Artigos

Introduction
Global lifestyle patterns have been changing, including health behaviors, dietary patterns, and habits related to physical activity and sedentary behavior. This can be observed in the adult population and, worryingly, already in childhood and adolescence. Highly caloric and low-nutrient foods are increasingly present in population’s consumption routine1. In Brazil, one of the milestones in the literature pointing to this change in the food consumption scenario is the Household Budget Survey (POF) 2008–2009, which showed an increase in processed foods consumption and a decrease in fresh foods consumption by the population from the age of 10 years onwards2. Such a change in dietary patterns may be associated with obesity, as well as other noncommunicable chronic diseases (NCDs)3,4.
In addition to diet, other lifestyle-related habits such as reduced physical activity, increased sedentary behavior, longer screen time, decreased sleep time and quality, and alcohol and tobacco consumption habits are associated with an increased risk for obesity, diabetes, and cardiovascular diseases1,5–7. The region of residence is also an important factor, as each place has different forms of social, economic and cultural organization that determine different environments and therefore different lifestyles8. Besides adult populations, young people from urban and rural areas have also been studied regarding their behaviors and eating habits9, as well as their daily activities10,11. It has been observed that adolescents from rural areas in Brazil have a higher frequency of physical activity; however, they consume fewer vegetables and fruits compared to those from urban areas9. However, when we began to review the literature directly evaluating this population, we found it limited and inconsistent.
Sedentary behavior has been associated with an increase in waist circumference and body mass index, high blood pressure, and high cholesterol levels12. It can also be linked to increased screen time and reduced sleep time5,13. Smoking and alcohol consumption, especially among younger populations, are habits that may start in adolescence and persist into adulthood7.
Despite clear differences in lifestyles influenced by cultural and socioeconomic peculiarities of each area, the literature suggests a convergence in dietary habits changes in rural and urban areas. For example, the consumption of ultra-processed foods has increased in rural areas, which was previously observed mainly in urban areas9. Considering the importance of assessing behaviors and habits of younger populations (since these habits often persist into adulthood) and the lack of data in the literature on health behaviors and lifestyle habits of adolescents in rural areas, this study aims to describe and compare dietary and behavioral patterns of adolescents from urban and rural areas.

Methods
This research is part of the Study of Cardiovascular Risks in Adolescents (ERICA), a Brazilian nationwide, multicenter, school-based cross-sectional study that aims to estimate the prevalence of metabolic syndrome and its components in Brazilian adolescents aged 12 to 17 years, attending public and private schools. Data were collected from February 2013 to November 2014. Further details about the ERICA sampling method have been described in previous publications14 . In summary, the eligible population for ERICA was divided into 32 geographical strata, composed of 27 capitals and 5 strata with other eligible cities in each region of Brazil. Public and private schools participating were selected by proportional probability to size, based on the ratio of the number of eligible students and the distance between the school and the state capital. In total, 1247 schools participated in the study across 273 municipalities with more than 100,000 inhabitants in Brazil.

The inclusion criteria for this study were students from rural areas who were between 12 and 17 years old and of both sexes. Urban students who were matched by sex, age, and region of residence were also included if they participated in ERICA. Students from rural areas who could not be matched with urban students, as well as those with incomplete data, pregnant participants, and those unable to complete the anthropometric assessment, were excluded from the sample. The final sample size was 2,488, selected via simple random sampling with 1:1 matching. Of these, 1,244 were from rural areas and 1,244 were from urban areas.
Further details are provided in figure 1.
Sociodemographic data were obtained using a self-administered questionnaire on an electronic data collection device (Personal Digital Assistant – PDA), containing 105 questions about sociodemographic profile. It inquired about gender, age (in complete years), self-declared skin color; labor and employment situation socioeconomic status was defined using the Brazilian Economic Classification Criteria, which considers ownership of assets, employment of a domestic worker, and education level of the head of the household.
Screen time was assessed using questions from the self-administered questionnaire. It was classified as ?2 hours per day or >2 hours per day15. Sleep time was also assessed through the self-administered questionnaire, which enabled the determination of sleep time on weekdays and weekends, categorized as <8 hours per day or ?8 hours per day16. Sleep hours were calculated by the difference between bedtime and wake-up time, assigning a weight to weekdays and weekends, of five and two times, respectively.
The use of tobacco and alcohol was also assessed with questions, which evaluated consumption within the past 30 days. Classified as "yes" for use if any reported at least one day of use, and classified as "no" if response was "never smoked cigarettes", "no days," or "never consumed alcoholic beverages", "no days”.
