The Veggie Van: Customer characteristics, fruit and vegetable consumption, and barriers to healthy eating among shoppers at a mobile farmers market in the United States

Appetite 133 (2019) 279–285

Contents lists available at ScienceDirect

Appetite journal homepage: www.elsevier.com/locate/appet

The Veggie Van: Customer characteristics, fruit and vegetable consumption, and barriers to healthy eating among shoppers at a mobile farmers market in the United States

T

Kelly R. Ylitaloa,∗, Christina Duringa, Katherine Thomasa, Kelly Ezellb, Patrick Lillardb, Joel Scottb a b

Public Health Program, Baylor University, Waco, TX, USA World Hunger Relief, Inc., Elm Mott, TX, USA

A R T I C LE I N FO

A B S T R A C T

Keywords: Mobile farmers market Veggie van Healthy eating Community nutrition program

Objective: Mobile farmers markets may improve local food environments by increasing access to healthy food, yet research is limited. The purpose of this study was to describe customer characteristics and barriers to healthy eating among customers at a mobile farmers market called the Veggie Van. Design: In 2016, a customer intercept design was used to survey English-speaking Veggie Van customers (n = 192; 70.5% survey response rate) aged ≥18 years on sociodemographic and health characteristics, normal daily consumption of fruit and vegetables (F/V) using the Health Information National Trends Survey screener, food acquisition and purchasing habits, and potential barriers to healthy eating. We compared customers to service area neighborhood residents. Within customers, we compared first-time and repeat customers, and those with low and high F/V consumption. Results: Veggie Van customers were more likely to identify as non-Hispanic white and have a bachelor's degree than neighborhood residents. Participants were mostly female (76.0%) and non-Hispanic white (53.7%). Approximately half (45.0%) were first-time customers and many (41.7%) did not meet F/V consumption recommendations. In the total sample, cost was the most frequently reported barrier to healthy eating. Among repeat customers, those with low F/V consumption were more likely to report cost as a barrier than those with high F/V consumption (p = 0.02). Only 8.9% reported no transportation to buy healthy food. Conclusions: Veggie Van customers may not represent neighborhood residents. Although few participants met F/ V recommendations, most had transportation to buy healthy food. Mobile markets have lower overhead costs and greater flexibility than traditional stores and can address geo-spatial barriers to food access, but should ensure that they are serving target customers.

1. Introduction Unhealthy eating behaviors are major contributors to the obesity epidemic in the United States (Economos et al., 2015). Less than 15% of adults in the United States meet daily recommendations for fruit and vegetable consumption (Moore et al., 2015) and poor diet is now the leading risk factor for disability-adjusted life years, even ahead of tobacco (Murray et al., 2013). Food environments in low-income and race/ethnic minority neighborhoods are often characterized by limited access to healthy food and increased access to low-nutrition energydense food (Cubbin, Haden, & Winkleby, 2001; Deaton & Lubotsky, 2013; Morland et al., 2002; Walker, Keane, & Burke, 2010). Since low-

income and race/ethnic minority groups are disproportionately affected by obesity and obesity-related chronic conditions in the United States, improving the food environment by increasing access to healthy foods like fruit and vegetables has the potential to mitigate chronic health problems. However, community-level food access interventions are often costly and not well understood. One potential strategy to improve the local food environment is a mobile farmers market. Mobile farmers markets are typically renovated buses or vans that sell local, affordable fruits and vegetables on a rotating schedule, serve multiple sites within a weekly schedule, and focus on areas where substantial proportions of residents have limited access to affordable and nutritious food (Robinson et al., 2016).

Abbreviations: USDA, United States Department of Agriculture; WHRI, World Hunger Relief, Inc.; SNAP, Supplemental Nutritional Assistance Program; HINTS, Health Information National Trends Survey; BMI, body mass index ∗ Corresponding author. Baylor University, Department of Public Health, One Bear Place # 97343, Waco, TX, 76798-7343, USA. E-mail address: [email protected] (K.R. Ylitalo). https://doi.org/10.1016/j.appet.2018.11.025 Received 31 January 2018; Received in revised form 4 September 2018; Accepted 27 November 2018 Available online 28 November 2018 0195-6663/ © 2018 Elsevier Ltd. All rights reserved.

Appetite 133 (2019) 279–285

K.R. Ylitalo et al.

the activity of interest, and have been used previously in food purchasing and mobile farmers market research (Caspi et al., 2017; Kraschnewski et al., 2014). English-speaking customers aged ≥18 years were eligible to participate in the survey. Data collection occurred for approximately eight months, January through May and then September through November 2016. On average, the Veggie Van served 175 customers, measured by sales transactions, per month. Sales transactions were not tracked with any personal, family, or household identifiers of the purchasers, so first time and repeat customers could not be differentiated. For the purpose of the survey, repeat customers were asked to self-defer from survey participation if they had previously participated in the survey during the survey period. Program staff and trained volunteers explained the survey to customers and obtained verbal informed consent. As participation incentive and compensation, survey participants received a $5 gift card to the Veggie Van to purchase additional fruit and vegetables. Overall, 192 customers completed the survey and the response rate was 70.5%. We did not collect information on personal characteristics of customers who declined to participate in the survey.

