Observational study: determinants for vegetable consumption among elementary school students in Indonesia [version 1; peer review: awaiting peer review]

Background : The prevalence of obesity in school children has been increasing worldwide, including in developing countries such as Indonesia. Childhood obesity is associated with an unbalanced diet, especially the lack of fruit and vegetable consumption. This study aimed to determine the factors that influence vegetable consumption behavior in school children in Indonesia. Methods : The cross-sectional study was conducted in eight public elementary schools in a suburban area of Bogor Regency involving 433 students from the fifth grade. Data collection was conducted from November to December 2018. Data on the students’ characteristics, knowledge, and attitude as well as availability of vegetables at home and at school were collected using a structured questionnaire. While, a 24-hour food recall was utilized to obtain information on vegetable consumption and overall diet. Data analysis was done using Chi-squared test and logistic regression. Results : The average daily intake of vegetables was 47.11 g, far below the recommend amount (250 g). Female students had significantly better vegetables intake compared with their male counterparts. The availability of vegetables at home and students’ positive attitude toward vegetables were also significantly associated with better vegetable intake. Logistic regression analysis showed that availability of vegetables at home almost tripled the likelihood of better vegetable consumption (OR 2.822, 95% CI 1.822–4.372). Additionally, availability


Introduction
The prevalence of childhood overweight and obesity has been increasing globally, with an estimated 124 million children and adolescents aged 5-19 years considered obese and 213 million classified as overweight in 2016. 1 This includes developing countries like Indonesia, which causes it to suffer from a double burden due to malnutrition. Within five years, the proportion of children aged 5-12 years who are overweight in the country increased from 18.8% (10.8% overweight and 8% obese) in 2013 to 20% (10.8% overweight and 9.2% obese) in 2018. 2 This can be associated with poor eating habits, as many Indonesian children lack the consumption of fiber, vegetables, and fruits. In addition, they often eat food with a high fat and sugar content, 3,4 generally contained in junk food, fast food, or frozen food. 5 They also lack physical activity, with an increasing amount of time spent watching TV or on gadgets. 6 Children tend to consume food that is appealing to their tastes and preferences; this lack of diversity leads to poor nutrient intake such as limited protein sources and a lack of fruits and vegetables for vitamins and minerals. 7 Therefore, support from the surrounding environment is pivotal; families and schools can help develop healthier eating habits that will affect the nutritional status, growth, and development of children. 8 Despite of their health benefits, children, caregivers, and schools are often unaware of the importance of adequate vegetable consumption for children. An Indonesian study by Sinaga et al. 9 found that the average consumption of vegetables among school children was only 32 g and 62.7 g of fruit, far below the recommended daily amount of 250 g and 150 g for vegetables and fruits, respectively. The negative health impact of skipping vegetables are not immediately evident, but the long-term effect can lead to various degenerative diseases, such as obesity, 10 hypertension, 11 cancer, 12 diabetes, 13 cardiovascular diseases, 14 and digestive problems. 15 Choosing food is human nature, especially in children who tend to choose food that looks attractive and delicious but is sometimes unhealthy; they usually prefer sweet and savory foods to healthy foods such as fruit and vegetables. Currently, unhealthy and non-nutritive foods are very easy to find with cheap and affordable prices, compared with healthy, balanced, nutritious foods. On the other hand, children's eating behavior tends to be formed early in their life and maintained into adulthood, therefore the role of the food environment at home is imperative. 16,17 Parents function as gatekeepers who can support and model healthy dietary behavior for their children. 18 Parents create food choices through the purchase and preparation of food and indirectly influence children's food choices by setting rules, providing information, and shaping behavior. 19 However, the food environment at home might not be the only factor influencing children's vegetable intake. There are other modifiable factors such as the food environment at school, as well as the children's knowledge and attitudes toward their diet. Many efforts have been implemented to try to improve vegetable and fruits consumptions among children in Indonesia. However, we are yet to see the results. Against this backdrop, this study aimed to determine the current average vegetable consumption of school children in Indonesia and the factors that influence this vegetable consumption behavior. We included personal level factors such as students' knowledge and attitude towards vegetables as well as their food environment at home and at school to identify which are the determining factors for their actual vegetable consumption. The result will provide important insight for program design and policy makers on their decision-making process in curbing childhood obesity.

