Preference for and consumption of traditional and fast foods among adolescents in Indonesia

This study aimed to examine the preference for and consumption of traditional and fast foods among adolescents in Indonesia and their relationship to social, economic, demographic, and regional factors. This cross - sectional study was conducted online in six provinces in Java and Bali (JB) and nine provinces outside Java and Bali (OJB), involving 4,500 junior high school, high school and university students. Questionnaire items that have been tested included personal and family information, food preferences (three Likert scales), frequency of food consumption (never, ≤ 1/month, 2 - 3 times/month, and ≥4 times/ month) for about 10 – 20 types of traditional food and 17 types of fast food that varies among the provinces. The t - test was applied to analyze the differences in scores and frequency of food consumption in Java and Bali (JB) and the provinces outside Java and Bali (OJB). Logistic regression was employed to discover factors that contributed to the frequency of consumption of traditional food and fast food. In the provinces outside Java and Bali, traditional foods—especially groups of food made with cereals, vegetables, seafood, bean and legumes, and traditional snacks—are preferred and consumed over fast food. In Java and Bali, fast food is preferred over traditional food. However, some food groups such as cereals, vegetables and seafood are consumed more in the form of traditional food and snacks. The results showed that in Java and Bali, most adolescents preferred fast food to traditional food (P<0.05). Although fast food was preferred, adolescents in Java and Bali more than often consumed traditional foods made with cereals, vegetables, seafood, beans and legumes, and traditional snacks; because traditional food is more affordable (P<0.05). Meanwhile, in provinces outside Java and Bali, adolescents generally prefer traditional food to fast food. This is in line with the frequency of consumption of traditional food which was also higher than that of fast food from most food groups (P<0.05). Consumption of traditional food and fast food is influenced by several social, economic, and demographic factors such as gender, place of residence, educational level of the adolescents, parental occupation and income, as well as online food purchase habits (P<0.05)


Introduction
The term "nutrition transition" was coined in the early 1990s, describing changes in food consumption patterns and energy expenditure among individuals. In some countries, this transition occurs because of changes in demographic, socio-economic and epidemiological conditions. In developing countries, this transition is indicated, among other things, by changes in food consumption patterns among the people. Diets based on traditional foods rich in grains, vegetables, fruit, and local foods have shifted to western-style diets high in sugar, saturated fat, simple carbohydrates and animalbased foods (Popkin, 1993).
Similarly, adolescents living in low-to-middleincome countries (LMICs) also experience a change in the pattern of food consumption from traditional food to Western food consumption, which is then followed by an increase in the prevalence of obesity (Popkin et al., 2012). A study conducted in 54 LMICs by Li et al. (2020) revealed that around 55% of adolescents consume fast food at least once a week. A national survey in Malaysia revealed that almost 70% of adolescents aged 10-18 years consume fast food 1-3 times a week. The term fast food mostly refers to western-style fast food that is easy, fast, and ready to eat but is not nutritionally balanced. This diet is high in sugar, sodium, and saturated fat but low in fiber, which leads to insufficient nutritional needs for a healthy life (Fuhrman, 2018;Ashdown-Franks et al., 2019;MacKay et al., 2021).
Analysis of multi-center data from 36 countries was conducted by comparing the age groups of children and adolescents. In both age groups of children and adolescents, those who eat fast food more often have a higher BMI than those who rarely eat it (Braithwaite et al., 2014). In cross-sectional and cohort studies, it has been shown that frequent consumption of fast food affects markers for metabolic syndromes, such as Body Mass Index (BMI), body fat, insulin, glucose, triglycerides and Low Density Lipoprotein (LDL) (Bahadoran et al., 2013;Marlatt et al., 2016).
A study conducted by Pengpid and Peltzer (2016) on adolescents in the Association of Southeast Asian Nations (ASEAN) countries showed that adolescents who consumed fast food twice or more a week were at risk of becoming overweight or obese. A study in Jakarta found that elementary school children who consumed fast food more than three times a week had a 4.2 times greater risk of obesity than those who consumed fast food less than three times a week (Febriani and Sudarti, 2019). The Indonesian National Health Survey reported an increase in obese adolescents (aged 16-18 years) from 7.3-13.5% in 2013 and 2018.
Indonesia is a developing country with rapid economic growth, followed by changes in food expenditure among its people. Expenditure on processed food (including fast food) increased from 16.72% in 2002 to 34.27% in 2020, and expenditure on processed food in urban areas was greater (38%) than that in rural areas (27%) (Rachman and Suryani, 2008;Statistics Indonesia, 2020). A cohort study in 13 provinces in Indonesia from 1993-2015 revealed a shift in food consumption towards Western-style food, such as animal -based food, packaged food, and ready-to-eat food. However, the preference for traditional foods, those were cereals and processed plant foods, remains high in both urban and rural areas (Colozza and Avendano, 2019).
