The community food environment as an effect modifier of the relationship between racial discrimination and food insecurity among adults in Southern Brazil

Abstract

Background

Racial discrimination is linked to unhealthy food environments and a higher prevalence of food insecurity. However, no study has explored their interrelated effects. We analyzed the relationship between racial discrimination, community food environment, and food insecurity in adults of different socioeconomic status. We also investigated the potential modifying effect of the food environment on the relationship between racism and food insecurity.

Methods

This was a cross-sectional study of 400 adults aged 20–70 years residing in the central area of ​​Porto Alegre, the capital of Rio Grande do Sul state. Race and racial discrimination were assessed by self-reported race/skin color using the Experiences of Discrimination scale (EOD), respectively. The food environment was assessed using the Nutrition Environment Measures Survey in Stores (NEMS-S) tool. Food insecurity was assessed using the short version of the Brazilian Food Insecurity Scale (EBIA for short, in Portuguese). Poisson regression with robust variance was employed for the multivariate analysis.

Results

The prevalence of food insecurity was higher in areas with a poorer food environment (areas 1 and 3; 56.6% and 58.8%, respectively). Racial discrimination was associated with food insecurity, where every 1-point increase in the racial discrimination score increased the likelihood of food insecurity by 7% (prevalence ratio [PR] 1.07; 95% CI, 1.03–1.20). When stratifying the analyses by food environment, racial discrimination was associated with food insecurity only in areas with a poorer food environment (PR 1.06; 95% CI, 1.01–1.10).

Conclusions

Experiences of racial discrimination were associated with a higher prevalence of food insecurity in the study population. The community food environment was an effect modifier of this relationship, highlighting the relevance of interventions in the food environment focused on areas with a greater presence of Black people as a way of combating racism and food insecurity.

Background

Food insecurity is defined as the limitation or instability in the availability of nutritionally adequate and safe foods without compromising access to other basic needs, or the inability to obtain food in a socially acceptable way [1,2,3]. In 2023, about 28.9% of the world’s population experienced moderate and severe food insecurity, totaling 2.33 billion people [4]. In Brazil, 27.6% of the population faced food insecurity in the same year, which equates to 3.2 million households [5].

Food insecurity has multiple social causes, such as poverty, unemployment, gender disparities, educational level and race disparities [6]. A recent survey conducted by the Brazilian Research Network on Food and Nutrition Security and Sovereignty showed that 65.0% of households headed by Black people, mainly women-headed households, are affected by food insecurity, against 46.8% of households headed by white people [7]. In the United States, households headed by Black people are also the most affected by food insecurity, accounting for 22.4%, against 9.3% of households headed by white people [8]. This disparity is attributed to the consequences of racial discrimination against Black people observed in both countries.

Racism can be defined as a system of unfair and avoidable oppression and discrimination, which assigns power and privilege to one group over another based on their race or ethnicity [9, 10]. Racism is often understood based on 3 main concepts: interpersonal racism (occurs as an ethical or psychological phenomenon, reduced to internalization or prejudiced expressions and behaviors by individuals), institutional racism (results from policies and practices carried out by institutions), and structural racism (occurs as the interconnection of organizational conditions and structures of society itself, at its social, economic, political, ideological, and ecological levels, resulting from political and historical processes) [11,12,13]. Generally, when manifestations of all 3 concepts are taken together in society, racism is defined as systemic. Given that Brazil is the country that most compulsorily received enslaved African people into its territory and was the last one to abolish slavery, the African Brazilian population still faces the consequences of the legacy of this historical process with a wide range of persistent inequalities in different social domains, including food insecurity [14].

Alongside the association of racism with food insecurity, Black people also live in more degraded social environments [15,16,17]. Food environment serves as an interface between the food system and dietary practices, including the availability, affordability, convenience, and desirability of food [18]. Studies in the United States and Brazil have shown that the food environment in neighborhoods with predominantly Black residents is associated with a reduced number and variety of retail food stores [15, 19,20,21,22,23], an increased number of fast food restaurants [16], and reduced availability of fruits and vegetables [24]. These factors, along with lower income and higher unemployment rates [25], can contribute to the worsening of food insecurity and health of the Black population and reinforce structural mechanisms of racism in urban dynamics [26].

Also, a food environment with poor availability of food, especially healthy foods, and high food prices further contributes to food insecurity [27]. The occurrence of food insecurity has already been associated with the purchase of food in grocery stores and convenience stores and limited physical access to food stores in the neighborhood, situations also observed in the Brazilian reality [24]. One of the few studies to explore the association between food environment and food insecurity demonstrated that food insecurity is more common in populations living in regions with unhealthy food environments [10].

