Complementary feeding practices and their determinants among children aged 6–23 months in rural Bangladesh: evidence from Bangladesh Integrated Household Survey (BIHS) 2018–2019 evaluated against WHO/UNICEF guideline -2021

Appropriate Infant and young child feeding (IYCF) practices are widely considered as one of the most immediate solutions for child undernutrition. Optimal Complementary feeding (CF) during early infancy plays a crucial role in children’s growth, development, and survival [1]. CF is the process of initiating supplementary foods and liquids along with breast milk when breast milk alone is no longer adequate to meet an infant’s nutritional needs [2]. The introduction of energy-rich, nutrient-dense, and timely feedings fulfills the nutrition gaps of the children and can avert stunting along with under-five child mortality [3]. Moreover, the child’s growth faltering is most noticeable from 3 to 24 months, as this time is called the window of opportunity [4]. Measuring and monitoring trends in appropriate IYCF is crucial for planning programs and policies for proper child growth and development.

Bangladesh has experienced rapid economic growth in recent years. The gross national per capita income has increased to $1470, and the country is expected to become a developing country in 2026 [5]. The poverty rates have also dropped by 3.8%, and different healthcare indicators have shown significant improvements [6]. However, the poverty rate is comparatively higher in rural areas: 20.5% in rural areas and 14.7% in urban areas according to the Household Income and Expenditure Survey (HIES)- 2022 [7]. Apart from economic progress, the government of Bangladesh has emphasized reducing child undernutrition and has implemented multiple national nutrition programs and policies. Child undernutrition is still a major issue, notwithstanding the advancements in socioeconomic conditions and legislative initiatives. The prevalence of stunting and underweight among under-five children is high at 28% and 10%, respectively, whereas 23% of children are still wasted [8]. The annual reduction rate has become stagnant in the past few years, and if the current reduction rate continues (3% point per year), the prevalence of stunting and wasting among under-five children will remain at more than 25% by 2025 [9]. However, the lowest socioeconomic groups, rural areas, and slum areas have the highest rates of stunting [8, 10].

Despite Bangladesh’s above-mentioned socioeconomic improvement over the past ten years, CF practices have shown a static trend [11]. Only 28% of the young children aged 6 to 23 months received a Minimum acceptable diet (MAD) and 38% met Minimum dietary diversity (MDD) in Bangladesh according to WHO/UNICEF 2008 guideline [8, 10]. The optimum CF practice rate differs across rural and urban areas, according to MICS and BDHS. MDD and MAD rates are around 8–10% higher in urban settings [10].

Understanding the predictors of poor feeding practices would help identify the factors affecting CF, facilitate the implementation of policies and programs, and design intervention strategies. Na et al. 2018 [11] identified poverty, parental education, residence, and child age as independent predictors of poor CF practices among Bangladeshi children. Similarly, household socioeconomic condition, maternal education level, geographic location, birth order, and gender of child were shown significant associations with various CF complements [12, 13].

To aid in assessing IYCF, the World Health Organization (WHO) developed set of indicators of appropriate feeding practices for children aged 6 to 23 months in 2008 [14]. However, the previous version of the guideline has been criticized for shortcomings identified by users over the preceding decade [15]. Therefore, IYCF-2008 indicator definitions were revised in 2021, and a few new indicators were added to accommodate user demand for new information, followed by two inter-agency meetings in 2017 and 2018 [16]. This new set of indicators did not include any categorization such as core and optional as the previous guideline and recommended assessing all 17 indicators in a population. The definition of minimum dietary diversity, minimum meal frequency, and minimum acceptable diet is altered in the revised guideline. In addition, four new CF indicators are included related to egg and/or flesh food, unhealthy food, and beverage consumption.

The prior IYCF guideline has been followed in the literature on CF practices in Bangladesh [11‐13, 17]. Till now, to our knowledge, no study has assessed the new CF indicators of 2021 except the one conducted by Roy et al. (2022) [17]. However, they only analyzed MDD as per the new guideline. Apart from this, child food poverty is measured based on MDD in UNICEF IYCF global database using DHS/MICS survey data [18]. In light of the revised recommendation, there is a lack of information on complementary feeding practices against all CF indicators among young children in Bangladesh. It is important to comprehensively assess all the CF domains according to the most valid and reliable indicators to design new intervention programs or pick the most relevant interventions. Addressing this knowledge gap, employing the recent WHO/UNICEF guiding principles, the present study, for the first time, evaluated CF practices for all important indicators among young children living in rural Bangladesh where malnutrition and poverty rate are more prevalent than the urban ones. It is also important to identify the predictors of CF practices separately for rural and urban areas to identify whether divergent intervention and policy strategies are required for the particular setting. Even if there is a previous guideline for CF, no other study has specifically sought to generate information on predictors of CF for rural and urban context separately. Therefore, in addition to evaluating complementary feeding practices, the current study attempted to capture the influence of the influencing factors of different CF component, particularly in the setting of rural Bangladesh employing secondary data from that region. However, since our data was limited to rural Bangladesh, we couldn’t investigate the rural–urban difference in these specific issues.

