The economic and social burdens posed by nutrition-related chronic health conditions (e.g., obesity, cardiovascular disease, and diabetes) are tremendous. It is well accepted that diet and nutrition influence health outcomes; much of the research focus has been on the relationship between nutrient intakes and disease processes [1]. The burden of nutrition-related conditions becomes greater for marginalized populations that face greater vulnerability to food insecurity, poor nutritional health, and adverse health outcomes [2]. One, such marginalized population is Hispanic families who reside in colonias along the Texas-Mexico border and are primarily poor. Colonias, which were developed from subdivided agricultural lands in response to a deficit in low-income housing [3], are substandard residential areas with inadequate roads, variable housing conditions (see Figure 1), and drainage which frequently does not provide access to safe water or sewer sources [4]. Almost 20% of these largely Hispanic households have a female household head, and half of all children are food insecure [5]. Almost 70% of all colonias in the Lower Rio Grande Valley of Texas are located in Hidalgo County, where the number of colonias continues to grow [6]. Residents of colonias face great structural and neighborhood disadvantage, such as inadequate roads, limited or non-existent public transportation, and poor access to community resources [7]. This makes it particularly difficult for children and adults in these areas to initiate or maintain healthy eating habits that are critical for the prevention of disease conditions.

Ecological approaches to behavior change and health recognize that there is a dynamic interaction between the individual and the environment [2,8,9]. For example, the consumption of healthy foods, such as fruits and vegetables, are recommended for overall nutritional health; however, they are often not easily accessible [10]. Local food environments, which are a primary source for food that is consumed, may have an effect upon health and well-being through food cost and availability [11-14]. In documenting a relationship between food access and health, studies have used several proxy measures for health: 1) access to fruits and vegetables or other healthy foods, 2) intake of fruits and vegetables, and 3) prevalence of obesity. In a Colorado study that involved data collection at multiple times, greater availability of produce was associated with greater increase in fruit and vegetable servings [15]. In a review of 12 descriptive studies, Robinson found that dietary behaviors and fruit and vegetable intake among African Americans in the U.S. were the result of several factors, including food access and availability [9]. Using data from the Atherosclerosis Risk in Communities (ARIC) study, Morland and colleagues found that regardless of race, fruit and vegetable intake was higher in census tracts with more supermarkets [11]. Additional observational studies supported this relationship [16-22]. In a study of urban New Orleans, Bodor and colleagues found that greater availability of fresh vegetables in the neighborhood, regardless of type of store, was associated with increased intake [23]. In contrast, a postal survey conducted among 426 respondents to a postal survey (42.6% response rate) in the U.K. found that travel distance to the nearest supermarket was not associated to either fruit or vegetable consumption [24]. In two U.K. natural experiments that involved the introduction of a new supermarket in an area that previously had limited access to food stores, the results were not consistent on a relationship between food access and consumption [25,26]. Morland and colleagues looked at the relationship between food stores and obesity, using ARIC data, and found that obesity was associated with lower numbers of supermarkets and higher numbers of convenience stores within census tracts [27]. A large study in urban areas of Massachusetts found a significant association between environmental variables, such as presence of a supermarket, and obesity [28]. When considering away-from-home foods, such as fast food, Thomson and colleagues found a relationship between frequency of fast food consumption and development of obesity among U.S. girls [29].

However, the results of food environment studies from the U.S., U.K., Canada, Australia, and New Zealand that examined physical access to food stores are mixed. Some studies determined that families who live in poorer areas (e.g., lower socioeconomic or greater deprivation) or minority neighborhoods have little or no access to supermarkets, where there are larger selections of healthy foods (e.g., fruits, vegetables, low-fat dairy and meats) [12,30-39]. Other studies found little or no difference in access to supermarkets between deprived and affluent areas [32,40-45], or better access to food stores from more deprived neighborhoods [46-49]. Although the availability of healthy foods in food stores or consumption of healthy foods by residents is beyond the scope of this paper, it stands to reason that proximity of local food environments may influence food choice and adherence to dietary recommendations through food cost and availability [11-14,18,50].

Individual and community concerns with food security, limited access to supermarkets, higher costs for food (food price and transportation cost), and nutritional status are receiving increased attention [12,18,51-53]. However, there has been limited study related to: 1) use of multiple dimensions of spatial access to food stores, such as proximity (minimum distance) and coverage (number within a specified area), especially in the U.S. [14,22,35,38,40,46,49,54]; 2) spatial equity – the distribution of food resources in relation to population need [46,55-59]; 3) the growing influence of convenience stores and non-traditional food stores, such as mass merchandisers (e.g., Target, Kmart, Wal-Mart) and dollar stores on food availability [45,60]; and 4) the increased reliance on fast-food meals for in-home and away-from-home consumption [42,61,62].

Using network-based spatial measures, our primary objective was to determine the extent to which neighborhood need is associated with spatial access to food stores and fast food outlets in areas of colonias by 1) identifying and geocoding all food stores and fast food outlets in a 197-census block group (CBG) area of Hidalgo County, using ground-truthed methods (direct observation and on-site GPS); 2) determining network-based potential spatial access using distance and coverage criteria; and 3) using multivariate models to examine the relationship between neighborhood need and potential spatial access to food stores and fast food outlets.

There were 315 food stores (21 traditional food stores, 255 convenience stores, and 39 non-traditional food stores) and 204 fast food restaurants identified and geocoded. Traditional food stores included 3 supercenters, 11 supermarkets, and 7 grocery stores, while non-traditional food stores included 33 dollar stores, 4 mass-merchandisers, and 2 pharmacies. Forty-four percent of the fast food restaurants were national chains. Figures 3 and 4 show the spatial distribution of food stores, fast food restaurants, and CBG deprivation. Figure 3 shows that few food stores are located in high deprivation areas. In contrast, Figure 4 shows that convenience food stores and fast food restaurants can be found in areas of high deprivation.

