Preventive Medicine 65 (2014) 7–12

Contents lists available at ScienceDirect

Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed

Changes in food and beverage environments after an urban corner store intervention☆ Erica Cavanaugh a,⁎, Sarah Green a, Giridhar Mallya c, Ann Tierney a, Colleen Brensinger a, Karen Glanz a,b a b c

Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA School of Nursing, University of Pennsylvania, Philadelphia, PA 19104, USA Philadelphia Department of Public Health, 1401 JFK Boulevard, Philadelphia, PA 19102, USA

a r t i c l e

i n f o

Available online 13 April 2014 Keywords: Food environment Corner stores Obesity prevention NEMS

a b s t r a c t Objective. In response to the obesity epidemic, interventions to improve the food environment in corner stores have gained attention. This study evaluated the availability, quality, and price of foods in Philadelphia corner stores before and after a healthy corner store intervention with two levels of intervention intensity (“basic” and “conversion”). Methods. Observational measures of the food environment were completed in 2011 and again in 2012 in corner stores participating in the intervention, using the Nutrition Environment Measures Survey for Corner Stores (NEMS-CS). Main analyses included the 211 stores evaluated at both time-points. A time-by-treatment interaction analysis was used to evaluate the changes in NEMS-CS scores by intervention level over time. Results. Availability of fresh fruit increased significantly in conversion stores over time. Specifically, there were significant increases in the availability of apples, oranges, grapes, and broccoli in conversion stores over time. Conversion stores showed a trend toward a significantly larger increase in the availability score compared to basic stores over time. Conclusion. Interventions aimed at increasing healthy food availability are associated with improvements in the availability of low-fat milk, fruits, and some vegetables, especially when infrastructure changes, such as refrigeration and shelving enhancements, are offered. © 2014 Elsevier Inc. All rights reserved.

Introduction Neighborhood food environments are getting increasing attention as intervention targets for addressing the obesity epidemic. Small neighborhood food stores are an obvious choice for intervention due to their convenience; their tendency to charge higher prices for healthier foods (Krukowski et al., 2010); limited availability of fruit, vegetables, and low-fat milk (Leone et al., 2011); and their contributions to daily caloric consumption (Borradaile et al., 2009). Reports on the feasibility and effectiveness of such interventions are now appearing in the literature (Adams et al., 2012; Dannefer et al., 2012; Gittelsohn et al., 2012)

☆ This project was funded by Cooperative Agreement #3U58DP002626-01S1 from the Centers for Disease Control and Prevention and Get Healthy Philly, an initiative of the Philadelphia Department of Public Health. The views expressed in this manuscript do not necessarily reflect the official policies of the Department of Health and Human Services or the City of Philadelphia; nor does mention of trade names, commercial practices, or organizations imply endorsement by the U.S. Government. ⁎ Corresponding author at: 802 Blockley Hall, 423 Guardian Drive, Philadelphia, PA 19104, USA. Fax: +1 215 573 5315. E-mail addresses: [email protected] (E. Cavanaugh), [email protected] (S. Green), [email protected] (G. Mallya), [email protected] (A. Tierney), [email protected] (C. Brensinger), [email protected] (K. Glanz).

http://dx.doi.org/10.1016/j.ypmed.2014.04.009 0091-7435/© 2014 Elsevier Inc. All rights reserved.

with the majority focusing on increasing healthy food access and/or access to fresh fruit and vegetables (Gittelsohn et al., 2012). Still, reports are few, and based mainly on self-report of store owners/managers or customers from a small number of stores. More information on objectively measured outcomes is needed to establish the feasibility and effectiveness of these types of interventions. This article describes the findings of store food environment assessments for more than 200 corner stores in Philadelphia, PA, conducted before and after a healthy corner store intervention as part of the Get Healthy Philly initiative (www.foodfitphilly.org) by an external evaluation team. Because stores received either a “basic” intervention or a more intensive “conversion” intervention, two evaluation questions were addressed: 1) do healthy food environments in participating stores improve over time, and 2) do stores receiving a more intensive intervention show greater improvement in their healthy food environments over time? Methods Evaluation design Observational measures of food environments were completed in a sample of corner stores participating in the Healthy Corner Store Initiative intervention,

