Beruflich Dokumente
Kultur Dokumente
Abstract
Background: Poor nutrition during childhood impedes physical and mental development of children,
which propagate the vicious cycle of intergenerational under nutrition. This paper is aimed at
understanding the determinants of stunting among children aged 0 to 59 months in Bangladesh.
Methods: The study used Bangladesh Demographic and Health Survey 2011 data and a multistage
stratified cluster-sampling design. Anthropometric data (for height and weight) were collected and
analysis was limited to 7647 children. Multiple binary logistic regression analysis was performed to
assess the association of stunting with potential socioeconomic and demographic factors.
Results: The prevalence of stunting has been found to be about 41% among children aged less than 60
months and higher in rural setting than in urban areas (43% vs 36%). Adjusted model revealed that
several factors were influencing stunting. The children living in moderately food-insecure households
had higher odds of becoming stunted (odds ratio [OR] ¼ 1.27, 95% confidence interval [CI]: 1.05-1.54,
P ¼ .01) compared to the children living in food-secure households. The derived ORs of stunting for
children delivered at institutions facilitated particularly by public (OR ¼ 0.80, 95% CI: 0.67-0.96; P ¼
.02) or private (OR ¼ 0.81, 95% CI: 0.67-0.97; P ¼ .02) sectors were less than for children delivered at
home. Similarly, wealth index, exposure of mother to the mass media, age of child, size of child at birth,
and parents’ education were significantly associated with stunting.
Conclusions: Moreover, the demographic characteristics and other indicators appeared to have
significant influence in the prevalence of stunting. Public health programs are needed to avert the risk
factors of stunting among children in Bangladesh.
1
Nutrition and Clinical Services Division, International Centre for Diarrheal Disease Research, Bangladesh (icddr,b), Mohakhali,
Dhaka, Bangladesh
2
Research School of Population Health, ANU College of Medicine, Biology and Environment, Australian National University,
Acton, Australia
3
Helen Keller International, Dhaka, Bangladesh
Corresponding Author:
Haribondhu Sarma, Nutrition and Clinical Services Division, International Centre for Diarrheal Disease Research, Bangladesh
(icddr,b), Mohakhali, GPO Box 128, Dhaka 1000, Bangladesh.
Email: hsarma@icddrb.org
292 Food and Nutrition Bulletin 38(3)
Keywords
stunting, prevalence, influencing factors, children aged below five, Bangladesh
Data for this study came from the nationally rep- Figure 1. Sampling and participants enrollment.
resentative BDHS 2011. This survey was con-
ducted by the National Institute of Population children aged 0 to 59 months (Figure 1). For this
Research and Training under the authority of the analysis, we first extracted the children’s data set
Ministry of Health and Family Welfare, in colla- from the BDHS 2011 original file; then, we
boration with The United States Agency for Inter- excluded 214 children whose height-for-age
national Development (USAID). The BDHS z score (HAZ) values were implausible (6 >
report is publicly available and may be accessed HAZ > 6) or missing, and finally, our final sample
upon request (http://dhsprogram.com/data/avail- size for this analysis was 7647 children (Figure
able-datasets.cfm). 1). The survey involved well-trained data collec-
The BDHS 2011 is nationally representative and tors for conducting interviews with the caregivers
covers the entire population, using the sampling of children aged 0 to 59 months and for collecting
frame prepared for the 2011 population and housing anthropometric data (for height and weight)
census, provided by the Bangladesh Bureau of Sta- using standardized procedures.20
tistics. The primary sampling unit for the survey is Chronic undernutrition status of children aged 0
an enumeration area (EA) that was created to have to 59 months has been assessed by HAZ. Accord-
an average of about 120 households. The survey is ing to the definition of WHO, HAZ explains the
based on a 2-stage stratified sample of households. height of a child in terms of the number of standard
In the first stage, 600 EAs were selected with prob- deviations (SD) above or below the median height
ability proportional to the EA size, with 207 clusters of healthy children in the same age group.21 A
in urban areas and 393 clusters in rural areas. A child was classified as stunted (dependent vari-
complete household listing operation was then car- able) if he/she had a z score below 2 SD.