National advocates contend that mobile markets have lower overhead costs and greater flexibility than traditional brick and mortar stores, and mobile markets can address both economic and geo-spatial barriers to food access (Best & Johnson, 2016; Robinson et al., 2016). Numerous policy and practice recommendations have identified mobile markets as a potential solution to healthy food access, particularly in low-income communities (Centers for Disease Control and Prevention, 2014; Johnson Foundation, 2014; The Network for Public Health Law, 2013), and between 2009 and 2015, the United States Department of Agriculture (USDA) invested over $1 billion in food infrastructure projects, including mobile food markets (US Department of Agriculture, 2013). Several popular mobile farmers markets include Arcadia's Mobile Market in Washington, D.C (Arcadia Center for Sustainable Food & Agriculture, 2017). and Fresh Truck in Boston (Fresh Truck, 2015), among others, which have been in operation for at least four years. However, multiple mobile farmers markets have closed due to operational constraints or have experienced substantial challenges to recruiting community partners (Best & Johnson, 2016; Indiana University Health, 2017; Tripicchio et al., 2017). Research on mobile farmers markets is limited. Understanding mobile farmers market customers may assist in future planning efforts to meet customer needs, tailor nutrition education opportunities with customers, and sustain mobile farmers market programs to reduce food access inequalities in the United States. In our mid-sized, Southern community, a mobile farmers market called the Veggie Van was implemented in 2015 and customers were evaluated in 2016 with a survey. The primary purpose of this research was to characterize Veggie Van customers in terms of sociodemographics, healthy eating behaviors and barriers, and motivations to use the market, and describe how these personal characteristics may influence mobile market shopping frequency and fruit and vegetable consumption.

2.2. Measures Key survey variables were sociodemographic characteristics, including age, sex, race/ethnicity (Hispanic, non-Hispanic white, nonHispanic black, or Other), highest level of education completed (High school diploma/GED or less, Some college, or College degree), household size, and receipt of benefits (SNAP, WIC, or free/reduced lunch for children in the home). Health variables included self-reported health (Idler & Benyamini, 1997), and self-reported height (in feet and inches) and weight (in pounds) were converted to body mass index (BMI). Respondents were asked to report their normal daily consumption of fruit and vegetables using questions from the Health Information National Trends Survey (HINTS) (Health Information National Trend Surveys, 2013). Fruit intake and vegetable intake were assessed separately with the following answer choices: none, ½ cup or less, ½ to 1 cup, 1 to 2 cups, 2 to 3 cups, 3 to 4 cups, and 4 or more cups. Compared to a 24-h food frequency recall, the 2-question HINTS fruit and vegetable screener has moderate validity (Pearson correlation coefficient is 0.51 and 0.32, respectively) and strong test-retest reliability (Intraclass correlation coefficient is 0.59 and 0.60, respectively) (Yaroch et al., 2012). Customer respondents were also asked about frequency of eating together with household members, frequency of eating takeout or fast food, and usual method of transportation for food shopping (personal vehicle, family or friend's vehicle, public transit (bus), bicycle, or walk). As part of survey development, potential household barriers to healthy eating were identified using empirically supported barriers in the literature as well as those identified by community leaders (Eikenberry & Smith, 2004; Fulkerson et al., 2004; Kamphuis et al., 2007; Yeh et al., 2008). Respondents were asked to Agree or Disagree with seven potential healthy eating barriers, including cost, time, location, transportation, knowledge, preparation skills, and taste (Eikenberry & Smith, 2004; Fulkerson et al., 2004; Kamphuis et al., 2007; Yeh et al., 2008). Finally, participants were asked about usage of the Veggie Van, including frequency of purchases and motivations for using the Van.

2. Materials and methods 2.1. Study design The Veggie Van in central Texas was implemented by a nonprofit organization, World Hunger Relief, Inc. (WHRI), to increase access to fresh, local produce throughout several neighborhoods. WHRI is a sustainable working farm in central Texas committed to the alleviation of hunger locally and globally (World Hunger Relief, 2016). Currently, WHRI trains interns and hosts several thousand individuals each year for educational programs focusing on sustainable agriculture, environmental responsibility, and hunger issues (World Hunger Relief, 2016). The Veggie Van service area, which included four zip codes in central Texas, was identified using community-wide input prior to implementation and data that showed these areas had a large proportion of the population that lived more than one mile from a supermarket. All zip codes had proportionately more households that used food stamps/Supplemental Nutritional Assistance Program (SNAP) compared to the state of Texas. The proportion of families below poverty ranged from 19.9 to 40.1%, over 50% of residents in the service area identified as a race/ethnic minority, and most residents spoke English (Table 1). Ten weekly stops, ranging from 45 min to 2 h in duration, occurred within a seven-mile radius and within these four zip codes on a rotating schedule. Locations included a medical center, community center, elementary school, and residental facility for older adults. The Veggie Van sold fresh, whole produce and select locally produced food products (e.g., honey) from their own garden or other local farms (see Fig. 1). The Veggie Van accepted SNAP food benefits. SNAP customers and older adults received half-price discounts, but otherwise pricing was consistent with supermarket chain store produce pricing in central Texas. Customer intercept interviews were used to conduct assessments of Veggie Van customers in 2016. Intercept surveys are commonly used to assess visitor experiences or satisfaction while people are engaged in

2.3. Statistical analysis Descriptive statistics, including means (standard deviation) for continuous variables and proportions for categorical variables, were calculated for all variables. We compared customer characteristics between first-time Veggie Van customers and repeat Veggie Van customers using student t-tests for continuous variables and chi-square tests for categorical variables. We categorized fruit and vegetable consumption using cutpoints from the 2015–2020 Dietary Guidelines for Americans, which recommends eating 2 ½ cups of vegetables and 2 cups of fruit per 280

Appetite 133 (2019) 279–285

K.R. Ylitalo et al.