Food preference analysis has been widely used to estimate food shopping choices in consumer research. Final food choices depend on several factors such as health, price, weight control, familiarity, ethical concern, and food preference (Vabø and Hansen, 2014). There is a strong correlation between food consumption habits, taste, and preferences (Contento et al., 1995). Meanwhile, Drewnowski and Hann (1999) showed that food consumption among university students, as measured by food frequency, was related to their food preferences. There are not many studies in Indonesia that examine preference for and consumption of certain foods in adolescent groups across social, economic, cultural, and ethnic aspects and across islands. Therefore, this study examined the preference for and consumption of traditional food and fast food and their relationship with social, economic, demographic, and regional factors in Indonesia.

Research design
This study used a cross-sectional design and online data collection. Data were collected from fifteen provinces in Indonesia, divided into the Java and Bali (JB) regions and outside the islands of Java and Bali (OJB). The JB region has consisted of six provinces, those are Jakarta, West Java, Central Java, Yogyakarta, East Java, and Bali; and the OJB covered nine provinces such as North Sumatra, West Sumatra, Riau Islands, Jambi, South Kalimantan, South Sulawesi, Gorontalo, West Nusa Tenggara, and East Nusa Tenggara. The difference in the prevalence of stunting among adolescents aged 13-18 ranges from the lowest in Bali (9.8%) to the highest in East Nusa Tenggara (41.0%).

Sampling method
The study sample was adolescents, consisting of groups of junior high school students (13-15 years), senior high school students (15-18 years), and university students (18-22 years old). The minimum sample size was based on Lemeshow et al. (1990) with the assumption that the proportion of adolescents who consume less fruit and vegetables in Indonesia is 96.4% (Basic Health Research, 2018), and the precision value (d) is 0.05. The sampling was performed from adolescents with less consumption of fruit and vegetables since a systematic review reported that fast food consumption is linked to poor dietary habits i.e. low intake of fruits and vegetables, among adolescents (Li et al., 2020;Rachmi et al., 2020). Thus, the minimum sample size of 93 people per school was obtained, and the minimum total sample size was 4,500 people. A sample size of less than 100 (2.2%) cannot be analyzed further because the returned questionnaires will provide incomplete data. This study collaborated with local university partners from 15 provinces to collect data. The university partners were tasked with determining the locations of the schools, managing research permits, and coordinating with schools in online data collection. The https://doi.org/10.26656/fr.2017.7(4).156 © 2023 The Authors. Published by Rynnye Lyan Resources FULL PAPER samples were determined using quota sampling by first selecting schools purposively to obtain schools located in urban and rural areas.

Ethical approval
This study has received ethical approval from the Commission on Ethics for Research on Human Subjects, Bogor Agricultural University, with approval number 493/IT3.KEPMSM-IPB/SK/2021. Each respondent gave written informed consent after the background and research objectives were explained to them.

Research instrument and data collection
Data were collected by conducting an online survey using google forms. The food list contained in the FFQ that was used in this study was determined through Focus Group Discussion (FGD) with local research partners in each research area to decide the food items for each of the respective regions. The initial questionnaire developed was tested twice among junior high, high school and college students with a total of 128 respondents. The first trial was aimed to observe the familiarity and the frequency of the food items listed in the questionnaire. Food items with low familiarity were then excluded from the questionnaire. Next, the second trial was performed to check the comprehensiveness of the questionnaire to ensure that it was ready to be used on a larger scale. Data that were collected on respondent characteristics included age, gender, place of residence, weight, height, screen time, food sources and online food shopping habits. Meanwhile, data on family characteristics included educational attainment, occupation, parents' income, number of family members and family assets.
Adolescents' food preferences were measured using a three-point Likert scale: Dislike, Neither Like nor Dislike, and Like. The data regarding food consumption during the last month were collected using the Food Frequency Questionnaire (FFQ). In previous studies (Abdullah et al., 2015;Van Rongen et al., 2020), when most of the population consumed fast food within 1-3 months or less, the modified frequency scores were grouped into four ordinal categories, those were never consuming these food items, ≤ 1 time per month, 2-3 times per month and ≥ 4 times per month.
