- Email: [email protected]
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey Mohammad Rocky Khan Chowdhury, MSc1,2, Mohammad Shafiur Rahman, MSc2,3, Mohammad Mubarak Hossain Khan, PhD4,*, Mohammad Nazrul Islam Mondal, PhD1, Mohammad Mosiur Rahman, PhD1,5, and Baki Billah, PhD6 Objective To identify the prevalence and risk factors of child malnutrition in Bangladesh. Study design Data was extracted from the Bangladesh Demographic Health Survey (2011). The outcome measures were stunting, wasting, and underweight. c2 analysis was performed to find the association of outcome variables with selected factors. Multilevel logistic regression models with a random intercept at each of the household and community levels were used to identify the risk factors of stunting, wasting, and underweight. Results From the 2011 survey, 7568 children less than 5 years of age were included in the current analysis. The overall prevalence of stunting, wasting, and underweight was 41.3% (95% CI 39.0-42.9). The c2 test and multilevel logistic regression analysis showed that the variables age, sex, mother’s body mass index, mother’s educational status, father’s educational status, place of residence, socioeconomic status, community status, religion, region of residence, and food security are significant factors of child malnutrition. Children with poor socioeconomic and community status were at higher risk of malnutrition. Children from food insecure families were more likely to be malnourished. Significant community- and household-level variations were found. Conclusions The prevalence of child malnutrition is still high in Bangladesh, and the risk was assessed at several multilevel factors. Therefore, prevention of malnutrition should be given top priority as a major public health intervention. (J Pediatr 2016;-:---).
M
alnutrition refers to inadequate dietary intake, infectious disease, or a combination of both.1-3 Three common indices of malnutrition for children are stunting (low height-for-age), wasting (low weight-for-height), and underweight (low weight-for-age).4-6 About 165 million or 26% of the world’s children less than 5 years of age are stunted, which slows down cognitive and physical development.7 Over 52 million children or around 8% of the world’s children less than 5 years of age suffer from wasting because of acute malnutrition, which noticeably increased the risk of death.6 Similarly, an estimated 101 million or 16% of world’s children less than 5 years of age are underweight.7 Malnutrition can exacerbate the impact of disease and nearly one-half of all child deaths globally are attributable to this cause.2 Children suffering malnutrition are more likely to die from common childhood illnesses such as diarrhea, pneumonia, malaria, measles, and AIDS.8 The primary causes of malnutrition include a lack of quality food, poor infant and child feeding and care practices such as suboptimal breastfeeding, deficiencies of micronutrients such as vitamin A or zinc, and recurrent attack of infections, often intensified by intestinal parasites.9,10 Child malnutrition also draws parallel links with demographic aspects,2,11-13 environmental aspects,2,14 socioeconomic aspects,6,12,15,16 parental characteristics,2,11 household possession,17 and geographical location.11 Bangladesh has made significant progress in the health and human development sectors since its independence in 1971.18,19 In Bangladesh, the child mortality per 1000 live births declined from 144 in 1990 to 41 in 2012, with an annual rate of reduction of 5.5%. Bangladesh has already achieved the MillenFrom the Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi, nium Development Goal 4 and proven its achievement to be more impressive Bangladesh; Department of Public Health, Faculty of Health Science, First Capital University of Bangladesh, than other South Asian countries, particularly, India, Pakistan, and Chuadanga, Bangladesh; Department of Global Health Afghanistan.20 Unfortunately, Bangladesh was not very successful in addressing Policy, School of International Health, The University of Tokyo, Tokyo, Japan; Department of Public Health the problems of child malnutrition. Severe malnutrition increased from 16% in Medicine, School of Public Health, Bielefeld University, Bielefeld, Germany; International Health Section, 2011 to 18% in 2013 which may be due to increasing rate of malnutrition among Division of Public Health, Graduate School of Medicine, Tokyo Medical and Dental University, Tokyo, Japan; and mothers, poor socioeconomic status of parents, relatively low rate of fully breastDepartment of Epidemiology and Preventive Medicine, 21 feeding babies, and food insecurity. The prevalence malnutrition in children School of Public Health and Preventive Medicine, 1
2
3
4
5
6
Monash University, Melbourne, Australia *Current affiliation: Department of Public Health, College of Applied Medical Sciences, King Faisal University, Al-Hassa, Saudi Arabia.
BDHS BMI WHO
Bangladesh Demographic Health Survey Body mass index World Health Organization
The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2016.01.023
1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
less than 5 years of age in Bangladesh is nearly 40%, which is one of the highest in the world, and causes nearly 60% of deaths in children less than 5 years of age.4 In Bangladesh, as well as in other developing countries, prevalence of malnutrition is one of the major indicators of child health in children less than 5 years of age. The factors related to malnutrition may help in policy formulation for the governments in these countries. Hence, a comprehensive study to determine the relationship between health-related factors and various malnutrition statuses is in demand. Several studies used fixed-effect models such as binary logistic regression, generalized Poisson regression, and ordinal logistic regression models to identify the determinants of malnutrition in Bangladesh.4,22,23 In an another study, a 2-level binary logistic regression model with a random intercept was used to identify the factors related to the malnutrition status in Bangladesh, including any community-level variations in the data.24 However, no previous studies addressed the possibility of both community- and household-level effects on malnutrition. In this study, therefore, multilevel binary logistic regression with a random intercept was used to identify individual-, household-, and community-level factors related to child malnutrition status of children less than 5 years of age in Bangladesh, including any community and household variations on malnutrition.
Methods Data was extracted from the Bangladesh Demographic Health Survey (BDHS) 2011, a nationally representative cross-sectional study. In BDHS, data collection was implemented in 5 phases, starting on July 8 and ending on December 27, 2011. The BDHS 2011 was conducted by the National Institute of Population Research and Training under the Ministry of Health and Family Welfare, implemented by a Bangladeshi research organization “Mitra and Associate.” Technical support was provided by Inner City Fund International of Calverton, Maryland, and financial support was given by US Agency for International Development. ORC Macro Institutional Review Board approved the data collection procedure. The 2011 BDHS was also reviewed and approved by the National Research Ethics Committee of the Bangladesh Medical Research Council (Dhaka, Bangladesh). Informed consent was obtained from each participant prior to subject’s enrollment. The BDHS sample was drawn from Bangladeshi adults residing in noninstitutional dwellings. The survey was operated in 7 administrative regions (divisions): Southern region (Barisal), Southeastern region (Chittagong), Central region (Dhaka), Western region (Khulna), Midwestern region (Rajshahi), Northwestern region (Rangpur), and Eastern region (Sylhet). Enumeration areas from the 2011 census were used as the primary sampling units for the survey. The survey was based on multistage stratified sampling techniques of households. The detail sampling design and all other issues related to the BDHS were discussed elsewhere.21 2
Volume All children listed in the survey were born in January 2006 or later. These children were eligible for height and weight measurements. Logistic support was given by United Nations Children’s Fund. The implausibility was defined based on World Health Organization (WHO) 2006 standards flag limits of z-score: stunting: < 6 or >6; wasting: < 5 or >5; and underweight: < 6 or >5. A total of 8761 children less than 5 years of age in the BDHS sample households were eligible for anthropometric measurements. However, of these figures, only anthropometric and age data available for 7647 children were considered complete and credible (Figure; available at www.jpeds.com). The mean age of children was recorded as 30.38 months. Other indicators, such as, height, weight, and hemoglobin level had mean of 83.25 cm, 10.47 kg, and 10.74 g/dL, respectively. Outcome Measures and Operational Definitions The primary outcomes were stunting, wasting, and underweight. A child was considered stunted, wasted, and underweight, respectively, if the height-for-age, weight-for-height, and weight-for-age indices were less than 2 SDs below the respective median of the WHO reference population.21,25,52 Covariates Most of the covariates in this study were considered based on previous literature review.2,16,22,26 Covariates were classified into 3 level characteristics: individual-level, household-level, and community-level. Individual-level characteristics were age, sex, birth order, mothers’ body mass index (BMI), mothers’ and fathers’ education, religion, and food security. Five household food security indicators were selected using the Household Food Insecurity Access Scale.27 The technical working group of the BDHS 201125 systematically reviewed and modified the indicators to suit Bangladesh. The questions used were: (1) “In the past 12 months, did you have 3 square (‘full-stomach’) meals a day?”; (2) “In the past 12 months, did you have to skip entire meals because there was not enough food?”; (3) “In the past 12 months, did you have less food in a meal because there was not enough food?”; (4) “In the past 12 months, did you or any of your family members eat wheat or another grain in place of rice?”; and (5) “In the past 12 months, did you ask for food from relatives or neighbors to make a meal?” Each indicator had the following 4 response options: never, rarely (1-6 times in the past 12 months), sometimes (7-12 times in the past 12 months), and often (few times each month), which were coded consecutively ranging from 3-0 for first question and 0-3 other than first question. A household was classified food insecure when the family experienced any of the 5 conditions within the recall period (ie, if the answer to first question was “sometimes,” “often,” or “never” and any of the other 4 questions was “rarely,” “sometimes,” or “often”). A household that did not meet these conditions (ie, “0”) was classified as food secure. Afterward, the individual food frequency scores for all the 5 frequency responses were summed in a single food security score for each ever-married woman of the household. To facilitate analysis, a composite score Chowdhury et al
- 2016
ORIGINAL ARTICLES
ranging from a minimum of “0” to a maximum of “15,” which was later classified as 4 categories, food secure (0), mild food insecurity (1-5), moderate food insecurity (610), and severe food insecurity (11-15).25 Household socioeconomic status was considered as the household-level characteristics. Household socioeconomic status, namely the wealth index was constructed from data on household assets, including ownership of durable goods (such as televisions and bicycles) and dwelling characteristics (such as source of drinking water, sanitation facilities, and construction materials). Principal components analyses were used to assign individual household wealth scores. These weighted values were then summed and rescaled to range from 0-1, and each household was assigned into quintiles: the first quintile: poorest, the second quintile: poorer, the third quintile: middle class, the fourth quintile: richer and the fifth quintile: richest.25 Community-level characteristics included types of residence, region of residence, and community economic status. Socioeconomic status of community was estimated by averaging the household wealth index at the community level, and then tertiles were obtained to represent community status as poor, average, and rich.28 Statistical Analyses Descriptive statistics were presented as mean (SD) and percentages where appropriate and c2 test was used to evaluate the association between outcome and independent variables (covariates). Multilevel binary logistic regression model with a random intercept at community, household, and individual levels were used for investigating the relationships between individual-level, household-level, and community-level characteristics and each of the outcome variables namely stunting, wasting, and underweight. Stata v 11.2/SE (Stata Corp, College Station, Texas) was used for all statistical analysis.
