systematic review of the efficacy and effectiveness of complementary feeding interventions

Systematic review of the Systematic review of the efficacy and efficacy and

effectiveness of effectiveness of complementary feeding complementary feeding

interventionsinterventionsKathryn G. Dewey, PhDKathryn G. Dewey, PhDSeth Adu-Afarwuah, PhDSeth Adu-Afarwuah, PhD

Program in International and Community Program in International and Community NutritionNutrition

University of California, DavisUniversity of California, Davis

Supported by the Mainstreaming Nutrition InitiativeSupported by the Mainstreaming Nutrition Initiative with funding from the World Bankwith funding from the World Bank

Complementary feeding is a Complementary feeding is a key window for interventionkey window for intervention

Age range of complementary feeding (6-Age range of complementary feeding (6-24 mo) is the time of peak incidence of24 mo) is the time of peak incidence of– Growth falteringGrowth faltering– Micronutrient deficienciesMicronutrient deficiencies– Morbidity, e.g. diarrheal diseaseMorbidity, e.g. diarrheal disease

After 2 years, difficult to reverse effects of After 2 years, difficult to reverse effects of malnutritionmalnutrition– StuntingStunting– Effects on brain function due to micronutrient Effects on brain function due to micronutrient

deficiency?deficiency? A food-based, comprehensive approach A food-based, comprehensive approach

may be more effective and sustainable may be more effective and sustainable than programs targeting individual than programs targeting individual nutrient deficienciesnutrient deficiencies

Guiding principles for Guiding principles for complementary feeding of the complementary feeding of the

breastfed child (PAHO/WHO 2003)breastfed child (PAHO/WHO 2003)

1.1. Age of introduction of complementary foodsAge of introduction of complementary foods2.2. Maintenance of breastfeedingMaintenance of breastfeeding3.3. Responsive feedingResponsive feeding4.4. Safe preparation & storage of complementary foodsSafe preparation & storage of complementary foods5.5. Amount of complementary foods neededAmount of complementary foods needed6.6. Food consistencyFood consistency7.7. Meal frequency and energy densityMeal frequency and energy density8.8. Nutrient content of complementary foodsNutrient content of complementary foods9.9. Use of vitamin/mineral supplements or fortified Use of vitamin/mineral supplements or fortified

productsproducts10.10. Feeding during and after illnessFeeding during and after illness

Scope of ReviewScope of Review

Interventions in developing countries Interventions in developing countries that targeted children 6-24 mothat targeted children 6-24 mo

Outcomes measured: growth, Outcomes measured: growth, morbidity, child development, morbidity, child development, micronutrient intake, micronutrient micronutrient intake, micronutrient statusstatus

Studies that assessed only the impact Studies that assessed only the impact on feeding practices were not includedon feeding practices were not included

Generally focused on reports from Generally focused on reports from 1996-20061996-2006

Number of papers includedNumber of papers included

SourceSource Efficacy Efficacy trialstrials

Effectiveness Effectiveness studies/program studies/program

reportsreports

PubMedPubMed 1212 00

Snowball Snowball techniquetechnique

1111 77

Personal Personal contactscontacts

66 66

Total (42)Total (42) 2929 1313

Intervention StrategiesIntervention Strategies Education as the main treatmentEducation as the main treatment Complementary food or a food product Complementary food or a food product

offering extra energy (with or without added offering extra energy (with or without added micronutrients) provided as the only micronutrients) provided as the only treatmenttreatment

Provision of food combined with some other Provision of food combined with some other strategy, usually education for mothersstrategy, usually education for mothers

Fortification of complementary foods Fortification of complementary foods (central or home-fortification) with (central or home-fortification) with micronutrients (with no difference in energy micronutrients (with no difference in energy provided to intervention vs. control groups)provided to intervention vs. control groups)

Increased energy density and/or nutrient Increased energy density and/or nutrient bioavailability of complementary foods via bioavailability of complementary foods via simple technologiessimple technologies

Effect size (ES) = Effect size (ES) =

DDifference betw. intervention & control ifference betw. intervention & control groupsgroups

