Abstract

The aim of the current ecological study is to assess the evolution of food deserts and food swamps in the metropolitan city of Belo Horizonte between 2008 and 2020. Food deserts were determined based on the density of healthy establishments per 10,000 inhabitants, whereas food swamps were set based on the density of ultra-processed food procurement establishments per 10,000 inhabitants. The rate of census tracts classified as food deserts has decreased between 2008 and 2020, whereas that of census tracts classified as food swamps has increased within this same period. Furthermore, despite the reduced number of food deserts, these areas have increased in census tracts living under lower socioeconomic vulnerability condition. Food swamps recorded sharp increase in census tracts living under higher vulnerability condition. The population living in the herein investigated city has been increasingly exposed to an unhealthy community food environment over 12 years. Monitoring changes in community food environment is key strategy to enable tracking the effectiveness and efficiency of actions taken in food environments to ensure the human right to adequate food.

Key words:
Food desert; Food supply; Time series studies

Resumo

O estudo tem por objetivo avaliar a evolução dos desertos e pântanos alimentares entre 2008 e 2020. Os desertos alimentares foram determinados pela densidade de estabelecimentos saudáveis por 10 mil habitantes, e os pântanos alimentares pela densidade de estabelecimentos de aquisição de alimentos ultraprocessados por 10 mil habitantes. O estudo identificou que, entre 2008 e 2020, diminuiu o percentual de setores censitários classificados como desertos alimentares e aumentou o de setores censitários classificados como pântanos alimentares. Além disso, observou-se que, apesar da redução no número de desertos alimentares, essas áreas tiveram aumento em setores censitários de menor vulnerabilidade socioeconômica. E os pântanos alimentares apresentaram aumento acentuado em setores censitários de maior vulnerabilidade. Ao longo de 12 anos, a população está cada vez mais exposta a um ambiente alimentar comunitário pouco saudável. O monitoramento das mudanças no ambiente alimentar comunitário é uma estratégia primordial que possibilita acompanhar a efetividade e eficácia das ações realizadas no ambiente alimentar para a garantia do direito humano à alimentação adequada.

Palavras-chave:
Desertos alimentares; Abastecimento de alimentos; Estudo de séries temporais

Introduction

Eating habits can be influenced by different factors, such as cultural and social issues, and food environment features. Food environment is part of the study field focused on assessing population’s access to food, and it is done by taking into consideration physical, economic and sociocultural dimensions¹1 Swinburn B, Sacks G, Vandevijvere S, Kumanyika S, Lobstein T, Neal B, Barquera S, Friel S, Hawkes C, Kelly B, L'abbé M, Lee A, Ma J, Macmullan J, Mohan S, Monteiro C, Rayner M, Sanders D, Snowdon W, Walker C; INFORMAS. INFORMAS (International Network for Food and Obesity/non-communicable diseases Research, Monitoring and Action Support): overview and key principles. Obes Rev 2013;14(Suppl. 1):1-12.. Different dimensions can be assessed at the time to investigate individuals’ access to food, namely: availability of and accessibility to establishments, food quality, food price, convenience and desirability²2 Turner C, Aggarwal A, Walls H, Herforth A, Drewnowski A, Coates J, Kalamatianou S, Kadiyala S. Concepts and critical perspectives for food environment research: a global framework with implications for action in low- and middle-income countries. Glob Food Security 2018; 18:93-101.^,33 Downs SM, Ahmed S, Fanzo J, Herforth A. Food environment typology: advancing an expanded definition, framework, and methodological approach for improved characterization of wild, cultivated, and built food environments toward sustainable diets. Foods 2020; 9(4):532..

Food environment has been the topic of discussions about public policies focused on food and nutrition over the years. Using metaphors to describe food environment features is one of the strategies used by government agencies to address this topic⁴4 Acheson D. Inequalities in health. BMJ 1998; 317(7173):1659.

5 United States Department of Agriculture. Access to affordable and nutritious food: measuring and understanding food deserts and their consequences [Internet]. 2009. [cited 2022 out 13]. Available from: https://www.ers.usda.gov/webdocs/publications/ 42711/12716_ap036_1_.pdf
https://www.ers.usda.gov/webdocs/publica...

