Socio-economic risk factors for early childhood underweight in Bangladesh

Socio-economic risk factors for early childhood underweight in Bangladesh Background: Underweight is a major cause of global disease burden. It is associated with child mortality and morbidity, and its adverse impact on human performance and child survival is well recognized. Underweight is a major public health problem in Bangladesh, which is amongst the highest underweight prevalent countries in the world. The objectives of our study were to determine the national and regional prevalence rates of underweight and severe underweight in Bangladesh, and to investigate the association of socioeconomic and demographic factors with child underweight and severely underweight among children under the age of five living in Bangladesh. Methods: We performed a cross sectional study using Multiple Indicators Cluster Survey 2012–13, Bangladesh data on 17,133 children under 5 years of age. Weight-for-age Z scores based upon World Health Organization (WHO) guidelines were used to define child underweight and severe underweight. The association of underweight and severe underweight with household socioeconomic factors and demographic characteristics was investigated using binary logistic regression model. Results: An estimated 31.67% children were underweight and 8.81% children were severely underweight. Children of mothers with incomplete secondary education [Odds Ratio (OR) = 0.84, 95% CI: 0.75, 0.94] and mothers with completed secondary education [Odds Ratio (OR) = 0.77, 95% CI: 0.64, 0.93] were less likely to be underweight than children of uneducated mothers who had no formal schooling. A similar association exists for father’s education, children from households in the highest wealth index quintile had lower likelihood of underweight [Odds Ratio (OR) = 0.44, 95% CI: 0.37, 0.53] than children from households in the lowest quintile. Consumption of non-iodized salt had higher risk of severe underweight for children aged between 24 and 35 months [Odds Ratio (OR) = 2.32, 95% CI: 1.80, 3.00]. Other risk factors of child severe underweight included living in Sylhet division and increases in the number of children under the age of five in a household. Conclusion: Underweight was associated with lower parental education, household position in lower wealth index, living in Sylhet division and consumption of non-iodized salt. Strategies are discussed considering the relative importance of risk factors for child underweight. Keywords: Determinants, Underweight, Bangladesh, Child, Early childhood Background children who are mild to moderately malnourished com- Developing countries have the greatest prevalence of pared with well-nourished children [5, 6]. The adverse early childhood malnutrition despite recent rapid eco- impacts of malnutrition are not limited to school attend- nomic development [1–3]. Malnutrition is the prime ance rates and academic accomplishments, but also culprit for early childhood death, causing the deaths of linked to decreases in the development of social skills 13 million infants and children under the age of five among children [7, 8]. Malnutrition is associated with every year [4]. There is a higher risk of dying among poorer academic performance in school-aged children [9] and lower survival capacity for a child in its adult- * Correspondence: tuhinchy.sust@gmail.com hood [10]. Department of Economics, Shahjalal University of Science & Technology, Sylhet Kumargaon, Sylhet 3114, Bangladesh Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Chowdhury et al. Globalization and Health (2018) 14:54 Page 2 of 12 Malnutrition status is investigated using three an- education. Mohsena et al. [24] explored the Bangladesh thropometric parameters: height for age, weight for Demographic and Health Survey (2004) Data and found height and weight for age [11]. Height for age is used that higher levels of maternal education are associated to identify stunting which represents linear growth with lower risk of underweight. In Bangladesh, paternal failure for a child [12]. Weight for height is used to education was also found to be an independent deter- identify wasting which indicates an unusually body tissue minant of child underweight by few studies [27, 33]. Ac- and fat mass for a child relative to its height or length cording to Das et al. [33], children of illiterate fathers [13]. Weight for age is the WHO-recommended indicator had a 30% higher likelihood of being underweight to determine whether the child is underweight or not than the children of fathers who had completed pri- [11]. As underweight is a composite indicator that encom- mary school. Another study that finds a positive asso- passes both stunting and wasting [14] implying that stunt- ciation between paternal education and risk of child ing, wasting or both can be reflected by underweight [15], underweight [26]. our primary interest is to use weight for age indicator to Socioeconomic status like poverty is also a signifi- measure early childhood underweight. cant predictor of child underweight. Children coming Underweight has become a major public health from poor families are more likely to be underweight problem particularly in developing nations, and its compared with children coming from wealthier fam- destructive impact on human performance and child ilies [4]. Despite the fact that Bangladesh has man- survival has been well recognized [16]. It is a prime aged to achieve an enormous economic growth in the concern as underweight is the leading cause of global last 20 years along with massive improvements in disease burden [17]. Underweight is associated with educational achievement, still a large proportion of its increased risk of contagious disease such as diarrhea population, nearly 43.3%, are living under the poverty and pneumonia [18]. The World Health Organization line of 1.25 dollar per day [28]. Having such a large (WHO)cites underweightas the single largest risk proportion of the population living in poverty, hence factor in developing countries to the burden of dis- unable to afford nutritious food or access improved ease, [19] as an estimated 52.5% of all deaths among health care facilities, makes the task of reducing the coun- young children are attributable to underweight [18]. try’s prevalence of underweight children extremely difficult. In the countries where child mortality is high, under- Indeed, a number of studies conducted in Bangladesh sug- weight is accountable for 15% of the total disability gested household socioeconomic status as a significant de- adjusted life years losses [17]. Underweight children terminant of child underweight, reporting that children are 8.4 times more likely to die and children who are from relatively higher socioeconomic condition had lower suffering from moderate and mildly underweight are likelihood of being underweight [28, 33]. A study in 4.6 times and 2.5 times respectively more likely to die Bangladesh carried out by Alom et al. [29]reportedthat before the age of five compared to children who are children who came from the richest wealth index had 44% well-nourished [5]. lower odds of being underweight compared with children In 2011, 101 million children globally were estimated as from the poorest wealth index. underweight indicating the prevalence rate of underweight Among the socioeconomic determinants, most of the as 16% [20]. The situation is worse in Bangladesh, where studies conducted in Bangladesh confirmed that mothers’ underweight prevalence rate was reported to be 41% [21]. literacy status, choice of latrine and poor household in- There are many socioeconomic and biological factors that come are the associated risk factors for underweight are associated with child underweight. Biological factors among children under five [25, 27, 29, 30]. Other risk fac- include lower birth weight [22], having an underweight tors estimated as significantly associated with the risk of mother or father, and having a mother or father of short child underweight are improper feeding practices, mater- stature. Among the socioeconomic factors, children with nal antenatal and postnatal care, diarrhea of children mothers who don’t have any formal education are at within last 2 weeks [32]; living in rural area [30]; mothers greater risk of being underweight [23]. having less exposure to media, children of mothers work- Even though Bangladesh has one of the highest preva- ing outside home [31]; unimproved toilet facilities, and liv- lence of underweight in the world, few studies have inginSylhet division[33]. The risk of being underweight identified the socioeconomic factors responsible for among under-five children was observed to increase with child underweight [24–30]. Most of the studies con- an increase in children’s age by several studies in ducted in Bangladesh suggested that maternal education Bangladesh. [29, 31, 33]. The causes for underweight are is a strong and significant predictor for child under- multidimensional and interrelated, and they often change weight [24, 28, 29, 31, 32]. A recent study by Devkota et from country to country, highlighting the importance of al. [28] established a positive association of WAZ local research studies. Therefore, the objectives of our (weight-for-age z score) with the level of maternal study is to determine the national and regional prevalence Chowdhury et al. Globalization and Health (2018) 14:54 Page 3 of 12 of underweight and severe underweight and their associ- The first stage of sampling involved determining census ated risk factors in Bangladesh. enumeration areas required within each district, which were defined as primary sampling units. From 30 No- vember, 2012 to 6 December, 2012, field operations were Method carried out in each enumeration area to make new list- Data sources ings of households; primarily because the 2011 census The analysis was based on the Bangladesh Multiple Indi- frame was not up-to-date. Experienced staff of the cators Cluster Survey (MICS) 2012–13 data conducted Bangladesh Bureau of Statistics (BBS) prepared the list by Bangladesh Bureau of Statistics along with UNICEF of households in the field, and at the BBS headquarter, who provided technical and financial support. Data were the listed households were numbered sequentially from collected from December 2012 to April 2013. The major 1 to n within each enumeration area. Here, Random Sys- objectives of the survey were to collect the most recent tematic Sample Selection Procedure was used to select data on women and child health, household socioeco- 20 households in each enumeration area. Households in- nomic status, access to mass media, ICT improved sani- cluded 23,402 children below 5 years of age and infor- tation and pure water, child nutrition, mortality and mation on 20,903 under five children was collected, child development. The sample was selected using indicating a response rate of 89.3%. For the purpose of two-staged stratified cluster sampling approach. In the data collection, structured questionnaires were used. time period during which data were collected, there The questionnaires had four sections comprised of ques- were seven divisions and 64 districts in Bangladesh and tions pertaining to household, women (age 15–49), chil- districts were defined as sampling strata [34]. The pro- dren under five and water quality testing. Thirty-two portion of women with 4+ ANC visits, where the term data collection teams collected data, with each team hav- ANC stands for antenatal care visit, was used as key in- ing seven members including a supervisor and inter- dicator in estimating the size of sample. The sample size viewers. The teams were provided with training for was determined using the following formula n= 14 days prior to collecting data. Recommended an- ½4ðrÞð1−rÞðdeff Þ , where 4 = factor for achieving 95% con- thropometric tools were used to measure child weight ½e ðpbÞðAve SizeÞðRRÞ fidence level; r = projected value of the indicator, which and height [34]. Collected data from the field were en- was assumed to be 26% and expressed as a proportion; tered using CSPro software and all questionnaires were n = sample size; deff = design effect for the indicator, and entered twice to ensure the quality of data [35]. the value of deff was estimated to be 1.4 based on previ- ous surveys; Ave Size = average household size, and it Statistical analysis was assumed as 4.5; pb refers to the percentage of There were 20,903 children under age 5 years about women in the total population, who gave birth in the last whom information was collected. However, 982 children 2 years, and it was taken to be 4%; RR = anticipated re- were not included as there was no information for sponse rate which was assumed to be 95% based on the weight-for-age-z score. For the households with more previous survey experience [35]. In estimating r, max- than one child under the age of five, the youngest one imum margin of error was allowed of ±8 to 9%, and was selected as the index child and the other children in thereby a sample size of 800 was decided for each dis- the household were not considered for the study. Finally, trict. For obtaining a more precise estimation result, data pertaining to 17,133 children below the age of five 1000 samples were drawn from each of the 20 UNDAF were subject to analysis. Weight-for-age-z score of districts. UNDAF (United Nation Development Assist- WHO growth standard was used to create the under- ance Framework) districts are those 20 districts that weight variable. Children were defined to be under- have been declared as prioritized districts by Bangladesh weight if their weight-for-age-z scores were below minus government. Twenty households per cluster were se- two and severely underweight if their weight-for-age-z lected for the survey, taking in consideration various fac- scores were below minus three. Data analysis was per- tors such as time, budget, design effect etc. Twenty formed using Stata version 12.1. Child, paternal and households per cluster suggest that for drawing 1000 household socioeconomic factors were included in the households as sample, 50 clusters would be needed in analysis as predictors of child underweight. Factors in- each of the 20 UNDAF districts, and 40 cluster would be cluded child’s age, gender, geographic area, education needed for drawing 800 households as sample size in level of mother, education level of father, regional div- each of the remaining 44 non UNDAF districts. In this ision, wealth index quintile used as the proxy indicator