Environmental, social and individual factors influence eating patterns, which in turn affect the risk of many chronic diseases. This study aimed to estimate associations between environmental factors and the consumption of fruit and vegetables among adults in a Brazilian urban context. Data from the surveillance system for risk factors for chronic diseases (VIGITEL) of Brazilian Ministry of Health were used. A cross-sectional telephone survey (VIGITEL – 2008–2010) was carried out with 5826 adults in the urban area of Belo Horizonte. Individual variables were collected. The frequency of fruit and vegetables consumption was assessed from number of servings, weekly frequency and an intake score was calculated. Georeferenced variables were used to characterize the food environment. The density of healthy food outlets (stores specialized in selling fruit and vegetables), unhealthy food outlets (bars, snack bars and food trucks/trailers) and the neighborhood family income were investigated and associated with fruit and vegetables intake score. Weighted multilevel linear regression was used to evaluate the associations between the environment variables and the fruit and vegetables intake score. Higher fruit and vegetables intake scores were observed in neighborhoods with higher density of healthy food outlets and higher income. Lower scores were observed in neighborhood with higher density of unhealthy food outlets. These associations were adjusted by individual variables such as gender, age, physical activity, sugar sweetened beverages consumption, education level and smoking. The food environment might explain some of the socioeconomic disparities with respect to healthy food intake and health outcomes. Healthy food stores are less common in socially disadvantaged neighborhoods, and therefore, healthy foods such as fruits and vegetables are less available or are of a lower quality in lower income areas. Food environment characteristics and neighborhood socioeconomic level had significant associations with fruit and vegetable intake score. These are initial findings that require further investigation within the middle income world populations and the role of the environment with respect to both healthy and unhealthy food acquisition and intake.

Background: Food deserts are neighborhoods with little or no access to healthy food, whereas food swamps are neighborhoods where unhealthy food options prevail over healthy ones. The main aims of the current study are to feature and compare the neighborhoods of food deserts and food swamps based on social inequality. Methods: Ecological study carried out in Belo Horizonte City, Minas Gerais State, Brazil. Information about commercial food establishments derived from two different databases. It was measured by secondary governmental databases, which were virtually conferred in the present study. Census tracts were considered as analysis units and classified as food deserts and food swamps, based on the Brazilian methodology. Take into consideration the density of establishments classified as selling fresh or minimally-processed food, mixed establishments, and establishments selling ultra-processed food. The Brazilian methodology evaluates food deserts by the density of healthy establishments (establishments classified as mostly selling fresh or minimally-processed food and mixed establishments) per 10 thousand inhabitants. And the metric to evaluate food swamps considers the density of unhealthy establishments (establishments mostly selling ultra-processed food) per 10 thousand inhabitants. Information about social inequalities comprised aspects such as income, population count, number of households, number of literate individuals, race, water and energy supply, and garbage collection. The Health Vulnerability Index (HVI) was used as a synthetic social vulnerability indicator. Results: Neighborhoods of food deserts presented worse essential service availability, lower income per capita, and smaller mean number of literate individuals. Census tracts classified as food swamps presented better socio-demographic conditions than those areas food deserts. Neighborhoods simultaneously classified as food deserts and food swamps presented lower income per capita and were more often observed in census sectors presenting medium and high HVI. Conclusion: The food environment in Belo Horizonte was featured by the strong presence of food deserts and food swamps. However, the potential influence of these areas on food intake has changed depending on social inequalities.

O objetivo foi avaliar o ambiente alimentar comunitário e a existência de pântanos alimentares no entorno das escolas de uma metrópole brasileira. Trata-se de um estudo ecológico realizado em escolas públicas e privadas de Belo Horizonte, Minas Gerais, Brasil, e teve como unidade de análise o buffer de 250m. Foram incluídas no estudo todas as escolas de Educação Infantil, de Ensinos Fundamental e Médio. Em relação às escolas, foram avaliadas a dependência administrativa e o tipo de ensino ofertado, bem como a renda per capita dos setores censitários das escolas. Contabilizou-se também as informações sobre os estabelecimentos de venda de alimentos para o consumo imediato que estavam dentro do buffer no entorno das escolas. Em relação ao ambiente alimentar foram avaliados apenas os estabelecimentos que comercializam alimentos para o consumo imediato no entorno escolar. A análise dos buffers revelou que 97,4% das escolas tinham ao menos um desses estabelecimentos no seu entorno. Os estabelecimentos mais disponíveis no entorno da escola foram lanchonetes, restaurantes e bares. As escolas localizadas em setores censitários de maior renda apresentavam maior média de todos os estabelecimentos no seu entorno, exceto das mercearias e supermercados. Ademais, 54,6% das escolas estavam em vizinhanças que são classificados como pântanos alimentares. Os resultados revelam que entre as categorias avaliadas ocorre um predomínio dos estabelecimentos que comercializam, predominantemente, alimentos ultraprocessados, como os bares e lanchonetes no entorno das escolas de Belo Horizonte, o que expõe as crianças e os adolescentes a um ambiente alimentar não saudável.

Objective. This study identified environmental variables associated with obesity in the adult population of a city in Brazil. Methods. It was conducted using the Surveillance System for Risk and Protective Factors for Chronic Diseases by Telephone Survey from 2008 to 2010. The body mass index (BMI) was calculated from the participants' self-reported weight and height. Obesity was defined as a BMI ≥ 30 kg/m2. The food establishments, georeferenced areas conducive to physical activity, total income of the neighbourhood, homicide rate and population density were used to characterise the environment. In addition, individual variables were considered. A multilevel logistic regression was performed. Results. A total of 5273 individuals were evaluated. The odds of obesity was found to be significantly decreased with increases in the number of establishments that sell healthy food, number of restaurants, number of places for physical activity and total income - in different models. In addition, these associations remained significant after adjustment for age, gender, education and consumption of meat with visible fat. Conclusions. This study contributes to a better understanding of the complex interaction between environmental and individual determinants of obesity and may aid in the development of effective interventions, such as the expansion of obesity control programmes.

Objective: The purpose of this ecological study was to characterize the community food environment according to the socioeconomic condition of census tracts (CTs) in the urban area of a medium-sized city of southeastern Brazil in 2016. Method: Food establishments were identified on the streets covered by raters and information about type was collected through objective assessment. Geocoding was carried out from address observed by raters. Food establishments were categorized into establishments with predominant sale of natural or minimally processed foods, mixed establishments, and establishments with predominant sale of ultra-processed foods. The distribution of the number of establishments, by category, was evaluated according to tertiles of per capita income of the CT. The kernel estimation was used to analyze the density of establishments by category. The spatial pattern of the categories of establishments was investigated using the univariate Ripley’s K-function. Results: A total of 656 establishments were evaluated. In all, 11.1% had predominant sale of natural or minimally processed foods, 44.5% were mixed, and 44.4% had predominant sale of ultra-processed foods. The average of establishments with predominant sale of natural or minimally processed foods, of ultra-processed foods, and all categories increased according to the income of the CT. There was a clustering of all categories of establishments in high-income CTs downtown. However, peripheral and low-income CTs were composed of a higher number of mixed establishments or those with predominant sale of ultra-processed foods than establishments with predominant sale of natural or minimally processed foods. Conclusions: On average, the number of all categories of establishments increased according to the per capita income of the CT and were clustered in central and higher-income regions of the city. These findings may have practical implications for the development of public policies to increase the availability of healthy foods and to reduce the sale of unhealthy foods.