Food categorized as fast food is food or drink that has been processed, ready to serve, gives a modern impression (because major international chains sell it) and is commonly found in the research location. Examples of fast food from each food group are sushi (cereal), french fries (tubers), vegetable or fruit salad (vegetables and fruit), burgers and crispy fried chicken (meat or poultry), fish and chips(fish), clay pot tofu (beans), tteokbokki (snack) and non-dairy beverages (boba or bubble drink or grass jelly drink). Meanwhile, the requirements for the type of traditional food selected were popular and unique in each region and may be known by the respondents. Therefore, each questionnaire was designed specifically for each province. As many as 1-4 types of traditional food and fast food were selected from 9 food groups, and therefore, there were 10-20 types of traditional food and 17 types of fast food that varied among regions.

Data analysis
For analysis and discussion, as many as 15 provinces were grouped into two regions as research locations, based on a national development strategy, that included economic growth, poverty and unemployment rates (Bappenas, 2021). Provinces in Java and Bali are grouped into Java-Bali (JB), while other provinces are outside Java-Bali (OJB).
Respondents were defined as liking certain types of food if they liked at least one type of food. The preference score indicates the level of liking, 0 for "not consuming", 1 for "disliking", 2 for "neither liking nor disliking", and 3 for "liking". Preference score data per food group is the average of all food scores in that group. The preference scores in JB and OJB are presented in mean ± standard deviation, obtained from the average value of all food groups from all provinces in the two regions.
Respondents were defined to consume certain types of food if they have consumed it at least once in the past month. The food consumption frequency score was 0 for "not consuming", 1 for "consuming ≤1 time per month", 2 for "consuming 2-3 times per month" and 3 for "consuming ≥4 times per month". The consumption score per food group is the mean of all types of food in the group. Consumption scores in the provinces in Java-Bali and the provinces outside Java-Bali are shown in mean ± standard deviation, obtained from the mean value of all food groups from all provinces in the two regions.
As in most medical research, hypothesis testing employs parametric analysis based on the assumption of "central limit theorem", which states that data distribution is assumed to be normal as the sample size gets larger (Kwak and Kim, 2017). Mean values of both preference and consumption scores from each province were then averaged for the respective region, JB and OJB. Differences in preference scores and consumption scores between regions were tested using an independent t-test. Next, a multiple logistic regression test was conducted to examine the influence of individual FULL PAPER characteristics and family socioeconomic status on traditional and fast food consumption. Table 1 presents the characteristics of respondents based on their place of residence, which were provinces in Java-Bali and outside Java-Bali. The number of respondents living in the JB region was slightly lower (40%) than that of respondents living in OJB (60%). Meanwhile, the respondent's gender was mostly female, with 70% and 67% in JB and OJB, respectively.

Characteristics of respondents
Most of the respondents' nutritional status was normal (55% in JB and 67% in OJB), while 15% were obese. This is in agreement with the result from Indonesia's Basic Health Research 2018 that reported nutritional status in most adolescents aged 13-15 years old and 16-18 years old in Indonesia was normal (Ministry of Health, 2018).
Most respondents (72% and 58%, respectively) in JB and OJB take part in hybrid learning that combines online and offline classes because this study was conducted during the Corona Virus Disease 2019 (COVID-19) pandemic as instructed by the government of Indonesia (Ministry of Education, Culture, Research and Technology of the Republic of Indonesia, 2021; Ministry of Internal Affairs, 2021). The education stages of the respondents, those were junior high school, senior high school, and university, had a proportional distribution in both regions of 33.3%. Most of the respondents allocated more than four hours of screen time per day, but this was more often found in JB (63%) than in OJB (57%). The COVID-19 pandemic has made students spend more time in front of mobile, computer, or laptop screens for academic activities from home (Nagata et al., 2020;Jahja et al., 2021).
Most of the respondents consumed daily food and drinks prepared by their mothers at home (85% and 86%), while the rest bought convenience food outside. The government's recommendation to limit community activities and stay at home during the COVID-19 pandemic has led to changes in family eating habits. The results of the research done by Philippe et al. (2021) showed that more parents prepared and cooked meals at home during the pandemic than before the pandemic. The percentage of respondents in Java and Bali who bought food online more than twice a month was 44% and 41% outside of Java and Bali. Research by Rubby and Briawan (2020) on adolescents in Yogyakarta revealed almost similar results-the frequency of consumption and online food purchases of 1.6 times per week. FULL PAPER Elementary, junior high, or high school graduates are the educational attainment of most of the respondents' parents, both fathers (69% and 70%) and mothers (70% and 69%). In JB and OJB, the percentages of fathers who are university graduates and equivalents were 30% and 29%, respectively. In comparison, the percentage of university graduates or the equivalent mothers was 28% in both regions. The socioeconomic status of the respondent's family was identified based on their parent's income. The income of fathers in the two regions was mostly less than IDR 5 million per month (83% and 84%), while the percentage of fathers with an income of more than five million per month in JB and OJB was not much different; 18% and 16% respectively. Most of the respondents' mothers are homemakers and have an income of less than one million per month, which was 64% and 66% in JB and OJB, respectively. The minimum wage in Jakarta in 2021 is around 4.5 million a month, while the average wage in Indonesia is around 2.6 million a month (Central Bureau of Statistics, 2021).

