Undernutrition

Maps of
Bangladesh
2012

Foreword
Over the past two decades, Bangladesh has made impressive economic and social gains. It has significantly
reduced the incidence of poverty, increased education levels, reached self-sufficiency in rice production and
lowered fertility rates and child mortality. However, challenges still remain to keep undernutrition levels below
the World Health Organizations (WHO) public health critical thresholds. Undernutrition costs Bangladesh more
than 7,000 Crore Taka (US$ 1 billion) in lost productivity every year, and even more in health care costs1.
With an increased focus on improving child nutrition, it is critical to understand the geographical variations and
spatial patterns of undernutrition in Bangladesh.
The Bangladesh Bureau of Statistics (BBS) and the World Food Programme (WFP) have a long-standing
partnership and experience in developing poverty maps and undernutrition maps. Responding to the need for
undernutrition maps, BBS, partnering with WFP and with generous financial contribution from the International
Fund for Agricultural Development (IFAD) initiated the undernutrition mapping exercise in May 2013. We are
thankful to Massey University, New Zealand for its technical expertise to conduct and complete this exercise.
The initiative has produced estimates of underweight and stunting, two measures of undernutrition, which
are not subject to major seasonal variation. These estimates for children under five years of age are statistically
valid down to upazila (sub-district) level. These are based on applying the Small Area Estimation technique on
data from the Population and Housing Census 2011, Child and Mother Nutrition Survey 2012 and Health and
Morbidity Status Survey 2012. The resulting estimates have been mapped out to create powerful visual tools,
which are easily understood and objectively verifiable. The mapping exercise was also marked with providing
hands-on training at BBS on poverty and undernutrition mapping, jointly with the World Bank.
The aim of the mapping exercise was to serve policy makers, planners and researchers with a tool to identify the
spatial variation of undernutrition rates in Bangladesh at finer administrative levels. We hope that the district and
upazila level estimates of undernutrition rates, along with the associated maps, will be used in a complementary
manner with key correlates of undernutrition to strengthen decision making and investments on targeted and
integrated nutrition-focused development.
Behind the colours and percentages of the maps are millions of children who require our attention to achieve full
mental and physical potential for a brighter future. We hope that these maps lead to an enhanced engagement of
the Government of Bangladesh and development partners in creating a Bangladesh free of child undernutrition.

Christa Rder

Golam Mostafa Kamal

Hubert Boirard

Country Representative
World Food Programme

Director General
Bangladesh Bureau of Statistics

Country Programme Manager

International Fund for Agricultural
Development

Acknowledgement
The Undernutrition Maps of Bangladesh are the final products
of a collaborative effort between the Bangladesh Bureau
of Statistics (BBS) of the Statistics and Informatics Division
(SID), Ministry of Planning and the World Food Programme
(WFP). The project was funded by the International Fund
for Agricultural Development (IFAD) and WFP. Massey
University, New Zealand provided technical expertise for the
analytical work.
The team acknowledges the leadership and support of Ms.
Suraiya Begum (Secretary, SID), Mr. Md. Nojibur Rahman
(former Secretary, SID) and Mr. Golam Mostafa Kamal (Director
General, BBS). The team would like to thank the members of
the Steering Committee and the Technical Committee, who
reviewed and endorsed both the methodology and results of
the undernutrition mapping exercise.
Professor Stephen Haslett and Associate Professor Geoffrey
Jones of the Institute of Fundamental Sciences, Massey
University spearheaded the analytical work in deriving
upazila level estimates of undernutrition rates. Dr. Marissa
Isidro assisted them in this work.
The BBS team included Mr. Md. Zahidul Hoque Sardar,
Dr. Dipankar Roy, Mr. Anwar Hossain, Mr. Md. Rezaul Karim,
Mr. Md. Jibon Miah and Mr. Dinesh Roy. The WFP team
included Ms. Nusha Choudhury, Ms. Kayenat Kabir and Mr.
Mahabubul Alam of the Vulnerability Analysis and Mapping
(VAM) unit working under the guidance of Ms. Christa Rder
(Country Representative).