Table 1 Characteristics of Veggie Van service area in central Texas, 2012–2016 5-Year American Community Survey Estimates, United States Census Bureau.

Total Population, n Race/ethnicity, % Hispanic Non-Hispanic white Non-Hispanic black Non-Hispanic other Adult language skillsa, % English language only Spanish language English very well or well English not well or not at all Other language English very well or well English not well or not at all Bachelor's degree or higher, % Households with food stamps/SNAP benefits, % Families below poverty, %

76704

76706

76707

76708

Total Service Area

Texas

7984

37,062

16,308

26,220

87,574

26,956,435

10.4 8.0 80.6 < 1.0

28.4 53.9 13.1 4.6

49.0 24.1 24.8 2.1

36.7 47.3 13.1 3.0

33.0 42.2 21.4 3.3

38.6 43.4 11.6 6.3

94.0 5.4 92.9 7.1 0.6 61.1 38.9 9.8 38.6 34.9

77.6 16.8 74.7 25.3 5.6 93.5 6.5 17.7 13.9 40.1

66.9 32.7 64.0 36.0 0.4 87.0 13.0 10.4 31.4 31.3

74.2 24.0 66.6 33.4 1.8 84.7 15.3 18.2 13.5 19.9

– – – – – – – – – –

64.9 29.0 72.1 27.9 6.1 85.1 14.9 28.1 13.1 16.7

a Included adults aged ≥18 years and measured as language spoken in the home. Other languages include Indo-European, Asian/Pacific Island, and other languages.

members, and weekly frequency of takeout or fast food. Many participants (41.7%) reported low daily fruit and vegetable consumption (Table 2). We also assessed frequency of Veggie Van visits and reasons for visiting the Veggie Van. Almost half (45.0%) of survey respondents were first-time visitors to the Veggie Van, and the remaining 55.0% were repeat customers: 22.0% typically visited once or more per week, 19.4% visited once to twice per month, and 13.6% visited less than once per month. Reasons for visiting the Van included to support a local business (37.5%), to buy healthy foods (48.4%), convenience (35.4%), and quality of produce (30.2%). Compared to repeat customers, firsttime customers were less likely to have college degrees (p = 0.04) and more likely to use free or reduced lunch benefits if there were children in the home (p = 0.02), but were otherwise similar in terms of age, sex, race/ethnicity, household size, and SNAP benefit status. First-time and repeat customers reported different reasons for utilizing the Veggie Van. Repeat customers were more likely to report supporting a local business (p = 0.01), convenience (p < 0.01), and quality of produce (p < 0.01) compared to first-time customers, but first-time and repeat customers were equally as likely to report buying healthy food as a reason to shop at the Veggie Van (Table 2). Among all customer respondents (n = 192), the most common barrier to healthy eating was cost: almost half (49.0%) reported that it costs too much to eat healthy food. Participants who reported low fruit and vegetable consumption were significantly more likely to report cost (57.5% vs. 42.9%; p = 0.049), and preparation/cooking skills (22.5% vs. 8.9%; p = 0.01) as barriers to healthy eating compared to participants who reported higher fruit and vegetable consumption. Barriers to healthy eating did not differ by frequency of van visit at the α = 0.05 level, but daily fruit and vegetable consumption may have modified this relationship. Among repeat customers, those who reported low fruit and vegetable consumption were significantly more likely to report cost, knowledge, and preparation/cooking skills as barriers to health eating compared to repeat customers who reported higher fruit and vegetable consumption (Table 3).

Fig. 1. The Veggie Van at a weekly stop in central Texas.

day for Healthy U.S.-Style Eating Patterns (U.S. Department of Health and Humas Services and U.S. Department of Agriculture, 2015). Consistent with other survey work using HINTS, we created a composite dietary variable that included consumption of fruit and/or vegetables (Redmond, Baer, Clark, Lipsitz, & Hicks, 2010). Accordingly, we categorized individuals who reported 4 cups or less of fruit and/or vegetables as “low fruit and vegetable consumption” and 4 ½ cups or more of fruit and/or vegetables as “high fruit and vegetable consumption. We compared barriers to healthy eating between first time and repeat customers, and by high and low fruit and vegetable consumption. Data were analyzed using SAS v9.4 (SAS Institute Inc., Cary, NC, USA) and statistical significance was defined at the two-sided α = 0.05 level. 3. Results Participants (n = 192) were, on average, 47 years of age, threefourths female (76.0%), and mostly non-Hispanic white (53.7%). Onethird (34.4%) of participants reported having a college degree. Approximately one-third (30.5%) reported using SNAP benefits, and almost half (47.2%) of households with children reported receiving free or reduced school lunches for children in the home. Average BMI was 30.1 kg/m2, indicating that many participants were obese, and less than half reported excellent or very good self-rated health. We assessed several different eating behaviors, including usual fruit and vegetable consumption, weekly frequency of sharing a table with household

4. Discussion The primary purpose of this study was to describe Veggie Van mobile farmers market customers and identify potential needs or barriers related to healthy eating, which could ultimately inform tailored nutrition education opportunities with customers. Veggie Van customers who responded to the survey were mostly non-Hispanic white, female, obese (BMI≥30 kg/m2), and had a college degree. Noteably, many 281

Appetite 133 (2019) 279–285

K.R. Ylitalo et al.

