1 Introduction

Malnutrition is a condition that stems from undernutrition of individuals, inadequate access to essential nutrients including protein-energy, micronutrients like iron, zinc, vitamin A and iodine [1]. The prevalence of childhood undernutrition is considered a complex global problem and the prevalence is highest in low and middle-income countries (LMICs) [2]. Globally, 149 and 45 million under 5 children were estimated by the World Health Organization as stunted and wasted respectively and about 45% in deaths of this age group are linked to undernutrition which mostly occur in LMICs [3]. Asian countries bear a significant portion of globally undernourished children in which 53% showed signs of stunting and 70% showed signs of wasting in 2020 [4]. Although Bangladesh has progressed significantly in combating childhood undernutrition during the past few decades, it continues as a serious public health issue of the country due to socioeconomic, demographic and environmental factors that contribute to the morbidity and mortality of under 5 children [5]. The socio-economic characteristics of a community, including gender inequity, educational background, and poverty [6] significantly determine health and nutritional outcomes affecting the prevalence and incidence of malnutrition at both the community and household levels [7].

Socioeconomic and demographic characteristics and inappropriate infant and young child-feeding practices have been identified as primary contributing factors to malnutrition for children under five years of age [5, 8,9,10]. In Bangladesh, socio-economic condition of households emerged as a critical determining factor of childhood malnutrition where children from households with comparatively higher socioeconomic status are less likely to be underweight [5, 11]. In addition to these, poor access to drinking water, sanitation and hygiene (WASH) facilities is another key factor for child undernutrition [12] driving to morbidities like diarrhoea that also increase the risk of undernutrition in child undernutrition [13]. Several previous studies conducted in Bangladesh reported that inadequate WASH facilities or poor hygiene practice increased child malnutrition at household level [14,15,16].

It is evident through several studies that undernutrition prevalence and socio-economic factors in developing countries vary by geographic area [17,18,19] and environmental conditions like topography and climate. Numerous studies highlighted that the reduction of agrobiodiversity adversely affected food accessibility leading to a decline in consumption of variety of nutritious foods among communities residing in saline-prone coastal region of Bangladesh [20]. Bangladesh, a country of South Asia is highly susceptible to numerous natural disasters [21]. Particularly, the coastal part of the country are considered as the zone of multiple vulnerabilities [22]. Among the 35 million people in the coastal regions in Bangladesh more than 30% are under-resourced [23] with comparatively lower coverage of WASH facilities (25.2% in central coastal zone) than mainland [24]. Weaker social network, lack of access to WASH facilities exacerbated by the rising frequency and severity of natural disasters attributed to climate change, climate variability and poor access to health facilities [25] are the prime factors for the higher vulnerability of coastal households [26]. Climate change adaptation and resilience projects in coastal Bangladesh failed to demonstrate their intended outcomes due to limited resources of coastal communities [27] along with inadequate government initiatives, follow-up procedures, and a highly centralised organizational structure [28].

Numerous studies have explored influencing factors such as climate change [29,30,31], socioeconomic inequalities [32] and child feeding practices and household food insecurity [33] of child malnutrition while water and sanitation facilities of child morbidities [34, 35] in specific zone (eastern or central or western) of the coastal region of Bangladesh. Despite the vulnerability of coastal households, no study has been conducted on socio-economic inequalities and poor sanitation associated with child malnutrition critically focused on diverse coastal regions of Bangladesh. Consequently, it is imperative to better understand the prevalence and distribution of households experiencing poor sanitation and the associations between these socio-economic conditions and childhood malnutrition. The primary hypothesis of the study was, childhood malnutrition is directly or indirectly associated with the socio-economic characteristics and sanitation facilities of the households in the coastal region of Bangladesh. This study aimed to understand the association between socio-economic characteristics, available sanitation facilities and childhood malnutrition in the coastal regions of Bangladesh.

2 Methods

2.1 Study site

A total of 17 coastal districts; five from Barisal division, six from each of the Khulna and Chattogram division were considered as the sites for this study (Fig. 1).