Methods
The data collection for this cross-sectional study was conducted from November to December 2018, thus was before the pandemic and health and safety measures to limit Coronavirus disease 2019 (COVID-19) transmission was not implemented. The study was conducted in Cibinong Sub District in Bogor Regency, West Java, Indonesia. Cibinong was chosen as the study area because it is among the most populated area in the district, and has the largest number of schools and students. We limited the research among public elementary schools because the study was formative stage research linked to an intervention model design and advocacy for the district education office. Out of 92 public elementary schools in Cibinong subdistrict, we randomly selected 10 schools to represent 10 villages in the study, but only eight of them agreed to be involved in this study. This commitment was central since we provided a follow-up intervention after the study. The public elementary schools that were willing to participate are spread over eight villages in Cibinong subistrict. We are aware that this approach might lead to selection bias, since we only involved schools that were willing to participate in our follow-up intervention which might indicate that the school community, including their students, might have better acceptance towards facing nutritional issues. However, the random selection process of villages, schools, and students involved in the study would help in minimizing it.
The total number of students from the eight schools was 4334 students, thus 10% of the students (433) were sampled as respondents. Only students from the fifth grade were involved because they were assumed as already being fluent in reading and writing, able to understand the questions in the questionnaire, and able to answer and remember what they ate during the 24-hour food recall interview. The inclusion criteria for schools and subjects were as follows: fifth graders of public elementary school, did not have any long-term physical illness or mental incapacity that would hinder their ability to involve in the study, willing to become research subjects by signing an informed consent and assent by parents or teachers, and were able to present during the research activity. Students who were not willing to participate and those allergic to certain vegetables were excluded.
We employed stratified random sampling, with probability proportionate to size method to obtain the number of samples per school. The random number was then generated in excel, based on the students' attendance list in each school. The sampling calculation process is presented below in Figure 1.
Data were collected using a structured questionnaire guided by eight trained enumerators, the data included student characteristics, availability of vegetables at home and at school (including snack foods containing vegetables), eating habits, vegetable consumption, students nutrition knowledge, and attitudes toward balanced nutrition. 20 The vegetable intake was calculated from a 24-hour food recall assisted with an Indonesian food photograph album to help students in recalling the food and the quantity. The summary of research variables is presented in Table 1. The questionnaire was tested for validity and reliability on 30 students with the same characteristics as the respondents in this study. The validity and reliability test showed that 10 items of the nutritional knowledge questions had a cronbach alpha (0.957), 10 items of the attitude statements (0.953), and six items of the consumption behavior statement (0.943).
Students' vegetable consumption was divided into "good" (consuming vegetables at every meal and every day) and "poor" (never or rarely consuming vegetables) based on the questionnaire. The variable of vegetable availability at home and at school was translated as the availability of home food and street food sold at school that contain vegetables such as pecel, urap, gado-gado, soup, etc. People who provide food at home can be mothers or helpers or other people who buy, select, process, and cook food at home. The knowledge variables were assessed from questions and statements regarding balanced nutrition, among others; symbols, slogans, pillars, and the content of balanced diet derived from the national guideline for behavior change communication materials for elementary school age children. The questions and statement also covered the benefits of consuming vegetables, the impact of not eating or not eating enough vegetables, and the recommended portion of vegetables.