Howlader, Sushil Ranjan; Sethuraman, Kavita;

Begum, Ferdousi; Paul, Dipika; Sommerfelt,
A. Elisabeth; Kovach, Tara. 2012. Investing in
Nutrition Now: A Smart Start for Our Children,
Our Future. Estimates of Benefits and Costs
of a Comprehensive Program for Nutrition
in Bangladesh, 20112021. PRO-FILES and
Nutrition Costing Technical Report. Washington,
DC: Food and Nutrition Technical Assistance III
Project (FANTA), FHI 360

Introduction
Undernutrition mapping is a statistical
exercise to estimate the percentage of the

Administrative Structure of Bangladesh

total population of children under five years

of age who are suffering from undernutrition
in a defined geographical area. As with the

Bangladesh Poverty Maps2, there has been

Division

considerable demand from policy makers,

World Food Programme (WFP) initiated the

undernutrition mapping exercise in May

Un

Statistics (BBS), in partnership with the

64

targeting. The Bangladesh Bureau of

District

to aid policy making and programme

7 7 59

disaggregated level undernutrition maps

ion & Ward

development partners and researchers for

2013 to produce reliable undernutrition

estimates and maps for key sub-divisional

Up

administrative units (district and upazila3)

applying

the

Small

Area

& Tha

na

54 4

Estimation

(SAE) technique on data mainly from the

Population and Housing Census (Census)

azila

Source: Bangladesh Population and Housing Census 2011

2011 and the Child and Mother Nutrition

Survey (CMNS) 2012.
Undernutrition encompasses low birth

By Division

weight, stunting, wasting, underweight

and

micronutrient

deficiencies.

Districts

Two

central measures of undernutrition which

are not subject to seasonal variation are

Barisal

6
40

considered for small area estimation in this

mapping exercise, both based on body
measurementsstunting

11

Chittagong

111

(low height-for-

age) and underweight (low weight-for-age)

17

Dhaka

163

among children under five.

10

Khulna
The third generation of upazila level poverty maps of
Bangladesh were published in August 2014 by BBS, the
World Bank and WFP.

Upazila and metropolitan thana are referred together as

Upazila in this report.

Rajshahi

Within a particular area, stunting is defined as the

proportion of such children with a standardized heightfor-age (HAZ) value below 2 standard deviations.
Children with standardized height-for-age below
3 standard deviations are considered severely
stunted. Similarly underweight is the proportion with
a standardized weight-for-age (WAZ) value below 2
deviations, and severe underweight below 3 deviations.

Rangpur

64
8
70
8
58

Sylhet

Upazilas

4
38

Source: Bangladesh Population and Housing Census 2011

Stunting can be regarded

as evidence of chronic
undernutrition.

Unlike

wasting (low weight-forheight), the development

of stunting is a slow
cumulative process that
may

not

be

evident

immediately. It has been

recommended

as

the

Sustainable Development

Undernutrition Mapping
Methodology
Small-area estimates (SAE) of stunting and underweight in children under
five years of age in Bangladesh are produced at upazila (sub-district) level
by combining survey data from CMNS 2012 and Health and Morbidity
Status Survey (HMSS) 2012 with auxiliary data derived from Census 2011.
SAE is a mathematical technique to extract more detailed information from
existing data sources by statistical modelling. The SAE method used for
the undernutrition mapping exercise is an extended and modified version
of the ELL method (developed by Elberts et al, 2003) which has also been
used for developing the Bangladesh Poverty Maps.

Goal (SDG) undernutrition

The CMNS, which is conducted by BBS every 4-5 years, is one of the

indicator.

three main regular surveys5 in Bangladesh which provides nationally

Underweight reflects both

chronic undernutrition and
acute undernutrition - it
is a combined indicator.
Progress
Millenium

towards

the

Development

Goal (MDG) of halfing

representative data on anthropometric measurements with which

stunting and underweight can be estimated. However, the data from
CMNS is representative only at division level because of the limited
sample size. The Population and Housing Census includes all households
in the country but it has more limited data in terms of topics covered and
does not include child weight or height needed to estimate stunting and
underweight directly.

the proportion of people

The SAE technique takes advantage of the combined strengths of

who suffer from hunger

the survey and the census. The data sources have a common set of

between 1990 and 2015

explanatory variables. Using these, two models, one for standardized

is monitored through this

height-for-age and the second for standardized weight-for-age, have

indicator.

been fitted to the CMNS data. These models provide predictions of the
two measures for every child under five years in the census. In turn, the
predictions have then been used for estimating stunting and underweight
and their accuracy for children under five, by converting height-for-age to
stunting and weight-for-age to underweight, followed by aggregation to
district and upazila levels.
In August 2014, the preliminary undernutrition maps and technical report
were reviewed by the Technical Committee members. The final versions
of the maps were endorsed in the meeting of the Steering Committee on
3 November 2014.
For visual display of the spatial variation in stunting and underweight at
district level and upazila level the estimates have been presented in the
four maps on the next pages.