Americans recommend eating 4 ½ cups (2 ½ cups of vegetables and 2 cups of fruit) per day. Our findings are supported by national data, which indicate that many Americans do not meet current guidelines (Moore et al., 2015). In a recent Health Information National Trends Survey, a national probability sample of adults in the United States, 56.0% (95% CI: 53.6–58.3) of adults consumed 2 ½ cups or less of fruits and vegetables (U.S. Department of Health and Humas Services and U.S. Department of Agriculture, 2015). Findings from mobile market interventions are mixed. A New York study designed to increase fruit and vegetable intake among senior adults with a Veggie Mobile program detected no increase in fruit consumption and a small but statistically significant increase in vegetable consumption; however, over half of the increase in vegetables was potatoes (AbuSabha, Namjoshi, & Klein, 2011). Indiana University researchers studied Garden on the Go customers and found that the proportion of participants who ate five or more servings of fruit and vegetables per day increased during the study period (Indiana University Health, 2012). Among schoolchildren in Oakland, researchers compared fruteros (mobile fruit and vegetable vendors) sales with those of nearby unhealthy competitor vendors and detected an increase in fruit and vegetable sales and a decrease in the sales of nearby unhealthy food vendors (Tester, Yen, & Laraia, 2012). In our study, repeat Veggie Van customers appeared marginally more likely to consume more vegetables than first-time customers. Although the cross-sectional nature of our customer intercept survey must be interpreted with caution due to the possibility of reverse causation, these findings are supported by other work with focus groups in four diverse, urban areas in the United States, which found people who used mobile food markets reported more fruit and vegetable consumption than people who did not shop at mobile markets (Zepeda, Reznickova, & Lohr, 2014). Veggie Van respondents in our study reported multiple reasons for shopping at the Veggie Van, including to support a local business, to buy healthy foods, convenience, and quality of produce. In particular, repeat customers to the Veggie Van were more likely to indicate the personal importance of supporting a local business as a reason to use the Van. In general, qualitative research has demonstrated positive community support and buy-in for mobile produce units. Researchers from the Arcadia Mobile Market reported that customers provided positive feedback on items bought through the market and were enthusiastic about the improved access to healthy food (Arcadia Center for Sustainable Food & Agriculture, 2012). Other work has shown that affordability, convenience, and value were all indicated as important variables that influence the decision to purchase from these mobile produce vans (Zepeda et al., 2014). In our study, the high prevalence of obesity and low self-rated health indicates potential need for health education at mobile markets. To date, limited quantitative and qualitative research has yet to demonstrate the translation of increased fruit and vegetable consumption from a mobile produce unit to quantifiable health improvements. Although Garden on the Go participants at Indiana University reported increases in fruit and vegetable consumption and perceived health status, there were no statistically significant changes in BMI, HbA1c levels, or the proportion of participants with hypertension during the 6 month study (Indiana University Health, 2012). Other work has shown many customers have poor self-efficacy in the preparation of fruits and vegetables, and in particular seasonal fruit and vegetables (HaynesMaslow, Auvergne, Mark, Ammerman, & Weiner, 2015). These findings are supported by our study, indicating that repeat customers with low fruit and vegetable consumption were more likely to report lack of preparation/cooking skills than those with higher fruit and vegetable consumption. Lack of knowledge or confidence hinders the sustained consumption of fruit and vegetables necessary to result in quantifiable health improvements. Health education addressing basic cooking skills, preparation of seasonal produce in a culturally sensitive manner, and strategies to decrease time required to cook at home are likely needed to increase and sustain the consumption of healthy food, and mobile

Table 2 Characteristics of Veggie Van customers, 2016. Sociodemographics

Total (n = 192)

First time customer (n = 86)

Age, years (std) 47.1 (17.6) 46.7 (18.1) Sex, % Male 24.0 24.4 Female 76.0 75.6 Race/ethnicity, % Non-Hispanic white 53.7 52.3 Non-Hispanic black 23.4 26.7 Hispanic 16.7 11.6 Other 6.3 9.3 Bachelor's degree or 34.4 25.6 higher, % Average household size, 2.7 (1.6) 2.6 (1.5) n (std) Households with 38.0 32.6 children, % Benefits 37.5 31.4 SNAP, % 30.5 31.3 WIC, % 10.2 8.4 Free or reduced lunch, 47.2 29.6 a % Health characteristics 30.1 (7.9) 30.3 (8.3) BMI, kg/m2 (std) Self-rated health, % Excellent/Very good 42.7 43.0 Good/Fair/Poor 57.3 57.0 Eating behaviors Fruit consumption per day, % None 5.2 4.7 1/2 cup or less 15.6 16.3 1/2 to 1 cup 24.0 26.7 1 to 2 cups 30.2 29.1 2 to 3 cups 13.0 15.1 3 to 4 cups 6.3 4.7 4 or more cups 5.7 3.5 Vegetable consumption per day, % None 3.1 3.5 1/2 cup or less 7.9 12.8 1/2 to 1 cup 13.6 18.6 1 to 2 cups 38.2 32.6 2 to 3 cups 15.7 16.3 3 to 4 cups 13.6 11.6 4 or more cups 7.9 4.7 Frequency eating at a table with household, % None 20.0 20.5 1–4/week 27.2 28.9 5 or more/week 52.8 50.6 Frequency of takeout or fast food, % None 24.6 22.4 1–2/week 52.4 48.2 3 or more/week 23.0 29.4 Reasons for Veggie Van visit Support local business, 37.5 26.7 % Buy healthy foods, % 48.4 45.4 Convenience, % 35.4 24.4 Quality of produce, % 30.2 16.3

