Introduction

Fruits and vegetables (FV) are markers of a healthy diet [1, 2] and are therefore associated with a lower risk of morbidity and mortality [3, 4]. Despite its importance to health, FV consumption remains below recommended levels in most countries [5–8]. In many low and middle-income countries, purchasing the recommended levels would require a substantial portion of household income, making FV unaffordable [5–8]. As a matter of fact, FV represent the largest share of the costs to a healthy diet [9].

In Brazil, the increase in the social gap for adequate consumption of FV is concerning [8], and low-income families commit household income almost three times more than high-income families with the costs to a healthy diet [10]. Inequalities in FV access and consumption [5, 8] reinforce the need to improve access to healthy foods, which can be achieved through actions in the food environment [11–15].

The food environment is the interface between individuals and the food system [16, 17]. It encompasses dimensions of access to food that can influence people’s food acquisition and consumption [17]. The dimensions of access include availability, prices, affordability, accessibility and convenience of food and food retailers [17–20]. The importance of availability (i.e., types of food retailers and food available) and affordability (i.e., food price and purchasing power) stand out for FV consumption [5, 13, 20, 21].

Thus, the retail food environment wield significant influences on access [20], and barriers and facilitators for healthy food choices differ significantly among types of food retailers [22, 23]. For instance, in Brazil, fresh food is mostly acquired in open-air markets, greengroceries, butcheries, small producers, and street vendors [24, 25]. Although supermarkets have a larger market share with high participation in the acquisition of all types of food for Brazilians, it is important to highlight that most of the ultra-processed food comes from supermarkets [24].

Food and Nutrition Security public policies must be implemented through broad and synergistic actions that boost the distribution and commercialization of healthy foods [8]. In this context, the Food and Nutrition Public Establishments (FNPE), which are physical structures and public spaces designed to strengthening local circuits of food production, supply and distribution, stand out [26, 27]. FNPE that operate in the retail sector are food retailers focused on selling fresh foods at prices below those practiced in the market [26, 27]. The FNPE includes specialized fresh food public retailers, open-air markets, municipal markets, and places for direct purchase from regional producers [26, 27].

FNPE have some similarities with initiatives such as farmers markets, mobile markets, farm stands, and cost-offset community-supported agriculture, which are part of public Food and Nutrition Security policies in many countries in Latin America [11], Europe [28–32] and also in the United States [33–35]. In Brazil, FNPE are structural elements of the National Food and Nutritional Security Policy [23, 24]. They gained prominence in 2003 under the Zero Hunger strategy, launched by the federal government [23, 24]. Through bidding processes launched annually by the Ministry of Social Development, several states and municipalities joined the program [23, 24]. Currently, there are FNPE installed in 1652 medium and large Brazilian cities [23, 24]. Their structure varies between cities, in Belo Horizonte, they are implemented in areas with greater traffic of people, and there is a higher concentration of FNPE in areas of low socioeconomic vulnerability [36].

The FNPE are implemented to promote regular and permanent access to adequate and healthy food [23, 24]. However, there is little monitoring and evaluation of the impacts of public policies on food environments, which are fundamental for strengthening these actions [37]. It is necessary to consider that the application of FNPE takes place in complex spaces, which may affect availability and price variation in different ways in the various types of food retailers in the neighborhood [38–41]. Therefore, it is worth investigating the availability and price of food in FNPE and neighboring food retailers. Areas with FNPE are expected to have greater availability of FV sold at prices lower than those practiced in areas without FNPE. Thus, the study aimed to evaluate the availability and price of FV, according to the type of food retailers, in areas with and without FNPE, to better understand the food environment in the surroundings of FNPE.

Methods

Retail food environment audit

The store audits of FNPE and food retailers was carried out in pairs, by a team of 10 researchers. The researchers were previously trained and monitored by field supervisors and the research coordinator.

Data collection used the “Food Store Observation Tool (ESAO-S)” and the “ESAO-S adapted to be used at open-air food markets”, tested and validated for Brazil [49]. Information was collected on availability (i.e., number of food retailers with available items) and price per kilo of the four FV (banana, orange, papaya, watermelon, pumpkin, tomato, carrot, and chayote) most consumed in Belo Horizonte, according to the 2008–2009 Household Budget Survey (HBS) [50]. Prices were assessed on the cheapest variety. Price comparisons were made for ’conventional’ products, excluding organic items.

Results

A total of ten areas with FNPE were included in the study, containing specialized fresh food public retailers (n = 2), street markets (n = 4), municipal markets (n = 2), and direct purchase programs from local producers (n = 2). Within the municipal markets, nine stores were investigated, considering the coexistence of different types of food retailers with independent commercialization. Thus, 17 FNPE units were audited. In total, 148 food retailers were audited, including the FNPE units.

