Healthy Bodegas: Increasing and Promoting Healthy Foods at Corner Stores in New York City

Abstract

Objectives. We assessed the effectiveness of an initiative to increase the stock and promotion of healthy foods in 55 corner stores in underserved neighborhoods.

Methods. We evaluated the intervention through in-store observations and preintervention and postintervention surveys of all 55 store owners as well as surveys with customers at a subset of stores.

Results. We observed an average of 4 changes on a 15-point criteria scale. The most common were placing refrigerated water at eye level, stocking canned fruit with no sugar added, offering a healthy sandwich, and identifying healthier items. Forty-six (84%) store owners completed both surveys. Owners reported increased sales of healthier items, but identified barriers including consumer demand and lack of space and refrigeration. The percentage of customers surveyed who purchased items for which we promoted a healthier option (low-sodium canned goods, low-fat milk, whole-grain bread, healthier snacks and sandwiches) increased from 5% to 16%.

Conclusions. Corner stores are important vehicles for access to healthy foods. The approach described here achieved improvements in participating corner stores and in some consumer purchases and may be a useful model for other locales.

Access to healthier food is associated with a healthier diet.^1–8 Supermarkets, which tend to offer a larger variety of healthier foods such as fresh produce,^9,10 are especially important. People who do not live near a supermarket are as much as 46% less likely to have a healthy diet than are people with the most supermarkets nearby.¹¹ The presence of supermarkets has also been associated with a lower prevalence of obesity.^12–14 Low-income and minority communities, which bear a disproportionately high burden of obesity, diabetes, and other diet-related diseases, generally have fewer supermarkets than do other communities.^7,9,10 A national study found that low-income areas had only 75% as many chain supermarkets as did middle-income areas and that Black neighborhoods had roughly half as many chain supermarkets as did White neighborhoods.⁷ These areas are instead highly populated with smaller convenience or corner stores.^8,15,16

These corner stores are less likely to carry healthy foods such as fresh fruits or vegetables,^15,17 heavily advertise unhealthy products, and are laden with convenience items that are often high in calories.^2,3,7,17 The presence of corner stores has been associated with increased risk of obesity,^4,18 a finding which is particularly relevant in New York City’s most underserved neighborhoods, where corner stores, often termed bodegas, can make up more than 80% of retail food outlets.²

Researchers have pointed to the potential for increasing consumption of healthy foods by making environmental changes in retail food outlets, including corner stores.^1,19 In urban settings, these changes may be particularly relevant because corner stores are regularly frequented by neighborhood residents. In New York City, bodegas receive an average of 703 customer visits from 8:00 am to 8:00 pm (556 from adults and 147 from youths; R. Dannefer, MPH, MIA, A. Sperling, MPH, S. Baronberg, MPH, and A. Abrami, MS, RD, unpublished data, 2009). A study in Philadelphia, Pennsylvania, revealed that children frequenting corner stores near their schools bought an average of 357 calories per visit and mostly purchased energy-dense products with little nutritional value such as chips, candy, and sugary drinks.²⁰

Given the ubiquity of corner stores in underserved communities, many interventions have been developed to increase their healthy offerings, using a variety of strategies to change the store environment.^21–24 For example, an intervention with corner stores and supermarkets in Baltimore, Maryland, provided store owners with monetary incentives, in-store materials, recommendations for stocking and promoting healthy foods, guidelines for interacting with customers, nutrition education, and taste tests. Participating corner stores demonstrated significant increases in stocking healthy foods, and consumer purchases of some healthy foods increased.²¹ Other programs have used promotional materials, local advertising and mass media messaging, cooking demonstrations, and taste tests to promote healthy foods. Positive findings from these interventions have included increases in stores’ stock of healthy foods and, in 1 case, increased purchases of some healthy foods.^22–24 These findings are complemented by research indicating that changes in the retail environment can effect changes in sales; for example, strategies such as increasing shelf space, in-store advertising, and locating foods in prime areas have all been demonstrated to increase sales for the promoted items.²²