6 Centers for Disease Control and Prevention. Census tract level state maps of the modified retail food environment index (mRFEI) [Internet]. 2011. [cited 2022 out 13]. Available from: https://stacks.cdc.gov/view/cdc/61367
https://stacks.cdc.gov/view/cdc/61367...

7 Johns Hopkins Center for a Livable Future. Mapping Baltimore City's Food environment: 2015 Report [Internet]. 2015. [cited 2022 out 13]. Available from: https://clf.jhsph.edu/sites/default/files/2019-02/Baltimore-Food-Environment-Report-20 15-1.pdf
https://clf.jhsph.edu/sites/default/file... ^-88 Câmara Interministerial de Segurança Alimentar e Nutricional . Mapeamento dos desertos alimentares no Brasil [Internet]. 2018. [acessado 2023 jun 11]. Disponível em: https://aplicacoes.mds.gov.br/sagirmps/noticias/arquivos/files/Estudo_tecnico_mapeamento_desertos_alimentares.pdf
https://aplicacoes.mds.gov.br/sagirmps/n... . Food deserts and food swamps are the mostly used metaphors to characterize the food environment. Discussions about food deserts were observed for the first time in a governmental document addressing health inequities in the UK, in 1998⁴4 Acheson D. Inequalities in health. BMJ 1998; 317(7173):1659.. The approach to food swamps has gained momentum in the United States due to a document issued by the Centers for Disease Control and Prevention, back in 2011⁶6 Centers for Disease Control and Prevention. Census tract level state maps of the modified retail food environment index (mRFEI) [Internet]. 2011. [cited 2022 out 13]. Available from: https://stacks.cdc.gov/view/cdc/61367
https://stacks.cdc.gov/view/cdc/61367... .

The term “food desert” is used to describe socioeconomically vulnerable areas with limited access to healthy food⁹9 Cummins S, Macintyre S. "Food deserts" - evidence and assumption in health policy making. BMJ 2002; 325(7361):436-438.. On the other hand, the term “food swamp” is linked to areas with high availability of unhealthy food types¹⁰10 Rose D, Bodor JN, Swalm CM, Rice JC, Farley TA, Hutchinson PL. Deserts in New Orleans? Illustrations of urban food access and implications for policy [Internet]. 2009. [cited 2022 out 13]. Available from: https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=abc8b418aa0783c8f3b0a0c4fca8f137ad806e0a
https://citeseerx.ist.psu.edu/document?r... .

Studies about food deserts and food swamps have advanced, both in conceptual issues and in assessment measures, overtime. However, the evolution of these areas over the years has been poorly explored in the literature¹¹11 Larsen K, Gilliland J. Mapping the evolution of "food deserts" in a Canadian city: supermarket accessibility in London, Ontario, 1961-2005. Int J Health Geogr 2008; 7(1):16.

12 James P, Seward MW, James O'Malley A, Subramanian S, Block JP. Changes in the food environment over time: examining 40 years of data in the Framingham Heart Study. Int J Behav Nutr Phys Act 2017; 14(1):84.

13 Hobbs M, Mackenbach JD, Wiki J, Marek L, McLeod GFH, Boden JM. Investigating change in the food environment over 10 years in urban New Zealand: a longitudinal and nationwide geospatial study. Soc Sci Med 2021; 269:113522.^-1414 Justiniano ICS, Menezes MC, Mendes LL, Pessoa MC. Retail food environment in a Brazilian metropolis over the course of a decade: evidence of restricted availability of healthy foods. Public Health Nutr 2022; 25(9):2584-2592.. Thus, the aim of the present study was to investigate the evolution of food deserts and food swamps in a Brazilian metropolis.

Methods

Study design and site

The present research is an ecologically designed study, based on secondary data about Belo Horizonte City, capital of Minas Gerais State. Belo Horizonte is the sixth most populous city in Brazil; its estimated population reached 2,530,701 inhabitants in 2021. Formal workers living in the aforementioned city earned 3.4 minimum wages (U$ 684.59) in 2020, on average, and its Municipal Human Development Index (HDI) reached 0.810 (https://cidades.ibge.gov.br/brasil/mg/belo-horizonte/panorama).

