Arlette Suzy Setiawan ^1* Arief Budiarto ² Ratna Indriyanti ¹

• ¹ Department of Pediatric Dentistry, Faculty of Dentistry, Universitas Padjadjaran, Bandung, West Java, Bandung, Indonesia
• ² Faculty of Psychology, Universitas Jendral Achmad Yani, Bandung, West Java, Indonesia

The high stunting rate in children under five persists worldwide. Globally, 149.2 million children under five suffered from stunting in 2020. 1 This number can increase given the limited nutritional intake during the COVID-19 pandemic, the impact of which is stunting, which may only manifest itself in the next few years. 2 The World Health Organization has set a maximum tolerance limit for stunting for toddlers at 20%. 1 Talking about malnutrition is often associated with children who do not get enough food. However, the reality in Indonesia and several other countries is more complex. Teenagers face a nutritional crisis. 3 Adolescence is a nutritionally vulnerable period when rapid physical growth increases nutritional needs. Dietary behaviours formed in adolescence can contribute to nutrition-related problems that have long-term health consequences. 3,4 The UN defines adolescents as those aged between 10 and 19, comprising 1.2 billion people worldwide or approximately 16% of the global population. Most of these adolescents live in low-or middleincome countries and face various social and economic challenges. Investing in the rights and development of adolescents can contribute significantly to their full participation in society, a competitive workforce, sustainable economic growth, better governance, and more dynamic civil society. It can also accelerate progress towards achieving the Sustainable Development Goals (SDGs) set by the United Nations. 5 Adolescents are increasingly seen as a window of opportunity with the recognition that investment in the health and well-being of young people is crucial for a country's future and overall development. Evidence shows that adolescence provides a second chance to influence developmental trajectories (including cognitive growth and development), shape future habits, and compensate for some negative childhood experiences -second only to early childhood. 5,6 Investing in adolescents protects future generations of adults and consolidates investments in early childhood health, survival, and education.Malnutrition, in all its forms, can have long-term detrimental effects on the development of children -and can even be passed down from generation to generation. In addition to poor cognitive performance and low work productivity, adolescent girls with anaemia, for example, are more likely to give birth to low birth weight babies at risk of growth stunting. 7 In countries with a high prevalence of stunting, adolescent girls may face additional challenges in their eating behaviour. Several factors that may contribute to this include: first, limited access to nutritious food -adolescent girls may live in areas where access to nutritious PAGE \* Arabic \* MERGEFORMAT 3 food is limited, resulting in a lack of essential nutrients and an increased risk of malnutrition. Second, poverty -adolescent girls who live in poverty may not have the resources to buy varied and nutritious food, leading to food insecurity and malnutrition. Third, beliefs and cultural practices -adolescent girls may be subject to beliefs and cultural practices that limit their food choices, such as food taboos and gender-based food consumption patterns. Fourth, lack of education -adolescent girls may not have access to education on nutrition and healthy eating habits, resulting in poor food choices. Lastly, time constraints -adolescent girls in countries with high stunting prevalence may have to spend much time on household chores, leaving them with little time to prepare food, which can lead to skipping meals or consuming fast food.Healthy habits developed during adolescence, such as healthy eating and physical activity, can last a lifetime and help break the cycle of intergenerational malnutrition. 8 However, programs aimed at improving nutrition during these formative years are still too few, especially concerning preventing stunting, so they must be quickly increased. Therefore, a study was conducted to evaluate the eating behaviour of adolescent girls in several countries with a high prevalence of stunting in young children, above the standard set by WHO. The study aimed to provide an accurate picture of the eating behaviour of adolescent girls and its impact and to serve as a basis for developing future policies and strategies in achieving the SDGs 2030 targets related to nutrition. This review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 9 Additionally, it was registered in the International Prospective Register of Systematic Reviews database (PROSPERO) with registration number CRD42023389909. The search strategy was determined based on research questions as well as inclusion and exclusion criteria based on the PICOS (Participants, Intervention, Comparison, Outcomes, and Study design model 10 as shown in Table 1: The search strategy used in PubMed for this systematic review is designed to identify relevant studies related to adolescent girls' eating behavior and its association with stunting. The strategy includes the following components; Participants (P): MeSH Term: "adolescent" and Keywords in Title/Abstract: "girls", Outcome (O): MeSH Term: "feeding behavior" and Keywords