The study captured information on vital nutritional indicators such as stunting, thinness, and anemia among unmarried adolescent girls and their predictors in rural Bangladesh. These findings extend beyond national boundaries, reflecting a comparable situation in economically disadvantaged countries across South-East Asia. Our results highlight the positive impact of non-food practices such as using sanitary latrines, practicing hand hygiene, deworming, and television viewing on improved nutritional status and higher Hb levels in adolescent girls. The primary focus of the study was to delve into the underlying causes of undernutrition and anemia among adolescent girls. This aligns with the objectives of the BRAC nutrition program, which seeks to improve conditions of thinness, stunting, and anemia among adolescent girls, as these issues are intricately tied to their reproductive health.

In Bangladesh, a couple of studies had already been conducted to gauge the nutritional status of adolescent girls [48,49,50]. However, these studies faced limitations such as small sample sizes that might not fully represent all societal segments or the inclusion of only specific indicators of nutritional status. Nonetheless, our current study reveals a troubling finding: the prevalence of stunting and thinness among Bangladeshi adolescent girls has shown minimal improvement over the past decade [48, 50].

The significance of this study is in its exclusive focus on unmarried adolescent girls in the community. These findings strongly support integrating WASH messages with nutrition interventions, aligning with SDG-2 and SDG-6 and also offer substantial evidence to enhance girls’ health and nutrition by aligning with government strategies [30]. In contrast, a study conducted in BRAC health program area included both married and unmarried girls, found associations between anemia and select social factors. These factors might have limited impact on pre-marital undernutrition, anemia, and reproductive health of adolescent girls. Moreover, a study by Chattopadhyay et al. established a link between nutritional status and WASH practices among adolescent girls, albeit lacking data on other essential underlying and causal factors of undernutrition, such as feeding patterns, deworming and morbidity [24], whereas another study found that suffering from diarrhea and dysentery preceding a month of interview were associated with thinness [50]. Notably, a national survey in Bangladesh covered only ages 6–14 years [37].

An important finding, we observed is that stunting predicts thinness, and conversely, thinness predicts stunting in this population. Another study in Bangladesh showed that thinness correlated with low upper arm muscle and fat area Z-scores, rather than stunting [50]. BMI reflects body fat, while height indicates genetic traits influenced by nutrition, environment, and factors like puberty timing [51]. Puberty age affects height gain, influenced by diet, activity, and childhood obesity. Yet, increased growth velocity before puberty can compensate for final height [51]. All these pathways are impacted by energy balance, nutrient quality, especially protein, fat, micronutrients, and healthcare [51]. In Bangladesh, lacking food security hampers preparatory nutrition for pubertal growth, leading to both thinness and stunting. Thus, we observed that both thinness and stunting were associated with each other.

It's important to recognize that factors underlying undernutrition in adolescent girls can differ between countries due to unique contextual elements and study focuses. For instance, in Ethiopia, a study found associations between low BIMZ and factors like age, dietary diversity, and health knowledge, while HAZ was linked to age, household food security, and girls' knowledge, based on community data [52]. In this context, the FFQ quantifies nutrient intake, whereas dietary diversity assesses qualitative aspects of food consumption. In our study, we emphasized the relationship between nutrient intake quantity and girls' nutritional status, recognizing the strong correlation between quantitative and qualitative dietary assessments [53]. Importantly, we did not include qualitative dietary measures. An Ethiopian school-based study identified associations between age and thinness, as well as connections between stunting and factors like menstrual status, food security, and sanitary latrine use [54]. Similarly, other African study tied poor personal hygiene to anemia and underweight, echoing our findings [55]. Another study noted that parental factors' impacted on undernutrition decreased as socioeconomic status improved, which is akin to our study's patterns [56].

Differences in methodology among previous studies have also led to variations in reported anemia prevalence. For example, a study incorporating both married and unmarried girls, using the same Hb measurement method as ours, showed significant differences in anemia prevalence. In contrast, the national micronutrient survey measured iron deficiency anemia using ferritin levels, indicating lower prevalence in children aged 6 to 14 [37]. Notably, the demand for iron notably increases at ages 14 to 15, compounded by iron deficiency from pregnancy and lactation [4]. Despite methodological disparities, these collective findings consistently emphasize the link between early marriage, teenage pregnancy, and their contribution to anemia among adolescent girls.

The current study provides an alarming indication of early marriage and the significant data gap concerning married adolescent girls in rural Bangladesh. Recruiting unmarried girls for interviews was a challenge for current study due to prevalent early marriages. In the rural context of Bangladesh, adolescents have an increased likelihood of being married, experiencing pregnancy, leaving their education, and coming from households with a low socioeconomic status [50]. A nationally representative dataset, covering both married and unmarried girls, was crucial for accurate recruitment planning. However, obtaining current census data was time-consuming, and older data didn't reflect the present community conditions of rural areas [5]. To address this predicament, we chose neighboring villages covered by the BRAC nutrition program to ensure sufficient participants for randomization and interviews.

In urban India, a significant prevalence of anemia and worm infestation was discovered among adolescent girls, emphasizing the need for extra nutritional support, iron-folic acid supplements, and effective worm prevention and management [57]. Notably, we observed protective effects against anemia in those receiving anti-helminthic medication within 6 months. As worms hinder iron absorption, deworming tablets could boost intestinal iron uptake, potentially reducing anemia [58]. Globally, 12% of worm-related disease burden falls on children aged 5 to 14, with anemia risk rising in girls [59]. Our findings stress the urgency of rapid WASH implementation and routine deworming tablet use since childhood. This proactive strategy has the potential to prevent adolescent iron deficiency anemia [60].

