Artigos

Conteúdo (article):

ASSOCIAÇÃO ENTRE PADRÕES ALIMENTARES COM FATORES SOCIOECONÔMICOS E AMBIENTE ALIMENTAR EM UMA CIDADE DO SUL DO BRASIL
ASSOCIATION BETWEEN DIETARY PATTERNS AND SOCIOECONOMIC FACTORS AND FOOD ENVIRONMENT IN A CITY IN THE SOUTH OF BRAZIL

Autores
Caroline Marques de Lima Cunha1 – https://orcid.org/0000-0002--0529-1498
Raquel Canuto2https://orcid.org/0000-0002-4042-1913
Priscila Bárbara Zanini Rosa1 - https://orcid.org/0000-0002-1774-073X
Luana Schimmelfenning Longarai3 – http://orcid.org/0000-0002-9692-7286
Ilaine Schuch2 https://orcid.org/0000-0001-6861-4365

1Programa de Pós-Graduação em Alimentação, Nutrição e Saúde. Universidade Federal do Rio Grande do Sul. Porto Alegre, RS. Brasil.
2Programa de Pós-Graduação em Alimentação, Nutrição e Saúde. Departamento de Nutrição. Universidade Federal do Rio Grande do Sul.Rua Ramiro Barcelos, 2400, Santa Cecília, CEP: 90.035-003. Porto Alegre/RS. ischuch@hcpa.edu.br
3Curso de Graduação em Nutrição. Universidade Federal do Rio Grande do Sul. Porto Alegre, RS. Brasil.

Resumo: Objetivou-se estudar a associação entre padrão alimentar, fatores demográficos, socioeconômicos e o ambiente alimentar em uma população de adultos e idosos de uma cidade do Sul do Brasil. Estudo transversal com pessoas de ambos os sexos e idade entre 20 e 70 anos. Os padrões alimentares foram identificados através da análise de componentes principais. Regressão de Poisson foi utilizada para estimar as razões de prevalência brutas e ajustadas e os intervalos de 95% de confiança (IC95%). Identificou-se quatro padrões alimentares: saudável, tradicional, carboidratos e açúcares refinados e fast food. Encontraram-se associações positivas entre: sexo feminino e maior renda com o padrão alimentar saudável; cor da pele preta ou parda e domicílios com seis ou mais moradores, com padrões alimentares tradicional e carboidratos e açúcares refinados; maior escolaridade com o padrão fast food. Realizar as principais refeições em casa associou-se ao padrão tradicional e, almoço ou jantar fora do domicílio com os padrões carboidratos e açúcares refinados e fast food. Menor posição socioeconômica leva ao maior consumo do padrão alimentar tradicional e/ou rico em carboidratos e açúcares e, uma maior posição socioeconômica parece permitir uma escolha entre padrão saudável ou fast food.
Palavras-chave: Padrões alimentares. Fatores socioeconômicos. Desigualdades socioespaciais em saúde

Abstract: The aim of this study was to investigate the association between dietary patterns and demographic and socioeconomic factors and the food environment among adults and older persons in a city in the south of Brazil. We conducted a cross-sectional study with people of both sexes aged between 20 and 70 years. Dietary patterns were identified using principal component analysis. Poisson regression was used to estimate crude and adjusted prevalence ratios and 95% confidence intervals. Four dietary patterns were identified: Healthy; Traditional; Refined Carbs and Sugars; and Fast Food. Positive associations were found between being female and higher income and the Healthy dietary pattern; being black or brown and living in a household with at least six members and the Traditional and Refined Carbs and Sugars patterns; and higher education and the Fast Food dietary pattern. Having main meals at home was associated with the Traditional pattern and having lunch or dinner away from home was the associated with Refined Carbs and Sugars and Fast Food patterns. Lower socioeconomic status leads to higher consumption of the Traditional and/or Refined Carbs and Sugars dietary patterns, while higher socioeconomic status appears to enable individuals to choose between Healthy or Fast Food patterns.
Keywords: Food patterns. Socioeconomic factors. Socio-spatial health inequalities