Results
Results showed that the average vegetable consumption was still low at 47.11 g, with the maximum consumption 155 g per day, which is far from the recommended daily consumption of 250 g of vegetables. Table 2 also shows the average energy intake was 1244.98 kcal, also far below the recommended daily allowance (RDA) for children aged 10-12 years of 1900 kcal for girls and 2000 kcal for boys. However, 4.2% of students exceeded the recommended energy consumption, having above 2000 kcal. Similiarly, the average protein intake was 43.42 g, which is still below the RDA of 50 g for boys and 55 g for girls. But 18.9% of students consumed protein exceeding the recommended amount. For fat intake, the average was 45.87 g which is below the RDA of 65 g, while 9.5% of students exceeded the RDA. The average fiber intake of all respondents was only 4.72 g, while the RDA value for children aged 10-12 years is 28 g for boys and 27 g for girls. 22 Table 3 shows the relationship between students vegetable consumption behavior and potential determining variables examined in this study. There is a significant relationship between vegetable consumption and sex (p = 0.005), with girls tending to eat more vegetables than boys. Furthermore, there is a significant relationship between eating vegetables and the availability of vegetables at home (p = 0.001). More students who stated that vegetables were rarely available at their home tended to have poor vegetable consumption. Table 4 shows the research variables that reached the requirements for multivariate logistic regression analysis (p = 0.25) those variables were knowledge, availability of vegetables at home and at school, sex, and student attitudes. Analysis

Variable
Operational definition

Student
Fifth grade students at selected SDN in Cibinong District, Bogor Regency, aged 10-12 years.

Student characteristics
Including age and sex of the students. Age is the length of time a student has lived (since he was born). Sex is the difference between male and female physiological and anatomical attributes.
Food availability Availability of vegetables at home and at school.

Knowledge
Students' abilities to recall and relate nutritional knowledge/information.

Attitude
Positive/negative or neutral feelings/perceptions that can be felt by students regarding nutrition.
Food/vegetable consumption behavior The behavior of consuming vegetables every day.
Food recall Interviewing students, then recording the type and amount of food consumed by the respondent in 24 hours.

Lack of vegetable consumption among children
Our study found that vegetable consumption and other macronutrient intake was low among children participating in our study. Regarding vegetables, it showed that girls had better vegetable consumption compared to boys, this finding   23 which stated that women tend to consume more vegetables. The national balanced diet guideline recommended a minimum of 250 g of vegetables per day. This amount can be converted to household size, one serving of vegetables is equal to one bowl of fresh vegetables or half a bowl of cooked vegetables. More than 80% of students in this study consumed far below this RDA.
Lack of vegetable consumption can be caused by various factors including the unavailability of vegetables at home 24 such as found in this study. However, the process can be two-fold, the lack of availability may be due to the lack of demand from children and the lack of demand might be caused by lack of creativity in preparing vegetables so that they are appealing to children. Vegetables are not naturally appetizing to children, thus the use of spices and herbs are needed to increase vegetable intake in children. 25 Studies have found that the absence of parental support for fruit and vegetable consumption, especially from mothers as educators as well as food providers, is a risk factor for a lack of consumption among children. 26,27 While a review on experimental research showed that the change in eating habits resulting from providing more fruits and vegetables to children has been proven to influence their long-term eating behavior. 28 Contento stated that consumption behavior is the choice of food and other actions related to food and nutrition that people do to achieve the desired effect of their own choices. 29 Furthermore, the food selection process involves six categories, namely taste, habits, health, diet, rules, and filling food. The choice category becomes an individual's consideration in choosing the food to be consumed. 30 In addition, life experiences have a major influence on food choices which include ideals, personal factors, resources, social contexts, and food contexts. These influences inform the development of personal systems for making food choices that combine value negotiation and behavioral strategies. 31 While in children, there are four things that influence their food choices: (1) main factors including hunger level, preferences (based on taste, familiarity, experience), time pressure and priorities (children tend to choose fast food when they have limited time), ability to prepare food, and desired impact; (2) secondary factors such as peers, media, supermarkets and schools; (3) action, which are the actions taken by parents in influencing children's food choices; (4) time, children's activities will form a pattern of food selection over time which is influenced by the past (experience) and future plans. 19 The challenge from mental exercise can also influence food choice in children, a study found that children choose healthier foods to a greater degree when they have to do mathematical exercises than when they do not. 32 Hence, food selection is clearly the end result of a decision-making process about the food that a person will consume.