Repeat customer (n = 105)

P

47.1 (17.1)

0.86

23.8 76.2

0.92

54.3 21.0 21.0 3.8 41.9

0.14

0.04

2.8 (1.7)

0.37

42.9

0.15

42.9 29.8 11.5 57.8

0.10 0.82 0.49 0.02

30.0 (7.7)

0.81

42.9 57.1

0.98

5.7 15.2 21.0 31.4 11.4 7.6 7.6

0.75

2.9 3.9 9.6 42.3 15.4 15.4 10.6

0.08

19.8 26.0 54.2

0.88

26.7 56.2 17.1

0.13

46.7

0.01

51.4 44.8 41.9

0.40 < 0.01 < 0.01

Note: Missing data (n = 1) from frequency (first time vs. repeat customer) of van visit. a Proportion with free/reduced lunch benefits calculated among 72 families with children living in the household.

respondents perceived their health as good, fair, or poor rather than excellent or very good, indicating a likely prevalence of chronic diseases (Idler & Benyamini, 1997; Mavaddat et al., 2014). First-time and repeat customers appeared to differ in multiple reasons for using the Veggie Van. One of the main goals of a mobile farmers market is to increase access to fresh produce and, in turn, increase fruit and vegetable consumption. Many Veggie Van respondents in our study did not eat 4 cups of fruit or vegetables per day. The 2015–2020 Dietary Guidelines for 282

Appetite 133 (2019) 279–285

K.R. Ylitalo et al.

Table 3 Proportion (%) of agreement with healthy eating statements for total sample, by fruit and vegetable consumption, and by frequency of Veggie Van visit, 2016. All customers (n = 192)

It costs too much to eat healthy food. I know the difference between healthy and unhealthy food. I don't have time to think about eating healthy food. There is nowhere to buy healthy food near me. I do not have transportation to buy healthy food. I do not know what healthy foods are. I don't seem to have the skills to prepare or cook healthy food. Most healthy foods just don't taste that great. Buying local and organic food is important to me.

First time customer (n = 86)

Repeat customer (n = 105)

Overall

Low F/V

High F/V

p

Low F/V

High F/V

p

Low F/V

High F/V

p

49.0 96.4 19.9 17.9 8.9 9.4 14.6 16.2 82.3

57.5 93.8 23.8 20.0 11.3 10.1 22.5 21.3 77.5

42.9 98.2 17.1 16.4 7.2 8.9 8.9 12.6 85.7

0.05 0.13 0.26 0.52 0.33 0.78 0.01 0.11 0.14

53.9 89.7 23.1 15.4 5.1 0.0 23.1 23.1 76.9

48.9 97.9 14.9 17.0 8.7 12.8 14.9 17.8 80.9

0.65 0.11 0.33 0.84 0.52 0.03 0.33 0.21 0.66

62.5 97.5 22.5 25.0 17.5 20.5 22.5 20.0 80.0

38.5 98.5 18.8 15.9 6.2 6.2 4.6 12.5 89.2

0.02 0.73 0.64 0.25 0.07 0.03 0.01 0.30 0.19

Notes: Low fruit and vegetable (F/V) consumption defined as 4 cups or less of fruit and/or vegetables per day. P values calculated from chi-square tests or Fisher's exact test for small cell sizes.

2009). Future research on mobile farmers market customers should consider the role of the built environment, including transportation and walking, on food purchasing at mobile markets. Mobile markets may be particularly affected by transportation and neighborhood “walkability,” conceptualized as mixed land use, street connectivity, and higher residential density, which is more conducive to spending time in an area (Saelens, Sallis, Black, & Chen, 2003). We hypothesize that community walkability impacts food purchasing at the Veggie Van. For example, Walk Score is an open-source measure of walkability and ranges from 0 to 100, with lower Walk Scores indicative of car dependne and higher Walk Scores indicative of higher walkability (Duncan, 2013; Hirsch et al., 2013). Our community has poor walkability (Walk Score® = 34), meaning that our community is car-dependent and most errands require a car (Duncan, 2013; Hirsch et al., 2013). Although 20–40% of families in target zip codes live in poverty (Table 1), most Veggie Van customer respondents reported using a personal vehicle to travel to a grocery store; only 8.9% reported transportation as a barrier to healthy eating. Although some customers may not represent target residents, the availability of personal vehicles is supported by other survey work within our community, which found that only 12.2% of residents reported transportation as a barrier to healthy food consumption (Ylitalo, Umstattd Meyer, Stone, Doyle, & Curtis, 2016). The availability of personal vehicles in our community suggests that Veggie Van customers and residents of target neighborhoods do not have geo-spatial barriers to healthy food access, and would actually need to alter typical food purchasing habits to access our mobile farmers market. Highly publicized, popular mobile farmers markets in Boston and Washington, D.C., are located in some of the most walkable cities in the United States. Future development of mobile farmers markets in other, smaller communities like ours should consider community walkability and usual methods of transportation when selecting program locations. Overall, the Veggie Van experienced several challenges to sustaining the program within the first two years of implementation, including low customer numbers and low quantities of food sales. This observation is consistent with other reports that small customer bases and low sales challenge the independent financial viability of most mobile markets, and as such, many mobile markets are dependent on additional, external funding like grants and donations (Best & Johnson, 2016; Robinson et al., 2016). In addition, many notable mobile farmers markets have closed permanently or intermittently, including Garden on the Go in Indianapolis, Fresh Moves Mobile Food Market in Chicago, and others (Fresh, 2017).