Results Among the children, 51.19% were male, 69.33% were living in rural areas, and 65.09% had food security. Prevalence of Malnutrition Table I presents the prevalence of malnutrition (stunting, wasting, and underweight). The prevalence of malnutrition for children less than 5 years of age was higher among children of fourth and above birth order, children of underweight mother, children of uneducated mother, children of uneducated father, socioeconomically poorest children, children of poor community, Muslim children, the children of Eastern region, and moderate to severe food insecure children. Risk Factors of Malnutrition All multilevel logistic regression models presented in Table II were statistically significant (likelihood-ratio c2 = 41.49, P < .001 for stunting; likelihood-ratio c2 = 32.38, P < .001 for wasting; and likelihood-ratio c2 = 42.01, P < .001 for
underweight).28 The analysis results also showed that the variances of the community and household effects were significant for all the measures of malnourishment (stunting, wasting, and underweight) with a P-value of <.001. The analysis results of stunting presented in Table II shows that children’s age group, mother’s BMI, fathers’ educational status, place of residence, socioeconomic status, religion, and food security had significant impact on stunting. The risk of being stunted decreases as the age increases. Children of normal (OR 0.75, 95% CI 0.65-0.87) and overweight (OR 0.51, 95% CI 0.40-0.64) mothers had lower odds of being stunted than underweight mothers. As father’s education level decreases, the odds of stunting increases. Children with lower socioeconomic and community status, living in urban areas, belonging to Muslim, and food insecure families had higher odds of being stunted. Children from Midwestern region had significantly higher odds of stunting than Southern region. The results of wasting shows that children of normal mothers had 36% less odds of being wasted (OR 0.64, 95% CI 0.55-0.76), and the children of overweight mothers had 65% less odds of being wasted (OR 0.35, 95% CI 0.26-0.48) than that of the children of underweight mothers. The Eastern region had significantly higher odds of being wasted (OR 1.37, 95% CI 1.02-1.84) than Southern region. The risk of underweight increases with growing age. Female children had 17% higher odds of being underweight (OR 1.17, 95% CI 1.03-1.32) than male. Children of normal mothers had 45% less odds of being underweight (OR 0.55, 95% CI 0.47-0.65), and the children of overweight mothers had 66% less odds of being underweight (OR 0.34, 95% CI 0.26-0.43). Furthermore, mild (OR 1.28, 95% CI 1.10-1.49) and moderate to severe (OR 1.31, 95% CI 1.04-1.66) food insecure children as well as children of parents with a poor educational status (OR 1.82, 95% CI 1.22-2.72) had a higher odds of being underweight. Compared with Muslim, NonMuslim children (OR 0.80, 95% CI 0.63-1.01) had 20% lower odds of being underweight. The children of Eastern region had 40% higher odds of being underweight (OR 1.40, 95% CI 1.10-1.79) than that of Southern region.
Discussion The prevalence of stunting, wasting, and underweight in the nationally representative Bangladeshi sample were 41.3%, 15.5%, and 36.2%, respectively. This showed that Bangladesh is among the highest in the world, and the respective prevalence indicates that the situation is critical in emergence based on WHO-recommended thresholds.29,30 According to latest Demographic Health Survey reports, competitive figures have been observed in some South Asian countries; stunting, wasting, and underweight were, respectively, 48%, 19.8%, and 42.5% in India; 44.8%, 10.8%, and 30% in Pakistan; and 40.5%, 10.9%, and 28.8% in Nepal.31 Demographic, economic, geographic, environmental, and cultural similarities may resemble the malnutrition state of children
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
3
THE JOURNAL OF PEDIATRICS
Volume -
www.jpeds.com
Table I. Prevalence of stunting, wasting, and underweight of children under 5 years of age (N = 7568) Stunting Characteristics Age group (mo) 0-11 12-23 24-35 36-47 48-59 Sex of child Male Female Birth order of child First Second Third Fourth and above Mother’s BMI Normal Underweight Overweight Mother’s education No education Primary Secondary Higher Father’s education No education Primary Secondary Higher Place of residence Urban Rural Socioeconomic status Poorest Poorer Middle Richer Richest Community status Poor Average Rich Religion Muslim Non-Muslim Region of residence Southern Southeastern Central Western Midwestern Northwestern Eastern Food Security Food secure Mild food insecure Moderate to severe food insecure Overall
Wasting
N (%)
P value
N (%)
302 (20.4%) 698 (49.4%) 665 (47.6%) 768 (47.2%) 649 (41.5%)
<.001
204 (14.7%) 228 (15.8%) 212 (14.5%) 265 (15.9%) 263 (16.5%)
1567 (40.6%) 1515 (42.0%)
.293
984 (37.8%) 864 (39.7%) 540 (41.3%) 694 (50.2%)
N (%)
P value
.587
292 (19.8%) 505 (36.1%) 570 (39.7%) 695 (43.1%) 650 (41.4%)
<.001
617 (15.9%) 555 (15.2%)
.426
1327 (34.2%) 1385 (38.4%)
.001
<.001
390 (14.7%) 337 (14.9%) 191 (14.7%) 254 (18.8%)
.012
847 (32.5%) 741 (33.4%) 485 (37.1%) 639 (46.6%)
<.001
1774 (39.6%) 1069 (51.2%) 239 (25.8%)
<.001
653 (14.3%) 439 (21.0%) 80 (8.7%)
<.001
1494 (33.0%) 1032 (50.0%) 186 (19.8%)
<.001
753 (50.9%) 1070 (47.1%) 1136 (35.7%) 123 (21.8%)
<.001
259 (17.9%) 407 (18.0%) 441 (13.5%) 65 (9.9%)
<.001
699 (49.0%) 967 (42.3%) 947 (29.0%) 99 (16.9%)
<.001
1075 (49.6%) 1010 (46.3%) 765 (34.9%) 232 (24.5%)
<.001
361 (17.0%) 376 (17.0%) 314 (14.7%) 121 (10.5%)
<.001
971 (45.9%) 874 (39.7%) 664 (30.0%) 203 (19.7%)
<.001
821 (36.5%) 2261 (42.6%)
.001
329 (13.9%) 843 (16.0%)
.083
678 (27.9%) 2034 (38.6%)
<.001
915 (53.5%) 692 (46.1%) 589 (40.2%) 520 (36.2%) 366 (25.8%)
<.001
298 (17.9%) 237 (15.8%) 252 (17.6%) 200 (13.5%) 185 (11.9%)
<.001
837 (50.3%) 624 (42.0%) 531 (35.5%) 411 (28.0%) 309 (20.3%)
<.001
1156 (50.2%) 999 (39.9%) 927 (33.7%)
<.001
411 (18.1%) 364 (15.4%) 397 (13.1%)
.001
1053 (46.9%) 879 (35.1%) 780 (26.8%)
<.001
2812 (41.8%) 270 (35.7%)
.015
1076 (15.8%) 96 (12.1%)
.010
2479 (36.8%) 233 (30.4%)
.011
342 (44.1%) 619 (41.0%) 544 (43.6%) 298 (33.6%) 292 (33.8%) 417 (42.9%) 570 (49.9%)
<.001
120 (14.8%) 232 (15.8%) 195 (15.5%) 129 (14.7%) 150 (16.2%) 127 (13.2%) 219 (18.3%)
.460
310 (39.3%) 551 (37.3%) 451 (36.5%) 253 (28.8%) 288 (34.1%) 331 (34.5%) 528 (45.4%)
.001
1760 (37.0%) 893 (47.0%) 429 (53.9%) 3082 (41.3%)
<.001
717 (14.5%) 308 (16.5%) 147 (19.3%) 1172 (15.5%)
.007
1521 (31.4%) 802 (42.5%) 389 (51.3%) 2712 (36.2%)
<.001
among these countries. The prevalence malnutrition in children less than 5 years of age is also higher in Sub-Saharan Africa.32 Many African countries, such as Burundi, represent the African region with its highest prevalence of chronic malnutrition (57.7%)31 and contribute to the deaths of 3.1 million children less than 5 years of age every year.33 4
Underweight P value
Our study showed that the prevalence of malnutrition is significantly higher among the age groups of 12-23 months, 36-47 months, and 48-59 months and lower in the age group 0-11 months, which is consistent with previous studies.2,3,16 However, the prevalence of malnutrition has been recorded in younger children in some South Asian and African Chowdhury et al
- 2016
ORIGINAL ARTICLES
Table II. Logistic regression analysis for child malnutrition and individual-, household-, and community-level factors Characteristics Age group (mo) 0-11 (ref) 12-23 24-35 36-47 48-59 Sex of child Male (ref) Female Birth order of child First (ref) Second Third Fourth and above Mother’s BMI Underweight (ref) Normal Overweight Mother’s education No education Primary Secondary Higher (ref) Father’s education No education Primary Secondary Higher (ref) Place of residence Urban (ref) Rural Socioeconomic status Poorest Poorer Middle Richer Richest (ref) Community status Poor Average Rich (ref) Religion Muslim (ref) Non-Muslim Region of residence Southern (ref) Southeastern Central Western Midwestern Northwestern Eastern Food Security Food secure (ref) Mild food insecure Moderate to severe food insecure Random-effect variance Level 2 (household) Level 3 (community)
aORs (95% CI) for stunting
P value
aORs (95% CI) for wasted
P value
aORs (95% CI) for underweight
P value
1.00 5.17 (4.09-6.53) 4.65 (3.70-5.85) 4.35 (3.51-5.39) 3.30 (2.66-4.10)
<.001 <.001 <.001 <.001
1.00 1.16 (0.92-1.47) 1.07 (0.84-1.35) 1.20 (0.95-1.52) 1.25 (0.99-1.58)
.211 .600 .127 .064
1.00 2.70 (2.15-3.38) 3.56 (2.83-4.47) 3.81 (3.07-4.74) 3.65 (2.93-4.54)
<.001 <.001 <.001 <.001
1.00 1.00 (0.89-1.12)
.950
1.00 0.92 (0.80-1.05)
.213
1.00 1.17 (1.03-1.32)
.015
1.00 1.03 (0.90-1.19) 1.02 (0.86-1.20) 1.04 (0.87-1.24)
.641 .848 .663
1.00 1.03 (0.85-1.25) 0.95 (0.77-1.18) 1.09 (0.86-1.37)
.737 .645 .474
1.00 1.02 (0.87-1.19) 1.07 (0.90-1.28) 1.09 (0.90-1.33)
.83 .42 .37
1.00 0.75 (0.65-0.87) 0.51 (0.40-0.64)
<.001 <.001
1.00 0.64 (0.55-0.76) 0.35 (0.26-0.48)
<.001 <.001
1.00 0.55 (0.47-0.65) 0.34 (0.26-0.43)
<.001 <.001
1.41 (0.99-2.01) 1.38 (0.99-1.92) 1.29 (0.96-1.73) 1.00
.057 .060 .091
1.32 (0.87-2.00) 1.36 (0.93-2.01) 1.09 (0.76-1.57) 1.00
.198 .117 .625
1.82 (1.22-2.72) 1.68 (1.16-2.43) 1.37 (0.98-1.93) 1.00
.003 .006 .068
1.73 (1.31-2.28) 1.78 (1.37-2.30) 1.44 (1.14-1.83) 1.00
<.001 <.001 .002
0.94 (0.68-1.32) 1.12 (0.82-1.52) 1.01 (0.75-1.37) 1.00
.740 .484 .940
1.28 (0.96-1.71) 1.36 (1.04-1.76) 1.22 (0.96-1.56) 1.00
.091 .022 .106
1.00 1.24 (1.03-1.49)
.021
1.00 1.08 (0.85-1.38)
.508
1.00 1.14 (0.96-1.36)
.137
2.50 (1.88-3.31) 2.04 (1.57-2.65) 1.75 (1.37-2.25) 1.50 (1.21-1.86) 1.00
<.001 <.001 <.001 <.001
1.02 (0.71-1.46) 1.01 (0.72-1.41) 1.22 (0.89-1.68) 0.96 (0.72-1.28) 1.00
.911 .965 .223 .776
2.22 (1.63-3.02) 1.90 (1.41-2.54) 1.66 (1.27-2.18) 1.18 (0.92-1.49) 1.00
<.001 <.001 <.001 .187
1.31 (1.05-1.64) 1.14 (0.95-1.37) 1.00
.018 .149
1.18 (0.90-1.56) 0.98 (0.76-1.26) 1.00
.235 .846
1.34 (1.07-1.67) 1.13 (0.94-1.37) 1.00
.009 .189
1.00 0.79 (0.63-0.97)
.026
1.00 0.81 (0.62-1.04)
.098
1.00 0.80 (0.63-1.01)
.065
1.00 1.19 (0.92-1.53) 1.22 (0.95-1.57) 0.83 (0.64-1.07) 0.63 (0.48-0.82) 0.96 (0.75-1.24) 1.40 (1.08-1.82)
.191 .116 .145 .001 .782 .012
1.00 1.18 (0.88-1.58) 1.13 (0.85-1.52) 1.15 (0.81-1.61) 1.27 (0.90-1.78) 0.87 (0.63-1.22) 1.37 (1.02-1.84)
.277 .404 .436 .168 .428 .037
1.00 1.17 (0.92-1.48) 0.99 (0.77-1.26) 0.78 (0.60-1.03) 0.77 (0.59-0.99) 0.74 (0.57-0.95) 1.40 (1.10-1.79)
.192 .918 .076 .043 .017 .006
1.00 1.29 (1.12-1.48) 1.38 (1.11-1.73)
<.001 .004
1.00 1.02 (0.85-1.21) 1.11 (0.85-1.45)
.864 .449
1.00 1.28 (1.10-1.49) 1.31 (1.04-1.66)
.002 .024
0.78 (0.22*) 0.12 (0.04*)
0.78 (0.28*) 0.19 (0.06*)
1.04 (0.26*) 0.06 (0.04*)