8

E. Cavanaugh et al. / Preventive Medicine 65 (2014) 7–12

at two time points approximately one year apart, using the Nutrition Environment Measures Survey for Corner Stores (NEMS-CS). This evaluation examined changes over time, and the difference between the “basic” and “conversion” (intensive intervention) stores over time. Healthy Corner Store Initiative intervention

stores. At this point, these stores were still in the early stages of the basic intervention and had also been designated for future conversion intervention. NEMS-CS measures were completed in 233 (94.7%) of the 246 stores at baseline (Cavanaugh et al., 2013). All stores that were assessed at baseline (n = 233) were included in the follow-up sample. Of these, only 211 were assessed at follow-up—161 were basic stores and 50 were conversion stores. For evaluation purposes, a store was considered a conversion store if the NEMS-CS data were collected after the post-conversion completion date. This date was two weeks after equipment delivery and provided the store time to complete their training and stock healthier products before a conversion was considered ‘complete’.

In March 2010, the Philadelphia Department of Public Health (PDPH) partnered with The Food Trust to implement the Healthy Corner Store Initiative (HCSI) on a city-wide scale in Philadelphia. The Food Trust's HCSI is a nationally recognized model to improve access to and availability of healthy foods in urban areas by providing technical assistance, training, and capital investments to store owners (The Food Trust, 2012). Corner stores were defined as businesses with food as their primary product, having less than 2000 square feet, fewer than 4 aisles, and 1 cash register (The Food Trust, 2012). Recruitments of stores began in areas of the city with high rates of poverty (N 19% of the population living below the federal poverty level). Proximity to schools and other childserving institutions was also considered. Store owners were approached individually to assess their willingness and ability to participate. In exchange for a $100 incentive, enrolled stores began by adding two new healthy products from two healthy food categories to their inventory and implementing a Healthy Food Identification marketing campaign, including window and door clings, in-store banners, shelf labels, and recipe cards. The seven healthy food categories include: fresh fruits and vegetables, canned/ frozen fruits and vegetables, low-fat dairy, lean meats, whole grains, healthy snacks, and healthy beverages. The first five food categories were emphasized in the intervention. Next, stores were offered two business trainings conducted one-on-one at stores over 30–60 min and focused on product procurement, promotion, and pricing and, as needed, SNAP certification. This set of activities constitutes a “basic” intervention. Stores that met the basic intervention goals and demonstrated an ability to plan and execute larger inventory changes were recruited into the high-intensity intervention. These stores received mini-grants for shelving and refrigeration to help them store, display, and expand their inventory of healthy products; and individualized businesses training. These efforts, in addition to the basic intervention components, constitute a “conversion” intervention. On average, the time from enrollment to completion of inventory and marketing changes was 3 months during which project staff made approximately 2 visits per store. Completion of the conversion intervention required an additional 2–3 months and 2–4 visits per stores. Business trainings occurred during the first 6–9 months after enrollment. To help store owners procure healthy foods, The Food Trust developed partnerships with local food distributors and suppliers that stock, identify, and sell new products at reasonable costs and volumes. A total of 630 stores were participating in the program by the end of the evaluation period.

The main analyses and comparisons were limited to stores that were evaluated at both time points. Descriptive statistics on availability, pricing, and quality for all healthy and less-healthy options were computed for all stores, and by store type (basic or conversion store). A total NEMS-CS score (−9–49 points) and scores for the dimensions of availability (0–25 points), price (− 9–18 points), and quality (0–6 points) were calculated for each store using a standard scoring system (available by request). Means and standard deviations were used to summarize NEMS-CS scores. McNemar's test for matched pair data and paired t-tests were used to evaluate the changes in availability of healthy and less-healthy items, and the NEMS-CS scores and subscales for availability, quality and price at baseline and followup assessments. A time-by-treatment interaction analysis was used to evaluate the changes in NEMS-CS scores by intervention level (basic or conversion) over time (baseline to follow-up). Due to the number of outcomes tested, a threshold of 0.01 was used to assess statistical significance. An exploratory analysis was also conduced to examine whether there was a systematic difference in NEMS-CS scores associated with time from enrollment to baseline assessment. This was important given that the interventions had begun before the evaluation, limiting our ability to have a ‘true baseline’. Analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC).