The other two are Bangladesh Demographic and Health Survey (BDHS) conducted under the
authority of the National Institute of Population Research and Training (NIPORT) of the Ministry of
Health and Family Welfare and the Multiple Indicator Cluster Survey (MICS) conducted by BBS. The
MICS 2009 did not survey for anthropometric measurements and the MICS 2012-13 dataset which
does have them was not available yet during the study period.

Stunting Map
District Level

Percentage of Stunted* Children

Under Five Years of Age
*Note: The percentage of children under five years of
age whose height-for-age is more than two standard
deviations below the median for the international
reference population (WHO standard)

Underweight Map
District Level

Percentage of Underweight* Children

Under Five Years of Age
*Note: The percentage of children under five years of
age whose weight-for-age is more than two standard
deviations below the median for the international
reference population (WHO standard)

Stunting Map
Upazila Level

Percentage of Stunted* Children

Under Five Years of Age
*Note: The percentage of children under five years of
age whose height-for-age is more than two standard
deviations below the median for the international
reference population (WHO standard)

Underweight Map
Upazila Level

Percentage of Underweight* Children

Under Five Years of Age
*Note: The percentage of children under five years of
age whose weight-for-age is more than two standard
deviations below the median for the international
reference population (WHO standard)

Results at a Glance
Division Level Undernutrition Estimates
The small area estimates for the child undernutrition measures, stunting and underweight, were accumulated to
high levels of aggregation for comparison with the direct estimates available from the CMNS 2012. The tables
below show a comparison of the estimates from CMNS 2012, the mapping exercise, the Bangladesh Demographic
and Health Survey (BDHS) 2011 and the Multiple Indicator Cluster Survey (MICS) 2012-2013. The division level
estimates from the mapping exercise have better agreement with both CMNS 2012 and BDHS 2011, than they
have among themselves.

Comparison of estimates of prevalence of stunting

CMNS 2012, Undernutrition Mapping Exercise (SAE), BDHS 2011 and MICS 2012-2013
Division

CMNS

SAE

BDHS

MICS

Share of stunted children

(as % of overall no. of stunted children-SAE result)

Barisal

31.0

39.7

45.1

41.4

5.5

Chittagong

45.9

42.1

41.3

43.1

22.1

Dhaka

42.6

40.4

43.3

42.1

31.3

Khulna

34.9

39.5

34.1

34.4

8.9

Rajshahi

39.3

41.0

33.7

39.4

11.7

Rangpur

36.2

42.1

42.9

43.7

11.4

Sylhet

51.3

44.6

49.3

50.6

9.1

Comparison of estimates of prevalence of underweight

CMNS 2012, Undernutrition Mapping Exercise (SAE), BDHS .2011 and MICS 2012-2013
Division

CMNS

SAE

BDHS

MICS

Barisal

26.7

33.4

40.0

35.2

Chittagong

39.4

36.8

37.4

32.2

Dhaka

33.5

32.3

36.6

30.8

Khulna

26.2

32.0

29.1

28.5

Rajshahi

37.3

34.0

34.2

29.9

Rangpur

32.7

35.7

34.5

32.6

Sylhet

39.5

38.5

44.9

39.7

Results from CMNS 2012 and the undernutrition mapping exercise show that both stunting and underweight
rates at division level are highest in Sylhet, followed by Chittagong. Stunting and underweight rates according
to BDHS 2011 and MICS 2012-13 are also highest in Sylhet. Sylhet and Chittagong also have the highest share
of the countrys total stunted children.
Barisal and Khulna have the lowest stunting and underweight rates as well as the lowest share of the countrys
total stunted children.