ref, reference. aORs with 95% CI were reported from a multilevel logistic regression model accounting for intercept at household and community levels. *Denotes the SE of random intercept and it measures the variability of average effect in each level (community and household) to experience malnutrition. The P value for each random effect variance is <.001.
countries.3,34 The prevalence of malnutrition was also recorded significantly higher among female children, children of fourth and above birth order, children of underweight mother, children of uneducated parents, children in rural
setting, and children with poor socioeconomic status. Similar results were obtained in other studies as well.12,16,26,32 Nevertheless, in Sri Lanka, malnutrition was more prevalent among boys35 followed by some other developing countries, for
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
5
THE JOURNAL OF PEDIATRICS
www.jpeds.com
example, Nigeria and Libya.11,36 The majority of Bangladeshis are Muslims. Our study observed that children in Muslim families had significantly higher prevalence of child malnutrition compared with non-Muslim families. In India, children in Hindu families had higher prevalence of malnutrition.26 Similarly in Ethiopia, children from religiously orthodox families were found to have higher prevalence of malnutrition.2 The prevalence of malnutrition in Bangladesh, India, and Ethiopia is higher among the most common religion in these countries. This may be due to some cultural factors, which are represented by the dominated religion in these countries. Poor community status led to the higher prevalence of stunting, wasting, and underweight in Bangladesh. Severely food insecure children also led to the higher prevalence of stunting, wasting, and underweight. Similar results were observed in Ethiopia and Vietnam.37 The significant regional differences in terms of prevalence of malnutrition have also been observed, which is consistent with other studies.6,16 The multilevel binary logistic regression analysis showed that the risk of being malnourished among children increases with increasing age. This analysis also showed that female children, children of underweight mothers, and children of uneducated parents had a higher risk of being malnourished. This finding is similar to that of other studies.2,11,30,33,37-40 This study also showed that children of poor households were more likely to be malnourished than the children of wealthier households, which was consistent with the findings of other similar studies.14,16,38,41,42 The place of residence and religion were also found to be significant in this study. Similar results were also obtained in other studies conducted in some developing countries.43,44 In our study, regional variations have been observed. which were not reported in previous studies.6,16 Malnutrition may conceal important intraregional differences because of diverse cultural norms that might affect nutritional practice and needs more investigation. Food insecurity, in this study, was found as another risk factor for child malnutrition. Despite ensuring significant progress in boosting up domestic food products, food insecurity remains a critical concern in Bangladesh. Thirtyone percent of the population lives below the national poverty line, most of them women and children.45 Soaring inequalities in income and consumption rates may agitate the present nutritional state of children in Bangladesh.45 Access of children to nutritional practice is being hampered by various causes, such as poverty, lack of parental education, especially maternal education, inequalities in urban-rural place of residence, disparity between male and female children, underutilization of health service, poor quality care, and environment.13,32,42,46-50 The results of this study also showed that the variances of the community and household effects were significant for each of stunting, wasting, and underweight. These indicated that children from different communities and households with the same levels of variables will show different influences on all the measures of malnourishment (stunting, wasting, and underweight). Hence, community- and household6
Volume level factors are required to address policy interventions for child malnutrition in Bangladesh. Policy planners can achieve more success in all measures of malnourishment by addressing community and household specific variables. We recommend that concerted efforts are essential in making strong collaborations with government, nongovernment, social, cultural, and religious institutions to strengthen antidiscrimination campaigns for female children, especially in rural areas. Malnutrition can be reduced by undertaking measures to strengthen actions concerning social determinants of health and specific health programs. An integrated nutrition-sensitive social protection system is to be recommended under the Bangladesh’s national protection structure “National Social Security Strategy” to support the most vulnerable citizens in order to improve the nutritional status in Bangladesh.51 The existing Bangladesh National Nutrition Policy (2013) needs to be further coordinated and strengthened to address the problem of malnutrition.52 Health extension programs, which support and educate mothers, need to be developed to improve upon current feeding practices to children. Adopting family planning methods should be encouraged at a community level. Ready-to-use foods from locally available food ingredients can also be one of the interventions used in reducing child malnutrition. Our study has several strengths and limitations. The main strength of this study was that the data came from a large nationally representative survey carried out in 2011. The survey covered both urban and rural areas. In addition, our study findings provide detailed information on risk factors of malnutrition among children less than 5 years of age. However, the cross-sectional nature of the study means that it was not possible to establish a causal relationship between risk factors and outcomes of malnutrition. Because of the unavailability of recent data, BDHS 2011 data was used in this study, which does not present the current nutritional status. Being overweight has not been included in this study as an indicator of malnutrition because of low prevalence. The implausibility of stunting, wasting, and underweight was defined based on most inclusive standards flag limits of WHO, which may result in highest reported prevalence. Another limitation involves information bias, which may result from collecting information about self-reporting age, education, occupation, and household assets etc. The prevalence of malnutrition in children less than 5 years of age in Bangladesh is one of the highest in the world. Several individual- (eg, age, sex, birth order, mothers’ BMI, mothers’ and fathers’ education, religion, and food security), household (eg, socioeconomic status), and community- (eg, types of residence, region of residence, and community economic status) level variables have been identified as risk factors of child malnutrition in Bangladesh. Therefore, current policies toward reducing child malnutrition need to address these factors. Also, there is community- and household-level variations in the data. Thus, community- and household-level strategies may require addressing the problem of malnutrition in Chowdhury et al
- 2016
ORIGINAL ARTICLES
Bangladesh. The findings of this study may also be relevant to other countries where child malnutrition is common. Further, this could be a useful guideline for clinicians in these countries who professionally assess child malnutrition. Finally, in Bangladesh, as well as in the rest of the developing world, further investigations (eg, prospective cohort studies) on malnutrition are needed. Such research will inspire ingenuity in developing effective strategies to improve the nutritional status of children less than 5 years of age. n Submitted for publication Sep 1, 2015; last revision received Dec 15, 2015; accepted Jan 7, 2016. Reprint requests: Baki Billah, PhD, Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia. E-mail: [email protected]. edu.au
References 1. UNICEF. The infection-undernutrition cycle. Learning objectives, Page: 1-28. http://www.unicef.org/nutrition/training/2.5/1.html. Accessed October 24, 2015. 2. Mengistu K, Alemu K, Destaw B. Prevalence of malnutrition and associated factors among children aged 6-59 months at Hidabu Abote district, North Shewa, Oromia regional state. J Nutr Disord Ther 2013;T1:001. 3. Olack B, Burke H, Cosmas L, Bamrah S, Dooling K, Feikin DR, et al. Nutritional status of under-five children living in an informal urban settlement in Nairobi, Kenya. J Health Popul Nutr 2011;29:357-63. 4. Islam MM, Alam M, Tariquzaman M, Kabir MA, Pervin R, Begum M, et al. Predictors of the number of under-five malnourished children in Bangladesh: application of the generalized Poisson regression model. BMC Public Health 2013;13:11. 5. de Onis M, Onyango AW, Borghi E, Garza C, Yang H. Comparison of the World Health Organization (WHO) Child Growth Standards and the National Center for Health Statistics/WHO international growth reference: implications for child health programmes. Public Health Nutr 2006;9:942-7. 6. Zere E, McIntyre D. Inequities in under-five child malnutrition in South Africa. Int J Equity Health 2003;2:7. 7. Improving Child Nutrition. The achievable imperative for global progress; UNICEF, United Nations Plaza, New York, NY 10017 USA, ISBN 978-92-806-4686-3; 2013. United Nations Publications Sales No.: E.13.XX.4. 8. UNICEF. Tracking Progress on Child and Maternal Nutritional survival and development priority, http://www.unicef.org/publications/files/ Tracking_Progress_on_Child_and_Maternal_Nutrition_EN_110309.pdf; 2010. Accessed January 17, 2015. 9. A life free from hunger: “Tackling Child Malnutrition” report by Save the Children International, Geneva. February 2012. https://everyone. savethechildren.net/sites/everyone.savethechildren.net/files/library/A-LifeFree-From-Hunger-Summary-Africa1.pdf. Accessed January 17, 2015. 10. Franz J. Improving nutrition: Early childhood development and mobile health. Impact and Innovation Series. The National Bureau of Asian Research; 2012, http://www.pacifichealthsummit.org/downloads/ Impact%20&%20Innovations/Improving_Nutrition.PDF. Accessed January 17, 2015. 11. Babatunde RO, Olagunju FI, Fakayode SB, Sola-Ojo FE. Prevalence and determinants of malnutrition among under-five children of farming households in Kwara State, Nigeria. J Agric Sci 2011;3:173-81. 12. Psaki S, Bhutta ZA, Ahmed T, Ahmed S, Bessong P, Islam M, et al. Household food access and child malnutrition: results from the eightcountry MAL-ED study. Popul Health Metr 2012;10:24. 13. Sah N. Determinants of child malnutrition in Nepal: a case analysis from Dhanusha, Central Terai of Nepal. J Nepal Health Res Council 2004;2: 50-4.
14. Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis M, Ezzati M, et al. Maternal and child undernutrition: global and regional exposures and health consequences. Lancet 2008;371:243-60. 15. Demissie S, Worku A. Magnitude and factors associated with malnutrition in children 6-59 months of age in pastoral community of Dollo Ado district, Somali region, Ethiopia. Sci J Public Health 2013;1:175-83. 16. Kandala NB, Madungu TP, Emina JBO, Nzita KPD, Cappuccio FP. Malnutrition among children under the age of five in the Democratic Republic of Congo (DRC): does geographic location matter? BMC Public Health 2011;11:261. 17. Anderson AK, Bignell W, Winful S, Soyiri I, Asiedu MS. Risk factors for malnutrition among children 5-years and younger in the AkuapimNorth district in the eastern region of Ghana. Curr Res J Biol Sci 2010; 2:183-8. 18. Khan MMH, Kr€amer A, Khandoker A, Pr€ ufer-Kr€amer L, Islamc A. Trends in socio-demographic and health-related indicators in Bangladesh, 1993-2007: will inequities persist? Bull World Health Organ 2011;89:583-93. 19. Chowdhury AMR. Rethinking interventions for women’s health. Lancet 2007;370:1292-3. 20. de Onis M, Brown D, Bl€ ossner M, Borghi E. Levels and trends in child malnutrition: UNICEF-WHO-The World Bank joint child malnutrition estimates, UNICEF, New York; WHO, Geneva; The World Bank, Washington, DC; 2012, WHO Library Cataloguing-in-Publication Data; ISBN 978 92 4 1504515; 2012. 21. Islam A, Biswas T. Chronic stunting among under-5 children in Bangladesh: a situation analysis. Adv Pediatr Res 2015;2:18. 22. Siddiqi MNA, Haque MN, Goni MA. Malnutrition of under-five children: evidence from Bangladesh. Asian J Med Sci 2011;2:113-9. 23. Das S, Rahman RM. Application of ordinal logistic regression analysis in determining risk factors of child malnutrition in Bangladesh. Nutr J 2011;10:124. 24. Alom J, Quddus MA, Islam MA. Nutritional status of under-five children in Bangladesh: a multilevel analysis. J Biosoc Sci 2012;44:525-35. 25. Bangladesh Demographic and Health Survey, 2011. Dhaka, Bangladesh and Calverton, Maryland, USA: National Institute of Population Research and Training (NIPORT) Mitra and Associates and Macro International; 2013, http://www.measuredhs.com/publications/publicationFR265-DHS-Final-Reports.cfm. Accessed December 25, 2014. 26. Das S, Sahoo H. An investigation into factors affecting child undernutrition in Madhya Pradesh. Anthropologist 2011;13:227-33. 27. Coates J, Swindale A, Bilinsky P. Household Food Insecurity Scale (HFIAS) for measurement of food access: Indicator guide. Food and Nutrition Technical Assistance (FANTA). Version 3. Washington, DC: Academy of Educational Development; 2007, http://www.fantaproject. org/downloads/pdfs/HFIAS_v3_Aug07.pdf. Accessed October 24, 2015. 28. Rahman MM, Gilmour S, Akter S, Abe SK, Saito E, Shibuya K. Prevalence and control of hypertension in Bangladesh: a multilevel analysis of a nationwide population-based survey. J Hypertens 2015;33:465-72. 29. de Onis M, Bl€ ossner M. WHO Global Database on Child Growth and Malnutrition. Geneva: World Health Organization; 1997, whqlibdoc. who.int/hq/1997/WHO_NUT_97.4.pdf. Accessed January 23, 2015. 30. Young H, Jaspars S. The meaning and measurement of acute malnutrition in emergencies. A primer for decision-makers. Network paper. Humanitarian Practice Network (HPN). United Kingdom: Overseas Development Institute; November 2006. Number 56. 31. ICF International, 2012. The DHS Program STAT compiler. http://www. statcompiler.com. Accessed December 30, 2014. 32. Rusinga O, Moyo S. Determinants of child malnutrition in Changazi Ward in Chimanimani District, Zimbabwe. J Emerg Trends Educ Res Policy Stud 2012;3:187-92. 33. UNICEF. A strike against chronic malnutrition in Burundi, 2013. http:// www.unicef.org/infobycountry/burundi_69651.html. Accessed January 23, 2015. 34. Mashal T, Takano T, Nakamura K, Kizuki M, Hemat S, Watanabe M, et al. Factors associated with the health and nutritional status of children under 5 years of age in Afghanistan: family behavior related to women
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
7
THE JOURNAL OF PEDIATRICS
35.
36.
37.
38.
39.
40.
41.
42. 43.
8
Volume -
www.jpeds.com
and past experience of war-related hardships. BMC Public Health 2008; 8:301. Peiris TDR, Wijesinghe DGNG. Nutritional status of under 5 yearold children and its relationship with maternal nutrition knowledge in Weeraketiya DS division of Sri Lanka. Trop Agr Res 2010;21: 330-9. Adel ET, Marie-Franc¸oise RC, Salaheddin MM, Najeeb E, Ahmed AM, Ibrahim B, et al. Nutritional status of under-five children in Libya: a national population-based survey. Libyan J Med 2008;3:13-9. Ali D, Saha KK, Nguyen PH, Diressie MT, Ruel MT, Menon P, et al. Household food insecurity is associated with higher child undernutrition in Bangladesh, Ethiopia, and Vietnam, but the effect is not mediated by child dietary diversity. J Nutr 2013;43:1-7. Fuchs C, Sultana T, Ahmed T, Iqbal Hossain M. Factors associated with acute malnutrition among children admitted to a diarrhea treatment facility in Bangladesh. Int J Pediatr 2014;2014:267806. Ruwali D. Nutritional status of children under five years of age and factors associated in Padampur VDC, Chitwan. Health Prospect 2011;10: 14-8. Rahman A, Chowdhury S, Hossain D. Acute malnutrition in Bangladeshi children: levels and determinants. https://eprints.usq.edu.au/4332/1/ Rahman_Chowdhury_Hossain.pdf. Accessed January 25, 2015. Girmay M, Charlotte H, Michael D, Atalay A, Fikru T, Bogale W, et al. Prevalence and predictors of under nutrition among infants aged six and twelve months in Butajira, Ethiopia. BMC Public Health 2010;10:27. Deolalikar AB. Poverty and child malnutrition in Bangladesh. J Dev Soc 2005;21:55-9. Wamani H, Tylleskar T, Astrom AN, Tumwine JK, Peterson S. Mothers’ education but not fathers’ education, household assets or land owner-
44.
45. 46.
47.
48. 49.
50.
51.
52.
ship is the best predictor of child health inequalities in rural Uganda. Int J Equity Health 2004;3:9. Hien NN, Kam S. Nutritional status and the characteristics related to malnutrition in children under five years of age in Nghean, Vietnam. J Prev Med Public Health 2008;41:232-40. Fotso JC. Urban–rural differentials in child malnutrition: trends and socioeconomic correlates in sub-Saharan Africa. Health Place 2007;13:205-23. UNICEF. Child malnutrition and household food insecurity remain major concerns for Bangladesh 2009. http://www.unicef.org/media/media_ 48981.html. Accessed January 25, 2015. Setboonsarng S. ADB Institute Discussion Paper No. 21. Child malnutrition as a poverty indicator: An evaluation in the context of different development interventions in Indonesia; the paper draws part of the Indonesian experience from an unpublished ADB Study on Indonesia in 2002. January 2005. Das S, Hossain MZ, Islam MA. Predictors of child chronic malnutrition in Bangladesh. Proc Pakistan Acad Sci 2008;45:137-55. Haque SE, Rahman M, Mostofa MG, Zahan MS. Reproductive health care utilization among young mothers in Bangladesh: does autonomy matter? Womens Health Issues 2012;22:e171-80. Choudhury KK, Hanifi MA, Rasheed S, Bhuiya A. Gender inequality and severe malnutrition among children in a remote rural area of Bangladesh. J Health Popul Nutr 2000;18:123-30. Henry FJ, Briend A, Fauveau V, Huttly SA, Yunus M, Chakraborty J. Gender and age differentials in risk factors for childhood malnutrition in Bangladesh. Ann Epidemiol 1993;3:382-6. Save the Children. Malnutrition in Bangladesh. Harnessing social protection for the most vulnerable, https://www.savethechildren.org.uk/ sites/default/files/images/Malnutrition_in_Bangladesh.pdf. Accessed October 24, 2015.