Although Bangladesh has made remarkable progress in recent years to improve child and maternal health, child and maternal undernutrition remain a major public health concern. Despite the country’s impressive socioeconomic development, complementary feeding practices among younger children were found to be stagnant or deteriorating in a recent study [11]. Utilizing a nationally representative dataset, the current study examined the CF practices according to WHO/UNICEF new guidelines among children under two years of age.

Timely introduction of solid, semi-solid, or soft foods at six to eight months (ISSSF) is associated with a lower risk of stunting (i.e., low height in relation to a child’s age) and underweight (low weight for a given age of a child) [1]. We found that 63.5% of rural Bangladeshi infants consumed solid, semi-solid, or soft foods during the previous day, which was lower than the latest Bangladesh Demographic and Health Survey (BDHS) [10]. Studies conducted in Nepal (53.3%) [22] and Afghanistan (56%) [28] showed lower ISSSF rates than those found in the current study in rural Bangladesh. We found no significant association with any demographic or socioeconomic factors, most likely due to the small sample size (n = 126 for ISSSF indicator).

Between the ages of 6 and 23 months, feeding children a variety of foods will help them get all the nutrients they need. Minimum dietary diversity serves as a proxy for adequate micronutrient density of foods [2], and lack of dietary diversity can increase the risk of micronutrient deficiencies, which could be detrimental to physical and cognitive development [29]. Only 18.3% of infants in the current study received a minimum diversified diet, which is less than half of the latest national survey [10]. This is because the revised recommendation contains eight food groups, while the BDHS/MICS estimates were based on only seven. Consistent with a previous study in Bangladesh [11], the current study also revealed that children ages 6–11 months were 75% and children ages 12–17 months were 48% less likely to meet MDD compared to the older age group. Similar findings were observed in studies conducted in Malawi [21], Ethiopia [30], and India [31]. Similar to a recent study conducted in Malawi [21], we found that children from poorer households had a lower likelihood of having an adequate diversified diet than children from wealthier households. Richer households are more likely to be food secure and able to afford a wider variety of food items.

The prevalence of MMF was observed in 52.4% of the children. This estimate is appreciably lower than what was previously reported; around 81% of children had been fed the minimum number of times, according to the latest BDHS survey [10]. The most reasonable explanation behind this discrepancy is the alteration in the definition of this indicator in our study. Because the definition of MMF has been updated in the revised guideline (Supplementary Table 1), our data cannot be directly compared to those of research that followed the prior standard. When defining MMF, previous guidelines solely included breastfeeding; the 2021 guideline expands this to include non-milk feeding. Compared to older children, younger children had lower odds of receiving MMF. This is inconsistent with a study conducted in Malawi [21] but corroborates findings observed in studies conducted in Ethiopia [32], Ghana [33], and Sri Lanka [34]. In our analysis, children from food-insecure households and with low household monthly income had a lower likelihood of receiving MMF—a finding consistent with existing literature from Bangladesh [11].

Only 16.3% of children had been served a minimum acceptable diet, which is less than half of the BDHS estimation (35%) [10]. Our data cannot be directly compared to BDHS because the indicator definition, like MMF, was somewhat modified. Similar to Na et al. 2018 [11] findings, younger children had lower odds of meeting MAD. Surprisingly, our study showed that, compared to children of mothers with secondary or higher education, children of those with primary education or below had higher odds of meeting MAD. This is inconsistent with Na et al. 2018 [11] study conducted in Bangladesh. Although child mothers possessed academic knowledge, it is possible that they lacked understanding regarding optimal complementary feeding practices. The present analysis also revealed the positive role of ANC in feeding children adequately, which is consistent with another study [35].