One of the country’s largest corner store programs to date is the Healthy Bodegas Initiative, coordinated by the New York City Department of Health and Mental Hygiene and funded by the New York City Center for Economic Opportunity. Established in 2006, the initiative has worked with more than 1000 corner stores in neighborhoods with some of the highest poverty rates in New York City²⁵ and with a disproportionate burden of chronic disease, including obesity and diabetes.^2,17 The initiative supports environmental change in the store by offering simple, low-cost, and effective mechanisms for selling and promoting healthier items, and it also works with community organizations and residents to support the purchase of healthier foods, essential to ensuring sustainable inventory change. This article presents results from the evaluation of work conducted by the Healthy Bodegas Initiative in 2009 to increase availability and promotion of healthy foods at 60 bodegas in New York City.

METHODS

Healthy Bodegas Initiative outreach staff worked intensively with store owners to make positive changes based on 16 health-promoting criteria. Criteria were developed from a range of sources, including nutritionists’ input as to what foods should be available to ensure a healthy diet, potential customers’ input on what healthy foods they would buy at corner stores, and trial and error from previous Healthy Bodegas Initiative campaigns regarding what changes would be feasible given the physical limitations and resources of bodegas in target neighborhoods. Criteria were also based on health department priorities including increasing access to high-quality fresh produce and limiting sugary drink consumption. Specifically, criteria addressed increasing fresh produce and stocking healthier snacks, beverages, and grocery and deli items, as well as changes to the overall store environment, including improving produce displays and placing refrigerated water at eye level. Criteria were grouped to correspond to 3 increasingly higher “healthy store” levels on the basis of importance and difficulty of implementation. Level 1 required meeting 7 health-promoting criteria, level 2 required meeting 13 criteria, and level 3 required meeting 16 criteria (Table 1). These levels offered stores achievable goals while ensuring that they stocked a minimum range of healthy options. Bodegas received incentives as they advanced levels, including produce display crates and reusable shopping bags to offer to customers who purchased fresh produce. Stores also received promotional materials to market their healthier products to customers.

TABLE 1—

Number of Corner Stores Meeting Criteria and Healthy Store Levels Before and After Participation in an Intervention to Increase Stock and Promotion of Healthy Foods: 2009

Healthy Store Level	Criteria	Stores Before Intervention (n=55), No. (%)	Stores After Intervention (n=55), No. (%)	Stores That Changed, No.	P
Level 1a		5 (9.1)	24 (43.6)	19	< .001
	Stock canned fruit in own fruit juice or no sugar added	39 (70.9)	53 (96.4)	14	< .001
	Stock ≥ 2 varieties of fresh fruits	43 (78.2)	46 (83.6)	3	.182
	Stock ≥ 2 varieties of fresh vegetables	46 (83.6)	48 (87.3)	2	.159
	Stock low-salt or no-added-salt canned vegetables or soup	16 (29.1)	52 (94.5)	36	< .001
	Stock low-fat milk (1% or skim)	46 (83.6)	47 (85.5)	1	.659
	Display approved posters promoting healthy foods	30 (54.5)	55 (100.0)	25	< .001
	Meet produce display standards	53 (96.4)	55 (100.0)	2	.159
Level 2b		3 (5.5)	14 (25.5)	11	.001
	Stock ≥ 4 varieties of fresh fruits	20 (36.4)	26 (47.3)	6	.057
	Stock ≥ 4 varieties of fresh vegetables	35 (63.6)	38 (69.1)	3	.41
	Stock 2 healthier snack items	46 (83.6)	55 (100.0)	9	.002
	Stock whole-grain bread	38 (69.1)	48 (87.3)	10	.011
	Implement ≥ 1 strategy to increase healthy offerings	8 (14.5)	35 (63.6)	27	< .001
	Healthy sandwich (ham, turkey, chicken, or roast beef; whole grain bread; mustard; ≤ 1 slice of cheese; tomato; lettuce)	0 (0.0)	30 (54.5)	30	…
	Obtain a stoop line permit to sell fresh produce outside	3 (5.5)	5 (9.1)	2	.159
	Sell sliced and peeled fruits and vegetables in a packaged container	6 (10.9)	9 (16.4)	3	.371
	Displays refrigerated water at eye level	19 (34.5)	35 (63.6)	16	< .001
Level 3c		0 (0.0)	3 (5.5)	3	…
	Stock ≥ 1 dark green leafy vegetable	3 (5.5)	3 (5.5)	0	…
	Identify healthier items	0 (0.0)	52 (94.5)	52	…
	Participate in 20-min in-store training	0 (0.0)	43 (78.2)	43	…