Therefore, racism appears to be related to both unhealthy food environments and higher prevalence of food insecurity. However, to our knowledge, no study has jointly explored the interrelated effects of racism and unhealthy food environments on the occurrence of food insecurity. The objective of this study was to analyze the relationship between racial discrimination, community food environment, and food insecurity in adults living in a capital city in southern Brazil. We also aimed to investigate the potential modifying effect of the food environment on the relationship between racial discrimination and food insecurity.

Methods

Study design

This cross-sectional study included a sample of the population residing in the territory covered by the Santa Cecília primary health care (PHC) unit, located in the central area of ​​Porto Alegre, the capital of Rio Grande do Sul state, the southernmost state of Brazil.

This was a 2-stage study. In the first stage, data were collected from the population residing in the coverage territory, followed by the identification and audit of all food retailers in the territory. This study is part of a larger research project titled “Study of the social and environmental determinants of food and nutrition: an ecosocial approach,” which was approved by the Research Ethics Committee of Universidade Federal do Rio Grande do Sul, Brazil, under number CAAE 46934015.3.0000.5347, in accordance with the Declaration of Helsinki. Each participant provided written informed consent prior to inclusion in the study.

Study population

Three PHC units are responsible for serving more than 260,000 residents of the central area of Porto Alegre, including the Santa Cecilia PHC unit, which serves 30,000 people. Part of these families live in four geographically well-defined areas of lower socio-economic status (average per capita income, R$ 1700·00), whereas the remaining families live in higher-income areas (average/capita income, R$ 4000·00).

Sample and sampling

The sample size was calculated (n = 400) for a larger study [28]. For the present study, this sample size had an 80% power to detect a 10% difference in the prevalence ratio (PR) of the association between experiences of racial discrimination (exposure) and food insecurity (outcome).

The inclusion criterion was individuals aged between 20 and 70 years of both sexes. The exclusion criteria were individuals with any physical or mental limitations that impeded data collection as well as pregnant women. A proportional sample of residents from lower and higher-income areas was obtained to ensure representation across different socioeconomic and environmental strata.

In the lower-income areas (areas 1 and 3), with only 250 households, all eligible individuals were invited to participate in the study (census sampling); 201 individuals who agreed to participate were included (refusal rate, 16%). In the higher-income areas (areas 2 and 4), a similar number of participants was included to ensure a proportional sample (n = 199). In these areas, a random sampling procedure was used to select the main sampling units (households) (refusal rate, 22%). Only one person per household was included. When more than one person in a household met the inclusion criteria, one individual was randomly selected for the interview, alternating between men and women in each household to enhance representation (i.e., whenever a woman was included, an attempt was made to include a man in the neighboring house and vice versa).

The food environment assessment included all food retailers in the 4 areas under study.

Data collection

Participant data were collected in person between October 2018 and June 2019 using a standardized, pre-tested, pre-coded questionnaire with questions about sociodemographic and economic status, in addition to questions about racial discrimination.

The first stage consisted of mapping the areas for addresses and locations of families served by the PHC unit, with the help of community health workers. The team then identified individuals who met the inclusion criteria and invited them to participate in the study. The questionnaire was administered either at the time of invitation or, if necessary, during an interview scheduled for completion at the participant’s own home or the PHC unit. The questionnaire was applied only after the participant had agreed to participate and signed the informed consent form.

Data from food retailers were collected between December 2019 and February 2020 by a team trained in the use and application of the tools. In the first stage, all food retailers in the territory covered by the PHC unit were mapped, identified, and audited by 2 researchers working in pairs, where one identified the food store and recorded its geographic coordinates while the other characterized it in terms of type, address, and business name.

In the second stage, the food stores were visited and a standardized, pre-tested, pre-coded form was used to collect data on the price, quality, and availability of food items sold in the store.

Assessment of food insecurity

Food insecurity was assessed using the short version of the Brazilian Food Insecurity Scale (EBIA for short, in Portuguese), developed in 2014, tested and validated for use in the country [29]. The EBIA is a psychometric scale and its short version consists of 5 yes/no questions that generate a dichotomous assessment (presence or absence). The family is classified as food secure, in the absence of positive responses, or as food insecure, in the presence of any positive response.

Assessment of race/skin color and racial discrimination

Self-reported race/skin color followed the Brazilian Institute of Geography and Statistics classification, in which the participants identified their race/skin color among the following options: white, Black, Brown (“pardo” in Portuguese), Yellow, or Indigenous. The responses were categorized into white, Black, and Brown for analysis due to the lack of Yellow or Indigenous individuals. It is important to note that self-reported race/skin color may function as a proxy for racism, that is, as a social marker of groups that share experiences of oppression and discrimination based on race/ethnicity and historical and social processes of racialization. At no time should this variable be understood as a biological marker [30].