Fig. 1
figure 1

Study area showing coastal divisions (Barisal, Chattogram and Khulna) of Bangladesh

Full size image

2.2 Data source

In this study, we used the latest data of Bangladesh Demographic and Health Survey 2017–2018 (BDHS’ 2018), which is a nationally representative cross-sectional survey. We extracted the data for 17 coastal districts from geospatial covariates dataset of BDHS using geographical coordinates for the population of interest for this study. Data for selected districts were extracted using shapefiles provided in the dataset to identify and select specific districts based on their geographic boundaries. These comprised ever-married women (15–49 years old) and their children (0–59 months old) from 17 coastal districts from three coastal divisions. The BDHS’18 survey is based on a two-stage stratified sample where 675 enumeration areas (EAs), 250 from the urban and 425 from the rural area were selected in the first stage with probability proportional to EA size. In the next stage, on average 30 households from each EA were selected through systematic sampling which led to 20,250 households. For the present study, a total of 2153 households were selected where data were collected in a form of household survey using a structured questionnaire.

2.3 Outcome measures

As the outcome variable child malnutrition indices were considered for our study. In BDHS 2018, child malnutrition was measured by standard anthropometric indicators of height-for age Z-score (HAZ), weight-for-age Z-score (WAZ) and weight-for-height Z-score (WHZ) as defined by the WHO [36]. Child’s height-for-age represents their linear growth and weight-for-height explains their nutritional status. Another indicator weight-for-age is a combined measure of height-for age and weight-for-height. Thus, stunting, wasting and underweight were measured as the outcome of interest and specified as being less than -2 standard deviation (< -2SD) less than the median value for HAZ, WHZ and WAZ respectively.

2.4 Exposure and covariates

Socioeconomic and demographic characteristics including age group of the children, children sex, place of residence, educational attainment of caregivers, wealth index of the households were considered as the major exposure variables for the study. Principal component analysis (PCA) was used to compute households’ wealth index [37] on the basis of number and types of consumer goods a household own (ownership of agricultural land and farm animals, radio, television, mobile phone, electric fan, almirah/wardrobe, refrigerator, water pump, computer, DVD/VCD player, air conditioner and IPS/generator), means of transportation (car or bicycle) and household characteristics (drinking water source, types of toilet facilities, roof materials, wall materials, flooring materials and number of sleeping rooms). Households’ wealth index is a combined measurement of a household’s cumulative living standard, was separated into five categories (i.e., poorest, poorer, middle, richer and richest) using PCA by DHS program. At first a wealth score was calculated based on the information of household characteristics by PCA and then divided the households into five equal quintiles accordingly. Drinking water source and types of toilet facilities were also considered as another exposure for the outcome variables. Piped into dwelling, tube well/borehole, piped to yard/plot and neighbour, standpipe/public tap, protected well, rainwater, bottled water and cart with small tank were categorised as improved sources of drinking water. While river, dam, lake, canal, ponds, irrigation channel, stream and unprotected well were considered as unimproved sources of drinking water. Pit latrine with slab, flush to piped sewer system, to pit latrine, to septic tank and ventilated improved pit latrine were categorised as improved toilet facilities whereas, unimproved toilet facilities included no facility/bush/field, hanging toilet (made up of structures with bamboo and corrugated metal that were hung above the ground on poles letting waste drop straight down into a soup of mud and trash below), pit latrine without slab/open pit and flush to somewhere else or not known. Child breastfeeding practice and morbidity status (children who had diarrhoea, fever and cough in the two weeks prior to the survey) was considered as covariates for this study. Body mass index (BMI) of the caregivers was also considered as a covariate and was measured as weight in kg divided by height in meters squared (kg/m2). Caregiver’s BMI was categorized into four groups as underweight (< 18.5 kg/m2), normal weight (18.5–22.9 kg/m2), overweight (23–24.9 kg/m2) and obese (≥ 25 kg/m2) according to the Asia–Pacific cut-off points [38].