INTRODUCTION
Dietary patterns represent a broad picture of food and nutrient consumption and are characterized on the basis of usual eating behavior1. Over recent decades, in various countries, including Brazil, dietary patterns have undergone important changes, characterized mainly by an increase in the intake of fats, sugars and ultra-processed foods and reduction in the consumption of nutrient-rich foods, such as fruits and vegetables. These changes are associated with chronic non-communicable diseases (CNCDs)2,3.
International and Brazilian studies have shown consistent associations between dietary patterns and social, economic and life-style characteristics and other factors that can also influence eating patterns, such as household size, marital status and skin color4, 5,6,7.
More recently, studies have shown that the food environment is an important social determinant of individual food consumption. Food environments have four dimensions: physical (availability, quality, and promotion), economic (cost), policy (rules), and sociocultural (norms and beliefs of an individual or group regarding foods)8.
The food environment influences dietary patterns through access, the availability, price and quality of foods, and other individual factors, such as culture, preferences, acceptability and knowledge of the food9, 10.The distance between food purchase locations and homes and the form of transport used when purchasing food are therefore also factors that make up the food environment.
Socioeconomic and environmental characteristics are not uniformly distributed across big cities. People from poorer and socially vulnerable areas such as favelas share sociodemographic characteristics and social environments that are different to those of individuals living in more affluent areas11, and may be more likely to have less healthy dietary patterns.
Studies that seek to identify dietary patterns and their association with socioeconomic factors and the food environment are scarce in Brazil, especially when it comes to people living in poor areas. The aim of this study was therefore to investigate the association between dietary patterns and demographic and socioeconomic factors and the food environment in a city in the south of Brazil.

METHODOLOGY
We conducted a cross-sectional population-based study in the catchment area of a primary care unit (PCU) in the center of Porto Alegre, Rio Grande do Sul. The study is part of a research project entitled “A study of the social and environmental determinants of diet and nutrition: an ecosocial approach”, approved by Rio Grande do Sul Federal University’s research ethics committee (approval number CAAE 46934015.3.0000.5347). The center of Porto Alegre has around 260,000 inhabitants12and three primary care units. The population within the catchment area of the PCU included in this study consists of approximately 12,000 families. Part of the population live in four poorer areas or favelas (per capita income of R$1,700.00), while the rest live in more affluent neighborhoods (per capita income of R$4,000.00)13.
To ensure a representative sample, we used proportional sampling to include the same proportion of individuals from the poor and affluent areas. In the poor areas (around 250 families), all eligible participants that accepted the invitation to participate in the study were included. To maintain the same proportion of individuals, we included the same number of participants from the affluent areas. Participants included men and women aged between 20 and 70 years. Individuals with physical or mental impairments that made it impossible to collect data and pregnant women were excluded. Only one person per household was included and we sought to alternate the sex of the respondent between households.
The data were collected between October 2018 and June 2019. First, with the help of community health workers, the areas were mapped and the households were identified using maps and addresses. The research team then visited the areas, identifying individuals who met the inclusion criteria and inviting them to participate in the study. The interviews were conducted during the initial visit or scheduled for a later date, preferably at the respondent’s home, or at the PCU when requested by the participant.
We used a standardized questionnaire devised to obtain information on the following socioeconomic and demographic characteristics: sex (female/male); age (years); self-declared race/skin color; classified according to the 2010 census categories12 (white/black/brown/yellow/indigenous); level of education (no schooling/junior high school completed or not completed/high school completed or not completed/completed higher education/post-graduation); marital status (single/living in stable union/married/widowed/separated/divorced); monthly family income in minimum wages (<1 MW/1 to 2 MWs/2 to 3 MWs/3 to 4 MWs/4 to 5 MWs/>5 MWs); benefits (not received/Bolsa Família/pension/Continuous Cash Benefit – CCB/other); and number of household members (≤ 3 to ≥ 6).
Food environment was assessed using a questionnaire with the seven most commonly consumed food items in Brazil: industrialized products (biscuits, soft-drinks and instant noodles), fruits, vegetables and legumes, meats, bread, rice and beans)9. The instrument investigated the purchase location for each item (growers\' market/fruit and vegetable markets, supermarket, market, public market, warehouse store, delicatessen/bakery, bar, grocery store, butcher, home vegetable garden, community vegetable garden, donations - family, neighbors, organizations, other) and form of transport t(on foot, bicycle, car/motorbike, public transport). In addition, eating place was identified using a question that first inquired whether the respondent ate lunch and dinner and then went on to ask the eating place (at home, at work, snack bar, restaurant, other).
Food consumption was assessed using a food frequency questionnaire (FFQ) consisting of 65 foods and adapted to 85 typical items in the local food culture. The food list was devised based on consumption data from the dietary records of adults in Niterói, Rio de Janeiro14and validated for the population of the metropolitan region of Porto Alegre15. The questionnaire investigated daily, weekly, monthly or annual frequency of consumption (zero to seven times).
The data were entered into a data collection form in EpiData® version 3.1 using double entry. Data analysis was carried out using Stata version 12.0 (StataCorp, College Station, United States) and SPSS (Statistical Package for the Social Sciences) version 18.0.
Dietary patterns were analyzed using the a posteriori method, dividing the collected empirical food data into clusters on the basis of statistical analysis. This method uses multivariate techniques to identify similarities in eating habits or consumed food groups based on their interrelations16, 17,18.
Before identifying the dietary patterns, the different frequencies of consumption of food items (including seasonal foods) were transformed into annual consumption frequencies. The frequency of consumption of each item was then determined and items with a frequency of consumption below 5% were excluded19. The food items were then divided into 48 groups based on the statistical correlations between the dietary items (p ≤0.05) and nutritional and cultural similarities. Other foods were not clustered, either because it made no sense to group them (artificial sweetener, for example), or because they may have been indicative of a certain dietary pattern, such as rice and beans, for example20.
The dietary patterns were theoretically derived using principal component analysis (PCA). The applicability of the method was verified using the Kaiser-Mayer-Olkin (KMO) test, which measures the strength of the relationship between variables, where a value of ≥ 0.60 is considered adequate. Bartlett’s test of sphericity was used to test the null hypothesis (no relationship between the variables), adopting a p-value of < 0.05 to indicate that the dataset was suitable for analysis. To calculate the sample power needed to identify dietary patterns, we followed the criteria proposed by Hair et al.21 (five individuals per food item included in the principal component analysis).
Varimax rotation was used to examine the structure exploratory factors represented in the FFQ. The number of factors to extract was determined using a scree plot graph – where the steepest points indicated the appropriate number of components to retain– and the Kaiser criterion, where eigenvalues above 1 were accepted. The food items with absolute factor loadings of ≥ 0.30 were considered to make a significant contribution to a given factor. The dietary patterns were named according to the foods loaded most on each factor and to cultural aspects. Each of the derived patterns were divided into terciles and dichotomized as follows: tercile 3 (high consumption) vs. terciles 1 and 2 (low consumption)22,23.
Crude and adjusted prevalence ratios and95% confidence intervals (95% CI) were obtained using Poisson regression with robust variance. Variables with a significance level of up to 20% in the crude analysis were included in the multivariate model. For the multivariate analysis, we used a conceptual model (first level, demographic variables; second level, socioeconomic variables). Only variables with a p-value of <0.20 were retained in the second level. After adjustment, all variables with a p-value of ≤ 0.05 were considered to be associated with the dietary patterns.