in Title/Abstract: N/A (MeSH term covers the outcome), Exposure (E): Keywords: "growth disorders," "growth," "stunting," "stunted". The Boolean operators "AND" and "OR" are used to combine these components effectively. The search seeks to find studies that involve adolescents (using MeSH terms and title/abstract keywords), address eating behavior (using MeSH term), and explore the association with growth disorders, including stunting (using various keywords). The search strategy aims to be comprehensive by including both MeSH terms and additional keywords to ensure that relevant studies are captured. This approach helps in identifying a wide range of articles that match the research focus of the systematic review.The search started in the PubMed database in January 2023 (as seen in Table 2) and has been filtered since 2000. Later, it was replicated in three other databases: CINAHL, EMBASE, and Scopus. Then, two clusters of search terms were designed. The first term includes terms related to stunting, while the second relates to eating behaviour. Finally, words were selected from the Medical Subject Headings (MeSH) thesaurus to develop an advanced representative search. Study Selection Process This review was conducted in three main stages. The first stage was identifying and removing duplications. The second stage was selecting studies based on inclusion/exclusion criteria by reviewing article titles and abstracts. The final stage was reviewing the eligible articles by reading their full text. Two independent reviewers conducted all of these stages, and if there was a disagreement, a third reviewer was assigned to resolve it. The inter-reviewer agreement for the overall study selection was good, with a Cohen Kappa Index of 0.82 (95% CI, 0.49 to 1.00). 11 The search strategy resulted in a total of 601 potentially relevant studies. Of these, 408 studies met the inclusion criteria based on a review of the title and abstract. After reading the complete text, 393 articles were excluded because they did not meet the inclusion criteria, such as not being conducted in countries with a high prevalence of stunting or not focusing on adolescent girls. Finally, only 15 studies met the eligibility criteria and proceeded to the next stage. The screening process can be seen in Figure 1. The variables reviewed in the selected studies have been described in Table 3: Table 3. Study characteristics Country and design of the studies Most of the studies in this review utilized a cross-sectional design, with only one using a mixedmethod explanatory design 12 . The studies were conducted in India [12][13][14] , Malaysia 15 , Indonesia 3 , Ethiopia 16- 18 , South Africa 19 , Guatemala 20 , Bangladesh 21 , Tanzania 22 , Nigeria 23 , Pakistan 24 , and Timor Leste 25 . Although the target population of this study is adolescent girls, not all studies focus on the same age range, the adolescent period is divided into three stages: early (ages 11-14), mid (ages 15-17), and late (ages 18-21). 3 In this study, two studies only focused on the early stage, 19,24 one study on the mid-stage, 15 three studies on the early to mid-stage, 3,18,25 two studies on the mid to late-stage 13,20 , and the majority of studies focused on the entire adolescent period from early to late 12,14,16,17,[21][22][23] . These studies are listed with the age range mentioned in the text, along with the study number and corresponding reference. There is a difference in sample size among these studies, ranging from 24 subjects in the qualitative design study20 to 4,993 subjects in the sizeable community-based study in Bangladesh. Therefore, this systematic review evaluated 9,831 participants. All studies utilize different assessment tools to evaluate changes in eating behaviour targets. Four studies employed self-designed questionnaires 14,17,23,25 , while the others used standardized, validated scales such as the Food Frequency Questionnaire 13,22 , Dutch Eating Behavior Questionnaire 15 , Adolescent Food Habit Checklist 3 , Household Food Security 16,18 , and Eating Attitude Test 19 . Qualitative research employs assessment tools in the form of open-ended questionnaires delivered through focus group discussions (FGD) 12 and interviews 20,21,24 .Objective, Outcomes, and Result Table 3 also displays this systematic review's objective, outcome, and result. All studies have a similar objective: to assess eating/dietary behaviour in adolescent girls. Four studies link diet habits to the activities of adolescent girls in school 17,20,24,25 , two studies are related to body image perception 12,13 , one study examines the transition of eating behaviour in urban, rural, and slum areas 14 , two studies assess eating disorders directly in adolescent girls 15,19 , three studies examine dietary diversity, patterns, or habits 3,16,21,22 , and the last study aims to provide data information to policymakers or planners 18 . The outcomes of this study can be summarized as dietary restraint, unhealthy dietary patterns, dietary types, and eating behaviour in adolescent girls.The results of the studies generally show the presence of various eating behaviours in adolescent girls in the countries where the studies were conducted. Adolescent girls who diet to maintain body shape may avoid or skip meals to remain thin 12,13,16,24 , especially in urban areas 13 . However, poor eating behaviour was also found among adolescent girls in rural areas, potentially leading to stunting. 