Characteristics of respondent
In the provinces outside Java-Bali, the number of families with > 4 family members (58%) was higher than the number of families with ≤ 4 family members (42%), while in the JB region, there were fewer families of > 4 family members (48%) compared to families with ≤ 4 family members (52%). Central Bureau of Statistics (2021) reported that the average number of household members in JB is between 3.3-3.6, while in OJB, the number of household members is between 3.6-4.8. One of the welfare indicators of a family is ownership of assets such as a house, farm, motorbike or car (Central Bureau of Statistics, 2011). By combining the indicators of ownership of car, farm, and house, and based on regional distribution, it was known that 56% of respondents in JB and 53% of respondents in OJB did not have a car, far, or their own house.

Food preference and consumption
In the provinces of Java-Bali, most adolescents preferred fast food to traditional food (P < 0.05). This was indicated by the preference score for fast food which was higher than the preference score for traditional food in almost all food groups, except for the group of beans and legumes. Although fast food was more popular, adolescents in Java-Bali more often consumed traditional foods in the cereal, vegetable, seafood, bean and legume groups and traditional snacks. Meanwhile, a small number of other food groups, those tubers, fruits, meat and poultry, and non-dairy beverages, were mostly consumed in the form of fast food ( Table 2). The high consumption of traditional foods in the midst of a prominent preference for fast food indicates that eating behavior is determined by many factors, not only by preference. Meanwhile, in the provinces outside Java-Bali, adolescents preferred traditional food to fast food. In the cereal, vegetable, seafood, bean and legume, and snack groups, the preference score for traditional processed foods was higher than the preference score for fast food. This is in agreement with the frequency of consumption of traditional foods from these food groups, which was also higher than the frequency of consumption of fast foods from the same food groups. However, the groups of tubers, meat and poultry, and non-dairy beverages were more preferred and consumed in the form of fast food than in the form of traditional foods (Table 2).
In the provinces in the Java-Bali region and outside Java-Bali, cereals were mostly consumed in the form of traditional foods, according to their preference and consumption scores (P < 0.05). Traditional cereal-based foods include fried rice, mie kocok (shaken noodles from west java, kocok means "shake", referring to the method of shaking noodles while being simmered in hot water), gomak noodles (noodles from Toba Batak that are prepared by grabbing the noodle with bare hand, gomak means "grab"), nasi lemak (rice with coconut milk), and mie jawa (Javanese noodle, made from yellow noodle, vegetables, egg and spices), while cereal-based fast foods include sushi, onigiri, ramen and ramyeon. Those cereal-based fast food are from Japan and South Korea and are popular among people in Indonesia. Various types of cereals are known as important sources of essential amino acids and are also associated with a reduced risk of coronary heart disease (Mirmiran et al., 2016;Han et al., 2019).
Food choice is influenced by various internal and external factors, including food characteristics (taste and shape), physiological needs, habits, knowledge, personal identity, social environment, culture, economic variables and political elements (Chen and Antonelli, 2020). Traditional food is generally more affordable than fast food, so it is widely chosen by people with low-tomoderate incomes. Aiding the maintenance of traditional food needs to be done continuously thus it will not be displaced by other types of food, especially fast food. Traditional food shows the cultural richness of Indonesia which consists of various ethnic groups that inhabit the islands that stretch from the west to the east end of Indonesia.
In the provinces in Java-Bali and outside Java-Bali, tubers were more preferred and consumed in the form of fast food than traditional food, according to the preference and consumption scores (P < 0.05 Globalization is one aspect that is able to shift preferences and consumption of local or traditional food to foods that are considered to be more modern (Mardatillah et al., 2019).
Fast food, as food sold by international franchises, began its expansion to Indonesia around the 1990s and its expansion to various big cities in Indonesia was very rapid. The distribution of fast food chains in various cities makes it easier for people to buy fast food as an alternative to authentic Indonesian food. Research in India discovered that 22.4% of adolescents who consumed fast food were overweight and 9.52% were obese (Shah et al., 2014). As many as 54.40% of students liked fast food and more than 60% of students were not aware of the fact that fast food is food that is not nutritionally balanced (Purohit et al., 2015).