2.5 Statistical analysis

Data analysis included descriptions of the study population and estimation of the status of household socio-economic characteristics, sanitation facilities, child nutritional outcomes and morbidity and maternal BMI were presented with frequencies, percentages and proportion with respective 95% confidence interval (CI) using a binomial exact method. Cross tabulation with Chi square (χ2) test was used to determine the prevalence/percentage and associations (p values in χ2 test) between the selected exposures or covariates and outcomes. Univariate logistic regression was performed and variables for multivariable logistic regression model were adjusted setting the significance level at p < 0.25 in the univariate logistic regression to assess the statistical associations of child malnutrition, sociodemographic and sanitation facilities status or other covariates and presented in adjusted odds ratios (AOR) with a 95% CI [39]. All the significant variables in univariate logistic regression were selected for multivariable logistic regression model because no multicollinearity among the independent variables were found. Model fits of the multivariable logistic regression model were assessed using Pearson χ2 (goodness-of-fit test) and Hosmer and Lemeshow’s goodness-of-fit test. In Hosmer and Lemeshow’s goodness-of-fit test we determined p-values for first, second and third model as 0.96, 0.73 and 0.32, which was 0.19, 0.06 and 0.27, in Pearson χ2 (goodness-of-fit test). Thus, non-significant p-values (> 0.05) suggest good fit of the model. Sampling weights were applied to ensure that the results are accurately represented at the division and national levels. In all regression models, missing data were handled using listwise deletion method and total observations used in three models were 2014 for stunting, 2033 for wasting and 2084 for underweight. Whole analysis was performed by the STATA, a statistical software package version 17 (Stata Corporation, College Station, TX, USA).

3 Results

3.1 Descriptive statistics

Description of the socio-economic and household sanitation facilities status, prevalence of child malnutrition, child morbidity status and maternal BMI have been presented in Table 1. The highest (22.8%) and lowest (17.5%) proportion of children belonged to the age group of 0–11 and 36–47 months respectively. Majority of the households (72.5%, n = 1324) were located in rural areas, and most of the household heads (84.7%) were male. Of the mothers over half (52.5%) had secondary level of education. In terms of wealth index distribution, it was found that the highest proportion (21.3%) of the households were from the poorest group followed by richest (20.9%), middle (19.9%), richer (19.1%) and poorer (18.8%). The study revealed that, majority of the households (96.4%) had access to an improved drinking water source while a notable portion (35.3%) of the households had unimproved toilet facilities (Table 1).

Table 1 Household sociodemographic characteristics, sanitation facilities, child malnutrition, morbidity and maternal BMI status in the coastal regions of Bangladesh
Full size table

Stunting, wasting and underweight prevalence among the children in the coastal regions of Bangladesh was 31.4%, 8.5% and 21.1%, respectively. A substantial portion of the children were suffering from fever (30.5%) and cough (33.4%) while a small portion of the children (5.6%) were found to suffer from diarrhoea. In case of caregivers’ nutritional status maximum proportion (42%) had normal BMI which was followed by obese (30.2%), overweight (15.8%) and underweight (12%) (Table 1).

3.2 Association of child malnutrition with socio-economic variabilities, sanitation facilities, child morbidities and maternal BMI

Table 2 described that statistically higher significant prevalence of stunting was found among children aged 24–35 months (41.9%; χ2= 56, p < 0.001), from rural areas (33.7%; χ2 = 17.37, p < 0.001), born to mothers with no formal education (45.8%; χ2 = 88.02, p < 0.001), from poorest households (44.9%; χ2 = 96.82, p < 0.001), having unimproved toilets (37.4%; χ2 = 26.48, p < 0.001) and born to mothers with underweight to normal BMI (35.3% to 35.6%; χ2 = 33.94, p < 0.001). The wasting prevalence was found significantly higher for the children from poorest households (12.4%; χ2 = 14.31, p < 0.005), had fever (10.5%; χ2 = 5.89, p < 0.005), practicing breastfeeding (9.3%; χ2 = 3.95, p < 0.005) and children born to underweight mothers (14.8%; χ2 = 20.42, p < 0.001). Furthermore, significantly higher prevalence of underweight was found with children aged 48–59 months (25.7%; χ2 = 19.89, p < 0.005), born to mothers with no formal education (38.9%; χ2 = 60.76, p < 0.001), from poorest households (31.3%; χ2 = 53.32, p < 0.001), having unimproved toilet facilities (27.4%; χ2 = 30.32, p < 0.001), suffered from fever (26%; χ2 = 15.30, p < 0.001) and low maternal BMI (29.3%; χ2 = 34.65, p < 0.001).