^aStores meeting criteria for level 1.

^bStores meeting criteria for levels 1 and 2.

^cStores meeting criteria for levels 1–3.

Sixty bodegas were recruited in the South Bronx and East New York, Brooklyn, 2 of the city’s poorest neighborhoods²⁵ with high rates of obesity and diabetes.^26-27 To be eligible, stores had to carry a minimum of 1 fruit and 1 vegetable (excluding lemons, limes, potatoes, and onions) and meet the definition of a bodega (defined as a food store that has no more than 2 cash registers, sells mostly food, does not specialize in any 1 item, such as candy or meat, and sells milk). Outreach staff visited stores 10 times over 5 months to assess progress toward meeting the criteria and to identify areas for improvement.

Staff provided technical assistance such as rearranging the store to increase the visibility of healthy items, posting materials promoting healthier items, and assisting with applications for microfinancing and permits to display produce outside the store. Fifty-five stores completed the campaign; 5 were lost to follow-up because of stores closing or changing ownership.

To encourage demand for the new, healthier products, staff linked nearby community groups with bodegas to support their inventory changes. Activities included bringing community members to a participating bodega to highlight the healthier items available and facilitate a conversation with the store owner about stocking healthier items, as well as distributing consumer request cards to community organizations to be used to request specific healthy foods. Additionally, cooking demonstrations and recipe giveaways using bodega products were conducted outside of some stores to promote their healthy offerings.

Evaluation included 3 core methods: in-store observations, store owner surveys, and consumer surveys. All baseline data were collected in January 2009, and postintervention data were collected in July and August 2009.

Healthy Bodegas Initiative outreach staff conducted in-store observations before and after the intervention at all 55 stores that completed the campaign to measure their success in meeting criteria and healthy store levels. Because of inventory fluctuations, stores were observed 3 times before and 3 times after the intervention and were considered to meet a particular criterion if they fulfilled it at 2 of the 3 observations. When staff arrived at the store, they checked in with store staff and then initiated the observation; store owners were not informed ahead of time of when staff would visit the store.

We used a survey of store owners and managers to assess the campaign’s success and learn about owners’ experiences participating in the campaign. The survey included quantitative and qualitative questions. Topics included sales of healthy foods, barriers to carrying healthy foods, and aspects of the intervention such as the incentives distributed and suggestions for improvement. Of the 55 store owners, 46 (84%) completed both the preintervention and postintervention survey.

We used a consumer survey to assess changes in purchases of foods promoted by the campaign. For these surveys, we selected a subset of 10 stores on the basis of the owners’ willingness to allow interviewers to survey their customers. The survey was completed at 8 of the 10 stores before and after the intervention (2 were lost to follow-up). It collected no identifying information and included questions about food and nonalcoholic beverages purchased on that visit, shopping behavior, and demographics. New York City Department of Health and Mental Hygiene staff and volunteers conducted consumer surveys twice at each store, from 8:00 am to 10:00 am and from 4:00 pm to 6:00 pm. Customers were asked to participate in a brief survey as they exited the store. To be eligible, customers must have purchased something on that visit and had to have been aged 18 years or older. The survey was conducted with 617 respondents: 294 at baseline and 323 at the conclusion of the campaign. The response rate, calculated by dividing the number of surveys conducted by the number of eligible customers approached for the survey, was 53% at baseline and 63% at follow-up.

