In public health practice, we know that access to healthy food and overall health are largely reliant on built environments where individuals work, live, and spend leisure time. It is well known in the public health literature that residents in rural and inner-city areas may not be able to access healthy foods due to cost, availability, access to transport, the built environment, and/or other factors.1 Research on food access in rural areas has emphasized long commutes and lack of access to fresh produce among vulnerable populations.2 Research in urban areas has emphasized concepts such as food deserts and food swamps, which document a lack of access to fresh produce as well, typically highlighting disparities by race and income.3–7 Overall, poorer US residents typically have to commute farther for healthy food at an affordable cost, do not have access to reliable personal vehicles, and when there is access to healthy food, often there is an overabundance of cheaper, unhealthy alternatives (eg, processed foods, foods high in saturated fat and sugars and low in vitamins and minerals).8–10 The inability to access healthy foods can lead to adverse childhood health consequences, including obesity and nutrient deficiencies, and increased risk of chronic disease later in life.11–14
The provision of healthy food access for economic reasons in the United States is not new. The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) was piloted by the US Department of Agriculture in (USDA) in 1972 and was established as a permanent program in 1975 to provide nutrition education, healthy foods referrals to health, and other social services during key periods of growth and development.9 Studies have shown that women who participated in the program were more likely to have infants born at normal weight. In addition, Medicaid costs were lower than those for nonparticipants. WIC has successfully increased breastfeeding among participating women, improved household food security status, and participating children have demonstrated higher academic achievement than comparable low-income nonparticipants.15–18
The Inter Tribal Council of Arizona, Inc (ITCA) was established in 1952 to provide a united voice for Tribes in Arizona. In 1975, ITCA developed a nonprofit corporation to provide Member Tribes health services, and ITCA WIC was included in 1985 to provide food and nutrition services for clients.19 For the remainder of this article, references to WIC unless otherwise stated are in reference to ITCA WIC. WIC participants meet the federal income requirements. The WIC program serves Tribal and non-Tribal Members in Arizona. At the time of the study, the WIC program included WIC clinics at 4 Native Health urban sites and 12 Tribal sites. It is unknown whether the urban residents are Tribal Members (most are presumed to be), and nearly all of the rural residents live on Tribal lands. WIC authorized stores include independent grocers, chain grocery stores, and tribally owned stores. All WIC authorized stores must provide a combination of competitively priced healthy foods including cereal and bread, a combination of 5 fresh dairy products (milk, cheese, or yogurt), 7 types of fresh fruit, 7 types of fresh vegetables, and 4 or more varieties of meat (with no more than 2 as canned meat).9,20 This criterion for WIC store approval ensures access to healthy foods based on availability and competitive cost of food items. WIC, by design, is supposed to increase access to healthy foods; yet, an evaluation of improving access to stores had not been conducted. Therefore, ITCA WIC requested that the ITCA Tribal Epidemiology Center (TEC) evaluate several components of the program, including how well the current WIC store network was serving its participants in 2014 and 2016, and whether any stores could be added to the network to improve access to healthy foods. This study adds to the program evaluation literature by providing an inexpensive and expedient methodological framework and secondary analysis that could be replicated by other WIC programs nationally to improve access to healthy foods for WIC participants based on location.
The foundation of this evaluation is calculating the distance of each WIC client from his or her nearest approved WIC store. How that distance is determined is a key component of this evaluation. We used Euclidean distance, which is simply defined as a straight line from one point to another, regardless of barriers or “as the crow flies.” Past research using Geographic Information Systems (GIS) to evaluate food access involved analyzing Euclidean distance from store locations,21,22 travel distance from stores,23,24 and travel time to stores25 aggregated by neighborhoods in urban areas.26 Euclidean distance has the strength of being easy to calculate and easy to replicate for repeat analysis but often underestimates overall travel distance.27 Travel distance is more difficult to determine because it requires the creation of a network for analysis, and travel time requires both network creation and can have additional challenges if multiple transportation modalities are incorporated. There has not been nearly the volume of research dedicated to rural food deserts and food access problems as there has been for urban studies. However, McEntee and Agyeman2 recommended using census tracts to visualize travel time to identify rural areas that have limited food access in order to help convey the serious travel time issues in rural areas. This evaluation on the WIC program incorporates many of the methods utilized in the existing literature for urban areas for identifying food deserts and modifies the techniques to analyze both a store network and individual stores in both rural and urban areas.
