Abstract

In this study, we evaluated socioeconomic inequalities in the consumption of in natura/minimally processed and ultra-processed foods among adolescents. We used data from the Brazilian National Survey of School Health (PeNSE), 2015. According to the self-reported consumption of beans, vegetables and fruits, a score of in natura/minimally processed foods was generated (0-21 points). Sodas, sweets, instant noodles, and ultra-processed meat were used for the score of ultra-processed foods (0-21 points). Equality indicators were gender, maternal education, and socioeconomic level. Absolute difference, ratios, concentration index and slope index of inequality were calculated. Adolescents (n=101,689, 51% girls, 14.2 years) reported a mean score of 9.97 and 11.46 for ultra-processed foods and in natura/minimally processed foods, respectively. Absolute and relative differences between adolescents with the highest and lowest socioeconomic level, there were differences of 2.64 points and 33% for consumption of in natura/minimally processed foods; and 1.48 points and 15% for ultra-processed foods. Adolescents from higher socioeconomic level ate more in natura/minimally processed foods and ultra-processed foods.

Key words:
Healthy eating; Adolescent; Food consumption; Fast foods

Resumo

Nesse estudo, avaliamos as desigualdades socioeconômicas no consumo de alimentos in natura/minimamente processados e ultraprocessados entre adolescentes. Foram utilizados dados da Pesquisa Nacional de Saúde do Escolar (PeNSE), 2015. De acordo com o consumo autorrelatado de feijão, hortaliças e frutas, foi gerado um escore de alimentos in natura/minimamente processados (0-21 pontos). Refrigerantes, doces, macarrão instantâneo e carnes ultraprocessadas prontos para o consumo foram utilizados para a pontuação dos alimentos ultraprocessados (0-21 pontos). Os indicadores de equidade foram gênero, educação materna e nível socioeconômico. Foram calculados a diferença absoluta, razões, índice de concentração e índice de inclinação de desigualdade. Os adolescentes (n=101.689, 51% meninas, 14,2 anos) relataram escore médio de 9,97 e 11,46 para alimentos ultraprocessados e in natura/minimamente processados, respectivamente. As diferenças absolutas entre os adolescentes de alto e baixo nível socioeconômico foram mais altos e mais baixos, houve diferenças de 2,64 pontos e 33% para o consumo de alimentos in natura/minimamente processados; e 1,48 pontos e 15% para alimentos ultraprocessados. Adolescentes de níveis socioeconômicos mais elevados comeram mais alimentos in natura/minimamente processados e alimentos ultraprocessados comparado aos seus pares.

Palavras-chave:
Alimentação saudável; Adolescente; Consumo de alimentos; Fast Foods

Introduction

Nutrition is crucial for the healthy development of adolescents¹1 World Health Organization (WHO). Diet, nutrition and the prevention of chronic diseases: report of a joint WHO/FAO expert consultation. Geneva: WHO/FAO; 2002.^,22 Brasil. Ministério da Saúde (MS). Guia alimentar para a população brasileira. Brasília: MS; 2014.. The Dietary Guidelines for the Brazilian Population suggests that Brazilians’ diet should be based on in natura or minimally-processed foods, which is nutritious, available in great variety, and produced by environmentally sustainable foods systems²2 Brasil. Ministério da Saúde (MS). Guia alimentar para a população brasileira. Brasília: MS; 2014.. Considering the classification of foods according to the degree and purpose of processing, just over half of the calories consumed by the Brazilian population came from in natura or minimally-processed foods. Among these, rice, beef, beans and poultry meat were the most frequently consumed³3 Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa de orçamentos familiares: 2017-2018: análise do consumo alimentar pessoal no Brasil. Rio de Janeiro: IBGE; 2020.. On the other hand, among the ultra-processed foods with the highest frequency of consumption were margarine, crackers, packaged snacks and breads³3 Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa de orçamentos familiares: 2017-2018: análise do consumo alimentar pessoal no Brasil. Rio de Janeiro: IBGE; 2020..

