Stunting determinants in Indonesia:

What are they and how can we measure them

in national surveys?

Rosalind S Gibson,
Emeritus Professor,
Department of Human Nutrition,
University of Otago,
New Zealand
 Outline
• Prevalence & phases in life cycle when stunting occurs
• Adverse consequences of stunting
WHO Framework: Causes of childhood stunting
• Household and family factors;
• Breastfeeding
• Inadequate complementary feeding
• Infection
WHO Framework: Context-Community and societal factors
• Political economy; Health and healthcare
• Education; Society and culture
• Agriculture & food; Water, sanitation & environment
• Conclusions
 Prevalence of stunting in Indonesian children
0-59 months by district in 2013

Indonesia has 5th highest burden of stunted children in the world

In 2013: >37% < 5y stunted; > 40% < 5 y in 15 of 33 provinces
Large disparities subnationally: 26% in Riau Islands to 52% in East Nusa
Tenggara suggesting variations in exposure to stunting determinants
Phases in life cycle when stunting occurs & etiology

Inter-generational
• Genetics
• Short maternal stature; SGA at birth
Pre-conception
• Adolescent pregnancy; Short birth interval
Prenatal period
• Poor maternal nutrition; infections & other
environmental exposures
Postnatal period (0-5 mos; 6-24 mos)
• Inappropriate breastfeeding & IYCF practices
• Inadequate amount & quality of complementary foods

Poor maternal nutrition & socio-cultural factors, infection and other

environmental exposures play a role in all four phases
Adverse consequences of stunting
Health Developmental Economic
Concurrent problems & short-term consequences

↑Mortality ↓ Cognitive, motor & ↑ Health

language expenditures
↑ Morbidity development
↑Opportunity
costs for care of
sick children

Long-term consequences
↓ Adult stature ↓School performance ↓ Work capacity
↓ Reproductive ↓Learning capacity ↓ Work
health outcomes Unachieved potential productivity
↑ Obesity etc

From de Onis & Branca (2016)

 WHO Framework: Causes of childhood stunting
Causes Stunted growth and development (4)
Household Inadequate Breast Infection
& family complementary feeding feeding
factors
Maternal Poor quality foods Inadequate Clinical &
factors Inadequate practices practices Subclinical
Home Food and water safety infection
environment

Context Community and societal factors (6)

Political Health & Education Society Agriculture Water,
economy health care & & food sanitation &
culture systems environment
 Household & family factors: Maternal factors (9)
Factor Indicator Description
Short maternal stature Maternal height < 145 cm* % WCBA with height <145 cm
Adolescent pregnancy Age (y) on birth of first child* % 15-19 y had 1st child
Short birth spacing Nos months preceding birth % < 6 mos (for stunting)
Poor nutrition: Pre- BMI; Mn intakes; Hb g/L; % BMI < 18.5; deficits in MN
conception, Preg,Lact plasma MN biomarkers /L intakes/biomarkers; Hb<110 g/L
Hypertension Blood pressure % High blood pressure

IUGR & preterm birth BW*; Gestational age from % BW< 2.5 kg AGA; Gestational
LDMP; fundal height; age <37wks
ultrasound
Birth weight & length Birth weight*; birth length % BW <2.5 kg; Birth LAZ <-1

Infection Serum CRP; AGP % CRP > 5 mg/L; AGP > 1g/L
Mental health ?

Factors in red associated with child stunting in Indonesia; * included in IDHS

 Lactating women in Sumedang district:
Micronutrient intakes &an prevalence of adequacy (PA) as
%
Nutrient Median IQR PA (%)
Vitamin A (RAE) 501 (319-841) 57
Thiamin (mg) 1.4 (0.98-1.84) 60
Riboflavin (mg) 1.7 (1.2-2.2) 69
Niacin (mg) 12.8 (10.1-15.5) 47
Vitamin B6 (mg) 1.3 (1.0-1.7) 25
Folate (µg) 618 (478-836) 79
Vitamin B12 (µg) 2.5 (1.8-3.1) 52
Vitamin C (mg) 38 (25-62) 28
Calcium (mg) 613 (509-750) 51
Iron (mg) 18.3 (12.5-23.0) 79
Zinc (mg) 12.8 (10.8-15.0) 97
MPA (SD) 57% (±28%)
MPA: Mean prevalence of adequacy; Nutrients in green: contributed by fortified wheat flour
Prevalence of adequacy < 50% for niacin, vit C, and B-6;51% for Ca; 52% for B-12

EARs from WHO/FAO (2004) except Ca & Fe from IOM and Zn from IZiNCG
Data from Rahmania et al. (2018)
 Household & family factors: Home environment (5)