This evaluation required the development of a methodological framework that could analyze the current WIC store network and the individual WIC store service areas and to potentially identify new WIC stores that could meet any gaps in the current store network in both rural and urban areas. The objective for the WIC program is to have a WIC store within 1 mile or less of participants living in an urban area and 5 miles or less of participants living in nonurban areas. Using these parameters, ITCA TEC was able to determine how many WIC participants are considered being “served,” or were within the desired distance of a WIC store, and the “underserved participants,” or those not within the desired distance of a WIC store. The average Euclidean distance, travel distance, and travel time for both nonurban and urban WIC participants both inside and outside the target areas were the criteria used to determine served and underserved participants based on store access. Once the served and underserved WIC participants locations were identified, then additional stores could be identified and authorized by WIC to further increase healthy food access for WIC participants.
The methodological framework used by TEC, a public health authority, was approved by ITCA WIC. The institutional review board (IRB) at the Arizona State University determined that the activity was not research involving human subjects as defined by the Department of Health and Human Services and the Food and Drug Administration, and IRB review and approval were therefore not required.
ITCA WIC participant and store data
ITCA WIC contracts with 11 tribes and 2 urban Indian health centers, with nearly 40 WIC clinics in Arizona. The WIC participant database includes participant ID, address, city, state, zip code, and location description. All participant and store data were managed in ESRI's ArcMap 10.3. The participant location descriptions were either geocoded, or a description was included if the participant had no mailing address, or if the participant was not staying at his or her listed address (ie, red brick home, Route 42, mile post 34). In 2014, a total of 11 213 WIC participants addresses were identified and geocoded, roughly 93.4% of participants were available for store access analysis. In 2016, 12 275 WIC participants out of 13 000 were identified and geocoded, accounting for a similar 94.4%. WIC store data include only the name of the store and the address. Moreover, there were 136 WIC authorized stores in 2014 and 144 in 2016. WIC provided both the authorized WIC store data and the list and location of 700 additional stores that could potentially be authorized.
Arizona Department of Transportation (ADOT) road data, Open Street map data, and US Census 2010 data were used to determine travel distances, travel times, and urban and nonurban status. Open Street map data were used as a supplement to ADOT data, because these possessed better documentation of roads in rural areas and on tribal lands. Only road features were included in the network data set, and only roads that had assigned speed limits were included within the attribute data.
Distance and time variables
Each WIC participant was assigned a travel distance (by road) and travel time (by car) using attribute data. The spatial join function was leveraged by location to connect each WIC participant to his or her closest (Euclidean distance) authorized WIC store using standard techniques.28,29 This process yielded a WIC participant Euclidean distance value for the proximity to its closest WIC authorized store, in addition to the travel distance and travel time attributes.
Served and underserved WIC participants
WIC participants were assigned to urban and nonurban areas using 2010 US Census blocks (Figure).30 Within urban and nonurban designations using the US Census data, WIC participants were determined to be served and underserved on the basis of their distance (Euclidean) to the nearest WIC store. Individual authorized stores can be evaluated utilizing the information collected to assess the existing service network and by creating buffers to visualize the Euclidean service area of the authorized WIC stores and by joining vendors to nearest participants. Urban WIC participants were considered served if the participant was within 1 mile (Euclidean distance) of an authorized WIC store. The urban WIC participants were considered underserved if they had to travel more than 1 mile to an authorized store. Nonurban WIC participants were considered served if the participant was within 5 miles (Euclidean distance) of an authorized WIC store and underserved if they had to travel more than 5 miles.
Store network improvement
We evaluated the capacity of the additional stores provided to fill the service area gaps of the existing WIC store network for underserved WIC participants. ITCA WIC requested all stores in 2014 to be considered for approval by WIC that would serve at least 10 previously underserved clients. Each potential store was given a 1- and 5-mile service area in 2014. The list of potential stores was analyzed to determine whether more than 10 previously underserved participants would be considered served if the WIC program were to authorize the store. In 2016, we looked for improvements in the store network.
We used this information to identify a list of 127 possible stores for consideration for WIC authorization. From the list of 127 potential stores, WIC authorized an additional 8 stores. These stores were chosen on the basis of increasing access to the greatest number of WIC participants and WIC participant requests for store approval. Of the total WIC participants available for analysis, 5630 WIC participants in 2014 and 6653 WIC participants in 2016 were considered served by the store network. We identified that 50% of the WIC participants were served in 2014 and 54% were served in 2016 (Table), indicating a modest increase in store access improvement overall. Store network coverage varied by urban and nonurban areas, and with the addition of the 8 new stores, access increased in urban areas from 39% to 41% and from 66% to 74% in nonurban areas from 2014 to 2016 (Table). The biggest drops in travel distance and travel time observed with new store additions from 2014 to 2016 were among nonurban WIC participants who were more than 10 miles from a store; average travel time was reduced by 16 minutes, and travel distance was decreased by about 8 miles (Table).