Analysis unit

Census tracts were herein adopted as analysis unit. According to the 2010 census, Belo Horizonte City has 3,936 census tracts¹⁵15 Instituto Brasileiro de Geografia e Estatística (IBGE). Censo 2010 - resultados [Internet]. 2011. [acessado 2021 dez 11]. Disponível em: https://censo2010.ibge.gov.br/resultados.html
https://censo2010.ibge.gov.br/resultados... . Uninhabited census tracts were excluded from the current study; based on the 2010 Census data, they corresponded to 106 sectors. Therefore, 3,830 census tracts were included in the analysis (Figure 1).

Social environment

Some sociodemographic variables were extracted from the 2010 Demographic Census to explain social features of census tracts classified as food deserts and swamps. These variables encompassed number of literate individuals, number of households and number of individuals declaring themselves as black and brown. In addition, income per capita was calculated based on information about total income and total population living in the census sector¹⁵15 Instituto Brasileiro de Geografia e Estatística (IBGE). Censo 2010 - resultados [Internet]. 2011. [acessado 2021 dez 11]. Disponível em: https://censo2010.ibge.gov.br/resultados.html
https://censo2010.ibge.gov.br/resultados... .

Health Vulnerability Index (HVI) was another variable included in the study. This composite indicator was developed by Belo Horizonte City Hall. It is used as tool to determine the vulnerability of the city’s census tracts. HVI comprises variables such as sanitation, housing, education, income, social issues and the surrounding areas of each census sector. Data composing the HVI derived from the Demographic Census¹⁶16 Prefeitura de Belo Horizonte. Índice de Vulnerabilidade da Saúde 2012 [Internet]. 2013. [acessado 2021 dez 11]. Disponível em: https://prefeitura.pbh.gov.br/sites/default/files/estrutura-de-governo/saude/2018/publicacaoes-da-vigilancia-em-saude/indice_vulnerabilidade2012.pdf
https://prefeitura.pbh.gov.br/sites/defa... . This indicator was categorized as low, medium or high vulnerability.

Food deserts

Food deserts were defined based on the density of healthy food establishments per 10,000 inhabitants⁸8 Câmara Interministerial de Segurança Alimentar e Nutricional . Mapeamento dos desertos alimentares no Brasil [Internet]. 2018. [acessado 2023 jun 11]. Disponível em: https://aplicacoes.mds.gov.br/sagirmps/noticias/arquivos/files/Estudo_tecnico_mapeamento_desertos_alimentares.pdf
https://aplicacoes.mds.gov.br/sagirmps/n... . The following types were taken as healthy food establishments: butcher shops, delivery stores, street vendors, hypermarkets, fruit and vegetable stores, dairy shops, mini-markets, bakeries, fishmongers, restaurants and supermarkets. Census tracts below the 25th percentile of healthy food establishment-density distribution⁸8 Câmara Interministerial de Segurança Alimentar e Nutricional . Mapeamento dos desertos alimentares no Brasil [Internet]. 2018. [acessado 2023 jun 11]. Disponível em: https://aplicacoes.mds.gov.br/sagirmps/noticias/arquivos/files/Estudo_tecnico_mapeamento_desertos_alimentares.pdf
https://aplicacoes.mds.gov.br/sagirmps/n... were classified as food deserts.

Food swamps

Food swamps were defined based on the density of unhealthy food establishments per 10,000 inhabitants¹⁷17 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707.. Bars, snack bars and candy stores were classified as unhealthy food establishments⁸8 Câmara Interministerial de Segurança Alimentar e Nutricional . Mapeamento dos desertos alimentares no Brasil [Internet]. 2018. [acessado 2023 jun 11]. Disponível em: https://aplicacoes.mds.gov.br/sagirmps/noticias/arquivos/files/Estudo_tecnico_mapeamento_desertos_alimentares.pdf
https://aplicacoes.mds.gov.br/sagirmps/n... . The cut-off point used to classify the investigated sectors as food swamps lied on density above the 25th percentile of unhealthy food establishment-density distribution¹⁷17 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707..

Data analysis

Community food environment features were expressed as mean and standard deviation, based on the following categories: (1) healthy food establishments, (2) unhealthy food establishments and (3) total number of establishments. Secondary data from the State Farm Bureau, from 2008, 2011, 2015, 2018 and 2020, were used to assess food environment. It is worth emphasizing that dates representing important economic and health milestones were included in the current study. The year of 2008 was included in the study because it corresponds to the two-year period after the implementation of the National Health Promotion Policy (PNSP). One of the several PNPS’ goals highlighted the importance of including urban and rural planning in public policies aimed at improving individuals’ quality of life¹⁸18 Brasil. Ministério da Saúde (MS). Política nacional de promoção da saúde. Brasília: MS; 2006..