Factor Indicator Description

Food insecurity HFIAS#: 8Q’s yielding % with resource scores
scores up to 120; 0=none >xxx to 120 over past 4 wks
Poor caring practices PAHO/UNICEF 2003: 7 % Responding positively
Questions from ProPAN per Q
Inadequate sanitation & HH sanitation facility & Non-improved supply
water supply drinking water supply* Other facility

Household wealth Wealth index subdivided % HHs in lowest wealth

by quintiles* quintile

Low caregiver education Education level * % caregivers with 1ary

school level or less
* Included in IDHS
NB: Improved water source does NOT ensure water is free of fecal contamination (See Bain et al. 2014)
HFIAS: Household Food Insecurity Access Scale. See Coates et al. (2007). Indicator Guide (v.3).
Washington DC FHI360/FANTA
ProPAN: Process for the promotion of child feeding. Field Manual. See PAHO/UNICEF, 2003
Household & family factors: Home environment
cont. (5)
Factor Indicator Description
Smoking (paternal & Parental use of % parents using tobacco
maternal) tobacco*
Paternal short stature Fathers height- % fathers with HAZ< -1
for-age Z score (for M >19 y)
Crowded household # persons per % HHs > 5
sleeping room persons/room??
Inappropriate intra-HH ?
food allocation

Inadequate child ?
stimulation & activity

* Included in IDHS
 Why focus on early neonatal & complementary feeding period:
Mean LAZ/HAZ-scores by age in relation to WHO growth standards

EURO: Europe
EMRO: Eastern Mediterranean
PAHO: Latin America &
Caribbean
AFRO: Africa
SEARO:SE Asia

• LAZ scores already low at birth in some regions & decline sharply during first 24 mos
• After 24 mos HAZ-scores show no further decline or any improvement thereafter

Maternal & Child Nutrition

pages 12-26, 17 MAY 2016 DOI: 10.1111/mcn.12231 From Victora et al. (2010)
http://onlinelibrary.wiley.com/doi/10.1111/mcn.12231/full#mcn12231-fig-0004
 Two pathways lead to stunting postnatally:
inadequate diet & infections

Inadequate Poor dietary Micronutrient

diet quality deficiencies

Inadequate Macronutrient
quantity of food deficiency
Energy deficit :
Lack of nutrients
Synergism & available at
interaction Symptomatic Reduced nutrient cellular level for
infections absorption growth

Infections Subclinical infections: Direct nutrient

Environmental enteropathy losses
GROWTH
FALTERING
Inflammatory &/or Increased nutrient
Adapted from Panter- immune response requirements
Brick et al. 2009
Core WHO indicators for inadequate diet pathway:
Optimal IYCF practices
1. Early initiation of breastfeeding (BF) (within 1 hr)
2. Exclusive BF of children under 6 mos
3. Continued BF at 1 year (12-15 mos)
4. Introduction of solid & semi-solid/soft foods at 6-8 mos
5. Minimum dietary diversity
Foods from >4 food groups for 6-23 mos
6. Minimum meal frequency
– BF: 6 mos > 2 or 12 mos > 3 solid/semi-solid/soft meals/24 hr
7. Minimum acceptable diet for age 6-23 mos
% 6-23 mos with both minimum DD + meal frequency
8. Consumption of Fe-rich or Fe-fortified foods

From WHO (2008, 2010)

Diet pathway: Breastfeeding: Inadequate practices (3)

Factor Indicator Description

Delayed initiation Early initiation of % born in last 24 mos

breastfeeding* who were put to breast
within 1hr of birth
Non-exclusive Exclusive % aged 0-5 mos*
breastfeeding breastfeeding under exclusively BF during
6 mos* previous day

Early cessation Continued % aged 12-15 mos who

of BF breastfeeding at 1 y* were fed any BM during
previous day

* included in IDHS
Rationale for complementary feeding indicators
• Introduction of solid & semi-solid/soft foods at 6-8 mos
– Earlier introduction displaces breast milk often with plant-based
complementary foods (CFs) of low energy & nutrient density

• Minimum dietary diversity (DD)

– Said to reflect dietary quality: +ve relationship with HAZ (independent of
SES) in children 6-23 mos

• Minimum meal frequency

– Infants have limited gastric capacity, yet are fed thin gruels with low
energy & nutrient density
– Increase in meal frequency will enhance energy & nutrient intakes

• Minimum acceptable diet (MDD+MMF)