Moreover, more extensive adoption of sanitary latrine usage and the practice of washing hands after defecation was noted among adolescent girls. It is noteworthy that, during the survey period, the BRAC WASH program was operational in 61 Districts across Bangladesh. In addition to the effort of the GoB, various other NGOs were also engaged in the study areas to enhance WASH practices. Despite these efforts, a significant proportion of adolescents‒specifically, one-fourth of them displayed gaps in adhering to handwashing with soap after defecation. Additionally, three-fourths of the participants exhibited gaps in handwashing before eating, a practice that could potentially heighten the risk of contracting a range of infectious diseases. Research shows that drinking groundwater is associated with less anemia among women and children even though absence of vitamin C and heme iron in the food [61].We observed that an average 67% of the adolescent girls were drinking ground water (tube well and deep-tube well water). However, from our analysis we have not found any associations between drinking safe water and better Hb level, while average vitamin C intake is 65 and 71 mg/day in the intervention and comparison areas, respectively.

The findings from an observational study indicate that increased nutrient intake, encompassing energy, protein, folate, B12, iron, vitamin C, and red meat, is associated with a reduced likelihood of anemia among women [62]. This study also speculates that fortifying with folic acid could potentially contribute to the lower prevalence of anemia [62]. Notably, in the context of Bangladesh, it is important to highlight that while edible oil is fortified with Vitamin-A, it lacks fortification with iron, folic acid, or other nutrients known to improve anemia [63]. The Vitamins B-6, B-12, A, C, folic acid, and riboflavin are of paramount importance in the synthesis of hemoglobin [64,65,66,67]. These nutrients, either directly or indirectly, exert an influence on iron absorption and metabolism. The rich dietary sources of these micronutrients include meat, poultry, fish, milk, legumes, and green leafy vegetables. An analysis of dietary diversity data reveals that roughly three-fourths of adolescent girls included fish, meat, and poultry in their diets during the last 24 h, with one-fourth consuming eggs, one-fourth incorporating green leafy vegetables, and fewer than 30% including dairy products. Moreover, their fruit intake and vitamin C consumption were higher. However, it is vital to recognize that this 24-h dietary pattern does not necessarily reflect their daily eating habits. It is important to emphasize that in the study areas, observations indicated that people typically purchased fish, meat, and poultry only on designated “hat day” (weekly market day) and stored them for the subsequent 2–3 days. On other days of the week, most of them relied on eggs, legumes, dried mung bean pills (known locally as “daler bori”), and vegetables, etc., with fish, meat, and poultry being absent from their plates. This pattern persisted until the next “hat day”. Despite this dietary variation, it is conceivable that adolescent girls who consumed more energy might have included more foods that influence hemoglobin production in their diets. This potential association was revealed in our analysis.

However, there are certain limitations inherent to this study. Given that it was a cross-sectional study, it becomes challenging to draw definitive conclusions regarding growth patterns over time. The nature of the study design allows us to generate hypotheses concerning the potential influence of certain variables on the nutritional status of adolescent girls; however, it does not facilitate the exploration of their causal relationships [12].

We converted cooked foods to raw using established methods for nutrient intake calculation [39]. However, our 7-day recall data from a semi-quantitative food frequency questionnaire might have recall bias due to the complexity of Bangladeshi cuisine and regional variations in cooking methods [37]. Although we employed a validated FFQ questionnaire and ensured rigorous training and quality control, the intricacies of Bangladeshi cuisine could lead to underestimation or overestimation of certain foods. Our robust approach is supported by a larger sample size, reflected in normal distribution and statistical measures (SD, SE, and 95% CI) for nutrient intake. Notably, we lack data on nutrient intake for unmarried adolescent girls aged 10–19 years across Bangladesh, limiting our ability to compare their average intake with the national micronutrient survey that evaluated different age groups [37].

We employed a validated method for measuring Hb levels, supported by good sensitivity and specificity [68]. In resource-limited settings, using the hemocue-Hb photometer to assess Hb in capillary or venous blood is widely accepted due to its portability, user-friendliness, and real-time results [69]. Research shows HemoCue's comparability to other methods like Sysmex, KX21N, and Cyanmethemoglobin [68]. Some blood subsamples measured additional indicators (ferritin, serum transferrin receptor, etc.) to enhance results’ validity for iron deficiency anemia. The complementary tests validate hemoglobin (Hb) levels, which are correlated with iron deficiency. Factors such as overhydration, hemoglobinopathies, and chronic health conditions can contribute to a decrease in Hb levels [56, 57]. While this decline may account for anemia, it may not specifically indicate iron deficiency anemia [70]. Despite these considerations, our meticulous process for Hb determination, led by trained field enumerators and specialist teams, aimed to ensure precision and reliability.

Another drawback of our study is the absence of measurements for the physical activity of adolescent girls, a key predictor of the BIMZ score. In urban areas of Bangladesh, overweight parents, limited exercise, and high sedentary lifestyles contribute to obesity among school children [71]. Conversely, rural areas demonstrate a different dynamic, where reduced physical activity in girls directly relates to overweight and obesity [72]. Intriguingly, Saha et al. found that factors like food practices, screen time, gaming, and sleep duration were not significant predictors of obesity of adolescent girls in rural areas [72]. While recognizing this limitation, it's important to emphasize that the lack of physical activity data doesn't diminish the significance of our other findings. They provide valuable insights into various factors influencing the nutritional status of adolescent girls. To comprehensively understand the complexities of teenage health, future research should incorporate assessments of physical activity.