way, the resulting sample size would be 55,200. But 4 of household socioeconomic status, number of house- clusters were not visited due to unfavorable weather hold members, toilet facility, salt iodization test, whether and, therefore, data was collected from 55,120 house- the child drank plain water yesterday (yes/no), whether holds. Clusters were selected using 2011 census frame. the child drank milk yesterday (yes/no), whether the Chowdhury et al. Globalization and Health (2018) 14:54 Page 4 of 12 child took vitamins yesterday (yes/no) and number of The analysis determined the following factors to be as- children in a household below the age of five. Wealth sociated with the risk of underweight: age of children, index score was computed by principal component ana- division, education level of mother and education level lysis based on household assets and materials used to of father, wealth index, whether the child drank plain build house [35]. All households were then categorized water yesterday and finally the use of iodized salt by into five categories depending on wealth score such as households. Children living in Khulna division were 15% poorest, second, middle, fourth and richest. Toilet facil- less likely to be underweight [Odds Ratio (OR) = 0.85, ity variable was created with four categories as flush la- 95% CI: 0.75, 0.96] compared with the children living in trine, pit latrine, hanging latrine and others. Descriptive Dhaka division. Children from Sylhet division were more statistics were created for all predictors as well as for re- prone to the risk of underweight than any other division sponse variable. As our dependent variable was dichot- in Bangladesh with higher odds on underweight [Odds omous, underweight versus not underweight and Ratio (OR) = 1.38, 95% CI: 1.20, 1.60] than the children severely underweight versus not severely underweight, from Dhaka division. Our results suggested child’s age as predictors for the risk of child underweight and severely one of the major determinants of underweight. Children underweight were investigated using binary logistic re- aged 12–23 months [Odds Ratio (OR) = 1.98, 95% CI: gression model. Statistical significance was set at p < 0.05 1.69, 2.32] had around two-fold odds on underweight and adjusted odds ratio was reported along with 95% compared with children below 5 months of age. The risk confidence interval. for underweight was highest among the children who were in their early middle childhood. Children aged 24– Results 35 months [Odds Ratio (OR) = 2.58, 95% CI: 2.20, 3.03] Socioeconomic and demographic characteristics of the had over two-fold risk on underweight relative to the study population children less than 5 months of age. Children between The final analysis of 17,133 children estimated that 31.67% the ages of 36–47 months [Odds Ratio (OR) = 2.44, 95% of these children were underweight including 8.81% as se- CI: 2.07, 2.88] and children between the ages of 48– verely underweight. Table 1 demonstrated socioeconomic 59 months [Odds Ratio (OR) = 2.35, 95% CI: 1.98, 2.77] and demographic characteristics of the study population. were more likely to be underweight compared with the The male to female ratio of children was 1069:1000. children who were below 5 months of age. Children of More than four-fifths of children came from rural area secondary incomplete mothers had 16% smaller odds on and around two-thirds of the surveyed children belonged underweight [Odds Ratio (OR) = 0.84, 95% CI: 0.75, to the households with four to six family members. 0.94] than the children of mothers with no formal edu- More than one-fifth of the children were between the cation. There were 23% lower odds on underweight ages of 12 and 23 months, and approximately 11% of among children whose mothers completed secondary children were aged less than 5 months. The highest education [Odds Ratio (OR) = 0.77, 95% CI; 0.64, 0.93] (25.36%) proportion of children lived in Dhaka division than the children whose mothers had not attended with the lowest from Sylhet division (8.21%). More than school at all. As with mother’s education, an increase in one-third of fathers did not go to school and likewise the level of father’s education reduced the odds on child one-fourth of mothers did not have any formal educa- underweight. Children with father’s education secondary tion. Only 11.37% of mothers and 15.38% of fathers incomplete and secondary complete had 12% [Odds Ra- completed secondary education. Nearly, one third of the tio (OR) = 0.88, 95% CI: 0.79, 0.98] and 28% [Odds Ratio study children belonged to households of the poorest (OR) = 0.72, 95% CI: 0.62, 0.84] lower odds on under- wealth quintile (29.06%) and only 13.87% children weight respectively in comparison with children whose belonged to richest wealth quintile. On the previous day, fathers had no formal education. Children who didn’t approximately 86% children drank plain water 24% drink plain water yesterday were 1.27 times more likely drank milk and 9.06% ate or drank vitamins. More than to be underweight [Odds Ratio (OR) = 1.27, 955 CI: 1.12, half of the households have iodized salt (51.43%). The 1.43] compared with children who drank water the pre- majority of households had access to pit latrine (62.34%) vious day. and almost one-third of households had access to flush Wealth index was also related to the risk of under- latrines. weight. Children belonging to households which were in the middle category [Odds Ratio (OR) = 0.82, 95% CI: Risk factors for underweight 0.73, 0.91] and in fourth category [Odds Ratio (OR) = Table 2 lists the socioeconomic and demographic factors 0.67, 95% CI: 0.59, 0.77] were less likely to be under- along with their association and the magnitude of im- weight compared with the children who belonged to the pact on the risk for underweight among children below poorest households in the wealth index. There were 56% 5 years old using binary logistic regression analysis. lower odds on underweight [Odds Ratio (OR) = 0.44, Chowdhury et al. Globalization and Health (2018) 14:54 Page 5 of 12 Table 1 Socioeconomic and demographic characteristics of the Table 1 Socioeconomic and demographic characteristics of the study population study population (Continued) Total sample Total sample (n = 17133) (n = 17133) Characteristics Proportion (%) Characteristics Proportion (%) Household member Fourth 16.52 <=3 14.29 Richest 13.87 4–6 63.64 Child drank plain water yesterday > = 7 22.06 Yes 85.46 Age (months) No 14.54 0–5 10.93 Child drank milk yesterday 6–11 10.80 Yes 23.44 12–23 21.88 No 76.56 24–35 20.60 Child ate vitamins yesterday 36–47 18.75 Yes 9.06 48–59 17.04 No 90.94 Gender Toilet facility Male 51.67 Flush latrine 23.94 Female 48.33 Pit latrine 62.34 Area Hanging latrine 7.78 Urban 15.87 Others 5.93 Rural 84.13 Salt idolization test Division Not iodized 0 PPM 27.15 Dhaka 25.36 More than 0 PPM and less than 15 PPM 20.49 Barisal 9.64 15 PPM or more 51.43 Chittagong 19.53 No salt in the household 0.92 Khulna 13.96 Child underweight status Rajshahi 10.28 Underweight 31.67 Rangpur 13.01 Severely underweight 8.81 Sylhet 8.21 Education of mother None 24.10 95% CI: 0.37, 0.53] among the children who came from Primary incomplete 14.55 richest households than the children who came from Primary complete 15.39 poorest households. Children living in households where salt was found as adequately iodized (15 ppm or more) Secondary incomplete 34.59 had a lower likelihood of underweight [Odds Ratio (OR) Secondary complete or higher 11.37 = 0.90, 95% CI: 0.82, 0.98] compared with children from Education of father those households where salt was found to be not iodized None 34.79 in salt iodization tests. Our study did not establish a re- Primary incomplete 15.73 lationship between child gender and underweight, nor Primary complete 13.75 did it establish that children from rural areas were more likely to be underweight compared with children from Secondary incomplete 20.35 urban areas. Secondary complete or higher 15.38 Wealth index quintile Risk factors for severe underweight Poorest 29.06 Table 3 lists the results of the binary logistic regression Second 21.86 analysis on the risk for child severe underweight associ- Middle 18.69 ated with different socioeconomic and demographic factors. Chowdhury et al. Globalization and Health (2018) 14:54 Page 6 of 12 Table 2 Risk factors for total underweight in children aged less than 5 years Factors Underweight (%) N OR 95% CI Household member <=3 33.24 2449 ref. 4–6 31.54 10,904 0.91 (0.82–1.00) > = 7 31.03 3780 1.00 (0.88–1.13) Age (Months) 0–5 21.10 1872 ref. 6–11 21.88 1851 1.17 (0.98–1.39) *** 12–23 31.05 3749 1.98 (1.69–2.32) *** 24–35 36.91 3530 2.58 (2.20–3.03) *** 36–47 35.62 3212 2.44 (2.07–2.88) *** 48–59 34.77 2919 2.35 (1.98–2.77) Gender Male 32.15 8853 ref. Female 31.16 8280 0.96 (0.90–1.03) Area Urban 26.52 2719 ref. Rural 32.64 14,414 0.96 (0.86–1.07) Division Dhaka 30.29 4345 ref. Barisal 33.96 1652 1.04 (0.91–1.19) Chittagong 33.59 3346 1.11 (0.99–1.24) ** Khulna 26.55 2392 0.85 (0.75–0.96) Rajshahi 30.14 1762 0.90 (0.79–1.02) Rangpur 32.17 2229 0.98 (0.87–1.11) *** Sylhet 38.52 1407 1.38 (1.20–1.60) Education of mother None 39.55 4129 ref. Primary incomplete 37.18 2493 1.02 (0.91–1.14) Primary complete 34.36 2637 0.96 (0.85–1.07) ** Secondary incomplete 27.19 5926 0.84 (0.75–0.94) ** Secondary complete or higher 17.92 1948 0.77 (0.64–0.93) Education of father None 38.25 5428 ref. Primary incomplete 35.66 2454 0.98 (0.88–1.09) Primary complete 33.15 2145 0.97 (0.87–1.09) Secondary incomplete 27.81 3175 0.88 (0.79–0.98) *** Secondary complete or higher 19.30 2399 0.72 (0.62–0.84) Wealth index quintile Poorest 40.11 4979 ref. Second 35.54 3745 0.90 (0.82–0.99) *** Middle 30.78 3203 0.82 (0.73–0.91) *** Fourth 25.48 2830 0.67 (0.59–0.77) *** Richest 16.46 2376 0.44 (0.37–0.53) Child drank plain water yesterday Yes 32.11 14,619 ref. *** No 29.03 2487 1.27 (1.12–1.43) Child drank milk yesterday Yes 26.94 4013 ref. No 33.10 13,105 1.07 (0.98–1.17) Child ate vitamins yesterday Yes 28.69 1551 ref. No 31.97 15,559 0.96 (0.84–1.09) Toilet facility Flush latrine 24.46 4100 ref. Pit latrine 32.53 10,677 1.04 (0.94–1.16) Hanging latrine 41.86 1333 1.11 (0.95–1.30) Chowdhury et al. Globalization and Health (2018) 14:54 Page 7 of 12 Table 2 Risk factors for total underweight in children aged less than 5 years (Continued) Factors Underweight (%) N OR 95% CI Others 38.29 1016 1.02 (0.86–1.22) Salt iodization test Not iodized 0 PPM 36.54 4619 ref. More than 0 PPM and less than 15 PPM 34.77 3486 0.96 (0.87–1.06) 15 PPM or more 27.90 8750 0.90 (0.82–0.98) No salt in the household 29.30 157 0.80 (0.55–1.16) Number of under five children 1.02 (0.93–1.11) *** ** * p < 0.001; p < 0.01; p < 0.05 Age of children was associated with the risk for child had 21% lower odds of being severely underweight when severe underweight with children aged 12–23 months their fathers had enrolled in secondary education compare having almost twice the odds on severe underweight to children whose fathers had no formal education [Odds [Odds Ratio (OR) = 1.81, 95% CI: 1.40, 2.34] than the Ratio (OR) = 0.79, 95% CI: 0.65, 0.95]. There were 34% children aged less than 5 months. Children aged 24– lower odds of severely underweight among children with 25 months had the highest risk for being severely under- fathers who had completed secondary education [Odds Ra- weight. Compared with children aged below 5 months, tio (OR) = 0.66, 95% CI: 0.50, 0.88] compared with children children aged 24–35 months [Odds Ratio (OR) = 2.32, whose fathers had no formal education. An increase in the 95% CI: 1.80, 3.00] had a more than two-fold risk of se- number of children under the age of five in households by vere underweight. Furthermore, children between the 1 was associated with 17% higher odds on severe under- ages of 48–59 months had also larger odds of severe weight [Odds Ratio (OR) = 1.17, 95% CI: 1.02, 1.33]. underweight [Odds Ratio (OR) = 1.48, 95% CI: 1.12, 1.95] compared with their under-five counterparts. Chil- Discussion dren from the poorest households were at higher risk for Our study estimated 31.67% children below 5 years of age being severely underweight. Children from households as underweight, with 8.8% children as severely under- belonging to the middle category in wealth index had weight in 2011–2012 in Bangladesh. The present study 20% lower likelihood of underweight [Odds Ratio (OR) findings indicated a concerning nutritional condition = 0.80, 95% CI: 0.67, 0.96] and children from households among children aged under 5 years, especially those aged that belonged to fourth category in the wealth index had 2–3 years. Child underweight and severely underweight 49% lower odds of underweight [Odds Ratio (OR) = 0.51, were predicted by child’s age, regional division, education 95% CI: 0.40, 0.64] compared with the children from the level of mother, education level of father, household pos- poorest households. In comparison with the children ition in wealth index, whether child drank plain water yes- from the poorest households, children from the wealthi- terday, and the presence of iodized salt in households. est households had 68% lower odds of severe under- Our study reveals maternal education as an important weight [Odds Ratio (OR) = 0.32, 95% CI: 0.23, 0.46]. determinant of child underweight. According to our There were 32% lower odds on severe underweight study findings, higher levels of maternal education are among children living in the Khulna division [Odds Ra- associated with lower odds of child underweight, which tio (OR) = 0.68, 95% CI: 0.55, 0.85] and 34% higher odds was supported by previous studies [24, 36, 37]. Likewise, on severe underweight among children living in the Syl- maternal education was estimated to be an independent het division [Odds Ratio (OR) = 1.34, 95% CI: 1.09, 1.66] predictor of children being severely underweight. Chil- compared with children living in the Dhaka division. Im- dren of secondary completed mothers had 31% lower proved mother’s education and father’s education levels odds of being severely underweight compared with chil- were shown to be important protective factors against dren of uneducated mothers. Similar findings have been child severe underweight. Children with mothers’ pri- observed in earlier studies carried out in Bangladesh, mary education complete [Odds Ratio (OR) = 0.79, 95% which have shown that children of uneducated mothers CI: 0.66, 0.95] and secondary incomplete [Odds Ratio were significantly more likely to be severely