions of children are still deprived of some of their basichuman rights to be fed and taken care of, to receive thenecessary medical and health care, to go to school and get

a formal education that prepares the child for a produc-tive adult life. The present review of the literature shows arelatively consistent positive effect of SFPs on the energyintake and micronutrient status of school-aged children,and a decline in infections and morbidities, particularly inprograms in which micronutrient fortification and de-worming are provided to school-aged children in regionswith multiple micronutrient deficiencies and heavy loadsof worm infections. Mixed findings were reported fromvarious studies in terms of the impact of school feedingon weight, height, and BMI gains among school-agedchildren. Long-term studies that assess the benefits ofschool meals, snacks, and take-home rations on children’sgrowth and household food consumption patterns arestill lacking. Furthermore, as SFPs shift from emergencyand post-crisis strategies to stable and long-term devel-opment projects and safety nets, complementary nutri-tion and health education curricula will need to befurther emphasized with provision of school meals andfood rations to address long-term sustainable improve-ments in nutrition.

In conclusion, the success and almost universality ofSFPs highlights the important yet challenging task of sus-taining these programs and ensuring that they expandand benefit a larger target population of school-aged chil-dren, preschoolers, and their households. Collaborativeefforts are required to ensure that SFPs not only expand,but become increasingly cost-effective and efficient andachieve intended outcomes in the short and long terms.Health professionals, educators, researchers, and commu-nity workers need to work together when developingnational and regional SFPs and designing trials to evalu-ate their impact on various intended outcomes.

Acknowledgments

This review was conducted in part at the World FoodProgramme (WFP) headquarters in Rome, Italy. Theauthors thank Mr. Andreas Psoroulas, the head of theWFP library, Ms. Nancy Walter, the Chief of the SchoolFeeding Policy at the WFP, and Ms. Carmen Burbano,co-author of the joint WFP and WB publication“Rethinking School Feeding: Social Safety Net, ChildDevelopment, and the Education Sector,” as well as allmembers of the School Feeding Division for providinghelpful information and sharing their expertise in thisarea.

Declaration of interest. The authors have no relevantinterests to declare.

REFERENCES

1. Food and Agriculture Organization. The State of Food Insecu-rity in the World 2009. Rome: FAO; 2009.

Nutrition Reviews® Vol. 69(2):83–9896

2. United Nations. Declaration of Millenium DevelopmentGoals: United Nations Resolution. 2000; Available at: http://www.un.org/millennium/declaration/ares552e.pdf.Accessed 8 March 2010.

3. World Bank. Education for All Fast Track Initiative. 2002;Available at: http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTEDUCATION/0,contentMDK:20278663~menuPK:617564~pagePK:148956~piPK:216618~theSitePK:282386,00.html. Accessed 5 August 2009.

4. United Nations Children’s Fund, World Food Programme. TheEssential Package: Twelve Interventions to Improve the Healthand Nutrition of School-Age Children. Rome: World Food Pro-gramme; 2005.

5. World Hunger Education Service. World Hunger PovertyFacts and Statistics 2010. 2010; Available at: http://www.worldhunger.org/articles/Learn/world%20hunger%20facts%202002.htm. Accessed January 1, 2011.

6. Weinreb L, Wehler C, Perloff J, et al. Hunger: its impacton children’s health and mental health. Pediatrics. 2002;110:e41.

7. Broca S, Stamoulis K. Micro- and macroevidence on theimpact of undernourishment. Nutrition Intake and EconomicGrowth. Economic and Social Development Department.Food and Agriculture Organization of the United Nations.2003. http://www.fao.org/docrep/006/y4850e/y4850e03.htm. Accessed January 5, 2010.

8. Victora CG, Adair L, Fall C, et al. Maternal and child undernu-trition: consequences for adult health and human capital.Lancet. 2008;371:340–357.

9. Food and Agriculture Organization. The State of Food Insecu-rity in the World: Eradicating World Hunger – Key to Achievingthe Millenium Development Goals. Rome: FAO; 2005.

10. World Food Programme. Figures on WFP School FeedingProgrammes. 2008; Available at: http://home.wfp.org/stellent/groups/public/documents/communications/wfp205343.pdf. Accessed 26 January 2010.

11. United Nations. Road Map toward the Implementation ofthe United Nations Millenium Declaration. 2000; Availableat: http://www.un.org/News/Press/docs/2001/pi1380.doc.htm. Accessed January 2, 2010.