The year of 2015 marked the economic crisis that has forced the country to experience strong recession and negative economic growth ¹⁹19 Paula LF, Pires M. Crise e perspectivas para a economia brasileira. Estud Av 2017; 31(89):125-144.. The year of 2020, in its turn, marked the first COVID-19 pandemic year. The aforementioned health crisis has worsened the economic crisis that had started in 2019, when the country’s Gross Domestic Product (GDP) recorded significant decrease²⁰20 Granemann S. Crise econômica e a Covid-19: rebatimentos na vida (e morte) da classe trabalhadora brasileira. Trab Educ Saude 2020; 19:e00305137..

Data on company’s name, address and National Classification of Economic Activities (CNAE) were requested to the State Department of Finance. The investigated establishments were georeferenced based on their full address. This information was used to find their latitude and longitude. Comparison of means between years was carried out through Anova test for repeated measures, with Sidak correction.

Food deserts and swamps were described based on absolute and relative frequency; chi-square test was used to compare means. In addition, mean variation in these parameters over the 12-year assessed period was calculated, based on variations observed between assessed periods and on averaging that variation. Data on food deserts and swamps were stratified based on the HVI of the investigated census tracts.

Furthermore, descriptive analysis was applied to socioeconomic features of the population living in census tracts classified as food deserts and swamps. Results were expressed as mean and standard deviation. Subsequently, the mean rate of variations observed over 12 years (2008-2020) was calculated based on the following formula:

Analyses were carried out in SPSS 19.0 statistical software and in QGIS 2.14.3 software, at 5% significance level (p value = 0.05).

Results

Table 1 describes the overall features of the community food environment observed in the investigated census tracts. There was significant increase in the mean number of total establishments in each census sector between 2008 and 2020 (p < 0.0001), as well as increase in the number of places categorized as healthy food establishments (p < 0.0001). The number of unhealthy food establishments did not show significant difference between 2015 and 2020 (Table 1).The rate of census tracts classified as food deserts has decreased between 2008 and 2020, whereas the rate of census tracts classified as food swamps has increased within this same period (Figure 2).

There was mean variation of -3.30% in the rate of census tracts classified as food deserts. The period between 2011 and 2015 was the one recording the most significant reduction in this parameter. However, there was increase in the rate of low-vulnerability census tracts; mean variation reached 4.54%. Mean variation in the rate of census tracts classified as food swamps reached 8.44%. The period between 2011 and 2015 was the one accounting for the highest increase in this parameter. However, there was reduction in low vulnerability census tracts classified as food swamps; mean variation reached -4.49% (Table 2).

Table 3 describes sociodemographic features in census tracts classified as food deserts and swamps. There was increase in mean income per capita recorded for census tracts classified as food deserts, between 2008 and 2020 - variation rate reached 52.35%. In addition, there was decrease in the mean number of individuals who declared themselves black or brown in census tracts classified as food deserts - variation rate reached -13.22%. On the other hand, census tracts classified as food swamps presented opposite behavior, i.e., decrease in mean income per capita (variation rate reached -26.71%) and increase in mean number of individuals who declared themselves black or brown (variation rate reached 19.26%).

Discussion

Community food environment

There was significant increase in the number of food-selling establishments in Belo Horizonte City, over 12 years. The number of both healthy and unhealthy food establishments recorded significant increase. Previous study focused on assessing the evolution of food establishments in the investigated city between 2008 and 2018 had already evidenced the trend of increased the number of the aforementioned establishments¹⁴14 Justiniano ICS, Menezes MC, Mendes LL, Pessoa MC. Retail food environment in a Brazilian metropolis over the course of a decade: evidence of restricted availability of healthy foods. Public Health Nutr 2022; 25(9):2584-2592..

