The global market for fortified
complementary infant foods:
Evidence from 108 locally-produced,
commercially marketed products in 22 countries
William A. Masters
with Winnie Bell and Marc Nene
Friedman School of Nutrition Science and Policy, Tufts University
http://sites.tufts.edu/willmasters | http://www.nutritioninnovationlab.org
29th International Conference of Agricultural Economists, Milan
Symposium on Nutritional Gains for the Poor from Agri-food Value Chains
10 August 2015
Much of the world’s growth faltering is
experienced by infants during complementary feeding
from 6 to 24 months of age
Mean z scores relative to WHO standards across 54 DHS/MCIS studies, by age (1-59 mo.)
What can explain the sudden and
severe shortfalls during this period?
Source: CG Victora, M de Onis, PC Hallal, M Blössner and R Shrimpton, “Worldwide timing of growth
faltering: revisiting implications for interventions.” Pediatrics, 125(3, Mar. 2010):e473-80.
Many factors could cause this pattern
of onset and duration
• For example:
– Exposure to water-borne pathogens, air pollution etc.
– Expression of gestational deficits
• ...and also inadequate total nutrient intake:
Mean intake as a percentage of WHO recommended needs, by age
Energy
Protein
Calcium
Iron
Vitamin A
Vitamin B1
Vitamin B2
Niacin
(kcal)
(g)
(mg)
(mg)
(IU)
(mg)
(mg)
(mg)
6–8 months
84%
89%
66%
33%
535%
59%
48%
53%
9–11 months
90%
99%
69%
44%
663%
67%
50%
41%
12–18 months
91%
88%
78%
75%
443%
60%
35%
44%
Note: Data shown are mean values for a sample of 400 children in Eastern Ghana.
Source: C.A. Nti and A. Lartey (2007). “Young child feeding practices and child nutritional status in rural Ghana.”
International Journal of Consumer Studies, 31: 326-332.
Why might sufficient nutrient intake
be difficult to achieve?
• Total quantity is small and inexpensive relative to family budget
...but nutrient density and digestibility must be higher than family diet
• Complementary foods have high-cost ingredients and processing
– Starchy staple must be fortified with protein, fats and micronutrients
– Processing at home using traditional methods is labor intensive
– Availability from artisanal or industrial manufacturers is limited
• Since Incaparina in the 1960s, donors have funded startups to produce and
sell high-quality complementary foods at lowest possible cost
– Many different recipes and production methods are possible
– Locally-adapted products are readily acceptable, have proven efficacy, and cost
much less than either home production or imports from industrialized countries
• So what products do we now see in places with widespread malnutrition?
Low-income country markets are
still dominated by Nestle’s Cerelac
Accra, Ghana (2010)
There are many alternatives:
Generics, e.g.
“weanimix” in Ghana:
Many small-scale local start-ups,
e.g. these from around Africa:
Cameroon
And various
multinationals,
typically sold in
supermarkets
and pharmacies:
This project asks:
What’s for sale, and is it any good?
• A global catalog of marketed complementary foods:
– Step 1: Invite collaborators to identify products on local markets in 2014
• Criterion: marketed as complement to breastmilk, for >6 months of age
• Typical product: a fortified composite flour
– Step 2: Pay them to send us a random sample for nutrient testing
• Goal: at least 20 countries, averaging 5 products from each country
• Test for protein, fats, calories, iron, zinc, and phosphorus (for phytates)
– Step 3: Test predictions from the economics of product quality
• Hypothesis: Consumers cannot detect nutrients, so composition will vary widely
• Remedy would be third-party quality assurance
• If products can be good but often aren’t, quality assurance can help
– both plant inspections and product sampling with nutrient testing
– 3rd-party label, e.g. “INQAP—OK for babies over 6 months”
– local advertising to establish this third-party brand
For now:
Results from 109 products in 22 countries
Energy density varies widely within and across countries
an
rk
a
in
a
Fa
C
am so
er
oo
n
C
C
h
ot
in
a
e
d'
Iv
o
D
ire
R
C
on
g
Et o
hi
op
ia
G
ha
na
H
ai
In
d o ti
ne
si
a
Ke
M
n
ad
y
ag a
as
ca
r
M
al
aw
i
M
M
au ali
rit
an
ia
M
or
oc
co
N
ep
al
R
w
an
d
Se a
So neg
al
ut
h
Af
ric
a
U
ga
nd
a
ts
w
Bu
Bo
Be
ni
n
380
400
420
440
460
Energy density by country of purchase
relative to the WFP SuperCereal+ benchmark
For now:
Results from 109 products in 22 countries
Macronutrients typically fall short of international standards
Protein content by country of purchase
