‘Nobody shops at the neighborhood store’: leveraging a community’s pediatric fresh produce prescription program to inform future participating store redemption locations

ABSTRACT

Research examining the nature of food shopping often considers proximity to the nearest or overall distance travelled to multiple stores. Such studies make up a portion of new work on so-called ‘food deserts’ and the issues inherent in the term, including that most people do not shop at their nearest store, and mobility challenges vary vastly from one person to the next. Increasing the knowledge base on shopping characteristics could be useful for behavioral interventions and programs aimed at increasing healthy food shopping. In this study, we examined the shopping characteristics of 627 caregivers whose children were enrolled in a pediatric fresh produce prescription program at one of three large pediatric clinics in Flint, Michigan. We compare these characteristics to the potential of a new food cooperative to improve geographic accessibility to healthy food. In particular, we propose the expansion of the prescription program to this new cooperative for health-related as well as local economic development reasons. Our work bridges topics of interest to researchers and practitioners working in nutrition, food access, and economic development.

KEYWORDS:

• Geographic information systems
• food access
• pediatric clinic
• produce prescription program
• food co-op

Introduction

Studies on food access and food consumption commonly consider some analysis combining either the distances to the nearest stores to one’s home or to the actual stores where people shop. Studies differentiating these phenomena are important because of the wealth of evidence indicating that people do not commonly shop near where they live due to a variety of reasons.

While this complicates simplistic conceptions of food deserts and food access, broad patterns still remain in that people in underserved areas frequently travel farther to reach healthy foods – often outside of their own communities. Additional inquiries into this topic, therefore, may help contextualize how environmental-based interventions may be able to be deployed to both provide more local healthy food options and serve in capturing part of the economic leakage rate seen in such communities (i.e. the phenomenon whereby communities that have experienced disinvestment see spending dollars physically leave the neighborhood due to a lack of nearby stores). While many people may travel farther to reach preferred choices, providing better choices more locally may help alleviate travel-related stressors of food shopping and increase the economic vitality of disinvested neighborhoods.

At the same time, produce prescription for health programs have become increasingly popular as a means of promoting healthy food choices while also addressing food access challenges (Buyuktuncer et al. 2014, Bryce et al. 2017, Trapl et al. 2018, Saxe-Custack et al. 2019a). These programs differ in design and scope, but most include physician-issued prescriptions that may be exchanged for fresh produce at local farmers’ markets, mobile markets, or food stores. Recent studies have shown participant desire for produce prescription programs to partner with multiple large grocery store chains. These partnerships often require large numbers of program participants, however, as chain stores must incorporate the local incentive programs into their point-of-sale systems (Riemer et al. 2021). Further complicating matters, most people do not shop at farmers’ markets or local food stores because their products are more expensive, operating hours are limited, or they are less convenient to visit. In the context of prescription programs, these ingrained habits lead to limited redemption rates at the farmers’ market and a local food hub where our program operates, due additionally to limited hours of operation and distance from participant homes (Saxe-Custack et al. 2018, 2019b).

The continued challenges in encouraging people to use smaller neighborhood stores to strengthen local economic development – and the potential benefits of prescription programs – suggest that advocacy and outreach for smaller stores would be important for magnifying the impact of prescription programs and for strengthening the potential linkages to local economic development. In this study, we examined the shopping characteristics of 627 caregivers whose children were enrolled in a pediatric fresh produce prescription program at one of three large pediatric clinics in Flint, Michigan.

We conducted a series of spatial network analyses to determine the distance to stores where they reported shopping, their actual nearest stores, and a soon-to-be completed food co-op. We then compare these distances against one another to examine opportunities for more proximate grocery shopping. As the prescription program is currently planning expansion into additional stores, this evidence will inform where they choose to locate new sites, which could provide evidence that we and other advocates could use to promote further improvements in food access and redemption rates.

Literature review

Food access & food consumption

Topics of food access and food consumption have evolved considerably since the first papers on ‘food deserts’ in the UK more than 20 years ago (Beaumont et al. 1995, Cummins and Macintyre 2002). No longer thought of as mere ‘regions’ on a map, researchers have interrogated space-time aspects and unique shopping differences among individuals, and advocated for the retirement of the term (Widener and Shannon 2014, Widener 2018). Newer terms such as food apartheid and food sovereignty have become more popular, as researchers and community partners alike recognize the ways that structural racism and inequality create such environments, and instead point toward building strength in localized food systems for the benefit of individual consumers (Gripper et al. 2022, Joyner et al. 2022). Fundamentally, while the concern remains the same – addressing inequalities in access to nutritious foods – the recognition is that individual-level and sociocultural differences drive food access in ways that are more complex than geography.