Table 2 Prevalence of stunting, wasting and underweight across the socio-economic variables, sanitation facilities status, child morbidities and maternal BMI in the coastal regions of Bangladesh
Full size table

3.3 Factors associated with child malnutrition

All significant variables of the univariate logistic regression model (Supplementary Table S1) were considered for the multivariable logistic regression model. Children who belong to the age group 12–23 months had 2.24 times higher odds (AOR = 2.24, 95% CI: 1.63–3.07) of stunting compared to the children age group 0–11 months. The odds was also found higher for the children belong to the age group 24–35 (AOR = 3.32, 95% CI: 2.35–4.67), 36–47 (AOR = 2.35, 95% CI: 1.56–3.55) and 48–59 (AOR = 1.97, 95% CI: 1.32–2.95) compared to the children age group 0–11 months. Mothers having higher education were 69% (AOR = 0.31, 95% CI: 0.18–0.52) less likely to have stunted children compared to the mothers having no education. We also found that, children from poorest households showed higher likelihood of being stunted (AOR = 2.20, 95% CI: 1.48–3.25) followed by children from poorer households (AOR = 1.83, 95% CI: 1.24–2.70) compared to the children from richest households. Children born to mothers who are obese (compared with their normal BMI) were 34% less likely of being stunted (AOR = 0.66, 95% CI: 0.51–0.85) (Table 3).

Table 3 Multivariable logistic regression of factors associated with child malnutrition in the coastal regions of Bangladesh
Full size table

Children who had fever showed 1.51 times higher odds (AOR = 1.51, 95% CI: 1.08–2.11) of being wasted when compared with the children who had no fever. Furthermore, children of the underweight mothers also showed 1.94 times higher odds (AOR = 1.94, 95% CI: 1.26–3.00) compared to children born to mothers with normal BMI (Table 3).

Children belonging age group 24–35 months had 1.75 times higher odds (AOR = 1.75, 95% CI: 1.24–2.47) which is almost similar (AOR = 1.73, 95% CI: 1.21–2.48) to the age group 36–47 months compared to the children of 0–11 months. However, the odds of being underweight was quite higher (AOR = 2.14, 95% CI: 1.52–3.03) for the children who belong to the age group 48–59 months compared to the children age group 0–11 months. The odds of being underweight also decreases with increasing mother’s educational level. Children from poorest households were 57% (AOR = 1.57, 95% CI: 1.02–2.42) more at odds of underweight compared with children from richest households. Children from households having unimproved toilet facilities 29% (AOR = 1.29, 95% CI: 1.01–1.65) more likely to be underweight compared to those having improved toilet facilities. Significance association was found for co-morbidity such as fever (AOR = 1.68, 95% CI: 1.27–2.21) with underweight.

Furthermore, children to the overweight mothers were 32% (AOR = 0.68, 95% CI 0.48–0.96) and obese mothers were 39% (AOR = 0.61, 95% CI 0.45–0.81) less likely to be underweight compared to the mothers having normal BMI (Table 3).

4 Discussion

Findings of our study revealed significance associations between child malnutrition, socioeconomic inequalities and unimproved toilet facilities. Childrens’ age, low maternal education, lower household wealth index, unimproved toilet facilities, child morbidity and maternal BMI were found as significant risk factors for child malnutrition. We also found that one-fifth of the households were in the poorest wealth index and more than one third had unimproved toilet facilities in the coastal region of Bangladesh. Previous studies also reported that majority of the households in the coastal region of the country have vulnerable livelihoods [40] and lower socioeconomic status, unhygienic toilet facility have significant associations with child malnutrition [41, 42].

Findings of the study showed that sociodemographic factors including age of children and mother’s educational level are significant predictors for child undernutrition (stunting and underweight) in the coastal regions of Bangladesh. Prevalence of stunting and underweight was significantly higher among older age group, with the highest prevalence occurring in the age group 24–35 months. This trend may be the effects of inappropriate initiation of complementary feeding with inadequate diversity in diet and lack of nutritional knowledge of their caregivers [43, 44]. Additionally the way of complementary feeding practiced by the caregivers of the younger children in Bangladesh is still found to be deteriorating [45]. Public health interventions should prioritize educating caregivers about proper feeding practices for infants and young children, emphasizing the importance of dietary diversity and timely initiation of complementary feeding.