According to the WHO’s guiding principles, “meat, poultry, fish, or eggs should be consumed daily, or as often as possible,” for breastfed and non-breastfed children [2]. Egg consumption is linked to higher calorie, protein, essential fatty acid, vitamin B₁₂, vitamin D, phosphorus, and selenium intakes, as well as higher recumbent length [36]. Evidence also suggests that the introduction of meat as an early complementary food improves protein and zinc intake [37]. In the present analysis, 23.3% of children received egg and/or flesh food (EFF) during the previous day. We couldn’t compare the MMF to another study because it was recently added to the updated guideline. However, the proportion of children who ate eggs was substantially lower than what the BDHS estimated (19.7% vs. 29.2%) [10]. On the other hand, the percentage of children that consume flesh food in the current survey is nearly identical to that which was reported in BDHS. Similar to other indicators, children from poorer households had lower odds of meeting EFF, which could be explained by the fact that egg and flesh foods are more affordable in affluent families.

Sweet drinks only provide energy and no additional nutrition. Children of all ages are more likely to become obese when they consume more sugar-sweetened beverages. The introduction of sugar-sweetened beverages at earlier ages (before 12 months of age) is associated with obesity at six years of age [38]. Free sugars, including 100% fruit juice and sugar-sweetened beverages, also increase the risk of dental caries [39]. SWB is a new CF indicator, and it was estimated at 2.5% for rural Bangladeshi infants. Children of mothers with less than a primary level of education were more likely to consume sugar-sweetened beverages than those with a secondary or higher level of education. Education may positively impact food choices because it makes mothers more concerned about their children’s health and more likely to steer clear of unhealthy foods. Apart from that, children from low-income families were less likely to consume sugar-sweetened beverages. The possible reason may be the relatively higher cost of sweetened beverages like fruit juices and candy.

Another new indicator in the updated guideline is Unhealthy food consumption (UFC). With the upward socioeconomic condition, consumption pattern is also changing, and intakes of added sugars, unhealthy fats, salt, and refined carbohydrates are increasing in low- and middle-income countries. Consuming these foods could limit the intake of vital vitamins and minerals and replace them with less nutritious foods. Moreover, consuming unhealthy snacks and beverages has been linked to a higher risk of nutrient deficiency and shorter length-for-age [40]. Besides, childhood exposure to sweet foods and drinks regularly may develop an innate preference for sweetness, leading to increased consumption of sweet-tasting foods and drinks as a later-learned preference [41]. This study found that 12.2% of cases involved unhealthy eating habits, indicating the need to educate rural residents about child-feeding practices.

Zero vegetable or fruit consumption (ZVF) is also a new indicator in the revised guideline. Low vegetable and fruit consumption is one of the dominant drivers of non-communicable diseases. Grimm et al. 2014 [42] also found that a young child’s low fruit and vegetable intake is related to lower intake in later life. Approximately 63% of rural Bangladeshi children had not consumed any fruits or vegetables (ZVF) in the previous 24 h, with the prevalence of ZVF being much higher among younger children. High ZVF prevalence suggests that special focus should be placed on raising awareness regarding the significance of consuming fruits and vegetables in early life.

BIHS is a household-based survey collecting information on various agriculture, nutrition, and development issues. Questions related to complementary feeding indicators were applicable to those households where there was at least one child less than 2 years of age. The number of eligible cases was only 665, therefore, we can’t generalize the findings for the entire rural Bangladesh. The sample size for measuring some of the CF indicators was low in the original datasets, which may limit the study’s power since the power of the study increases with the sample size [43]. Additionally, due to the extremely low number of non-breastfed children, we were unable to disaggregate our analysis between breastfed and non-breastfed children. The predictive analysis could not be carried out for MMFF for the same reason. In addition, a causal inference couldn’t be drawn due to the cross-sectional nature of the study. We were unable to include other variables that might have explained our outcome variables because we used secondary data. We identified a wide range of CF predictors that have practical implications for policymakers when developing suitable nutrition projects.

Four of the nine CF indicators (ISSF, ISSF, MDD, MMF, and MAD) were found in various national surveys and published publications, as recommended by the prior guideline. Poorer feeding practice for these four CF components can be attributed to the adoption of the revised guideline. We also observed unhealthy feeding practices among approximately 12% of children, with 2.5% consuming sugary beverages. A comparatively higher proportion of children get appropriate feeding in terms of ISSF (63.2%) and MMF (52%). Eggs were only given to 19.7% of children, and most of them were fed cereal-based staples such as grains, roots, and tubers. Overall, child age, education level of mothers, ANC, household food security, monthly household income, and residential area were significantly associated with various components of complementary feeding practice. Quality counseling to child mothers or primary caregivers, as well as persuasive behavioral change advocacy to other family and community decision-makers, with an emphasis on younger children, may be an effective strategy to enhance complementary feeding practice.