In addition to these evaluation methods, and because of noted limitations inherent in owner and customer recall, we attempted to collect invoices at select stores for purchases of fresh produce and milk for a 3-week period before and after the intervention. However, unlike larger stores, most bodegas do not have electronic cash registers or systematic methods for record keeping and were often unable to provide complete invoices for the designated data collection weeks. Additionally, of the invoices we were able to collect, some were illegible, and store owners mentioned making purchases for which receipts were never procured. Therefore, we excluded this aspect of the evaluation. Data analysis was conducted using the SPSS 18 statistical software package (PASW Statistics, Chicago, IL).

RESULTS

Store observations showed that bodega owners made on average 4 health-promoting changes—with some making as many as 7—out of the 15 criteria involving improvements to the stock and promotion of healthy foods. (Although there were 16 criteria overall, when calculating the average number of changes we did not count the last criterion on receiving our in-store training because this was not a change to the store.) In particular, stores stocking no-sugar-added canned fruit increased from 71% (n = 39) before the intervention to 96% (n = 53) after the intervention, stores stocking 4 varieties of fresh fruit increased from 36% (n = 20) to 47% (n = 26), and stores with refrigerated water at eye level increased from 35% (n = 19) to 64% (n = 35). Additionally, 55% of stores (n = 30) offered a healthy sandwich (ham, turkey, chicken, or roast beef on whole-grain bread with no more than 1 slice of cheese and lettuce, tomato, and mustard), and 95% (n = 52) posted signs to identify healthier items at the conclusion of the intervention, which no stores did before the intervention (Table 1). Two thirds of stores advanced at least 1 healthy store level, and some advanced as many as 3. Because stores often met some but not all criteria for each level at baseline, they were able to advance levels as they complied with the unmet criteria for each level (Table 1). As shown in Table 1, many of the increases in stock and promotion of healthy foods were statistically significant when using a t test, as were increases in the number of stores meeting healthy store levels 1 and 2.

Of the 46 store owners who completed surveys both before and after the intervention, most (78%; n = 36) reported that the intervention helped improve their sales of healthier foods. Some barriers to carrying healthy foods included consumer demand, mentioned by 28% (n = 13) of respondents, and lack of resources such as space and refrigeration, mentioned by 13% (n = 6). Suggestions at the conclusion of the campaign included providing more posters and organizing more promotions, cooking demonstrations, and food tastings.

Of customers who participated in the consumer survey, 80% reported shopping at that specific bodega once a week or more. Food and beverage purchases were comparable before and after the intervention; 74% (n = 218) of customers purchased food or nonalcoholic beverages before the intervention, compared with 71% (n = 228) after it (Table 2). Although the number of purchases of healthy items was low, purchases increased for some healthy foods. Among the 124 people at baseline and the 153 after the intervention who purchased beverages, the percentage purchasing at least 1 bottle of water increased from 6% (n = 8) to 12% (n = 18). Among the 111 customers at baseline and after the intervention who purchased items for which we promoted a healthier option (low-sodium canned goods, low-fat milk, whole-grain bread, healthier snacks and sandwiches), the percentage purchasing healthier options increased from 5% (n = 6) before the intervention to 16% (n = 18) after it.

TABLE 2—

Overall Consumer Purchases at Corner Stores Before and After an Intervention to Increase Stock and Promotion of Healthy Foods: 2009

	Customers Purchasing Item Before Intervention (n = 218), No. (%)	Customers Purchasing Item After Intervention (n = 228), No. (%)
Beverage (milk, juice, water, soda, other sweetened beverage, coffee or tea)	124 (56.9)	153 (67.1)
Snack	63 (28.9)	63 (27.6)
Grocery Item (bread, eggs, meat, cereal, rice, pasta, etc.)	65 (29.8)	39 (17.1)
Fresh produce	41 (18.8)	39 (17.1)
Sandwich or other deli item	26 (11.9)	30 (13.2)

Note. Excludes customers who did not purchase food or nonalcoholic beverages.