Food access in American cities is, of course, typically a question of whether individuals have access to a car. Drewnowski et al. (2012) seminal article on this topic in Seattle found that only one in seven respondents shopped at the nearest store. Additional research has repeatedly shown lower-income individuals bypassing poorer quality or higher-priced stores to reach other stores, often by using a private automobile (Zachary et al. 2013, Rogers et al. 2018, Sadler et al. 2019).

Lack of access in underserved areas

The idea behind food deserts – and food swamps and related terms – gained such traction because of repeated findings that socioeconomic status correlated strongly with food access in several domains (Wood and Horner 2016, Kraft et al. 2020, Brown et al. 2021). The general tenets that healthy food is less accessible and unhealthy food is more accessible in poorer and minoritized neighborhoods prevail in this work. While the methodological choices for measuring food deserts vary based on population density and urban area size, this pattern tends to persist across the urban-rural continuum. Furthermore, the measures used throughout the food desert literature are often insufficient for defining places where a large percentage of people may have challenges accessing healthy food (Sadler et al. 2016). Since our study takes place in a mid-sized declining city, we motivate much of our literature around this particular framing (Bedore 2010, Deener 2017).

Relevantly, recent research has likewise linked historical structurally racist practices to inequalities in food access, including redlining (Li and Yuan 2022, Swope et al. 2022), as well as patterns of white flight and blockbusting (Sadler et al. 2021). Underscoring this work, however, is an increasing recognition that ‘geography is not everything’ when it comes to eating healthy, and that additional structural supports and interventions must be conducted to address apparent geographic disparities in healthy food consumption.

Benefits of ‘prescription for health’ programs

The benefits accrued from consuming nutritious foods (including fruits and vegetables) are diverse, and include aiding in human growth and development (Bradlee et al. 2010, Carter et al. 2013, Li et al. 2013, Donin et al. 2016, Albani et al. 2017), cognitive development and functioning (Cohen et al. 2016) and chronic disease prevention (Ness et al. 2005, Liu et al. 2014). But the majority of children in the United States consume fewer fruits and vegetables than would be recommended to attain these benefits (Kim et al. 2014). This evidence is particularly pronounced among children living in low-income households (Di Noia and Byrd-Bredbenner 2014, Drewnowski and Rehm 2015). Children are more likely to live in poverty than the population overall (Mattingly et al. 2011), and – owing to a lack of upward mobility in American culture – children who grow up in poverty are more likely to continue this pattern into adulthood (Moore et al. 2002).

Childhood is also a critical period for establishing dietary norms (Cruz et al. 2018, Appannah et al. 2018). However, issues of food access and affordability complicate the more simplistic education-to-action assumption (Morland et al. 2002, Walker et al. 2010, Zenk et al. 2011). Innovative programs designed to increase the affordability of fresh produce – such as farmers’ market incentive programs – have been successful in improving fruit and vegetable intake among adults (Lindsay 2013, Savoie-Roskos et al. 2016). The health-care field has leaned into this domain, and now offers fruit and vegetable ‘prescription’ programs (Buyuktuncer et al. 2014, Bryce et al. 2017, Trapl et al. 2018, Saxe-Custack et al. 2019a). These programs operate similarly to medical prescriptions, with physicians writing scripts for fresh produce and other healthy foods. Their early success has meant the rapid adoption of sustainable funding streams, including through the Gus Schumacher Nutrition Incentive Program in the US Farm Bill (Mozaffarian et al. 2019, White 2020). However, as these programs remain in their early stages, multiple knowledge gaps remain surrounding the appropriate duration and intensity of exposure, ideal program timing and length, implementation components, preferred redemption methods and sites, and overall effectiveness.

Most current food prescription programs use fruit and vegetable prescriptions as a disease-management strategy for adults with diet-related chronic health conditions (Bryce et al. 2017, Cavanagh et al. 2017, Trapl et al. 2018). However, prescription programs for youth frequently emphasize the important role of fruits and vegetables in the prevention of chronic diseases during childhood (Cruz et al. 2018, Appannah et al. 2018). Providing produce prescriptions to youth during office visits is but one means of attempting to counteract geographical barriers in food access by at least reducing economic barriers to accessing food that may be proximate but out of reach. The design of these programs likewise reflects a recognition among the medical field of the importance of fresh, nutrient-dense foods in health promotion and disease prevention.

