1. EE: Going Beyond Nutrition to Understand
Child Growth and Development
Laura Smith
Rebecca Stoltzfus,Francis Ngure, Brie
Reid, Gretel Pelto, Mduduzi Mbuya,
Andrew Prendergast, Jean Humphrey
Division of Nutritional Sciences

2. (Victora et al. 2010)
The “Window of Opportunity” for Improving Nutrition is
very small… Pre-pregnancy until 18-24 months of age

3. What is causing all this stunting?

4. Cause #1: Malnourished
Mother
• Malnourished mothers give birth to babies that are smaller
and shorter than normal
• 50% of Guatemalan babies are born stunted (Ruel 2001)
– Prevalence of stunting at birth not well documented
– Good length data on newborns is very hard to get!
Estimates of 30-50% of stunting is due to intra-uterine factors.
Effective macronutrient interventions for pregnant women are not
well established.

5. Cause #2: Poor Diet
• Systematic review of the efficacy and effectiveness of
complementary feeding interventions in developing countries
– Dewey & Adu-Afarwuah, 2008
– 42 studies/programs, most published 1996-2006
• Children who received interventions gained:
– 0.0 – 0.76 Z scores weight-for-age
– 0.0 – 0.64 Z scores length-for-age
The best studies caused a 0.7 Z score improvement. BUT:
the average growth deficit of African and Asian children is -2.0 Z
At best, diet solved 1/3 of the problem.

6. Cause #3: Diarrhea
• Between 6-18 months of age, children in developing countries
have around 9 episodes of diarrhea.
• Many authors reported that diarrhea accounts for 10-80% of
growth faltering
• But others contend that children grow at “catch-up rates”
between episodes, and thus recover these deficits
The Lancet Nutrition Series (2008) concluded that by
implementing sanitation and hygiene interventions with 99%
coverage, child malnutrition would be reduced by only 2.4%

7. However:
Evidence exists that the effect of WASH
interventions on linear growth is
independent of its effect on diarrhea.
In several studies, WASH had a bigger effect
on growth than it did on diarrhea

8. Peru:
(Checkley, et al)
• Children assessed for diarrhea and growth from birth
to 2 years
• Household sanitation and water assessed
• What predicted height deficit at 2 years?
16% explained by how much diarrhea the
child had experienced
40% explained by the level of sanitation and
water in child’s household

9. Rural Ethiopia: HH Hygiene Index was the
variable most strongly associated with stunting
Alive and Thrive baseline data; F Ngure (2013, in prep)

10. Cause #4:
The Environmental
Enteropathy Hypothesis
• A subclinical condition of the small intestine, called
environmental enteropathy (EE)
• Characterized by:
– Flattening of the villi of the gut, reducing its surface area
– Thickening of the surface through which nutrients must be absorbed
– Increased permeability to large molecules and cells (microbes)
• Likely causes:
– Too many microbes in the gut
– Effects of toxins on the gut
Decreased nutrient absorption + Infiltration of microbes

11. Microbial translocation
Microbial products cross
into blood stream
The lining of the gut
is only one cell thick
If the gut is injured and
becomes permeable, gaps
open up between cells
Chronic immune
activation
Diverts nutrients from
growth to infection-
fighting

12. EE is a major cause of post-natal stunting,
anemia and immune competence
EE can be prevented or reduced by
preventing infants and young children from
ingesting human and animal feces through a
package of interventions that improve
sanitation and hygiene.
Environmental Enteropathy and
Stunting Hypothesis:

13. Chronic immune activation
↑ pro-inflammatory
cytokines
Immunosenescence
(premature aging) of adaptive
cell-mediated immune system
↑Hepcidin ↓Growth Factor
(IGF-1)
Anemia Stunting
Impaired response to
vaccines and infections

14. HAZ changes over first 18 months in stunted and non-
stunted infants
Birth 6wks 3mo 6mo 12mo 15mo9mo 18mo

15. IGF-1 and IGFBP3 were lower in stunted
infants, beginning at 6 wk
0 3 6 9 12 15 18
0
20
40
60
80
Age (months)
MedianIGF-1(ng/mL)
0 3 6 9 12 15 18
0
500
1000
1500
IGFBP3
Months
P values for all time points 6 w to 12 mo,
p<0.001
Values for healthy European children range
from 54-170 ng/mL
P values for all time points 6 w to 18 mo,
p<0.001
stunted stunted