markets could be one potential avenue for this type of education. More research is needed to understand the health effects of mobile markets, particularly among race/ethnic minority and low-income populations who reside in food deserts. 4.1. Limitations There are several limitations to our study. The original purpose of the Veggie Van was to increase access to healthy food among low-income neighborhoods with limited access to supermarkets. Although weekly Van locations were in these neighborhoods, we suspect that Veggie Van customers may not represent neighborhood residents. Our hypothesis is supported by Census data and our survey data related to educational attainment. According to Census data, 9.8–18.2% of the Veggie Van service area population have a bachelor's degree (Table 1), while 34.4% of customer survey respondents reported a bachelor's degree (Table 2). Unfortunately, we did not collect information on residential addresses and therefore cannot determine if customer respondents lived within the Veggie Van service area. Qualitative work has shown that people who live near mobile markets but do not shop there are not aware of them, have misconceptions about who can shop there, and may not trust outside vendors (Zepeda et al., 2014). Future surveys at mobile farmers markets should collect respondent information on residential zip code, and consider implementing surveys of non-shoppers in the immediate vicinity of the mobile market location(s) for comparison. Furthmore, we did not have resources to offer the survey in multiple languages. One-third of the service area identifies as Hispanic, and only 16.7% of survey respondents were Hispanic. Although many area residents speak Spanish, two-thirds or more of Spanish speakers in each zip code also speak English well or very well (Table 1). Nevertheless, we may have missed Spanish, Indo-European, Asian/Pacific Island, or other language speakers who were not able to participate and thus introduced selection bias in our study. Future work on mobile farmers markets should include multi-lingual research teams and surveys. Any survey, inlcuding ours, that aims to balance comprehensive information with participant time burden is subject to unmeasured variables. In our study, we did not collect data on Veggie Van purchase history or food security. Future studies may wish to include this information to characterize the customer population. Finally, all variables were measured by self-report. Other work has shown that individuals over-report their consumption of healthy foods (Andersen, Bere, Kolbjornsen, & Klepp, 2004), yet mobile market shoppers tend to eat more fruits and vegetables than non-shoppers (Zepeda et al., 2014). As such, it is difficult to know in which direction our results may be misclassified. A social-ecological perspective on health and disease posits the importance of the environment on human behavior (Saarloos, Kim, & Timmermans, 2009). The built environment plays a key role in shaping individual healthy eating behaviors (Walker et al., 2010; Saarloos et al.,

4.2. New opportunities Mobile farmers markets may wish to consider health education opportunitites. In our study, customers who reported low fruit and vegetable consumption were significantly more likely to indicate that personal preparation and cooking skills were barriers to healthy eating. 283

Appetite 133 (2019) 279–285

K.R. Ylitalo et al.

(K.R.Y. pilot grant). The funder had no role in the design, analysis, or writing of this article.

One pilot program demonstrated a short-term increase in nutrition knowledge through health education offered through a mobile farmers market (Ellsworth, Ernst, & Snelling, 2015). Other work has also shown that younger and low-income residents may be unsure of how to prepare or cook certain produce but express interest in receiving storage and cooking recommendations by produce vendors (Haynes-Maslow et al., 2015). In central Texas, the Veggie Van provided free recipe cards and instructions on how to use produce. Future work at mobile markets should consider expanding nutrition education to cooking classes, and measuring long-term changes in nutrition knowledge, eating behavior, and health outcomes. In the first two years of operation, Veggie Van sites included schools, parks, community recreation centers, medical facilities, apartment complexes, and elderly residential facilities. Locations were selected based on qualitative information gained from community input meetings prior to program implementation in 2015. After two years, the most successful Veggie Van sites included the local hospital and a lowincome retirement facility, where older adults received half-price discounts on produce and could use SNAP benefits. However, overall low customer numbers and exhaustion of pilot grant support have led to Veggie Van restructuring. A new focus of our Veggie Van is delivering fresh produce “prescriptions” to a large, local, federally-qualified health center, where patients who are at-risk for food insecurity and malnutrition receive vouchers for fresh vegetable boxes that include recipe cards. Future research will evaluate the longitudinal impact of fresh produce prescriptions delivered by our Veggie Van on fruit and vegetable consumption, health resiliency, and health management of patients. Other mobile markets may consider becoming involved with fruit and vegetable “prescriptions” at health clinics as well, which is supported by a growing body of literature. For example, in the Detroit area, the Fresh Produce program observed significant decreases in HbA1c over a 13-week produce prescription program for adult patients with type 2 diabetes at a federally qualified health center (Bryce et al., 2017). In Portland, a small feasbility study found that patients increased their variety of vegetable consumption during a 23-week community-supported agriculture program, also at a federally-qualified health center (Izumi et al., 2018). Connecting the food sector and the health care sector may be an exciting opportunity to increase fruit and vegetable consumption and support healthy behavior changes, but more work is needed to determine the long-term sustainability and impact on health outcomes.