RESULTS
The total sample comprised 400 participants. First, all residents of poorer areas that accepted the invitation to participate in the study were included (n = 201), followed by 199 residents selected from the more affluent areas.
Table 1 shows the general characteristics of the study population by area. Most of the respondents were women and the mean age of the sample was 47 years (SD = 13.98). Most of the sample did not live with a partner, were white, had completed high school, had a family income of three to five minimum wages, and lived in households with up to three members. Half of the respondents received some kind of welfare benefit. The respondents from poorer areas were younger, more likely to be brown or black and receive welfare benefits, had a lower level of education and income, and lived in households with a higher number of members.
Table 2 shows the four dietary patterns identified in the analysis (Healthy, Traditional, Refined Carbs and Sugars, and Fast Food) and their respective components, factor loadings and level of explained variance. Both the KMO coefficient and Bartlett’s test of sphericity indicate that the correlations between the items were sufficient and the variables were suitable for factor analysis. The dietary pattern with the highest explained variance was Healthy (10.84%), meaning it is the pattern that best represents the consumption of the study population.
Table3 shows the prevalence of high consumption of each dietary pattern according to demographic and social characteristics. High consumption of the Healthy dietary pattern was associated with being female, receiving or living with someone who received a pension, CCB or other welfare benefit, and living in more affluent areas. High consumption of the Traditional dietary pattern was associated with being young and brown, lower levels of income and education, receiving or living with someone who received benefits from the family benefit program (Programa Bolsa Família -PBF), living in poorer areas, and living in households with at least six members. High consumption of the Refined Carbs and Sugars dietary pattern was associated with being younger and black, lower levels of income and education, receiving or living with someone who received benefits from the PBF, living in more vulnerable areas, and living in households with at least six members. Finally, high consumption of the Fast Food dietary pattern was associated with being young and white, higher levels of education, family income above five minimum wages, and living in more affluent areas.
After adjustment, the Healthy dietary pattern was directly associated with being female and having a monthly family income above five minimum wages, and inversely associated with living in a household with four to five members. The Traditional dietary pattern was directly associated with being black or brown and living in a household with at least six members, inversely associated with living in more affluent areas, and showed an inversely proportional association with age and level of education. The Refined Carbs and Sugars dietary pattern was directly associated with being brown and black and livening in a household with at least six members, and inversely associated with being older and living in affluent areas. Finally, the Fast Food dietary pattern was directly associated with higher levels of education and inversely associated with older age and being brown or black (Table 4).
Table 5 shows the frequency distribution of the food environment variables and their association with dietary patterns. Individuals who showed greater adherence to the Traditional dietary pattern were more likely to purchase fruits, vegetables and legumes in a supermarket, market or warehouse store and went on foot or by bicycle to purchase industrialized products, meats and bread. The most frequent places of purchase for bread among individuals with higher consumption of the Refined Carbs and Sugars dietary pattern were delicatessen/bakery, bar or grocery store. Individuals with higher consumption of the Fast Food dietary pattern went by car/motorbike or public transport to purchase the majority of food items. Having main meals at home was associated with the Traditional dietary pattern, while eating lunch or dinner away from home was associated with the Refined Carbs and Sugars and Fast Food patterns.