It was also observed that mother’s educational level had protective effect on stunting and underweight which is often regarded as a critical factor for child’s nourishment [46, 47] because this corresponds to women’s knowledge regarding mothering and subsequent care-taking of under 5 children [48]. A previous study reported that mothers having formal education have knowledge on child health and nutrition and are capable of identifying illness and subsequent treatment for their children while, mothers having no formal or lower education usually don’t know about quality foods, healthy and hygienic living conditions which could have an impact on their children’s health and nutritional status as well as their own [18]. Therefore, it is necessary to arrange community nutrition and public health intervention programmes on a regular basis to ensure proper health practices including exclusive breast feeding up to six months and timely initiation of colostrum feeding appropriately in the coastal regions of Bangladesh [49].

Household wealth index was also identified as a significant risk factor for child stunting and underweight. Children living in the households with the poorest wealth index were more likely to be stunted because households with lower socioeconomic status might have less resources to access nutritious food for their children [50]. It is well established by several previous studies both globally and in the Bangladesh context that socio-economic inequality is a significant predictor of child health and well-being [41, 51, 52] which is in concordance with the findings of our study.

Unimproved toilet facility and child morbidity were other significant risk factors for child undernutrition in this study. Previous studies identified negative impacts of unimproved toilet facility as well as poor sanitation facilities on both child morbidity and their nutritional status [53, 54]. Children exposed to unhygienic environments resulting from household inadequate water, sanitation and hygiene facilities faced increased risks of morbidities due to heightened exposure to infectious diseases [55, 56]. Consequently, childhood morbidities play a significant role in their nutritional status and it is evidence indicated that children with different morbidities such as diarrhoea, fever, and acute respiratory infection have more likelihood of being underweight than those having no morbidities [57].

Maternal BMI was observed as a significant predictor for child malnutrition in this study where obesity of mothers was found as a protective factor for stunting and underweight. As mother’s BMI represents her own physical attributes, stature and nourishment [50] so, it is not surprising that maternal BMI is a significant determinant for child undernutrition. In this study, low maternal BMI of mothers was identified as significant risk factor for child wasting because malnourished mother cannot provide sufficient breast-milk due to their nutritional deficiency [10].

Households with lower socio-economic status cannot afford adequate food which may leads to greater risk of maternal undernutrition as well as child undernutrition. Childhood malnutrition is significantly correlated with household’s socio-economic inequalities [58] as wealthier households have access and can manage to pay for additional healthy and nutritious food [59]. Policy makers should develop and implement research based initiatives to improve nutrition education and behaviours of adults for health promotion specially targeting low socioeconomic groups [60] in the coastal region of Bangladesh. Findings of the study supported our initial hypothesis revealing that childhood malnutrition is directly associated with socioeconomic characteristics and sanitation facilities of the households in the coastal region of Bangladesh. Therefore, addressing socioeconomic inequalities through targeted interventions, social safety nets, poverty alleviation programs, ensuring social protection and improving living condition should be a priority in efforts to combat childhood malnutrition in these particular regions.

4.1 Strengths and limitations of the study

Our study is strengthened by inclusion of the broad range of critical key factors of childhood malnutrition and its unique focus on extensive coastal regions of Bangladesh, where communities face distinct challenges. However, this study has several limitations. The study is cross-sectional, which limits our ability to prove a causal relationship between exposures with key nutritional outcomes. Recall bias on responding childhood morbidity and household socio-economic status among the respondents may also be a limitation of this study. Among 19 coastal districts we extracted the data of 17 districts from three coastal divisions where there may be some study participants who are not exclusively exposed to the coastal influence. Furthermore, the data did not include the critical influencing indicator of malnutrition like household food insecurity status.

5 Conclusion

This study shows that the sociodemographic conditions and sanitation facilities status are key determinants of child malnutrition in the coastal region of Bangladesh. This region is highly susceptible to child malnutrition. Given the geographical vulnerabilities of the coastal area compared to the mainland, policymakers should prioritize the development of targeted and sustainable interventions. To reduce childhood malnutrition in this region further research addressing other critical influencing factors like household food insecurity and climate change needs to be incorporated. Additionally, longitudinal research should be designed along with a cross-sectional setting to get a clear snapshot of the situation and understand the dynamic interaction of the determining factors associated with child malnutrition. Addressing these research gaps through further research will enable to formulate of more precise, holistic and sustainable policies to consistently reduce child malnutrition in the coastal regions of Bangladesh.