underweight (OR) = 0.78, 95% CI: 0.65, 0.92] were less likely to be se- than children whose mothers had a relatively higher verely underweight compared with children whose level of education [25, 38–43]. There are some plausible mothers had no formal education. Children of mothers explanations as to why higher maternal education re- who had completed secondary education had 31% lower duces the chance of underweight and severely under- odds [Odds Ratio (OR) = 0.69, 95% CI: 0.49, 0.97] of being weight among children under the age of five. First, severely underweight compared with children whose education improves information processing capacity by mothers had not completed secondary education. Children the mother [44, 45]. For instance, an educated mother Chowdhury et al. Globalization and Health (2018) 14:54 Page 8 of 12 Table 3 Risk factors for severe underweight in children aged less than 5 years Factors Severe Underweight (%) N OR 95% CI Household member <=3 8.82 2449 4–6 8.69 10,904 0.89 (0.75–1.05) > = 7 9.13 3780 0.98 (0.80–1.20) Age (Months) 0–5 6.57 1872 6–11 6.86 1851 1.25 (0.94–1.66) *** 12–23 9.02 3749 1.81 (1.40–2.34) *** 24–35 11.16 3530 2.32 (1.80–3.00) *** 36–47 9.74 3212 2.04 (1.56–2.65) *** 48–59 7.33 2919 1.48 (1.12–1.95) Gender Male 9.25 8853 Female 8.33 8280 0.90 (0.80–1.00) Area Urban 6.91 2719 Rural 9.16 14,414 0.96 (0.80–1.15) Division Dhaka 8.63 4345 Barisal 9.87 1652 1.06 (0.86–1.32) Chittagong 10.10 3346 1.06 (0.89–1.26) ** Khulna 5.60 2392 0.68 (0.55–0.85) Rajshahi 7.72 1762 0.86 (0.69–1.06) Rangpur 8.03 2229 0.82 (0.67–1.00) ** Sylhet 13.08 1407 1.34 (1.09–1.66) Education of mother None 12.79 4129 Primary incomplete 10.87 2493 0.93 (0.78–1.10) Primary complete 8.99 2637 0.79 (0.66–0.95) ** Secondary incomplete 6.77 5926 0.78 (0.65–0.92) Secondary complete or higher 3.70 1948 0.69 (0.49–0.97) Education of father None 11.81 5428 Primary incomplete 9.98 2454 0.93 (0.79–1.09) Primary complete 9.60 2145 1.02 (0.86–1.22) Secondary incomplete 6.46 3175 0.79 (0.65–0.95) ** Secondary complete or higher 3.96 2399 0.66 (0.50–0.88) Wealth index quintile Poorest 12.59 4979 Second 10.28 3745 0.93 (0.80–1.08) Middle 8.30 3203 0.80 (0.67–0.96) *** Fourth 5.55 2830 0.51 (0.40–0.64) *** Richest 3.11 2376 0.32 (0.23–0.46) Child drank plain water yesterday Yes 8.65 14,619 *** No 9.77 2487 1.45 (1.21–1.73) Child drank milk yesterday Yes 7.28 4013 No 9.28 13,105 0.94 (0.81–1.09) Child ate vitamin yesterday Yes 7.67 1551 No 8.93 15,559 1.01 (0.81–1.25) Toilet facility Flush latrine 6.44 4100 Pit latrine 8.73 10,677 0.91 (0.76–1.08) Hanging latrine 14.40 1333 1.01 (0.80–1.28) Chowdhury et al. Globalization and Health (2018) 14:54 Page 9 of 12 Table 3 Risk factors for severe underweight in children aged less than 5 years (Continued) Factors Severe Underweight (%) N OR 95% CI Others 11.81 1016 0.96 (0.73–1.26) Salt iodization test Not iodized 0 PPM 10.46 4619 More than 0 PPM and less than 15 PPM 10.33 3486 1.04 (0.89–1.21) 15 PPM or more 7.38 8750 0.94 (0.82–1.08) No salt in the household 7.01 157 0.79 (0.42–1.47) Number of under five children 1.17 (1.02–1.33) *** ** * p < 0.001; p < 0.01, p < 0.0 can read, interpret and apply health tips from newspa- more knowledge of improved child care practices [53], pers and other sources, and is more likely to understand are better able to take responsibility for their children medical information such as medication levels suggested during illness [54], and more likely to have greater con- by doctor. These skills enable mothers to make informed trol over household decision-making [55]. Even though decisions about their children’s nutrition and health care, we do not know exactly why increased maternal educa- leading to improved child health. tion is associated with lower likelihood of child under- Second, maternal education is associated with an in- weight, the aforementioned result underscore the need crease in the frequency of mothers’ antenatal care visits for female education to reduce underweight prevalence and their utilization of child health services, leading to among children. decreased likelihood of underweight and severely under- Similar to maternal education, paternal education was a weight children. It has been evident from studies that significant predictor of underweight and severe under- the more educated mothers are the more likely they are weight. The result was supported by other studies that to access child health facilities more regularly compared suggested low paternal education was associated with in- to less educated mothers [46–48]. Antenatal care is often creased odds of child underweight [26, 27, 56, 57]. Educa- associated with improved child nutrition [26], and a tion level of fathers is generally associated with household study that employed data from 17 countries, aiming at income due to the fact that more educated fathers are determining the relationship between maternal educa- more likely to earn more money; indeed, father’seduca- tion and child survival, found that uneducated women tion is typically used as a proxy indicator of household in- represented 55% more non-use of antenatal services come [58]. Data on Indonesian children suggested that than their educated counterparts [49]. preventive factors against child malnutrition, such as use Third, there is general recognition that more educated of improved latrine, vitamin A capsules for children, ac- mothers are more likely to earn more money [44, 50]. cess to health facilities from local health community This implies increased capability to invest in their chil- clinics, and consumption of iodized salt, were strongly re- dren’s health, and more possibilities for living in an en- lated with higher level of paternal education [12]. There is vironment with relatively higher food security. Finally, strong evidence that more educated men are more likely maternal education improves child nutritional status to marry more educated women. As a result, children of through its impact on appropriate feeding practices by more educated fathers have an increased chance of having mothers. Appropriate feeding practices have been docu- mothers who are also more educated. The view is sup- mented as very crucial for improving nutrition condi- ported by Grossman’s model of the demand for good tions of children; particularly during their infancy [51], health which claims that more efficient investment in and a study conducted in Bangladesh revealed the asso- health is associated with more education [59]. ciation between improved feeding practices with educa- Another factor estimated to be significantly associated tion level of mother by demonstrating their findings with child under-nutrition is the age of the children. that; compared with uneducated mothers, educated Children below 5 months of age had the lowest risk of mothers were more concerned to breastfeed their chil- being underweight. One possible explanation could be dren in protected place along with cleanness; that they that during this period children are still being breastfed, were more aware to take breaks from the work at hand thereby gaining essential nutrition that renders them less while breastfeeding their children and were more con- vulnerable to malnutrition. On the other hand, study re- scious ensuring hygiene in the time children were pro- sults indicate that children’s risk of being underweight vided food [52]. Furthermore, higher level of education, increased with age, this risk reaching its peak between completed by mothers, is associated with behavioral the ages of 24–35 months. This suggests that during that changes that improve child health. Women who have period of children’s development, they had more than completed secondary education are more likely to have double the risk for underweight than the children who Chowdhury et al. Globalization and Health (2018) 14:54 Page 10 of 12 were less than 5 months of age. The similar result was media and poor maternal health were responsible for in- reported by a study which found the risk of being under- creased risk and higher prevalence of underweight among weight and severely underweight as highest between the under five children in the Sylhet division [4]. ages of 22–24 months [4]. Another study conducted The study had some limitations that should be ad- with Ethiopian children also supports our findings, dressed. First of all, the study was cross-sectional in de- which determined the association of the odds on child sign indicating that it was impossible to establish any underweight and severely underweight with age of chil- causal relationship between underweight and the esti- dren [60]. The reason for increased risk of underweight mated risk factors. Nevertheless, the findings have great associated with age may be that as children’s age in- significance as it is claimed that the results from the creased, they consumed complementary food along with same specific area using cross-sectional data have very breast milk. Contamination is more likely in comple- limited variation over time. Secondly, household income mentary food, therefore implicated in spreading infec- is a well-established determinant of child underweight. tious diseases among children and in turn rendering But it is nearly impossible to have reliable data on them vulnerable to the risk of underweight [61]. household income and expenditure, particularly in de- Moreover, there is strong evidence that the prevalence veloping countries such as Bangladesh. That being the of childhood disease is associated with child age, particu- case, we elected to use wealth index as a proxy indicator larly in most of the developing countries [62]. Children of household income, bearing in mind that wealth index living in higher wealth index households had smaller risk might not sufficiently represent household income. of being underweight and the finding was supported by Finally, data did not include child birth weight, other research work that claimed that children in lower mother’s weight, father’s weight, parents’ age, father’s oc- wealth index households had greater chance of under- cupation, or mother’s occupation. The magnitude of esti- weight than the children in higher wealth index house- mated risk factors could possibly be changed if those holds [56, 61, 63–66]. The explanation could be that factors were adjusted for our regression model. Despite the children in higher wealth index households were more study’s limitations, its strength was in assessing a large likely to belong to relatively food secure families, had number of samples, hence the findings can be generalized comparatively more educated fathers and mothers, and to all Bangladeshi children below 5 years old in 2011– lived in better neighborhood with improved health facil- 2012. This group of children needs to be re-examined in 5 ities. These factors might possibly have reduced the risk and 10 years to determine whether underweight and severe of underweight among the children living in higher underweight leads to lasting detriments for these children. wealth index households. It is widely known that mater- nal knowledge on child health has profound impact on Conclusion child well nourishment, and a study performed in Overall, in Bangladesh approximately one-third of the Bangladesh documented that mothers with better child study population is estimated as underweight. Under- health knowledge had mostly come from the richest weight prevalence not only reflects poor nutritional con- households [67]. It implies the significance of household ditions for children in a country, in general, it also economic condition on child nutrition through im- reflects high socio-economic costs for low quality of life, proved maternal knowledge on child health. Respon- vulnerability to different diseases and greater risk of dents who reported their children had drunk plain water childhood death. Due to the fact that it is more likely yesterday had lower odds on underweight for their chil- that underweight children who survive to adulthood will dren but there was no relationship between underweight experience greater health problems, the prevalence of and the type of toilet facilities. The risk of underweight underweight may also contribute to loss of productivity. was similar between male and female children, and be- The study revealed that the Sylhet division has the tween rural and urban children, which was unexpected. highest risk of severe underweight for under five chil- One explanation for this may be that poor people from dren, followed by the Chittagong division. With both di- rural areas were migrating to urban slum areas where visions their poverty rates are among the lowest in the economic conditions and health facilities were not re- country according to the latest Bangladesh Poverty markably different from rural areas. The risk for children Maps. This paper reveals that other than household pos- to be underweight depended on the division they lived ition in the lower wealth index, lower parental education in. Children from the Sylhet division had higher probability is one of the most important risk factors for child under- of underweight relative to children from the Dhaka division. weight. Sylhet is the home of most Bangladeshis that mi- The reason is not clear why children living in the Sylhet grate to the United Kingdom and are engaged in unskilled division were at more risk of underweight. Nevertheless, or semi-skilled blue-collar jobs and send remittances. Al- one possible reason might be that poor socioeconomic con- though remittances have improved land and housing in ditions such as few skilled workers, limited exposure to Sylhet, investment in other social sectors like education or Chowdhury et al. Globalization and Health (2018) 14:54 Page 11 of 12 business is reported to be insignificant [68]. In terms of Somali region, Ethiopia. Sci J Public Health. 2013; https://doi.org/10.11648/j. sjph.20130104.12. other important social indicators like ‘parental education’ 2. Masibo PK, Makoka D. Trends and determinants of undernutrition among Sylhet is reported to be poor-performing, according to the young Kenyan children: Kenya demographic and health survey; 1993, 1998, Bangladesh Bureau of Statistics, the overall literacy rate in 2003 and 2008–2009. Public Health Nutr. 2012; https://doi.org/10.1017/ S1368980012002856. the Sylhet division was the lowest in the country, at 39.2% 3. Pasricha SR, Biggs BA. Undernutrition among children in south and South- in 2007, with an even lower female literacy rate of 35.1% East Asia. J Paediatr Child Health. 