DISCUSSION
Being female and higher income were associated with consumption of the Healthy dietary pattern, while being younger, brown or black, lower level of education, larger number of household members, and living in poorer areas were associated with consumption of less healthy patterns or a smaller dietary share of fruits and vegetables. In addition, buying foods in supermarkets, going on foot or by bicycle to purchase foods, and having main meals at home were associated with consumption of the Traditional dietary pattern, while going by car or bus to purchase foods and eating main meals away from home were associated with dietary patterns including ultra-processed foods that are related to risk of CNCDs.
With regard to composition, the Healthy dietary pattern was rich in fruits, vegetables and whole-grain cereals, while the Traditional dietary pattern was composed of staple foods such as rice, beans, pasta, potatoes and red meat. At the same time, we identified two dietary patterns composed predominantly of ultra-processed foods. The results corroborate the findings of previous studies in Brazil reporting an increase in the consumption of foods rich in carbohydrates, sugar and fats, while at the same time identifying a dietary pattern based on foods traditionally consumed in the country such as rice and beans59.
With regard to composition, despite having different names, the patterns found in the present study were similar to those identified in other studies conducted in Brazil – more specifically in the states of Rio Grande do Sul24, 25, São Paulo26, 27,28, Rio de Janeiro29, Espírito Santo30and Ceará31 – including the ELSA-Brazil study32and National Food Survey33.
Our results show that being female and higher income were directly associated with the Healthy dietary pattern, corroborating the findings of other studies26, 28, 34, 35, 36,37. Women are more aware of and value the relationship between diet and health, and are mainly responsible for cooking38. Data from the VIGITEL survey show that the prevalence of regular consumption of fruits and vegetables is higher among women (39.2%) than men (27.7%)39. In addition, studies have reinforced that people of higher socioeconomic status are better able to purchase food items40.
Being older was inversely associated with the Traditional, Refined Carbs and Sugars, and Fast Food dietary patterns. Authors such as Avelar and Rezende41 have shown that increasing age is associated with being concerned about diet quality. Other studies highlight that the share of energy intake from ultra-processed foods is high among adults and adolescents because they consume large amounts of packaged foods like candies, soft-drinks and biscuits and products rich in sugars, fats and sodium42, 43,44. Data from recent studies in Brazil show that ultra-processed foods account for a large share of calorie intake (49.2%43and 50.6%44) in younger individuals. Similar results were found in Canada (51.2%)45, the United States (57.9%)46 and United Kingdom (50.4%)47, where diets were made up predominantly of industrialized foods.
Few studies have investigated the influence of race/skin color on food consumption. A study comparing the dietary patterns of white and black Americans showed that a dietary pattern based on processed meats, fried foods, refined grains, sugar, margarine, candies and fats was more common among black people48. In Brazil, national surveys showed that being black or brown was associated with higher consumption of beans, meat and milk with high fat content, and lower consumption of fruits and vegetables, indicating that food choices may be associated with price. In a society characterized by racism, as a category of socioeconomic status, race/skin color is a determining factor in individual life trajectories49.
Having a higher level of education was directly associated with the Fast Food dietary pattern, which is consistent with other studies22,24, and inversely associated with the Traditional dietary pattern. In a study with women in the south of Brazil, Lenz et al. found that a consumption pattern associated with high risk of CNCDs, based on candies, biscuits, cheeses, mayonnaise, fried food and other items, was more common among women with a higher level of education22. These findings show that higher level of education is not necessarily associated with healthy food choices. A recent systematic review reinforced the complexity of the relationship between level of education and the dietary patterns of Brazilians, showing that higher level of education was directly associated with a “dual” dietary pattern, with individuals consuming both healthy products and high-sugar and high-fat foods and drinks49.
Income was associated only with higher consumption of the Healthy dietary pattern, being the only socioeconomic variable associated with this pattern after adjustment. The foods that make up the Healthy dietary pattern (fruits, legumes, vegetables and whole-grain cereals) are usually more expensive than ultra-processed foods and the consumption of these foods is higher among individuals with higher income49.The literature shows that individuals of lower socioeconomic status generally live in areas lacking services, hampering access to food purchase locations such as markets, and growers’ markets/fruit and vegetable markets50, 51.
In addition, availability and access to healthy foods is more restricted in vulnerable areas and, when available, these foods tend to be of lower quality and more expensive. On the other hand, exposure to unhealthy foods is greater in these areas, since foods are sold in small establishments and convenience stores50, 51. These findings corroborate the results of the current study, which show that the Traditional and Refined Carbs and Sugars dietary patterns were more common among individuals living in poorer areas.
According to Boyle, Stone-Francisco, and Samuels52and Story et al.53, the food environment can be characterized as follows: the physical presence of a food that affects a person’s diet; a person’s proximity to food store locations; the distribution of food stores, food service, and any physical entity by which food may be obtained; and a connected system that allows access to food. Our findings show that people with high consumption of the Traditional dietary pattern bought fruits and vegetables in supermarkets, markets and/or warehouse stores. In these locations, this type of food is usually more expensive than in growers’ markets/fruit and vegetable markets, meaning that people place priority on basic food items, reducing purchases of fresh fruits and vegetables50, 51. Studies in the United States show that industrialized products like snacks, desserts and soft-drinks are generally purchased in large supermarket chains54, 55. In Brazil, studies suggest that 54% of household food expenditure was in supermarkets or hypermarkets, reaching 67% in the state of Rio Grande do Sul10,13.
The findings show that the choice of forms of transport for food purchases may be related to socioeconomics characteristics, as individuals with high consumption of the Traditional dietary pattern went on foot or by bicycle to buy food, while those with high consumption of the Fast Food dietary pattern used a car/motorbike or public transport.
One of the goals of studies investigating the relationship between food environment and health is the identification of eating places. The results of the present study show that most individuals ate main meals at home and eating lunch or dinner away from home was associated with the Refined Carbs and Sugars and Fast Food patterns. Other studies in Brazil show that eating away from home is associated with an increased share of ultra-processed foods in the diet, as some of the most commonly chosen food items are soft-drinks and fast food snacks56-58.
National survey data demonstrate that spending on eating out as a proportion of overall monthly food spending in urban areas rose from 25.7% in 2002-2003 to 33.9% in 2017-201859, following global trends. The increase in consumption of foods away from home may indicate that structural changes in the economy and society have restricted the amount of time available, consequently increasing the demand for ready-to-eat foods56, 60. The World Health Organization highlights that increased consumption of this type of food is potentially associated with a rise in the incidence CNCDs such as diabetes andhypertension61.
This study has some limitations. First, the use of a retrospective method may have led to recall errors. It is also important to highlight that the method used to assess dietary patterns is limited by the subjectivity of the decisions taken by the researchers. Furthermore, cross-sectional studies are unable to determine temporal relationship between exposure variables and outcomes. The fact that the study was conducted with a population from the catchment area of a primary care unit means that the sample is not representative of the general population. This means that caution should be taken when generalizing the findings to other groups.
However, this study is one of the first to investigate dietary patterns among adults and older persons in the center of big cities, including people living in poor areas.