District and Upazila Level Undernutrition Estimates

Division

Stunting Rate
(at District Level)

Underweight Rate
(at District Level)

Minimum

Maximum

Minimum

Maximum

Barisal

37.0
(Jhalokati)

42.9
(Bhola)

28.8
(Jhalokati)

38.1
(Bhola)

Chittagong

40.1
(Chandpur)

47.7
(Bandarban)

32.8
(Rangamati)

39.7
(Lakshmipur)

Dhaka

34.2
(Dhaka)

44.3
(Kishoreganj)

22.8
(Dhaka)

39.1
(Netrakona)

Khulna

36.7
(Narail)

41.7
(Kushtia)

29.1
(Khulna)

34.2
(Chuadanga)

Rajshahi

36.9
(Joypurhat)

45.1
(Sirajganj)

30.1
(Joypurhat)

36.7
(Sirajganj)

Rangpur

40.6
(Panchagarh)

43.4
(Nilphamari)

33.0
(Dinajpur)

38.7
(Kurigram)

Sylhet

43.8
(Maulvibazar)

46.1
(Sunamganj)

36.7
(Maulvibazar)

40.9
(Sunamganj)

Stunting rates are highest in Bandarban district of Chittagong division, which is a hilly area; followed by Coxs
Bazar district in the coastal belt of Chittagong division and Sunamganj district of Sylhet division which largely
consists of haor (wetland) areas. Both stunting and underweight rates are lowest in Dhaka district where the
national capital is situated.
39 out of 64 districts in Bangladesh have stunting rates above the WHO critical threshold level for stunting (40
percent). And 55 districts are above WHO critical threshold level for underweight (30 percent).
At upazila level 300 out of a total of 544 upazilas (55 percent) in Bangladesh have a stunting rate above 40
percent and 440 upazilas (81 percent) have an underweight rate above 30 percent.
The stunting rate varies from as low as 28 percent in Dakhshinkhan thana, Dhaka district to as high as 51
percent in Ukhiya upazila, Coxs Bazar district.
The upazilas surrounding the divisional cities of Dhaka, Chittagong, Khulna and Rajshahi have the lowest rates
of stunting in the respective divisions.
Upazilas within a district have considerable spatial variation in undernutrition rates. The difference in stunting
rate between the highest and lowest ranked upazila within the respective districts range from 5 to 13 percent.
Within Dhaka district, which has the lowest stunting rate in Bangladesh, the stunting rates varies from as low
as 28 percent in Dakhshinkhan thana to as high as 41 percent in Keraniganj thana.

Undernutrition Rate vs. Size of Undernourished Population

The upazila level stunting/underweight maps are based on the percentage of stunted/underweight children of
the total population of children under five in the respective upazilas. A map based on headcount rate can be
different from a map based on absolute numbers. The area circled in red is Dhaka district which has the lowest
rate of stunting among all the districts in Bangladesh, yet it has a large number of stunted children residing in the
area as the total population of children under five in Dhaka itself is high. Whereas in the hilly Bandarban district
(circled in blue), the stunting rate is very high, but the number of stunted children is relatively small.

Density of Stunted Children

Causes of Undernutrition
Undernutrition is the outcome of insufficient intake of
energy, protein and/or micronutrients; poor absorption
or rapid loss of nutrients due to illness or increased
energy expenditure. The causes of undernutrition
are multi-level and multi-sectoral. The underlying
causes at household level are insufficient access to
food, inadequate maternal and child care practices
WFP/Kauser Haider

and poor water, sanitation, and inadequate health

services. Basic causes at societal level include power
relationships and socio-economic conditions that are
not specific to nutrition but can have powerful impacts6.
The next section explores the upazila level stunting
map in Bangladesh in relation to the widely accepted
underlying and basic causes of undernutrition.

Outcome

Undernutrition, death
and disability

Immediate causes
Inadequate dietary
intake

Underlying causes
at household/family
level

Basic causes at
societal level

Insufficient
access to FOOD

Disease

Inadequate
maternal and child
CARE practices

Poor water, sanitation

and inadequate
HEALTH services

Quantity and quality of actual resources: human, economic and

organisational and the way they are socially and culturally controlled

Potential resources: environment, technology, people

Adapted from UNICEF 1990 model

6

UNICEF 1998, The State of Worlds Children

10

Undernutrition and Poverty

While the eastern divisions namely Sylhet and Chittagong fare better at poverty status than their western
counterparts of Khulna and Rajshahi, the scenario reverses in the undernutrition maps. In Sylhet division, stunting
and underweight rates are not only high in the upazilas of the north eastern haor region, but in all upazilas stunting
and underweight rates are above WHO critical threshold level. In contrast, although Khulna division has one of

Stunting Map

11

the highest poverty rates, it has one of the lowest stunting rates. The BDHS 2011 reports one in four children
(26 percent) under five years old stunted and one in five children (21 percent) underweight, even in the highest
household wealth quintile. Clearly factors contributing to undernutrition lie beyond household income levels.