2010; https://doi.org/10.1111/j.1440-1754. [69]. This result calls for government and non-government 2010.01839.x. 4. Rahman A. Significant risk factors for childhood malnutrition: evidence from interventions to increase education nationwide as much as an Asian developing country. Sci J Public Health. 2016; https://doi.org/10. possible, especially for the Sylhet division. This will also 11648/j.sjph.s.2016040101.14. markedly increase the mother’s age when children are 5. Pelletier DL, Frongillo EA Jr, Schroeder DG, Habicht JP. A methodology for estimating the contribution of malnutrition to child mortality in developing born, and decrease family size, which are particularly im- countries. J Nutr. 1994;124(10):2106S. portant for reducing the prevalence of underweight 6. Schroeder DG, Brown KH. Nutritional status as a predictor of child survival: children. summarizing the association and quantifying its global impact. Bull World Health Organ. 1994;72(4):569. 7. Alaimo K, Olson CM, Frongillo EA. Food insufficiency and American school- Endnotes aged children’s cognitive, academic, and psychosocial development. Stunting: height for age < − 2 standard deviations Pediatrics. 2001;108(1):44–53. 8. Winicki J, Jemison K. Food insecurity and hunger in the kindergarten (SD) of the WHO Child Growth Standards median classroom: its effect on learning and growth. Contemp Econ Policy. 2003; Wasting: weight for height < − 2 standard deviations https://doi.org/10.1093/cep/byg001. (SD) of the WHO Child Growth Standards median 9. Glewwe P, Jacoby HG, King EM. Early childhood nutrition and academic achievement: a longitudinal analysis. J Public Econ. 2001;81(3):345–68. Underweight: weight for age < − 2 standard deviations 10. Mosley WH, Gray R. Childhood precursors of adult morbidity and mortality (SD) of the WHO Child Growth Standards median in developing countries: implications for health programs. Epidemiol Transit. 1993;1:69–100. Acknowledgements 11. Shetty P. Malnutrition and undernutrition. Medicine. 2003;31(4):18–22. We would like to thank Professor Lindsay Brown for providing us a critical 12. Semba RD, de Pee S, Sun K, Sari M, Akhter N, Bloem MW. Effect of parental review and comments on the paper. formal education on risk of child stunting in Indonesia and Bangladesh: a cross-sectional study. Lancet. 2008;371(9609):322–8. (35) Availability of data and materials 13. Ricci JA, Becker S. Risk factors for wasting and stunting among children in The analysis was based on the Bangladesh Multiple Indicators Cluster Survey metro Cebu, Philippines. Am J Clin Nutr. 1996;63(6):966–75. (MICS) 2012–13 data conducted by Bangladesh Bureau of Statistics along 14. United Nations Children’s Fund. Improving Child Nutrition: The with UNICEF who provided technical and financial support, data is available Achievable Imperative for Global Progress. New York: United Nations on request at http://microdata.worldbank.org/index.php/catalog/2533 Children’sFund; 2013. 15. World Health Organization. Nutrition Landscape Information System (NLIS) Authors’ contributions country profile indicators: interpretation http://apps.who.int/iris/bitstream/ TRC managed the dataset and analyze the data and write the script. SC 10665/44397/1/9789241599955_eng.pdf conceptualized and synthesized the analysis, interpreted findings and helped 16. Walker SP, Grantham-McGregor SM, Powell CA, Chang SM. Effects of growth in writing. MR helped to interpret and outline the findings. SS and KD restriction in early childhood on growth, IQ, and cognition at age 11 to 12 helped in writing, editing and interpreting the results. All authors read and years and the benefits of nutritional supplementation and psychosocial approved the final manuscript. stimulation. J Pediatr. 2000;137(1):36–41. 17. Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ. Comparative risk Ethics approval and consent to participate assessment collaborating group. Selected major risk factors and global and Not applicable. regional burden of disease. Lancet. 2002;360(9343):1347–60. 18. Caulfield LE, de Onis M, Blössner M, Black RE. Undernutrition as an underlying cause of child deaths associated with diarrhea, pneumonia, Competing interests malaria, and measles. Am J Clin Nutr. 2004;80(1):193–8. The authors declare that they have no competing interests. 19. Wolde M, Berhan Y, Chala A. Determinants of underweight, stunting and wasting among schoolchildren. BMC Public Health. 2015; https://doi.org/10. Publisher’sNote 1186/s12889-014-1337-2. Springer Nature remains neutral with regard to jurisdictional claims in 20. United Nations Children’s Fund, World Health Organization, The World Bank. published maps and institutional affiliations. UNICEFWHO-World Bank Joint Child Malnutrition Estimates. New York, Geneva, Washington, DC: UNICEF, WHO, The World Bank; 2012. http://www. Author details who.int/nutgrowthdb/jme_unicef_who_wb.pdf. Accessed 26 May 2018. Department of Economics, Shahjalal University of Science & Technology, 21. National Institute of Population Research and Training (NIPORT). Bangladesh Sylhet Kumargaon, Sylhet 3114, Bangladesh. Institute for Resilient Regions, Demographic and Health Survey 2007. Available from https://www.unicef. University of Southern Queensland, QLD, Springfield 4300, Australia. School org/bangladesh/BDHS2007_Final.pdf. Accessed 26 May 2018. of Teacher Education and Early Childhood, University of Southern 22. Adeba A, Garoma S, Gemede HF, Garoma W. Prevalence of Stunting Queensland, Springfield Campus, Springfield, Australia. School of Medicine, and Associated Factors of Children among 6–59 Months Age in Guto University of California, Riverside, USA. Gida District, East Wollega Zone, Oromia, Ethiopia. Food Sci Qual Manag. 2014;29 Received: 19 February 2018 Accepted: 15 May 2018 23. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, underweight and overweight in children aged 2.0–4.9 years in Indonesia: prevalence trends and associated risk factors. PLoS One. 2016;11(5):e0154756. References 24. Mohsena M, Mascie-Taylor CN, Goto R. Association between socio-economic 1. Demissie S, Worku A. Magnitude and factors associated with malnutrition in status and childhood undernutrition in Bangladesh; a comparison of possession children 6–59 months of age in pastoral community of Dollo Ado District, score and poverty index. Public Health Nutr. 2010;13(10):1498–504. (A1) Chowdhury et al. Globalization and Health (2018) 14:54 Page 12 of 12 25. Islam MA, Rahman MM, Mahalanabis D. Maternal and socioeconomic factors 51. Kumar D, Goel NK, Mittal PC, Misra P. Influence of infant-feeding practices and the risk of severe malnutrition in a child: a case-control study. Eur J Clin on nutritional status of under-five children. Indian J Pediatr. 2006;73(5):417– Nutr. 1994;48(6):416–24. (12 IN A15) 21. (39 IN A13) 26. Siddiqi MN, Haque MN, Goni MA. Malnutrition of under-five children: 52. Guldan GS, Zeitlin MF, Beiser AS, Super CM, Gershoff SN, Datta S. Maternal evidence from Bangladesh. Asian J Med Sci. 2011;2(2):113–9. (A6) education and child feeding practices in rural Bangladesh. Soc Sci Med. 27. Nahar B, Ahmed T, Brown KH, Hossain MI. Risk factors associated with 1993;36(7):925–35. (A7) severe underweight among young children reporting to a Diarrhoea 53. Samuel Oyekale A. Factors explaining acute malnutrition among under- five treatment facility in Bangladesh. J Health Popul Nutr. 2010;1(32):476–83. children in Sub-Sahara Africa (SSA). Life Sci J. 2012;9(4):2101–7. Available from: http://www.lifesciencesite.com/lsj/life0904/314_11976life0904_2101_ 28. Devkota S, Panda B. Socioeconomic gradients in early childhood health: 2107.pdf. evidence from Bangladesh and Nepal. Int J Equity Health. 2016;15(1):78. 54. Bhutta ZA, Ahmed T, Black RE, Cousens S, Dewey K, Giugliani E, Haider BA, 29. Alom J, Quddus MA, Islam MA. Nutritional status of under-five children in Kirkwood B, Morris SS, Sachdev HP, Shekar M. What works? Interventions for Bangladesh: a multilevel analysis. J Biosoc Sci. 2012;44(5):525–35. maternal and child undernutrition and survival. Lancet. 2008;371(9610):417–40. (30) 30. Srinivasan CS, Zanello G, Shankar B. Rural-urban disparities in child nutrition 55. Keats A. Women’s schooling, fertility, and child health outcomes: Evidence in Bangladesh and Nepal. BMC Public Health. 2013;13(1):581. from Uganda’s free primary education Program. Mimeo: Wesleyan 31. Ahsan KZ, El Arifeen S, Al-Mamun MA, Khan SH, Chakraborty N. Effects of University; 2014. p. 31. individual, household and community characteristics on child nutritional 56. Janevic T, Petrovic O, Bjelic I, Kubera A. Risk factors for childhood status in the slums of urban Bangladesh. Archives Public Health 2017;75(1):9. malnutrition in Roma settlements in Serbia. BMC Public Health. 2010; 32. Das S, Rahman RM. Application of ordinal logistic regression analysis in https://doi.org/10.1186/1471-2458-10-509. determining risk factors of child malnutrition in Bangladesh. Nutr J. 2011; 57. Rayhan MI, Khan MS. Factors causing malnutrition among under five 10(1):124. children in Bangladesh. Pak J Nutr. 2006;5(6):558–62. (A4) 33. Das S, Gulshan J. Different forms of malnutrition among under five children 58. Bairagi R. Is income the only constraint on child nutrition in rural in Bangladesh: a cross sectional study on prevalence and determinants. Bangladesh? Bull World Health Organ. 1980;58(5):767. (34) BMC Nutrition. 2017 Jan 3;3(1):1. 59. Grossman M. On the concept of health capital and the demand for health. 34. Bangladesh Multiple Indicator Cluster Survey 2012-2013. http://microdata. J Polit Econ. 1972;80(2):223–55. (36) worldbank.org/index.php/catalog/2533/sampling. Accessed 26 May 2018. 60. Yimer G. Malnutrition among children in southern Ethiopia: levels and risk 35. Bangladesh Multiple Indicator Cluster Survey 2012-2013. ProgotirPathey: factors. Ethiop J Health Dev. 2000;14(3) (Yimer in A16) Final Report. Dhaka: Bangladesh Bureau of Statistics (BBS) and UNICEF 61. Hong R, Banta JE, Betancourt JA. Relationship between household wealth Bangladesh; 2014. inequality and chronic childhood under-nutrition in Bangladesh. Int J Equity 36. Alasfoor D, Traissac P, Gartner A, Delpeuch F. Determinants of persistent Health. 2006;5(1):15. (A14) underweight among children, aged 6-35 months, after huge economic 62. Casas JA, Dachs JN, Bambas A. Health disparities in Latin America and the development and improvements in health services in Oman. J Health Popul Caribbean: the role of social and economic determinants. Equity Health. Nutr. 2007:359–69. 2001;8(37):22–49. 37. Mukabutera A, Thomson DR, Hedt-Gauthier BL, Basinga P, Nyirazinyoye L, 63. Khor GL, Sharif ZM. Dual forms of malnutrition in the same households in Murray M. Risk factors associated with underweight status in children under Malaysia-a case study among Malay rural households. Asia Pac J Clin Nutr. five: an analysis of the 2010 Rwanda demographic health survey (RDHS). 2003;12(4):427–37. (38) BMC Nutrition. 2016; https://doi.org/10.1186/s40795-016-0078-2. 64. Larrea C, Freire W. Social inequality and child malnutrition in four Andean 38. Rahman A, Chowdhury S. Determinants of chronic malnutrition among countries. Rev Panam Salud Publica. 2002;11(5–6):356–64. (26 IN A14) preschool children in Bangladesh. J Biosoc Sci. 2007;39(2):161–73. (9 IN A15; 65. Wang Y, Monteiro C, Popkin BM. Trends of obesity and underweight in 10 IN 805) older children and adolescents in the United States, Brazil, China, and 39. Henry F, Briend A, Fauveav V, Huttly SR, Yunus M, Chakraborty J. The risk Russia. Am J Clin Nutr. 2002;75(6):971–7. (27 IN A14) approach to intervention in severe malnutrition in rural Bangladesh. Am J 66. Doak C, Adair L, Bentley M, Fengying Z, Popkin B. The underweight/ Epidemiol. 1992;136(4):460–3. (18 IN A15) overweight household: an exploration of household sociodemographic and 40. Henry FJ, Briend A, Fauveau V, Huttly SA, Yunus M, Chakraborty J. Gender dietary factors in China. Public Health Nutr. 2002;5(1a):215–21. (28 IN A14) and age differentials in risk factors for childhood malnutrition in 67. Sarma H, Khan JR, Asaduzzaman M, Uddin F, Tarannum S, Hasan MM, Bangladesh. Ann Epidemiol. 1993;3(4):382–6. (19 IN A15) Rahman AS, Ahmed T. Factors influencing the prevalence of stunting 41. HENRY F, Briend A, Fauveau V, HUTTLY SR, Yunus M, Chakraborty J. Risk among children aged below five years in Bangladesh. Food Nutr Bull. 2017; factors for clinical marasmus: a case-control study of Bangladeshi children. 1(805):0379572117710103. Int J Epidemiol. 1993;22(2):278–83. (20 IN A15) 68. De Bruyn T. Dynamics of remittance utilisation in Bangladesh. IOM research 42. Hossain MI, Yasmin R, Kabir I. Nutritional and immunisation status, weaning series, no. 18. Geneva: International Organization for Migration; 2005. p. 40. practices and socio-economic conditions of under five children in three 69. Mohsena M, Goto R, Taylor MNCG. Regional variation in maternal and villages of Bangladesh. Indian J Public Health. 1999;43(1):37–41. (21 IN A15) childhood undernutrition in Bangladesh: evidence from demographic and 43. Chisti MJ, Hossain MI, Malek MA, Faruque AS, Ahmed T, Salam MA. health surveys. WHO South-East Asia J Public Health. 2015;4:2. Characteristics of severely malnourished under-five children hospitalized with diarrhoea, and their policy implications. Acta Paediatr. 2007;96(5):693–6. (22 IN A15) 44. Strauss J, Thomas D. Human resources: empirical modeling of household and family decisions. Handb Dev Econ. 1995;3(26):1883–2023. 45. Glewwe P. Why does mother’s schooling raise child health in developing countries? Evidence from Morocco. J Hum Resour. 1999;1(27):124–59. 46. Tekce B, Shorter FC. Determinants of child mortality: a study of squatter settlements in Jordan. Popul Dev Rev. 1984;10(Suppl 1):257–80. (11 in A8) 47. Benyoussef A, Wessen AF. Utilization of health services in developing countries—Tunisia. Soc Sci Med (1967). 1974;8(5):287–304. (12 IN A8) 48. Caldwell JC, Reddy PH, Caldwell P. The social component of mortality decline: an investigation in South India employing alternative methodologies. Popul Stud 1983;37(2):185–205. (13 IN A8). 49. Bicego GT, Boerma JT. Maternal education and child survival: a comparative study of survey data from 17 countries. Soc Sci Med. 1993; 36(9):1207–27. (A8) 50. Handa S. Maternal education and child height. Econ Dev Cult Chang. 1999; 47(2):421–39. (28) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Globalization and Health Springer Journals

Socio-economic risk factors for early childhood underweight in Bangladesh

Free
12 pages

Loading next page...