Pooled SDPooled SD

Impact on growth outcomes Impact on growth outcomes via educational approaches: via educational approaches:

efficacy trialsefficacy trials

Author, dateAuthor, date SiteSite NN ES, ES,

weighweightt

ES, ES,

lengthlength

Bhandari, Bhandari, 20012001

IndiaIndia 188188 0.140.14 0.080.08


IndiaIndia 829829 0.020.02 0.050.05

Penny, 2005Penny, 2005 PeruPeru 338338 0.34*0.34* 0.49*0.49*

Roy, 2005Roy, 2005 BangladeshBangladesh 183183 0.58*0.58* 0.090.09

Santos, 2001Santos, 2001 BrazilBrazil 404404 0.090.09 0.040.04

Vitolo, 2005Vitolo, 2005 BrazilBrazil 397397 -- NSNS

Impact on growth outcomes Impact on growth outcomes via educational approaches: via educational approaches:

effectiveness/program studieseffectiveness/program studies


weighweightt

ES, ES,

lengthlength

Gulden, 2000Gulden, 2000 ChinaChina 495495 0.96*0.96* 0.64*0.64*

Guyon, 2005Guyon, 2005 MadagascarMadagascar NANA -0.06-0.06 0.140.14

Kilaru, 2005Kilaru, 2005 IndiaIndia 242242 0.160.16 --

Maluccio, Maluccio, 20042004

NicaraguaNicaragua NANA -- 0.120.12

Impact on growth outcomes Impact on growth outcomes via provision of via provision of

complementary foodcomplementary foodAuthor, dateAuthor, date SiteSite NN ES, ES,

weighweightt

ES, ES,

lengthlength

Adu-Afarwuah, Adu-Afarwuah, 20062006

GhanaGhana 194194 0.31*0.31* 0.26*0.26*

Beckett, 2000Beckett, 2000 IndonesiaIndonesia 7878 0.030.03 0.020.02

Kuusipalo, 2006Kuusipalo, 2006 MalawiMalawi 112112 0.51*0.51* 0.67*0.67*

Lartey, 1999Lartey, 1999 GhanaGhana 190190 0.57*0.57* 0.69*0.69*

Obatolu, 2003Obatolu, 2003 NigeriaNigeria 6060 2.99*2.99* 1.81*1.81*

Oelofse, 2003Oelofse, 2003 South AfricaSouth Africa 3030 -0.02-0.02 -0.04-0.04

Owino, in pressOwino, in press ZambiaZambia 106106 0.300.30 0.37*0.37*

Santos, 2005Santos, 2005 BrazilBrazil 191191 0.100.10 -0.02-0.02

Impact on growth outcomes Impact on growth outcomes via comp. food + education:via comp. food + education:

efficacy trialsefficacy trials


weighweightt

ES, ES,

lengthlength


IndiaIndia 178178 0.32*0.32* 0.250.25

Roy, 2005Roy, 2005 BangladeshBangladesh 189189 0.66*0.66* 0.32^0.32^

^significant in children < 15 mo and low in WAZ or LAZ at baseline

Impact on growth outcomes Impact on growth outcomes via comp. food + education: via comp. food + education:

effectiveness/program studieseffectiveness/program studiesAuthor, dateAuthor, date SiteSite NN ES, ES,

weightweightES, ES,

lengthlength

Gartner, 2007Gartner, 2007 SenegalSenegal 16761676 NSNS NSNS

Hossain, 2005Hossain, 2005 BangladesBangladeshh

23882388 NSNS NSNS

Lutter, 2006Lutter, 2006 EcuadorEcuador 319319 0.24*0.24* 0.140.14

Lopez de Romana,Lopez de Romana, 20002000

PeruPeru NANA NANA NSNS

Rivera, 2004Rivera, 2004 MexicoMexico 650650 NANA 0.12^0.12^

Schroeder, 2002Schroeder, 2002 VietnamVietnam 230230 0.18^^0.18^^

0^^0^^^significant only in children < 6 mo at baseline

^^significant only in children < 15 mo and underweight or stunted at baseline

Impact on growth outcomes Impact on growth outcomes via fortification of comp. foods: via fortification of comp. foods:

efficacy trialsefficacy trialsAuthor, dateAuthor, date SiteSite NN ES, ES,

weightweightES, ES,

lengthlength

Adu-Afarwuah, Adu-Afarwuah, 20072007

GhanaGhana 296296 0.030.03 -0.02-0.02

Dhingra, 2004Dhingra, 2004 IndiaIndia 570570 0.37*0.37* 0.45*0.45*

Faber, 2005Faber, 2005 South South AfricaAfrica

292292 0.000.00 0.110.11

Giovannini, 2006Giovannini, 2006 CambodiaCambodia 127127 -0.22-0.22 0.070.07

Lartey, 1999Lartey, 1999 GhanaGhana 190190 0.200.20 0.080.08

Smuts, 2005Smuts, 2005 South South AfricaAfrica

9999 0.290.29 0.050.05

Impact on growth outcomes Impact on growth outcomes via interventions to increase via interventions to increase