Experiences of racial discrimination were assessed using an adapted Portuguese version of the Experiences of Discrimination scale (EOD), developed in 1990 by Nancy Krieger and updated in 2005 by the author. The EOD is an 18-item self-report questionnaire that measures perceived experiences of discrimination based on race/ethnicity or skin color for population health research [31]. The scale was adapted and validated for use in the Brazilian population [32], consisting of 13 items that cover 2 dimensions: experiences of discrimination (9 items) and worry about discrimination (4 items). Dimension 1 encompasses personal experiences of maltreatment or unfair treatment based on self-identified race, ethnicity, or skin color. The 9 items cover discrimination encountered across the following situations: at school, getting a job, at work, getting housing, getting medical care, getting service in a store or restaurant, getting credit, on the street or in a public setting, and from the police or in the courts. The response options for each situation are: never, once, 2 or 3 times, and 4 or more times. Although we applied the complete scale, we only used dimension one in this study. The scores assigned to each response (0 to “never,” 1 to “once,” 2.5 to “2–3 times,” and 4 to “4 or more times”) were summed across the items, for a total score ranging from 0 to 30. Higher scores indicate more experiences of racial discrimination throughout the lifespan. The scale showed good validity for this population in a previous analysis [13].

Assessment of the community food environment

The community food environment encompasses the distribution of food sources, that is, the number, type, location, and accessibility of food retailers, as well as their hours of operation and drive-through windows [33]. The assessment of the community food environment was based on the number and type of food retailers, categorized into 7 types, and the quality of food retailers through audit and application of a tool based on the Nutrition Environment Measures Survey in Stores (NEMS-S), developed by Glanz et al. and adapted and validated to assess the food environment in urban areas in Brazil [34].

The tool consists of a scoring system that classifies food stores according to the availability and price of 108 food items and the quality of fruits and vegetables, if available in the stores. The tool was adapted to the local food context by replacing some foods with others from the same food category, a replacement based on previous publications [35, 36]. The tool also allows us to classify food composition and characterize food items as unhealthy, intermediate, or healthy, assigning a negative score to unhealthy foods and a positive score to intermediate and healthy foods. The total score ranges from − 30 to 100 points for each food store, with higher scores indicating healthier food items available in the store. In this study, scores were obtained for the total territory and for each of the 4 household areas, categorized according to the NEMS-S score.

Assessment of socioeconomic and demographic variables

A questionnaire was used to assess the participants’ socioeconomic and demographic characteristics by asking questions about sex (female/male), age in completed years (categorized into age groups: 19–36, 37–49, 50–59, and > 60 years), marital status (with a partner [married/ consensual union] or without a partner [single/ separated/ divorced/ widowed]), education in completed years of schooling (< 8, 8–10, 11, and > 11 years), and monthly family income (categorized according to the Brazilian minimum monthly salary: < 1, 1–2, 3–5, and > 5 minimum monthly salaries). Of note, the Brazilian minimum monthly salary denotes government regulation for a minimum monthly rate paid for a worker who works, on average, 44 h per week for 4 weeks in a month.

Statistical analysis

The data were double-entered and checked using EpiData version 3.5. Descriptive statistics were used to compare the characteristics of different household areas. Categorical variables were expressed as absolute (n) and relative frequencies (%), and numerical variables were expressed by measures of central tendency (mean and median) and dispersion (SD, IQR, and minimum and maximum). Pearson’s chi-square test or Kruskal-Wallis H test were used to evaluate the heterogeneity of proportions as needed.

The multivariate model was based on a previously constructed directed acyclic graph (DAG) using DAGtty (available on www.dagitty.net), to minimize possible biases [37] (Fig. 1) Crude and adjusted PRs were calculated for the associations between food environment (NEMS-S score), race/skin color, and racial discrimination using Poisson regression with robust variance, including the respective 95% CIs and Wald test for linear restrictions. The adjusted analysis controlled for the demographic and socioeconomic characteristics associated with food insecurity in the bivariate analysis and single-level theoretical model. Income and education were not considered confounding factors, as they were associated with the outcome and exposures and were part of the causal chain of associations. Therefore, they were not included in the adjusted model. Stratified analyses were performed to investigate the potential modifying effect of the food environment on the relationship between racial discrimination and food insecurity after a statistically significant interaction test (p < 0.001).