 
/lp/springer_journal/socio-economic-risk-factors-for-early-childhood-underweight-in-xH0XwrhrJ0
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Public Health; Development Economics; Social Policy; Quality of Life Research; Epidemiology
eISSN
1744-8603
D.O.I.
10.1186/s12992-018-0372-7
Publisher site
See Article on Publisher Site

Abstract

Background: Underweight is a major cause of global disease burden. It is associated with child mortality and morbidity, and its adverse impact on human performance and child survival is well recognized. Underweight is a major public health problem in Bangladesh, which is amongst the highest underweight prevalent countries in the world. The objectives of our study were to determine the national and regional prevalence rates of underweight and severe underweight in Bangladesh, and to investigate the association of socioeconomic and demographic factors with child underweight and severely underweight among children under the age of five living in Bangladesh. Methods: We performed a cross sectional study using Multiple Indicators Cluster Survey 2012–13, Bangladesh data on 17,133 children under 5 years of age. Weight-for-age Z scores based upon World Health Organization (WHO) guidelines were used to define child underweight and severe underweight. The association of underweight and severe underweight with household socioeconomic factors and demographic characteristics was investigated using binary logistic regression model. Results: An estimated 31.67% children were underweight and 8.81% children were severely underweight. Children of mothers with incomplete secondary education [Odds Ratio (OR) = 0.84, 95% CI: 0.75, 0.94] and mothers with completed secondary education [Odds Ratio (OR) = 0.77, 95% CI: 0.64, 0.93] were less likely to be underweight than children of uneducated mothers who had no formal schooling. A similar association exists for father’s education, children from households in the highest wealth index quintile had lower likelihood of underweight [Odds Ratio (OR) = 0.44, 95% CI: 0.37, 0.53] than children from households in the lowest quintile. Consumption of non-iodized salt had higher risk of severe underweight for children aged between 24 and 35 months [Odds Ratio (OR) = 2.32, 95% CI: 1.80, 3.00]. Other risk factors of child severe underweight included living in Sylhet division and increases in the number of children under the age of five in a household. Conclusion: Underweight was associated with lower parental education, household position in lower wealth index, living in Sylhet division and consumption of non-iodized salt. Strategies are discussed considering the relative importance of risk factors for child underweight. Keywords: Determinants, Underweight, Bangladesh, Child, Early childhood Background children who are mild to moderately malnourished com- Developing countries have the greatest prevalence of pared with well-nourished children [5, 6]. The adverse early childhood malnutrition despite recent rapid eco- impacts of malnutrition are not limited to school attend- nomic development [1–3]. Malnutrition is the prime ance rates and academic accomplishments, but also culprit for early childhood death, causing the deaths of linked to decreases in the development of social skills 13 million infants and children under the age of five among children [7, 8]. Malnutrition is associated with every year [4]. There is a higher risk of dying among poorer academic performance in school-aged children [9] and lower survival capacity for a child in its adult- * Correspondence: tuhinchy.sust@gmail.com hood [10]. Department of Economics, Shahjalal University of Science & Technology, Sylhet Kumargaon, Sylhet 3114, Bangladesh Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Chowdhury et al. Globalization and Health (2018) 14:54 Page 2 of 12 Malnutrition status is investigated using three an- education. Mohsena et al. [24] explored the Bangladesh thropometric parameters: height for age, weight for Demographic and Health Survey (2004) Data and found height and weight for age [11]. Height for age is used that higher levels of maternal education are associated to identify stunting which represents linear growth with lower risk of underweight. In Bangladesh, paternal failure for a child [12]. Weight for height is used to education was also found to be an independent deter- identify wasting which indicates an unusually body tissue minant of child underweight by few studies [27, 33]. Ac- and fat mass for a child relative to its height or length cording to Das et al. [33], children of illiterate fathers [13]. Weight for age is the WHO-recommended indicator had a 30% higher likelihood of being underweight to determine whether the child is underweight or not than the children of fathers who had completed pri- [11]. As underweight is a composite indicator that encom- mary school. Another study that finds a positive asso- passes both stunting and wasting [14] implying that stunt- ciation between paternal education and risk of child ing, wasting or both can be reflected by underweight [15], underweight [26]. our primary interest is to use weight for age indicator to Socioeconomic status like poverty is also a signifi- measure early childhood underweight. cant predictor of child underweight. Children coming Underweight has become a major public health from poor families are more likely to be underweight problem particularly in developing nations, and its compared with children coming from wealthier fam- destructive impact on human performance and child ilies [4]. Despite the fact that Bangladesh has man- survival has been well recognized [16]. It is a prime aged to achieve an enormous economic growth in the concern as underweight is the leading cause of global last 20 years along with massive improvements in disease burden [17]. Underweight is associated with educational achievement, still a large proportion of its increased risk of contagious disease such as diarrhea population, nearly 43.3%, are living under the poverty and pneumonia [18]. The World Health Organization line of 1.25 dollar per day [28]. Having such a large (WHO)cites underweightas the single largest risk proportion of the population living in poverty, hence factor in developing countries to the burden of dis- unable to afford nutritious food or access improved ease, [19] as an estimated 52.5% of all deaths among health care facilities, makes the task of reducing the coun- young children are attributable to underweight [18]. try’s prevalence of underweight children extremely difficult. In the countries where child mortality is high, under- Indeed, a number of studies conducted in Bangladesh sug- weight is accountable for 15% of the total disability gested household socioeconomic status as a significant de- adjusted life years losses [17]. Underweight children terminant of child underweight, reporting that children are 8.4 times more likely to die and children who are from relatively higher socioeconomic condition had lower suffering from moderate and mildly underweight are likelihood of being underweight [28, 33]. A study in 4.6 times and 2.5 times respectively more likely to die Bangladesh carried out by Alom et al. [29]reportedthat before the age of five compared to children who are children who came from the richest wealth index had 44% well-nourished [5]. lower odds of being underweight compared with children In 2011, 101 million children globally were estimated as from the poorest wealth index. underweight indicating the prevalence rate of underweight Among the socioeconomic determinants, most of the as 16% [20]. The situation is worse in Bangladesh, where studies conducted in Bangladesh confirmed that mothers’ underweight prevalence rate was reported to be 41% [21]. literacy status, choice of latrine and poor household in- There are many socioeconomic and biological factors that come are the associated risk factors for underweight are associated with child underweight. Biological factors among children under five [25, 27, 29, 30]. Other risk fac- include lower birth weight [22], having an underweight tors estimated as significantly associated with the risk of mother or father, and having a mother or father of short child underweight are improper feeding practices, mater- stature. Among the socioeconomic factors, children with nal antenatal and postnatal care, diarrhea of children mothers who don’t have any formal education are at within last 2 weeks [32]; living in rural area [30]; mothers greater risk of being underweight [23]. having less exposure to media, children of mothers work- Even though Bangladesh has one of the highest preva- ing outside home [31]; unimproved toilet facilities, and liv- lence of underweight in the world, few studies have inginSylhet division[33]. The risk of being underweight identified the socioeconomic factors responsible for among under-five children was observed to increase with child underweight [24–30]. Most of the studies con- an increase in children’s age by several studies in ducted in Bangladesh suggested that maternal education Bangladesh. [29, 31, 33]. The causes for underweight are is a strong and significant predictor for child under- multidimensional and interrelated, and they often change weight [24, 28, 29, 31, 32]. A recent study by Devkota et from country to country, highlighting the importance of al. [28] established a positive association of WAZ local research studies. Therefore, the objectives of our (weight-for-age z score) with the level of maternal study is to determine the national and regional prevalence Chowdhury et al. Globalization and Health (2018) 14:54 Page 3 of 12 of underweight and severe underweight and their associ- The first stage of sampling involved determining census ated risk factors in Bangladesh. enumeration areas required within each district, which were defined as primary sampling units. From 30 No- vember, 2012 to 6 December, 2012, field operations were Method carried out in each enumeration area to make new list- Data sources ings of households; primarily because the 2011 census The analysis was based on the Bangladesh Multiple Indi- frame was not up-to-date. Experienced staff of the cators Cluster Survey (MICS) 2012–13 data conducted Bangladesh Bureau of Statistics (BBS) prepared the list by Bangladesh Bureau of Statistics along with UNICEF of households in the field, and at the BBS headquarter, who provided technical and financial support. Data were the listed households were numbered sequentially from collected from December 2012 to April 2013. The major 1 to n within each enumeration area. Here, Random Sys- objectives of the survey were to collect the most recent tematic Sample Selection Procedure was used to select data on women and child health, household socioeco- 20 households in each enumeration area. Households in- nomic status, access to mass media, ICT improved sani- cluded 23,402 children below 5 years of age and infor- tation and pure water, child nutrition, mortality and mation on 20,903 under five children was collected, child development. The sample was selected using indicating a response rate of 89.3%. For the purpose of two-staged stratified cluster sampling approach. In the data collection, structured questionnaires were used. time period during which data were collected, there The questionnaires had four sections comprised of ques- were seven divisions and 64 districts in Bangladesh and tions pertaining to household, women (age 15–49), chil- districts were defined as sampling strata [34]. The pro- dren under five and water quality testing. Thirty-two portion of women with 4+ ANC visits, where the term data collection teams collected data, with each team hav- ANC stands for antenatal care visit, was used as key in- ing seven members including a supervisor and inter- dicator in estimating the size of sample. The sample size viewers. The teams were provided with training for was determined using the following formula n= 14 days prior to collecting data. Recommended an- ½4ðrÞð1−rÞðdeff Þ , where 4 = factor for achieving 95% con- thropometric tools were used to measure child weight ½e ðpbÞðAve SizeÞðRRÞ fidence level; r = projected value of the indicator, which and height [34]. Collected data from the field were en- was assumed to be 26% and expressed as a proportion; tered using CSPro software and all questionnaires were n = sample size; deff = design effect for the indicator, and entered twice to ensure the quality of data [35]. the value of deff was estimated to be 1.4 based on previ- ous surveys; Ave Size = average household size, and it Statistical analysis was assumed as 4.5; pb refers to the percentage of There were 20,903 children under age 5 years about women in the total population, who gave birth in the last whom information was collected. However, 982 children 2 years, and it was taken to be 4%; RR = anticipated re- were not included as there was no information for sponse rate which was assumed to be 95% based on the weight-for-age-z score. For the households with more previous survey experience [35]. In estimating r, max- than one child under the age of five, the youngest one imum margin of error was allowed of ±8 to 9%, and was selected as the index child and the other children in thereby a sample size of 800 was decided for each dis- the household were not considered for the study. Finally, trict. For obtaining a more precise estimation result, data pertaining to 17,133 children below the age of five 1000 samples were drawn from each of the 20 UNDAF were subject to analysis. Weight-for-age-z score of districts. UNDAF (United Nation Development Assist- WHO growth standard was used to create the under- ance Framework) districts are those 20 districts that weight variable. Children were defined to be under- have been declared as prioritized districts by Bangladesh weight if their weight-for-age-z scores were below minus government. Twenty households per cluster were se- two and severely underweight if their weight-for-age-z lected for the survey, taking in consideration various fac- scores were below minus three. Data analysis was per- tors such as time, budget, design effect etc. Twenty formed using Stata version 12.1. Child, paternal and households per cluster suggest that for drawing 1000 household socioeconomic factors were included in the households as sample, 50 clusters would be needed in analysis as predictors of child underweight. Factors in- each of the 20 UNDAF districts, and 40 cluster would be cluded child’s age, gender, geographic area, education needed for drawing 800 households as sample size in