The distribution of neighborhood socioeconomic characteristics shows diverse levels of neighborhood need in the study area (Table 1). In data not shown, in 25% of the 197 CBG, at least 16% of households did not have an available vehicle; and in 10% of CBG, at least 24% of households were without an available vehicle. Table 1 also shows the distribution of proximity and coverage. Considering both proximity and coverage, neighborhoods had best access to convenience stores, fast food restaurants, and dollar stores.

This study extends our understanding of spatial access to food resources from neighborhoods that are more socioeconomically deprived or have greater proportions of households without available transportation by examining two dimensions of access: 1) proximity (distance) to the nearest food store or fast food restaurant and 2) coverage (number) of food stores or fast food restaurants within a specified distance [40]. In contrast to studies that narrowly define the food environment as supermarkets and/or fast food restaurants [38], our examination of access from primarily colonia neighborhoods in an area of high or persistent poverty along the South Texas border with Mexico focuses on all food stores (i.e., supercenters, supermarkets, grocery stores, convenience stores, mass merchandisers, dollar stores, and pharmacies) and fast food restaurants.

Overall, neighborhoods had the best spatial access to a convenience store or fast food restaurant, both in the distance to the nearest convenience store or fast food restaurant and in the number of shopping opportunities at convenience stores or fast food restaurants within a certain distance of the neighborhood. Among traditional food stores, supercenter/supermarket provided greater proximity and coverage than any individual type of traditional food store. When neighborhood deprivation was considered, the results differed between proximity-determined and coverage-determined access. Using distance to the nearest store as a measure of access, CBG with increased deprivation were correlated with poor access to the nearest supercenter or grocery and good access to a convenience store. More limited availability of a vehicle was correlated with better proximity to a supercenter/supermarket, convenience store, dollar store, and fast food restaurant. However, after controlling for population density, increased deprivation was associated with greater distance to the nearest supercenter/supermarket, grocery store, mass merchandiser, dollar store, and pharmacy. Vehicle availability was not associated with any food store or fast food restaurant. This suggests that neighborhoods with greater need have poorer spatial access to the type of stores where greater selections and lower prices are available [83].

Initially, study results of coverage criterion showed that neighborhoods of greater deprivation were associated with better access to grocery stores within 1-network mile, which is opposite of the relationship using proximity. As the buffer distance increased, food store and fast food restaurant access declined with increasing deprivation. For neighborhoods with decreasing availability of a vehicle, access to food stores and fast food restaurants was greater within a 1-mile buffer, and reverse with increasing distance. After controlling for population density, neighborhoods of increased deprivation had access to fewer fast food restaurants within 1-network mile; none of the food stores was significant. At the same distance, neighborhoods with decreasing vehicle availability were more likely to have more supercenter/supermarkets, grocery stores, or fast food restaurants within one mile. At three miles, increasing deprivation was associated with fewer grocery stores, mass merchandisers, dollar stores, pharmacies, and fast food restaurants; all controlled for population density. At this same distance, vehicle availability was not associated with any aspect of the food environment.

Regardless of measure, access to the food environment, especially to supercenter/supermarkets, where the largest selections of affordable healthy foods are marketed [84], was a major problem for many residents of colonia neighborhoods in South Texas. Access in terms of distance to the nearest food store and the variety of stores within a given area was relatively better to convenience stores and fast food restaurants, where the opportunities for healthy foods are limited [65,85]. This study found similar pictures of spatial distribution for proximity and coverage. The value of using both measures is in the disparate information that each provides. Proximity provides a measure of network distance to the nearest store of a specific type. This can also provide a sense of travel time from starting point. Coverage describes choice; that is, the number of stores and potential difference in selection and/or price that are available within a given area.

Unlike much larger urban areas [35,40], none of the neighborhoods in this study had more than two supercenter/supermarkets within one mile and 94% did not have more than three within three miles. Being able to walk to a supermarket was out of the question for almost all the neighborhoods in this study. This compares with more urban areas in which a supermarket is walkable [38,40,42]. The difference in population density may explain much of this; for example, the Zenk study in Detroit estimated a median population density of 5367.4 while this study reports a median of 2528.9 [35].

Access and availability to the food environment will play a pivotal role for the nutritional health of families, especially among the increasing Hispanic population within neighborhoods of colonias. Many of these neighborhoods experience overall socioeconomic deprivation; many are home to large proportions of residents who have a low household income, are unemployed, or lack access to a vehicle. The lack of public transportation, especially in many of the areas, further marginalizes a large, disadvantaged population and limits their options for food resources. Indeed, it is difficult to initiate or maintain healthful eating habits without access to healthy foods. Knowing more about the food environment is essential for combining environmental approaches with traditional health interventions to make it easier for individuals to make healthier food choices [90]. The preparation for policy change to strengthen food assistance programs or program delivery activities, or interventions to improve nutritional health should include an understanding of the community – where people live and where they shop for food [90]. Additionally, it is important to understand not only the distance that people must travel to the nearest store to make a purchase, but also how much diversity in stores they have in order to compare price, quality, and selection. Future research should examine how spatial access to the food environment influences the utilization of food stores and fast food restaurants, and the strategies used by low-income Hispanic families in this area to obtain food for the household.

CBG: census block group; CFEP: Colonia Food Environment Project.

The authors declare that they have no competing interests.

JS conceptualized the study and supervised the data collection. He carried out statistical analyses. SH carried out mapping of results. JS drafted the manuscript. DH carried out spatial analysis; JCH directed multivariate modeling. All authors approved the final manuscript.