16. Development of the WASH Intervention
(Efficacy = “Proof of concept”)
WASH Goal:
All infants never ingest any faeces between birth
to 18 months

17. Conventional WASH formative research
(2008-2009)
Sanitation HIGHLY
valued don’t have a
latrine because lack
money; a Blair VIP is a
source of status
• Infant stools less offensive than adults’
• Handwashing is seldom with soap
• Frequently feed cold leftover food

18. • 6 hour observation of 20 babies, recorded what and how
often went in the mouth and if visibly dirty
• Returned and collected samples of most frequent and
dirtiest things mouthed for micro analysis
Baby Observation Study (2011)

19. Findings
Most frequent:
38 time in 6 hours
75% visiby dirty
Dirtiest
Soil (3 ate avg 11 bites)
chicken faeces, stones

20. If allowed, toddlers consume
poultry feces
Peruvian shantytown families:
– Households who owned free-range poultry:
• Average ingestion of poultry feces by toddlers per
12-hour observation period was 3.9 times
– Marquis GM et al., Am J Public Health 1990
Rural Zimbabwe:
– Not selected for poultry ownership:
• 3 of 7 toddlers directly ate chicken feces during a
6-hour observation period.
– Ngure F et al., submitted, 2012

21. % HH with E
coli + sample
E coil/
Per gram
Average E Coli
Per Day
Infant Food 0% 0 0
Drinking Water
54% 2 800
Soil in
laundry area
60-80% 70 1,400
Chicken feces
100% 10,000,000 10,000,000
Clearly, kids must stop eating dirt and chicken poop!

22. 24
Babies are fed on
*Ground in the yard
(60-80% E coli+) or
*Kitchen floor
(81% E coli+)

23. Source: World Bank, accessed 6.23.11
http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTWAT/EXTTOPSANHYG/

24. Laundry
Water
Nappy Handling
Protective Play Space
for babies!

25. A new way of thinking about WASH in the first 1000 days
• Protective play space, to protect developing child from
contaminated soil and animal feces (especially chickens)
• Infant handwashing with soap, when outside of protective
play space.
• Caregiver handwashing with soap after fecal contact and
before preparing/serving food
• Safe disposal of feces—especially of children
• Water treatment
• Avoid feeding leftovers, or reheat

26. Control
Infant Feeding:
Education +
Nutributter
WASH:
Integrated Water,
Hygiene & Sanitation
WASH
+
Infant Feeding
2x2 Factorial Design

27. Objective
To measure the independent and combined effects of
WASH and infant nutrition on stunting and anemia
among children from birth to 18 months of age
And, on a sample of 1600 infants, measure the
hypothesized “causal pathway” of EE
1000 HIV- mothers
600 HIV+ mothers

28. Protective play space
Goal: Culturally-acceptable, economical product that could be
locally fabricated, which protects babies and toddlers from
ingesting soil while allowing physical and cognitive
development
Engaged a marketing expert (Malinda Sanna, Spark) for
consumer research
Design process led by team from Cornell’s Department of
Design and Environmental Analysis and Human Development

29. Early
Childhood
Development
Nutrition
Stimulation
Social
Protection
Hygiene
Programmatic approaches for nutrition,
stimulation and social protection are
well developed.
UNICEF 2006 Programming Experiences in Early
Childhood Development
Lancet 2011 Child Development Series
Hygiene for babies
(Baby WASH)
needs to be further
developed and tested
Environmental Protection?

30. SHINE Investigators:
MoHCW
Goldberg Mangwadu – Director of Environmental Health, MoHCW (Co-PI)
Cynthia Chasokela – Director of Nursing
Zvitambo
Jean Humphrey (Co-PI) Mduduzi Mbuya,
Naume Tavengwa, Kuda Mutasa, Robert Ntozini
Johns Hopkins Bloomberg School of Public Health
Larry Moulton, Jim Tielsch (J Humphrey)
Cornell
Rebecca Stoltzfus
University of London
Andrew Prendergast
University of British Columbia
Amee Manges
Funding
Gates, DFID, CIDA,
NIH,
Wellcome Trust,
UNICEF

32. Zimbabwe SHINES
Zimbabwe Sanitation Hygiene Infant Nutrition
Efficacy Study
Observation: babies with healthier guts and less inflammation grow better.
Biological hypothesis: babies who are protected from fecal ingestion will
grow better.
Randomized trial hypotheses: babies whose households receive a
comprehensive Water Sanitation & Hygiene (WASH) intervention will grow
better.
WASH
Intervention
Reduced
fecal
ingestion
Better
Growth
San & Hyg
Behaviors
Healthier Gut
& Less
Inflammation
WASH + Nutrition will have more benefit than either alone.