level of mother, education level of father, regional div- each of the remaining 44 non UNDAF districts. In this ision, wealth index quintile used as the proxy indicator way, the resulting sample size would be 55,200. But 4 of household socioeconomic status, number of house- clusters were not visited due to unfavorable weather hold members, toilet facility, salt iodization test, whether and, therefore, data was collected from 55,120 house- the child drank plain water yesterday (yes/no), whether holds. Clusters were selected using 2011 census frame. the child drank milk yesterday (yes/no), whether the Chowdhury et al. Globalization and Health (2018) 14:54 Page 4 of 12 child took vitamins yesterday (yes/no) and number of The analysis determined the following factors to be as- children in a household below the age of five. Wealth sociated with the risk of underweight: age of children, index score was computed by principal component ana- division, education level of mother and education level lysis based on household assets and materials used to of father, wealth index, whether the child drank plain build house [35]. All households were then categorized water yesterday and finally the use of iodized salt by into five categories depending on wealth score such as households. Children living in Khulna division were 15% poorest, second, middle, fourth and richest. Toilet facil- less likely to be underweight [Odds Ratio (OR) = 0.85, ity variable was created with four categories as flush la- 95% CI: 0.75, 0.96] compared with the children living in trine, pit latrine, hanging latrine and others. Descriptive Dhaka division. Children from Sylhet division were more statistics were created for all predictors as well as for re- prone to the risk of underweight than any other division sponse variable. As our dependent variable was dichot- in Bangladesh with higher odds on underweight [Odds omous, underweight versus not underweight and Ratio (OR) = 1.38, 95% CI: 1.20, 1.60] than the children severely underweight versus not severely underweight, from Dhaka division. Our results suggested child’s age as predictors for the risk of child underweight and severely one of the major determinants of underweight. Children underweight were investigated using binary logistic re- aged 12–23 months [Odds Ratio (OR) = 1.98, 95% CI: gression model. Statistical significance was set at p < 0.05 1.69, 2.32] had around two-fold odds on underweight and adjusted odds ratio was reported along with 95% compared with children below 5 months of age. The risk confidence interval. for underweight was highest among the children who were in their early middle childhood. Children aged 24– Results 35 months [Odds Ratio (OR) = 2.58, 95% CI: 2.20, 3.03] Socioeconomic and demographic characteristics of the had over two-fold risk on underweight relative to the study population children less than 5 months of age. Children between The final analysis of 17,133 children estimated that 31.67% the ages of 36–47 months [Odds Ratio (OR) = 2.44, 95% of these children were underweight including 8.81% as se- CI: 2.07, 2.88] and children between the ages of 48– verely underweight. Table 1 demonstrated socioeconomic 59 months [Odds Ratio (OR) = 2.35, 95% CI: 1.98, 2.77] and demographic characteristics of the study population. were more likely to be underweight compared with the The male to female ratio of children was 1069:1000. children who were below 5 months of age. Children of More than four-fifths of children came from rural area secondary incomplete mothers had 16% smaller odds on and around two-thirds of the surveyed children belonged underweight [Odds Ratio (OR) = 0.84, 95% CI: 0.75, to the households with four to six family members. 0.94] than the children of mothers with no formal edu- More than one-fifth of the children were between the cation. There were 23% lower odds on underweight ages of 12 and 23 months, and approximately 11% of among children whose mothers completed secondary children were aged less than 5 months. The highest education [Odds Ratio (OR) = 0.77, 95% CI; 0.64, 0.93] (25.36%) proportion of children lived in Dhaka division than the children whose mothers had not attended with the lowest from Sylhet division (8.21%). More than school at all. As with mother’s education, an increase in one-third of fathers did not go to school and likewise the level of father’s education reduced the odds on child one-fourth of mothers did not have any formal educa- underweight. Children with father’s education secondary tion. Only 11.37% of mothers and 15.38% of fathers incomplete and secondary complete had 12% [Odds Ra- completed secondary education. Nearly, one third of the tio (OR) = 0.88, 95% CI: 0.79, 0.98] and 28% [Odds Ratio study children belonged to households of the poorest (OR) = 0.72, 95% CI: 0.62, 0.84] lower odds on under- wealth quintile (29.06%) and only 13.87% children weight respectively in comparison with children whose belonged to richest wealth quintile. On the previous day, fathers had no formal education. Children who didn’t approximately 86% children drank plain water 24% drink plain water yesterday were 1.27 times more likely drank milk and 9.06% ate or drank vitamins. More than to be underweight [Odds Ratio (OR) = 1.27, 955 CI: 1.12, half of the households have iodized salt (51.43%). The 1.43] compared with children who drank water the pre- majority of households had access to pit latrine (62.34%) vious day. and almost one-third of households had access to flush Wealth index was also related to the risk of under- latrines. weight. Children belonging to households which were in the middle category [Odds Ratio (OR) = 0.82, 95% CI: Risk factors for underweight 0.73, 0.91] and in fourth category [Odds Ratio (OR) = Table 2 lists the socioeconomic and demographic factors 0.67, 95% CI: 0.59, 0.77] were less likely to be under- along with their association and the magnitude of im- weight compared with the children who belonged to the pact on the risk for underweight among children below poorest households in the wealth index. There were 56% 5 years old using binary logistic regression analysis. lower odds on underweight [Odds Ratio (OR) = 0.44, Chowdhury et al. Globalization and Health (2018) 14:54 Page 5 of 12 Table 1 Socioeconomic and demographic characteristics of the Table 1 Socioeconomic and demographic characteristics of the study population study population (Continued) Total sample Total sample (n = 17133) (n = 17133) Characteristics Proportion (%) Characteristics Proportion (%) Household member Fourth 16.52 <=3 14.29 Richest 13.87 4–6 63.64 Child drank plain water yesterday > = 7 22.06 Yes 85.46 Age (months) No 14.54 0–5 10.93 Child drank milk yesterday 6–11 10.80 Yes 23.44 12–23 21.88 No 76.56 24–35 20.60 Child ate vitamins yesterday 36–47 18.75 Yes 9.06 48–59 17.04 No 90.94 Gender Toilet facility Male 51.67 Flush latrine 23.94 Female 48.33 Pit latrine 62.34 Area Hanging latrine 7.78 Urban 15.87 Others 5.93 Rural 84.13 Salt idolization test Division Not iodized 0 PPM 27.15 Dhaka 25.36 More than 0 PPM and less than 15 PPM 20.49 Barisal 9.64 15 PPM or more 51.43 Chittagong 19.53 No salt in the household 0.92 Khulna 13.96 Child underweight status Rajshahi 10.28 Underweight 31.67 Rangpur 13.01 Severely underweight 8.81 Sylhet 8.21 Education of mother None 24.10 95% CI: 0.37, 0.53] among the children who came from Primary incomplete 14.55 richest households than the children who came from Primary complete 15.39 poorest households. Children living in households where salt was found as adequately iodized (15 ppm or more) Secondary incomplete 34.59 had a lower likelihood of underweight [Odds Ratio (OR) Secondary complete or higher 11.37 = 0.90, 95% CI: 0.82, 0.98] compared with children from Education of father those households where salt was found to be not iodized None 34.79 in salt iodization tests. Our study did not establish a re- Primary incomplete 15.73 lationship between child gender and underweight, nor Primary complete 13.75 did it establish that children from rural areas were more likely to be underweight compared with children from Secondary incomplete 20.35 urban areas. Secondary complete or higher 15.38 Wealth index quintile Risk factors for severe underweight Poorest 29.06 Table 3 lists the results of the binary logistic regression Second 21.86 analysis on the risk for child severe underweight associ- Middle 18.69 ated with different socioeconomic and demographic factors. Chowdhury et al. Globalization and Health (2018) 14:54 Page 6 of 12 Table 2 Risk factors for total underweight in children aged less than 5 years Factors Underweight (%) N OR 95% CI Household member <=3 33.24 2449 ref. 4–6 31.54 10,904 0.91 (0.82–1.00) > = 7 31.03 3780 1.00 (0.88–1.13) Age (Months) 0–5 21.10 1872 ref. 6–11 21.88 1851 1.17 (0.98–1.39) *** 12–23 31.05 3749 1.98 (1.69–2.32) *** 24–35 36.91 3530 2.58 (2.20–3.03) *** 36–47 35.62 3212 2.44 (2.07–2.88) *** 48–59 34.77 2919 2.35 (1.98–2.77) Gender Male 32.15 8853 ref. Female 31.16 8280 0.96 (0.90–1.03) Area Urban 26.52 2719 ref. Rural 32.64 14,414 0.96 (0.86–1.07) Division Dhaka 30.29 4345 ref. Barisal 33.96 1652 1.04 (0.91–1.19) Chittagong 33.59 3346 1.11 (0.99–1.24) ** Khulna 26.55 2392 0.85 (0.75–0.96) Rajshahi 30.14 1762 0.90 (0.79–1.02) Rangpur 32.17 2229 0.98 (0.87–1.11) *** Sylhet 38.52 1407 1.38 (1.20–1.60) Education of mother None 39.55 4129 ref. Primary incomplete 37.18 2493 1.02 (0.91–1.14) Primary complete 34.36 2637 0.96 (0.85–1.07) ** Secondary incomplete 27.19 5926 0.84 (0.75–0.94) ** Secondary complete or higher 17.92 1948 0.77 (0.64–0.93) Education of father None 38.25 5428 ref. Primary incomplete 35.66 2454 0.98 (0.88–1.09) Primary complete 33.15 2145 0.97 (0.87–1.09) Secondary incomplete 27.81 3175 0.88 (0.79–0.98) *** Secondary complete or higher 19.30 2399 0.72 (0.62–0.84) Wealth index quintile Poorest 40.11 4979 ref. Second 35.54 3745 0.90 (0.82–0.99) *** Middle 30.78 3203 0.82 (0.73–0.91) *** Fourth 25.48 2830 0.67 (0.59–0.77) *** Richest 16.46 2376 0.44 (0.37–0.53) Child drank plain water yesterday Yes 32.11 14,619 ref. *** No 29.03 2487 1.27 (1.12–1.43) Child drank milk yesterday Yes 26.94 4013 ref. No 33.10 13,105 1.07 (0.98–1.17) Child ate vitamins yesterday Yes 28.69 1551 ref. No 31.97 15,559 0.96 (0.84–1.09) Toilet facility Flush latrine 24.46 4100 ref. Pit latrine 32.53 10,677 1.04 (0.94–1.16) Hanging latrine 41.86 1333 1.11 (0.95–1.30) Chowdhury et al. Globalization and Health (2018) 14:54 Page 7 of 12 Table 2 Risk factors for total underweight in children aged less than 5 years (Continued) Factors Underweight (%) N OR 95% CI Others 38.29 1016 1.02 (0.86–1.22) Salt iodization test Not iodized 0 PPM 36.54 4619 ref. More than 0 PPM and less than 15 PPM 34.77 3486 0.96 (0.87–1.06) 15 PPM or more 27.90 8750 0.90 (0.82–0.98) No salt in the household 29.30 157 0.80 (0.55–1.16) Number of under five children 1.02 (0.93–1.11) *** ** * p < 0.001; p < 0.01; p < 0.05 Age of children was associated with the risk for child had 21% lower odds of being severely underweight when severe underweight with children aged 12–23 months their fathers had enrolled in secondary education compare having almost twice the odds on severe underweight to children whose fathers had no formal education [Odds [Odds Ratio (OR) = 1.81, 95% CI: 1.40, 2.34] than the Ratio (OR) = 0.79, 95% CI: 0.65, 0.95]. There were 34% children aged less than 5 months. Children aged 24– lower odds of severely underweight among children with 25 months had the highest risk for being severely under- fathers who had completed secondary education [Odds Ra- weight. Compared with children aged below 5 months, tio (OR) = 0.66, 95% CI: 0.50, 0.88] compared with children children aged 24–35 months [Odds Ratio (OR) = 2.32, whose fathers had no formal education. An increase in the 95% CI: 1.80, 3.00] had a more than two-fold risk of se- number of children under the age of five in households by vere underweight. Furthermore, children between the 1 was associated with 17% higher odds on severe under- ages of 48–59 months had also larger odds of severe weight [Odds Ratio (OR) = 1.17, 95% CI: 1.02, 1.33]. underweight [Odds Ratio (OR) = 1.48, 95% CI: 1.12, 1.95] compared with their under-five counterparts. Chil- Discussion dren from the poorest households were at higher risk for Our study estimated 31.67% children below 5 years of age being severely underweight. Children from households as underweight, with 8.8% children as severely under- belonging to the middle category in wealth index had weight in 2011–2012 in Bangladesh. The present study 20% lower likelihood of underweight [Odds Ratio (OR) findings indicated a concerning nutritional condition = 0.80, 95% CI: 0.67, 0.96] and children from households among children aged under 5 years, especially those aged that belonged to fourth category in the wealth index had 2–3 years. Child underweight and severely underweight 49% lower odds of underweight [Odds Ratio (OR) = 0.51, were predicted by child’s age, regional division, education 95% CI: 0.40, 0.64] compared with the children from the level of mother, education level of father, household pos- poorest households. In comparison with the children ition in wealth index, whether child drank plain water yes- from the poorest households, children from the wealthi- terday, and the presence of iodized salt in households. est households had 68% lower odds of severe under- Our study reveals maternal education as an important weight [Odds Ratio (OR) = 0.32, 95% CI: 0.23, 0.46]. determinant of child underweight. According to our There were 32% lower odds on severe underweight study findings, higher levels of maternal education are among children living in the Khulna division [Odds Ra- associated with lower odds of child underweight, which tio (OR) = 0.68, 95% CI: 0.55, 0.85] and 34% higher odds was supported by previous studies [24, 36, 37]. Likewise, on severe underweight among children living in the Syl- maternal education was estimated to be an independent het division [Odds Ratio (OR) = 1.34, 95% CI: 1.09, 1.66] predictor of children being severely underweight. Chil- compared with children living in the Dhaka division. Im- dren of secondary completed mothers had 31% lower proved mother’s education and father’s education levels odds of being severely underweight compared with chil- were shown to be important protective factors against dren of uneducated mothers. Similar findings have been child severe underweight. Children with mothers’ pri- observed in earlier studies carried out in Bangladesh, mary education complete [Odds Ratio (OR) = 0.79, 95% which have