Fresh food retailers were less prevalent than mixed food retailers, and there was no difference in the prevalence of types of retailers between areas (p>0.05) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Types of food retailers according to areas.

Belo Horizonte, Brazil. 2019.

https://doi.org/10.1371/journal.pone.0294473.t001

Regarding the availability of FV, in mixed food retailers there was low and similar availability between areas (34.0% vs. 26.2%; p>0.05). The availability of vegetables was higher in fresh food retailers in areas with FNPE when compared to fresh food retailers in areas without FNPE (100% vs. 82.8%; p = 0.024) (Table 2). Availability was tested for each FV investigated, agreeing with the results of total values.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Availability of fruits and vegetables according to areas.

Belo Horizonte, Brazil, 2019.

https://doi.org/10.1371/journal.pone.0294473.t002

In general, similar values were observed when comparing the prices of the same food in the same type of food retailer–fresh food or mixed food–in areas with and without FNPE; in mixed food retailers, one of the items had lower prices in areas without FNPE (R$2.90 vs. R$3.85; p<0.05) (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Price of fruits and vegetables and price differentials between food retailers according to areas.

Belo Horizonte, Brazil, 2019.

https://doi.org/10.1371/journal.pone.0294473.t003

However, the price difference between fresh food and mixed food retailers varied significantly between areas with and without FNPE (p<0.001 for six of the eight foods investigated). In general, price differences comparisons show that fresh food retailers are cheaper than mixed ones in areas with FNPE. On the other hand, in areas without FNPE, mixed retailers apparently are cheaper than fresh food retailers. For example, in areas with FNPE, the average banana price at fresh food retailers subtracted from the average banana price at mixed food retailers results in a difference of -R$0.55, (i.e. 16.3% lower at fresh food retailers). But, in areas without FNPE, the subtraction of the average banana price at fresh food retailers subtracted from the average banana price at mixed food retailers is R$0.45 (i.e. 13.4% higher at fresh food retailers) (Table 3).

Discussion

Through store audits of a sample of 148 food retailers in areas with and without FNPE, it was possible to make progress in measuring the availability and price of FV in areas where a public policy on Food and Nutrition Security is operating. Fresh food retailers were less prevalent than mixed food retailers and were in similar proportions between areas with and without FNPE. In fresh food retailers, the availability of vegetables was higher in areas with FNPE. In mixed food retails, the availability of FV did not vary according to the areas, remaining low in both. In fresh food retailers, similar prices were observed in regions with and without FNPE. In mixed food retailers, one of the investigated items had higher prices in areas with FNPE. In general, price differences comparisons show that fresh food retailers are cheaper than mixed ones in areas with FNPE. On the other hand, in areas without FNPE, mixed retailers apparently are cheaper than fresh food retailers.

Most audited food retailers were the mixed type, reinforcing the hypothesis of similar distribution of establishment types in the country [25] and also in different urban centers [22, 51, 52]. However, our study adds new information by observing such distribution also concerning the presence of FNPE. There was no difference between the prevalence of these food retailers (fresh and mixed) according to the areas with and without FNPE, suggesting that the presence of the FNPE did not attract (or repel) food retailers in their surroundings. Environmental changes that drive responses to local food systems are complex and dynamic [53]. In addition to demand, the choice of a location for installing a new food retailer is based on strategies aimed at maximizing profits, to ensure the economic viability of the point of sale [54, 55]. In this logic, food retailers prioritize the richest areas of cities [54, 56–60]. In general, FV consumption is more common among better educated and wealthier people [61], and regular FV consumption is lower among lower-income individuals living in neighborhoods with fewer fresh food retailers [61].

Although FNPE did not interfere with the prevalence of establishments in their surroundings, their presence was associated with the availability of vegetables. In several countries, initiatives similar to FNPE demonstrate effectiveness in increasing the availability of FV in the regions where they are implemented [28–30, 33, 38]. It is known that the increased availability of FV in food retail is positively associated with increased consumption of these foods [17].

About the price of FV, it was possible to observe that the price difference between fresh food and mixed food retailers varied significantly between areas with and without FNPE, apparently because mixed retailers in areas with FNPE charge higher prices than those observed in the mixed retailers in areas without FNPE, as seem for one of the investigated items. Although our study is not able to present explanations for this scenario, two hypotheses must be considered: i) the presence of FNPE influences the type of food offered in the mixed retailers around them, causing them to focus on products of higher quality and higher income customers, who would be willing to pay higher values for convenience aspects [62]; and ii) the FNPEs are located in areas with a large natural flow of customers, concentrating in regions of low social vulnerability [36], a condition that also attracts food retailers more intensely [54, 56–60], precisely because of the possibility of profit maximization and implementation of higher prices [54, 62].