14 There is also a tendency among adolescent girls to eat only when they feel hungry. 3 Unfortunately, unhealthy dietary habits associated with excess body weight can lead to the consumption of unhealthy foods. 22 A study also reported a healthy eating pattern of three meals a day, but the meals were not well-balanced. 23 For example, in a PAGE \* Arabic \* MERGEFORMAT 3 study conducted in Timor Leste, adolescent girls were found to consume insufficient calories and protein in their meals. 25 Eating behaviour is also related to Dietary Diversity (DD), which refers to consuming various food groups or types within a specific period. Consuming diverse foods is advantageous for obtaining a variety of macro and micronutrients and is essential for ensuring sufficient nutrient intake. 26 The adolescent girls in this reviewed study were found to have low dietary diversity, associated with several factors, including spending too much time on social media, consuming more sweet foods, and again, fear of gaining weight. 16 Irregular or disrupted eating schedules can also factor in developing unhealthy eating behaviors. Busy school schedules often cause adolescent girls to skip meals or rush through them, leading to loss of appetite or more meal skipping and irregular snacking. 20 Inadequate dietary diversity was found in adolescent girls who consume few fruits and a large number of starchy staples. 17 The type of food also became the focus of the review study. Adolescent girls in Ethiopia more frequently consume staple foods such as rice or wheat than non-staple foods such as eggs, meat, and fish. Fruits are more commonly consumed according to their seasonal availability. 18,21 Possible eating disorders were reported in one study. 19 The Newcastle-Ottawa Quality Assessment Scale (NOS) 27,28 was utilized to evaluate the quality of the studies. This tool was designed to aid systematic reviews in reaching quality criteria while minimizing bias risks. The NOS comprises three main categories: selection, comparability, and outcome, each consisting of several items about the features of observational studies, with several answer options. At least one answer within each category is accompanied by a star (6) which signifies a low risk of bias. Upon completion of the scale, the stars are tallied, and if the score is less than seven stars, there may be a high risk of bias. Additionally, if the reviewers select answers without stars, it is important to examine possible study biases. NOS scores are categorized into three groups: very high risk of bias (0-3 stars), high risk of bias (4-6 stars), and low risk of bias (7-9 stars). 27 Two independent reviewers assessed the quality of each study, with a third reviewer assigned in case of disagreements. The final consensus is presented in Table 4. The Cohen Kappa Index was used to measure the level of agreement among reviewers in assessing the overall quality of the studies. The findings indicated a strong agreement (Cohen Kappa Index = 0.76; 95% CI, 0.60 to 0.89) in evaluating the individual scores of each study. Based on this analysis, it can be inferred that all fifteen studies had a minimal risk of bias. During adolescence, there is a need for greater nutritional intake to support rapid growth and ensure a healthy lifestyle for teenagers. An additional study indicated that teenagers' choices regarding their lifestyle and diet could either negatively affect their health during this transition period or contribute positively to their overall well-being as they become adults. 29 In this systematic review involving fifteen studies, a comprehensive overview of eating behaviour in adolescent girls in countries with high stunting prevalence was obtained. The observed eating behaviours from this systematic review include eating disorder 15,19 , weight reduction diet 12,13 , restrained eating 15,[20][21][22][23] , craving-induced eating 3 , unhealthy dietary pattern 23 , and low diversity diet 16,17 . The first crucial finding of this systematic review underscores the connection between eating disorders and extreme weight-loss behaviours among adolescent girls. Only four per cent of the surveyed teenage girls were diagnosed with an eating disorder using the EAT-26 measurement tool. Notably, individuals who engaged in extreme weight-loss behaviours in the past year were likelier to score higher on the EAT-26, indicating a potential risk of eating disorders. 19 These results emphasise the need for early identification and support for individuals at risk for or already experiencing disordered eating patterns. Psychological, genetic, environmental, and cultural factors may contribute to the development of eating disorders. 30 Additionally, the study found that perceiving oneself as overweight was a predictor of higher total scores on the EAT-26 and a score equal to or greater than 20, often used as a cut-off point for identifying individuals at risk for an eating disorder. This finding aligns with previous research showing a correlation between perceiving oneself as overweight and experiencing eating problems during adolescence. It also suggests that individuals who perceive themselves as overweight may be more likely to engage in weightloss behaviours. 