ried out in all selected EAs to provide a sampling Previous studies have shown that different risk
frame for the second stage of selection of house- factors (independent variables) were associated
holds. In the second stage, a systematic sample of with stunting in Bangladesh.9-11 Based on those
30 households per EA was selected, on average, to studies, we classified the potential risk factors
provide statistically reliable estimates of key demo- into different groups (such as parental, child,
graphic and health variables for the country as a household, and community factors). Mothers’
whole, for urban and rural areas separately, and for and fathers’ education (no education, primary,
each of the 7 administrative divisions. The survey secondary, or higher), fathers’ occupation (unem-
was conducted in 18 000 residential households ployed, farm worker, day laborer, semi-skilled
with 6210 and 11 790 in urban and rural areas, worker, and service holder/businessman), moth-
respectively. A total of 18 222 ever married women ers’ exposure (yes, no) to the mass media (radio,
were identified in these households, and 17 842 television, newspapers, and magazines), and birth
were interviewed with a response rate of 98% (Fig- place of the child (home, public sector facility,
ure 1). Interviews were conducted in the same private sector facility, and Non-governmental
dwelling place of the respondents. Among those organization [NGO] facility) were included in the
households, the survey identified a total of 7861 parental factors; age of child (0-11, 12-23, 24-35,
294 Food and Nutrition Bulletin 38(3)
36-47, and 48-59 months) and the size of child at all personal information of the respondents in the
birth (average or larger, very small, and small) database prior to making these available online.
were considered the child-level factors.
Household-level factors included household food
insecurity (secure, mildly insecure, moderately
Results
insecure, and severely insecure), wealth index Among the study children, almost 42% were living
(richest, richer, middle, poor (poorer and poorest), in poor (poorer 21%, poorest 21%) households, and
and types of toilet facilities (improved, non- about 51% of households had non-improved toilet
improved); the community-level factor included facilities. About 64% fell in the category of food-
the place of residence (urban, rural). secure households and 26% in mildly food-
insecure households. About 75% of the children
were delivered at home, and the size at birth was
Statistical Analysis average or larger for 83% of the children. Large
proportions of the parents (20% mothers and 30%
A description about the sampled participants was
fathers) had no formal education (Table 1). The
revealed using univariate analysis and reported in
prevalence of stunting in below five children was
terms of percentages, along with respective 95%
higher in the rural regions (43%) compared to the
confidence interval (CI). Bivariate analysis was
urban regions (36%). Administrative division–wise
done to see the differentials in the prevalence of
distribution of prevalence shows that Sylhet was a
stunting over geographical characteristics
high-risk area for stunting, where the prevalence
(administrative division and place of residence).
was about 50%, and Khulna and Rajshahi division
As our dependent variable is dichotomous in
were the low-risk areas for stunting (Figure 2).
nature (stunted or not stunted), we fitted simple
Table 2 shows the ORs derived from simple
and multiple binary logistic regression analysis to
and multiple logistic regression analysis for
see the association of the independent variables
assessing the association between stunting status
with dependent variable (stunting status of chil-
of children aged 0 to 59 months and sociodemo-
dren). To explore the factors that influenced
graphic characteristics of their households.
stunting, we built multiple regression model
Results from multiple logistic regression analysis
adjusted with variables having significant associ-
are described here. Child’s age in months was
ation with dependent variable in simple logistic
statistically associated with stunting; children
regression model only to best fit the regression
aged 12 to 23 months were more likely to be
model and predict the outcome variable. A
stunted (OR ¼ 3.98, 95% CI: 3.32-4.76; P ¼
P value less than .05 has been considered for
.00) than those aged 0 to 11 months. Children
statistical significance. Variations in the errors
living in the poorest households were 2.17 times
due to complex survey design were controlled
more likely to be stunted (OR ¼ 2.17, 95% CI:
during analysis. The results of logistic regression
1.70-2.76; P ¼ .00), and those who lived in
analyses have been presented in terms of odds
poorer households were 1.79 times more likely to
ratio (OR), 95% CI, and P value. Data of this
be stunted (OR ¼ 1.79, 95% CI: 1.43-2.23; P ¼ .00)
study were analyzed using R (version 3.1.0) and
compared to those born in the richest households.
STATA (version 12.0).
Exposure of mothers to the mass media seemed to
play an important role in the stunting status of chil-
dren as children of mothers with no media exposure
Ethical Approval were 1.20 times more likely to be stunted (OR ¼
The institutional review board of the Bangladesh 1.20, 95% CI: 1.06-1.36; P ¼ .00) compared to
Medical Research Council approved the BDHS children of mothers exposed to the media.