• Consumption of Fe-rich or Fe-fortified foods
– Fe-rich foods (meat/poultry/fish) are nutrient-dense
– Prevent Fe-deficiency anemia & impaired cognition; growth; immunity etc
Inadequate complementary feeding: poor quality foods (5)
Factor Indicator Description
Poor Consumption Fe-rich/ Fe- % 6-23 mos receiving Fe-rich or
micronutrient fortified foods (FFs) (WHO) and FFs foods or vitamin A –rich
quality vitamin A-rich foods* foods in previous day

Low Dietary Minimum Dietary Diversity* % 6-23 mos receiving

Diversity (WHO) foods from > 4 food groups in
previous day
Low intake Intake of ASFs: flesh % 6-23 mos children
animal-source +liver/organ meats (Sentinel food consuming ASFs (>10 g) in
foods grp) (WHO)* previous day

Anti-nutrient Intake unrefined cereals + % 6-23 months consuming

content? legumes/d; diets with average Phy:Zn
Phytate: Zn molar ratios in diets molar ratio > 18.
Low energy Adequacy of energy intake % children with energy intake
content of meeting their daily energy
CFs requirement or estimated need
* Included in IDHS but does not include intake of fortified foods
IYCF & sentinel food indicators at 9 mos in relation to LAZ
scores of Sumedang, W Java infants at 12 mos
IYCF indicators
ᵝ 95%Confidence I p-value

Minimum dietary diversity 0.01 -0.18, 0.20 0.91

Minimum acceptable diet 0.01 -0.18, 0.20 0.91
Iron-rich/iron fortified foods 0.22 0.01, 0.44 0.04*
Sentinel food indicators
Flesh foods 0.07 -0.12,0.26 0.45
Eggs 0.02 -0.18, 0.21 0.88
Animal-source foods 0.10 -0.10, 0.30 0.32
Fortified infant foods 0.29 0.09, 0.48 0.04*

* Also significant for WAZ at 12 mos Diana et al. (2017)

Inadequate complementary feeding: inadequate practices (6)

Factor Indicator Description

Infrequent Minimum meal frequency (MMF)* (can % 6-23 mos -receiving solid,
feeding be used as proxy for dietary energy semi-solid or soft foods at least
intake) (WHO) minimum number times in
previous day
Unacceptable
diet
Minimum Acceptable Diet (MMF + DD )* % 6-23 mos receiving MAD

Feeding Introduction of solid-semi-solid or soft % 6-8 mos who received

insufficient foods (WHO) at 6-8 mos solid,semi-solid or soft foods in
amounts previous day
Thin food Consumption of “thick” foods (proxy % fed foods with “thick” and not
consistency for adequacy of energy intake) runny consistency (picture)

Inadequate See Propan: 6 Qs Children fed as % children fed as recommended

feeding during & recommended during & after an illness during & after illness
after illness
Non-responsive See Propan: 4 Qs Children fed by % children fed by caregiver
feeding caregiver responsive to child feeding responsive to child feeding (i.e.
(score +1 or -1/Q) total score of > 2)
 Two pathways lead to stunting postnatally:
inadequate diet & infections

Inadequate Poor dietary Micronutrient

diet quality deficiencies

Inadequate Macronutrient
quantity of food deficiency
Energy deficit :
Lack of nutrients
Synergism & available at
interaction Symptomatic Reduced nutrient cellular level for
infections absorption growth

Infections Subclinical infections: Direct nutrient

Environmental enteropathy losses
GROWTH
FALTERING
Inflammatory &/or Increased nutrient
Adapted from Panter- immune response requirements
Brick et al. 2009
 Inadequate complementary feeding:
Infection pathway: Food and water safety (4)

Factor Indicator Description

Contaminated watera Practice of boiling water for % of caregivers boiling
child feeding water to feed child during
previous day
Contaminated foodb PROPAN: Practice of % of caregivers washing
washing raw food, covering raw food, covering food
food prepared for child, and for child until it is eaten,
washing child’s hands washing child’s hands
Poor hygiene practices PROPAN: 6 key % caregivers adhering to
handwashing moments 6 key handwashing
identified moments
Unsafe storage and Propan Q (See a,b ) % caregivers adhering to
preparation of foods safe storage and
preparation of foods
 Infection: Clinical & sub-clinical infection (9)
Factor Indicator Description
Diarrhea > 3 loose stools/day* % < 5 y with diarrhea in past 2 wks