Fig. 1

Directed acyclic graph (DAG) representing the association between racial discrimination, community food environment, and food insecurity

All data were analyzed using Stata (StataCorp, College Station, TX, USA), version 18.0, and a p-value less than 5% (p < 0.05) was considered statistically significant.

Results

Table 1 shows the sociodemographic, food insecurity, racial discrimination, and community food environment characteristics of the total sample and by household area. The mean (SD) of age was 47 (13.98) years (data not present in tables). Most participants were women (75%), did not have a partner (62.8%), had 11 years of schooling (39.9%), and had a monthly family income of 3 to 5 minimum monthly salaries (48.4%). Food insecurity was present in more than half of the participants (51.1%) and was also more prevalent in areas 1 (56.5%) and 3 (58.8%). Additionally, 37.8% of the sample self-identified as Black or Brown. When stratifying by household area, area 1 (lower socioeconomic status) had a higher proportion of Black and Brown residents, lower education, and lower income.

Table 1 Description of sociodemographic characteristics, racial discrimination, food insecurity, and community food environment

Regarding self-perceived racial discrimination, the mean (SD) EOD score was 2.4 (5.43), with areas 1 and 3 showing the highest means of discrimination, with mean (SD) scores of 3.25 (6.20) and 2.94 (5.98), respectively. As measured by the NEMS-S tool, the mean (SD) score of the community food environment for the availability of healthy and unhealthy items in the food retailers was 17.6 (23.9), with lower scores being observed in lower income areas (area 1: 6.6 points, SD 19.6; area 3: 17.9 points, SD 25.7) (Table 1).

Table 2 shows the associations between socioeconomic variables and food insecurity. Food insecurity was more prevalent in women (55.3%), in those aged 19 to 49 years (56.9%), in lower-income people (77.2%), and in Brown people (57.5%), although without statistical significance. The prevalence of food insecurity was higher in areas with a community food environment of poorer quality (area 1: 56.5%; area 3: 58.8%). The mean racial discrimination score was higher in food-insecure households (2.93 points, SD 5.91) compared to food-secure households (1.84 points, SD 4.82).

Table 2 Participants’ sociodemographic and economic characteristics and associations with food insecurity

As shown in Table 3, there was an association between experiences of racial discrimination and a higher prevalence of food insecurity, even after adjusting for potential confounders (PR 1.07; 95% CI, 1.03–1.20). Regarding household areas, participants residing in area 4 had a lower prevalence of food insecurity than those residing in area 1. When stratifying the analyses of the association between racial discrimination and food insecurity by the community food environment, the association remained statistically positive only for the households in areas with more unhealthy community food environment (PR 1.06; 95% CI, 1.01–1.10), even after adjusting for potential confounders.

Table 3 Crude and adjusted prevalence ratios (PRs) and their respective 95% CIs for the association between race/skin color and experiences of racial discrimination and the prevalence of food insecurity according to the community food environment

Discussion

Our results indicate that Black and Brown race/skin color is positively associated with a less healthy community food environment. Experiences of racial discrimination were positively associated with food insecurity and, when evaluating this association stratified by household area, we observed a positive association only for households in areas with a poorer food environment (this association remained statistically significant even after adjusting for potential confounders). This result demonstrates that the food environment can be a modifier of the effect of the association. Based on our literature review, this is the first study to evaluate how the food environment can modify the effect of the association between racism and food insecurity.

Previous studies have demonstrated a consistent association between a greater presence of Black people in the neighborhood and a poorer community food environment [15, 16, 19,20,21,22,23,24], in line with the results obtained in the present study. A scoping review conducted in 2023, including only studies carried out in the United States, concluded that 30% of the included studies associated one or more indicator of structural racism, such as gentrification and racial residential segregation, with geographic access to food retailers, while 70% of studies documented disparities in access to food retailers according to the neighborhood racial/ethnic composition [38]. Likewise, in Brazil, studies based on secondary data conducted in the Southeast [39], Northeast [23], and South [15] regions documented that a greater presence of Black people in the neighborhood was associated with the presence of food deserts, that is, regions characterized by limited availability and access to healthy and fresh food. In Porto Alegre (South region), areas with a higher percentage of Black and Indigenous people are twice as likely to be classified as a food desert [15].

The literature has characterized the poor availability of healthy food options in areas with a greater proportion of Black residents as a manifestation of structural racism [15, 26]. By using race/skin color as a variable to describe the household areas with a greater or lesser percentage of Black residents, we were able to identify an important part of the action of structural racism, which occurs through spatial segregation by providing residents with poor access to food products, services, and options in different areas or keeping access narrowed to ecologically degraded areas. In this study, structural racism operated through Black people’s difficulty in having access to healthy foods, resulting in a higher prevalence of food insecurity [11].