DISCUSSION

Our findings indicate that although improving the inventory of healthy foods in corner stores is feasible, changing customer purchases is more difficult. In particular, the results demonstrate the importance of outreach staff making multiple store visits (which is very resource intensive, because each full-time outreach worker managed 30 bodegas), which allows staff to develop relationships with store owners to implement changes over time. Also, simple changes, such as adding healthier versions of products already for sale and moving healthier items to make them more prominent, were the most successful. Although the number of stores meeting healthy store levels 1 and 2 increased significantly, achieving level 3 proved challenging, namely because of the criterion of stocking a leafy green vegetable. Of note, owners were reluctant to purchase a perishable item that they believed would not sell, indicating that working with communities to increase demand for healthier items at bodegas is key. As new healthier products appear in the store’s inventory, customers may need motivation and time to adjust their purchasing choices. The initiative addressed this issue by working to build partnerships between community organizations and bodegas. However, interested organizations were often located several blocks from participating stores, which made it difficult for them to patronize stores regularly.

On the basis of our findings, we made some changes to our intervention model. We revised our criteria to give stores more choice in the changes they make. Additionally, we recognized the importance of engaging and empowering community members to participate in the process. Bodegas are now recruited on the basis of proximity to potential collaborating organizations so that organizations can support stores they already patronize and sustain inventory changes even after the intervention ends. In response to store owner feedback, we have increased the number of cooking demonstrations at bodegas and expanded our selection of in-store promotional materials. We have also expanded our incentives to include fruit salad starter kits (includes produce peeler, knife, cutting board, and containers), produce storage shelves and baskets, blenders, and small refrigerators, all of which are low cost and make it easier for the store to stock healthier items. Because of funding limitations, we were unable to offer stores financial support for larger scale infrastructure such as air conditioners, refrigeration, or shelving for large produce displays. This investment might have had a substantial impact on the store owners’ capacity to increase healthy offerings and should be considered if funding allows. However, the utility cost for such items may be prohibitive for some stores and should be taken into consideration.

Limitations

Limitations of this study include fluctuations in the inventory of products assessed through the in-store observations (although we attempted to address this by observing each store 3 times) and variability in applying the criterion for low-sodium canned goods, which led to inconsistencies in how this was measured, most likely overestimating the change seen in this area. Additionally, this evaluation relied on self-report from store owners regarding increases in sales of healthy foods rather than on sales data because of the challenges with invoice collection mentioned previously.

Several limitations were also associated with the consumer surveys conducted in this evaluation. Because interviewers were at stores for just 4 hours and a substantial number of customers did not agree to do the surveys, the surveys captured a very small portion of store purchases. Additionally, we were only able to conduct consumer surveys at 8 of the 55 stores that completed the campaign, so results may not be representative. Because of the small number of purchases captured in our consumer survey data, we were unable to explore the relationship between changes in consumer purchases and positive changes made by stores. Additionally, the consumer surveys relied on nonrandom selection of locations owing to dependence on store owner cooperation. Given these limitations, in particular the limited consumer purchases captured through this method, future evaluations may need to target data collection at other or more times of the day. Finally, this particular evaluation did not include a control group of stores or customers.

Conclusions

Given the abundance of corner stores and their important role in underserved communities, public health advocates and community leaders are increasingly seeking strategies to expand their healthy food offerings. The innovative approach described in this article was successful in significantly improving healthy food inventory in participating stores and modestly improving customer purchases. Behavior change among customers may take time as new products are introduced and promoted. Targeted outreach, community engagement, and dialogue with store owners may be sufficient to effect more immediate change in smaller stores that rely heavily on customer demand and loyalty. However, to fully address the shortage of healthy foods in underserved neighborhoods, a full range of efforts targeting all areas of food retail, from seasonal to year round, large to small, must be implemented.