31 These findings highlight the importance of recognising the relationship between extreme weight-loss behaviours, body perception, and the risk of developing an eating disorder. Identifying individuals who exhibit these behaviours and perceptions can help in early intervention and support for individuals at risk for or already experiencing disordered eating patterns. 19 Our second key finding revolves around body perception and its role in the risk of developing eating disorders among adolescent girls. Notably, perceiving oneself as overweight predicted higher total scores on the EAT-26. This aligns with prior research showing a correlation between perceiving oneself as overweight and engaging in extreme weight-loss behaviours. Recognising this relationship is crucial for targeted interventions and support.Adolescent girls commonly experience dissatisfaction with their body image and negative selfperception, affecting their eating behaviour. In the study we reviewed, most participants reported skipping at least one meal per day. Among this group, some intentionally skipped meals to lose weight, commonly called dieting. When comparing the practice of dieting among all the reasons for skipping meals with the participants' self-perception of their body shape, it was observed that those who engaged in dieting were only individuals who perceived their body shape as fat or normal, while none of those who considered themselves as thin reported dieting. 12 PAGE \* Arabic \* MERGEFORMAT 3 In our third finding, we delve into body image's influence on adolescent girls' dieting behaviours. It was observed that those who engaged in dieting were exclusively individuals who perceived their body shape as fat or normal, while none of those who considered themselves thin reported dieting. This underscores the complexity of body image perceptions and their impact on adolescent dietary choices.Initially, the process of perceiving body image is believed to be narrow. Still, over time it can develop into body distortion, where there is a significant alteration in the perception of body size. This means that individuals may perceive themselves as fat even when dangerously underweight. 32 This tendency to desire thinness was also observed quantitatively in the current study. There was a lack of agreement between adolescents' perceptions of themselves and their actual body size. Therefore, it became evident that their beliefs about their bodies were based on a mistaken notion. It was concerning that those girls who perceived themselves as normal or fat in the present study were underweight. A study conducted in Korea on adolescents 33 found that underweight or normal-weight individuals who perceived themselves as overweight were at a higher risk of developing eating disorders such as anorexia nervosa. Our fourth important finding highlights restrained eating (RE) behaviors among young adolescents. This vulnerability is linked to increased weight concerns, body dissatisfaction, and problematic eating behaviors. Adolescents who exhibited RE behaviors tended to eat only when hungry or craving, which contradicts previous notions. Understanding the factors contributing to craving-induced diet behavior, such as limited access to food at school, is critical for developing effective interventions.The restrained eating (RE) identified in four reviewed studies 15,[20][21][22][23] indicates that young adolescents are especially susceptible to RE. This vulnerability stems from their increased likelihood of experiencing weight concerns, body dissatisfaction, and problematic eating behaviours. The restrained eating behaviour exhibited by adolescents in Mahriani's33 3 study involves only eating when hungry or craving, contradicting findings that suggest adolescents frequently experience hunger. 34,35 However, Mahriani's 3 study reveals that the craving-induced diet behaviour is primarily attributed to adolescents' lack of knowledge regarding the benefits of proper nutrition and the difficulty in accessing food while they are at school. Our fifth finding sheds light on the unhealthy eating patterns prevalent during adolescence, characterized by skipped meals, increased consumption of processed and fast foods, and decreased intake of fruits and vegetables. Despite knowledge about the importance of a balanced diet, adolescents frequently indulged in junk food, indicating that factors beyond knowledge influence their dietary choices. During adolescence, characterized by rapid growth, individuals experience significant physical, psychological, hormonal, cognitive, and social changes. These transformations lead to fluctuations in the nutritional requirements, eating habits, and food preferences of the body. The eating pattern of adolescents differs from that of children. The higher academic and socio-economic pressures associated with this growth stage often result in skipped meals, an increased intake of processed and fast foods, and a decrease in the consumption of fruits and vegetables. 36 The unhealthy eating patterns identified in the reviewed study indicate that adolescents tend skipped meals or to have inadequate consumption of vegetables and fruits while frequently indulging in junk food. 23 Furthermore, despite knowing the importance of a balanced diet for maintaining a healthy life, this knowledge did not reflect in the nutritional status of the respondents. This suggests that factors other than knowledge influence the dietary choices of adolescent girls. 37 The last crucial finding pertains to the presence of a low-diversity diet among adolescent girls, reflected in sub-optimal dietary intake and micronutrient deficiencies. 