Local food consumption and strengthened economies

In addition to the health benefits of prescription programs, they can also be a part of efforts to strengthen local food economies by contributing dollars to smaller-scale farmers and retailers who may be economically uncompetitive with multi-national chains.

Many efforts have been made in the recent past to counteract the increasing strength of the conventional food retailing system and the deleterious impact it has had on low-income urban areas in particular. On the retailing side, locally owned convenience stores can help foster ‘mutual interdependencies’ (Rybaczewska and Sparks 2020, p. 7), whereby small stores can grow and offer better, healthier foods for communities with the right community support. Such places can become community ‘hubs’. Other efforts have been directed at creating actual food hubs, which provide stronger local economic benefits than only promoting direct sales from farmers’ markets (Hughes and Isengildina-Massa 2015, Shideler et al. 2018).

The intent behind efforts to strengthen local food systems is to minimize leakage and thus strengthen the local economy (Vitiello and Wolf-Powers 2014). For example, at locally owned food co-ops, ‘communities benefit from the multiplier effect when co-op members spend money on local products and keep them in the community’ (Katchova and Woods 2013, p. 232). Others have explored these impacts at farmers’ markets and other food retail stores (Pearson et al. 2011, Sadler et al. 2013). The long-term net effect of such efforts – to reinforce the strength of local food economies, in particular via prescription programs at small-scale sites – include spillover effects by way of people spending more money at other nearby businesses (Kneafsey et al. 2013).

Study context

Our study took place in the city of Flint, Michigan, and its surrounding suburbs. Flint has long been in decline due to a combination of anti-urban state government policies predisposing the city to stagnation (Sadler and Highsmith 2016), institutionally supported local segregation strategies that later provoked massive white flight (Highsmith and Erickson 2015), and economic change resulting in the departure of tens of thousands of industrial jobs (Highsmith 2014). Taken together, these factors led to more than half of the city’s population leaving from 1970 to 2020. The remaining population is characterized by extremely high poverty living in poor quality, older homes. The city is also majority African American, with most of the remainder being white. More recently, a lead-in-water crisis resulted from a short-sighted state government decision to undemocratically change the city’s water source in 2014, provoking further population abandonment and increasing already-high rates of government mistrust (Hanna-Attisha et al. 2016).

Approximately 60% of youth in Flint live in poverty, and few full-service grocery stores operate within the city limits (Sadler 2016, Sadler et al. 2019). As expected from the literature above, many children here experience the double-burden of insufficient intake of nutrient-dense foods coupled with a high intake of poor-quality, calorie-dense foods (Saxe-Custack et al. 2019a, 2019). Stores in low-income neighborhoods in Flint tend to offer lower-quality foods with fewer fresh, healthy food options than higher-income neighborhoods (Saxe-Custack et al. 2018, 2019b, Shaver et al. 2018). Furthermore, the early stages of the COVID-19 pandemic had a particularly strong impact in Flint’s Genesee County (Wojciechowski et al. 2022, Sadler et al. 2022), with a spatial and racial disparity in the city and among African American residents. Nutrition was also negatively impacted (Saxe-Custack et al. 2022). Even in ordinary times – with constrained wealth and many families living in poverty and/or being under- or un-employed – Flint faces remarkable health challenges. The pandemic only compounded these ordinary times, with results that included more hungry children and increased food insecurity (Kim and Groden 2022).

The fruit and vegetable prescription program

Owing to broader issues related to nutrition and poverty, the Hurley Children’s Center Pediatric Fruit and Vegetable Prescription Program (prescription program) was serendipitously being planned even before the city’s lead-in-water disaster; it has since taken on added importance in the wake of the water crisis. The Hurley Children’s Center opened at the Flint Farmers’ Market in 2015 and debuted its prescription program in February 2016. It was the first of its kind in Michigan, with patients receiving either a voucher to the market or a basket of fresh produce at each doctor’s visit. All children received these prescriptions, irrespective of income or health status. After several program iterations responding to parental and provider feedback, the prescription program is now a $15 fresh produce prescription. This prescription is distributed to every pediatric patient at each office visit and may be redeemed for fresh produce at the Flint Farmers’ Market or Flint Fresh Mobile Market and Food Hub. Initial results indicated that the prescription program improved caregiver perceptions of children’s dietary habits and overall food security (Saxe-Custack et al. 2018); these results were measured and confirmed when the program was expanded to a second clinic (Saxe-Custack et al. 2019, 2021).