Chowdhury et al
- 2016
ORIGINAL ARTICLES
A total of 8550 children under 5 years of age were eligible for anthropometric measurements
Height and weight data are missing for 115
8435 Children have complete and credible anthropometric data
Age data are missing for 790 children
7647 Children have complete and credible anthropometric and age data
BMI are absence for mothers of 48
7599 Children have complete and credible anthropometric and age data
25 missing data were recorded for other variables
children
children
7568 Children deemed eligible for the present study (Final data)
Figure. Sample selection.
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
8.e1
M
alnutrition refers to inadequate dietary intake, infectious disease, or a combination of both.1-3 Three common indices of malnutrition for children are stunting (low height-for-age), wasting (low weight-for-height), and underweight (low weight-for-age).4-6 About 165 million or 26% of the world’s children less than 5 years of age are stunted, which slows down cognitive and physical development.7 Over 52 million children or around 8% of the world’s children less than 5 years of age suffer from wasting because of acute malnutrition, which noticeably increased the risk of death.6 Similarly, an estimated 101 million or 16% of world’s children less than 5 years of age are underweight.7 Malnutrition can exacerbate the impact of disease and nearly one-half of all child deaths globally are attributable to this cause.2 Children suffering malnutrition are more likely to die from common childhood illnesses such as diarrhea, pneumonia, malaria, measles, and AIDS.8 The primary causes of malnutrition include a lack of quality food, poor infant and child feeding and care practices such as suboptimal breastfeeding, deficiencies of micronutrients such as vitamin A or zinc, and recurrent attack of infections, often intensified by intestinal parasites.9,10 Child malnutrition also draws parallel links with demographic aspects,2,11-13 environmental aspects,2,14 socioeconomic aspects,6,12,15,16 parental characteristics,2,11 household possession,17 and geographical location.11 Bangladesh has made significant progress in the health and human development sectors since its independence in 1971.18,19 In Bangladesh, the child mortality per 1000 live births declined from 144 in 1990 to 41 in 2012, with an annual rate of reduction of 5.5%. Bangladesh has already achieved the MillenFrom the Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi, nium Development Goal 4 and proven its achievement to be more impressive Bangladesh; Department of Public Health, Faculty of Health Science, First Capital University of Bangladesh, than other South Asian countries, particularly, India, Pakistan, and Chuadanga, Bangladesh; Department of Global Health Afghanistan.20 Unfortunately, Bangladesh was not very successful in addressing Policy, School of International Health, The University of Tokyo, Tokyo, Japan; Department of Public Health the problems of child malnutrition. Severe malnutrition increased from 16% in Medicine, School of Public Health, Bielefeld University, Bielefeld, Germany; International Health Section, 2011 to 18% in 2013 which may be due to increasing rate of malnutrition among Division of Public Health, Graduate School of Medicine, Tokyo Medical and Dental University, Tokyo, Japan; and mothers, poor socioeconomic status of parents, relatively low rate of fully breastDepartment of Epidemiology and Preventive Medicine, 21 feeding babies, and food insecurity. The prevalence malnutrition in children School of Public Health and Preventive Medicine, 1
2
3
4
5
6
Monash University, Melbourne, Australia *Current affiliation: Department of Public Health, College of Applied Medical Sciences, King Faisal University, Al-Hassa, Saudi Arabia.
BDHS BMI WHO
Bangladesh Demographic Health Survey Body mass index World Health Organization
The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2016.01.023
1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
less than 5 years of age in Bangladesh is nearly 40%, which is one of the highest in the world, and causes nearly 60% of deaths in children less than 5 years of age.4 In Bangladesh, as well as in other developing countries, prevalence of malnutrition is one of the major indicators of child health in children less than 5 years of age. The factors related to malnutrition may help in policy formulation for the governments in these countries. Hence, a comprehensive study to determine the relationship between health-related factors and various malnutrition statuses is in demand. Several studies used fixed-effect models such as binary logistic regression, generalized Poisson regression, and ordinal logistic regression models to identify the determinants of malnutrition in Bangladesh.4,22,23 In an another study, a 2-level binary logistic regression model with a random intercept was used to identify the factors related to the malnutrition status in Bangladesh, including any community-level variations in the data.24 However, no previous studies addressed the possibility of both community- and household-level effects on malnutrition. In this study, therefore, multilevel binary logistic regression with a random intercept was used to identify individual-, household-, and community-level factors related to child malnutrition status of children less than 5 years of age in Bangladesh, including any community and household variations on malnutrition.
Methods Data was extracted from the Bangladesh Demographic Health Survey (BDHS) 2011, a nationally representative cross-sectional study. In BDHS, data collection was implemented in 5 phases, starting on July 8 and ending on December 27, 2011. The BDHS 2011 was conducted by the National Institute of Population Research and Training under the Ministry of Health and Family Welfare, implemented by a Bangladeshi research organization “Mitra and Associate.” Technical support was provided by Inner City Fund International of Calverton, Maryland, and financial support was given by US Agency for International Development. ORC Macro Institutional Review Board approved the data collection procedure. The 2011 BDHS was also reviewed and approved by the National Research Ethics Committee of the Bangladesh Medical Research Council (Dhaka, Bangladesh). Informed consent was obtained from each participant prior to subject’s enrollment. The BDHS sample was drawn from Bangladeshi adults residing in noninstitutional dwellings. The survey was operated in 7 administrative regions (divisions): Southern region (Barisal), Southeastern region (Chittagong), Central region (Dhaka), Western region (Khulna), Midwestern region (Rajshahi), Northwestern region (Rangpur), and Eastern region (Sylhet). Enumeration areas from the 2011 census were used as the primary sampling units for the survey. The survey was based on multistage stratified sampling techniques of households. The detail sampling design and all other issues related to the BDHS were discussed elsewhere.21 2
Volume All children listed in the survey were born in January 2006 or later. These children were eligible for height and weight measurements. Logistic support was given by United Nations Children’s Fund. The implausibility was defined based on World Health Organization (WHO) 2006 standards flag limits of z-score: stunting: < 6 or >6; wasting: < 5 or >5; and underweight: < 6 or >5. A total of 8761 children less than 5 years of age in the BDHS sample households were eligible for anthropometric measurements. However, of these figures, only anthropometric and age data available for 7647 children were considered complete and credible (Figure; available at www.jpeds.com). The mean age of children was recorded as 30.38 months. Other indicators, such as, height, weight, and hemoglobin level had mean of 83.25 cm, 10.47 kg, and 10.74 g/dL, respectively. Outcome Measures and Operational Definitions The primary outcomes were stunting, wasting, and underweight. A child was considered stunted, wasted, and underweight, respectively, if the height-for-age, weight-for-height, and weight-for-age indices were less than 2 SDs below the respective median of the WHO reference population.21,25,52 Covariates Most of the covariates in this study were considered based on previous literature review.2,16,22,26 Covariates were classified into 3 level characteristics: individual-level, household-level, and community-level. Individual-level characteristics were age, sex, birth order, mothers’ body mass index (BMI), mothers’ and fathers’ education, religion, and food security. Five household food security indicators were selected using the Household Food Insecurity Access Scale.27 The technical working group of the BDHS 201125 systematically reviewed and modified the indicators to suit Bangladesh. The questions used were: (1) “In the past 12 months, did you have 3 square (‘full-stomach’) meals a day?”; (2) “In the past 12 months, did you have to skip entire meals because there was not enough food?”; (3) “In the past 12 months, did you have less food in a meal because there was not enough food?”; (4) “In the past 12 months, did you or any of your family members eat wheat or another grain in place of rice?”; and (5) “In the past 12 months, did you ask for food from relatives or neighbors to make a meal?” Each indicator had the following 4 response options: never, rarely (1-6 times in the past 12 months), sometimes (7-12 times in the past 12 months), and often (few times each month), which were coded consecutively ranging from 3-0 for first question and 0-3 other than first question. A household was classified food insecure when the family experienced any of the 5 conditions within the recall period (ie, if the answer to first question was “sometimes,” “often,” or “never” and any of the other 4 questions was “rarely,” “sometimes,” or “often”). A household that did not meet these conditions (ie, “0”) was classified as food secure. Afterward, the individual food frequency scores for all the 5 frequency responses were summed in a single food security score for each ever-married woman of the household. To facilitate analysis, a composite score Chowdhury et al
- 2016
ORIGINAL ARTICLES
ranging from a minimum of “0” to a maximum of “15,” which was later classified as 4 categories, food secure (0), mild food insecurity (1-5), moderate food insecurity (610), and severe food insecurity (11-15).25 Household socioeconomic status was considered as the household-level characteristics. Household socioeconomic status, namely the wealth index was constructed from data on household assets, including ownership of durable goods (such as televisions and bicycles) and dwelling characteristics (such as source of drinking water, sanitation facilities, and construction materials). Principal components analyses were used to assign individual household wealth scores. These weighted values were then summed and rescaled to range from 0-1, and each household was assigned into quintiles: the first quintile: poorest, the second quintile: poorer, the third quintile: middle class, the fourth quintile: richer and the fifth quintile: richest.25 Community-level characteristics included types of residence, region of residence, and community economic status. Socioeconomic status of community was estimated by averaging the household wealth index at the community level, and then tertiles were obtained to represent community status as poor, average, and rich.28 Statistical Analyses Descriptive statistics were presented as mean (SD) and percentages where appropriate and c2 test was used to evaluate the association between outcome and independent variables (covariates). Multilevel binary logistic regression model with a random intercept at community, household, and individual levels were used for investigating the relationships between individual-level, household-level, and community-level characteristics and each of the outcome variables namely stunting, wasting, and underweight. Stata v 11.2/SE (Stata Corp, College Station, Texas) was used for all statistical analysis.