The ITCA TEC program evaluated the WIC programs' authorized store network in order to further increase access to healthy foods for the WIC population. By using a constructed framework to identify served and underserved WIC participants in urban and nonurban areas, we were able to increase store coverage in order to serve the most participants with healthy foods. From the analysis of potential stores, WIC authorized an additional 8 stores. This modestly increased store network coverage by 2% for urban participants and 8% for nonurban participants. From 2014 to 2016, among the underserved participants, travel times and distances for both urban and nonurban WIC participants were reduced modestly.
In 2016, 74% of nonurban WIC participants were within 5 miles of an authorized store. The primary limiting factor in many of the service gaps for 2016 for nonurban WIC participants is simply a lack of potential stores. In 2014, there was only one major nonurban store in or near the tribal land that had not been authorized. When the store was authorized, 180 previously underserved participants were then served. This resulted in reduced travel times and distances to access healthy foods for these nonurban WIC participants. All qualifying stores on or near tribal lands that met WIC qualifications at the time of the review were authorized.
In urban areas, WIC has efficiently identified stores in the Phoenix and Tucson metropolitan areas that are within an average of 1.8 miles for all urban WIC participants. For some of the WIC participants, there are many WIC stores along public transportation routes such as bus lines and light rail, although public transit was not included in the distance analysis. There are certain store configurations within the Phoenix and Tucson metropolitan areas that may fit the participants marginally better based on pure location proximity. WIC has based many decisions on authorizing larger grocery stores that tend to be favored over smaller stores with limited selection.31,32 The smaller target service area within the urban areas makes it difficult to greatly improve the store network without authorizing a large number of additional stores in urban areas. Additional urban store approvals require more labor for ITCA WIC to conduct monitoring of WIC authorized stores as required by federal regulations. The state of Arizona and Navajo Nation also serve as WIC state agencies serving overlapping areas of the state with ITCA and share many of the same vendors. However, WIC participants from the 3 programs can only shop at stores authorized by the program they are enrolled in. A practical solution to this problem is for USDA to allow WIC state agencies to easily share store management and monitoring rather than duplicating efforts with other WIC programs.
There are several strengths and limitations to this analysis. We established a methodological framework that uses secondary WIC data that can be replicated easily and are cost-effective. We analyzed WIC participant movement from 2014 to 2016, and if participants moved, they tended to live within the same areas. So, it is unlikely that increased store coverage was due to WIC participants moving closer to WIC authorized stores. However, about 5% to 6% of WIC participant addresses could not be geocoded because of missing address information or nonstandard addresses (eg, PO boxes, Red house, 2 miles from store). We do not know whether these participants were served or underserved, and they were not included in the analysis. Many participants, particularly in urban areas, do not necessarily go to the store that is closest to them; participants may be willing to increase travel distance and times to go to a preferred store. We know that personal preference plays a significant role in store choice.21,33,34 We did not include public transportation options in the analysis, which could influence some of the results.
By utilizing the evaluation framework based on WIC participant and store locations, we determined how well the existing WIC store network served the current WIC participants. In addition, we were able to evaluate potential stores based on the spatial relationship to existing WIC participants. By approving 8 additional stores, WIC was able to increase healthy food access and reduce travel times and distances for participants. This evaluation framework based in GIS and spatial analysis methodology can be utilized every 2 years in order to ensure the best WIC store coverage over time, since stores open and close, and participants move.
Implications for Policy & Practice
- WIC is a nationwide program. This framework uses secondary data and is a cost-effective method to evaluate WIC participant access to stores with healthy food. Other WIC programs nationally can use this framework for evaluation to increase healthy food access.
- Authorizing additional stores for WIC improves overall population access to healthy foods beyond those that are WIC eligible. This is particularly important in remote and rural areas where stores may be less inclined to serve a variety of fresh produce and dairy products if they were not required to do so in order to comply with WIC requirements.
- Authorizing only needed stores ensures that WIC state agency resources are utilized effectively and efficiently.
- Reducing travel distances and travel times for WIC participants reduces stress and commuting expenses that can be used for other important items.
1. Walker RE, Keane CR, Burke JG. Disparities and access
to healthy food in the United States: a review of food deserts literature. Health Place. 2010;16(5):876–884.
2. McEntee J, Agyeman J. Towards the development of a GIS method for identifying rural food deserts: geographic access
in Vermont, USA. App Geogr. 2010;30(1):165–176.
3. Chen T, Gregg E. Food Deserts and Food Swamps: A Primer. Vancouver, BC, Canada: National Collaborating Centre for Environmental Health; 2017:7.
4. Cooksey-Stowers K, Schwartz MB, Brownell KD. Food swamps predict obesity rates better than food deserts in the United States. Int J Environ Res Public Health. 2017;14(11):E1366.
5. Cummins S, Macintyre S. Food deserts—evidence and assumption in health policy making. BMJ. 2002;325:436–438.
6. Hendrickson D, Smith C, Eikenberry N. Fruit and vegetable access
in four low-income food deserts communities in Minnesota. Agric Hum Values. 2006;23:371–383.