The prescriptions themselves are created by each clinic’s electronic medical record system, which also tracks monthly distribution rates. Redemption is monitored in partnership with the farmers’ market and food hub, with vendors collecting paper prescriptions and turning them into management for payment. Data from these vendors are then entered into a database for tracking purposes.

Study objectives

The objectives of this study are not only to illustrate a method for calculating the distance to actual and nearest grocery stores among a population of caregiver-child dyads participating in the prescription program but also highlight the potential benefits of shopping at a new (for many more proximate) cooperative grocery store (‘co-op’). The intent behind this work is to provide the evidence base for potentially expanding the prescription program to this co-op, thus providing dyads with a more proximate high-quality food retail source and strengthening the local food economy by increasing local food shopping.

Methods

Survey collection

The current study includes a series of exploratory spatial network analyses to determine the distance to stores where participants in a previous study of prescription program effectiveness reported shopping, their ‘actual’ nearest stores, and a soon-to-be completed food co-op. Our study included a consecutive sample of 691 caregiver-child dyads across three large pediatric clinics in Flint that offered the pediatric prescription program (170 at Clinic 1, 202 at Clinic 2, and 310 at Clinic 3).

Following caregiver consent and child assent, caregiver-child dyads completed a series of demographic and survey questions that included information related to their addresses and primary food stores. A trained research assistant collected data with dyads during regular clinic visits while inside identified patient rooms at each partnering clinic. All self-reported data were collected using a secure digital platform (Research Electronic Data Capture) available to dyads at each clinic.

Spatial analysis

Once all data were collected, addresses were geocoded to their exact location. The dataset was manually scoured to identify the grocery stores where people shopped. Store addresses were then identified and geocoded, and store IDs were assigned to each store in the participant database and the store database.

A series of network analyses were then run to determine the distance from every individual to every store where they reported shopping and linking on store IDs. This entailed using the New Route function in ESRI’s Network Analyst extension (Environmental Systems Research Institute 2023), with home locations and associated stores assigned as ‘stops’ in the route. Closest Facility analyses were also run on each caregiver to determine the three nearest grocery stores to where they lived, as well as the distance to a food co-op discussed later. After each calculation, the route layer was joined to the original dataset via person IDs, resulting in a dataset highlighting store IDs and distances for actual and nearest stores.

Results

Descriptive statistics

Overall, 691 caregivers were present in the initial dataset. Of these, eight did not have an address available; thus, we worked from an initial sample of 683 caregivers. Overall, 398 caregivers lived in the city of Flint (360 of whom provided a store), while 285 lived outside the city (267 of whom provided a store). In total, 627 caregivers mentioned at least one store where they regularly shopped. Overall, 108 caregivers shopped at two stores (64 of 360—or 18.6%—in the city of Flint and 44 of 267—or 16.5%—outside the city), while just 37 caregivers (5.9%) noted regularly shopping at three or more stores.

Overall, 740 store mentions were given by the 627 caregivers in the dataset. Wal-Mart was the most frequently mentioned store (299 people, or 47.7% of the sample), followed by Kroger (280 people, or 44.7%) and Meijer (94 people, or 15.0%). No other stores were mentioned by more than 4.6% of the sample. Only 6.0% of the sample overall mentioned shopping at local stores (including Landmark, Great Giant, Fresh Choice, VG’s, Hutchinson’s, and others).

Spatial analysis

Flint residents tended to live slightly closer to the nearest store where they shopped, at 6017 m (or 3.74 miles); caregivers outside the city were 6736 m (or 4.18 miles) from their nearest store. Conversely, while only 4.4% of Flint residents lived within a mile of where they shopped, 12.8% of caregivers outside the city lived within a mile of their store.

When considering the second nearest store where people shopped, Flint residents were 7211 m (or 4.48 miles) away, while caregivers outside the city were substantially farther, at 9452 m (or 5.87 miles) away. These numbers increased to 8693 m (or 5.40 miles) and 11,684 m (or 7.26 miles) for the third nearest store.