2011. Informed consent was obtained from each The result indicates that children whose par-
respondent in the survey before interviewing, and ents’ attained secondary educations were less
again, separately before taking weight and height likely to be stunted compared to those whose par-
measurements.3 The DHS Program also removed ents had no formal education. Children who were
Sarma et al 295
wealth, food insecurity, mass media exposure, households may also be affected by stunting. The
age of the child, size of the child at birth, and World Bank report suggests that in the richer
place of delivery. Our analysis also reveals that, households stunting may also arise due to inade-
sometimes, these factors are also interrelated. quate knowledge on food, feeding practices, inap-
Mass media exposure, age of the child, size of propriate food allocation, and poor hygiene
child at birth, and place of delivery are directly practices.23 Lack of maternal education, lack of
dependent on household wealth index as well as maternal knowledge about child nutrition, and
on parental education. lack of financial ability to purchase appropriate
Our study indicates that household wealth is foods are the factors associated with stunting.
significantly associated with stunting. Children One study in Northeastern Peninsular Malay-
from the poorest households are more likely to sia showed that food insecurity was associated
be stunted than those from middle, richer, and with stunting and underweight.24 Another study
richest households. These children may become from Kenya found that stunting was highest
stunted due to their mothers’ lower educational among severely food-insecure households ranked
status and lack of sources of knowledge on child in the poorest tertile and lowest among food-
nutrition. Moreover, a study in Bangladesh illus- secure households in the middle wealth tertile.25
trates that the mothers who have more knowledge Similarly, our study found that mildly and mod-
about child health and nutrition are educated and erately food-insecure households were more
come from richer households; they also have likely to have stunted children than those reported
greater access to media compared to mothers as being food secure. Similar findings were also
from poorer households.3 Financial ability of a obtained from other studies in Bangladesh,
household is also important for ensuring that the Ethiopia, Vietnam, and Colombia.26,27 In Ban-
child gets the required amount of nutritious foods. gladesh, approximately 25% of the population is
A study in Bangladesh found that mothers from food insecure.28 Approximately 31% of the rural
poor households did not have the financial ability population in Bangladesh suffers from “chronic
to purchase nutritious foods for their children.22 poverty,” which is characterized by low con-
A study in Nepal revealed that richer households sumption, lack of access to basic health services,
have more financial ability or access to nutritious and undernutrition.29 Another study also illu-
foods and are more capable of ensuring proper strated that, in Bangladesh, household food
care for the children residing in those house- security is significantly associated with child
holds.7 On the contrary, the children from richer feeding practices.30
Sarma et al 297
Table 2. Odds Ratios in Simple and Multiple Logistic Regressions Assessing the Impacts of Selected Variables on
Stunting of Children below five in Bangladesh.
Place of residence
Urban 1 1
Rural 1.39 1.26-1.54 .00 0.91 0.80-1.04 .20
Household wealth index
Richest 1 1
Richer 1.69 1.44-1.98 .00 1.34 1.11-1.61 .00
Middle 2.15 1.84-2.51 .00 1.55 1.27-1.90 .00
Poorer 2.88 2.47-3.37 .00 1.79 1.43-2.23 .00
Poorest 3.89 3.34-4.55 .00 2.17 1.70-2.76 .00
Food security status of household
Secure 1 1
Mildly insecure 1.65 1.48-1.83 .00 1.18 1.04-1.33 .01
Moderately insecure 2.31 1.96-2.72 .00 1.27 1.05-1.54 .01
Severely insecure 2.30 1.61-3.29 .00 1.32 0.89-1.94 .17
Type of toilet facilities at household
Improved 1 1
Non-improved 1.71 1.55-1.88 .00 1.11 0.99-1.24 .09
Exposure of mothers to media
Yes 1 1
No 1.84 1.68-2.03 .00 1.20 1.06-1.36 .00
Age of child (in completed months)
0-11 1 1