Results Among the children, 51.19% were male, 69.33% were living in rural areas, and 65.09% had food security. Prevalence of Malnutrition Table I presents the prevalence of malnutrition (stunting, wasting, and underweight). The prevalence of malnutrition for children less than 5 years of age was higher among children of fourth and above birth order, children of underweight mother, children of uneducated mother, children of uneducated father, socioeconomically poorest children, children of poor community, Muslim children, the children of Eastern region, and moderate to severe food insecure children. Risk Factors of Malnutrition All multilevel logistic regression models presented in Table II were statistically significant (likelihood-ratio c2 = 41.49, P < .001 for stunting; likelihood-ratio c2 = 32.38, P < .001 for wasting; and likelihood-ratio c2 = 42.01, P < .001 for
underweight).28 The analysis results also showed that the variances of the community and household effects were significant for all the measures of malnourishment (stunting, wasting, and underweight) with a P-value of <.001. The analysis results of stunting presented in Table II shows that children’s age group, mother’s BMI, fathers’ educational status, place of residence, socioeconomic status, religion, and food security had significant impact on stunting. The risk of being stunted decreases as the age increases. Children of normal (OR 0.75, 95% CI 0.65-0.87) and overweight (OR 0.51, 95% CI 0.40-0.64) mothers had lower odds of being stunted than underweight mothers. As father’s education level decreases, the odds of stunting increases. Children with lower socioeconomic and community status, living in urban areas, belonging to Muslim, and food insecure families had higher odds of being stunted. Children from Midwestern region had significantly higher odds of stunting than Southern region. The results of wasting shows that children of normal mothers had 36% less odds of being wasted (OR 0.64, 95% CI 0.55-0.76), and the children of overweight mothers had 65% less odds of being wasted (OR 0.35, 95% CI 0.26-0.48) than that of the children of underweight mothers. The Eastern region had significantly higher odds of being wasted (OR 1.37, 95% CI 1.02-1.84) than Southern region. The risk of underweight increases with growing age. Female children had 17% higher odds of being underweight (OR 1.17, 95% CI 1.03-1.32) than male. Children of normal mothers had 45% less odds of being underweight (OR 0.55, 95% CI 0.47-0.65), and the children of overweight mothers had 66% less odds of being underweight (OR 0.34, 95% CI 0.26-0.43). Furthermore, mild (OR 1.28, 95% CI 1.10-1.49) and moderate to severe (OR 1.31, 95% CI 1.04-1.66) food insecure children as well as children of parents with a poor educational status (OR 1.82, 95% CI 1.22-2.72) had a higher odds of being underweight. Compared with Muslim, NonMuslim children (OR 0.80, 95% CI 0.63-1.01) had 20% lower odds of being underweight. The children of Eastern region had 40% higher odds of being underweight (OR 1.40, 95% CI 1.10-1.79) than that of Southern region.
Discussion The prevalence of stunting, wasting, and underweight in the nationally representative Bangladeshi sample were 41.3%, 15.5%, and 36.2%, respectively. This showed that Bangladesh is among the highest in the world, and the respective prevalence indicates that the situation is critical in emergence based on WHO-recommended thresholds.29,30 According to latest Demographic Health Survey reports, competitive figures have been observed in some South Asian countries; stunting, wasting, and underweight were, respectively, 48%, 19.8%, and 42.5% in India; 44.8%, 10.8%, and 30% in Pakistan; and 40.5%, 10.9%, and 28.8% in Nepal.31 Demographic, economic, geographic, environmental, and cultural similarities may resemble the malnutrition state of children
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
3
THE JOURNAL OF PEDIATRICS
Volume -
www.jpeds.com
Table I. Prevalence of stunting, wasting, and underweight of children under 5 years of age (N = 7568) Stunting Characteristics Age group (mo) 0-11 12-23 24-35 36-47 48-59 Sex of child Male Female Birth order of child First Second Third Fourth and above Mother’s BMI Normal Underweight Overweight Mother’s education No education Primary Secondary Higher Father’s education No education Primary Secondary Higher Place of residence Urban Rural Socioeconomic status Poorest Poorer Middle Richer Richest Community status Poor Average Rich Religion Muslim Non-Muslim Region of residence Southern Southeastern Central Western Midwestern Northwestern Eastern Food Security Food secure Mild food insecure Moderate to severe food insecure Overall
Wasting
N (%)
P value
N (%)
302 (20.4%) 698 (49.4%) 665 (47.6%) 768 (47.2%) 649 (41.5%)
<.001
204 (14.7%) 228 (15.8%) 212 (14.5%) 265 (15.9%) 263 (16.5%)
1567 (40.6%) 1515 (42.0%)
.293
984 (37.8%) 864 (39.7%) 540 (41.3%) 694 (50.2%)
N (%)
P value
.587
292 (19.8%) 505 (36.1%) 570 (39.7%) 695 (43.1%) 650 (41.4%)
<.001
617 (15.9%) 555 (15.2%)
.426
1327 (34.2%) 1385 (38.4%)
.001
<.001
390 (14.7%) 337 (14.9%) 191 (14.7%) 254 (18.8%)
.012
847 (32.5%) 741 (33.4%) 485 (37.1%) 639 (46.6%)
<.001
1774 (39.6%) 1069 (51.2%) 239 (25.8%)
<.001
653 (14.3%) 439 (21.0%) 80 (8.7%)
<.001
1494 (33.0%) 1032 (50.0%) 186 (19.8%)
<.001
753 (50.9%) 1070 (47.1%) 1136 (35.7%) 123 (21.8%)
<.001
259 (17.9%) 407 (18.0%) 441 (13.5%) 65 (9.9%)
<.001
699 (49.0%) 967 (42.3%) 947 (29.0%) 99 (16.9%)
<.001
1075 (49.6%) 1010 (46.3%) 765 (34.9%) 232 (24.5%)
<.001
361 (17.0%) 376 (17.0%) 314 (14.7%) 121 (10.5%)
<.001
971 (45.9%) 874 (39.7%) 664 (30.0%) 203 (19.7%)
<.001
821 (36.5%) 2261 (42.6%)
.001
329 (13.9%) 843 (16.0%)
.083
678 (27.9%) 2034 (38.6%)
<.001
915 (53.5%) 692 (46.1%) 589 (40.2%) 520 (36.2%) 366 (25.8%)
<.001
298 (17.9%) 237 (15.8%) 252 (17.6%) 200 (13.5%) 185 (11.9%)
<.001
837 (50.3%) 624 (42.0%) 531 (35.5%) 411 (28.0%) 309 (20.3%)
<.001
1156 (50.2%) 999 (39.9%) 927 (33.7%)
<.001
411 (18.1%) 364 (15.4%) 397 (13.1%)
.001
1053 (46.9%) 879 (35.1%) 780 (26.8%)
<.001
2812 (41.8%) 270 (35.7%)
.015
1076 (15.8%) 96 (12.1%)
.010
2479 (36.8%) 233 (30.4%)
.011
342 (44.1%) 619 (41.0%) 544 (43.6%) 298 (33.6%) 292 (33.8%) 417 (42.9%) 570 (49.9%)
<.001
120 (14.8%) 232 (15.8%) 195 (15.5%) 129 (14.7%) 150 (16.2%) 127 (13.2%) 219 (18.3%)
.460
310 (39.3%) 551 (37.3%) 451 (36.5%) 253 (28.8%) 288 (34.1%) 331 (34.5%) 528 (45.4%)
.001
1760 (37.0%) 893 (47.0%) 429 (53.9%) 3082 (41.3%)
<.001
717 (14.5%) 308 (16.5%) 147 (19.3%) 1172 (15.5%)
.007
1521 (31.4%) 802 (42.5%) 389 (51.3%) 2712 (36.2%)
<.001
among these countries. The prevalence malnutrition in children less than 5 years of age is also higher in Sub-Saharan Africa.32 Many African countries, such as Burundi, represent the African region with its highest prevalence of chronic malnutrition (57.7%)31 and contribute to the deaths of 3.1 million children less than 5 years of age every year.33 4
Underweight P value
Our study showed that the prevalence of malnutrition is significantly higher among the age groups of 12-23 months, 36-47 months, and 48-59 months and lower in the age group 0-11 months, which is consistent with previous studies.2,3,16 However, the prevalence of malnutrition has been recorded in younger children in some South Asian and African Chowdhury et al
- 2016
ORIGINAL ARTICLES
Table II. Logistic regression analysis for child malnutrition and individual-, household-, and community-level factors Characteristics Age group (mo) 0-11 (ref) 12-23 24-35 36-47 48-59 Sex of child Male (ref) Female Birth order of child First (ref) Second Third Fourth and above Mother’s BMI Underweight (ref) Normal Overweight Mother’s education No education Primary Secondary Higher (ref) Father’s education No education Primary Secondary Higher (ref) Place of residence Urban (ref) Rural Socioeconomic status Poorest Poorer Middle Richer Richest (ref) Community status Poor Average Rich (ref) Religion Muslim (ref) Non-Muslim Region of residence Southern (ref) Southeastern Central Western Midwestern Northwestern Eastern Food Security Food secure (ref) Mild food insecure Moderate to severe food insecure Random-effect variance Level 2 (household) Level 3 (community)
aORs (95% CI) for stunting
P value
aORs (95% CI) for wasted
P value
aORs (95% CI) for underweight
P value
1.00 5.17 (4.09-6.53) 4.65 (3.70-5.85) 4.35 (3.51-5.39) 3.30 (2.66-4.10)
<.001 <.001 <.001 <.001
1.00 1.16 (0.92-1.47) 1.07 (0.84-1.35) 1.20 (0.95-1.52) 1.25 (0.99-1.58)
.211 .600 .127 .064
1.00 2.70 (2.15-3.38) 3.56 (2.83-4.47) 3.81 (3.07-4.74) 3.65 (2.93-4.54)
<.001 <.001 <.001 <.001
1.00 1.00 (0.89-1.12)
.950
1.00 0.92 (0.80-1.05)
.213
1.00 1.17 (1.03-1.32)
.015
1.00 1.03 (0.90-1.19) 1.02 (0.86-1.20) 1.04 (0.87-1.24)
.641 .848 .663
1.00 1.03 (0.85-1.25) 0.95 (0.77-1.18) 1.09 (0.86-1.37)
.737 .645 .474
1.00 1.02 (0.87-1.19) 1.07 (0.90-1.28) 1.09 (0.90-1.33)
.83 .42 .37
1.00 0.75 (0.65-0.87) 0.51 (0.40-0.64)
<.001 <.001
1.00 0.64 (0.55-0.76) 0.35 (0.26-0.48)
<.001 <.001
1.00 0.55 (0.47-0.65) 0.34 (0.26-0.43)
<.001 <.001
1.41 (0.99-2.01) 1.38 (0.99-1.92) 1.29 (0.96-1.73) 1.00
.057 .060 .091
1.32 (0.87-2.00) 1.36 (0.93-2.01) 1.09 (0.76-1.57) 1.00
.198 .117 .625
1.82 (1.22-2.72) 1.68 (1.16-2.43) 1.37 (0.98-1.93) 1.00
.003 .006 .068
1.73 (1.31-2.28) 1.78 (1.37-2.30) 1.44 (1.14-1.83) 1.00
<.001 <.001 .002
0.94 (0.68-1.32) 1.12 (0.82-1.52) 1.01 (0.75-1.37) 1.00
.740 .484 .940
1.28 (0.96-1.71) 1.36 (1.04-1.76) 1.22 (0.96-1.56) 1.00
.091 .022 .106
1.00 1.24 (1.03-1.49)
.021
1.00 1.08 (0.85-1.38)
.508
1.00 1.14 (0.96-1.36)
.137
2.50 (1.88-3.31) 2.04 (1.57-2.65) 1.75 (1.37-2.25) 1.50 (1.21-1.86) 1.00
<.001 <.001 <.001 <.001
1.02 (0.71-1.46) 1.01 (0.72-1.41) 1.22 (0.89-1.68) 0.96 (0.72-1.28) 1.00
.911 .965 .223 .776
2.22 (1.63-3.02) 1.90 (1.41-2.54) 1.66 (1.27-2.18) 1.18 (0.92-1.49) 1.00
<.001 <.001 <.001 .187
1.31 (1.05-1.64) 1.14 (0.95-1.37) 1.00
.018 .149
1.18 (0.90-1.56) 0.98 (0.76-1.26) 1.00
.235 .846
1.34 (1.07-1.67) 1.13 (0.94-1.37) 1.00
.009 .189
1.00 0.79 (0.63-0.97)
.026
1.00 0.81 (0.62-1.04)
.098
1.00 0.80 (0.63-1.01)
.065
1.00 1.19 (0.92-1.53) 1.22 (0.95-1.57) 0.83 (0.64-1.07) 0.63 (0.48-0.82) 0.96 (0.75-1.24) 1.40 (1.08-1.82)
.191 .116 .145 .001 .782 .012
1.00 1.18 (0.88-1.58) 1.13 (0.85-1.52) 1.15 (0.81-1.61) 1.27 (0.90-1.78) 0.87 (0.63-1.22) 1.37 (1.02-1.84)
.277 .404 .436 .168 .428 .037
1.00 1.17 (0.92-1.48) 0.99 (0.77-1.26) 0.78 (0.60-1.03) 0.77 (0.59-0.99) 0.74 (0.57-0.95) 1.40 (1.10-1.79)
.192 .918 .076 .043 .017 .006
1.00 1.29 (1.12-1.48) 1.38 (1.11-1.73)
<.001 .004
1.00 1.02 (0.85-1.21) 1.11 (0.85-1.45)
.864 .449
1.00 1.28 (1.10-1.49) 1.31 (1.04-1.66)
.002 .024
0.78 (0.22*) 0.12 (0.04*)
0.78 (0.28*) 0.19 (0.06*)
1.04 (0.26*) 0.06 (0.04*)
ref, reference. aORs with 95% CI were reported from a multilevel logistic regression model accounting for intercept at household and community levels. *Denotes the SE of random intercept and it measures the variability of average effect in each level (community and household) to experience malnutrition. The P value for each random effect variance is <.001.