Helminths Use of deworming * % with deworming in past 6 mos

Environmental Elevated inflammatory % < 5 y with elevated levels of
enteropathy stool biomarkers myeloperoxidase
Acute Symptoms of ARI (cough + % < 5 y with symptoms of ARI in past 2
respiratory short rapid breathing &/or wks
infections-ARI chest-related difficulty in
breathing) *
Malaria Fever + antimalarials* %< 5 y w.antimalarials for fever in past 2 wks
Reduced Yesterday at the main meal, did % 6-23 mos eating at their main meal all the food
appetite your child eat all the food you their caregiver thought they should eat yesterday
thought he should?
Inflammation Serum CRP & AGP % w. CRP > 5 mg/L; AGP >1 g/L in past 2 wks

Fever Fever* %< 5 y ill with fever in past two wks

Partial or no Immunisations* % children being fully immunised

vaccines
 Contextural factors
Community and societal factors (6)
Political Health & Education Society Agriculture Water,
economy health care & & food sanitation &
culture systems environment

• These contextural factors are linked with underlying causes of

stunting with indirect pathways to stunting; they vary across settings
• Contextural factors can act to impede or enable progress
• Context MUST be taken into account: can increase likelihood that
programs to reduce stunting will be successful
• Multisectoral approach needed to address contextural factors
• Indonesian studies have only linked child stunting with context for:
– health and health care factors
– water, sanitation and environment factors, specifically urbanization
• Many other factors not explored so their possible association with
child stunting or linear growth in Indonesia is unknown
 Political economy (6)
Factor Indicator Description
Food prices & trade
policy
Marketing regulations
Political stability
Poverty, income, Wealth index % HHs in lowest wealth quintile
wealth subdivided by
quintiles*
Financial services
Employment & Occupation of father & % of children in HHs of
livelihoods mother* unemployed father

Only two factors (in red) explored and associated with stunting in Indonesia
These two factors: Poverty, income and wealth; Employment & livelihoods- overlap
with “home environment”, specifically with household wealth and food insecurity.
 Health and health care (5)
Factor Indicator Description
Access to Problems in accessing health % reporting problem in accessing
health care care when women are sick* health care when sick

Qualified Least qualified person % delivered by least qualified

healthcare providing assistance during person
providers delivery*
Availability of ?
supplies
Infrastructure Access to health facility for % obtaining antenatal care from a
antenatal care* health facility

Healthcare Timing of postnatal checkups % receiving checkups 2 days after

systems and & type of provider for mother giving birth in past 2 y
policies and newborn infant *

Most have been associated with risk of stunting in children 0-23 mos in Indonesia.
Health care system underlies multiple causal factors in pathway to child growth.
Most factors have been included in Indonesia as Qs from IDHS survey, except supplies.
 Education (4)
Factor Indicator Description
Access to quality Educational % with highest level school
education attainment of male & attended or completed
female head of HH* Median # yrs schooling completed
% literate
Qualified teachers
Qualified health
educators
Infrastructure:(schools
& training institutions)

• No associations explored with stunting in Indonesia despite use in IDHS of the

indictor on access to quality education of male and female head of household

• Interventions designed to improve care-giver education on optimal IYCF

associated with reductions in stunting and improved linear growth
• (See Bhutta et al., Lancet 382:452-477 (2013)
 Society and culture (4)
Factor Indicator Description
Beliefs and
norms
Social support
networks
Child
caregivers
(parental &
non-parental)
Women’s
status

• No associations with stunting in Indonesia except a study with child-care services

• Cultural considerations must be taken into account in design of BCC
• Female empowerment is an important factor underlying healthy child growth
• IDHS has some Qs on female empowerment but not directly related to child health
• Qs on women participating in decisions about their own health care; making major
HH purchases
 Agriculture and food systems (3)

Factor Indicator Description

Food
production and
processing
Availability of
MN-rich foods
Food safety &
quality
•No associations with stunting in Indonesia to date.
•Increased efforts to produce higher-quality foods (eg. biofortified foods) and
biodiversify exist but so far little evidence of their impact on child growth

•Increased efforts to develop small livestock (poultry, goats, fish, insects) that
yield micronutrient-rich foods have potential to improve IYCF but must be coupled
with behaviour change to ensure they are fed to children

Higher quality evidence linking this subelement with child growth urgently needed
 Water, sanitation, and environment (4)

Factor Indicator Description

Water and
sanitation
infrastructure
& services
Population
density
Climate
change
Urbanization