energy density of energy density of complementary foodscomplementary foods


weightweightES, ES,

LengthLength

Hossain, 2005Hossain, 2005 BangladesBangladeshh

100100 0.250.25 0.320.32

John, 1993John, 1993 IndiaIndia 4242 1.37*1.37* 0.71*0.71*

Mamiro, 2004Mamiro, 2004 TanzaniaTanzania 258258 NANA -0.04-0.04

Moursi, 2003Moursi, 2003 CongoCongo 7575 -0.13-0.13 0.40*0.40*

Owino, 2007Owino, 2007 ZambiaZambia 113113 -0.07-0.07 -0.25-0.25

Effect size for weight growth of Effect size for weight growth of different intervention different intervention

strategiesstrategies

ED = Education; FD = Complementary food; FD+ED = Education + complementary food; FT = Fortification of comp. foods; EN = Increased energy density

Excluding Obatolu, 2003 (outlier)

-0.50

-0.25

0.00

0.25

0.50

0.75

1.00

1.25

1.50

ED FD FD+ED FT EN

Intervention strategy

Eff

ects

siz

e fo

r w

eigh

t gr

owth

Max Min Mean

Effect size for linear growth of Effect size for linear growth of different intervention different intervention

strategiesstrategies

ED = Education; FD = Complementary food; FD+ED = Education + complementary food; FT = Fortification of comp. foods; EN = Increased energy density

Excluding Obatolu, 2003 (outlier)

-0.50

-0.25

0.00

0.25

0.50

0.75

1.00

ED FD FD+ED FT EN

Intervention strategy

Effe

ct s

ize

for

line

ar

gro

wth

Max Min Mean

Impact of complementary Impact of complementary feeding interventions on feeding interventions on

morbidity outcomesmorbidity outcomes

Only 10 of 42 papers presented data Only 10 of 42 papers presented data on morbidityon morbidity

Most showed no impact, but Most showed no impact, but generally not designed or powered to generally not designed or powered to evaluate morbidity as a primary evaluate morbidity as a primary outcomeoutcome


morbidity outcomesmorbidity outcomes 4 studies showed reduced morbidity:4 studies showed reduced morbidity:

– Education only; Brazil (Vitolo, 2005)Education only; Brazil (Vitolo, 2005) Reduced diarrhea and URIReduced diarrhea and URI

– Food + education; Vietnam (Schroeder, Food + education; Vietnam (Schroeder, 2002)2002) Reduced URIReduced URI

– Fortified CF; India (Sazawal, 2007)Fortified CF; India (Sazawal, 2007) Reduced diarrhea, LRI and feverReduced diarrhea, LRI and fever

– Fortified CF; Pakistan (Sharieff, 2006)Fortified CF; Pakistan (Sharieff, 2006) Reduced diarrhea and feverReduced diarrhea and fever


morbidity outcomesmorbidity outcomes

2 studies showed increased 2 studies showed increased morbidity:morbidity:– Food + education; India (Bhandari, Food + education; India (Bhandari,

2001)2001) Increased fever & dysenteryIncreased fever & dysentery Possibly due to reduced breastfeedingPossibly due to reduced breastfeeding

– Increased energy density (Moursi, 2003)Increased energy density (Moursi, 2003) Increased URIIncreased URI

Impact of complementary Impact of complementary feeding interventions on child feeding interventions on child

developmentdevelopment Only 4 of 42 papers presented data on Only 4 of 42 papers presented data on

developmentdevelopment Home-fortification improved gross motor Home-fortification improved gross motor

development in Ghana (Adu-Afarwuah, development in Ghana (Adu-Afarwuah, 2007): 2007): % walking at 12 mo: % walking at 12 mo: – 25% in non-intervention group25% in non-intervention group– 36% with foodlet; 39% with Sprinkles; 49% with 36% with foodlet; 39% with Sprinkles; 49% with

fat-based fortified productfat-based fortified product No significant impact of fortification in South No significant impact of fortification in South

Africa (Oelofse, 2003) or India (Dhingra, Africa (Oelofse, 2003) or India (Dhingra, 2004)2004)

Provision of extra energy increased mental Provision of extra energy increased mental scores in Indonesia, but only in the subgroup scores in Indonesia, but only in the subgroup with low initial length-for-age (Pollitt, 2002)with low initial length-for-age (Pollitt, 2002)


micronutrient intakemicronutrient intake Education alone can increase intake of Education alone can increase intake of