Timing of measures

Results

Funding for evaluation became available after the HCSI intervention began. Stores were selected for the evaluation to baseline as possible (see Sample of corner stores section below). Baseline data collection was completed between January and April of 2011 and a follow-up data collection was completed between January and April of 2012. The assessments were conducted approximately one year apart, to control for seasonality. Trained evaluation staff from the University of Pennsylvania completed all data collection at both time points.

Of the 233 stores rated at baseline, 211 of those stores were also rated at the follow-up time point (90.6%; n = 161 basic stores and n = 50 conversion stores). Seven stores (3.0%) were excluded at the store owner's request and 15 stores (6.4%) were found to have closed. There was no significant difference in the total NEMS-CS score at baseline for stores rated at baseline only (mean = 14.6, SD = 6.53) and those rated at both time points (mean = 16.5, SD = 5.57).

Sample of corner stores A multi-stage, stratified sampling procedure was used to achieve a sufficient sample of basic and conversion stores. The initial sample of 220 stores was chosen from all the stores enrolled in the intervention that were located in highpoverty areas (N 19% of residents living below the federal poverty level, approx. 92% of participating stores). At that time, all stores were designated to receive the basic intervention, and determinations regarding receipt of the highintensity intervention had not been completed. To ensure a sufficient sample of stores that would eventually receive the conversion intervention, the Conversion Potential Rating (CPR) was used as a proxy. The CPR was determined at the time of intervention enrollment by The Food Trust intervention staff. It estimated a store's potential to progress through the program, reflecting the degree of store owner commitment, the availability of space for new displays, and surrounding assets, such as schools. An additional 26 stores were chosen at a later date—for a total sample of 246 stores—to increase the sample of conversion

Measurement tool: Nutrition Environment Measures Survey for Corner Stores (NEMS-CS) The NEMS for Corner Stores (NEMS-CS) was adapted from the Nutrition Environment Measures Survey for Stores (NEMS-S) for use in corner stores (Cavanaugh et al., 2013). NEMS-S is a validated observational measure of retail store nutrition environments focusing on the availability of healthful food choices, quality of fresh produce, and price of healthy vs. comparable less healthy options, in 11 common categories (Glanz et al., 2007, www.med. upenn.edu/nems/). Statistical methods

Characteristics of stores Of the 211 stores that were rated at both time points, 76.3% (n = 161) were basic stores and 23.7% (n = 50) were conversion stores. More than 95% of the stores had only one cash register and used a majority (N50%) of the store space for food. The stores were located in various parts of the city, with 26.5% in Central Philadelphia (n = 56), 19.4% in West/Southwest Philadelphia (n = 41), 18.5% in North Philadelphia (n = 39), 18.0% in Northeast Philadelphia (n = 38), and 17.5% in South/South Central Philadelphia (n = 37). All 211 stores (100%) were located in high priority zip codes. The number of days between enrollment into the intervention and the baseline assessment date ranged from 8 to 350 days, with a mean of 192.44 (SD = 83.7) days and a median of 211 days. There was a

E. Cavanaugh et al. / Preventive Medicine 65 (2014) 7–12

very small, non-significant trend in total NEMS-CS scores, with stores measured more than 4 months after enrollment receiving scores about 1 to 1.5 points higher. Among the conversion stores, the number of days between the completion of the intervention and the follow-up NEMS-CS evaluation ranged from 0 to 410 days, with a mean of 146.7 (SD = 113.7) days and a median of 130 days. Most of the conversion stores (71.5%) had been converted for at least 2 months prior to follow-up assessment; 34.5% had been converted for 2–6 months (n = 28) and 37.0% of the conversion stores had been converted for more than six months (n = 30). To assess for potential attrition bias, differences between stores rated at baseline only (n = 233) versus both time points (n = 211) were explored. There was no significant difference in the total NEMS-CS score at baseline for stores rated at baseline only (mean = 9.36, SD = 5.54) and those rated at both time points (mean = 11.45, SD = 4.84).

