The Sri Lankan Problem

1

Comparative performance in South Asia

0

5

10

15

20

25

30

35

40

45

50

Stunting (%)

0

10

20

30

40

50

60

70

80

IMR

2

Stagnation and disparities in Stunting Sri Lanka (recent years)

31.1

26.4

17.6 17.5

0

5

10

15

20

25

30

35

1985 1990 1995 2000 2005 2010

Stunting (DHS)

Underweight (DHS)

3

IHP PO

LICY BRIEF ISSU

E 02 Novem

ber 2015

6

<14.014.1 - 16.016.1 - 20.020.1 - 25.0>25

<9.59.6 - 12.512.6 - 15.715.8 - 19.9>20.0

<18.818.9 - 22.422.5 - 26.726.8 - 32.8>32.9

Low birth weight incidence (%) 2010 Stunting prevalence in children aged 6–59 months (%) 2012

Underweight prevalence in children aged 6–59 months (%) 2007-2012

Source: Birth registration data (Registrar General’s Department). Source: National Nutrition and Micronutrient Survey 2012. Source: DHS 2006/2007, National Nutrition and Micronutrient Survey 2012.

Low birth weight by ethnicity (%) 2007

Source: DHS 2006/2007.

5

0

10

15

20

25

30

35

40

SinhalaSri LankaTamil

Estate Tamil Muslim

Stunting by wealth decile (%) 2007

Source: DHS 2006/2007.

35

30

25

20

15

10

5

0Poorest Q2 Q3 Q4 Richest

PoorestQ2Q3Q4Richest

Trends in underweight by quintile (%) 1987-2007

Source: DHS surveys 1987 - 2007.

45

40

35

30

25

20

15

10

5

01985 1990 1995 2000 2005 2010

Figure 2: The distribution and recent trends in child undernutrition in Sri Lanka

Income inequalities in Sri Lanka in undernutrition –amongst the worst in world

22 21

14 16

12

28

22

16 15

9

27 26

18 17

9

-

5

10

15

20

25

30

Poorest Q2 Q3 Q4 Richest

Perc

enta

ge (%

)

Low Birth Weight

Stunting

Underweight

4

Bangladesh

India

Sri Lanka

Maldives

Nepal

Pakistan

BEST

BESTWORST

WORST

-2

-1

0

1

2

3

Rela

tive

perfo

rman

ce in

redu

cing

child

stu

ntin

g

-2 -1 0 1 2 3Relative performance in reducing infant mortality

Comparative performance in undernutrition vs. IMR (controlling for income)

5

The Sri Lankan Paradox

•   Poor performance in child undernutrition despite otherwise excellent health outcomes

•   Large inequality in undernutrition outcomes despite generally equitable access to healthcare and comparatively equitable health outcomes

•   Slow improvement in nutrition outcomes despite good economic growth

•   No obvious impact from large scale food supplementation program (Thriposha) since 1970s

•   But part of wider South Asian enigma –   Discrepancy in nutrition performance and health outcomes/

economic growth shared with rest of South Asia

6

What do the data tell us?

7

8

Series

2 www.thelancet.com Published online June 6, 2013 http://dx.doi.org/10.1016/S0140-6736(13)60937-X

Health Organization, Department of Nutrition for

Health and Development, Geneva, Switzerland

(M de Onis MD); Imperial College of London, St Mary’s Campus,

School of Public Health, MRC-HPA Centre for

Environment and Health, Department of Epidemiology and Biostatistics, London, UK

(Prof M Ezzati PhD); Institute of Child Health, University College

London, London, UK (Prof S Grantham-McGregor FRCP); The

University of the West Indies, Mona, Jamaica

(Prof S Grantham-McGregor); Emory University, Atlanta, GA, USA (Prof R Martorell PhD); and London School of Hygiene and Tropical Medicine, London, UK

(Prof R Uauy PhD)

Correspondence to:Prof Robert Black, Johns Hopkins University, Bloomberg School of

Public Health, Baltimore, MD 21205, USA

rblack@jhsph.edu

The present Series is guided by a framework (fi gure 1) that shows the means to optimum fetal and child growth and development, rather than the determinants of undernutrition as shown in the conceptual model developed by UNICEF and used in the 2008 Series.1 This new framework shows the dietary, behavioural, and health determinants of optimum nutrition, growth, and development and how they are aff ected by underlying food security, caregiving resources, and environmental con ditions, which are in turn shaped by economic and social conditions, national and global contexts, resources, and governance. This Series examines how these determinants can be changed to enhance growth and development. These changes include nutrition-specifi c interventions that address the immediate causes of suboptimum growth and development. The framework shows the potential eff ects of nutrition-sensitive inter-ventions that address the underlying determinants of malnutrition and incorporate specifi c nutrition goals and actions. It also shows the ways that an enabling environment can be built to support interventions and programmes to enhance growth and development and their health consequences. In the fi rst paper we assess the prevalence of nutritional conditions and their health and development consequences. We deem a life-course perspective to be essential to conceptualise the nutritional eff ects and benefi ts of interventions. The nutritional status of women at the time of conception and during pregnancy is important for fetal growth and development, and these factors, along with nutritional status in the fi rst 2 years of life, are important determinants of both