Large mixed food retailers, such as supermarkets and hypermarkets, have ample negotiating power [63]. In the last two decades, there has been a greater focus on agriculture and food production to supply these mixed food retailers [11], which gives them an advantage in commercial negotiation and the possibility of greater price control in the markets where they are installed [54]. However, in these food retailers, there is greater availability and dominance of ultra-processed foods [22–24, 52], which exposes consumers to a simultaneous offer of healthy and unhealthy foods. This availability and promotion of healthy and unhealthy foods in the same place seem to have ambiguous impacts on consumer eating behaviors, leading to ’matched’ purchases and, consequently, the acquisition and consumption of unplanned foods, which are commonly ultra-processed [61, 62]. Strategies present in mixed food retailers, such as greater availability, advertising, product promotion (tasting), price promotion, and physical modifications (such as consumer circulation through all sections and store; arrangement in islands or queues at checkouts or on shelves at eye level), encourage the purchase of ultra-processed foods [22, 24, 52, 64].

In general, price differences comparisons show that fresh food retailers are cheaper than mixed ones in areas with FNPE. On the other hand, in areas without FNPE, mixed retailers apparently are cheaper than fresh food retailers. It is known that food prices influence the choice of purchasing place [20] and the tendency to buy in fresh food retailers is more significant when these places have lower prices than other food retailers [65]. Increasing economic access to fresh food retailers is a crucial factor for the efficiency of intervention policies in the food environment [34]. Thus, the scenario observed in areas with FNPE could contribute to an increase in the propensity of choosing fresh food retailers as shopping places for consumers. The lowest price in fresh food retailers could be associated with the presence of FNPE. There are reports in the literature of reductions of up to 8% in food prices in food retailers adjacent to a retailer that charged FV prices lower than the local average [65]. However, it was not possible to find a difference between the prices charged in fresh food retailers in areas with FNPE compared to those in areas without FNPE. Naturally, only the larger FNPE has the actual capacity to interfere in the commercial structure of their surroundings, and given the great spatial variability of food prices, the identification of differences of small expressive magnitude (less than 20%) is conditioned to the use from samples with a much higher number of establishments than that of the present study.

In general terms, this study suggests that the food environment in areas with FNPE can lean toward healthy eating habits, by improving the availability of vegetables. Also, only in areas with FNPE, fresh food retailers presented lower price values than those observed in mixed food retailers, which can lead consumers to choose these food retailers. The price difference between fresh food retailers and mixed food retailers is not necessarily attributed to the presence of the FNPE and could be related to the socioeconomic status of the areas where these FNPE are mostly implemented. Broadly speaking, Food and Nutrition Security public policies aimed at making fresh food retailers more accessible and affordable could have wide benefits in FV consumption. Monitoring public policies in this field is essential for strengthening and expanding these actions in different contexts.

Some limitations of the study need to be pointed out. Data regarding the availability and price of FV were assessed simultaneously in a cross-sectional study. Thus, temporal variability in food availability and price may not have been measured. However, Brazil has a tropical climate, where the most consumed FV is produced all year round. And the availability of these foods tends to be consistent over two weeks, with a single observation reflecting commercial activities at the selected food retailers [66]. Another limitation related to food retailers that refused to participate in the study (34.4%), of these, 31% were fresh food retailers, and 69% were mixed food retailers, so the availability, and prices can be different. However, the distribution of the types of refusal food retailers did not differ from the distribution of the analyzed food retailers (p>0.01). Additionally, the circular buffers, drawn to delimit the investigation areas, establish boundaries of the food environment, which can be larger or smaller [67], but which are arbitrarily defined by the researcher, based on the literature [51, 61, 68].

Finally, the study design does not allow inferring a causal relationship between the presence of FNPE and the differences found between the areas. It is significant to consider the need for different approaches and designs, such as those used in this study, to try to clarify complex issues in environmental studies and to move forward [69]. In addition, monitoring and understanding the impacts of public intervention policies on the food environment are essential to register trends and provide relevant information to civil society and government authorities, however, the employment of traditional empirical methods to test the effectiveness of these strategies remains challenging [34].

Conclusion

There was a higher prevalence of mixed food retailers, with no difference between areas with and without FNPE. The presence of FNPE was associated with the availability of vegetables. Overall, FV prices were similar in areas with and without FNPE. The price difference between fresh food retailers and mixed food retailers varied significantly between areas; in general, fresh food retailers presented lower price values than those observed in mixed food retailers, only in areas with FNPE. Adopting Food and Nutrition Security public policies can help to decrease the global burden of nutrition-related chronic diseases and inequities in health and access to healthy foods, and the application of relevant policies requires political leadership, broad social dialogue, and scientific information, as pointed out in the study. We used validated tools for auditing the retail food environment, aimed at monitoring public policies for physical and economic access to healthy foods; the importance of studies such as this one is highlighted, especially in low- and middle-income countries.

Supporting information