9

increases were not statistically different from one another. There were no significant changes over time in the availability of whole wheat bread, lower sugar cereal, baked chips, and 100 calorie snack packs in basic or conversion stores. In the fruits and vegetables category, there was a trend towards an increase in the availability of any fresh fruit in conversion stores as compared to basic stores over time (p = 0.029), with 78.0% (n = 39) of conversion stores having any fresh fruit available at follow-up. There were no other significant changes over time in basic or conversion stores in the fruits and vegetables category. Table 2 summarizes the availability of specific fresh fruits and vegetables at baseline and follow-up assessments by intervention level. For basic stores, there were no significant changes over time, but for conversion stores there were significant increases in the availability of apples, oranges, and broccoli and a significant decrease in cabbage availability. When comparing conversion stores to basic stores over time, there were significant increases in the availability of apples, oranges, grapes, and broccoli in conversion stores.

Comparison of basic and conversion stores over time Comparison of fresh fruit and vegetable quantities over time Table 1 summarizes the availability of healthier food and healthy corner store marketing materials at baseline and follow-up by intervention level. In the meats and frozen meal category, there were no significant changes over time in the availability of healthier products in basic or conversion stores. The vast majority of the stores did not have lean ground beef, light or fat-free hot dogs, or reduced-fat frozen dinners available at either time point. In the beverage category, there was an increase in low-fat milk (1% or skim) availability in conversion stores compared to basic stores over time (p = 0.031), although this did not meet the 0.01 threshold for statistical significance. There were no changes in the availability of diet soda, 100% juice, or bottled water over time. Many stores had non-carbonated, zero or low-calorie drinks available, and there was a significant decrease in availability at follow-up for basic stores. In the baked goods category, there was a significant increase in the availability of fat-free or low-fat baked goods over time for both basic (from 7% to 20%) and conversion stores (from 0% to 16%); these

Table 3 summarizes the quantity of the top three fresh fruit and vegetables available at baseline and follow-up assessments by intervention level. Among basic stores, there were no significant differences over time in the quantities of fresh fruit available, but there were increases in the quantities of sweet peppers. Among conversion stores, the quantity of available sweet peppers increased significantly over time. NEMS-CS scores for availability, price and quality by store type Table 4 shows the NEMS-CS scores—total score and the dimensions of availability, price, and quality—at baseline and follow-up assessments by intervention level. The total scores, as well as the scores for availability and health-favorable pricing, were generally low at both time points. Among basic stores, the price score (p b 0.001) and total score (p b 0.001) increased significantly over time. Among conversion stores, the price score (p b 0.001), and total score (p b 0.001) increased significantly over time. When comparing conversion stores to basic stores

Table 1 Availability of healthier food in Philadelphia corner stores at baseline (2011) and follow-up (2012), by store type. Basic stores (N = 161) Baseline n HCS materials Meats and frozen meals Lean ground beef Light, fat-free, or turkey hot dogs Reduced-fat frozen dinners Beverages Lower-fat milk (2%, 1%, or skim) Low-fat milk (1% or skim) Diet soda 100% juice Non-carbonated, zero or low-calorie drinks Bottled water Baked goods, snacks and grains Fat-free or low-fat baked goods 100% whole wheat bread Baked chips 100 calorie snack packs Lower sugar cereal Fruit and vegetables Any fresh fruit Any frozen fruit Canned fruit in 100% fruit juice, light syrup, or water Any fresh vegetables Frozen vegetables Canned vegetables in water, no sauce a

Significant within group differences (p b 0.01).