16,17 This sub-optimal nutrition during adolescence can lead to delayed puberty, contracted pelvis, and unfavourable birth outcomes, including stunted growth in newborns. Recognising the importance of dietary diversity is essential for addressing these nutritional challenges.The dietary diversity score (DDS) is a measure that counts the number of different food groups consumed over a specific period. 38 It reflects the diet quality at the household or individual level. Optimal nutrition is crucial during adolescence since this period is responsible for gaining 50% of adult weight, 20% of adult height, and 50% of skeletal mass. 39 However, research indicates that 45-60% of adolescent girls have sub-optimal dietary intake 40 , leading to various micronutrient deficiencies, including Vitamin A, iron, and iodine 17 . Similar findings were observed in Ethiopia, where 29% and 30% of adolescent girls had thinness and anaemia, respectively. 41 These nutritional deficiencies can result in delayed puberty, contracted pelvis, and unfavourable birth outcomes, including stunted growth in newborns. 7 While this systematic review provides valuable insights into eating behaviours among adolescent girls in countries with high stunting prevalence, it is essential to acknowledge its limitations: Limited Generalizability: The findings of this review are based on studies conducted in specific countries or regions with high stunting prevalence. As such, the generalizability of the results to a broader global context may be limited. Cultural, socioeconomic, and environmental factors that influence eating behaviors can vary significantly between regions, potentially affecting the applicability of these findings elsewhere.Variability in Study Designs: The included studies in this review may have employed diverse methodologies, assessment tools, and criteria for measuring eating behaviors. This variability in study designs could introduce heterogeneity into the review, making it challenging to draw uniform conclusions. Language Bias: The review's search strategy may have been limited to studies published in English, which could introduce language bias. Relevant studies published in other languages may have been excluded, potentially affecting the comprehensiveness of the review.Publication Bias: There is a possibility of publication bias, where studies with statistically significant or more sensational findings are more likely to be published. This bias could skew the review's findings, potentially overemphasizing certain aspects of eating behaviors among adolescent girls.Cross-Sectional Nature of Studies: Many of the included studies may have adopted cross-sectional designs, which provide a snapshot of eating behaviors at a specific point in time. Cross-sectional studies are limited in their ability to establish causality or track changes in eating behaviors over time.Self-Reported Data: Several studies may have relied on self-reported data from adolescents. Selfreporting can introduce recall and social desirability biases, as participants may underreport or overreport certain behaviours due to social or cultural pressures.Lack of Longitudinal Data: Longitudinal data that track eating behaviors and their consequences over an extended period are crucial for understanding trends and potential long-term effects. The absence of such longitudinal data may limit the review's ability to explore the persistence and impact of certain eating behaviours.Limited Focus on Interventions: This review primarily describes eating behaviours rather than evaluating interventions or prevention programs. While identifying eating behaviors is valuable, a more comprehensive review might explore the effectiveness of interventions in addressing these behaviours.Evolving Nature of the Field: The field of nutrition and eating behaviors continually evolves. The review's findings are based on studies available up to a specific date, with the most recent sources cited in this systematic review published in 2022. 3 Since then, new research may have emerged that could provide additional insights or alter the conclusions.Potential Biases in Included Studies: The individual studies in the review may have biases and limitations, which can propagate to the review itself. The quality and rigor of the original studies can impact the overall strength of the review's conclusions. In conclusion, the eating behaviours of adolescent girls in countries with a high prevalence of stunting exhibit concerning patterns, including a low diversity diet, sub-optimal dietary intake, and the development of various micronutrient deficiencies. These challenges are compounded by eating disorders, such as retained and craving-induced diet, which further impact their nutritional status. These eating disorders contribute to an unhealthy relationship with food and can harm their physical and mental wellbeing. Addressing these multifaceted issues requires a comprehensive approach encompassing nutritional education, promoting balanced diets, addressing the underlying factors contributing to eating disorders, and providing appropriate support and intervention for those affected. By addressing these eating behaviours and disorders, we can strive to improve adolescent girls' overall health and well-being in these countries and reduce the prevalence of stunting and its associated health consequence.