33. Community-based
Two entire rural districts
total popn: 180,000K

34. Standard Care
1. Revived VHW network
2. Strengthened PMTCT care
3. EBF Promotion
4. Latrine at end of study
Infant Nutrition
1. Standard care
2. Promote optimal use of local foods
for complementary feeding
3. 20 g Nutributter daily provided for
infants (6-18 mo)
Sanitation/Hygiene
1. Standard care
2. VIP latrine
3. 2 Tippy Taps and soap
4. Water Guard
5. Protected play area
6. Promote safe feces disposal hand
washing, water Rx, hygienic infant
feeding
Infant Nutrition &
Sanitation/Hygiene
1. Standard care
2. All Nutrition interventions
3. All Sanitation/Hygiene interventions

35. Implementation
• Enroll 4,800 pregnant women at 10-12
weeks gestation and follow them until their
babies are 18 months old.
• Interventions delivered by 360 Village
Health Workers on bicycle and Oxfam
• Outcomes measured by 34 Research
Nurses on motorbike

36. Outcomes
• Among all 4800 infants assess growth,
anemia, intervention uptake, relevant
behaviors
• Among 1600 infants, assess causal
pathway (indicators of EE) at 3, 6, 12,
18 months

37. The context: IO
study
Laundry area
Bare soil and
animal waste

38. Micro team

39. Results: IO phase summary
• Barely no HW for infants:
21 times (in 13/23 HH) in 130 hours
HWWS 6/21: all as part of a bath.
• Adult caregivers’ HWWS after contact with stool-
7.5 % (4 HWWS/53 opportunities)
• 9/23 (39 %) infants took:
3 infants- active soil ingestion
2 took chicken feces
4 took stones from the dirt

40. Results: Micro phase summary
A one year old ingesting chicken feces, soil and
400 ml of contaminated water:
On average 10 million E. coli counts/g of
Chicken feces
Frequency of E. coli: 22/22 HH (100 %)
69 E. coli counts/g of wet shaded area soil:
Frequency: 18/22 HH (82 %)
800 E. coli counts from water.
Frequency: 12/ 22 HH (55 %)
 Other bacteria population

41. Key messages
• Chicken feces and soil ingestion are
predominant pathways
• Exploratory soil ingestion and geophagia could
be more prevalent in rural Zimbabwe than
observed for 23 HH - 2 FGDs indicated this.
• Water contamination is significant.
• Infants HW barely practiced.

42. Key messages
• Caregiver’s HWWS after fecal contact is not
common.
• WASH interventions need to focus more on
protecting infants from eating earth and chicken
feces
• The idea of a washable mat or playing pen as
a WASH intervention input

44. IYCF Intervention
Promote optimal use of locally available foods,
responsive feeding, nutrient density, feeding during
illness
Provide 20 g/d Nutributter 6-
18 mo

45. Prior to Trial Launch

46. Village Health Worker Revitalization
Full complement recruited, trained (8 months) Provide
tools (bike, kit, uniforms)

47. Campaign to promote Exclusive
breastfeeding for all to 6 months
1. Health worker
training
2. Social marketing

48. Implement WHO 2010 Prevention of Mother to Child
Transmission of HIV and Infant Feeding Guidance

49. EE is caused by environmental exposure
• Fetuses and newborns do not have it
• People acquire it moving into developing country
and lose it moving out
• Widely believed:
Result of high exposure to fecal contamination in
an environment of poor sanitation and hygiene

50. • EE is a major cause of child stunting
• EE can be prevented or reduced by
preventing infants and young children from
ingesting human and animal feces through a
package of interventions that improve
sanitation and hygiene.
Environmental Enteropathy and
Stunting Hypothesis:

51. Fecal contaminationEnvironment
Altered bacterial load, composition
and/or timing of colonisation
Intestinal inflammation and reduced
surface area
Increased intestinal permeability
Microbial translocation
Microbiome
Intestinal
pathology
Microbial translocation
StuntingAnemia
Activates innate
immune system
 proinflammatory
cytokines
Immune
activation
Activates adaptive
immune system
Immunosenescence
Impaired responses
to vaccinations and
infections
 IGF-1 Hepcidin Erythropoiesis
 Iron absorption and
mobilization

52. IGF-1
• 70 amino acid polypeptide
• Mostly produced by liver
• Mediates the effects of growth hormone
• Important in fetal and postnatal growth
• Endocrine function (paracrine, autocrine)
• Highly protein bound
– 6 binding proteins
– Mostly IGFBP3
– Acid labile subunit

53. Inflammatory markers were higher in stunted infants
P values for all time points 6 w to 12 mo,
p<0.001
0 3 6 9 12 15 18
5.0
5.5
6.0
6.5
7.0
7.5
8.0
LnCRP
Months
0 3 6 9 12 15 18
0
5
10
15
20
Age
IL-6
0 3 6 9 12 15 18
11
12
13
14
15
Months
LnAGP
0 3 6 9 12 15 18
0
1×1006
2×1006
Months
sCD14
P values from
multivariate models
that included
maternal education,
MUAC, sex, birth
weight, and mixed
feeding.
P=0.007 P=0.064
NSP=0.023
stunted

54. Stunted Non stunted
0
100
200
300
400
500
MaternalIGF-1(ng/mL)
Mean 95.9 Mean 114.3
P=0.02
0 1000 2000 3000 4000 5000
0
5
10
15
20
InfantIGF-1atbirth(ng/mL)
Birth weight (grams)
R=0.50
P<0.001
0 200000 400000 600000 800000
0
10
20
30
40
50
InfantIGF-1atbirth(ng/mL)
Infant AGP (MU/L)
R=-0.39
P<0.001
105 106
104
105
106
Maternal AGP (MU/L)
InfantAGP(MU/L)
R=0.52
P<0.001
Maternal and infant values were correlated at birth

55. Intestinal Fatty Acid Binding Protein (I-FABP)
• Very small (15kD) cytoplasmic protein
• Found in enterocytes
• Involved in intracellular transport of LCFA
• Rapidly released into blood after cellular damage
• Predominantly expressed in cells at tips of villi in
small intestine
• Elevated in celiac disease compared to healthy
contols
Derikx JP, J Clin Gastroenterol 2009

56. High I-FABP levels in stunted and non-
stunted Zimbabwean infants
Non stunted Stunted
0
2000
4000
6000
8000
I-FABP(pg/mL)
Non stunted Stunted
0
2000
4000
6000
8000
I-FABP(pg/mL)
6 months 12 months
Median 859 Median 978 Median 1148Median 1070
Healthy controls 172.7 pg/mL (±20.2) – mean age 22 yrs (range 1-61 yrs)
Coeliac disease Median 784.7 pg/ml (±145.5)
Advanced HIV (adults) 174.4 pg/mL
P=0.36 P=0.13

57. I-FABP was higher in stunted infants
0 3 6 9 12 15 18
0
1×1006
2×1006
Months
I-FABP
stunted
P value from
multivariate model
that included
maternal education,
MUAC, sex, birth
weight, and mixed
feeding.
P=0.030

58. Summary
• Growth hormone axis is perturbed very early in life in
apparently healthy Zimbabwean infants with poor
linear growth.
• Small intestinal damage and low-grade inflammation
are evident post-natally and associated with stunting
by 18 mo of age
• Diarrhea (measured by clinic visits) was not
associated with stunting.
• Circulating levels of pro-inflammatory mediators and
IGF-1 in mother-infant dyad at birth are associated
with in utero growth.

59. Zimbabwe SHINES
Zimbabwe Sanitation Hygiene Infant Nutrition
Efficacy Study
Observation: babies with healthier guts and less inflammation grow better.
Biological hypothesis: babies who are protected from fecal ingestion will
grow better.
Randomized trial hypothesis: babies whose households receive a
comprehensive Water Sanitation & Hygiene (WASH) intervention will grow
better.
WASH
Intervention
Reduced
fecal
ingestion
Better
Growth
San & Hyg
Behaviors
Healthier Gut
& Less
Inflammation