Conflicts of interest None. Author contributions K.R.Y. designed the study, analyzed the data, and wrote the paper. C.D. and K.T. collected data, assisted with writing and editing, and approved the final version. K.E., P.L., and J.S. provided logistical support for data collection, helped with data interpretation, assisted with writing and editing, and approved the final version. Ethical standards disclosure The Baylor University Institutional Review Board determined that this survey project was exempt from full board review. Verbal informed consent was obtained from all participants. Acknowledgments Dr. Ylitalo gratefully acknowledges the dedication of undergraduate epidemiology students from PUBH 3351 and graduate epidemiology students from PUBH 5337 at Baylor University for their data collection efforts. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.appet.2018.11.025. References AbuSabha, R., Namjoshi, D., & Klein, A. (2011). Increasing access and affordability of produce improves consumption of vegetables in low-income seniors. Journal of the American Dietetic Association, 111, 1549–1555. Andersen, L. F., Bere, E., Kolbjornsen, N., & Klepp, K. I. (2004). Validity and reproducibility of self-reported intake of fruit and vegetable among 6th graders. European Journal of Clinical Nutrition, 58(5), 771–777. Arcadia Center for Sustainable Food & Agriculture (2012). Mobile markets: Applying the food Truck model to food access. http://arcadiafood.org/sites/default/files/files/ arcadia%20%20mobile%20market%20report%20Digital.pdf. Arcadia Center for Sustainable Food & Agriculture (2017). Mobile markets. http:// arcadiafood.org/programs/mobile-market. Best, A. L., & Johnson, J. L. (2016). Alternate food markets, NGOs, and health policy: Improving food access and food security, trust bonds, and social network ties. World Medical & Health Policy, 8, 157–178. Bryce, R., Guajardo, C., Ilarraza, D., et al. (2017). Participation in a farmers' market fruit and vegetable prescription program at a federally qualified health center improves hemoglobin A1C in low income uncontrolled diabetics. Prev Med Rep: 7, (pp. 176– 179). Caspi, C. E., Lenk, K., Pelletier, J. E., et al. (2017). Association between store food environment and customer purchases in small grocery stores, gas-marts, pharmacies and dollar stores. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 76. Centers for Disease Control and Prevention (2014). Healthier food retail: An action guide for public health practitioners. https://www.cdc.gov/nccdphp/dnpao/state-localprograms/pdf/healthier-food-retail-guide-full.pdf. Cubbin, C., Haden, W. C., & Winkleby, M. A. (2001). Neighborhood context and cardiovascular disease risk factors: The contribution of material deprivation. Ethnicity & Disease, 11, 87–700. Deaton, A., & Lubotsky, D. (2013). Mortality, inequality and race in American cities and states. Social Science & Medicine, 56, 1139–1153. Duncan, D. (2013). What's your Walk Score®? American Journal of Preventive Medicine, 45, 244–245. Economos, C. D., Hatfield, D. P., King, A. C., et al. (2015). Food and physical activity environments. American Journal of Preventive Medicine, 48, 620–629. Eikenberry, N., & Smith, C. (2004). Healthful eating: Perceptions, motivations, barriers, and promoters in low-income Minnesota communities. Journal of the American Dietetic Association, 104, 1158–1161. Ellsworth, D., Ernst, J., & Snelling, A. (2015). A mobile farmers' market bring nutrition education to low-income students. Health Education, 115(2), 171–177. Epicurious (2017). Fresh moves delivers vegetables to the south side of Chicago by bus.

5. Conclusions Mobile farmers markets are often touted by policy statements and action guides in the United States as a panacea to increase access to healthy food in food deserts. We recommend continued evaluation on the efficacy and effectiveness of mobile markets. Mobile farmers markets specifically designed to target food deserts should implement periodic customer intercept surveys to ensure mobile markets are serving target customers and to identify perceived barriers related to healthy food access. For any mobile market, periodic surveys can also be used to identify customer needs and inform program planning and nutrition education opportunities such as cooking demonstrations. New opportunities for mobile farmers markets could include linking the food and health care sectors to increase fruit and vegetable consumption for low-income populations. More research is needed to evaluate the longterm social, economic, and health effects of increasing access to fruits and vegetables through mobile farmers markets, particularly in geographically diverse communities with limited walkability and varied transportation patterns. Financial support This study was supported, in part, by funds from the Baylor University Research Committee and the Vice Provost for Research 284

Appetite 133 (2019) 279–285

K.R. Ylitalo et al.