Urbanization only factor in Indonesia investigated: reduced stunting risk vs. rural areas
Fecal pollution is often pervasive & of direct import for children 0-23 mos learning to
feed themselves
Note: Access to improved water source does NOT ensure it is free from fecal
contamination. See Bain et al. (2014)
 Unique factors associated with child stunting
in Indonesia
• Low household wealth: assessed in IDHS
– may be partially represented by food security but this is not currently
measured in IDHS survey
• Paternal short stature: may be correlated with maternal
stature
– Not measured in IDHS
• Paternal and maternal smoking: measured in IDHS
• Crowded households: not measured in IDHS
• Fever: assessed in IDHS
• Partial or no receipt of vaccines: assessed in IDHS
• Boys at greater risk to stunting than girls: assessed in IDHS
 Potential indicators omitted from IDHS
• Maternal factors:
– Length at birth
• Home environment:
– Food insecurity via HFIAS
– Poor caring practices via 7 Q in Propan (PAHO/UNICEF)
– Check for fecal contamination despite improved water source
– Crowded household via # persons per sleeping room
• CF: poor quality foods:
– Poor MN quality: via 24-hr recall with 40-50 repeats per stratum
– (include intake of fortified foods
– Low energy content: via 24-hr recall with 40-50 repeats per stratum
• CF: inadequate practices:
– Thin food consistency: picture of runny vs thick CF
– Non-responsive feeding via 7 Q in Propan
Potential indicators omitted from IDHS

• CF: Food and water safety:

– Clean infant feeding & play area (to avoid exposure to fecal
contamination)
– Contaminated water & foods via PROPAN Qs
• Infection: Clinical & sub-clinical infection:
– Environmental enteropathy via elevated inflammatory stool
biomarker( myeloperoxidase)
– Inflammation via serum C-reactive protein & alpha-aid glycoprotein
Contextural factors
• Education
– Care giver education on IYCF: KAP survey
 Conclusions
• Lack of sensitivity and specificity of indicators may contribute to
inconsistencies between their associations with child growth and
stunting across studies
• Variability across settings may limit use of universal indicators
• Misclassification may occur when 24-hr recall period used for IYCF
• When % of respondents achieving indicator is small, power may be
insufficient to detect growth differences
• Differences in model covariates used (e.g., wealth; SES; education)
when assessing associations) (e.g. multicollinearity) may impact
outcomes
• Cross-sectional data often used so causal sequencing of any
relationships cannot be determined
• Minor differences in constructing indicators may limit cross-
country comparisons
 Key References
Bain R, Cronk R, Wright J et al. (2014). Fecal contamination of drinking-
water in low- and middle-income countries: A systematic review and meta-
analysis. PLoS Med 11(5):e1001644.doi:10.1371/journal.pmed1001644

Beal T, Tumilowicz A, Sutrisna A, Izwardy D (2017). A review of child

stunting determinants in Indonesia. Matern Child Nutr 14:e12617.

Haselow N, Stormer A, Pries A (2016). Evidence-based evolution of an

integrated nutrition-focused agriculture approach to address the underlying
determinants of stunting. Matern Child Nutr 12(Suppl10:155-168.

Stewart C, Iannotti L, Dewey KG et al. (2013). Contextualising

complementary feeding in a broader framework for stunting prevention.
Matern Child Nutr 9(Suppl.2):27-45.

• *
Thank you for your attention!
 Considerations when selecting indicators for
national surveys
• Term “indicator” relates to use or application of indices and are often
constructed from them
• An indicator is supposed to indicate some underlying reality that is
difficult /impossible to measure
• Must be reproducible with substantial agreement among observers
• For population-level assessment a balance is needed between
sensitivity and specificity to minimize misclassification by the indicator
– Sensitivity : ability to identify those who are genuinely malnourished
– Specificity: ability to identify those who are genuinely well-nourished
• Choice of indicators depends on objectives of their use:
– Assessment (national surveys); targeting; monitoring and evaluation
• For assessment to make national/subnational comparisons: sensitivity
more important than specificity as there are no risks to false positives
(low specificity) and false negatives are avoided
• Lack of sensitivity and specificity of indicators may contribute to
inconsistencies observed between indicators and stunting
 Baby-WASH interventions designed to interrupt feco-
oral transmission in first two years & reduce stunting

Reduction in risk of diarrhea Intervention Objective

Household sanitary Reduce fecal load
• Hand washing w. soap: 48%
facility in living
• improved water quality: 17% environment

• Improved excreta disposal: 36% Hand washing facility Reduce hand fecal
w. soap transmission

Exclusive breast Protect infant from

feeding (0-6 mos) harmful liquids etc

Safe water & storage Improve drinking

at point of use water quality*

Clean infant feeding& Avoid child fecal

play area ingestion
Access to improved water source does Hygienic food handling
NOT ensure it is free from fecal & preparation
contamination. See Bain et al. (2014)
From Humphrey et al.