Even so, when considering the distances from people to their actual three nearest stores, the distances were considerably shorter. Flint residents lived an average of 1620 m (or 1.01 miles) from their nearest store, while people outside the city lived 3222 m (or 2.00 miles) away. These distances increased to 2358 m (or 1.47 miles) and 5126 m (or 3.18 miles) for the second nearest store and 3344 m (or 2.08 miles) and 6151 m (or 3.82 miles) for the third nearest store.

The percentage of people who live within a mile of a grocery store is substantially higher than the percentage of people who shop at them: 55.5% of Flint residents and 32.5% of caregivers outside the city are within a mile of a store.

Despite living closer to stores than people outside the city, Flint residents on average travel 4394 m (or 2.73 miles) farther than they need to reach a grocery store. This is greater than the overshoot distance for people outside the city of 3424 m (or 2.13 miles).

Table 1 illustrates the percentage of people for whom the actual stores they shop at are from one of the following chains (Kroger, Meijer, and Wal-Mart). In total, 84% of people who live in the city of Flint shop at a chain as their closest store, and 89% of people outside of Flint shop at a chain as their closest. Overall, 96% of both sample shop in at least one chain. Overall, 49% of Flint residents shop at a Kroger, 20% shop at a Meijer, and 56% shop at a Wal-Mart. These numbers are broadly similar among people outside of Flint.

Table 1. Percentage of caregivers shopping at various store types, sorted by the four nearest stores.

	Closest	2nd Closest	3rd Closest	4th Closest	Any
Flint
Any Chain	84.4%	93.8%	95.7%	100.0%	95.7%
Kroger	37.9%	12.5%	4.3%	0.0%	49.5%
Meijer	4.4%	18.8%	26.1%	42.9%	20.1%
Wal-Mart	34.9%	48.4%	47.8%	14.3%	55.5%
n	398	64	23	7	398
Outside of Flint
Any Chain	89.0%	97.7%	92.9%	100.0%	96.1%
Kroger	38.5%	11.4%	21.4%	25.0%	47.3%
Meijer	12.7%	20.5%	28.6%	25.0%	23.0%
Wal-Mart	32.2%	47.7%	21.4%	25.0%	46.6%
n	283	44	14	4	283

When considering the kinds of stores that are closest to where people live, the landscape changes quite a bit (Table 2). In Flint, only 33% of people are closest to a chain store (this figure is 64% outside of Flint). Very few people live closest to a Wal-Mart (0% in Flint and 4% outside). This is easily contrasted with how many people shop at Wal-Mart above. Overall, 69% of Flint residents have a chain store among their nearest three stores, while 92% of people outside of Flint have a chain store among their nearest three.

Table 2. Percentage of caregivers closest to various stores, sorted by the three nearest stores.

	Closest	2nd Closest	3rd Closest	Any
Flint
Any Chain	32.7%	45.0%	31.4%	68.8%
Kroger	12.1%	12.6%	5.3%	29.9%
Meijer	6.0%	1.8%	2.3%	10.1%
Wal-Mart	0.0%	0.0%	2.8%	2.8%
n	681
Outside of Flint
Any Chain	64.0%	70.3%	60.4%	92.2%
Kroger	33.6%	28.6%	13.8%	72.8%
Meijer	10.2%	6.4%	3.2%	18.0%
Wal-Mart	3.5%	6.4%	17.7%	26.9%
n	681

Table 3 shows how many people are shopping ‘optimally’, with respect to shopping at their nearest stores. Only 9% of Flint residents and 27% of people outside of Flint shop at their nearest store. 19% of Flint residents shop at their second closest store, while 48% people outside the city shop at their second closest store.

Table 3. Percentage of caregivers shopping at their nearest stores.