countries.3,34 The prevalence of malnutrition was also recorded significantly higher among female children, children of fourth and above birth order, children of underweight mother, children of uneducated parents, children in rural
setting, and children with poor socioeconomic status. Similar results were obtained in other studies as well.12,16,26,32 Nevertheless, in Sri Lanka, malnutrition was more prevalent among boys35 followed by some other developing countries, for
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
5
THE JOURNAL OF PEDIATRICS
www.jpeds.com
example, Nigeria and Libya.11,36 The majority of Bangladeshis are Muslims. Our study observed that children in Muslim families had significantly higher prevalence of child malnutrition compared with non-Muslim families. In India, children in Hindu families had higher prevalence of malnutrition.26 Similarly in Ethiopia, children from religiously orthodox families were found to have higher prevalence of malnutrition.2 The prevalence of malnutrition in Bangladesh, India, and Ethiopia is higher among the most common religion in these countries. This may be due to some cultural factors, which are represented by the dominated religion in these countries. Poor community status led to the higher prevalence of stunting, wasting, and underweight in Bangladesh. Severely food insecure children also led to the higher prevalence of stunting, wasting, and underweight. Similar results were observed in Ethiopia and Vietnam.37 The significant regional differences in terms of prevalence of malnutrition have also been observed, which is consistent with other studies.6,16 The multilevel binary logistic regression analysis showed that the risk of being malnourished among children increases with increasing age. This analysis also showed that female children, children of underweight mothers, and children of uneducated parents had a higher risk of being malnourished. This finding is similar to that of other studies.2,11,30,33,37-40 This study also showed that children of poor households were more likely to be malnourished than the children of wealthier households, which was consistent with the findings of other similar studies.14,16,38,41,42 The place of residence and religion were also found to be significant in this study. Similar results were also obtained in other studies conducted in some developing countries.43,44 In our study, regional variations have been observed. which were not reported in previous studies.6,16 Malnutrition may conceal important intraregional differences because of diverse cultural norms that might affect nutritional practice and needs more investigation. Food insecurity, in this study, was found as another risk factor for child malnutrition. Despite ensuring significant progress in boosting up domestic food products, food insecurity remains a critical concern in Bangladesh. Thirtyone percent of the population lives below the national poverty line, most of them women and children.45 Soaring inequalities in income and consumption rates may agitate the present nutritional state of children in Bangladesh.45 Access of children to nutritional practice is being hampered by various causes, such as poverty, lack of parental education, especially maternal education, inequalities in urban-rural place of residence, disparity between male and female children, underutilization of health service, poor quality care, and environment.13,32,42,46-50 The results of this study also showed that the variances of the community and household effects were significant for each of stunting, wasting, and underweight. These indicated that children from different communities and households with the same levels of variables will show different influences on all the measures of malnourishment (stunting, wasting, and underweight). Hence, community- and household6
Volume level factors are required to address policy interventions for child malnutrition in Bangladesh. Policy planners can achieve more success in all measures of malnourishment by addressing community and household specific variables. We recommend that concerted efforts are essential in making strong collaborations with government, nongovernment, social, cultural, and religious institutions to strengthen antidiscrimination campaigns for female children, especially in rural areas. Malnutrition can be reduced by undertaking measures to strengthen actions concerning social determinants of health and specific health programs. An integrated nutrition-sensitive social protection system is to be recommended under the Bangladesh’s national protection structure “National Social Security Strategy” to support the most vulnerable citizens in order to improve the nutritional status in Bangladesh.51 The existing Bangladesh National Nutrition Policy (2013) needs to be further coordinated and strengthened to address the problem of malnutrition.52 Health extension programs, which support and educate mothers, need to be developed to improve upon current feeding practices to children. Adopting family planning methods should be encouraged at a community level. Ready-to-use foods from locally available food ingredients can also be one of the interventions used in reducing child malnutrition. Our study has several strengths and limitations. The main strength of this study was that the data came from a large nationally representative survey carried out in 2011. The survey covered both urban and rural areas. In addition, our study findings provide detailed information on risk factors of malnutrition among children less than 5 years of age. However, the cross-sectional nature of the study means that it was not possible to establish a causal relationship between risk factors and outcomes of malnutrition. Because of the unavailability of recent data, BDHS 2011 data was used in this study, which does not present the current nutritional status. Being overweight has not been included in this study as an indicator of malnutrition because of low prevalence. The implausibility of stunting, wasting, and underweight was defined based on most inclusive standards flag limits of WHO, which may result in highest reported prevalence. Another limitation involves information bias, which may result from collecting information about self-reporting age, education, occupation, and household assets etc. The prevalence of malnutrition in children less than 5 years of age in Bangladesh is one of the highest in the world. Several individual- (eg, age, sex, birth order, mothers’ BMI, mothers’ and fathers’ education, religion, and food security), household (eg, socioeconomic status), and community- (eg, types of residence, region of residence, and community economic status) level variables have been identified as risk factors of child malnutrition in Bangladesh. Therefore, current policies toward reducing child malnutrition need to address these factors. Also, there is community- and household-level variations in the data. Thus, community- and household-level strategies may require addressing the problem of malnutrition in Chowdhury et al
- 2016
ORIGINAL ARTICLES
Bangladesh. The findings of this study may also be relevant to other countries where child malnutrition is common. Further, this could be a useful guideline for clinicians in these countries who professionally assess child malnutrition. Finally, in Bangladesh, as well as in the rest of the developing world, further investigations (eg, prospective cohort studies) on malnutrition are needed. Such research will inspire ingenuity in developing effective strategies to improve the nutritional status of children less than 5 years of age. n Submitted for publication Sep 1, 2015; last revision received Dec 15, 2015; accepted Jan 7, 2016. Reprint requests: Baki Billah, PhD, Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia. E-mail: [email protected]. edu.au
References 1. UNICEF. The infection-undernutrition cycle. Learning objectives, Page: 1-28. http://www.unicef.org/nutrition/training/2.5/1.html. Accessed October 24, 2015. 2. Mengistu K, Alemu K, Destaw B. Prevalence of malnutrition and associated factors among children aged 6-59 months at Hidabu Abote district, North Shewa, Oromia regional state. J Nutr Disord Ther 2013;T1:001. 3. Olack B, Burke H, Cosmas L, Bamrah S, Dooling K, Feikin DR, et al. Nutritional status of under-five children living in an informal urban settlement in Nairobi, Kenya. J Health Popul Nutr 2011;29:357-63. 4. Islam MM, Alam M, Tariquzaman M, Kabir MA, Pervin R, Begum M, et al. Predictors of the number of under-five malnourished children in Bangladesh: application of the generalized Poisson regression model. BMC Public Health 2013;13:11. 5. de Onis M, Onyango AW, Borghi E, Garza C, Yang H. Comparison of the World Health Organization (WHO) Child Growth Standards and the National Center for Health Statistics/WHO international growth reference: implications for child health programmes. Public Health Nutr 2006;9:942-7. 6. Zere E, McIntyre D. Inequities in under-five child malnutrition in South Africa. Int J Equity Health 2003;2:7. 7. Improving Child Nutrition. The achievable imperative for global progress; UNICEF, United Nations Plaza, New York, NY 10017 USA, ISBN 978-92-806-4686-3; 2013. United Nations Publications Sales No.: E.13.XX.4. 8. UNICEF. Tracking Progress on Child and Maternal Nutritional survival and development priority, http://www.unicef.org/publications/files/ Tracking_Progress_on_Child_and_Maternal_Nutrition_EN_110309.pdf; 2010. Accessed January 17, 2015. 9. A life free from hunger: “Tackling Child Malnutrition” report by Save the Children International, Geneva. February 2012. https://everyone. savethechildren.net/sites/everyone.savethechildren.net/files/library/A-LifeFree-From-Hunger-Summary-Africa1.pdf. Accessed January 17, 2015. 10. Franz J. Improving nutrition: Early childhood development and mobile health. Impact and Innovation Series. The National Bureau of Asian Research; 2012, http://www.pacifichealthsummit.org/downloads/ Impact%20&%20Innovations/Improving_Nutrition.PDF. Accessed January 17, 2015. 11. Babatunde RO, Olagunju FI, Fakayode SB, Sola-Ojo FE. Prevalence and determinants of malnutrition among under-five children of farming households in Kwara State, Nigeria. J Agric Sci 2011;3:173-81. 12. Psaki S, Bhutta ZA, Ahmed T, Ahmed S, Bessong P, Islam M, et al. Household food access and child malnutrition: results from the eightcountry MAL-ED study. Popul Health Metr 2012;10:24. 13. Sah N. Determinants of child malnutrition in Nepal: a case analysis from Dhanusha, Central Terai of Nepal. J Nepal Health Res Council 2004;2: 50-4.
14. Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis M, Ezzati M, et al. Maternal and child undernutrition: global and regional exposures and health consequences. Lancet 2008;371:243-60. 15. Demissie S, Worku A. Magnitude and factors associated with malnutrition in children 6-59 months of age in pastoral community of Dollo Ado district, Somali region, Ethiopia. Sci J Public Health 2013;1:175-83. 16. Kandala NB, Madungu TP, Emina JBO, Nzita KPD, Cappuccio FP. Malnutrition among children under the age of five in the Democratic Republic of Congo (DRC): does geographic location matter? BMC Public Health 2011;11:261. 17. Anderson AK, Bignell W, Winful S, Soyiri I, Asiedu MS. Risk factors for malnutrition among children 5-years and younger in the AkuapimNorth district in the eastern region of Ghana. Curr Res J Biol Sci 2010; 2:183-8. 18. Khan MMH, Kr€amer A, Khandoker A, Pr€ ufer-Kr€amer L, Islamc A. Trends in socio-demographic and health-related indicators in Bangladesh, 1993-2007: will inequities persist? Bull World Health Organ 2011;89:583-93. 19. Chowdhury AMR. Rethinking interventions for women’s health. Lancet 2007;370:1292-3. 20. de Onis M, Brown D, Bl€ ossner M, Borghi E. Levels and trends in child malnutrition: UNICEF-WHO-The World Bank joint child malnutrition estimates, UNICEF, New York; WHO, Geneva; The World Bank, Washington, DC; 2012, WHO Library Cataloguing-in-Publication Data; ISBN 978 92 4 1504515; 2012. 21. Islam A, Biswas T. Chronic stunting among under-5 children in Bangladesh: a situation analysis. Adv Pediatr Res 2015;2:18. 22. Siddiqi MNA, Haque MN, Goni MA. Malnutrition of under-five children: evidence from Bangladesh. Asian J Med Sci 2011;2:113-9. 23. Das S, Rahman RM. Application of ordinal logistic regression analysis in determining risk factors of child malnutrition in Bangladesh. Nutr J 2011;10:124. 24. Alom J, Quddus MA, Islam MA. Nutritional status of under-five children in Bangladesh: a multilevel analysis. J Biosoc Sci 2012;44:525-35. 25. Bangladesh Demographic and Health Survey, 2011. Dhaka, Bangladesh and Calverton, Maryland, USA: National Institute of Population Research and Training (NIPORT) Mitra and Associates and Macro International; 2013, http://www.measuredhs.com/publications/publicationFR265-DHS-Final-Reports.cfm. Accessed December 25, 2014. 26. Das S, Sahoo H. An investigation into factors affecting child undernutrition in Madhya Pradesh. Anthropologist 2011;13:227-33. 27. Coates J, Swindale A, Bilinsky P. Household Food Insecurity Scale (HFIAS) for measurement of food access: Indicator guide. Food and Nutrition Technical Assistance (FANTA). Version 3. Washington, DC: Academy of Educational Development; 2007, http://www.fantaproject. org/downloads/pdfs/HFIAS_v3_Aug07.pdf. Accessed October 24, 2015. 28. Rahman MM, Gilmour S, Akter S, Abe SK, Saito E, Shibuya K. Prevalence and control of hypertension in Bangladesh: a multilevel analysis of a nationwide population-based survey. J Hypertens 2015;33:465-72. 29. de Onis M, Bl€ ossner M. WHO Global Database on Child Growth and Malnutrition. Geneva: World Health Organization; 1997, whqlibdoc. who.int/hq/1997/WHO_NUT_97.4.pdf. Accessed January 23, 2015. 30. Young H, Jaspars S. The meaning and measurement of acute malnutrition in emergencies. A primer for decision-makers. Network paper. Humanitarian Practice Network (HPN). United Kingdom: Overseas Development Institute; November 2006. Number 56. 31. ICF International, 2012. The DHS Program STAT compiler. http://www. statcompiler.com. Accessed December 30, 2014. 32. Rusinga O, Moyo S. Determinants of child malnutrition in Changazi Ward in Chimanimani District, Zimbabwe. J Emerg Trends Educ Res Policy Stud 2012;3:187-92. 33. UNICEF. A strike against chronic malnutrition in Burundi, 2013. http:// www.unicef.org/infobycountry/burundi_69651.html. Accessed January 23, 2015. 34. Mashal T, Takano T, Nakamura K, Kizuki M, Hemat S, Watanabe M, et al. Factors associated with the health and nutritional status of children under 5 years of age in Afghanistan: family behavior related to women
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
7
THE JOURNAL OF PEDIATRICS
35.
36.
37.
38.
39.
40.
41.
42. 43.
8
Volume -
www.jpeds.com
and past experience of war-related hardships. BMC Public Health 2008; 8:301. Peiris TDR, Wijesinghe DGNG. Nutritional status of under 5 yearold children and its relationship with maternal nutrition knowledge in Weeraketiya DS division of Sri Lanka. Trop Agr Res 2010;21: 330-9. Adel ET, Marie-Franc¸oise RC, Salaheddin MM, Najeeb E, Ahmed AM, Ibrahim B, et al. Nutritional status of under-five children in Libya: a national population-based survey. Libyan J Med 2008;3:13-9. Ali D, Saha KK, Nguyen PH, Diressie MT, Ruel MT, Menon P, et al. Household food insecurity is associated with higher child undernutrition in Bangladesh, Ethiopia, and Vietnam, but the effect is not mediated by child dietary diversity. J Nutr 2013;43:1-7. Fuchs C, Sultana T, Ahmed T, Iqbal Hossain M. Factors associated with acute malnutrition among children admitted to a diarrhea treatment facility in Bangladesh. Int J Pediatr 2014;2014:267806. Ruwali D. Nutritional status of children under five years of age and factors associated in Padampur VDC, Chitwan. Health Prospect 2011;10: 14-8. Rahman A, Chowdhury S, Hossain D. Acute malnutrition in Bangladeshi children: levels and determinants. https://eprints.usq.edu.au/4332/1/ Rahman_Chowdhury_Hossain.pdf. Accessed January 25, 2015. Girmay M, Charlotte H, Michael D, Atalay A, Fikru T, Bogale W, et al. Prevalence and predictors of under nutrition among infants aged six and twelve months in Butajira, Ethiopia. BMC Public Health 2010;10:27. Deolalikar AB. Poverty and child malnutrition in Bangladesh. J Dev Soc 2005;21:55-9. Wamani H, Tylleskar T, Astrom AN, Tumwine JK, Peterson S. Mothers’ education but not fathers’ education, household assets or land owner-
44.
45. 46.
47.
48. 49.
50.
51.
52.
ship is the best predictor of child health inequalities in rural Uganda. Int J Equity Health 2004;3:9. Hien NN, Kam S. Nutritional status and the characteristics related to malnutrition in children under five years of age in Nghean, Vietnam. J Prev Med Public Health 2008;41:232-40. Fotso JC. Urban–rural differentials in child malnutrition: trends and socioeconomic correlates in sub-Saharan Africa. Health Place 2007;13:205-23. UNICEF. Child malnutrition and household food insecurity remain major concerns for Bangladesh 2009. http://www.unicef.org/media/media_ 48981.html. Accessed January 25, 2015. Setboonsarng S. ADB Institute Discussion Paper No. 21. Child malnutrition as a poverty indicator: An evaluation in the context of different development interventions in Indonesia; the paper draws part of the Indonesian experience from an unpublished ADB Study on Indonesia in 2002. January 2005. Das S, Hossain MZ, Islam MA. Predictors of child chronic malnutrition in Bangladesh. Proc Pakistan Acad Sci 2008;45:137-55. Haque SE, Rahman M, Mostofa MG, Zahan MS. Reproductive health care utilization among young mothers in Bangladesh: does autonomy matter? Womens Health Issues 2012;22:e171-80. Choudhury KK, Hanifi MA, Rasheed S, Bhuiya A. Gender inequality and severe malnutrition among children in a remote rural area of Bangladesh. J Health Popul Nutr 2000;18:123-30. Henry FJ, Briend A, Fauveau V, Huttly SA, Yunus M, Chakraborty J. Gender and age differentials in risk factors for childhood malnutrition in Bangladesh. Ann Epidemiol 1993;3:382-6. Save the Children. Malnutrition in Bangladesh. Harnessing social protection for the most vulnerable, https://www.savethechildren.org.uk/ sites/default/files/images/Malnutrition_in_Bangladesh.pdf. Accessed October 24, 2015.
Chowdhury et al
- 2016
ORIGINAL ARTICLES
A total of 8550 children under 5 years of age were eligible for anthropometric measurements
Height and weight data are missing for 115
8435 Children have complete and credible anthropometric data
Age data are missing for 790 children
7647 Children have complete and credible anthropometric and age data
BMI are absence for mothers of 48
7599 Children have complete and credible anthropometric and age data
25 missing data were recorded for other variables
children
children
7568 Children deemed eligible for the present study (Final data)
Figure. Sample selection.
Risk Factors for Child Malnutrition in Bangladesh: A Multilevel Analysis of a Nationwide Population-Based Survey
8.e1