CONCLUSION
Variables that indicate lower socioeconomic status were associated with higher consumption of less healthy but more affordable dietary patterns (Traditional and/or Refined Carbs and Sugars, which are composed predominantly of ultra-processed foods).
On the other hand, higher socioeconomic status was associated with higher consumption of the Healthy pattern, which is rich in fruits and vegetables, and/or the Fast Food pattern, which is rich in fatty snacks and more expensive foods. It is therefore concluded that people of higher socioeconomic status have the opportunity to choose between healthy and unhealthy dietary patterns, while individuals of lower socioeconomic status are restricted to more affordable, monotonous and generally poor-quality dietary patterns.
The food environment influences the purchase of particular foods and thus dietary patterns. Individuals who adhered to the Traditional dietary pattern purchased food in supermarkets, markets and warehouse stores, and thus bought less fruits, vegetables and legumes, probably because they are more expensive, prioritizing staple foods such as rice and beans. Eating lunch or dinner away from home was associated with the consumption of less healthy dietary patterns (Fast Food and Refined Carbs and Sugars).
The findings of this study can contribute to improving health and food policies and reorganizing the logistic structures of food supply, thus helping to address unfair differences.

REFERENCES
1. Hu FB. Dietary patterns analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol 2002; 13 (1):3-9.