Dietary intake was assessed by a 24-hour dietary recall, using the multiple-pass method17. Adolescents were interviewed at school by trained researchers who used a specific software for inputting dietary consumption data directly into netbooks using ERICA-REC24h, which was created specifically for ERICA18. The ERICA-REC24h software contains a list of 1626 food items using the database. The software contained a list of foods from the food and beverage acquisition database of the 2008–2009 Household Budget Survey (POF)2. Photographs included in the software were used to help the adolescents estimate the size of portions consumed. The preparation and quantity of food consumed were also recorded in detail. The interviewers could add foods that were mentioned by students but were not in the database.
To assess the dietary quality of adolescents, the Diet Quality Index for Brazilian Adolescents (DQIA-BR) was used. The tool is based on three indicators of healthy eating according to WHO guidelines: dietary quality, which assesses food choices; dietary diversity, which evaluates the variety of food groups in the usual diet; and dietary equilibrium, which measures the balance between the consumption of adequate and inadequate foods. Of all food groups, eight were recommended groups: 1) bread, potatoes, and grains; 2) vegetables; 3) fruits; 4) milk products; 5) cheese; 6) meat, fish, and eggs; 7) beans; and 8) fats and oils. Two were non-recommended food groups: 9) snacks and candies and 10) sugar-sweetened beverages, fruit juices, and alcoholic beverages. The DQIA-BR score ranges from ?33% to 100%. The higher the score, the better the dietary quality. More detailed information about the tool can be found in Ronca et al., 202019,20.
Anthropometric measurements were taken by trained researchers. Waist circumference (WC) was measured at the midpoint between the lower rib margin and the top of the iliac crest using a 1.5-meter anthropometric tape with a precision of 1mm. Its classification was performed according to the cutoff points described by the International Diabetes Federation (IDF) 200721 for ages 12 to 16 years, considering increased WC for those equal to or greater than the 90th percentile. For those aged 16 to 17 years, the cutoff point for adults was used: ?80 cm for females and ?94 cm for males.
Body weight was measured using a digital scale with a capacity of 200kg and a precision of 50g, and height was measured with a portable stadiometer. Body mass index (BMI) was calculated by dividing weight (kg) by the square of height (m²). For BMI classification, age- and sex-specific curves from the World Health Organization (WHO) were used22. The cutoff points were: underweight BMI Z-score < ?1; normal weight BMI Z-score ? ?1 and ? 1; overweight BMI Z-score > 1.
To determine the level of physical activity, an adapted version of the Self-Administered Physical Activity Checklist was used23. The product of time and frequency was calculated for each activity. Then, the total was obtained by adding these values together. Adolescents who did not accumulate at least 300 minutes/week of physical activity were classified as inactive during leisure time activities24.
The statistical analysis was performed using the Statistical Analysis System (SAS) software version 9.4. Proportions were calculated to describe the sample and are presented as absolute numbers and percentages. The Mann-Whitney test was used to compare the DQIA-BR scores and their subcomponents between areas due to the non-parametric distribution of the data. Pearson’s chi-square test was applied to compare behavioral variables and nutritional status between areas.
The adjusted models were built using a backward selection approach, with variables removed when p > 0.20. The covariates included for adjustment were sex, age, skin color, socioeconomic status, and maternal education. Binary or multinomial logistic regression models were used for categorical variables, whereas quantile regression models were applied to dietary variables described by their median. This method estimates the median of outcome variables, rather than the mean, across the distribution, eliminating the need to categorize participants. A significance level of 5% was adopted.
All students included in this analysis signed an informed consent form, and their guardians signed an assent form. ERICA was approved by the Research Ethics Committee of the Institute of Health, Federal University of Rio de Janeiro (CAAE: 05185212.2.1001.5286 - UFRJ coordinating center), and by the Ethics Committees of each of the 27 federative units.

Results
Table 1 presents the characteristics of the study sample. Most of the participants in the sample were female (54.8%), aged from 12 to 14 years (55%), of mixed race (62%), and residing in the Northeast (48.8%) and North (33.8%) regions. The rural population had a higher percentage of mothers with incomplete primary education (26.3%), whereas in most of the urban population mothers had completed high school (25.5%). The rural population showed a higher number of working adolescents compared to the urban population (22.3% vs. 12%). Anthropometric characteristics did not differ among rural and urban adolescents; approximately 92% of the sample had values considered adequate. Both rural and urban adolescents were mostly classified within normal weight as evaluated by their BMI.