7. Richards R, Smith C. Food deserts and how to tackle them: a study of one city's approach. Health Educ J. 2000;59:137–149.
8. Caraher M, Dixon P, Car-Hill R. Access
to healthy foods: part 1. Barriers to accessing healthy foods: differentials by gender, social class, income and mode of transport. Health Educ J. 1998;57:191.
9. USDA. WIC at a Glance. Washington, DC: USDA; 2015.
10. Whelan A, Wrigley N, Warm D, Cannings E. Life in a “food desert.” Urban Stud. 2002;39(11):2083–2100.
11. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr. 2004;79(1):6–16.
12. Morland K, Diez Roux A, Wing S. Supermarkets, other food stores, and obesity: the Atherosclerosis Risk in Communities Study. Am J Prev Med. 2006;30(4):333–339.
13. Papas MA, Alber AJ, Ewing R, Heizisouer KJ, Gary TL, Klassen AC. The built environment and obesity. Epidemiol Rev. 2007;29:129–143.
14. Pearson T, Russell J, Campbell MJ, Barker ME. Do “food deserts” influence fruit and vegetable consumption?—A cross-sectional study. Appetite. 2005;45(2):195–197.
15. Metallinos-Katsaras E, Brown L, Colchamiro R. Maternal WIC participation improves breastfeeding rates: a statewide analysis of WIC participants. Matern Child Health J. 2015;19(1):136–143.
16. Metallinos-Katsaras E, Gorman K, Wilde P, Kallio J. A longitudinal study of WIC participation on household food insecurity. Matern Child Health J. 2011;15(5):627–633.
17. Chapman-Novakofski K. Good outcomes with WIC continue. J Nutr Educ Behav. 2012;44(3):195.
18. Jackson MI. Early childhood WIC participation, cognitive development and academic achievement. Soc Sci Med. 2015;126:145–153.
19. Inter Tribal Council of Arizona: 21 Tribal Nations. ITCA Web site. http://itcaonline.com
. Accessed April 30, 2019.
20. Inter Tribal Council of Arizona, Inc. Special Supplemental Nutrition Program for Women
, Infants, and Children
. Phoenix, AZ: Inter Tribal Council of Arizona Inc; 2014. http://itcaonline.com/?page_id=53
. Accessed April 30, 2019.
21. Hillier A, Cannuscio CC, Karpyn A, McLaughlin J, Chilton M, Glanz K. How far do low-income parents travel to shop for food? Empirical evidence from two urban neighborhoods. Urban Geogr. 2011;32(5):712–729.
22. Grengs J. Does public transit counteract the segregation of carless households? Measuring spatial patterns of accessibility. Transp Res Rec. 2001;1753:3–10.
23. Bader MD, Marnie P, Yousefzadeh P, Neckerman KM. Disparities in local food environments: implications of measurement strategies. Econ Geogr. 2010;86(4):409–430.
24. Jiao J, Moudon A, Ulmer J, Hurvitz P, Drewnowski A. How to identify food deserts: measuring physical and economic access
to supermarkets in King County, Washington. Am J Public Health. 2012;102(10):32–39.
25. Burns CM, Inglis AD. Measuring food access
in Melbourne: access
to healthy and fast foods by car, bus and foot in an urban municipality in Melbourne. Health Place. 2007;13(4):877–885.
26. McKenzie BS. Access
to supermarkets among poorer neighborhoods: a comparison of time and distance measures. Urban Geogr. 2014;35(1):133–151.
27. Sparks A, Bania N, Leete L. Comparative approaches to measuring food access
in urban areas: the case of Portland, Oregon. Urban Stud. 2011;48(8):1715–1737.
28. Jacox E, Samet H. Spatial join techniques. J ACM Trans Database Syst. 2007;32(1):44.
29. Zhang F, Zhou J, Liu R, Du Z, Ye X. A new design of high-performance large-scale GIS computing at a finer spatial granularity: a case study of spatial join with spark for sustainability. Sustainability. 2016;8(9):926.
31. Kaufman PR. Rural poor have less access
to supermarkets, large grocery stores. Rural Dev Perspect. 1999;13(3):19–25.
32. Chung C, Myers SL. Do the poor pay more for food? An Analysis of grocery store availability and food price disparities. J Consum Aff. 1999;33(2):276–296.
33. Chen Y. Neighborhood form and residents' walking and biking distance to food markets: evidence from Beijing, China. Transp Policy. 2017. doi:10.1016/j.tranpol.2017.09.015.
34. Shannon J. Beyond the supermarket solution: linking food deserts, neighborhood context, and everyday mobility. Ann Am Assoc Geographers. 2016;106(1):186–202.