12. Bundy D, Burbano C, Grosh M, Gelli A, Jukes M, Drake L. Whyimplement school feeding? Rethinking School Feeding: SocialSafety Nets, Child Development, and the Education Sector.Washington, DC: World Bank; 2009:15–34.

13. Bergeron G, Del Rosso JM. Food for Education IndicatorGuide. Food and Nutrition Assistance Project. 2001; Avail-able at: http://www.fantaproject.org/downloads/pdfs/FFE.pdf. Accessed 28 June 2010.

14. Bundy D, Burbano C, Grosh M, Gelli A, Jukes M, Drake L.Rethinking School Feeding: Social Safety Nets, Child Develop-ment, and the Education Sector. Washington, DC: World Bank;2009.

15. World Food Programme. School Feeding Policy: InformalConsultation. 2009; Available at: http://home.wfp.org/stellent/groups/public/documents/resources/wfp205242.pdf. Accessed 26 June 2010.

16. Kristjansson E, Robinson V, Petticrew M, et al. School feedingfor improving the physical and psychosocial health of disad-vantaged students. Cochrane Database Syst Rev. 2006;1. doi:10.4073/csr.2006

17. Neumann CG, Bwibo NO, Murphy SP, et al. Animal sourcefoods improve micronutrient status, growth and cognitivefunction in Kenyan school children: background, study designand baseline findings. J Nutr. 2003;133(Suppl):S3941–S3949.

18. Powell C, Walker SP, Chang SM, Grantham-McGregor SM.Nutrition and education: a randomized trial of the effectsof breakfast in rural primary school children. Am J Clin Nutr.1998;68:873–879.

19. Du X, Zhu K, Trube A, et al. School-milk intervention trialenhances growth and bone mineral accretion in Chinesegirls aged 10–12 years in Beijing. Br J Nutr. 2004;92:159–168.

20. Agarwal DK, Agarwal KN, Upadhyay SK. Effect of mid-daymeal programme on physical growth & mental function.Indian J Med Res. 1989;90:163–174.

21. Bailey KV. Rural nutrition studies in Indonesia IX: feeding trialon school boys. Trop Geogr Med. 1962;14:129–139.

22. Devadas RP, Jamala S, Surabhi A, Murthy NK. Evaluation ofa food supplement to school children. Indian J Nutr Diet.1979;16:335–341.

23. Simeon D. School feeding in Jamaica: a review of its evalua-tion. Am J Clin Nutr. 1998;67(Suppl):S790–S794.

24. Ahmed AU. The Impact of Feeding Children in School: Evidencefrom Bangladesh. Washington, DC: International Food PolicyResearch Institute; 2004.

25. Grillenberger M, Neumann CG, Murphy SP, et al. Foodsupplements have a positive impact on weight gain and theaddition of animal source foods increases lean body massof Kenyan schoolchildren. J Nutr. 2003;133(Suppl):S3957–S3964.

26. Neumann CG, Murphy SP, Gewa C, Grillenberger M,Bwibo NO. Meat supplementation improves growth, cogni-tive, and behavioral outcomes in Kenyan children. J Nutr.2007;137:1119–1123.

27. Kruger M, Badenhorst CJ, Laubscher JA, Spinnler Benade AJ.Effects of iron fortification in a school feeding scheme andanthelmintic therapy on the iron status and growth of six- toeight year-old schoolchildren. Food Nutr Bull. 1996;17:11–21.

28. Jacoby E, Cueto S, Pollitt E. Benefits of a school breakfastprogramme among Andean children in Huaraz. Bull NutrFood. 1996;17:54–64.

29. Jacoby E, Cueto S, Pollitt E. When science and politics listen toeach other: good prospects from a new school breakfastprogram in Peru. Am J Clin Nutr. 1998;67:(Suppl):S795–S797.

30. Afridi F. Child welfare programs and child nutrition: evidencefrom a mandated school meal program in India. J Dev Econ.2010;92:152–165.

31. Arsenault JE, Mora-Plazas M, Forero Y, et al. Provision of aschool snack is associated with vitamin B-12 status, lineargrowth, and morbidity in children from Bogota’, Colombia. JNutr. 2009;139:1744–1750.

32. Van Stuijvenberg ME, Kvalsvig JD, Faber M, Kruger M,Kenoyer DG, Benade AJS. Effect of iron-, iodine-, and beta-carotene-fortified biscuits on the micronutrient status ofprimary school children: a randomized controlled trial. Am JClin Nutr. 1999;69:497–503.