2. Martins APB, Levy RB, Claro RM, Moubarac JC, Monteiro CA. Participação crescente de produtos ultraprocessados na dieta brasileira (1987-2009). Rev Saúde Publica 2013; 47 (4):656-665.

3. Louzada MLC, Martins APB, Canella DS, Baraldi LG, Levy RB, Claro RM, et al. Ultra-processed foods and the nutritional dietary profile in Brazil. Rev Saúde Publica 2015; 49:38.

4. Sichieri R, Castro JFG, Moura AS. Fatores associados ao padrão de consumo alimentar da população brasileira urbana. Cad Saude Publica 2003; 19 (Suppl 1):S47-S53.

5.Krieger JP, Pestoni G, Cabaset S, Brombach C, Sych J, Schader C, et al. Dietary patterns and their sociodemographic and lifestyle determinants in Switzerland: results from the National Nutrition Survey menu CH. Nutrients 2018; 11 (1):62.

6. Chen L, Zhu H, Gutin B, Dong Y. Race, gender, family structure, socioeconomic status, dietary patterns, and cardiovascular health in adolescents. Curr Dev Nutr 2019; 3(11): nzz117.

7. Pérez-Tepayo S, Rodríguez-Ramírez S, Unar-Munguía M, Shamah-Levy T. Trends in the dietary patterns of Mexican adults by sociodemographic characteristics. Nutr J 2020; 19:51.

8. Swinburn B, Sacks G, Vandevijvere S,Kumanyika S, Lobstein T, Neal B, et al. INFORMAS (International Network for Food and Obesity/non-communicable diseases Research, Monitoring and Action Support): overview and key principles. Obes Rer 2013; 14 (Suppl 1):1-12.

9. Backes V. Ambiente alimentar urbano de São Leopoldo: identificação, descrição e relação com a obesidade [tese]. São Leopoldo (RS): Universidade do Vale do Rio dos Sinos, Programa de Pós-Graduação em Saúde Coletiva; 2017.

10. Herforth A, Ahmed S. The food environment, its effects on dietary consumption, and potential for measurement within agriculture-nutrition interventions. Food Sec 2015; 7:505-520.

11. Azevedo SJS. Segregação e oportunidades de acesso aos serviços básicos de saúde em Campinas: vulnerabilidades sociodemográficas no espaço intra-urbano. Campinas (SP): Unicamp, Núcleo de Estudos de População. Textos Nepo 65; 2014.

12. Instituto Brasileiro de Geografia e Estatística. Censo Demográfico 2010. Características da população e dos domicílios: resultados do universo. Rio de Janeiro: IBGE; 2011.

13. PNUD - Programa das Nações Unidas para o Desenvolvimento; IPEA – Instituto de Pesquisa Econômica e Aplicada. Fundação João Pinheiro. Atlas do Desenvolvimento Humano no Brasil. 2013 [acessado em 19 de março de 2020]. Disponível em: http://www.atlasbrasil.org.br/2013

14. Anjos LA, Wahrlich V, Vasconcellos MTL, Souza DR, Olinto MTA, Waissmann W et al. Development of a food frequency questionnaire in a probabilistic sample of adults from Niterói, Rio de Janeiro, Brazil. Cad Saude Publica 2010; 26 (11):2196-2204.

15. Machado FCS, Henn RL, Olinto MTA, Anjos LA, Wahrlich V, Waissmann W. Reprodutibilidade e validade de um questionário de frequência alimentar por grupos de alimentos, em adultos da Região Metropolitana de Porto Alegre, Brasil. Rev Nutr 2012; 25 (1):65-77.

16. Panagiotakos D. α-priori versus α-posterior methods in dietary pattern analysis: a review in nutrition epidemiology. Nutrition Bulletin 2008; 33 (4):311-315.

17. Olinto MTA. Padrões alimentares: análise de componentes principais. In: Kac G, Sichieri R, Gigante DP, organizadores. Epidemiologia nutricional [online]. Rio de Janeiro: Fiocruz/Atheneu; 2007. p. 213-225.
18. Newby PK, Tucker KL. Empirically derived eating patterns using factor or cluster analysis: a review. Nutr Rev 2004; 62(5):177-203.

19. Hoffmann M, Mendes KG, Canuto R, Garcez AS, Theodoro H, Rodrigues AD et al. Padrões alimentares de mulheres no climatério em atendimento ambulatorial no Sul do Brasil. Cienc Saude Colet 2015; 20 (5):1565-1574.

20. Hu FB, Rimm E, Smith-Warner SA, Feskanich D, Stampfer MJ, Ascherio A, et al. Reproducibility and validity of dietary patterns assessed with a food-frequency questionnaire. Am J Clin Nutr 1999;69 (2): 243-249.