	Closest	2nd Closest	3rd Closest
Flint
Closest Shopped Store	8.9%	18.6%	23.6%
2nd Closest Shopped Store		1.6%	4.7%
3rd Closest Shopped Store is 3rd Closest			4.3%
Outside of Flint
Closest Shopped Store	27.2%	47.5%	61.5%
2nd Closest Shopped Store		18.6%	23.3%
3rd Closest Shopped Store is 3rd Closest			0.0%

The North Flint Food Market

As shown above, most caregivers whose children are enrolled in the prescription program do not shop at locally owned grocery stores. The reasons for this tend to revolve around higher prices and poorer quality of goods and have been discussed at more length in a general sample of Flint residents in Sadler et al. (2019). An additional likely reason for the lack of shopping at local stores is that prescription program vouchers have only been able to be redeemed at the Flint Farmers’ Market or the Flint Fresh local food hub (which provides a ‘veggie box’ program).¹ This itself has proven to be a hurdle for engagement in the program, as redemption rates hover between 30 and 40% (personal communication, program manager). Many caregivers, whose children receive prescriptions from their pediatricians, continue to request additional redemption sites that are open every day and easy to access. Recently, program staff and partnering pediatric clinics have discussed the need to expand the prescription program to a local food store to improve redemption rates.

Connecting this program to a larger chain store such as Meijer, Kroger, or Wal-Mart is attractive from the perspective of increasing redemption rates, and after this research was conducted our team established a connection with Meijer to bring this program to at least four area stores. These stores have been slow to begin participating because of the cost of modifying their point-of-sale systems to include the prescription program vouchers. Additionally, considerable economic leakage rates can be observed by these shopping habits. Considering that 96% of the sample shops at chain stores – which by design send most profits out of the community – a prescription program model that recaptured some of these sales could be a double benefit for participants in the program and food retailing in the city.

In a study of Flint’s farmers’ market in 2011, researchers found an economic multiplier effect of 1.39, meaning for every dollar spent at the market, additional 39 cents were recycled into the community (Sadler et al. 2013). Converting a portion of caregivers from this study (and new caregivers going forward) to shop at a local food store could lead to thousands of dollars of monthly sales being recycled in the community at a rate far higher than could be found if all of these individuals continued to shop at a chain store.

The North Flint Food Market (co-op) potentially serves both of these needs – of having an independent retailer to partner with and indirectly serving the purpose of decreasing the economic leakage rate of food retail in the community. The co-op was conceived by a community development corporation after two major grocery chains (a Kroger and a Meijer) closed in 2014 and 2015 on this side of town. One goal of the co-op is providing better quality and lower priced foods than are currently available at the other existing grocery stores in north Flint (all of which are also locally/independently owned). The idea of a co-op was promoted to encourage buy-in from the local community and strengthen the possibility that a store could succeed here.

Because of the ownership model of this store, our team wondered if this may provide a stronger rationale for them being a part of the prescription program. To get an indication of how convenient the co-op may be for caregivers, we also calculated the distance from our caregivers to the store’s site (shown in Figure 1, along with the former locations of the now-closed Kroger and Meijer and jittered locations of our current prescription program enrollees). Note that this store was not open at the time of writing, but that the thought was that it could provide an important new source of healthy food for caregivers.

Figure 1. Locations of grocery stores and program participants.

As shown in Table 4, the co-op will be closer than the nearest shopped at store for 41% of Flint residents and 22% of the out-county residents. Among the sample who shopped at more than three stores, the co-op would be closer than the third farthest store for 70% of people in Flint and 64% of people outside of Flint. When considering all store options, the co-op would be the closest store for 8% of Flint residents and 5% of people outside of Flint. It would be within the three nearest stores for 18% of Flint residents and 9% of people outside of Flint.

Table 4. Percentage of caregivers for whom the North Flint Food Market would be closer than their currently shopped-at or nearest (actual) stores.

	Closer	Closer than 2nd	Closer than 3rd
Flint
As Shopped at	41.1%	60.9%	69.6%
	360	64	23
Actual	8.3%	9.4%	17.5%
	360	360	360
Outside of Flint
As Shopped at	21.5%	39.5%	64.3%
	265	43	14
Actual	5.3%	8.7%	8.7%
	265	265	265

Discussion

Healthy food prescription programs offer an important new dimension to promoting healthy food consumption among vulnerable populations, especially those with children. The complexities of offering such programs in our study area necessitated their initial deployment at farmers’ markets and local food stores, despite the fact that most people shop at conventional chain stores. In this paper, we illustrate the types of and distances to the stores where people enrolled in a prescription program in Flint shop, compared those locations to their nearest stores, and illustrated the potential of a new co-op in improving food access via the prescription program for participants.

Most notably, we found that the vast majority of participants shopped exclusively at chain stores, despite most bypassing multiple stores closer to their homes. This accords with past work indicating that many locally owned stores have higher prices, poorer quality goods, and/or uninviting atmospheres (Sadler et al. 2019, Saxe-Custack et al. 2019b) – and thus that Flint residents are bypassing their nearest stores. We also found that Flint residents are bypassing stores more frequently than people living outside the city. Both present potential complications for promoting shopping at locally owned, smaller-scale stores.