Upper Poverty Line Map 2010

12

Undernutrition and Care Practices

The first 1000 days (from conception to two years of age) is the most crucial time for a childs growth potential.
Appropriate infant and young child feeding (IYCF) practices include initiating timely feeding of solid or semisolid
foods at the age of six months and increasing the amount and variety of foods and frequency of feeding as the

Stunting Map

13

child gets older, while maintaining frequent breastfeeding. A comparison of the maps below show that Khulna
division which has a low stunting rate compared to other divisions also has better IYCF practices; while Sylhet and
Chittagong divisions, which have high stunting rates, do not fare well on this indicator.

Quality of Infant and Young Child Feeding Practices

14

Undernutrition and the Public Environment

In Bangladesh, the use of unsafe water for bathing and washing is a major cause of diarrhoea and other infections
that lead to undernutrition, as is the use of open space as toilet. A combined spatial overview of household access
to drinking water and adequate sanitation is provided. Areas along the coastal belt, in the Chittagong Hill Tracts,

Stunting Map

15

in Sylhet division and in the drought zone in the westernmost part of Rajshahi division are among the regions
performing worst on access to drinking water and sanitary toilet.

Household Access to Sanitation and Drinking Water

16

Undernutrition and Female Education

A mother completing secondary school is associated with having significant protective effect against her child
being stunted, and this is also significantly correlated with increased dietary diversity7.The comparative maps
show pockets in Barisal division and a number of districts in Dhaka and Chittagong divisions which have high
female secondary education completion rates and low stunting rates among children. Mothers with no schooling

Stunting Map

IFPRI (2013). Agriculture, Income, and Nutrition Linkages in India. Available from: http://www.ifpri.org/sites/default/files/publications/ifpridp01195.pdf

17

are more than two times more likely to have stunted children than educated mothers8. While higher education
levels of mothers may also be indicative of the income status of the household, poverty is definitely not the only
determining factor of undernutrition

Secondary Education Completion Rate of Female Adult

BDHS 2011 results show that 51% of children of mothers with no schooling were stunted compared to 23% for children of mothers with secondary or
higher education.

18

Concluding Remarks
Maps are easily understood and objectively verifiable, but they do not tell us directly
about the underlying causes of a situation, in this case of undernutrition in a particular
area. These maps should thus be used with caution and in a complementary manner
with other maps and information on possible underlying causes and up-to-date local
information. The comparison of the undernutrition maps with key causal maps indicate
that the commonly identified causes do not have the same level of effect in all areas, but
there is an interplay of different factors. As the causes of undernutrition are manifoldso are the solutions, there is no one solution, but the solutions are multi-sectoral.
The regional concentration of undernutrition suggests that geographic targeting of
nutrition interventions can have a high pay off in achieving significant reductions in the

Kayenat Kabir

number and rate of undernourished children.

Report prepared by Kayenat Kabir, Maps by Mahabubul Alam

Cover Photo-Huzzatul Mursalin, Graphic Design-Mohammad Inamul Shahriar
Artwork- Farzana Yasmin Sumona, Grade 5, Runner up WFP School Feeding Drawing Competition 2014

World Food Programme

Bangladesh Bureau of Statistics

IDB Bhaban (17th floor)

E/8-A, Rokeya Sharani
Sher-e-Bangla Nagar
Dhaka-1207, Bangladesh
Tel: 880 2 9183022-25
Fax: 880 2 9183020
www.wfp.org/countries/bangladesh

Parishankhyan Bhaban
E-27/A, Agargaon
Sher-e-Bangla Nagar
Dhaka-1207, Bangladesh
Tel: 880 2 9112589
Fax: 880 2 9111064
www.bbs.gov.bd

International Fund for

Agricultural Development
Bangladesh Office
E/8-A, Rokeya Sarani
Sher-e-Bangla Nagar
Dhaka 1207, Bangladesh
Tel: +880 2 9183022 (Ext 2115)
www.ifad.org