21. Hair JF, Anderson RE, Tatham RL, Black WC. Análise de agrupamentos. In: Análise multivariada de dados. Porto Alegre: Bookman; 2005. cap.9. p. 380-419.

22. Lenz A, Olinto MTA, Dias-da-Costa JS, Alves AL, Balbinotti M, Pattussi MP, et al. Socioeconomic, demographic and lifestyle factors associated with dietary patterns of women living in Southern Brazil. Cad Saude Publica 2009; 25 (6):1297-1306.

23. Alves ALS, Olinto MTA, Costa JSD, Bairros FS, Balbinotti MAA. Padrões alimentares de mulheres adultas residentes em área urbana no sul do Brasil. Rev Saude Publica 2006; 40 (5): 865-873.

24. Ternus DL, Henn RL, Bairros F, Costa JS, Olinto MTA. Padrões alimentares e sua associação com fatores sociodemográficos e comportamentais: Pesquisa Saúde da Mulher 2015, São Leopoldo (RS). Rev Bras Epidemiol 2019; 22:e190026.

25. Poltronieri TS, Gregoletto MLO, Cremonese C. Padrões alimentares e fatores associados em docentes de uma instituição privada de ensino superior. Cad Saude Colet 2019; 27 (4):390-403.

26. Gimeno SGA, Mondini L, Moraes SA, Freitas ICM. Padrões de consumo de alimentos e fatores associados em adultos de Ribeirão Preto, São Paulo, Brasil: Projeto OBEDIARP. Cad Saude Publica 2011; 27 (3):533-545.

27. Ferreira-Nunes PM, Papini SJ, Corrente JE. Padrões alimentares e ingestão de nutrientes em idosos: análise com diferentes abordagens metodológicas. Cienc Saude Colet 2018; 23 (12):4085-4094.

28. Arruda SP, da Silva AA, Kac G, Goldani MZ, Bettiol H, Barbieri MA. Socioeconomic and demographic factors are associated with dietary patterns in a cohort of young Brazilian adults. BMC Public Health 2014; 14:654.

29. Aguiar OB, Vasconcelos AGG, Barreiro PLD. Identificação de padrões alimentares: comparação das técnicas de análise de componentes principais e de principais eixos fatoriais. Rev Bras Epidemiol 2019; 22:e190048.

30. Cattafesta M, Zandonade E, Bissoli NS, Salaroli LB. Padrões alimentares de trabalhadores bancários e sua associação com fatores socioeconômicos, comportamentais e laborais. Cienc Saude Colet 2019; 24 (10):3909-3922.

31. Nogueira VC, Arruda SPM, Sampaio HAC, Rodrigues BC, Silva EB, Farias BO, et al. Fatores socioeconômicos, demográficos e de estilo de vida associados a padrões alimentares de trabalhadores em turnos. Cienc Saude Colet 2019; 24 (3):761-769.

32. Cardoso LO, Carvalho MS, Cruz OG, Melere C, Luft VC, Molina MCB, et al. Eating patterns in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil): an exploratory analysis. Cad Saude Publica 2016; 32 (5):e00066215.

33. Massarani FA, Cunha DB, Muraro AP, Souza BSN, Sichieri R, Yokoo EM. Agregação familiar e padrões alimentares na população brasileira. Cad Saude Publica 2015; 31 (12):2535-2545.

34. Olinto MTA, Willett WC, Gigante DP, Victora CG. Sociodemographic and lifestyle characteristics in relation to dietary patterns among young Brazilian adults.Public Health Nutrition 2011; 14 (1):150–159.

35. Assumpção D, Domene SMÁ, Fisberg RM, Canesqui AM, Barros MBA. Diferenças entre homens e mulheres na qualidade da dieta: estudo de base populacional em Campinas, São Paulo. Cienc Saude Colet 2017; 22 (2):347-358.

36. Lins APM, Sichieri R, Coutinho WF, Ramos EG, Peixoto MVM, Fonseca VM. Alimentação saudável, escolaridade e excesso de peso entre mulheres de baixa renda. Cienc Saude Colet 2013; 18 (2):357-366.

37. Figueiredo ICR, Jaime PC, Monteiro CA. Fatores associados ao consumo de frutas, legumes e verduras em adultos da cidade de São Paulo. Rev Saude Publica 2008; 42 (5):777-785.

38. Baker AH, Wardle J. Sex differences in fruit and vegetable intake in older adults. Appetite 2003; 40 (3):269-275.

39. Brasil. Ministério da Saúde (MS). Vigitel Brasil 2018. Vigilância de fatores de risco e proteção para doenças crônicas por inquérito telefônico. Brasília: MS; 2019.