relative to the WFP SuperCereal+ benchmark
15
10
5
0
0
10
20
30
Fat content (g/100g dry matter)
40
20
Fat content by country of purchase
relative to the WFP SuperCereal+ benchmark
Be
n
ni
o
a
n
a
na
na oire ngo
pi
an Fas roo
hi
io
ha
o
w
Iv
e
C
th
C
G
ts ina
d'
m
E
o
a
R
e
B
rk
C
D
ot
C
Bu
i
ti
ia
ya car law
ai
es en
s
H
a
K
on
ga
M
d
a
In
ad
M
M
i
al
M
au
a
rit
ni
a
M
l
l
a
a
a
co
pa nd
ga fric
nd
oc Ne
a
ne
A
ga
w
or
U
R
Se uth
So
Be
n
ni
o
a
n
a
na
na oire ngo
pi
an Fas roo
hi
io
ha
o
w
Iv
e
C
th
C
G
ts ina
d'
m
E
o
a
R
e
B
rk
C
D
ot
C
Bu
ti
ai
H
I
o
nd
a
a
ar awi
si
ny
sc
al
ne K e
ga
M
a
ad
M
M
i
al
M
au
a
rit
ni
a
M
l
l
a
a
a
co
pa nd
ga fric
nd
oc Ne
a
ne
A
ga
w
or
U
R
Se uth
So
For now:
Results from 109 products in 22 countries
Zinc and Iron also often fall short of international standards
Iron content by country of purchase
relative to the WFP SuperCereal+ benchmark
a
Fa
s
am o
er
oo
n
C
C
h
ot
in
a
e
d'
Iv
oi
D
r
e
R
C
on
go
Et
hi
op
ia
G
ha
na
H
ai
In
d o ti
ne
si
a
Ke
M
ad nya
ag
as
ca
r
M
al
aw
i
M Ma
li
au
rit
an
ia
M
or
oc
co
N
ep
al
R
w
an
da
Se
So neg
al
ut
h
Af
ric
a
U
ga
nd
a
rk
in
C
a
an
Bu
Be
Bo
ts
w
ni
n
0
40
20
0
5
10
15
Iron content (mg/100g dry matter)
20
60
Zinc content by country of purchase
relative to the WFP SuperCereal+ benchmark
Be
n
ni
o
a
n
a
na
na oire ngo
pi
an Fas roo
hi
io
ha
o
w
Iv
e
C
th
C
G
ts ina
d'
m
E
o
a
R
e
B
rk
C
D
ot
C
Bu
ti
ai
H
I
o
nd
a
a
ar awi
si
ny
sc
al
ne K e
ga
M
a
ad
M
M
i
al
M
au
a
rit
ni
a
M
l
l
a
a
a
co
pa nd
ga fric
nd
oc Ne
a
ne
A
ga
w
or
U
R
Se uth
So
For now:
Results from 109 products in 22 countries
0
20
40
60
80
100
Macronutrients by level of energy density
Energy
mainlybenchmark
on fat content
relativedensity
to WFPdepends
SuperCereal+
levels
380
400
420
440
Energy density (kcal/100g dry matter)
460
Fat content (g/100g dry matter)
Protein content (g/100g dry matter)
Carbohydrate content (g/100g dry matter)
For now:
Results from 109 products in 22 countries
0
5
10
15
20
Zinc content by level of iron content
relative
to WFP
SuperCereal+
benchmark levels
For zinc and iron,
both
typically
fall below international
standards
0
20
40
Iron content (mg/100g dry matter)
60
Packaging lists the product’s energy content
for 75 products in 18 countries
Actual energy content often exceeds labelled energy, and varies less
300
350
400
450
500
Energy density by labeled value
300.00
350.00
400.00
Labeled energy density (kcal/100g)
450.00
Note: the WFP’s SuperCereal Plus has minimum energy content of 410 kcal/100g
Packaging lists protein for 77 products and
fat for 72 products, both in 19 countries
Both protein and fat typically fall short at the higher labeled values
Fat content by labeled value
15
10
5
0
0
10
20
30
Actual fat content (g/100g dry matter)
40
20
Protein content by labeled value
5.00
10.00
15.00
20.00
25.00
Labeled protein content (g/100g dry matter)
30.00
0.00
5.00
10.00
15.00
Labeled fat content (g/100g dry matter)
Note: the WFP’s SuperCereal Plus has minimum fat content of 9 g/100g,
and minimum protein content of 16 g/100g.
20.00
Packaging lists iron for 53 products in 17 countries
Iron content is widely scattered around labeled values
0
20
40
60
Iron content by labeled value
0.00
10.00
20.00
30.00
40.00
Labeled iron content (mg/100g dry matter)
50.00
Note: the WFP’s SuperCereal Plus has minimum iron content of 12.5 g/100g.
Conclusions:
These products can have adequate nutrients
but they usually don’t
To remedy this market failure, would need quality assurance
Its effectiveness could be measured by an RCT:
1) Establish an Infant Nutrition Quality Assurance Project (INQAP)
2) Recruit manufacturers to participate, and issue time limited
INQAP-OK stickers
3) Roll out billboards and demonstrations at randomly-chosen
market locations
4) Use household surveys and growth monitoring to track food
purchases & infant bodyweights
5) Use child’s age at the time of market advertisements to identify
causal effect of certification on growth
Acknowledgements
For funding: USAID, for an International Food Policy Research Institute
Linkage Grant and the Feed the Future Nutrition Innovation Lab.
For samples: Ambroise Agbota, Jean-Paul Anoh, Mahaman Bamba, Aaron
Cheng, Gnangbo Christian, Amelia Darrouzet-Nardi, Claude Emile, Issa
Fadiga, Matt Hazel, Dorothy Nzembi Kimanthi, Yolande Kouame,
Emmanuel Kouame, Rachid Laajaj, Perrine Loock, Janeen Madan, Will
Masters, Menno Mulder-Sibanda, Fatouman Ouattara, Galase
Ramolefhe, Marianne Santoso and Robin Shretha.
For testing: Midwest Labs, Omaha.
Thank you!