12-23 3.65 3.10-4.30 .00 3.98 3.32-4.76 .00
24-35 3.35 2.85-3.94 .00 3.46 2.88-4.14 .00
36-47 3.29 2.81-3.86 .00 3.27 2.75-3.90 .00
48-59 2.63 2.24-3.09 .00 2.52 2.12-3.01 .00
Size of child at birth
Average or larger 1 1
Very small 1.73 1.40-2.13 .00 1.64 1.41-1.92 .00
Small 1.65 1.43-1.89 .00 1.65 1.30-2.08 .00
Place of delivery
Home 1 1
Public sector facility 0.54 0.47-0.63 .00 0.80 0.67-0.96 .02
Private sector facility 0.42 0.36-0.49 .00 0.81 0.67-0.97 .02
NGO facility 0.69 0.50-0.96 .03 1.25 0.86-1.82 .23
Mothers’ educational level
No education 1 1
Primary 0.78 0.69-0.89 .00 0.95 0.82-1.10 .50
Secondary 0.48 0.43-0.55 .00 0.88 0.75-1.04 .15
Higher 0.23 0.18-0.29 .00 0.71 0.52-0.96 .03
Fathers’ educational level
No education 1 1
Primary 0.82 0.73-0.92 .00 1.03 0.89-1.18 .73
Secondary 0.51 0.45-0.58 .00 0.85 0.73-1.00 .05
Higher 0.28 0.23-0.33 .00 0.63 0.50-0.81 .00
(continued)
298 Food and Nutrition Bulletin 38(3)
Table 2. (continued)
Simple Logistic Regression Multiple Logistic Regression
Fathers’ occupation
Unemployed 1 1
Farm worker 1.16 0.84-1.61 .37 0.75 0.52-1.08 .12
Day laborer 1.10 0.79-1.53 .57 0.74 0.52-1.07 .11
Semi-skilled worker 0.76 0.55-1.06 .10 0.78 0.54-1.12 .18
Service holder/businessman 0.64 0.46-0.89 .01 0.75 0.52-1.08 .13
Abbreviations: CI, confidence interval; OR, odds ratio.
Exposure to mass media is important as than-average size at birth. One study illustrated
sources of knowledge. The mass media provides that stunting passes from one generation to
information that is essential to amplifying peo- another in an intergenerational cycle; this indi-
ple’s knowledge and awareness regarding issues cates that the children born to stunted mothers are
in their day-to-day life. Our study illustrates that likely to be stunted themselves.23 Considering
the children of mothers who did not listen to the such a reality, size of children at birth plays an
radio, did not watch television, and did not read important role in the prevalence of stunting.
newspapers or magazines at least once a week Besides, studies in Nepal, Bangladesh, and Ethio-
had higher risk of becoming stunted compared pia have also stated that children born with
to the children of those mothers who were smaller than average size were more prone to
exposed to any type of mass media. Moreover, stunting.7,10,32 Moreover, in Bangladesh, size of
the BDHS 2011 data suggest that almost 49% child at birth is positively associated with the
of women have no access to mass media, and less level of maternal education. A study showed that,
than 1% women were found to have all the media in Bangladesh, educated mothers tend to give
sources.3 Mass media advertisements provide birth to average or large-sized children; on the
information on health, nutrition, proper hygiene other hand, smaller children are born to the moth-
practices, proper child feeding practices, and on ers who have no education at all.3 Furthermore,
overall knowledge about health, which may con- the size of child at birth may also depend on
tribute to reduction of the prevalence of stunting multidimensional factors, including maternal
among below five children. It is well proven in nutrition status, antenatal checkup, and food
Bangladesh that mothers who have less access to intake during pregnancy, intrauterine growth, size
mass media are less likely to be exposed to appro- for gestational age, timely delivery, maternal
priate messages and campaigns on heath and vac- sociodemographic status, and household eco-
cines provided by the NGOs or government nomic status.10 These need to be considered in
organizations compared to those who have more the prevention of stunting.
access to the mass media.3 A study conducted Our study indicated that children aged 12 to
among children below the age of 2 years in Ban- 23 months are mostly at risk of stunting compared
gladesh revealed that mothers of undernourished to children aged 0 to 11 months and 24 to
children were less educated and had less access to 59 months. Appropriate complementary feeding
the mass media,31 which indicates that mothers’ practices are essential for mental and physical
education and exposure to the mass media is an development of children of these age groups.