Conversion stores (N = 50) Follow-up

%

n

Baseline %

n a

Interaction

Follow-up %

n

%

p-Value

90.7%

48

96.0%

49

98.0%

0.330

0.0% 5.0% 5.0%

1 3 2

2.0% 6.0% 4.0%

0 4 4

0% 8.0% 8.0%

NA 0.383 0.174

78.3% 14.3% 93.8% 96.3% 50.3%a 95.7%

40 11 46 47 33 49

80.0% 22.0% 92.0% 94.0% 66.0% 98.0%

41 14 47 47 26 49

82.0% 28.0% 94.0% 94.0% 52.0% 98.0%

0.304 0.031 0.620 0.542 0.497 0.455

31 83 57 4 137

20.%a 51.6% 35.4% 2.5% 85.1%

0 34 16 1 40

0.0% 68.0% 32.0% 2.0% 80.0%

8 33 16 1 41

16.3%a 66.0% 32.0% 2.0% 82.0%

0.854 0.219 0.904 0.953 0.940

84 4 119 129 67 150

52.2% 2.5% 73.9% 80.1% 41.9% 93.8%

37 0 34 39 20 48

74.0% 0.0% 68.0% 78.0% 40.0% 96.0%

39 1 33 43 20 49

78.0% 2.0% 66.0% 86.0% 40.0% 98.0%

0.029 0.704 0.624 0.278 0.730 0.316

129

80.1%

146

1 12 12

0.6% 7.5% 7.5%

134 27 152 153 120 156

83.2% 16.8% 94.4% 95.0% 75.0% 97.5%

126 23 151 155 81 154

11 88 61 8 138

7.0% 54.7% 37.9% 5.0% 85.7%

86 6 124 126 71 150

53.4% 3.7% 77.0% 78.3% 44.1% 93.2%

0 8 8

10

E. Cavanaugh et al. / Preventive Medicine 65 (2014) 7–12

Table 2 Availability of fruits and vegetables in Philadelphia corner stores at baseline (2011) and follow-up (2012), by store type. Basic Stores (N = 161) Baseline n Fruit Bananas Apples Oranges Grapes Strawberries Pears Peaches Cantaloupe Watermelon Honeydew Vegetables Tomatoes Sweet Peppers Lettuce Carrots Cucumbers Celery Cabbage Broccoli Cauliflower Corn

Conversion Stores (N = 50) Follow-up

%

Baseline

n

Interaction Follow-up

%

n

%

n

%

p-Value

70 58 55 22 13 13 13 9 8 6

43.5% 36.0% 34.2% 13.7% 8.1% 8.1% 8.1% 5.6% 5.0% 3.7%

71 52 46 17 15 17 7 6 3 3

44.1% 32.3% 28.6% 10.6% 9.3% 10.6% 4.4% 3.7% 1.9% 1.9%

31 18 17 4 6 4 3 4 5 1

62.0% 36.0% 34.0% 8.0% 12.0% 8.0% 6.0% 8.0% 10.0% 2.0%

29 25 27 9 3 5 2 2 3 1

58.0% 50.0% 54.0% 18.0% 6.0% 10.0% 4.0% 4.0% 6.0% 2.0%

0.556 0.005 b0.001 0.004 0.166 0.900 0.887 0.767 0.328 0.867

108 106 97 30 23 21 18 13 12 7

67.1% 65.8% 60.3% 18.6% 14.3% 13.0% 11.2% 8.1% 7.5% 4.4%

105 96 98 26 16 19 13 11 9 4

65.2% 59.6% 60.9% 16.2% 9.9% 11.8% 8.1% 6.8% 5.6% 2.5%

33 27 33 14 8 6 7 1 1 2

66.0% 54.0% 66.0% 28.0% 16.0% 12.0% 14.0% 2.0% 2.0% 4.0%

39 29 36 17 6 8 2 5 1 2

78.0% 58.0% 72.0% 34.0% 12.0% 16.0% 4.0% 10.0% 2.0% 4.0%

0.037 0.737 0.218 0.023 0.769 0.230 0.076 0.004 0.584 0.519

No items had significant within group differences (p b 0.01).

over time, the increase in availability score for conversion stores compared to basic stores approached significance (p = 0.019). There were no significant differences in the price score, quality score, or total NEMS-CS scores by intervention level over time. Exploratory analyses were conducted in the 50 conversion stores to see if changes in food environment scores were associated with the time between conversion implementation and follow-up assessment. Comparisons were made between stores with less than 2 months, 2–6 months, and more than 6 months between the conversion completion date and the NEMS-CS follow-up date. There were no significant differences in availability, price, quality, or total scores by time between conversion and follow-up assessments.

Discussion Using a standardized nutrition environment assessment of 211 corner stores, we found that there were modest improvements in the food environments in both basic and intensive (“conversion”) stores where healthy corner store interventions were conducted. Further, the conversion intervention—which included shelving and refrigeration enhancements—led to greater improvements in the availability of low-fat milk, fresh fruits, and some vegetables over time. To the authors' knowledge, this study represents the largest published assessment of a corner store intervention using objective observational measures. Overall, the findings show that a healthy corner store initiative implemented at a

Table 3 Quantity of select fruit and vegetables available in Philadelphia corner stores at baseline (2011) and follow-up (2012), by store type. Basic stores (N = 161) Quantity

Conversion stores (N = 50)