The higher participation of ultra-processed foods in the diet is associated with a lower quality diet. In adolescents, the intake of these foods has increased over the years and has been related to sociodemographic factors⁴4 Baraldi LG, Martinez Steele E, Canella DS, Monteiro CA. Consumption of ultra-processed foods and associated sociodemographic factors in the USA between 2007 and 2012: evidence from a nationally representative cross-sectional study. BMJ Open 2018; 8(3):e020574., such as sex, maternal education and socioeconomic level. In addition, consumption of ultra-processed foods has been associated with increased risk of overweight and obesity, cardiovascular diseases, and cancer⁵5 Chen X, Zhang Z, Yang H, Qiu P, Wang H, Wang F, Zhao Q, Fang J, Nie J. Consumption of ultra-processed foods and health outcomes: a systematic review of epidemiological studies. Nutr J 2020; 19(86):1-10..

Identifying possible health inequalities among subgroups⁶6 World Health Organization (WHO). Handbook on health inequality monitoring: with a special focus on low- and middle-income countries. Geneva: WHO; 2013., such as sex and socioeconomic level, is important to plan and implement policies supporting the adoption and maintenance of healthy diets. Some studies have indicated better eating habits in adolescent girls⁷7 Nilsen SM, Krokstad S, Holmen TL, Westin S. Adolescents' health-related dietary patterns by parental socio-economic position, the Nord-Trøndelag Health Study (HUNT). Eur J Public Health 2010; 20(3):299-305. and those with higher level of maternal educational⁸8 van Ansem WJ, Schrijvers CT, Rodenburg G, van de Mheen D. Maternal educational level and children's healthy eating behaviour: role of the home food environment (cross-sectional results from the INPACT study). Int J Behav Nutr Phys Act 2014; 11:113., possibly due better accessibility to heathy foods options⁹9 Fahlman MM, McCaughtry N, Martin J, Shen B. Racial and socioeconomic disparities in nutrition behaviors: targeted interventions needed. J Nutr Educ Behav 2010; 42(1):10-16.. Although these findings highlight inequities in fruit and vegetable consumption, they describe the social organization and eating habits of high-income countries⁷7 Nilsen SM, Krokstad S, Holmen TL, Westin S. Adolescents' health-related dietary patterns by parental socio-economic position, the Nord-Trøndelag Health Study (HUNT). Eur J Public Health 2010; 20(3):299-305.

8 van Ansem WJ, Schrijvers CT, Rodenburg G, van de Mheen D. Maternal educational level and children's healthy eating behaviour: role of the home food environment (cross-sectional results from the INPACT study). Int J Behav Nutr Phys Act 2014; 11:113.^-99 Fahlman MM, McCaughtry N, Martin J, Shen B. Racial and socioeconomic disparities in nutrition behaviors: targeted interventions needed. J Nutr Educ Behav 2010; 42(1):10-16.. Exploring the differences in dietary behaviors across socioeconomic level in low- to middle-income countries may provide overall evidence of the magnitude of existing inequalities⁶6 World Health Organization (WHO). Handbook on health inequality monitoring: with a special focus on low- and middle-income countries. Geneva: WHO; 2013.. In this study, we assessed socioeconomic inequalities in the consumption of in natura or minimally processed and ultra-processed foods in Brazilian adolescents.

Methods

We used cross-sectional data from the Brazilian National Survey of School Health (Pesquisa Nacional da Saúde do Escolar - PeNSE) carried out in 2015. Additional information about the PeNSE methods and sampling procedures has been published elsewhere¹⁰10 Oliveira MM, Campos MO, Andreazzi MAR, Malta DC. Characteristics of the National Adolescent School-based Health Survey - PeNSE, Brazil. Epidemiol Serv Saude 2017; 26(3):605-616.. Briefly, PeNSE enrolled two independent samples of students attending high school and 9^th grade, mostly aged between 14 and 15 years. In this study, we analyzed data from 9^th grade students from public and private schools from 27 Brazilian federative units.