undernutrition in childhood and obesity and related diseases in adulthood. Thus, we organise this paper to consider prevalence and con sequences of nutritional conditions during the life course from adolescence to pregnancy to childhood and discuss the implications for adult health. In the second paper, we describe evidence supporting nutrition-specifi c interventions and the health eff ects and costs of increasing their population coverage. In the third paper we examine nutrition-sensitive inter-ventions and approaches and their potential to improve nutrition. In the fourth paper we examine the features of an enabling environment that are needed to provide support for nutrition programmes and how they can be favourably changed. Finally, in a Comment6 we will examine the desired national and global response to address nutritional and developmental needs of women and children in LMICs.

Prevalence and consequences of nutritional conditionsAdolescent nutritionAdolescent nutrition is important to the health of girls and is relevant to maternal nutrition. There are 1·2 billion adolescents (aged 10–19 years) in the world, 90% of whom live in LMICs. Adolescents make up 12% of the population in industrialised countries, com pared with 19% in LMICs (appendix p 2 shows values for ten countries studied in depth).7 Adolescence is a period of rapid growth and maturation from childhood to adulthood. Indeed, some researchers have argued that adolescence is a period with some potential for height catch-up in children with

Figure 1: Framework for actions to achieve optimum fetal and child nutrition and development

Morbidity andmortality in childhood

Cognitive, motor,socioemotional development

Breastfeeding, nutrient- rich foods, and eating routine

Nutrition specificinterventionsand programmes• Adolescent health and preconception nutrition• Maternal dietary supplementation• Micronutrient supplementation or fortification• Breastfeeding and complementary feeding• Dietary supplementation for children• Dietary diversification• Feeding behaviours and stimulation• Treatment of severe acute malnutrition• Disease prevention and management• Nutrition interventions in emergencies

Feeding and caregivingpractices, parenting,stimulation

Low burden ofinfectious diseases

Food security, includingavailability, economicaccess, and use of food

Feeding and caregivingresources (maternal, household, and community levels)

Knowledge and evidencePolitics and governance

Leadership, capacity, and financial resourcesSocial, economic, political, and environmental context (national and global)

Access to and use of health services, a safe and hygienic environment

School performance and learning capacity

Adult stature

Obesity and NCDs

Work capacity and productivity

Benefits during the life course

Optimum fetal and child nutrition and development Nutrition sensitiveprogrammes and approaches• Agriculture and food security• Social safety nets• Early child development• Maternal mental health• Women’s empowerment• Child protection• Classroom education• Water and sanitation• Health and family planning services

Building an enabling environment• Rigorous evaluations• Advocacy strategies• Horizontal and vertical coordination• Accountability, incentives regulation, legislation • Leadership programmes• Capacity investments• Domestic resource mobilisation

See Online for appendix

Black et al. (2013)

BIRTH WEIGHT STUNTING

Significant Risk Factors

Household income +++ +++ Maternal height ++ ++ Altitude above sea-level + + Indian Tamil ethnicity + +

IYCF Indicators

Breast-feeding>12 months + Early initiation of BF

Exclusive BF >5 months

Dietary diversity (9-23 mos)

Iron rich diet/Meal frequency

Other Factors

Complementary food indicators

Receipt of Thriposha

Vitamin A/Iodine/Access to healthcare

Indoor air quality + Sanitation/electricity/water

9

Determinants of Stunting/Low Birth Weight, Sri Lanka DHS 2006/07

BIRTH WEIGHT STUNTING

Significant Risk Factors

Household income +++ +++ Maternal height ++ ++ Altitude above sea-level + + Indian Tamil ethnicity + +

IYCF Indicators

Breast-feeding>12 months + Early initiation of BF

Exclusive BF >5 months

Dietary diversity (9-23 mos)

Iron rich diet/Meal frequency

Other Factors

Complementary food indicators

Receipt of Thriposha

Vitamin A/Iodine/Access to healthcare

Indoor air quality + Sanitation/electricity/water

10

Determinants of Stunting/Low Birth Weight, Sri Lanka DHS 2006/07

Impact of key determinants on stunting – simulation results

•   Improving breastfeeding/complementary feeding practices will have no impact

•   Only three factors have significant impact

•   Increasing incomes or eliminating LBW will have modest impacts

•   Largest impact would be from increasing in maternal height

11

Simulation using DHS 2006/07 data

0.20

0.18

0.48

-0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6

Improving IYCF indicators to optimal

Raising incomes to that of richest quintile

Eliminating LBW

Increasing maternal height to 161 cm

Improvement in child growth (Z-scores)

represents 10,000 affected children

Altitude OtherIndoor air pollution

Breast/Complementary

feed/Disease/Sanitation

STUNTED

UNDERWEIGHT

Poverty/Food insecurity Parents height

12 Source: Results of population attributable risk analysis by IHP and Subramanian

The nexus between poverty and food insecurity

– Association or causal relationship?