Baseline

Follow-up

p-Value

n

%

n

%

0 1–9 10+ 0 1–9 10+ 0 1–9 10+

93 13 55 104 12 45 106 13 42

57.7% 8.1% 34.2% 64.6% 7.5% 28.0% 65.8% 8.1% 26.1%

90 12 59 110 14 37 116 8 37

55.9% 7.5% 36.7% 68.3% 8.7% 23.0% 72.1% 5.0% 23.0%

Top 3 available vegetables Tomatoes 0 1–9 10+ Sweet peppers 0 1–9 10+ Lettuce 0 1–9 10+

55 35 71 111 40 10 66 58 37

34.2% 21.7% 44.1% 68.9% 24.8% 6.2% 41.0% 36.0% 23.0%

56 27 78 66 49 46 63 56 42

34.8% 16.8% 48.5% 41.0% 30.4% 28.6% 39.1% 34.8% 26.1%

Top 3 available fruits Bananas

Apples

Oranges

0.933

0.254

0.224

0.612

≤0.001

0.658

Quantity

Baseline

Follow-up

p-Value

n

%

n

%

0 1–9 10+ 0 1–9 10+ 0 1–9 10+

19 5 26 32 6 12 34 5 11

38.0% 10.0% 52.0% 64.0% 12.0% 24.0% 68.0% 10.0% 22.0%

23 4 23 25 8 17 23 10 17

46.0% 8.0% 46.0% 50.0% 16.0% 34.0% 46.0% 20.0% 34.0%

0 1–9 10+ 0 1–9 10+ 0 1–9 10+

17 8 25 34 13 3 17 23 10

34.0% 16.0% 50.0% 68.0% 26.0% 6.0% 34.0% 46.0% 20.0%

12 12 26 23 14 13 14 21 15

24.0% 24.0% 52.0% 46.0% 28.0% 26.0% 28.0% 42.0% 30.0%

0.719

0.092

0.054

0.461

0.009

0.423

E. Cavanaugh et al. / Preventive Medicine 65 (2014) 7–12

11

Table 4 NEMS-CS scores in Philadelphia corner stores at baseline (2011) and follow-up (2012), by store type. Basic stores (N = 161) Baseline

Total availability score Total price score Total quality score Total score

Conversion stores (N = 50) Follow-up

Baseline

Interaction Follow-up

Mean

SD

Min, max

Mean

SD

Min, max

Mean

SD

Min, max

Mean

SD

Min, max

p-Value

7.3 0.2 3.8 11.3

3.20 1.01 2.27 4.92

0, 18 −3, 3 0, 6 0, 26

6.8 1.5 3.7 12.0

3.19 1.57⁎ 2.11 5.22⁎

0, 17 −2, 5 0, 6 0, 27

7.1 0.5 4.3 11.9

3.09 1.33 2.14 4.58

1, 16 −3, 4 0, 6 1, 21

7.5 1.6 4.6 13.8

2.68 1.79⁎ 1.80 4.77⁎

3, 15 −2, 6 0, 6 5, 26

0.019 0.657 0.224 0.059

Modified NEMS-CS Scoring. Possible ranges: Availability (0–25 points); Price (−9–18 points); Quality (0–6 points); Total (−9–49 points). ⁎ Significantly higher score in all stores at follow-up compared to baseline (p b 0.001).