shown that children of uneducated mothers CI: 0.66, 0.95] and secondary incomplete [Odds Ratio were significantly more likely to be severely underweight (OR) = 0.78, 95% CI: 0.65, 0.92] were less likely to be se- than children whose mothers had a relatively higher verely underweight compared with children whose level of education [25, 38–43]. There are some plausible mothers had no formal education. Children of mothers explanations as to why higher maternal education re- who had completed secondary education had 31% lower duces the chance of underweight and severely under- odds [Odds Ratio (OR) = 0.69, 95% CI: 0.49, 0.97] of being weight among children under the age of five. First, severely underweight compared with children whose education improves information processing capacity by mothers had not completed secondary education. Children the mother [44, 45]. For instance, an educated mother Chowdhury et al. Globalization and Health (2018) 14:54 Page 8 of 12 Table 3 Risk factors for severe underweight in children aged less than 5 years Factors Severe Underweight (%) N OR 95% CI Household member <=3 8.82 2449 4–6 8.69 10,904 0.89 (0.75–1.05) > = 7 9.13 3780 0.98 (0.80–1.20) Age (Months) 0–5 6.57 1872 6–11 6.86 1851 1.25 (0.94–1.66) *** 12–23 9.02 3749 1.81 (1.40–2.34) *** 24–35 11.16 3530 2.32 (1.80–3.00) *** 36–47 9.74 3212 2.04 (1.56–2.65) *** 48–59 7.33 2919 1.48 (1.12–1.95) Gender Male 9.25 8853 Female 8.33 8280 0.90 (0.80–1.00) Area Urban 6.91 2719 Rural 9.16 14,414 0.96 (0.80–1.15) Division Dhaka 8.63 4345 Barisal 9.87 1652 1.06 (0.86–1.32) Chittagong 10.10 3346 1.06 (0.89–1.26) ** Khulna 5.60 2392 0.68 (0.55–0.85) Rajshahi 7.72 1762 0.86 (0.69–1.06) Rangpur 8.03 2229 0.82 (0.67–1.00) ** Sylhet 13.08 1407 1.34 (1.09–1.66) Education of mother None 12.79 4129 Primary incomplete 10.87 2493 0.93 (0.78–1.10) Primary complete 8.99 2637 0.79 (0.66–0.95) ** Secondary incomplete 6.77 5926 0.78 (0.65–0.92) Secondary complete or higher 3.70 1948 0.69 (0.49–0.97) Education of father None 11.81 5428 Primary incomplete 9.98 2454 0.93 (0.79–1.09) Primary complete 9.60 2145 1.02 (0.86–1.22) Secondary incomplete 6.46 3175 0.79 (0.65–0.95) ** Secondary complete or higher 3.96 2399 0.66 (0.50–0.88) Wealth index quintile Poorest 12.59 4979 Second 10.28 3745 0.93 (0.80–1.08) Middle 8.30 3203 0.80 (0.67–0.96) *** Fourth 5.55 2830 0.51 (0.40–0.64) *** Richest 3.11 2376 0.32 (0.23–0.46) Child drank plain water yesterday Yes 8.65 14,619 *** No 9.77 2487 1.45 (1.21–1.73) Child drank milk yesterday Yes 7.28 4013 No 9.28 13,105 0.94 (0.81–1.09) Child ate vitamin yesterday Yes 7.67 1551 No 8.93 15,559 1.01 (0.81–1.25) Toilet facility Flush latrine 6.44 4100 Pit latrine 8.73 10,677 0.91 (0.76–1.08) Hanging latrine 14.40 1333 1.01 (0.80–1.28) Chowdhury et al. Globalization and Health (2018) 14:54 Page 9 of 12 Table 3 Risk factors for severe underweight in children aged less than 5 years (Continued) Factors Severe Underweight (%) N OR 95% CI Others 11.81 1016 0.96 (0.73–1.26) Salt iodization test Not iodized 0 PPM 10.46 4619 More than 0 PPM and less than 15 PPM 10.33 3486 1.04 (0.89–1.21) 15 PPM or more 7.38 8750 0.94 (0.82–1.08) No salt in the household 7.01 157 0.79 (0.42–1.47) Number of under five children 1.17 (1.02–1.33) *** ** * p < 0.001; p < 0.01, p < 0.0 can read, interpret and apply health tips from newspa- more knowledge of improved child care practices [53], pers and other sources, and is more likely to understand are better able to take responsibility for their children medical information such as medication levels suggested during illness [54], and more likely to have greater con- by doctor. These skills enable mothers to make informed trol over household decision-making [55]. Even though decisions about their children’s nutrition and health care, we do not know exactly why increased maternal educa- leading to improved child health. tion is associated with lower likelihood of child under- Second, maternal education is associated with an in- weight, the aforementioned result underscore the need crease in the frequency of mothers’ antenatal care visits for female education to reduce underweight prevalence and their utilization of child health services, leading to among children. decreased likelihood of underweight and severely under- Similar to maternal education, paternal education was a weight children. It has been evident from studies that significant predictor of underweight and severe under- the more educated mothers are the more likely they are weight. The result was supported by other studies that to access child health facilities more regularly compared suggested low paternal education was associated with in- to less educated mothers [46–48]. Antenatal care is often creased odds of child underweight [26, 27, 56, 57]. Educa- associated with improved child nutrition [26], and a tion level of fathers is generally associated with household study that employed data from 17 countries, aiming at income due to the fact that more educated fathers are determining the relationship between maternal educa- more likely to earn more money; indeed, father’seduca- tion and child survival, found that uneducated women tion is typically used as a proxy indicator of household in- represented 55% more non-use of antenatal services come [58]. Data on Indonesian children suggested that than their educated counterparts [49]. preventive factors against child malnutrition, such as use Third, there is general recognition that more educated of improved latrine, vitamin A capsules for children, ac- mothers are more likely to earn more money [44, 50]. cess to health facilities from local health community This implies increased capability to invest in their chil- clinics, and consumption of iodized salt, were strongly re- dren’s health, and more possibilities for living in an en- lated with higher level of paternal education [12]. There is vironment with relatively higher food security. Finally, strong evidence that more educated men are more likely maternal education improves child nutritional status to marry more educated women. As a result, children of through its impact on appropriate feeding practices by more educated fathers have an increased chance of having mothers. Appropriate feeding practices have been docu- mothers who are also more educated. The view is sup- mented as very crucial for improving nutrition condi- ported by Grossman’s model of the demand for good tions of children; particularly during their infancy [51], health which claims that more efficient investment in and a study conducted in Bangladesh revealed the asso- health is associated with more education [59]. ciation between improved feeding practices with educa- Another factor estimated to be significantly associated tion level of mother by demonstrating their findings with child under-nutrition is the age of the children. that; compared with uneducated mothers, educated Children below 5 months of age had the lowest risk of mothers were more concerned to breastfeed their chil- being underweight. One possible explanation could be dren in protected place along with cleanness; that they that during this period children are still being breastfed, were more aware to take breaks from the work at hand thereby gaining essential nutrition that renders them less while breastfeeding their children and were more con- vulnerable to malnutrition. On the other hand, study re- scious ensuring hygiene in the time children were pro- sults indicate that children’s risk of being underweight vided food [52]. Furthermore, higher level of education, increased with age, this risk reaching its peak between completed by mothers, is associated with behavioral the ages of 24–35 months. This suggests that during that changes that improve child health. Women who have period of children’s development, they had more than completed secondary education are more likely to have double the risk for underweight than the children who Chowdhury et al. Globalization and Health (2018) 14:54 Page 10 of 12 were less than 5 months of age. The similar result was media and poor maternal health were responsible for in- reported by a study which found the risk of being under- creased risk and higher prevalence of underweight among weight and severely underweight as highest between the under five children in the Sylhet division [4]. ages of 22–24 months [4]. Another study conducted The study had some limitations that should be ad- with Ethiopian children also supports our findings, dressed. First of all, the study was cross-sectional in de- which determined the association of the odds on child sign indicating that it was impossible to establish any underweight and severely underweight with age of chil- causal relationship between underweight and the esti- dren [60]. The reason for increased risk of underweight mated risk factors. Nevertheless, the findings have great associated with age may be that as children’s age in- significance as it is claimed that the results from the creased, they consumed complementary food along with same specific area using cross-sectional data have very breast milk. Contamination is more likely in comple- limited variation over time. Secondly, household income mentary food, therefore implicated in spreading infec- is a well-established determinant of child underweight. tious diseases among children and in turn rendering But it is nearly impossible to have reliable data on them vulnerable to the risk of underweight [61]. household income and expenditure, particularly in de- Moreover, there is strong evidence that the prevalence veloping countries such as Bangladesh. That being the of childhood disease is associated with child age, particu- case, we elected to use wealth index as a proxy indicator larly in most of the developing countries [62]. Children of household income, bearing in mind that wealth index living in higher wealth index households had smaller risk might not sufficiently represent household income. of being underweight and the finding was supported by Finally, data did not include child birth weight, other research work that claimed that children in lower mother’s weight, father’s weight, parents’ age, father’s oc- wealth index households had greater chance of under- cupation, or mother’s occupation. The magnitude of esti- weight than the children in higher wealth index house- mated risk factors could possibly be changed if those holds [56, 61, 63–66]. The explanation could be that factors were adjusted for our regression model. Despite the children in higher wealth index households were more study’s limitations, its strength was in assessing a large likely to belong to relatively food secure families, had number of samples, hence the findings can be generalized comparatively more educated fathers and mothers, and to all Bangladeshi children below 5 years old in 2011– lived in better neighborhood with improved health facil- 2012. This group of children needs to be re-examined in 5 ities. These factors might possibly have reduced the risk and 10 years to determine whether underweight and severe of underweight among the children living in higher underweight leads to lasting detriments for these children. wealth index households. It is widely known that mater- nal knowledge on child health has profound impact on Conclusion child well nourishment, and a study performed in Overall, in Bangladesh approximately one-third of the Bangladesh documented that mothers with better child study population is estimated as underweight. Under- health knowledge had mostly come from the richest weight prevalence not only reflects poor nutritional con- households [67]. It implies the significance of household ditions for children in a country, in general, it also economic condition on child nutrition through im- reflects high socio-economic costs for low quality of life, proved maternal knowledge on child health. Respon- vulnerability to different diseases and greater risk of dents who reported their children had drunk plain water childhood death. Due to the fact that it is more likely yesterday had lower odds on underweight for their chil- that underweight children who survive to adulthood will dren but there was no relationship between underweight experience greater health problems, the prevalence of and the type of toilet facilities. The risk of underweight underweight may also contribute to loss of productivity. was similar between male and female children, and be- The study revealed that the Sylhet division has the tween rural and urban children, which was unexpected. highest risk of severe underweight for under five chil- One explanation for this may be that poor people from dren, followed by the Chittagong division. With both di- rural areas were migrating to urban slum areas where visions their poverty rates are among the lowest in the economic conditions and health facilities were not re- country according to the latest Bangladesh Poverty markably different from rural areas. The risk for children Maps. This paper reveals that other than household pos- to be underweight depended on the division they lived ition in the lower wealth index, lower parental education in. Children from the Sylhet division had higher probability is one of the most important risk factors for child under- of underweight relative to children from the Dhaka division. weight. Sylhet is the home of most Bangladeshis that mi- The reason is not clear why children living in the Sylhet grate to the United Kingdom and are engaged in unskilled division were at more risk of underweight. Nevertheless, or semi-skilled blue-collar jobs and send remittances. Al- one possible reason might be that poor socioeconomic con- though remittances have improved land and housing in ditions such as few skilled workers, limited exposure to Sylhet, investment in other social sectors like education or Chowdhury et al. Globalization and Health (2018) 14:54 Page 11 of 12 business is reported to be insignificant [68]. In terms of Somali region, Ethiopia. Sci J Public Health. 2013; https://doi.org/10.11648/j. sjph.20130104.12. other important social indicators like ‘parental education’ 2. Masibo PK, Makoka D. Trends and determinants of undernutrition among Sylhet is reported to be poor-performing, according to the young Kenyan children: Kenya demographic and health survey; 1993, 1998, Bangladesh Bureau of Statistics, the overall literacy rate in 2003 and 2008–2009. Public Health Nutr. 2012; https://doi.org/10.1017/ S1368980012002856. the Sylhet division was the lowest in the country, at 39.2% 3. Pasricha SR, Biggs BA. Undernutrition among children in south and South- in 2007, with an even lower female literacy rate of 35.1% East Asia. J Paediatr Child Health. 