In the vegetable group, adolescents in Java-Bali preferred vegetables that were processed into fast food, but they still consumed vegetables more often in the form of traditional foods. Meanwhile, outside Java-Bali, vegetables were preferred and consumed in the form of traditional foods (P < 0.005). The vegetable-based traditional foods in question were gado-gado (mix raw/ half cooked vegetables served with peanut sauce and usually served with boiled egg and tofu), karedok gado (mix raw vegetables served with peanut sauce), pecel (mix cooked vegetables served with peanut sauce), mashed cassava leaves, gulai pucuk ubi (cassava leaves with coconut milk), and gudeg (raw jackfruit boiled with palm sugar and coconut milk for hours), while vegetablebased fast food is the vegetable salad with common salad dressing. This is different from the preference and consumption of fruits in the provinces in Java-Bali, which tended to score higher if the fruit was in the form of fast food than if the fruit was in the form of traditional food (P < 0.005), and there was no significant difference in preferences and consumption in the provinces outside Java-Bali. A study conducted by Anggraeni and Sudiarti (2018) has shown that adolescents with fruit and vegetable availability at home have higher fruit and vegetable consumption.
The results of Basic Health Research (2013) revealed that 93.6% of the population aged ≥ 10 years consumed less vegetables and fruit, where the portion of vegetables and fruit consumed was less than five portions per day for seven days a week. Meanwhile, in the Guidelines for The results of another study on adolescent high school students in four major cities in Indonesia, those were-Palembang, Bali, Yogyakarta, and Pontianakshowed that vegetables were the food group with a low level of preference (Hendra et al., 2019). Dishes made with vegetables need to be modified to increase public acceptance. Taste and price are the two aspects that most determine a person's food preferences (Ernawati and Prakoso, 2019).
The preference score for the meat and poultry group among adolescents in the provinces in Java-Bali and outside Java-Bali was higher and the frequency score for the consumption of the two food items in the form of fast food was higher than the consumption in the form of traditional foods (P < 0.05). The meat and poultry-based fast foods include burger, kebab, and fried chicken, while the traditional ones include empal (sweet, dry, and spicy mashed beef), beef jerky, rendang (slow cooked beef with coconut milk and spices), opor (boiled chicken with coconut milk and spices), gulai (spicy chicken/beef with coconut milk), semur (chicken/beef with brown gravy made from soy sauce and spices), soto (indonesian soup), sate kere (skewered tempe/fermented soy bean), and garang asem (chicken cooked with banana leaves and has a sour and spicy taste). Fast food consumption is influenced by the number of fast food restaurants in an area and people's income. Additionally, fast food for some people is also considered a prestigious food (Xiao et al., 2018).
In the Java-Bali provinces, the preference for seafood products in the form of fast food was higher than the preference for products made with similar food items in the form of traditional foods. Moreover, the consumption of fast food made with seafood was more frequent than the consumption of traditional foods made with seafood (P < 0.05). In the provinces outside Java-Bali, traditional foods made with seafood were preferred and consumed more, based on the preference and consumption scores (P < 0.05). Examples of seafoodbased traditional foods are pepes ikan (fish wrapped in banana leaves), sambal teri Medan (Indonesian sauce served with ancovies), fish curry, sour and spicy Hemibagrus stew, kerutup ikan (fish cooked with specific spices, consumed widely in Jambi), mangut lele (catfish cooked with coconut milk), milkfish cake, while seafood-based fast food is fish and chips. Overall, it can be said that seafood processed into traditional foods is consumed more than seafood processed into fast food (fish and chips).
In Java-Bali and outside Java-Bali, beans and legumes processed into traditional foods were preferred and consumed more often than those processed into fast food. Traditional foods made with beans and legumes are fried tempeh and tofu, tahu gejrot (mashed tofu with spices), tahu gunting (chopped tofu), sambal tempeh (Indonesian spicy sauce with tempe/fermented soy bean), mung bean bakpia (pastry filled with mung bean), tahu petis (tofu served with petis/sauce made by shrimp or fish extract), tahu tek (fried tofu with lontong/rice wrapped in banana leaves, and served with peanut sauce and petis), pigeon peas, and spicy hyacinth bean, while fast food made with beans and legumes is clay pot tofu.