The sampling strategy included stratification per cluster and multi-stage selection. The primary and secondary sample units were schools and classes, respectively. School selection was proportional to the number of 9^th grade classes, while classes were selected at random. Schools with at least two 9^th grade classes had one class selected while schools with three or more classes had two classes selected. All students enrolled in the selected classes were invited to participate in the study. Those who agreed to participate answered a standardized questionnaire utilizing smartphones provided by the research team¹⁰10 Oliveira MM, Campos MO, Andreazzi MAR, Malta DC. Characteristics of the National Adolescent School-based Health Survey - PeNSE, Brazil. Epidemiol Serv Saude 2017; 26(3):605-616.. The questionnaire was based on the School-Based Student Health Survey¹⁰10 Oliveira MM, Campos MO, Andreazzi MAR, Malta DC. Characteristics of the National Adolescent School-based Health Survey - PeNSE, Brazil. Epidemiol Serv Saude 2017; 26(3):605-616. and the Youth Risk Behaviour Surveillance System, which were adapted to the Brazilian context¹¹11 Guedes DP, Lopes CC. Validação da versão brasileira do youth risk behavior survey 2007. Rev Saude Publica 2010; 44:840-850.. Adolescents responded to the questionnaire at school, guided by a researcher.

Assessment of diet

Adolescents reported their consumption of in natura or minimally processed foods and ultra-processed foods through the following question “In the last seven days, how many days did you consume…?” regarding each one of the following foods: beans, vegetables, fresh fruits, fried salty foods (e.g. french fries, fried chicken), sweets (e.g. candies, bubble gum), ultra-processed meat (e.g. ham, chicken nuggets), instant noodles, fast foods, and soda. The answers ranged from zero to seven days, and portion size was not reported. These questions have been validated for Brazilian adolescents¹²12 Tavares LF, Castro IRR, Levy RB, Cardoso LO, Passos MD, Brito FSB. Relative validity of dietary indicators from the Brazilian National School-Based Health Survey among adolescents in Rio de Janeiro, Brazil. Cad Saude Publica 2014; 30(5):1029-1041..

Dietary scores were calculated for each foods group, summing the weekly frequency of each diet indicator included (0-7 days/week). For the in natura or minimally-processed foods score, the indicator variables were beans, vegetables, and fresh fruits consumption. Consumption of soda, sweets, and ultra-processed foods were indicators for the ultra-processed foods score. As the weekly frequency of three items were used as indicators for each score, the score variables ranged from 0 to 21, meaning they adolescents ate none of the items weekly, or every item every day, respectively. Fast foods and fried salty foods were not included in any score, as their description does not clearly discriminate the degree and purpose of foods processing²2 Brasil. Ministério da Saúde (MS). Guia alimentar para a população brasileira. Brasília: MS; 2014..

Equality indicators

Equality indicators were sex (male or female), maternal education, and socioeconomic level. Maternal education was obtained by asking the question “What level of education did your mother achieve?” and it was categorized as follows: 0-8, 9-11, ≥12 years of study, and unknown. The socioeconomic level was obtained through the possession of goods (landline phone, cell phone, computer, internet service, car, motorcycle, number of bathrooms), and the presence of a maid in the home. The choice of such variables was based on the original research report¹⁰10 Oliveira MM, Campos MO, Andreazzi MAR, Malta DC. Characteristics of the National Adolescent School-based Health Survey - PeNSE, Brazil. Epidemiol Serv Saude 2017; 26(3):605-616. which uses the variables of sex, maternal education, possession of goods, number of bathrooms at home and availability of maid services at home to characterize socioeconomic aspects. The method used in this article has been widely used in national surveys¹³13 Barros AJD, Victora CG. A nationwide wealth score based on the 2000 Brazilian demographic census. Rev Saude Publica 2005; 39(4):523-529.^,1414 Azeredo CM, de Rezende LFM, Mallinson PAC, Ricardo CZ, Kinra S, Levy RB, Barros AJD. Progress and setbacks in socioeconomic inequalities in adolescent health-related behaviours in Brazil: results from three cross-sectional surveys 2009-2015. BMJ Open 2019; 9(3):e025338.. In addition, the World Health Organization¹⁵15 World Health Organization (WHO). Handbook on health inequality monitoring: with a special focus on low- and middle-income countries. Geneva: WHO; 2013. suggests the use of more complex analyses, such as principal components analysis, which considers statistical methods to determine the weights of items in the index.