Our sample traveled slightly farther for groceries than an earlier focus group-based sample of Flint adults (3.74 miles compared to 3.38 miles) (Sadler et al. 2019). Furthermore, fewer shopped within one mile of their homes (4.4% compared to 11%). These remarkably low numbers are in contrast to the objective finding that 55.5% of Flint residents and 32.5% of people outside the city live within a mile of a store.

Our findings that the new North Flint Food Market co-op would be closer to many caregivers than their currently shopped-at stores are important in multiple ways: offering another redemption site for the prescription program – and one in a grocery store – could help improve redemption rates and, therefore, benefits accrued from the program. It could also help in the promotion of local economic development via shopping at smaller stores (recognizing of course that many shoppers still prefer larger chain stores and will have the option to redeem at such stores as well). Furthermore, the local farmers’ market is a year-round market but is open only 3 days per week (Tuesday, Thursday, and Saturday). Caregivers have requested that the prescription program partner with a food store that offers extended hours of operation, particularly for working parents. Finally, and in addition to creating employment opportunities and job training for community members, the co-op is invested in providing high-quality fresh foods. Together, this suggests that the co-op could be a suitable partner for the expansion of fresh produce prescriptions. While it is not a chain grocery store, it does offer a full range of products and is situated in a neighborhood where food access has been of concern to many in the past (Sadler et al. 2019). And given that our program will also be expanding to Meijer stores in the area (which are not spatially coincident with the co-op), there is an exciting opportunity to study the ongoing impact of the rollout of this program across the community.

Our study is not without limitations. Foremost, prescription programs remain in their early stages in many places; the methods for increasing redemption and maintaining behaviors initiated through the programs need refinement. Second, our focus is narrowly on the possibilities for improving healthy food access and consumption among caregivers whose children attend three pediatric clinics in Flint, Michigan. Thus, we acknowledge that our findings are not generalizable beyond this study population (either across the Flint region or more broadly). Similar studies should be conducted with other programs to determine consistency in findings. Third, one caregiver and one child completed surveys in the current study; therefore, households with multiple food shoppers may not have been adequately represented.

Even so, our work is broadly useful for researchers and practitioners working in nutrition, food access, and economic development. Our spatial analysis techniques are easily replicable, and others can illustrate similar patterns of improved food access via the expansion of prescription programs to new food retail formats. Since we have tracked and recorded prescription redemption patterns across clinics for approximately 5 years, we are able to assess changes in redemption resulting from the addition of the North Flint Food Market. This information is critical to our understanding of participant engagement in fresh produce prescription programs.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development; Michigan Health Endowment Fund; Community Foundation of Greater Flint.

Notes on contributors

Richard C. Sadler

Richard C. Sadler is a medical geographer with expertise in environmental science, GIS, food systems planning, and land use policy in legacy cities. His research integrates urban planning and public health topics related to neighborhood effects on health. Recently, this has revolved around uncovering elements of structural racism in the housing environment and determinants of urban development that exacerbate racial and socioeconomic inequality. Sadler combines spatial analysis and community-based participatory research approaches. Throughout his work, the overarching goal is to strengthen the understanding between the built environment and health behaviors/outcomes with the goal of shaping land use policy to build healthier cities.

Amy Saxe-Custack

Amy Saxe-Custack is a Registered Dietitian and Assistant Professor in the Charles Stewart Mott Department of Public Health at Michigan State University. She serves as the Nutrition Director for the Pediatric Public Health Initiative, a joint effort between Michigan State University and Hurley Children’s Hospital, to address the impact of Flint’s population-wide lead exposure on children. A long-time resident of the Flint community, Dr. Saxe-Custack is dedicated to the evaluation and expansion of nutrition programs that target children and families living in Flint, with a particular focus on improving access to fresh foods.

Notes

1. The locations of the North Flint Food Market co-op, the Flint Farmers’ Market and the Hurley Children’s Center are shown in Figure 1, and images of their exteriors are shown in Figures 2 and 3.

Figure 2. Side and Front views of the North Flint Food Market.

Figure 3. Front view of the Flint Farmers’ Market and Hurley Children’s Center.