important factor for their children’s health and Studies conducted in rural Bangladesh and in
well-being. India among children below 2 years of age sug-
Our study revealed that the likelihood of stunt- gest that the lack of knowledge about time of
ing was higher among children who had smaller- initiation of complementary feeding, dietary
Sarma et al 299
diversity, and nutritional knowledge increases the Knowledge and awareness on child nutrition
risk factors of stunting.31,33 Other studies con- among parents of below five children are important
ducted in rural Uganda, Vietnam, and Ethiopia factors for ensuring adequate and appropriate child
also reported that mothers and caregivers who feeding practices and child care. In this regard,
have little knowledge or inadequate knowledge parental education is important. Our study found
about IYCF practice34-36 ultimately lead to induce that parents’ education was a risk factor for stunt-
stunting in their children. This means that stunting ing among children aged 6 to 59 months. A study
has a relationship with child feeding practices. So, conducted in rural Bangladesh among children
it can be said that children aged 12 to 23 months below the age of 2 years revealed that educated
are at a risk of stunting because of their inappropri- mothers were more conscious about their chil-
ate child feeding practices. Stunting is a cumula- dren’s health and nutrition, enabling them to take
tive process that can begin in the uterus and better care, better use of the health services, and
continue up to the first 3 years of life.32 However, also to help ensure better hygiene practice.31 These
after the second year of a child’s life, it seems to findings were also supported by another study con-
steadily reduce again with time.10,11 The World ducted in Indonesia.37 In Bangladesh, mothers and
Bank report illustrates that the lack of care, access fathers from richer families are more likely to be
to health service, and inappropriate feeding prac- educated than those from the poorest families.3 Our
tices also lead to stunting.23 So, adequate knowl- analyses derived that there is a negative association
edge and appropriate feeding practices are between educational status of parents and stunting,
essential for the prevention of stunting. which has been observed in all BDHS from 1999 to
Place of delivery is found to be another crucial 2007.4,38,39 Besides, educated parents might have
factor for stunting. In Bangladesh, 71% of child- better income as well as better household access to
births still occur in the home while only 29% of food, higher proportion of wealth allocated for
births take place in health facilities.3 Our study children’s welfare, and improved standard of liv-
indicates that children born in health facilities ing,10 which would ultimately help to provide bet-
with the help of health professionals were less ter care to the children. So, education is one of the
likely to be stunted compared to the children born essential needs toward preventing stunting.
at home with the help of traditional birth atten- One of the limitations of this study is that we
dants (trained or untrained). Similarly, several depend only on secondary data sources which do
other studies conducted in Nepal, Bangladesh, not contain detailed information on dietary diver-
and India among below five children also suggest sity or dietary practice of the study children. Our
that children born in health facilities were less study depends on data sources from other studies
likely to be stunted compared to children born for clarification or further explanation. This study
at home.7,10,33 There are some factors associated was unable to present sufficient information on
with delivery at health facilities and at home. children’s birth weights as well as gestational age
Studies conducted in Bangladesh and Nepal from intrauterine growth restriction as a cause for
revealed that educated mothers from richer fam- size at birth. Also, these data do not represent the
ilies go to health facilities for antenatal visits and population or the children residing in urban slums,
child delivery.3,7 Health professionals at hospitals although children of urban slums are more prone to
and clinics provide counseling on antenatal care stunting than other children residing in the urban
for safe delivery and advise to receive postnatal areas. Despite these limitations, the strength of this
care. Besides, the counseling is given on IYCF study is that our results have identified the most
practices to the new mothers helping them to take vital factors for stunting, which will be important
proper care of their children. However, the assur- contributions to the available literature on the asso-
ance of safe delivery and counseling on IYCF ciation of socioeconomic and demographic vari-
practices are not provided in the rural context ables with stunting of below five children.
during home births in Bangladesh. Considering Our results revealed that stunting is most com-
the current situation, below five children are at mon among the children aged 12 to 23 months in
risk of stunting in Bangladesh. Bangladesh. The demographic characteristics of
300 Food and Nutrition Bulletin 38(3)
the study participants and other indicators, from BDHS 2011. J Food Sec. 2013;1(2):
including household socioeconomic status (poor- 52-57.
est), place of delivery (home delivery), size at 3. Bangladesh Demographic and Health Survey.
birth (small size), and parents’ education (no for- Bangladesh Demographic and Health Survey.
mal education) appeared to influence the preva- Fairfax, VA: ICF International; 2013.
lence of stunting significantly. Several potential 4. Bangladesh Demographic and Health Survey.
educational interventions by the policymakers are Bangladesh Demographic and Health Survey.
needed for developing information that can serve ORC Macro. USA: 2005.
as a platform to prevent stunting among children 5. Bangladesh Demographic and Health Survey. Ban-
aged below five years in this country. gladesh Demographic and Health Survey. Fairfax,
VA: ICF International; 2013 (policy brief).