Statistical analysis

Descriptive statistics were provided for the whole sample and for those in the first and fifth quintile of the socioeconomic level, which were calculated using principal component analysis, as a reduction method to define this variable¹³13 Barros AJD, Victora CG. A nationwide wealth score based on the 2000 Brazilian demographic census. Rev Saude Publica 2005; 39(4):523-529.. The variable was calculated based on the first component, and the weights of each asset were used to calculate a score, which was grouped in quintiles, with the first quintile being the lowest socioeconomic level. The first component explained 29.71% of the variance of the socioeconomic level indicators used.

The average score of the dependent variables (in natura or minimally processed foods and ultra-processed foods) were described according to the categories of each equality indicator (socioeconomic level, sex, maternal education). Simple measures of inequality were estimated between sex, maternal education categories, and between the lowest and the highest socioeconomic level (pairwise comparisons). Simple linear regression models were used to calculate mean differences and 95% confidence intervals between dichotomous equality indicators (e.g. participants in the lowest socioeconomic level scored an average -2.64 (95%CI -2.87; -2.40) in the ultra-processed foods score compared to those in the highest socioeconomic level). For ratios, Poisson regression with robust variance were used (e.g. participants in the lowest socioeconomic level had a 25% (0.75; 95% CI 0.73; 0.78) lower ultra-processed foods score compared to those in the highest socioeconomic). The coefficients of the dependent variables for each social indicator and their respective 95% confidence intervals were reported.

The slope index of inequality and the concentration index were calculated to evaluate inequalities between socioeconomic level in relation to the consumption of in natura/minimally-processed foods and ultra-processed foods. Slope index of inequality analyses the distance of extreme categories (lowest and highest socioeconomic level) against the midpoint of the cumulative range of the socioeconomic level. Where the slope indicates the extent of inequality (with values close to zero indicating no inequalities), positive values suggesting that the dietary indicator is more prevalent in adolescents with higher socioeconomic level, and negative values suggesting it is more prevalent adolescents with lower socioeconomic level. The concentration index provides a gradient value across subgroups ranked by socioeconomic level, where negative values suggest the outcome measure is concentrated in lower socioeconomic level, with values ranging between -1 and +1, and values above 0.2 representing reasonable levels of relative inequality⁶6 World Health Organization (WHO). Handbook on health inequality monitoring: with a special focus on low- and middle-income countries. Geneva: WHO; 2013..

All analyses were performed using Stata, version 15.0 for Windows.

Ethical considerations

This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Comissão Nacional de Ética em Pesquisa (CONEP) (Brazilian National Ethics Committee, protocol No. 1.006.467, of 30.03.2015). Written informed consent was obtained from all subjects/patients.

Results

A total of 101,689 adolescents participated in the study (51% girls; 14.2 years). The proportion of boys was higher among those in the highest socioeconomic level (52%) compared to those in the lowest socioeconomic level (46%). Maternal education was higher among adolescents in the highest socioeconomic level (43% of mothers had ≥12 years of education) compared to the lowest socioeconomic level (3% of mothers had ≥12 years of education) (Table 1).

Table 2 shows the mean differences and ratios of consumption of in natura/minimally processed foods and ultra-processed foods by socioeconomic level, sex, and maternal education. Consumption of in natura/minimally processed foods and ultra-processed foods were higher among participants in the highest socioeconomic level compared to the lowest socioeconomic level. Participants whose mother had ≥12 years of education consumed more ultra-processed foods compared to those who had mothers with ≤8 years of education. Differences found in the scores of in natura or minimally processed foods and ultra-processed foods between boys and girls were smaller than 1 point.