Our findings regarding the association between racial discrimination and food insecurity corroborate those of international studies. In 2019, Phojanakong et al. performed a cross-sectional study in the United States showing that household food insecurity in mothers who reported experiences of discrimination was twice that in mothers without such experiences [40]. Our study demonstrated that each 1-point increase in the discrimination score was associated with a 7% increase in the likelihood of food insecurity. In line with this finding, Burke et al., in a cross-sectional study of Black people living in South Carolina, United States, found that each 1-point increase in the self-perceived racial discrimination scale was associated with a 5% increase in the odds of being food insecure [25].

In Brazil, to our knowledge, there are no studies relating measures of experiences of racial discrimination to food insecurity. However, nationally representative data have previously demonstrated a higher prevalence of food insecurity among Black individuals, which indicates, as stated earlier, its relationship with structural racism [41,42,43,44,45,46]. In our study, although prevalence rates were also higher among Brown and Black people, the result was not statistically significant. Black people in Brazil have lower wages, about 30% lower than those of white people, occupy only 29.5% of management positions, have higher poverty rates, reaching 34.5% for Black people and 38.4% for Brown people, and experience more physical, psychological, and sexual violence [40, 47]. The whole scenario, consisting of different indicators of institutional and structural racism, increases the vulnerability of the Black population to food insecurity in a country already widely affected by it [11].

Considering the multiple causes of food insecurity, the community food environment was shown to be an important factor in the relationship with racism, since the association between food insecurity and racial discrimination was present among residents of areas with a poorer food environment. Physical access to food retailers that primarily sell healthy and culturally appropriate foods is essential to ensure food security [48]. However, people experiencing racism may have difficulty accessing these food stores, a factor that, along with other socioeconomic constraints, contributes to increased food insecurity. Racial residential segregation in the city of Porto Alegre, a product of institutional and structural racism, has hampered urban mobility, access to retailers and services, and the availability of better employment and housing opportunities for Black people for decades [49].

There are two main approaches to measuring the impact of racism in the field of health, both of which were employed in this study to analyze food insecurity. The first approach involves assessing the association between race/skin color and the corresponding study outcome to identify health disparities among specific racial/ethnic groups or populations. The second is based on psychometric questionnaires that measure experiences of discrimination, which are considered a more direct way of assessing racial discrimination in its multiple settings (interpersonal, internalized, or institutional) [30, 50, 51]. Direct assessment of discrimination using psychometric instruments is a growing trend in research. Their use is supported by the need to understand the stressful experiences related to discrimination and their association with negative health outcomes. This allows for the investigation and construction of more detailed mechanisms on how discrimination operates in the study population [50]. In the present study, only the variable ‘experiences of racial discrimination’ was associated with food insecurity, unlike the variable ‘race/skin color’. This finding suggests that instruments that measure racial discrimination should be incorporated into future studies investigating the health of the Black population [52]. Conversely, it also suggests the need for more complex tools and indicators to better understand the broader manifestations of racism, including systemic and institutional racism.

Finally, the findings revealed a high prevalence of food insecurity across the entire study population. The results are in line with reports in Brazil and worldwide. In Brazil, food insecurity increased from 20.6% in 2017 to 32.8% in 2022, while globally its prevalence increased from 21.7% in 2015 to 29.6% in 2022 [53]. This demonstrates the growing importance of monitoring and understanding the factors associated with food insecurity to implement public food and nutrition policies focused on the most vulnerable populations.

Limitations and strengths of the study

This study has strengths (1) the use of validated questionnaires for the study population to measure the outcome and exposures, and (2) the assessment of the community food environment through an audit of all food retailers, with this measure being more robust than assessment using secondary data, which only evaluates the presence of retailers. Limitations include (1) the enrollment of a study population that is not representative of the general population, but the explanatory mechanisms of associations may be used to help understand similar contexts in low- and middle-income countries, (2) the use of the variable ‘race/skin color’ as a proxy for structural racism, which hinders the development of detailed mechanisms about its effects and the most appropriate interventions, and (3) the use of the short version of the EBIA, which did not allow us to classify individuals according to different levels of food insecurity.

Conclusion

This study aimed to analyze the relationship between racial discrimination and food insecurity. We found that racial discrimination, as an expression of racism in different situations, was associated with a higher prevalence of food insecurity in the study population. We also found that the food environment, in areas with a greater presence of Black people, was an effect modifier of this relationship, potentially indicating the impact of structural racism. The results highlight the need for public food security policies focused on areas with a greater presence of Black people to reduce racial inequalities in health and promote access to adequate and healthy food for all.