Changes in nutrition outcomes in Ethiopia, 2000-2016

Kalle HirvonenInternational Food Policy Research Institute

The Future of Ethiopia’s Agriculture: Towards a Resilient System to End

Hunger and Undernutrition

Addis Ababa HiltonDecember 15, 2017Addis Ababa

1

2

Chronic undernutrition1. Typically measured using Height-for-Age Z-score (HAZ)

• For given age and sex: HAZ measures the distance in height to the median child of a healthy and well-nourished population

2. A child is considered stunted if this distance is -2 standard deviations or more: HAZ ≤ -2.

3. In a healthy and well-nourished population this is rare: about 2% have HAZ ≤-2.

4. Higher stunting prevalence rates imply chronic under-nutrition.

5. According to WHO (1995) stunting rates of

o > 40 % = very high malnutrition severity

o 30-39 % = high malnutrition severity

o 20-29 % = medium malnutrition severity

o <20 = low severity malnutrition severity

3

Why should we care?1. Stunting is a marker for a number of development outcomes:

i. Sufficient nutrition is a critical input for brain development (fastest in the first years of life)

ii. Considerable body of research shows that stunted children attain less schooling and score poorly on tests measuring cognitive ability

iii. All this is then linked to economic productivity, welfare and poverty 20 years later when these children are adults.

GoE (2015) estimates that child undernutrition reduces GDP by 55.5 billion ETB or 16% of GDP.

4

Ethiopia since 2000: one of the fastest reductions in stunting rate in the world

Source: Demographic and Health Survey (DHS) – Ethiopia 2000, 2005, 2011, 2016

WHO: very high malnutrition

severity

WHO: high malnutrition

severityWHO: very high malnutrition

severity

WHO: very high malnutrition

severity

5

Stunting-age relationship: Global evidence

Born small

Intensive breastfeeding

Introduction of solid foods

Source: DHS data for 65 middle and low income countries; more than 1 million observations.

Adapted from Headey, Hoddinott & Menon (2017): “The age dynamics of growth faltering among pre-school children”

6

Stunting-age relationship: Ethiopia

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

7

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

Hypotheses

Hypothesis 1: Marginal improvements in birth sizes

8

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

Hypotheses

Hypothesis 1: Marginal improvements in birth sizes

Hypothesis 2: ‘More Intensive’ breastfeeding

9

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

Hypotheses

Hypothesis 1: Marginal improvements in birth sizes

Hypothesis 2: ‘More Intensive’ breastfeeding

Hypothesis 3: Little improvement in CF practices

10

Marginal improvements in birth sizes?

1. Difficult to verify: few children are born at health facilities

2. Self-reported birth sizes by the mothers suggest little improvement but these data are questionable

3. But we can get clues from women’s BMI (who are in reproductive age)

• There’s considerable evidence that (pre-pregnancy) BMI is positively associated with infant birth weight

11

Marginal improvements in birth sizes ✔

Women’s (15-49 y) Body-Mass Index (BMI)

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

Rural Urban

12

‘More Intensive’ breastfeeding ✔

2000 2016

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

13

Little improvement in CF practices?

1. Minimum meal frequency (%): Proportion of breastfed and non-breastfed children 6-23.9 months of age who receive solid, semi-solid, or soft foods or milk feeds the minimum number of times or more.

2. Minimum dietary diversity (%): Proportion of children 6-23.9 months of age who receive foods from 4 or more food groups (out of 7).

o Problem: only available in 2011 and 2016

14

Larger share of the younger children achieve recommended meal frequency

Source: Demographic and Health Survey (DHS) – Ethiopia 2000 & 2016

% of children achieving recommended meal frequency

15

Few children consume a sufficiently diverse diet

% of children achieving recommended dietary diversity

Source: Demographic and Health Survey (DHS) – Ethiopia 2011 & 2016

16

Summary (so far)1. Considerable reduction in stunting rates since 2000.

o One of the fastest reductions in the world

2. Little improvement in avg birth sizes; 20% are born stuntedo Calls for focus on maternal health & nutrition

3. In 2016, growth faltering occurs later – likely due to improvements in ‘intensive’ breastfeeding practices. o Median duration of exclusive breastfeeding increased

from 2.5 to 3.6

4. However, growth still ‘collapses’ around the time when complementary foods are introduced (~ 6 mo).o Dietary diversity is alarmingly poor

17

How is this linked to agriculture?

1. Complementary feeding practices are shaped by two factors:

i. Caregiver’s knowledge (Health Extension program, Alive & Thrive)

ii. Access to food

a. Physical access (e.g. Abay & Hirvonen 2017; Stifel & Minten 2017)

b. Financial access (affordability & prices)

2. All these factors are important AND complementary (e.g. Hirvonen et al 2017).

3. Global evidence shows that what young children are consuming is strongly influenced by prices (Headey, Hirvonen & Hoddinott 2017).

• e.g. doubling the price of eggs is associated with a 15-percentage point drop in the probability of children 6-23 mo consuming them.

Source: Bachewe, Hirvonen, Minten & Yimer (2017) based on CSA retail price data

% change in real prices, 2007-2016, by food group

(real) Prices of nutritious foods have risen fast

19

This is how it’s linked to agriculture

• Prices of nutritious foods (fruits, vegetables & animalsource foods) are directly shaped by agriculture policy– production, marketing & trade.

Agricultural policy has a major role to play in futurereduction in stunting rates in Ethiopia.