The slope index of inequality indicated that in natura/minimally-processed foods and ultra-processed foods were more prevalent in the highest socioeconomic level (Table 3). However, values were close to zero (0.0143 for in natura/minimally-processed foods score, and 0.0346 for ultra-processed foods score), suggesting a small difference in the consumption of in natura/minimally-processed foods and ultra-processed foods across socioeconomic level. Similar results were observed for the concentration index of inequality analyses, with small but statistically significant differences in the in natura/minimally-processed foods score (CIX=0.0215) and ultra-processed foods (CIX=0.0475), suggesting that both dietary score were concentrated among adolescentes with higher socioeconomic level.

Discussion

This study examined differences in adolescents’ consumption of in natura/minimally-processed foods and ultra-processed foods according to socioeconomic level, sex, and maternal education. The largest differences in consumption of in natura/minimally-processed foods and ultra-processed foods were found between the lowest and the highest socioeconomic level. Adolescents in the highest socioeconomic level reported consuming more in natura/minimally processed foods, but also more ultra-processed foods, compared to adolescents in the lowest socioeconomic level.

Similar results have been found in high-income countries⁹9 Fahlman MM, McCaughtry N, Martin J, Shen B. Racial and socioeconomic disparities in nutrition behaviors: targeted interventions needed. J Nutr Educ Behav 2010; 42(1):10-16.^,1616 Novakovic R, Cavelaars A, Geelen A, Nikolic M, Altaba II, Viñas BR, Ngo J, Golsorkhi M, Medina MW, Brzozowska A, Szczecinska A, Cock D, Vansant G, Renkema M, Majem LS, Moreno LA, Glibetic M, Gurinovic M, Veer PV, Groot LCPVM. Socio-economic determinants of micronutrient intake and status in Europe: a systematic review. Public Health Nutr 2014; 17(5):1031-1045.

17 Giskes K, Avendano M, Brug J, Kunst AE. A systematic review of studies on socioeconomic inequalities in dietary intakes associated with weight gain and overweight/obesity conducted among European adults. Obes Rev 2010; 11(3):413-429.^-1818 Darmon N, Drewnowski A. Does social class predict diet quality? Am J Clin Nutr 2008; 87(5):1107-1117.. However, in high-income countries, socioeconomically disadvantaged groups have lower consumption of fruits and vegetables and higher consumption of energy-dense products¹⁶16 Novakovic R, Cavelaars A, Geelen A, Nikolic M, Altaba II, Viñas BR, Ngo J, Golsorkhi M, Medina MW, Brzozowska A, Szczecinska A, Cock D, Vansant G, Renkema M, Majem LS, Moreno LA, Glibetic M, Gurinovic M, Veer PV, Groot LCPVM. Socio-economic determinants of micronutrient intake and status in Europe: a systematic review. Public Health Nutr 2014; 17(5):1031-1045.

17 Giskes K, Avendano M, Brug J, Kunst AE. A systematic review of studies on socioeconomic inequalities in dietary intakes associated with weight gain and overweight/obesity conducted among European adults. Obes Rev 2010; 11(3):413-429.^-1818 Darmon N, Drewnowski A. Does social class predict diet quality? Am J Clin Nutr 2008; 87(5):1107-1117.. In our study, Brazilian adolescents in the lowest socioeconomic level reported consuming less in natura/minimally processed foods, but also less ultra-processed foods compared to those in the highest socioeconomic level. The availability of in natura/minimally processed foods, such as fruits and vegetables, may be limited in disadvantaged neighbourhoods¹⁹19 Sallis JF, Glanz K. Physical activity and food environments: solutions to the obesity epidemic. Milbank Q 2009; 87(1):123-154.