40. Estima CCP, Philippi ST, Alvarenga MS. Fatores determinantes de consumo alimentar: por que os indivíduos comem o que comem? Rev Bras Nutr Clin 2009; 24 (4):263-268.

41. Avelar AE, Rezende DC. Hábitos alimentares fora do lar: um estudo de caso em Lavras-MG. Organizações Rurais & Agroindustriais 2013; 15 (1):137-152.

42. Bielemann RM, Motta JVS, Minten GC, Horta BL, Gigante DP. Consumo de alimentos ultraprocessados e impacto na dieta de adultos jovens. Rev Saude Publica 2015;49:28.

43. D’Avila HF, Kirsten VR. Consumo energético proveniente de alimentos ultraprocessados por adolescentes. Rev Paul Pediatr2017; 35 (1):54-60.

44. Enes CC, Camargo CM, Justino MIC. Ultra-processedfoodconsumptionandobesity in adolescents. Rev Nutr 2019; 32:e180170.

45. Moubarac JC, Batal M, Louzada ML, Martinez Steele E, Monteiro CA. Consumption of ultra-processed foods predicts diet qualityin Canada. Appetite 2017; 108:512-520.

46. Martinez Steele E, Baraldi LG, Louzada MLC, Moubarac JC, Mozaffarian D, Monteiro CA. Ultra-processed foods and added sugars in the US diet: evidence from a nationally representative cross-sectionalstudy. BMJ Open 2016;6 (3):e009892.

47. Monteiro CA, Moubarac JC, Levy RB, Canella DS, Louzana MLC, Cannon G. Household availability of ultra-processed foods and obesity in nineteen European countries. Public Health Nutr 2018; 21 (1):18-26.

48. Judd SE, Gutiérrez OM, Newby PK, Howard G,Howard VJ,Locher JL, et al. Dietary patterns are associated with incident stroke and contribute to excess risk of stroke in Black Americans. Stroke 2013; 44 (12):3305-11.

49. Canuto R, Fanton M, Lira PIC. Iniquidades sociais no consumo alimentar no Brasil: uma revisão crítica dos inquéritos nacionais. Cienc Saude Colet 2019; 24 (9):3193-3212.

50. Filomena S, Scanlin K, Morland KB. Brooklyn, New York foodscape 2007-2011: a five-year analysis of stability in food retail environments. Int J Behav Nutr Phys Act 2013; 10:46.

51. Mook K, Laraia BA, Oddo VM, Jones-Smith JC. Food security status and barriers to fruit and vegetable consumption in two economically deprived communities of Oakland, California, 2013-2014. Prev Chronic Dis 2016; 13:E21.

52. Boyle M, Stone-Francisco, S, Samuels SE. Environmental strategiesand policies to suppor the althy eating and physical activity in low-income communities. Journal of Hunger & Environmental Nutrition 2007; 1 (2): 3-25.

53. Story M, Nanney MS, Schwartz MB. Schools and obesity prevention: creating school environments and policies to promote healthy eating and physical activity. Milbank Q 2009; 87 (1): 71-100.

54. Stern D, Ng SW, Popkin BM. The nutrient content ofU.S. household food purchases by storetype. Am J Prev Med 2016;50 (2):180-190.

55. Taillie LS, Ng SW, Popkin BM. Global growth of "big box" stores and the potential impact on human health and nutrition. NutrRev2016; 74 (2):83-97.

56. Queiroz PWV, Coelho AB. Alimentação fora de casa: uma investigação sobre os determinantes da decisão de consumo dos domicílios brasileiros. Análise Econômica 2017; 35:67-104.

57. Andrade GC. Consumo de alimentos ultraprocessados fora de domicílio no Brasil [dissertação]. São Paulo: Universidade de São Paulo, Faculdade de Medicina; 2017.

58. Bezerra IN, Cavalcante JB, Moreira TMV, Mota CC, Sicheiri R. Alimentação fora de casa e excesso de peso: uma análise dos mecanismos explicativos. Rev Bras Promoç Saude 2016; 29 (3):455-461.

59. Instituto Brasileiro de Geografia e Estatística (IBGE). Pesquisa de orçamentos familiares 2017-2018: primeiros resultados. Rio de Janeiro: IBGE; 2019.

60. Bezerra IN, Sichieri R. Características e gastos com alimentação fora do domicílio no Brasil. Rev Saude Publica 2010; 44 (2):221-229.

61. WHO/FAO. Diet, nutrition and prevention of chronic diseases. Geneva: World Health Organization, 2003.