Authors’ Note 6. World Food Programme. WFP Bangladesh Nutri-
Jahidur Rahman Khan is currently affiliated with the tion Strategy 2012-2016. Dhaka, Bangladesh:
Center for Bioinformatics Learning Advancement and World Food Programme Bangladesh; 2012.
Systematics Training (cBLAST), University of Dhaka, 7. Tiwari R, Ausman LM, Agho KE. Determinants of
Bangladesh. This study was conducted by team mem- stunting and severe stunting among under-fives:
bers of Nutrition Program Evaluation Unit under Nutri- evidence from the 2011 Nepal Demographic and
tion and Clinical Services Division of icddr,b as part of Health Survey. BMC Pedia. 2014;14:239.
their research activities. 8. UNICEF. Improving Child Nutrition: the Achiev-
able Imperative for global progress. UNICEF.
Acknowledgments New York: 2013.
The authors acknowledge the contributions of the 9. ACC/SCN. Nutrition: A Foundation for Develop-
Bangladesh Demographic and Health Survey 2011 team ment. Geneva, Switzerland: WHO; 2002.
for their efforts in providing open access to their dataset; 10. Rahman A, Chowdhury S. Determinants of
icddr,b is grateful to the Government of Bangladesh,
chronic malnutrition among preschool children
Canada, Sweden, and the United Kingdom for providing
in Bangladesh. J Biosoc Sci. 2007;39(2):161.
core/unrestricted support. The authors thank Ms Antara
Kabir for support in editing the manuscript. 11. Mostafa KS. Socio-economic determinants of
severe and moderate stunting among under-five
Declaration of Conflicting Interests children of rural Bangladesh. Malays J Nutr.
2011;17(1):105-118.
The author(s) declared no potential conflicts of interest
with respect to the research, authorship, and/or publi- 12. Jesmin A, Yamamoto SS, Malik AA, Haque MA.
cation of this article. Prevalence and determinants of chronic malnutrition
among preschool children: a cross-sectional study
Funding in Dhaka City, Bangladesh. J H Pop and Nutr.
The author(s) disclosed receipt of the following finan- 2011;29(5):494-499.
cial support for the research, authorship, and/or publi- 13. United Nations. Transforming our world: The
cation of this article: The Children’s Investment Fund 2030 agenda for sustainable development. United
Foundation-UK (CIFF) was the funder for operating Nations. 2015. https://sustainabledevelopment.u
research work of this unit. Authors invested their valu- n.org/content/documents/21252030%20Agenda
able time for this study. So, the study was indirectly %20for%20Sustainable%20Development%20
funded by CIFF for the time invested by authors. web.pdf. Accessed May 25, 2016.
14. World Health Organization. Nutrition Landscape
References Information System (NLIS) country profile indica-
1. UNICEF. UNICEF-WHO-World Bank Joint Child tors: interpretation guide. WHO. Switzerland.
Malnutrition Estimates. New York, NY: United 2010. Available at: http://apps.who.intl/iris/bit
Nations Children’s Fund; Geneva, Switzerland: stream (Accessed May 29, 2016)
World Health Organization; Washington, DC: The 15. Hong R, Banta JE, Betancourt JA. Relationship
World Bank; 2012. between household wealth inequality and chronic
2. Hasan MM, Ahmed S, Chowdhury MAH. Food childhood under-nutrition in Bangladesh. Int J
insecurity and child undernutrition: evidence Equity Health. 2006;5:15.
Sarma et al 301
16. Ermias A. The prevalence of stunting and associ- but the effect is not mediated by child dietary
ated factors among children age 6-59 months at diversity. J Nutr. 2013;143(12):2015-2021.