Snacks were preferred and consumed more often in the form of traditional snacks than fast food snacks, both in the provinces in Java-Bali and outside Java-Bali (P < 0.05). The types of snacks consumed by adolescents were quite diverse. Fast food snacks included JCo/ Dunkin Donuts, sweet cakes, waffles/pancakes, tempura and tteokbokki. The traditional snacks that were consumed were cilok (ball shaped snack made from tapioca flour), seblak (spicy and hot dish from Cianjur), fritters, ombusombus (cake made from rice flour with palm sugar in the middle and wrapped in banana leaves), lemang (rice with coconut milk wrapped by banana leaves), lompong sagu (snack made from sago flour and coconut, wrapped by banana leaves), dodol (a sweet, chewy, caramel-like, and sticky snack), pempek (indonesian fish cake served with special vinegar sauce, "cuko") and jenang sumsum (rice pudding). According to Intani (2014), the expansion of modern food (fast food) was initially able to marginalize the existence of traditional food. Fast food is simple in terms of ingredients, equipment and processing and presentation methods. However, over time, slowly, some traditional foods began to rise in popularity. This phenomenon is shown by the increasing number of traditional food sellers even in the shopping mall or shopping center areas.
In the provinces in Java-Bali and outside Java-Bali, non-dairy fast food beverages were preferred and consumed more frequently, according to preference scores and consumption scores (P < 0.05). Non-dairy fast food beverages include boba/grass jelly drinks, franchised tea/coffee (Starbucks), and carbonated soft drinks (Cola/Fanta), while traditional types of drinks are bandrek/bansus (warm ginger drink), es doger (shaved ice beverages with coconut milk, fermented sticky rice and cassava, and red colored syrup), es cincau https://doi.org/10.26656/fr.2017.7(4).156 © 2023 The Authors. Published by Rynnye Lyan Resources FULL PAPER (indonesian grass jelly beverages), traditional coffee drinks, passion fruit juice, tamarillo juice, cinnamon syrup, and wedang angsle (sweet and hot dish made from ginger soup, coconut milk, sticky rice, green beans, and slices of bread). There are many types of traditional, non -dairy beverages in the target regions of the study.
It can be said that non-dairy fast food beverages have become part of the lifestyle trend of young people and have shifted the trend of consumption of traditional drinks. Young people need a comfortable place to mingle with their peers while consuming fast food drinks. There were 53.1% of the Indonesian population aged ≥ 10 years consumed sugary drinks ≥ 1 time a day (Basic Health Research, 2013). In fact, consumption of sugary drinks is correlated with the incidence of type 2 diabetes, heart disease, kidney disorders, liver disease, nutritional damage, and gout (arthritis) (Bomback et al., 2010;Malik and Hu, 2015;Malik and Hu 2019;Valenzuela et al., 2021). The Center for Disease Control and Prevention (CDC) recommends consuming sugary drinks, one of which is soft drinks, of no more than 130 kcal because people who consume sugary drinks usually do not reduce their food portions, resulting in excess energy intake (Pan and Hu, 2011).

Factors that influence fast food consumption among adolescents in Indonesia
Logistic regression analysis was conducted on the variables that influenced the consumption of fast food from nine food groups based on the place of residence, socio-economic status, and demographic conditions of the respondents and their parents. The results of the analysis are presented in Table 3.
The place of residence of the respondents (Java-Bali (JB) vs. outside Java-Bali (OJB)) was a significant variable in the level of consumption of fast food from the categories of cereals, tubers, vegetables and fruit and snacks. Meanwhile, the place of residence (urban areas vs rural areas) had a significant influence on the level of consumption of fast food made with cereals, tubers, fruit, meat and beans. Living in Java-Bali and urban areas is convenient for obtaining fast food. This is in agreement with the research conducted by Widaningrum et al. (2020) in Jakarta, which revealed that fast food restaurants are located in the vicinity of entertainment venues, educational, religious, and health facilities, and office buildings and are therefore very easy to access on a daily basis.
Fruit-based fast food that tends to be consumed by respondents living in rural areas indicates its popularity. Moreover, it is expanding faster and consumers could reach it easily. Research conducted by Story et al. (2002) discovered that the consumption of fruits and vegetables among individuals can be influenced by four factors, those were individual factors (knowledge and reasons an individual consumes fruits and vegetables), social environmental factors (family and peers), physical environmental factors, and mass media factors (marketing). Fruit and vegetable consumption was significantly and negatively related to the incidence of chronic heart disease. For an increase in one serving of fruit and vegetable consumption, there is a 4.0% reduction in the risk of developing chronic heart disease (Dauchet et al., 2006). This conclusion is in agreement with the research conducted by Mikkilä et al. (2004), which revealed that the diet of children aged 3-18 years who consume more foods high in saturated fat than fruits and vegetables can increase the risk of cardiovascular disease later in their life. Meanwhile, according to Drapeau et al. (2004), the consumption of fruits and vegetables can prevent obesity because it can reduce hunger and does not cause excess fat intake.