Table 2 presents data on the DQIA-BR. Both rural and urban adolescents had low DQIA-BR scores. In adjusted analysis, individuals from rural areas showed a higher consumption of legumes (? = 4.73; 95% CI: 1.24–8.23) and a lower consumption of sugar-sweetened beverages compared to the urban population (525.2ml vs. 536.2, p=0.0158; ? = -13.3, 95% CI:-23.9; -2.62).
Table 3 presents data on the assessed behaviors. The rural population was mostly considered active, standing out in terms of physical activity level, meaning they engage in more than 300 minutes of activity per week. Regarding screen time, the rural population also showed better results, as most participants (59%) spend ?2 hours per day in front of screens. Additionally, they had longer sleep duration, with 76.5% getting ?8 hours of sleep per day. The adjusted analyses showed no material differences after accounting for potential confounders.

Discussion
This study described dietary patterns, behaviors, and lifestyles of adolescents in urban and rural areas, observing those that could contribute to a higher risk of cardiovascular diseases in a representative sample of Brazilian individuals. The differences found in lifestyle between urban and rural adolescents were better dietary quality, higher levels of physical activity, less screen time, and more sleep among those living in rural areas.
A difference was observed in the level of education of the mothers of each individual, with those in rural areas tending to have lower levels of education. This finding aligns with data from the Institute of Applied Economic Research (IPEA) in 2019, which shows lower expected years of education in the rural population25. This data is possibly explained by the disparity in service provision across the national territory, reflected in both urban and rural regions, where rural schools often present poorer facilities, lack of materials and equipment, accessibility challenges, and consequently higher dropout rates compared to urban areas25,26.
There was also a difference in economic class: the rural population was more likely to be in the lower classes than the urban population. This data is in line with what has been demonstrated in national research, where populations in rural areas live with lower incomes than individuals residing in urban areas27. One of the factors that may contribute to this reality is the distance that these populations have from access to information, and consequently from access to the benefits of public policies, as well as difficulties in accessing education26.
Adolescents from rural areas demonstrated a higher consumption of legumes and a lower consumption of sugar-sweetened beverages, with no differences observed in other variables. Souza et al., when comparing dietary habits of adolescents from urban and rural areas in the state of Pernambuco in 2019, did not observe differences in dietary intake between the two groups, with both showing low consumption of fruits and vegetables9. The dietary quality of the young population has become increasingly inadequate, with an increase in the consumption of high-calorie foods low in fiber and various micronutrients, which does not bode well. This fact may be related to the nutritional transition that Brazil has been undergoing in recent decades. One of the factors influencing this transition is globalization, in addition to increased participation of women in family income, reducing the time they previously devoted to family meals28.
Adolescents from rural areas were also more active than those from urban areas. Regis et al., in their work evaluating urban vs rural lifestyle of adolescents aged 14 to 19 years in the state of Pernambuco, also reported a higher levels of physical activity among rural adolescents, along with longer use of electronic devices by students from urban areas10. These two behaviors had been observed together in other studies, as physical activity declines and electronic media use increases among young individuals5, which also agrees with the findings of this study, where the population considered inactive, the adolescents from urban areas were also the ones with the longest screen time.
Adolescents from rural areas, in addition to having a higher frequency of physical activity, also had a higher frequency of individuals with some type of work during adolescence in their sociodemographic characteristics. Leon. E.B et al., in their study aimed to associate work, sociodemographic factors, health behaviors, and cardiovascular risks in Brazilian adolescents. They observed an association between work and physical activity, indicating that more active students, more often work11.
When assessing anthropometric variables, the groups did not show differences between them; however, other authors reported a higher probability of children from rural areas being overweight29. On the other hand, a systematic review including studies from around the world showed a greater tendency for children from urban areas to be overweight compared to those from rural areas, also observing a relationship between individuals, weight and socioeconomic development levels, in which less developed regions have fewer overweight children30. The discrepancy in the literature may be explained by factors related to the level of socioeconomic development of each region, as well as the lack of standardization in the definition of rural characteristics, resulting in different samples.
Overall, this study observed that adolescents living in rural areas exhibited some healthier behaviors compared to their urban counterparts, in agreement with previous findings in the literature that have also reported such differences9,10. With regard to dietary habits, both groups showed similar patterns, characterized by a low intake of fruits and vegetables. However, the rural population reported a lower consumption of sugar-sweetened beverages and a higher consumption of legumes. This difference in consumption is related to socioeconomic, cultural, and environmental factors that shape the context in which these adolescents live. In rural areas, barriers such as distance can limit access to establishments that sell ultra-processed foods, thereby reducing their consumption. In this context, the higher intake of foods such as legumes reflects the persistence of more traditional eating habits, characterized by a greater reliance on home-prepared meals and less influence from the modern food environment31. Furthermore, sleep duration, screen time, and the lower prevalence of sedentary behavior may be associated with the fact that rural adolescents are more frequently engaged in work outside the home, both due to less favorable economic conditions and the more pressing need to contribute to household income11.