A national study, have pointed out the trend of community food environment to present worsened features over 10 years. The aforementioned authors reported increase by 154% in the number of unhealthy food establishments over the investigated period-of-time¹⁴14 Justiniano ICS, Menezes MC, Mendes LL, Pessoa MC. Retail food environment in a Brazilian metropolis over the course of a decade: evidence of restricted availability of healthy foods. Public Health Nutr 2022; 25(9):2584-2592..

In addition, other studies conducted in Brazil have already shown higher prevalence of unhealthy food procurement establishments in more urbanized regions, such as metropolises⁸8 Câmara Interministerial de Segurança Alimentar e Nutricional . Mapeamento dos desertos alimentares no Brasil [Internet]. 2018. [acessado 2023 jun 11]. Disponível em: https://aplicacoes.mds.gov.br/sagirmps/noticias/arquivos/files/Estudo_tecnico_mapeamento_desertos_alimentares.pdf
https://aplicacoes.mds.gov.br/sagirmps/n... ^,1717 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707.^,2121 Castro Junior PCP. Ambiente alimentar comunitário medido e percebido: descrição e associação com índice de massa corporal de adultos brasileiros [tese]. Rio de Janeiro: ENSP; 2018.. It is also noteworthy that previous studies conducted in middle- and low-income countries have shown that the sociodemographic features of a given investigated area can influence the availability of establishments; thus, areas presenting greater socioeconomic vulnerability have lower availability of all establishment types¹⁴14 Justiniano ICS, Menezes MC, Mendes LL, Pessoa MC. Retail food environment in a Brazilian metropolis over the course of a decade: evidence of restricted availability of healthy foods. Public Health Nutr 2022; 25(9):2584-2592.^,1717 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707..

A cohort study conducted in a high-income country has shown increased density of convenience stores, fast food and full-service restaurants, as well as decreased density of supermarkets, between 1971 and 2008¹²12 James P, Seward MW, James O'Malley A, Subramanian S, Block JP. Changes in the food environment over time: examining 40 years of data in the Framingham Heart Study. Int J Behav Nutr Phys Act 2017; 14(1):84.. There was increase in the number of food establishments, as well as in food and beverage availability, in New York City, between 2010 and 2015. In addition, a study pointed out increased food and beverage trade in establishments that did not have the trade of these products as end activity²²22 Lucan SC, Maroko AR, Patel AN, Gjonbalaj I, Abrams C, Rettig S, Elbel B, Schechter CB. Change in an urban food environment: storefront sources of food/drink increasing over time, and not limited to 'food stores' and restaurants. J Acad Nutr Diet 2018; 118(11):2128-2134..

Based on another study conducted in New Zealand, there was increase in the number of fast-food outlets and supermarkets within a decade (2005-2015). The authors of the aforementioned study have also observed that the distance traveled by car to fast-food stores and supermarkets has reduced, even in the most vulnerable regions¹³13 Hobbs M, Mackenbach JD, Wiki J, Marek L, McLeod GFH, Boden JM. Investigating change in the food environment over 10 years in urban New Zealand: a longitudinal and nationwide geospatial study. Soc Sci Med 2021; 269:113522..

Food deserts

Findings regarding food deserts have shown decreased rate of census tracts presenting these features in the herein assessed 12-year period. This finding contradicts what international studies have been showing, i.e., that the number of food deserts shows stability²³23 Smets V, Cant J, Vandevijvere S. The changing landscape of food deserts and swamps over more than a decade in Flanders, Belgium. Int J Environ Res Public Health 2022; 19(21):13854. or increases number of food deserts²⁴24 Semple H, Giguere A. The evolution of food deserts in a small midwestern city: the case of Ypsilanti, Michigan: 1970 to 2010. J Plann Edu Res 2018; 38(3):359-370.. However, it is worth emphasizing that these studies considered supermarkets as marker to determine food deserts, and that they were conducted in high-income countries.

There is evidence that using supermarkets as the only marker of healthy food availability is not the most appropriate strategy for middle-income countries¹⁷17 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707.^,2525 Wagner J, Hinton L, McCordic C, Owuor S, Capron G, Arellano SG. Do urban food deserts exist in the Global South? An analysis of Nairobi and Mexico City. Sustainability 2019; 11(7):1963.. A study conducted in Mexico has shown that families living in food deserts (defined exclusively based on supermarkets’ availability) had access to multiple food retailers²⁵25 Wagner J, Hinton L, McCordic C, Owuor S, Capron G, Arellano SG. Do urban food deserts exist in the Global South? An analysis of Nairobi and Mexico City. Sustainability 2019; 11(7):1963.. A study conducted in Brazil has emphasized the need of taking into consideration the specificities of each location at the time to design and propose ways to measure food deserts¹⁷17 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707..