20 Larson NI, Story MT, Nelson MC. Neighborhood environments: disparities in access to healthy foods in the U.S. Am J Prev Med 2009; 36(1):74-81.^-2121 Pearce J, Hiscock R, Blakely T, Witten K. A national study of the association between neighbourhood access to fast-food outlets and the diet and weight of local residents. Health Place 2009; 15(1):193-197.. Moreover, other individual characteristics, such as knowledge and self-efficacy for changing dietary behaviors are likely less prevalent in adolescents with low socioeconomic level⁹9 Fahlman MM, McCaughtry N, Martin J, Shen B. Racial and socioeconomic disparities in nutrition behaviors: targeted interventions needed. J Nutr Educ Behav 2010; 42(1):10-16., which may also explain these differences.

In our study, participants in the highest socioeconomic level consumed more ultra-processed foods compared to those participants in the lowest socioeconomic level, which differs from studies carried out in high-income countries. This may be due to ultra-processed foods being cheaper and easily accessible in high-income countries²⁰20 Larson NI, Story MT, Nelson MC. Neighborhood environments: disparities in access to healthy foods in the U.S. Am J Prev Med 2009; 36(1):74-81.^,2121 Pearce J, Hiscock R, Blakely T, Witten K. A national study of the association between neighbourhood access to fast-food outlets and the diet and weight of local residents. Health Place 2009; 15(1):193-197., but more costly for the Brazilian population²²22 Claro RM, Maia EG, Costa BV L, Diniz DP. Food prices in Brazil: prefer cooking to ultra-processed foods. Cad Saude Publica 2016; 32(8):e00104715., especially in smaller stores, compared to supermarkets²³23 Machado PP, Claro RM, Canella DS, Sarti FM, Levy RB. Price and convenience: The influence of supermarkets on consumption of ultra-processed foods and beverages in Brazil. Appetite 2017; 116(1):381-388.. In addition, living in poor neighborhoods may be a barrier for accessing cheaper healthy foods in supermarkets, compared to those who live close to such establishments²⁴24 Black C, Moon G, Baird J. Dietary inequalities: What is the evidence for the effect of the neighbourhood food environment? Health Place 2014; 27:229-242.^,2525 Morland K, Wing S, Diez Roux A. The contextual effect of the local food environment on residents' diets: the atherosclerosis risk in communities study. Am J Public Health 2002; 92(11):1761-1767. or those who can afford to own a car²⁶26 Aggarwal A, Cook AJ, Jiao J, Seguin RA, Moudon AV, Hurvitz PM, Drewnowski A. Access to Supermarkets and Fruit and Vegetable Consumption. Am J Public Health 2014; 104(5):917-923.. These differences in price may play an important role in the accessibility to ultra-processed foods. In high-income countries, ultra-processed foods are more aggressively advertised²⁷27 Yancey AK, Cole BL, Brown R, Williams JD, Hillier A, Kline RS, Ashe M, Grier SA, Backman D, McCarthy WJ, A cross-sectional prevalence study of ethnically targeted and general audience outdoor obesity-related advertising. Milbank Q 2009; 87(1):155-184., which usually is very appealing and, thus, preferred by children and adolescents²⁸28 Santos MM, Marreiros CS, Silva HBS, Oliveira ARS, Cruz KJC. Associations between taste sensitivity, preference for sweet and salty flavours, and nutritional status of adolescents from public schools. Rev Nutr 2017; 30(3):369-375.. The impact of the advertisement is not clear in Brazil, as marketing these products to children and adolescents is prohibited²⁹29 São Paulo. Projeto de Lei nº 193, de 29 de março de 2008. Regulamenta a publicidade infantil de alimentos no Estado de São Paulo. Diário Oficial; 2008.. In addition, our results indicate that adolescents living in households of higher socioeconomic level had higher consumption of all evaluated foods groups. The lowest consumption of foods in the most vulnerable group does not occur uniformly because this group has restricted access to all types of foods. In addition, the amount of foods consumed is low. The ideal consumption of fresh/minimally processed foods should be 21 points, that is, daily consumption of these foods. However, the group with the highest socioeconomic level consumed just over half. Therefore, despite being a higher consumption than the group with lower socioeconomic level, it is still an insufficient consumption, indicating even more precarious foods consumption among the most vulnerable adolescents.