– Why hasn’t Thriposha solved the problem?

13

Short-fall in energy intake by income decile, 2010

14

HOUSEHOLD LEVEL

15

POVERTY

Lack of Food

Inadequate Food

Diversity

Thin adults Child

Stunting/Wasting

Gap between Thriposha contribution and family deficit in calories/day

1,000 – 3,500

200 0

200

400

600

800

1,000

1,200

1,400

1,600

Shortfall in energy intake in poor Calories supplied by Thriposha

16

Is an expanded supplementary food scheme feasible?

Could it work? •   Experience of Jaffna Nutrition Rehabilitation

Programme (NRP) •   Global evidence is yes

Is it politically feasible? •   Will politicians be willing to spend the money? •   Can policy-makers be persuaded to spend more on

supplementary food given experience with Thriposha?

17

. . . Did the nutrition community miss the boat in Sri Lanka?

18

That problem of parental height

19

represents 10,000 affected children

Altitude OtherIndoor air pollution

Breast/Complementary

feed/Disease/Sanitation

STUNTED

UNDERWEIGHT

Poverty/Food insecurity Parents height

20

The meaningfulness of parental height

Increasing importance of maternal height as a risk factor in Sri Lankan data may reflect success in addressing other issues, plus short height of Sri Lankan women (151-154 cm) But raises several questions:

1.  How long will it take for Sri Lankans to close the adult height deficit if we break the “cycle of undernutrition”? –   Net increase in height unlikely to be more than 1 cm/decade!

2.  Can child growth compensate for short parental height? –   No evidence for any intervention with this size of impact –   Parental height may impose limit on potential improvement

3.  Should lower child linear growth in better-off linked to maternal height be of concern? Or does it represent normal physiological growth?

21

Bangladesh

India

Sri Lanka

Maldives

Nepal

Pakistan

BEST

BESTWORST

WORST

-2

-1

0

1

2

3

Rela

tive

perfo

rman

ce in

redu

cing

child

stu

ntin

g

-2 -1 0 1 2 3Relative performance in reducing infant mortality

Comparative performance in undernutrition vs. IMR (controlling for income)

22

Japan

Sri Lanka

BEST

WORST

BESTWORST

-3

-2

-1

0

1

2

Rela

tive

perfo

rman

ce in

redu

cing

child

stu

ntin

g

-2 -1 0 1 2 3Relative performance in reducing infant mortality

Comparative performance in undernutrition vs. IMR (controlling for income and maternal height)

23

Q3: Is lower child growth/higher

stunting due to parental height always bad?

– Some observations

24

Is lower child height in non-poor a problem?

Growth in Sri Lankan kids matches children in populations where this pattern of growth is not thought to be of concern 1.  Linear growth in better-off top quintile matches

children of South Asian-origin parents in UK and Netherlands

2.  Linear growth and stunting rates in better-off top third matches children in Japan

What does a Japanese perspective tell us?

25

Impact of using Japanese national reference on assessment of stunting (%), Sri Lanka by

quintile 31

23

17 14

9

17

11 8

6 3

-

5

10

15

20

25

30

35

Poorest Q2 Q3 Q4 Richest

WHO Standard Japanese Reference

26

Impact of using Japanese national reference on assessment of HAZ>2 (%), Sri Lanka by

quintile

1.9 1.5 1.4 1.3

1.9

3.0

2.4 2.7

3.2

5.5

-

1.0

2.0

3.0

4.0

5.0

6.0

Poorest Q2 Q3 Q4 Richest

WHO Standard Japanese Reference

27

Impact of using Japanese national reference Assessment of linear growth (HAZ): Poorest 40% vs Richest 40%,

Sri Lanka 2007

28

-1.0

-0.8

-0.6

-0.4

-0.2

-

0.2

0 1 2 3 4 Age (years)

Poorest Richest

Implications

Applying a growth reference more relevant to Sri Lanka’s population (short stature) leads to very different conclusions about:

•   Inequality in child undernutrition

•   Potential for improvement at different income levels

•   Trade-off between undernutrition and obesity

•   Interpretation of growth faltering

29

Reflections

30

For Sri Lanka policy 1.  An effective undernutrition strategy must focus on family food

insecurity with an expanded supplementary food intervention. Anything else is a waste of time.

2.  An effective undernutrition strategy must target the poor. Anybody else is a waste of time.

3.  Sri Lanka should consider the use of nationally-specific growth references as in Japan and be realistic about the limits for improvement

For South Asia

1.  The South Asian enigma may simply reflect patterns of growth linked to parental height.

2.  There may be more global success in the region, i.e., Sri Lanka and Maldives, than recognized which can be a basis for learning

31