Furthermore, the present study has evidenced change in the socioeconomic features of census tracts classified as food deserts. The social pattern of food deserts has improved over the years, since food deserts are more often observed in lower-vulnerability census tracts.

A study conducted in a high-income country focused on assessing the evolution of food deserts between 1970 and 2010. Results have evidenced changes in the socio-demographic features of the population living in food deserts over the years. The black population accounted for the highest rate of individuals living in food deserts, overtime. Furthermore, household income in food desert areas was lower in all analyzed periods; consequently, the population living below the poverty line was larger in these areas²⁴24 Semple H, Giguere A. The evolution of food deserts in a small midwestern city: the case of Ypsilanti, Michigan: 1970 to 2010. J Plann Edu Res 2018; 38(3):359-370..

The literature on food deserts has evidences that reinforce the need of always including social issues at the time to assess the environments. Some measurement procedures already use social indicators to assess food deserts²⁶26 Dutko P, Ploeg MV, Farrigan T. Characteristics and influential factors of food deserts [Internet]. 2012. [cited 2022 jun 11]. Available from: https://www.ers.usda.gov/webdocs/publications/45014/30940_err140.pdf
https://www.ers.usda.gov/webdocs/publica... . Social changes observed in food deserts may be associated with changes in establishments’ concentration location. There is evidence in the literature about changes in establishments’ location. Semple and Giguere (2018) assessed the layout of food establishments over the years; they observed that these establishments were more concentrated in the downtown area of the investigated city at the beginning of the evaluation period, but they migrated to the most outlying areas over time²⁴24 Semple H, Giguere A. The evolution of food deserts in a small midwestern city: the case of Ypsilanti, Michigan: 1970 to 2010. J Plann Edu Res 2018; 38(3):359-370..

Food swamps

There was increase in the rate of census tracts classified as food swamps, over the years. These findings are similar to the ones observed in a study conducted in Belgium, which presented increase in the number of food swamp areas between 2008 and 2020, despite the different features presented by the two countries²³23 Smets V, Cant J, Vandevijvere S. The changing landscape of food deserts and swamps over more than a decade in Flanders, Belgium. Int J Environ Res Public Health 2022; 19(21):13854..

Unhealthy food trading presents ascending curve in several Latin American countries presenting community food environment features close to the Brazilian ones²⁷27 Pan American Health Organization (PAHO). Ultra-processed food and drink products in Latin America: Sales, sources, nutrient profiles, and policy implications [Internet]. 2019. [cited 2022 out 31]. Available from: https://iris.paho.org/handle/10665.2/51094
https://iris.paho.org/handle/10665.2/510... . Moreover, it is importante emphasizing the lower price and longer lifespan of this food type. The demand for cheaper and more durable food grows in times of both economic and sanitary crises²⁸28 Mendes LL, Canella DS, Araújo ML, Jardim MZ, Cardoso LO, Pessoa MC. Food environments and the COVID-19 pandemic in Brazil: analysis of changes observed in 2020. Public Health Nutr 2022; 25(1):32-35..

Furthermore, the rate of food swamps in census tracts presenting higher socioeconomic vulnerability has increased overtime. It is importante highlighting that food swamps are still more often observed in less vulnerable areas.

There is no consensus in the literature about correlation between social features and incidence of food swamps. There is evidence that census tracts classified as food swamps present more favorable socioeconomic conditions than those classified as food deserts, in Brazil¹⁷17 Honório OS, Horta PM, Pessoa MC, Jardim MZ, Carmo AS, Mendes LL. Food deserts and food swamps in a Brazilian metropolis: comparison of methods to evaluate the community food environment in Belo Horizonte. Food Sec 2021; 14:605-707..

Final considerations

The present study has some limitations. The first one refers to the use of secondary data, which only enables making inferences about the Evolution of community food environments registered in inspection agencies. Nevertheless, it is essential emphasizing that knowledge about formal food environments can be used as important indicator of community food environment features. The second limitation refers to the use of socioeconomic data from 2010 as vulnerability indicator.