citywide scale can lead to increases in availability of some categories of healthy foods. The Philadelphia Healthy Corner Store Initiative (HCSI) enrolled over 600 stores out of approximately 1800 in the entire city. The basic intervention—which included inventory standards, promotional materials, and business training—resulted in small improvements in low-fat/low-calorie baked good availability, sweet pepper and celery quantity, and NEMS price and total scores. These changes are notable but may not be sufficient to impact customer purchasing behaviors and food intake. The more intensive “conversion” intervention that led to larger improvements over time than the basic intervention shows promise. However, because stores were not randomly assigned to intervention levels, the intensive intervention may be more effective only in stores with a higher degree of readiness to change. These findings are consistent with a recent review of 16 corner store interventions, including 11 set in low-income urban settings, that demonstrated that availability, sales, and purchasing of healthy foods increased post-intervention (Gittelsohn et al., 2012). Most of these studies assessed relatively small samples, and assessment methodologies varied widely. Two evaluations of larger interventions showed divergent results (Adams et al., 2012; Dannefer et al., 2012). Dannefer et al. assessed 55 of 1000 corner stores enrolled in the Healthy Bodegas Initiative in New York City via in-store observations and owner and customer surveys. They found improvements in the availability of canned fruit, low/no-salt canned goods, healthy snacks, and whole grain bread. There were no significant changes in the availability of fresh fruits or fresh vegetables like those seen in the present study. Adams et al. assessed 87 of the 87 convenience stores enrolled in the Change4Life Programme in Northeast England. Intervention fidelity was poor, resulting in inappropriate use of promotional materials and refrigeration units. Interestingly, similar challenges for retail store healthy food interventions may occur in settings outside the US. Based on the review by Gittelsohn and others, strategies to improve healthy food availability have included inventory standards and, less commonly, store owner training and small infrastructural enhancements (refrigeration, shelving) (Gittelsohn et al., 2012). Strategies used to promote healthy food purchases included in-store promotional materials, community engagement, and, less commonly, pricing modifications (incentives and coupons). The Philadelphia HCSI used many of these approaches, including inventory standards (focusing on 5 healthy food categories—fresh fruits and vegetables, canned/frozen fruits and vegetables, low-fat dairy, lean meats, and whole grains), promotional materials, business training, equipment provision and community engagement. Business trainings focused on storage and procurement of healthy foods, particularly fresh produce, pricing and placement strategies, and SNAP certification, which may contribute to the financial sustainability of small stores (13% of stores were not SNAP certified at entry). With regard to equipment, approximately 43% of the conversion stores chose refrigeration units only, 12% chose shelving only, and 45% chose both. Most store owners placed the new equipment in high-visibility locations, like at store entrances or near the cash register. Implementation of the basic

intervention cost approximately $300–$400 per store, while the conversion intervention averaged $1390 per store (both estimates do not include program staff-related costs). A unique aspect of the Philadelphia HCSI was the partnership developed by The Food Trust with a locally-based national distributor, Jetro. Approximately 80% of enrolled stores reported using Jetro regularly as a source of food products. Jetro management engaged in the initiative by labeling and co-locating foods that met HCSI standards, offering competitive prices, and making produce available in smaller portions more appropriate for corner stores than larger grocers. The Philadelphia HCSI might have achieved more substantial changes with larger incentives for owners, such as renovation grants or low/ no interest loans. Other promising approaches might include an explicit focus on lower prices for consumers (Andreyeva et al., 2010; Gittelsohn et al., 2012), requirements to reduce unhealthy food availability and/or promotion (Farley et al., 2009; Glanz et al., 2012; Lucan et al., 2010). However, these interventions would require more external funding and might not be sustainable in the long term. While 100 stores received conversions through CPPW funds in 2010 and 2011, an additional 196 stores received conversions through other funding in 2011 and 2012, bringing the total to 296. In approximately fifty stores, sugary drink counter-marketing emphasizing calorie equivalents will soon be added to refrigerated cases (Bleich et al., 2012), and in December 2012, Philadelphia City Council passed an amendment to the zoning code limiting retail advertising (in a content-neutral fashion) to 20% of window and door space. Last, The Food Trust and the Philadelphia Department of Public Health have begun the implementation of a Healthy Corner Store Certification Program, which includes more extensive inventory, promotional, and pricing requirements with more tailored and repeated business trainings and façade-improvement grants as incentives. Strengths of this evaluation included the large sample of stores, validated assessment tools, and two levels of intervention to compare to one another. However, this study also had several limitations. First, it did not measure sales at the store- or product-level or customer purchases. Second, this study also did not assess the intervention's impact on consumer knowledge, attitudes, and norms; store revenue, profit, and loss; or, specifically, facilitators and barriers to implementation. Third, it did not assess effects on customers' larger dietary patterns or diet-related health outcomes. Fourth, stores were not randomly assigned to the high-intensity intervention. Rather, they were designated to receive the conversion intervention based on their demonstrated commitment to the initiative. Therefore, environmental improvements reflect an interaction between store-level factors, some of which were not measured, and the impact of the intervention. Lastly, while the evaluation was planned to include true pre-intervention baseline measures, the baseline NEMS assessments were completed on average 192 days after stores were enrolled in the intervention and after many stores had implemented components of the basic intervention (inventory changes, marketing materials). Therefore, findings of this study reflect store changes during and after the intervention rather than a true pre–post analysis. Because of the delayed baseline data collection, this

12

E. Cavanaugh et al. / Preventive Medicine 65 (2014) 7–12

evaluation yields a more conservative estimate of the impact of the intervention than it would if data were collected before the intervention began.