Our findings corroborate studies analyzing Brazilian and North American adolescents, which found no significant differences between the eating behavior of boys and girls⁹9 Fahlman MM, McCaughtry N, Martin J, Shen B. Racial and socioeconomic disparities in nutrition behaviors: targeted interventions needed. J Nutr Educ Behav 2010; 42(1):10-16.^,3030 Louzada MLC, Martins APB, Canella DS, Baraldi LG, Levy RB, Claro RM, Moubarac JC, Cannon G, Monteiro CA. Ultra-processed foods and the nutritional dietary profile in Brazil. Rev Saude Publica 2015; 49:38.. However, studies with European adolescents have found that girls eat more healthy foods and avoid eating sugar-rich foods compared to boys⁷7 Nilsen SM, Krokstad S, Holmen TL, Westin S. Adolescents' health-related dietary patterns by parental socio-economic position, the Nord-Trøndelag Health Study (HUNT). Eur J Public Health 2010; 20(3):299-305.^,3131 Inchley J, Currie D, Young T, Samdal O, Torsheim T, Augustson L, Mathison F, Aleman-Diaz A, Molcho M, Weber M, BarnekowV. Growing up unequal: gender and socioeconomic differences in young people's health and well-being: Health Behaviour in School-Aged Children (HBSC) Study: international report from the 2013/2014 survey. Copenhagen: WHO Regional Office for Europe; 2016.. These findings point out the differences between Europeans’ eating behavior compared to Americans’, with the latter being similar to our findings. Similar foods consumption between boys and girls may be the result of the widespread consumption of rice and beans among the Brazilian population³⁰30 Louzada MLC, Martins APB, Canella DS, Baraldi LG, Levy RB, Claro RM, Moubarac JC, Cannon G, Monteiro CA. Ultra-processed foods and the nutritional dietary profile in Brazil. Rev Saude Publica 2015; 49:38., as well as the implementation of national policies, such as the National School Foods Program (Programa Nacional de Alimentação Escolar - PNAE), which has incorporated in natura/minimally-processed foods into the school meals, and discourage the commercialization of energy dense processed foods in school canteens.