Limitations of the study include that the sample includes rural populations from cities with populations of at least 100,000, which may not represent rural populations from smaller cities. It is important to note that individuals living in rural areas of densely populated municipalities differ from those living in rural areas of smaller towns. This difference limits the generalizability of the findings to the overall rural adolescent population, due to the sample's non-representative nature. However, this analysis is one of the strengths of the present study because the rural population of larger cities is underrepresented in the literature. Another limitation is that the data were collected from 2013 to 2014, a period when technology use and access were different from what is currently seen in 2023, representing a snapshot of reality from the time of data collection. However, for the purpose of comparison between the two areas, which is the main objective of our study, it was important that both data collections were conducted simultaneously. Most of the data collected were self-reported by individuals, which is also important to consider as it may be more susceptible to participants' memory biases and perceptions of reality, potentially leading to under- or overestimation of certain responses. Another limitation is that dietary intake was assessed only once through a single 24-hour dietary recall, which may not adequately represent participants’ habitual dietary patterns due to intra-individual variability in daily food consumption. Conducting repeated recalls would allow for a more accurate estimation of habitual dietary intake. However, a strong point is that ERICA was pioneering in addressing important aspects of cardiovascular risk habits among younger Brazilian populations with a robust and well-structured methodology.
There is still a significant gap in the literature regarding the lifestyles of rural adolescent populations and comparisons between rural and urban areas. These analyses are important to have a good understanding of their habits and lifestyles, in order to better guide public policies aimed at improving the living conditions of these populations, from the socioeconomic level to the adoption of healthier habits.
The findings of this study indicate sociodemographic and lifestyle differences between adolescents in urban and rural areas of Brazil. However, when it comes to dietary habits, both groups displayed similar patterns, characterized by low fruit and vegetable intake and a high consumption of calorie-dense, nutrient-poor foods, which may suggest an increased risk of cardiovascular diseases. However, more studies are needed to better understand the convergence in the behaviors of young populations in rural and urban areas of Brazil.

Declaração de Disponibilidade de Dados
Os dados de pesquisa estão disponíveis mediante solicitação ao autor de correspondência.

References
1. Pan-Americana Da Saúde O. Risk Factors for Noncommunicable Chronic Diseases in the Americas: Considerations on Strengthening Regulatory Capacity [Internet]. 2016. Disponível em: www.paho.org/publications/copyright-forms
2. IBGE. Family Budget Survey: Analysis of Personal Food Consumption in Brazil. 2011.
3. Wang Y, Lobstein T. Worldwide trends in childhood overweight and obesity. International Journal of Pediatric Obesity. 2006;1(1):11–25.
4. Ritter J do A, Cureau FV, Ronca DB, Blume CA, Teló GH, Camey SA, et al. Association between diet quality index and cardiometabolic risk factors in adolescents: Study of Cardiovascular Risks in Adolescents (ERICA). Nutrition. 2021;90.
5. Souza Neto JM de, Costa FF da, Barbosa AO, Prazeres Filho A, Santos EVOD, Farias Júnior JC de. Physical activity, screen time, nutritional status and sleep in adolescents in northeast Brazil. Revista paulista de pediatria?: orgao oficial da Sociedade de Pediatria de Sao Paulo. 2021;39:e2019138.
6. Figueiredo VC, Szklo AS, Costa LC, Kuschnir MCC, Da Silva TLN, Bloch KV, et al. ERICA: Smoking prevalence in Brazilian adolescents. Revista de Saude Publica. 2016;50:1s–10s.
7. Coutinho ESF, França-Santos D, Da Silva Magliano E, Bloch KV, Barufaldi LA, De Freitas Cunha C, et al. ERICA: Patterns of alcohol consumption in Brazilian adolescents. Revista de Saude Publica. 2016;50:1s–9s.
8. Rodrigues J. Rural and urban in Brazil: a methodological proposal for the classification of cities. 2014.
9. Souza ETC, Borba JMC, Silva MSP da, Melo APR de. Teenager’s eating habits from urban and rural areas of a county of Pernambuco. Research, Society and Development. 27 de março de 2023;12(4):e7212440969.