Fe (by ~24%) and Zn (by ~26%)Fe (by ~24%) and Zn (by ~26%) Soaking + germination had no Soaking + germination had no

significant effect in Tanzania (Mamiro, significant effect in Tanzania (Mamiro, 2004)2004)

Fortification strategies can have a much Fortification strategies can have a much larger impactlarger impact– Fe intake increased by 145-207%Fe intake increased by 145-207%– Zn intake increased by 201-271%Zn intake increased by 201-271%– Vit A intake increased by 107-2300%Vit A intake increased by 107-2300%

Impact of complementary Impact of complementary feeding interventions on Fe feeding interventions on Fe

statusstatus Education alone improved Fe status Education alone improved Fe status

in some sites (India; China) but not in some sites (India; China) but not others (Brazil; Nicaragua)others (Brazil; Nicaragua)– Overall impact +4 g/L Hb, -5 PP anemiaOverall impact +4 g/L Hb, -5 PP anemia

Provision of Fe-fortified food Provision of Fe-fortified food – Overall impact +4-6 g/L Hb, -13-17 PP Overall impact +4-6 g/L Hb, -13-17 PP

anemia (12 studies)anemia (12 studies) Home fortificationHome fortification

– Overall impact +8 g/L Hb, - 21 PP Overall impact +8 g/L Hb, - 21 PP anemia (7 studies) anemia (7 studies)

Impact of complementary Impact of complementary feeding interventions on Zn feeding interventions on Zn

statusstatus

Only 5 studies reported plasma ZnOnly 5 studies reported plasma Zn Four studies evaluated a fortified Four studies evaluated a fortified

comp. food (3-6.5 mg/d Zn); none comp. food (3-6.5 mg/d Zn); none had signif. impacthad signif. impact

One study (South Africa) showed One study (South Africa) showed increased plasma Zn with home-increased plasma Zn with home-fortification (10 mg/d Zn)fortification (10 mg/d Zn)

Impact of complementary Impact of complementary feeding interventions on vitamin feeding interventions on vitamin

A statusA status 7 studies reported vitamin A status7 studies reported vitamin A status All involved a fortified comp. food or home All involved a fortified comp. food or home

fortificationfortification Amount of vitamin A provided ranged Amount of vitamin A provided ranged

widely: 83-658 ug RE/dwidely: 83-658 ug RE/d Significant increase in serum vitamin A in 4 Significant increase in serum vitamin A in 4

of the 5 studies with fortified comp. foodsof the 5 studies with fortified comp. foods No significant impact in 2 studies of home-No significant impact in 2 studies of home-

fortification, probably due to vit A fortification, probably due to vit A supplementation programssupplementation programs

ConclusionsConclusions

No single universal “best” package of No single universal “best” package of components in complementary components in complementary feeding interventionsfeeding interventions

Impact is context-specificImpact is context-specific– Initial prevalence of malnutritionInitial prevalence of malnutrition– Degree of household food insecurityDegree of household food insecurity– Energy density of traditional Energy density of traditional

complementary foodscomplementary foods– Availability of micronutrient-rich local Availability of micronutrient-rich local

foodsfoods

Conclusions - GrowthConclusions - Growth

Growth may not be the most Growth may not be the most sensitive indicator of impactsensitive indicator of impact

Impact may be greater in younger Impact may be greater in younger age groups: should begin CF age groups: should begin CF programs during infancyprograms during infancy

Effect sizes generally modest (0.1-Effect sizes generally modest (0.1-0.5), but potential larger if optimal 0.5), but potential larger if optimal design and implementation (0.5-0.6)design and implementation (0.5-0.6)

Conclusions – Growth (cont.)Conclusions – Growth (cont.)

Educational approaches more likely Educational approaches more likely to have impact if there is an to have impact if there is an emphasis on nutrient-rich animal-emphasis on nutrient-rich animal-source foodssource foods

Provision of food – variable resultsProvision of food – variable results– Greater impact in Africa & S Asia – due Greater impact in Africa & S Asia – due

to food insecurity?to food insecurity?– 2 studies compared food + education 2 studies compared food + education

vs. education only: somewhat greater vs. education only: somewhat greater impact when food includedimpact when food included

Conclusions – Growth (cont.)Conclusions – Growth (cont.) Most of the foods provided were fortified, Most of the foods provided were fortified,

so can’t distinguish impact of increased so can’t distinguish impact of increased energy/protein/fat from micronutrientsenergy/protein/fat from micronutrients– In Ghana, impact on weight gain partially In Ghana, impact on weight gain partially

explained by increased energy intake, but explained by increased energy intake, but impact on length gain related to change in impact on length gain related to change in plasma fatty acid profileplasma fatty acid profile

Micronutrient fortification alone has little Micronutrient fortification alone has little effect on growtheffect on growth– Exception: relatively large study in India in Exception: relatively large study in India in

which many children stunted at baseline & which many children stunted at baseline & fortified product resulted in reduced fortified product resulted in reduced morbidity morbidity

Conclusions – Growth (cont.)Conclusions – Growth (cont.)