Home and school environment influence
Findings regarding the influence of social environment found that interventions which focused on promoting a positive environment, by improving the availability and accessibility of healthier food as well as their taste preference toward these healthier choices, was seen as an effective approach. 33 The Voices for Healthy Kids, a multi-sectoral collaboration promoting food, environment, physical activity, and a healthy weight for all children and adolescents in the United States of America, proved that multi-sectoral collaboration involving many groups and stakeholders is key in responding to increasing prevalence of obesity in children and adolescents. 34 Availability and accessibility, fruit and vegetable preference, and taste have been the most consistent determinants for consumption. 35 Therefore, this study suggests that mothers should not to give up and should always provide vegetables at home, because children's food preferences and desire to try new foods are influenced by the people around them. 36 A study showed that children whose mothers work are less likely to have breakfast than children with non-working mothers. 37 There are developmental differences between the perceptions of children and adolescents about the factors that influence food choices. In adolescents, parental control begins to decrease and adolescents have more independence in their food choices compared to children. 38 Therefore, to improve diets, especially vegetable intake, it is very important to gather input from children themselves regarding the factors that are thought to influence their food choices.
Our result showed that the availability of vegetables at home tripled the odds for having better vegetable consumption. Studies have shown that experiences in early life and the family have a prominent role in encouraging healthy eating patterns that will last throughout a person's life. Parental eating habits and feeding strategies are the most dominant determinants of children's eating behavior and food choices. 39 Thus, it is imperative for parents, as the closest environment for children, to always monitor children's eating habits to avoid nutritional problems such as malnutrition, obesity, diabetes, and other diseases due to deficiency or excess nutrition. Mothers play the main role in providing and managing food in the family, mothers are most dominantly responsible for feeding children 40 and other family members.
In particular, parents play the roles of health promoters, role models, and educators in children's lives, influencing their cognition and eating choices. Food availability and the influence of parenting patterns have a strong relationship in the consumption of healthy and unhealthy foods in children. 41 Children are more likely to eat foods that are readily available and they tend to eat more when larger portions are provided. 42 Parents function as gatekeepers who can support and control healthy dietary behavior in children. Providing nutritional knowledge and parental support regarding the nutritional needs of children is important, for example through the provision of supplies to school and the management of nutritious and safe food for consumption at home. 43 It is suggested that parents create food choices through the purchase and preparation of food and indirectly influence children's food choices by setting a number of rules, providing information, and shaping their behavior. 19 Likewise, parents also play an important role in the formation of eating habits and preferences, for example the tendency to want to or be afraid to try new foods. 44 Previous eating experiences can be a major determinant of children's eating development and acceptance of diverse foods. Following the law of supply and demand, parents often prepare only foods that their children like, which sometimes ignores the importance of nutritional diversity in food and offers fewer experiences of trying new foods. Thus, it is advised for parents to provide healthy food, and wherever possible accompany children and eat with family members, creating positive and enjoyable eating situations to help children develop healthy eating habits. Such an approach can be translated into a behavior change communication strategy through school-based nutrition programs where parents, especially mothers, are involved and provided not only with knowledge about the importance of a balanced diet for their children, but also to improve their ability to prepare healthy food that is affordable and appetizing for their children.
Nowadays many parents only provide energy-dense food which mostly comes from carbohydrates, then protein sources such as eggs and chicken because they are obtained at a low price, taste good, and are easy to manage and cook. Therefore, these foodstuffs are more widely chosen and consumed by children compared to foods with sufficient fiber including vegetables, fruit, and other protein sources such as fish (not preferred because processing starts from the cleaning process, has a fishy smell and contains bones). The practice of feeding children food that only contains carbohydrates and protein, if consumed continuously in excess, will have an impact on weight gain. To meet the challenge of providing children with healthy, balanced nutritious food choices, parents need exposure to information and knowledge about nutrition. The better a person's nutritional knowledge, the more attention will be paid to the quality and quantity of food they consume. 45 The provision of early nutrition and food education in children plays an important role in shaping food choices, diet quality, and nutritional status. 46 The provision of early nutrition and food education in children plays an important role in shaping food choices, diet quality, and nutritional status. Therefore, it is important to involve parents, especially mothers, in children's nutrition education activities. It is a good opportunity to organize nutrition education activities in schools because it targets students, parents, and also teachers who have an important role in shaping and determining children's eating habits.
In addition to the food environment at home, school is the second most frequented environment for children, they spend approximately five to six hours with learning activities, playing, and exercising, so they need adequate food intake and healthy food environment at school to supply their need. The selection of healthy, safe, and nutrient-rich snacks at school, therefore, is also important. However, not all school canteens provide healthy and safe snacks. A Bogor District Health Office survey report on 2018, showed that almost 80% of the snacks sold in canteens and around schools in Bogor Regency did not meet the clean and hygienic requirements. Referring to our finding, that most children did not meet their macronutrient needs showed that food insecurity might still be an issue. The adverse impact of food insecurity in children can affect their academic development and skills as well as their nutritional status. 47 There is also a strong association between food safety and malnutrition, where children who eat unsafe foods are almost three times more likely to suffer from malnutrition than children who eat safe foods. 48 According to Brown et al. (2011) 49 schools are responsible for providing quality food for students. This must be included in the school education program, because school-age children are starting to develop their independence and distance themselves from parental supervision, especially in terms of choosing to eat at school. They can choose and make their own decisions regarding what snacks they consume during breaks, this highlighted the need for regulation of snacks in schools. 50 Our preliminary study in the eight schools involved in this study did not find any school policies regulating food safety and hygiene, we only found an oral appeal from government agencies to schools in their efforts to minimize children's consumption of unhealthy food and drinks. While globally, the movement for malnutrition and obesity prevention in children should also cover not only hygiene but also sugar contents. However, it requires hard work and better enforcement. A study found that in some cases, the sale of energy drinks is prohibited in schools which is supported by local authorities, but children perceive that schools allow energy drinks to be brought and consumed during class. This proves the lack of enforcement of the existing rules in schools regarding snacks that are safe for consumption by children. 51