10. Regis MF, Oliveira LMFT de, Santos ARMD, Leonidio A da CR, Diniz PRB, Freitas CMSM de. Urban versus rural lifestyle in adolescents: associations between environment, physical activity levels and sedentary behavior. Einstein (Sao Paulo, Brazil). 1o de outubro de 2016;14(4):461–7.
11. de Leon EB, Tavares BM, Fernandes TG, Gonçalves RL, Franco FFS, de Souza MBCA, et al. Study of cardiovascular risks in adolescents (Erica): Factors associated with work in adolescence. Ciencia e Saude Coletiva. 2021;26(7):2601–12.
12. Carter S, Hartman Y, Holder S, Thijssen DH, Hopkins ND. Sedentary behavior and cardiovascular disease risk: Mediating mechanisms. Exercise and Sport Sciences Reviews. 2017;45(2):80–6.
13. Hale L, Guan S. Screen time and sleep among school-aged children and adolescents: A systematic literature review. Sleep Medicine Reviews. 1o de junho de 2015;21:50–8.
14. de Vasconcellos MTL, do Silva PLN, Szklo M, Kuschnir MCC, Klein CH, De Abreu GA, et al. Sampling design for the Study of Cardiovascular Risks in Adolescents (ERICA). Cadernos de Saude Publica. 2015;31(5):921–30.
15. COUNCIL ON COMMUNICATIONS AND MEDIA, Strasburger VC, Hogan MJ, Mulligan DA, Ameenuddin N, Christakis DA, et al. Children, Adolescents, and the Media. Pediatrics. 1o de novembro de 2013;132(5):958–61.
16. Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, et al. National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health. março de 2015;1(1):40–3.
17. Conway JM, Ingwersen LA, Vinyard BT, Moshfegh AJ. Effectiveness of the US Department of Agriculture 5-step multiple-pass method in assessing food intake in obese and nonobese women. Am J Clin Nutr. maio de 2003;77(5):1171–8.
18. Barufaldi LA, Abreu G de A, Veiga GV da, Sichieri R, Kuschnir MCC, Cunha DB, et al. Software to record 24-hour food recall: application in the Study of Cardiovascular Risks in Adolescents. Rev Bras Epidemiol. 2016;19(2):464–8.
19. Ronca DB, Blume CA, Cureau FV, Camey SA, Leotti VB, Drehmer M, et al. Diet quality index for Brazilian adolescents: the ERICA study. European Journal of Nutrition. 2020;59(2):539–56.
20. Word Health Organization. CINDI dietary guide. 2000.
21. International Diabetes Federation. The IDF consensus definition of the Metabolic Syndrome in children and adolescents. 2007.
22. Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bulletin of the World Health Organization. setembro de 2007;85(9):660–7.
23. Farias Júnior J, Lopes A, Mota J, Santos M, Ribeiro J, Hallal P. Physical Activity Checklist Validity and reproducibility of a physical activity questionnaire for adolescents: adapting the Self-Administered Physical Activity Checklist. Rev Bras Epidemiol. 2012;15(1):198–210.
24. Cureau FV, Da Silva TLN, Bloch KV, Fujimori E, Belfort DR, De Carvalho KMB, et al. ERICA: Leisure-time physical inactivity in Brazilian adolescents. Revista de Saude Publica. 2016;50:1s–11s.
25. Pereira CN, Nunes De Castro C. Education: contrast between urban and rural environments in Brazil [Internet]. 2019. Disponível em: http://atlasbrasil.org.br/2013/pt/o_atlas/idhm/
26. Pereira CN, Castro CN de. Education in Rural Areas: differences between rural and urban areas. Texto para Discussão. 1o de março de 2021;1–64.
27. IBGE. Brazilian census 2010. 2011.
28. Fox A, Feng W, Asal V. What is driving global obesity trends? Globalization or “modernization”? Global Health. dezembro de 2019;15(1):32.
29. Liu J, Bennett KJ, Harun N, Probst JC. Urban-Rural Differences in Overweight Status and Physical Inactivity Among US Children Aged 10-17 Years. National Rural Health Association. 2008;
30. Paciorek CJ, Stevens GA, Finucane MM, Ezzati M. Children’s height and weight in rural and urban populations in low-income and middle-income countries: a systematic analysis of population-representative data. www.thelancet.com/lancetgh [Internet]. 2013;1. Disponível em: www.thelancet.com/lancetgh
31. Silveira VNC, Dos Santos AM, França AKTC. Determinants of the consumption of ultra-processed foods in the Brazilian population. Br J Nutr. 28 de outubro de 2024;132(8):1104–9.