Good nutritional health, which relies on the adequate consumption of healthful foods, is essential for the prevention and management of nutrition-related health conditions such as obesity, diabetes, and cardiovascular disease (1-3). It is widely recognized that access to the food environment may have an effect on dietary intake and overall health (4-8). Common among these and other studies on food access is a focus on potential spatial access (proximity or distance) to food stores, primarily supermarkets (9-15). Use of the retail food environment or realized access is the outcome of a complex interaction between consumers and characteristics of the food environment, such as proximity or density of retail food sources (10), and presence of culturally specific foods in accessible markets (16-18). Further, studies of food access have expanded to include diverse outlets such as convenience stores, non-traditional food stores, and alternative retail food sources (11, 18-22).

The Mexican-origin population is the fastest growing racial/ethnic subpopulation in the United States, and the subpopulation at greatest risk for food insecurity, obesity, and diabetes (23-26). This population growth has been occurring in the burgeoning colonias along the U.S. border with Mexico, especially in Texas and in new destination immigrant communities throughout the U.S. (27-29). New destination communities are characterized by a high concentration of Mexican-origin residents and rapid growth in locations previously unsettled by Mexican immigrants (28, 29). As an archetype for new destination communities, colonias are smaller, more dispersed communities comprised of disproportionately poor families of Mexican-origin with limited access to resources (30, 31). Two distinguishing features of colonias in Texas include their existence since the 1960’s and current inclusion of native and immigrant residents of Mexican-origin (30). There is limited work that acknowledges the existence of alternative food sources (11, 19, 21, 22, 32), and apparently no studies that examine factors associated with the purchase of foods from three common alternative food sources for Mexican-origin families in Texas border colonias: vendedores (mobile food vendors), pulgas (flea markets), and vecinos o amigos (neighbors or friends) (25). The purpose of this research is to examine the use of these three alternative food sources by 610 Mexican-origin women from Texas border colonias and determine factors associated with use.

Data were collected from 610 adult women in 44 colonias as part of the 2009 Colonia Household and Community Food Resource Assessment (C-HCFRA). Spanish is the primary language spoken in these areas. Details of training, participant eligibility and recruitment, and administration of a face-to-face survey in Spanish by trained promotora-researchers (promotoras de salud trained in research methods) has been previously described in detail (25). The survey and all protocols were approved by the BLINDED FOR REVIEW Institutional Review Board, and each participant provided written informed consent, which was conducted in Spanish.

Table 1 shows the sample distributions for demographic characteristics, food environment access, availability, and use of alternative food sources. Briefly, the average age was approximately 40 y, 60% were married, 67.7% were born in Mexico, almost 20% were the lone adult in the household, and 26.9% participated in both SNAP and WIC. More than 16% of households were considered adult food insecure and 49% child food insecure. Almost 70% of the women had a car available during the day; 61.3% traveled at least 10 miles one-way to purchase groceries; and 29.5% rode with a friend or family member to purchase groceries. More than 90% strongly agreed or agreed that there was little variety in types of foods, few grocery stores or supermarkets, or high food prices in their community. Fifty-two percent of households used at least one type of alternative food source for the purchase of food. More than 30% of all participants and 59% of households that used alternative food sources purchased food from pulgas (flea markets). Among those who purchased food at a pulga, 62% purchased fresh fruit and vegetables. More than 25% of participants purchased food from at least two of the three types of alternative food sources.

Results of regression models that examined the sample characteristics associated with use of alternative retail food sources are shown in Table 2. Interaction variables were not statistically significant in any of the models and were removed. The odds for purchasing food from a mobile food vendor were greater for households that participated in school nutrition programs, were child food insecure, or also purchased food at a pulga or from a neighbor or friend. With inclusion of purchasing food from a mobile vendor or neighbor, the odds for purchasing food from a pulga were greater for participants in school nutrition programs, food secure households, purchasing food from a mobile food vendor, or purchasing food from a neighbor or friend. Women more likely to purchase food from a neighbor or friend (Model 2) were younger, in a lone adult household, participated in the school nutrition program, were food secure, purchased food from a mobile food vendor, or purchased food at a pulga. Recipients of SNAP benefits were less likely to purchase food from a neighbor or friend. WIC participation was not associated with any of the alternative food sources.