Strengths and limitations
The study identifies the determinants of vegetable consumption among elementary school students by gathering information directly from them. It also involved a large number of participants within the same age range from various schools in the study area. The use of validated tools by trained enumerators also helped to improve the data quality. However, the study was conducted at a subnational level thus it limits the generalizability to other settings. Additionally, we did not record respondent's food consumption data for more days (for example three or seven days of food recall), due to considerations of elementary school children's memory, so it was limited to only 24 hours of food recall. We also did not include teachers and parents in the study to validate our findings since they also have an important role in the growth and development of elementary school children. Hence, for future research, in addition to increasing the scale of the study, we suggest including mothers and teachers to triangulate results.

Conclusion
The average consumption of vegetables, energy, protein, fat, and fiber among the elementary school children in our study were below the RDA. Especially for their average vegetable's consumption, which was far below the recommended portion. There is a significant relationship between vegetable consumption and sex, availability of vegetables at home, and student attitudes. Vegetable availability at home is the strongest determinant influencing vegetable consumption among elementary school students involved in the study. Hence, parental support or the home food environment is key in improving their diet quality to prevent obesity. Suggestions for future research are to include mothers and teachers as respondents, as well as enrolling participants in intervention program activities to increase vegetable consumption in elementary school children

Ethics
This research permit has been approved by the Health Research Ethics Committee, Faculty of Medicine, UPN Veteran Jakarta, which was obtained in the form of ethical approval No. B/1708/2/2019/KEPK.

Consent
Written informed consent for publication of the participants' details was obtained from the participants in addition to assent by parents or teachers.