33. Van Stuijvenberg ME. Using the school feeding system as avehicle for micronutrient fortification: experience from SouthAfrica. Food Nutr Bull. 2005;26(Suppl 2):S213–S219.

34. Hurrell RF. Preventing iron deficiency through food fortifica-tion. Nutr Rev. 1997;55:210–222.

35. Walter T, Hertrampf E, Pizarro F, et al. Effect of bovine-hemoglobin-fortified cookies on iron status of schoolchildren: a nationwide program in Chile. Am J Clin Nutr.1993;57:190–194.

36. Seikmann J, Allen LH, Bwibo NO, Demment MW, Murphy SP,Neumann CG. Kenyan school children have multiple micro-nutrient deficiencies, but increased plasma vitamin B12 is

Nutrition Reviews® Vol. 69(2):83–98 97

the only detectable micronutrient response to meat or milksupplementation. J Nutr. 2003;133(Suppl):S3972–S3980.

37. Murphy SP, Gewa C, Lliang LJ, Grillenberger M, Bwibo NO,Neumann CG. School snacks containing animal source foodsimprove dietary quality for children in rural Kenya. J Nutr.2003;133(Suppl):S3950–S3956.

38. Gelli A, Meir U, Espejo F. Does provision of food in schoolincrease girls’ enrollment? evidence from schools insub-saharan Africa. Food Nutr Bull. 2007;28:149–155.

39. Ahmed AU, del Ninno C. The Food for Education Program inBangladesh: An Evaluation of Its Impact on Educational Attain-ment and Food Security. Washington, DC: International FoodPolicy Research Institute; 2002.

40. McConnell PE, Shaw JB. Position of the American DieteticAssociation: child and adolescent food and nutritionprograms. J Am Diet Assoc. 1996;96:913–917.

41. Centers for Disease Control and Prevention. Guidelines forschool health programs to promote lifelong healthy eating.Morb Mortal Wkly Rep. 1996;45(RR-9):1–33.

42. Orphans and Vulnerable Children (OVC) Toolkit. EducationSector Programs. World Bank. 2009; Available at: http://info.worldbank.org/etools/docs/library/248484/toolkiten/sector/education4.htm. Accessed 26 February 2010.

43. The Caroline Walker Trust and the National HeartForum. Eating Well at School: Nutritional and PracticalGuidelines. 2005; Available at: http://www.cwt.org.uk/pdfs/EatingWellatSchool.pdf. Accessed 29 July 2010.

44. Centers for Disease Control and Prevention, Division ofAdolescent and School Health. Coordinated SchoolHealth Program. 2008; Available at: http://www.cdc.gov/healthyyouth/CSHP/. Accessed 29 July 2010.

45. Bundy D, Burbano C, Grosh M, Gelli A, Jukes M, Drake L. Whatis school feeding? Rethinking School Feeding: Social SafetyNets, Child Development, and the Education Sector. Washing-ton, DC: The World Bank; 2009:9–14.

46. International Initiative for Impact Evaluation (3ie). GlobalDevelopment Network. 2007; Available at: http://www.3ieimpact.org/. Accessed 9 July 2010.

47. White H. Theory-Based Impact Evaluation: Principles andPractice (Working Paper 3). 2009; Available at: http://www.3ieimpact.org/admin/pdfs_papers/51.pdf. Accessed 23 July2010.

48. Bundy D, Burbano C, Grosh M, Gelli A, Jukes M, Drake L.What is school feeding? Rethinking School Feeding: SocialSafety Nets, Child Development, and the Education Sector.Washington, DC: The World Bank; 2009:9–14.

49. Bundy D, Burbano C, Grosh M, Gelli A, Jukes M, Drake L.Institutional and procurement arrangements. RethinkingSchool Feeding: Social Safety Nets, Child Development, and theEducation Sector. Washington, DC: The World Bank; 2009:69–81.

50. Espejo F, Burbano C, Galliano E. Home-Grown School Feeding:A Framework to Link School Feeding with Local AgriculturalProduction. Rome: World Food Programme (WFP); 2009.

Nutrition Reviews® Vol. 69(2):83–9898

Copyright of Nutrition Reviews is the property of Wiley-Blackwell and its content may not be copied or

emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.

However, users may print, download, or email articles for individual use.