2010; https://doi.org/10.1111/j.1440-1754. [69]. This result calls for government and non-government 2010.01839.x. 4. Rahman A. Significant risk factors for childhood malnutrition: evidence from interventions to increase education nationwide as much as an Asian developing country. Sci J Public Health. 2016; https://doi.org/10. possible, especially for the Sylhet division. This will also 11648/j.sjph.s.2016040101.14. markedly increase the mother’s age when children are 5. Pelletier DL, Frongillo EA Jr, Schroeder DG, Habicht JP. A methodology for estimating the contribution of malnutrition to child mortality in developing born, and decrease family size, which are particularly im- countries. J Nutr. 1994;124(10):2106S. portant for reducing the prevalence of underweight 6. Schroeder DG, Brown KH. Nutritional status as a predictor of child survival: children. summarizing the association and quantifying its global impact. Bull World Health Organ. 1994;72(4):569. 7. Alaimo K, Olson CM, Frongillo EA. Food insufficiency and American school- Endnotes aged children’s cognitive, academic, and psychosocial development. Stunting: height for age < − 2 standard deviations Pediatrics. 2001;108(1):44–53. 8. Winicki J, Jemison K. Food insecurity and hunger in the kindergarten (SD) of the WHO Child Growth Standards median classroom: its effect on learning and growth. Contemp Econ Policy. 2003; Wasting: weight for height < − 2 standard deviations https://doi.org/10.1093/cep/byg001. (SD) of the WHO Child Growth Standards median 9. Glewwe P, Jacoby HG, King EM. Early childhood nutrition and academic achievement: a longitudinal analysis. J Public Econ. 2001;81(3):345–68. Underweight: weight for age < − 2 standard deviations 10. Mosley WH, Gray R. Childhood precursors of adult morbidity and mortality (SD) of the WHO Child Growth Standards median in developing countries: implications for health programs. Epidemiol Transit. 1993;1:69–100. Acknowledgements 11. Shetty P. Malnutrition and undernutrition. Medicine. 2003;31(4):18–22. We would like to thank Professor Lindsay Brown for providing us a critical 12. Semba RD, de Pee S, Sun K, Sari M, Akhter N, Bloem MW. Effect of parental review and comments on the paper. formal education on risk of child stunting in Indonesia and Bangladesh: a cross-sectional study. Lancet. 2008;371(9609):322–8. (35) Availability of data and materials 13. Ricci JA, Becker S. Risk factors for wasting and stunting among children in The analysis was based on the Bangladesh Multiple Indicators Cluster Survey metro Cebu, Philippines. Am J Clin Nutr. 1996;63(6):966–75. (MICS) 2012–13 data conducted by Bangladesh Bureau of Statistics along 14. United Nations Children’s Fund. Improving Child Nutrition: The with UNICEF who provided technical and financial support, data is available Achievable Imperative for Global Progress. New York: United Nations on request at http://microdata.worldbank.org/index.php/catalog/2533 Children’sFund; 2013. 15. World Health Organization. Nutrition Landscape Information System (NLIS) Authors’ contributions country profile indicators: interpretation http://apps.who.int/iris/bitstream/ TRC managed the dataset and analyze the data and write the script. SC 10665/44397/1/9789241599955_eng.pdf conceptualized and synthesized the analysis, interpreted findings and helped 16. Walker SP, Grantham-McGregor SM, Powell CA, Chang SM. Effects of growth in writing. MR helped to interpret and outline the findings. SS and KD restriction in early childhood on growth, IQ, and cognition at age 11 to 12 helped in writing, editing and interpreting the results. All authors read and years and the benefits of nutritional supplementation and psychosocial approved the final manuscript. stimulation. J Pediatr. 2000;137(1):36–41. 17. Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ. Comparative risk Ethics approval and consent to participate assessment collaborating group. Selected major risk factors and global and Not applicable. regional burden of disease. Lancet. 2002;360(9343):1347–60. 18. Caulfield LE, de Onis M, Blössner M, Black RE. Undernutrition as an underlying cause of child deaths associated with diarrhea, pneumonia, Competing interests malaria, and measles. Am J Clin Nutr. 2004;80(1):193–8. The authors declare that they have no competing interests. 19. Wolde M, Berhan Y, Chala A. Determinants of underweight, stunting and wasting among schoolchildren. BMC Public Health. 2015; https://doi.org/10. Publisher’sNote 1186/s12889-014-1337-2. Springer Nature remains neutral with regard to jurisdictional claims in 20. United Nations Children’s Fund, World Health Organization, The World Bank. published maps and institutional affiliations. UNICEFWHO-World Bank Joint Child Malnutrition Estimates. New York, Geneva, Washington, DC: UNICEF, WHO, The World Bank; 2012. http://www. Author details who.int/nutgrowthdb/jme_unicef_who_wb.pdf. Accessed 26 May 2018. Department of Economics, Shahjalal University of Science & Technology, 21. National Institute of Population Research and Training (NIPORT). Bangladesh Sylhet Kumargaon, Sylhet 3114, Bangladesh. Institute for Resilient Regions, Demographic and Health Survey 2007. Available from https://www.unicef. University of Southern Queensland, QLD, Springfield 4300, Australia. School org/bangladesh/BDHS2007_Final.pdf. Accessed 26 May 2018. of Teacher Education and Early Childhood, University of Southern 22. Adeba A, Garoma S, Gemede HF, Garoma W. Prevalence of Stunting Queensland, Springfield Campus, Springfield, Australia. School of Medicine, and Associated Factors of Children among 6–59 Months Age in Guto University of California, Riverside, USA. Gida District, East Wollega Zone, Oromia, Ethiopia. Food Sci Qual Manag. 2014;29 Received: 19 February 2018 Accepted: 15 May 2018 23. Rachmi CN, Agho KE, Li M, Baur LA. Stunting, underweight and overweight in children aged 2.0–4.9 years in Indonesia: prevalence trends and associated risk factors. PLoS One. 2016;11(5):e0154756. References 24. Mohsena M, Mascie-Taylor CN, Goto R. Association between socio-economic 1. Demissie S, Worku A. Magnitude and factors associated with malnutrition in status and childhood undernutrition in Bangladesh; a comparison of possession children 6–59 months of age in pastoral community of Dollo Ado District, score and poverty index. Public Health Nutr. 2010;13(10):1498–504. (A1) Chowdhury et al. Globalization and Health (2018) 14:54 Page 12 of 12 25. Islam MA, Rahman MM, Mahalanabis D. Maternal and socioeconomic factors 51. Kumar D, Goel NK, Mittal PC, Misra P. Influence of infant-feeding practices and the risk of severe malnutrition in a child: a case-control study. Eur J Clin on nutritional status of under-five children. Indian J Pediatr. 2006;73(5):417– Nutr. 1994;48(6):416–24. (12 IN A15) 21. (39 IN A13) 26. Siddiqi MN, Haque MN, Goni MA. Malnutrition of under-five children: 52. Guldan GS, Zeitlin MF, Beiser AS, Super CM, Gershoff SN, Datta S. Maternal evidence from Bangladesh. Asian J Med Sci. 2011;2(2):113–9. (A6) education and child feeding practices in rural Bangladesh. Soc Sci Med. 27. Nahar B, Ahmed T, Brown KH, Hossain MI. Risk factors associated with 1993;36(7):925–35. (A7) severe underweight among young children reporting to a Diarrhoea 53. Samuel Oyekale A. Factors explaining acute malnutrition among under- five treatment facility in Bangladesh. J Health Popul Nutr. 2010;1(32):476–83. children in Sub-Sahara Africa (SSA). Life Sci J. 2012;9(4):2101–7. Available from: http://www.lifesciencesite.com/lsj/life0904/314_11976life0904_2101_ 28. Devkota S, Panda B. Socioeconomic gradients in early childhood health: 2107.pdf. evidence from Bangladesh and Nepal. Int J Equity Health. 2016;15(1):78. 54. Bhutta ZA, Ahmed T, Black RE, Cousens S, Dewey K, Giugliani E, Haider BA, 29. Alom J, Quddus MA, Islam MA. Nutritional status of under-five children in Kirkwood B, Morris SS, Sachdev HP, Shekar M. What works? Interventions for Bangladesh: a multilevel analysis. J Biosoc Sci. 2012;44(5):525–35. maternal and child undernutrition and survival. Lancet. 2008;371(9610):417–40. (30) 30. Srinivasan CS, Zanello G, Shankar B. Rural-urban disparities in child nutrition 55. Keats A. Women’s schooling, fertility, and child health outcomes: Evidence in Bangladesh and Nepal. BMC Public Health. 2013;13(1):581. from Uganda’s free primary education Program. Mimeo: Wesleyan 31. Ahsan KZ, El Arifeen S, Al-Mamun MA, Khan SH, Chakraborty N. Effects of University; 2014. p. 31. individual, household and community characteristics on child nutritional 56. Janevic T, Petrovic O, Bjelic I, Kubera A. Risk factors for childhood status in the slums of urban Bangladesh. Archives Public Health 2017;75(1):9. malnutrition in Roma settlements in Serbia. BMC Public Health. 2010; 32. Das S, Rahman RM. Application of ordinal logistic regression analysis in https://doi.org/10.1186/1471-2458-10-509. determining risk factors of child malnutrition in Bangladesh. Nutr J. 2011; 57. Rayhan MI, Khan MS. Factors causing malnutrition among under five 10(1):124. children in Bangladesh. Pak J Nutr. 2006;5(6):558–62. (A4) 33. Das S, Gulshan J. Different forms of malnutrition among under five children 58. Bairagi R. Is income the only constraint on child nutrition in rural in Bangladesh: a cross sectional study on prevalence and determinants. Bangladesh? Bull World Health Organ. 1980;58(5):767. (34) BMC Nutrition. 2017 Jan 3;3(1):1. 59. Grossman M. On the concept of health capital and the demand for health. 34. Bangladesh Multiple Indicator Cluster Survey 2012-2013. http://microdata. J Polit Econ. 1972;80(2):223–55. (36) worldbank.org/index.php/catalog/2533/sampling. Accessed 26 May 2018. 60. Yimer G. Malnutrition among children in southern Ethiopia: levels and risk 35. Bangladesh Multiple Indicator Cluster Survey 2012-2013. ProgotirPathey: factors. Ethiop J Health Dev. 2000;14(3) (Yimer in A16) Final Report. Dhaka: Bangladesh Bureau of Statistics (BBS) and UNICEF 61. Hong R, Banta JE, Betancourt JA. Relationship between household wealth Bangladesh; 2014. inequality and chronic childhood under-nutrition in Bangladesh. Int J Equity 36. Alasfoor D, Traissac P, Gartner A, Delpeuch F. Determinants of persistent Health. 2006;5(1):15. (A14) underweight among children, aged 6-35 months, after huge economic 62. Casas JA, Dachs JN, Bambas A. Health disparities in Latin America and the development and improvements in health services in Oman. J Health Popul Caribbean: the role of social and economic determinants. Equity Health. Nutr. 2007:359–69. 2001;8(37):22–49. 37. Mukabutera A, Thomson DR, Hedt-Gauthier BL, Basinga P, Nyirazinyoye L, 63. Khor GL, Sharif ZM. Dual forms of malnutrition in the same households in Murray M. Risk factors associated with underweight status in children under Malaysia-a case study among Malay rural households. Asia Pac J Clin Nutr. five: an analysis of the 2010 Rwanda demographic health survey (RDHS). 2003;12(4):427–37. (38) BMC Nutrition. 2016; https://doi.org/10.1186/s40795-016-0078-2. 64. Larrea C, Freire W. Social inequality and child malnutrition in four Andean 38. Rahman A, Chowdhury S. Determinants of chronic malnutrition among countries. Rev Panam Salud Publica. 2002;11(5–6):356–64. (26 IN A14) preschool children in Bangladesh. J Biosoc Sci. 2007;39(2):161–73. (9 IN A15; 65. Wang Y, Monteiro C, Popkin BM. Trends of obesity and underweight in 10 IN 805) older children and adolescents in the United States, Brazil, China, and 39. Henry F, Briend A, Fauveav V, Huttly SR, Yunus M, Chakraborty J. The risk Russia. Am J Clin Nutr. 2002;75(6):971–7. (27 IN A14) approach to intervention in severe malnutrition in rural Bangladesh. Am J 66. Doak C, Adair L, Bentley M, Fengying Z, Popkin B. The underweight/ Epidemiol. 1992;136(4):460–3. (18 IN A15) overweight household: an exploration of household sociodemographic and 40. Henry FJ, Briend A, Fauveau V, Huttly SA, Yunus M, Chakraborty J. Gender dietary factors in China. Public Health Nutr. 2002;5(1a):215–21. (28 IN A14) and age differentials in risk factors for childhood malnutrition in 67. Sarma H, Khan JR, Asaduzzaman M, Uddin F, Tarannum S, Hasan MM, Bangladesh. Ann Epidemiol. 1993;3(4):382–6. (19 IN A15) Rahman AS, Ahmed T. Factors influencing the prevalence of stunting 41. HENRY F, Briend A, Fauveau V, HUTTLY SR, Yunus M, Chakraborty J. Risk among children aged below five years in Bangladesh. Food Nutr Bull. 2017; factors for clinical marasmus: a case-control study of Bangladeshi children. 1(805):0379572117710103. Int J Epidemiol. 1993;22(2):278–83. (20 IN A15) 68. De Bruyn T. Dynamics of remittance utilisation in Bangladesh. IOM research 42. Hossain MI, Yasmin R, Kabir I. Nutritional and immunisation status, weaning series, no. 18. Geneva: International Organization for Migration; 2005. p. 40. practices and socio-economic conditions of under five children in three 69. Mohsena M, Goto R, Taylor MNCG. Regional variation in maternal and villages of Bangladesh. Indian J Public Health. 1999;43(1):37–41. (21 IN A15) childhood undernutrition in Bangladesh: evidence from demographic and 43. Chisti MJ, Hossain MI, Malek MA, Faruque AS, Ahmed T, Salam MA. health surveys. WHO South-East Asia J Public Health. 2015;4:2. Characteristics of severely malnourished under-five children hospitalized with diarrhoea, and their policy implications. Acta Paediatr. 2007;96(5):693–6. (22 IN A15) 44. Strauss J, Thomas D. Human resources: empirical modeling of household and family decisions. Handb Dev Econ. 1995;3(26):1883–2023. 45. Glewwe P. Why does mother’s schooling raise child health in developing countries? Evidence from Morocco. J Hum Resour. 1999;1(27):124–59. 46. Tekce B, Shorter FC. Determinants of child mortality: a study of squatter settlements in Jordan. Popul Dev Rev. 1984;10(Suppl 1):257–80. (11 in A8) 47. Benyoussef A, Wessen AF. Utilization of health services in developing countries—Tunisia. Soc Sci Med (1967). 1974;8(5):287–304. (12 IN A8) 48. Caldwell JC, Reddy PH, Caldwell P. The social component of mortality decline: an investigation in South India employing alternative methodologies. Popul Stud 1983;37(2):185–205. (13 IN A8). 49. Bicego GT, Boerma JT. Maternal education and child survival: a comparative study of survey data from 17 countries. Soc Sci Med. 1993; 36(9):1207–27. (A8) 50. Handa S. Maternal education and child height. Econ Dev Cult Chang. 1999; 47(2):421–39. (28)

Journal

Globalization and HealthSpringer Journals

Published: May 30, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off