Maternal education was not strongly associated with consumption of in natura/minimally processed foods and ultra-processed foods. These findings differed from studies in European countries where adolescents (aged 8-12 years) with more educated mothers reported eating more fruits and vegetables, while their counterparts with less educated mothers reported eating more fried foods⁸8 van Ansem WJ, Schrijvers CT, Rodenburg G, van de Mheen D. Maternal educational level and children's healthy eating behaviour: role of the home food environment (cross-sectional results from the INPACT study). Int J Behav Nutr Phys Act 2014; 11:113.^,3232 Cribb VL, Jones LR, Rogers IS, Ness AR, Emmett PM. Is maternal education level associated with diet in 10-year-old children? Public Health Nutr 2011; 14(11):2037-2048.. Mothers with more education may have more knowledge related to healthy eating behavior, which could impact their children’s diet⁸8 van Ansem WJ, Schrijvers CT, Rodenburg G, van de Mheen D. Maternal educational level and children's healthy eating behaviour: role of the home food environment (cross-sectional results from the INPACT study). Int J Behav Nutr Phys Act 2014; 11:113.. A review of interventions at the family setting showed that increased knowledge can improve dietary habits (e.g., increase fruit intake and reduce fat intake)³³33 Black AP, D'Onise K, McDermott R, Vally H, O'Dea K. How effective are family-based and institutional nutrition interventions in improving children's diet and health? A systematic review. BMC Public Health 2017; 17(1):818-837.. Additionally, maternal education could be highly related to income⁸8 van Ansem WJ, Schrijvers CT, Rodenburg G, van de Mheen D. Maternal educational level and children's healthy eating behaviour: role of the home food environment (cross-sectional results from the INPACT study). Int J Behav Nutr Phys Act 2014; 11:113.^,99 Fahlman MM, McCaughtry N, Martin J, Shen B. Racial and socioeconomic disparities in nutrition behaviors: targeted interventions needed. J Nutr Educ Behav 2010; 42(1):10-16.^,2020 Larson NI, Story MT, Nelson MC. Neighborhood environments: disparities in access to healthy foods in the U.S. Am J Prev Med 2009; 36(1):74-81.^,3232 Cribb VL, Jones LR, Rogers IS, Ness AR, Emmett PM. Is maternal education level associated with diet in 10-year-old children? Public Health Nutr 2011; 14(11):2037-2048., which may also positively impact the availability of healthy foods at home³¹31 Inchley J, Currie D, Young T, Samdal O, Torsheim T, Augustson L, Mathison F, Aleman-Diaz A, Molcho M, Weber M, BarnekowV. Growing up unequal: gender and socioeconomic differences in young people's health and well-being: Health Behaviour in School-Aged Children (HBSC) Study: international report from the 2013/2014 survey. Copenhagen: WHO Regional Office for Europe; 2016. and overall foods diversity. This has been evidenced in high-income countries, which may be explained by more disposable income by families with higher socioeconomic level, but not related specifically to maternal education³⁴34 Powell LM, Bao Y. Food prices, access to food outlets and child weight. Econ Hum Biol 2009; 7(1):64-72.. However, a study with a national sample of Brazilian adolescents (aged 13-15 years) found that those with mothers that are more educated ate more sweets and drank softer drinks³⁵35 Ferreira NL, Claro RM, Lopes ACS. Consumption of sugar-rich food products among Brazilian students: National School Health Survey (PeNSE 2012). Cad Saude Publica 2015; 31(12):2493-504., which differs from results found in high-income countries. Our study revealed small differences when comparing adolescents’ diet by maternal education, which may reflect that it is not only health-related knowledge that affects eating behaviors but inequalities due to access and distribution of foods³⁶36 Rose DD. Interventions to reduce household food insecurity: a synthesis of current concepts and approaches for Latin America. Rev Nutr 2008; 21(s10):159s-173s.. Educational interventions may have a positive effect on healthy foods consumption³³33 Black AP, D'Onise K, McDermott R, Vally H, O'Dea K. How effective are family-based and institutional nutrition interventions in improving children's diet and health? A systematic review. BMC Public Health 2017; 17(1):818-837., and policies aimed at the school setting (e.g. PNAE) may improve diversity and distribution of healthy foods among children and adolescents in low socioeconomic level. Finally, changes in the social structure may be needed to consistently improve these inequalities in the long-term³⁷37 Landmann-Szwarcwald C, Macinko J. A panorama of health inequalities in Brazil. Int J Equity Health 2016; 15:174..

This study has some strength, such as the nationally representative sample, which included adolescents from diverse backgrounds. We used two different scores to report the dietary consumption of Brazilian adolescents, and calculated simple and complex measures of inequality by socioeconomic level, sex and maternal education. Our study also has some limitations. Consumption of in natura/minimally and ultra-processed foods were self-reported and limited to a relatively small number of dietary indicators. In addition, the questionnaire did not include foods preparation and amount consumed. However, information related to weekly foods consumption frequency revealed to be useful to identify differences between socioeconomic level in a large representative sample of adolescents.

In conclusion, adolescents from higher socioeconomic level ate more in natura/minimally processed foods, but also ultra-processed foods, compared to adolescents from lower socioeconomic level. Differences in the consumption of in natura/minimally processed foods and ultra-processed foods between sex and maternal education subgroups were small. Future studies should focus on understanding the causal pathways between socioeconomic level and the consumption of in natura/minimally processed foods and ultra-processed foods, in order to inform effective policies and interventions.

References

Publication Dates

History