In terms of respondent characteristics, Table 3 shows that gender is a determining variable in the level of consumption of fast food from all food categories, except vegetables and fruit, among adolescents in Indonesia. The tendency of female respondents to consume fast food indicates their preference for modern food. Eating habits are influenced by several environmental factors such as the cultural environment, the natural environment and the population. Adolescents usually have their own choice of favorite foods because by that time, eating habits have already been formed (Fuster et al., 2019). Boys tend to like filling food, while girls tend to like snacks (Monge-Rojas et al., 2015;Blum et al., 2019).
The educational background of the respondents was an important variable on the level of consumption of fast food from the categories of vegetables, seafood, beans and snacks. The habit of adolescent respondents in shopping online for food was also a variable that had a significant influence on the consumption of fast food in Indonesia from all food categories. Meanwhile, daily screen time only influenced the consumption of fast food made from vegetables, fruit, meat, seafood and beans. Learning methods (online or offline learning) during the current COVID-19 pandemic situation influenced the consumption of fast food from the fruit category.
Lifestyle, especially consumption patterns, has undergone many changes due to the development of internet technology. Various services can be accessed online, one of which is food purchasing service (Grunert and Ramus, 2005). Online food delivery applications are currently widely used by various groups of people, https://doi.org/10.26656/fr.2017.7(4).156 © 2023 The Authors. Published by Rynnye Lyan Resources FULL PAPER including adolescents. Online food delivery applications provide satisfaction to adolescents, especially in terms of efficiency, privacy, desire fulfillment and food variety (Suhartanto et al., 2019). They provide various types of food from food stalls or restaurants across a wide area. The foods offered are varied, ranging from main dishes, side dishes, vegetables, fruits, snacks, drinks, and all can be purchased easily. A study in Bangladesh discovered that factors related to students' preferences for fast food consumption include service effectiveness, easy access, food taste, price and fast service (Goon et al., 2014). Fast food consumption habits are based on the fast food industry's ability to meet consumer needs, which include easy access, easy ordering process, and fast and convenient service (Untaru and Ispas, 2013).
Regarding the socio-economic status of the respondents' families, Table 3 shows that the educational background of the respondents' fathers had a significant influence on the consumption of fast food from the cereal and tuber category among adolescents, while the educational background of their mothers had a significant influence on the consumption of fast food made with fruits and beans. Father's occupation was a factor that had a significant influence on the consumption of fast food from all food categories, except vegetables. Meanwhile, the mother's occupation only had a significant influence on the consumption of meat-based fast food and snacks. Parental income was also a variable that had a major influence on the consumption of fast food from all food categories, except for the fruit group and bean group, among adolescents in Indonesia. Meanwhile, the ownership of family assets only had a significant influence on the consumption of fast food in  The factors that influence the consumption of fast food include the socio-economic and socio-demographic characteristics of the family. Consumption of fast food has rapidly become common in the last ten years especially among children and adolescents (Das, 2015). The results of several studies reveal that consumption of fast food is higher among children, adolescents, young adults and those with high incomes (Pereira et al., 2005). According to Ariani et al. (2018), the income factor affects the level of food consumption. The higher the income level of a household, the greater the level of animal protein consumption per capita becomes. Children from families with higher socio-economic status tend to prefer fast food to traditional food. Fried chicken is often chosen by respondents in various studies as the type of fast food that is most often consumed and preferred because it is considered to have a good taste and therefore respondents declared their intention to eat it again (Ishak et al., 2020). Fried chicken is made with broiler chicken, which is processed by frying. This kind of fast food that is cooked by frying will absorb more oil than fast food that is cooked using other cooking methods, such as baking. In addition, the addition of wheat flour and spices add a more savory taste and a crunchy texture to fried chicken. Although the nutritional content tends to be damaged because of the frying process, fried food tastes more savory, contains higher calories, and often tastes better than foods that are cooked by other methods, for example boiling.
Apart from animal protein, it can be observed that family income has a significant influence on the level of consumption of fast food from the non-dairy beverages category. Consuming soft drinks regularly and excessively can have a bad impact on the body. Consumption of sweetened soft drinks is associated with the increasing incidence of obesity, diabetes mellitus and dyslipidemia (Bahadoran et al., 2013;Marlatt et al., 2016). Soft drinks are generally consumed together with fast food. In fact, according to the results of research by Bowman and Vinyard (2004), people who consume fast food and soft drinks tend to have a higher BMI. The consumption of high-calorie drinks is associated with an increase in energy intake (Woodward-Lopez et al., 2011).