Conclusion This study represents the largest published assessment of a healthy corner store intervention aimed at increasing healthy food availability. Overall, it shows that such interventions—when carefully planned— can be implemented at scale to improve the availability of low-fat milk, fruits, and some vegetables, particularly when additional training and refrigeration and shelving enhancements are offered. As small store interventions proliferate, strategies from this intervention and others that have been carefully evaluated should inform program design and implementation.

Conflict of interest statement The authors declare that there are no conflicts of interest.

Acknowledgments We acknowledge the following individuals who contributed to this research: Olivia Hamilton, Lisa Colby, Marcia Commins, Luke Bingamin, Amy vonSydow Green, Scott Schmidt, Gabriela Abrishamian-Garcia, Andrea Cheung, Nathan Angerett, Winnie Wang, and The Food Trust staff.

References Adams, J., Halligan, J., Burges Watson, D., et al., 2012. The Change4Life convenience store programme to increase retail access to fresh fruit and vegetables: a mixed methods process evaluation. PLoS One 7, e39431. Andreyeva, T., Long, M.W., Brownell, K.D., 2010. The impact of food prices on consumption: a systematic review of research on the price elasticity of demand for food. Am. J. Public Health 100, 216–222. Bleich, S.N., Herring, B.J., Flagg, D.D., Gary-Webb, T.L., 2012. Reduction in purchases of sugar-sweetened beverages among low-income Black adolescents after exposure to caloric information. Am. J. Public Health 102, 329–335. Borradaile, K.E., Sherman, S., Vander Veur, S.S., et al., 2009. Snacking in children: the role of urban corner stores. Pediatrics 124, 1293–1298. Cavanaugh, E., Mallya, G., Brensinger, C., Tierney, A., Glanz, K., 2013. Nutrition environments in corner stores in Philadelphia. Prev. Med. 56, 149–151. Dannefer, R., Williams, D.A., Baronberg, S., Silver, L., 2012. Healthy bodegas: increasing and promoting healthy foods at corner stores in New York City. Am. J. Public Health 102, e27–e31. Farley, T.A., Rice, J., Bodor, J.N., Cohen, D.A., Bluthenthal, R.N., Rose, D., 2009. Measuring the food environment: shelf space of fruits, vegetables, and snack foods in stores. J. Urban Health 86, 672–682. Gittelsohn, J., Rowan, M., Gadhoke, P., 2012. Interventions in small food stores to change the food environment, improve diet, and reduce risk of chronic disease. Prev. Chronic Dis. 9, E59. Glanz, K., Sallis, J.F., Saelens, B.E., Frank, L.D., 2007. Nutrition Environment Measures Survey in stores (NEMS-S): development and evaluation. Am. J. Prev. Med. 32, 282–289. Glanz, K., Bader, M.D., Iyer, S., 2012. Retail grocery store marketing strategies and obesity: an integrative review. Am. J. Prev. Med. 42, 503–512. Krukowski, R.A., West, D.S., Harvey-Berino, J., Elaine Prewitt, T., 2010. Neighborhood impact on healthy food availability and pricing in food stores. J. Community Health 35, 315–320. Leone, A.F., Rigby, S., Betterley, C., et al., 2011. Store type and demographic influence on the availability and price of healthful foods, Leon County, Florida, 2008. Prev. Chronic Dis. 8, A140. Lucan, S.C., Karpyn, A., Sherman, S., 2010. Storing empty calories and chronic disease risk: snack-food products, nutritive content, and manufacturers in Philadelphia corner stores. J. Urban Health 87, 394–409. The Food Trust, 2012. Philadelphia's Healthy Corner Store Initiative, 2010–2012. http:// thefoodtrust.org/uploads/media_items/hcsi-y2report-final.original.pdf.

Changes in food and beverage environments after an urban corner store intervention.

In response to the obesity epidemic, interventions to improve the food environment in corner stores have gained attention. This study evaluated the av...
232KB Sizes 2 Downloads 3 Views