Burden of diseases, injuries, and risk factors. Journal of the American Medical Association, 10, 591. Redmond, N., Baer, H. J., Clark, C. R., Lipsitz, S., & Hicks, L. S. (2010). Sources of health information related to preventive health behaviors in a national study. American Journal of Preventive Medicine, 38(6), 620–627 e2. Robinson, J. A., Weissman, E., Adair, S., et al. (2016). An oasis in the desert? The benefits and constraints of mobile markets operating in syracuse, New York food deserts. Agriculture and Human Values, 33, 877–893. Saarloos, D., Kim, J. E., & Timmermans, H. (2009). The built environment and health: Introducing individual space-time behavior. International Journal of Environmental Research and Public Health, 6(6), 1724–1743. Saelens, B. E., Sallis, J. F., Black, J. B., & Chen, D. (2003). Neighborhood-based differences in physical activity: An environment scale evaluation. American Journal of Public Health, 93(9), 1552–1558. Tester, J. M., Yen, I. H., & Laraia, B. (2012). Using mobile fruit vendors to increase access to fresh fruit and vegetables for schoolchildren. Preventing Chronic Disease, 9, E102. The Network for Public Health Law (2013). Food trucks and fruit carts: How mobile vending can create greater access to healthy foods. https://www.networkforphl.org/ the_network_blog/2013/09/03/232/food_trucks_and_fruit_carts_how_mobile_ vending_can_create_greater_access_to_healthy_foods. Tripicchio, G. L., Smith, J. G., Armstrong-Brown, J., et al. (2017). Recruiting community partners for Veggie van: Strategies and lessons learned from a mobile market intervention in North Carolina, 2012–2015. Preventing Chronic Disease, 14, E36. U.S. Department of Health and Humas Services and U.S. Department of Agriculture. 20152020 (December 2015). Dietary guidelines for Americans (8th ed.). . Available at: http://health.gov/dietaryguidelines/2015/guidelines/. US Department of Agriculture (2013). Measuring effects of mobile markets on healthy food choices. https://www.ams.usda.gov/sites/default/files/media/MobileMarkets. pdf. Walker, R. E., Keane, C. R., & Burke, J. G. (2010). Disparities and access to healthy food in the United States: A review of food deserts literature. Health & Place, 16, 876–884. World Hunger Relief, Inc. Welcome 2016. http://www.worldhungerrelief.org/. Yaroch, A. L., Tooze, J., Thompson, F. E., et al. (2012). Evaluation of three short dietary instruments to assess fruit and vegetable intake: The national cancer institute's food attitudes and behaviors survey. Journal of the Academy of Nutrition and Dietetics, 112, 1570–1577. Yeh, M. C., Ickes, S. B., Lowenstein, L. M., et al. (2008). Understanding barriers and facilitators of fruit and vegetable consumption among a diverse multi-ethnic population in the USA. Health Promotion International, 23, 42–51. Ylitalo, K. R., Umstattd Meyer, M. R., Stone, K., Doyle, E. I., & Curtis, R. (2016). Using the Community Assessment for Public Health Emergency Response (CASPER) to assess barriers to healthy eating and active living in a low-income community. Evaluation and Program Planning, 59, 41–46. Zepeda, L., Reznickova, A., & Lohr, L. (2014). Overcoming challenges to effectiveness of mobile markets in US food deserts. Appetite, 79, 58–67.

http://www.epicurious.com/expert_advice/fresh_moves_mobile_grocery_store_ chicago_article. Fresh Truck (2015). Driving, food, health and community. http://www.freshtruck.org/. Fulkerson, J. A., Sherwood, N. E., Perry, C. L., et al. (2004). Depressive symptoms and adolescent eating and health behaviors: A multifaceted view in a population-based sample. Preventive Medicine, 38, 865–875. Haynes-Maslow, L., Auvergne, L., Mark, B., Ammerman, A., & Weiner, B. J. (2015). Lowincome individuals' perceptions about fruit and vegetable access programs: A qualitative study. Journal of Nutrition Education and Behavior, 47(4), 317–324.e1. Health Information National Trend Surveys (2013). Survey instruments. http://hints. cancer.gov/instrument.aspx. Hirsch, J. A., Moore, K. A., Evenson, K. R., et al. (2013). Walk Score® and Transit Score® and walking in the multi-ethnic study of atherosclerosis. American Journal of Preventive Medicine, 45, 158–166. Idler, E. L., & Benyamini, Y. (1997). Self-rated health and mortality: A review of twentyseven community studies. Journal of Health and Social Behavior, 38, 21–37. Indiana University Health (2012). Garden on the Go demonstration study report. https:// iuhealth.org/images/uploads/Final_Report_Garden_on_the_Go_Demonstration_Study. pdf. Indiana University Health (2017). Garden on the Go. http://iuhealth.org/about-iuhealth/in-the-community/healthy-weight-and-nutrition/garden-on-the-go/. Izumi, B. T., Higgins, C. E., Baron, A., et al. (2018). Feasibility of using a communitysupported agriculture program to increase access to and intake of vegetables among federally qualified health center patients. Journal of Nutrition Education and Behavior, 50(3), 289–296.e1. Robert Wood Johnson Foundation (2014). Rural childhood obesity prevention toolkit. http://www.leadershipforhealthycommunities.org/wp-content/uploads/2014/12/ LHC_Rural_Toolkit_FINAL.pdf. Kamphuis, C. B., van Lenthe, F. J., Giskes, K., et al. (2007). Perceived environmental determinants of physical activity and fruit and vegetable consumption among high and low socioeconomic groups in The Netherlands. Health & Place, 13, 493–503. Kraschnewski, J. L., George, D. R., Rovniak, L. S., et al. (2014). Characterizing customers at medical center farmers' markets. Journal of Community Health, 39(4), 727–731. Mavaddat, N., Valderas, J. M., van der Linde, R., et al. (2014). Association of self-rated health with multimorbidity, chronic disease and psychosocial factors in a large middle-aged and older cohort from general practice: A cross-sectional study. BMC Family Practice, 15, 185. Moore, L. V., Dodd, K. W., Thompson, F. E., et al. (2015). Using behavioral risk factor surveillance system data to estimate the percentage of the population meeting US department of agriculture food patterns fruit and vegetable intake recommendations. American Journal of Epidemiology, 181, 979–988. Morland, K., Wing, S., Diez Roux, A., et al. (2002). Neighborhood characteristics associated with the location of food stores and food service places. American Journal of Preventive Medicine, 22, 23–29. Murray, C. J., Atkinson, C., Bhalla, K., et al. (2013). The state of US health, 1990-2010:

285