Factors influencing the consumption of traditional foods among adolescents in Indonesia
Logistic regression analysis was conducted on the variables that influence the consumption of traditional foods from nine food groups based on the place of residence, socio-economic status, and demographic conditions of the respondents and their parents. The results of the analysis are presented in Table 4 below.
Living in the provinces of Java-Bali significantly influenced the consumption of traditional foods from the categories of cereals, tubers, vegetables, meat and snacks among adolescents in Indonesia. Meanwhile, living in urban areas had a wider influence on almost all traditional food categories, except seafood, snacks and non-dairy beverages. A study in the Special Region of Yogyakarta showed that traditional foods that are still widely consumed by people are usually vegetable-based and consumed as a main course (Harsana and Triwidayati, 2020). Another study discovered that adolescents who live in rural areas consume more fruits and vegetables because of their sufficient availability (Anggraeni and Sudiarti, 2018).
In terms of respondent characteristics, gender was a variable that had a significant influence on the consumption of traditional foods from all food groups, except for vegetable-based traditional foods, among adolescents in Indonesia. Meanwhile, the educational background of the respondents had a significant influence on the consumption of traditional foods from the cereal, vegetable, seafood, bean and snack groups. Furthermore, online shopping habits significantly influenced the consumption of traditional foods from all food groups. In contrast, the learning method (online or offline learning) only had a significant influence on the consumption of traditional foods from the fruit group. In addition, the total time spent per day in front of the screen (screen time) had a significant influence on the consumption of traditional foods from the vegetable, fruit, meat and seafood groups.
In terms of gender, a study showed that tubers were mostly consumed as snacks and it was more consumed in urban areas by female respondents (Rachmawati and Sari, 2020). In terms of the respondent's educational background, among junior high and high school students, meal planning is more often done by parents because they are not like university students who do not live with their parents and prefer to buy ready-to-eat food (Blum et al., 2019). Parents of respondents who have sufficient income will support their purchase of vegetables. Apart from that, living in rural areas gives the advantage of sufficient availability of vegetables (Anggraeni and Sudiarti, 2018). Peer pressure was also a factor that influenced the level of consumption of traditional foods among adolescents. A study conducted on junior high school students in Bandung showed that the low consumption of traditional drinks (for example es goyobod, es cendol, es cingcau, sekoteng) was due to the influence of peers (Yulia et al., 2017). FULL PAPER educational background of the respondents' fathers only influenced the level of consumption of traditional foods from the cereal and tuber groups, while the educational background of their mothers had a significant influence on the level of consumption of traditional foods made with vegetables, fruit, seafood and beans. In addition, the father's occupation influenced the level of consumption of traditional foods from all food categories, while the mother's occupation only influenced the level of consumption of traditional foods made with tubers and meat as well as traditional snacks. Interestingly, the mother's income significantly influenced the level of consumption of traditional foods from all food categories among adolescents, while the father's income only influenced the consumption of traditional foods from the groups of cereal, tuber, meat, seafood, snack and nondairy beverages. The aspect of asset ownership did not have a strong influence and its influence was only on the level of consumption of traditional foods from the tuber group and the snack group. A study on adolescent high school students in the cities of Indonesia (Palembang, Yogyakarta, Bali and Pontianak) showed that in these cities, chicken meat was the main type of food group favored by students (Hendra et al., 2019). The important role of family income also appears in the seafood-based food group in which the research of Fuada et al. (2018) showed that economic status, affordability, and prices of fish are influential factors in fish consumption in Indonesia. The same thing was also discovered in another study regarding snacks, where good parental income allows respondents to try various kinds of traditional foods in the form of snacks thus respondents are more interested in consuming them (Bastami et al., 2019).

Conclusion
Traditional foods, especially those made with cereals, vegetables, seafood, beans, legumes and traditional snacks are more preferred and consumed than fast food in the provinces outside Java-Bali. In the provinces of Java-Bali, fast food is preferred over traditional food. However, some food groups such as cereals, vegetables, seafood, and snacks are mostly consumed as traditional foods.
Consumption of fast food and traditional food among adolescents is influenced by several socio-economic and demographic factors. Female and junior high school respondents are more likely to consume fast food. The respondent's place of residence (Java-Bali and urban areas) greatly influences the respondents' consumption of fast food. Consumption of fast food, in general, is also higher if the father works and has an income of more than or equal to IDR 5 million per month. Fast food consumption will be higher if respondents have a habit of buying food online. Meanwhile, in terms of traditional food, female junior high school respondents who live in the provinces of Java-Bali consume traditional foods more often. The residence of the respondents has an influence on their food consumption habits. Higher consumption of traditional foods was found in respondents whose parents had jobs with an income of more than IDR 5 million per month and who had the habit of shopping online.