We are aware that the herein used socioeconomic information may be obsolete, but these data comprise the latest official records available in Brazil for the whole investigated population. On the other hand, the current study presentes some strong points, since it was one of the first studies focused on assessing the evolution of the so-called food desert and swamp areas in a major Brazilian metropolis. Furthermore, it is noteworthy that two different periods-of-time (economic crisis - 201519; COVID-19 pandemic - 202020) accounting for substantial impact on Brazilian citizens’ lives were included in the study.

There is evidence that infrastructure inequalities are mainly associated with experienced historical moments. These moments trigger changes that range from adaptation to innovation processes taking place in all regions of the city²⁹29 Deener A. The origins of the food desert: urban inequality as infrastructural exclusion. Soc Forces 2017; 95(3):1285-1309.. It is therefore food environment analyses must include different points in time in order to capture the influence of importante historical moments changes in the food environment.

The first year of the pandemic was included in the current study due to its impact on the daily lives of the world’s population. This period representes the time when measures to stop the pandemic were most rigorously applied. Some food retailing areas, such as food-selling apps, have boosted during the first pandemic year³⁰30 Horta PM, Souza JPM, Rocha LL, Mendes LL. Digital food environment of a Brazilian metropolis: food availability and marketing strategies used by delivery apps. Public Health Nutr 2021; 24(3):544-548.. On the other hand, some food-selling establishment categories, mainly the ones selling ready-to-eat food, were affected by measures taken to stop the pandemic³¹31 Oliveira TC, Abranches MV, Lana RM. Food (in) security in Brazil in the context of the SARS-CoV-2 pandemic. Cad Saude Publica 2020; 36(4):e00055220.^,3232 Martinelli SS, Cavalli SB, Fabri RK, Veiros MB, Reis ABC, Amparo-Santos L. Strategies for the promotion of healthy, adequate and sustainable food in Brazil in times of Covid-19. Rev Nutr 2020; 33:e200181..

The current findings enabled observing increased availability of establishments selling unhealthy foods throughout the assessed period, a fact that rose the number of food swamps. On the other hand, the number of food desert areas has decreased within the same period. However, decreased number of food deserts does not mean increased access to healthy food.

Findings in the current study reinforced the need of developing public and community actions focused on increasing fresh food availability in residential areas, mainly in the lower income ones. In addition to increase healthy food availability, initiatives to provide access to healthy food to the most vulnerable populations should be promoted by civil society, as well as by the government and private institutions.

One strategy to increase access to healthy food would be to expand the availability of Public Food and Nutrition Security Facilities (EPSAN). Examples of EPSANs include: popular restaurants, agro-ecological and organic producers’ fairs. In addition, strategies such as food banks can also contribute to improving access to healthy food. Another important initiative to increase the availability of healthy food is urban agriculture, such as the production units in Belo Horizonte³³33 Araújo ML de, Pessoa MC, Honório OS, Schubert MN, Schneider S, Grisa C. Dinâmicas de abastecimento nos sistemas alimentares em Belo Horizonte. Confins 2023; 59. DOI: https://doi.org/10.4000/confins.52625
https://doi.org/10.4000/confins.52625... .

On the other hand, promoting changes in food swamp areas requires significant efforts, since changing this environment type demands multidimensional interventions. Changing food swamps requires public regulatory policies focused on limiting the trade of unhealthy food in certain areas or on prohibiting the opening of unhealthy food establishments.

The findings here show that, in the long term, the food environment in cities will increasingly be made up of establishments selling unhealthy food. And that to reverse this scenario, greater action by the public authorities is needed, since reducing the availability of unhealthy food requires regulatory action. For example, an important step to advance regulation would be to approve the taxation of ultra-processed foods.

Furthermore, it is note that for public policies to shape the food system and food environment, long-term action is needed. Another point is that food policies regulating are heavily influenced by food commodities, which often makes such legislations limited.

Acknowledgements

This work was supported by the Coordenação para o Aperfeiçoamento do Pessoal de Ensino Superior - Brazil (CAPES) and Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG). Productivity scholarship granted by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ) to researcher Larissa Loures Mendes (312979/2021-5).

Referencess

Publication Dates

History