Mizan-Aman Town, Bench Maji zone, SNNPR 28. Magnani R, Oot L, Sethuraman K, Kabir G, Rahman
region, Ethiopia, [Doctoral dissertation]AAU; S. USAID Office of Food For Peace Food Security
Ethiopia: 2015 Country Framework for Bangladesh (FY 2015–
17. Jiang Y, Su X, Wang C, et al. Prevalence and risk 2019). Washington, DC: FHI 360/FANTA; 2015.
factors for stunting and severe stunting among chil- 29. Sen B, Hulme D. Chronic Poverty in Bangladesh:
dren under three years old in mid-western rural areas Tales of Ascent, Descent, Marginality and Persis-
of China. Child Care Health Dev. 2015;41(1):45-51. tence. Dhaka, Bangladesh: Bangladesh Institute of
18. Asfaw M, Wondaferash M, Taha M, Dube L. Pre- Development Studies; 2006.
valence of under nutrition and associated factors 30. Saha KK, Frongillo EA, Alam DS, Arifeen SE,
among children aged between six to fifty nine Persson LÅ, Rasmussen KM. Household food
months in Bule Hora district, South Ethiopia. BMC security is associated with infant feeding prac-
Public Health. 2015;15:41. tices in rural Bangladesh. J Nutr. 2008;138(7):
19. Ramli KEA, Inder KJ, Bowe SJ, Jacobs J, Dibley 1383-1390.
MJ. Prevalence and risk factors for stunting and 31. Ahmed T, Roy S, Alam N, Hossain MI. Determi-
severe stunting among under-fives in north Maluku nants of under nutrition in children under 2 years
province of Indonesia. BMC Pediatr. 2009;9:64. of age from rural Bangladesh. Indian Pediatr.
20. Onis M. WHO Child Growth Standards based on 2012;49(10):821-824.
length/height, weight and age. Acta paediatrica. 32. Yimer G. Malnutrition among children in Southern
2006;95(S450):76-85. Ethiopia: Levels and risk factors. Ethio J H Deve.
21. De Onis M. WHO child growth standards: length/ 2000; 14(3): 283-292.
height-for-age, weight-for-age, weight-for-length, 33. Menon P, Bamezai A, Subandoro A, Ayoya MA,
weight-for-height and body mass index-for-age. Aguayo V. Age-appropriate infant and young child
WHO. Switzerland: 2006. feeding practices are associated with child nutri-
22. Osei A, Pandey P, Spiro D, Nielson J, Shrestha R, tion in India: insights from nationally representa-
Talukder Z, Haselow N. Household food insecur- tive data. Matern Child Nutr. 2015;11(1):73-87.
ity and nutritional status of children aged 6 to 23 34. Joyce N, Joshua KM. Barriers to infant and child-
months in Kailali District of Nepal. Food and Nutr feeding practices: a qualitative study of primary care-
Bull. 2010; 31(4):483-94. givers in rural Uganda. J Nutr. 2015;33(1):106-116.
23. World Bank. Repositioning Nutrition as Central to 35. Nguyen PH, Menon P, Ruel M, Hajeebhoy N. A
Development: A Strategy for Large Scale Action. situational review of infant and young child feed-
Washington, DC: The International Bank for Recon- ing practices and interventions in Viet Nam. Asia
struction and Development/The World Bank; 2006. Pac J Clin Nutr. 2011;20(3):359-374.
24. Ali Naser I, Jalil R, Muda W, et al. Association 36. Semahegn A, Tesfaye G, Bogale A. Complemen-
between household food insecurity and nutritional tary feeding practice of mothers and associated
outcomes among children in Northeastern of Penin- factors in Hiwot Fana Specialized Hospital, East-
sular Malaysia. Nutr Res Pract. 2014;8(3):304-311. ern Ethiopia. Pan Afr Med J. 2014;18:143.
25. Mutisya M, Kandala NB, Ngware MW, Kabiru 37. Semba RD, De Pee S, Sun K, Sari M, Akhter N,
CW. Household food (in)security and nutritional Bloem MW. Effect of parental formal education on
status of urban poor children aged 6 to 23 months risk of child stunting in Indonesia and Bangladesh:
in Kenya. BMC Public Health. 2015;15:1052. a cross-sectional study. Lancet. 2008;371(9609):
26. Hackett M, Melgar-Quiñonez H, Álvarez MC. 322-328.
Household food insecurity associated with stunt- 38. BDHS. Bangladesh Demographic and Health Sur-
ing and underweight among preschool children in vey. ORCM. USA: 2001.
Antioquia, Colombia. Rev Panam Salud Publica. 39. BDHS. Bangladesh Demographic and Health Sur-
2009;25(6):506-510. vey. Dhaka, Bangladesh and Calverton, Maryland,
27. Ali D, Saha KK, Nguyen PH, et al. Household food USA: National Institute of Population Research
insecurity is associated with higher child under and Training, Mitra and Associates, and Macro
nutrition in Bangladesh, Ethiopia, and Vietnam, International: 2007.