This study is apparently the first study to examine the extent of use of three types of alternative food sources for the purchase of food by Mexican-origin households in Texas border colonias and the sample characteristics associated with use of each food source. The findings indicate the importance of these alternative food sources to Mexican-origin families. Notably, vendedores (mobile food vendors) who market food in the colonias were primarily a source for snacks, such as nieves (ice cream) or elotes (roasted corn), which reflects findings from urban settings such as Los Angeles (21) and Oakland, CA (22) where mobile food vendors primarily served as a near-school source of snack foods. Pulgas were a source for fresh fruits and vegetables, along with other prepared food items, reinforcing recent observations in South Texas (19); and prepared foods were purchased from neighbors or friends.

Demographic characteristics were only associated with purchasing food from a neighbor or friend. Younger women and households with a lone adult were more likely to purchase food from a neighbor or friend. Possible explanations may be that younger women or lone adult households may not have an available car or the necessary resources such as time to prepare “scratch” meals (36, 37). Food secure households were more likely to purchase food from pulgas or from neighbors, which suggests that this may be a strategy used to remain food secure. Earlier research identified the lack of culturally specific fruits and vegetables as an obstacle to the consumption of healthy foods among Latinos in Connecticut and North Carolina (16). The use of pulgas as an inexpensive source of culturally specific fresh fruits and vegetables (19) allows women to overcome this obstacle to food acquisition, and offers a plausible explanation for greater levels of food security among households who shopped at pulgas (25). Conversely, the most food insecure households were more likely to purchase snacks or ice cream from mobile food vendors. For food insecure households, use of mobile food vendors may be a strategy to complement reduced household food supplies and purchase snacks or ice creams from mobile food vendors when small amounts of money were available. In a prior study of household food availability using multiple, comprehensive inventories of household food supplies, it was observed that ice cream and other snack foods were not available in the homes (38). Individual perception of the community food environment was only associated with purchasing food from a pulga; that is, participants who believed that there were few places in their community to buy groceries were more likely to frequent a pulga, where most households purchased fresh fruits and vegetables. Interestingly, the use of alternative food sources was significantly associated with using other alternative food sources. This suggests that households that rely on alternative food sources may use multiple types of alternative food sources to prevent or mitigate food insecurity (39-41). In addition, these findings suggest a community-based food system in colonias integrates consumers and vendors associated with local and culturally relevant sources of food, such as mobile food vendors, pulgas, and residentially-based entrepreneurs (e.g., neighbors/friends who sell food) (42).

Although there are not published studies of mobile food vendors or neighbors marketing foods within the colonia context, this study presents the consumer side of the pulga retail counter. Dean and colleagues described the variety of foods available in pulgas and the demographic and business characteristics of pulga vendors (19, 32). This study builds on that research and presents a picture of the extensive reliance by Mexican-origin households on pulgas as a source for food purchases.

There are a number of strengths to this study, including a large sample of hard-to-reach Mexican-origin households in Texas border colonias and a comprehensive survey instrument administered face-to-face by trained promotora-researchers. This study was also unique in the examination of multiple alternative food sources. There are also a number of limitations, such as the cross-sectional study design, lack of data on frequency of use of each alternative food source, and the type of food items purchased from neighbors or friends (e.g., home-prepared foods, hot dishes, or store-bought items).

Notwithstanding these limitations, two findings are of special importance. First, Mexican-origin households in this sample relied to a large extent on pulgas for fresh fruits and vegetables. This suggests that practitioners and researchers who develop interventions or programs to increase the consumption of healthful foods, such as fruits and vegetables, be aware of all potential food sources (traditional, convenience, non-traditional, and alternative), especially those preferred by community residents. Second, almost 25% of households purchased food from a neighbor or friend. Future research should help practitioners and researchers understand why and what types of foods are purchased from neighbors or friends. Our findings support further broadening the food environment concept to account for the breadth and importance of alternative food sources, and the importance of linking use of food sources with proximity to more accurately depict access to food sources.