Interventions to increase energy Interventions to increase energy density – results mixeddensity – results mixed– 3 of 5 studies had no impact on energy 3 of 5 studies had no impact on energy

intake or growthintake or growth– 2 of 5 studies had positive impact on 2 of 5 studies had positive impact on

growthgrowth May be effective when traditional CF May be effective when traditional CF

has low energy density & infant unable has low energy density & infant unable to compensate by increasing volume to compensate by increasing volume of food consumed or feeding of food consumed or feeding frequencyfrequency

Conclusions - MorbidityConclusions - Morbidity

Few studies had adequate N to Few studies had adequate N to evaluate morbidityevaluate morbidity

Mixed resultsMixed results– Beneficial impact in 4 studiesBeneficial impact in 4 studies– Adverse impact in 2 studiesAdverse impact in 2 studies

CF interventions need to include CF interventions need to include counseling on maintaining counseling on maintaining breastfeeding, responsive feeding and breastfeeding, responsive feeding and hygienic practiceshygienic practices

Conclusions - DevelopmentConclusions - Development

Very few studies evaluated Very few studies evaluated behavioral developmentbehavioral development

Promising results in Ghana using Promising results in Ghana using specially designed fortified fat-based specially designed fortified fat-based complementary foodscomplementary foods

Future evaluations of CF Future evaluations of CF interventions should include interventions should include assessment of child development – assessment of child development – may be more sensitive than growthmay be more sensitive than growth

Conclusions – Micronutrient Conclusions – Micronutrient intakeintake

Very difficult to achieve adequate Fe Very difficult to achieve adequate Fe intake from local foods without intake from local foods without fortification, at 6-12 mofortification, at 6-12 mo

Fortification increased Fe intake by 5-Fortification increased Fe intake by 5-11 mg/d11 mg/d

Can achieve adequate Zn and Vit A Can achieve adequate Zn and Vit A intakes from local foods, but requires intakes from local foods, but requires careful attention to dietary choicescareful attention to dietary choices

Fortification can help ensure Zn and Fortification can help ensure Zn and Vit A intakes when nutrient-rich local Vit A intakes when nutrient-rich local foods are costly or unavailable (e.g. foods are costly or unavailable (e.g. seasonally)seasonally)

Conclusions – Micronutrient Conclusions – Micronutrient statusstatus

Education can have a positive impact on Fe Education can have a positive impact on Fe status if Fe-rich foods are emphasizedstatus if Fe-rich foods are emphasized

A larger impact on Fe status can be expected A larger impact on Fe status can be expected from use of fortified products (reduction of 13-from use of fortified products (reduction of 13-21 percentage points in prevalence of anemia)21 percentage points in prevalence of anemia)

Little or no impact of fortification on plasma Little or no impact of fortification on plasma zinc – due to low absorption?zinc – due to low absorption?

Results mixed regarding fortification with Vit AResults mixed regarding fortification with Vit A– Positive impact in several studiesPositive impact in several studies– Little impact in some studies, probably due to Little impact in some studies, probably due to

concomitant vit A capsule distribution programsconcomitant vit A capsule distribution programs

Overall ConclusionsOverall Conclusions

Educational approaches can be effective, Educational approaches can be effective, but in many situations a greater impact but in many situations a greater impact may be seen when combined with home-may be seen when combined with home-fortification or provision of fortified foodsfortification or provision of fortified foods

To be most cost-effective and avoid To be most cost-effective and avoid displacement of breast milk, the amount of displacement of breast milk, the amount of food provided should be modest: no more food provided should be modest: no more than 200 kcal/d at 6-12 mothan 200 kcal/d at 6-12 mo

Biggest challenge: going to scale with a Biggest challenge: going to scale with a combination of the most cost-effective combination of the most cost-effective components, while assuring adequate components, while assuring adequate delivery and sustainabilitydelivery and sustainability

systematic review of the efficacy and effectiveness of complementary feeding interventions

Documents