Prepared Food Availability in U.S. Food Stores A National Study Shannon N. Zenk, PhD, Lisa M. Powell, PhD, Zeynep Isgor, PhD, Leah Rimkus, MPH, Dianne C. Barker, MHS, Frank J. Chaloupka, PhD Introduction: Prepared, ready-to-eat foods comprise a significant part of Americans’ diets and are increasingly obtained from food stores. Yet, little is known about the availability and healthfulness of prepared, ready-to-eat food offerings at stores. This study examines associations among community characteristics (racial/ethnic composition, poverty level, urbanicity) and availability of both healthier and less-healthy prepared foods in U.S. supermarkets, grocery stores, and convenience stores. Methods: Observational data were collected from 4,361 stores in 317 communities spanning 42 states in 2011 and 2012. Prepared food availability was assessed via one healthier food (salads or salad bar), three less-healthy items (pizza, hot dog/hamburger, taco/burrito/taquito), and one cold sandwich item. In 2014, multivariable generalized linear models were used to test associations with community characteristics.
Results: Overall, 63.6% of stores sold prepared foods, with 20.0% offering prepared salads and 36.4% offering at least one less-healthy item. Rural stores were 26% less likely to carry prepared salads (prevalence ratio [PR]¼0.74, 95% CI¼0.62, 0.88) and 14% more likely to carry at least one less-healthy prepared food item (PR¼1.14, 95% CI¼1.00, 1.30). Convenience stores in high-poverty communities were less likely to carry prepared salads than those in low-poverty communities (PR¼0.64, 95% CI¼0.47, 0.87). Among supermarkets, prepared salads were more likely to be carried in majority-white, low-poverty communities than in non-white, high-poverty communities.
Conclusions: Increasing the healthfulness of prepared foods within stores may offer an important opportunity to improve the food environment. (Am J Prev Med 2015;49(4):553–562) & 2015 American Journal of Preventive Medicine
Introduction
I
n 2007–2008, more than one third of American children, adolescents, and adults reported consuming fast food on a given day.1 Fast food comprised an average of 11.3% of American adults’ total daily calories in 2007–2010.2 Fast food intake is associated with higher intake of calories, fat, sodium, and sugar; lower intake of vegetables and micronutrients; and higher body weight From the College of Nursing (Zenk), School of Public Health (Powell), Institute for Health Research and Policy (Isgor, Rimkus), and Department of Economics (Chaloupka), University of Illinois at Chicago, Chicago, Illinois; and Barker Bi-Coastal Health Consultants Inc. (Barker), Calabasas, California Address correspondence to: Shannon N. Zenk, PhD, University of Illinois at Chicago, 845 S. Damen Ave., 9th Floor, Chicago IL 60612. E-mail: [email protected]. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2015.02.025
& 2015 American Journal of Preventive Medicine
and weight gain.1,3–9 Increasingly, sources of fast food, or food prepared and served quickly (www.merriam-web ster.com/dictionary/fast-food),10 extend well beyond traditional fast food restaurants (where customers order and pay at the counter and food is consumed either on premises or off premises as carryout11) to food stores. According to a 2010 report, almost two thirds of Americans indicated that they purchased prepared ready-to-eat/heat food from a grocery store or supermarket in the past month (www.packagedfacts.com/ Prepared-Foods-Ready-2694891/). In another survey, 82% reported that they purchased a prepared food or drink from a convenience store at least once a month (http://nrn.com/archive/c-store-growth-poses-threat-quickservice). Close location of sources of fast food and other prepared, ready-to-eat food to where individuals live, work, and play may contribute to their convenience and
Published by Elsevier Inc.
Am J Prev Med 2015;49(4):553–562
553
554
Zenk et al / Am J Prev Med 2015;49(4):553–562
utilization, and consequently poorer diet and weight outcomes. A 2005 study in the 20 largest U.S. cities found, for example, that 46% of neighborhoods had at least one fast food restaurant within walking distance.12 This likely underestimates the accessibility of fast food and other prepared food because it does not take into account availability in stores. A study of six rural Texas counties showed that the median distance to the nearest outlet at which fast food or other prepared food could be purchased decreased from 4.5 miles to 2.7 miles when all opportunities (traditional fast food restaurants and non-traditional outlets such as supermarkets, grocery stores, and convenience stores) were included.11 Although some research shows no association,13,14 other studies have found that living in neighborhoods with greater accessibility of fast food restaurants is associated with greater utilization, lower dietary quality, and higher body weight.6,13,15–20 In general, research suggests that low-income, racial/ ethnic minority, and some central-city neighborhoods in the U.S. may be disproportionately exposed to fast food restaurants as compared with more-advantaged neighborhoods.12,21–27 These communities may also have fewer healthy food options at food stores, even in supermarkets.28–33 However, limited empirical evidence exists on the healthfulness of prepared food offerings at food stores and how it may differ by community characteristics.34–36 As the first nationwide examination of prepared food availability in food stores, this study examines associations among community characteristics and the availability of both healthier and less-healthy prepared, ready-to-eat foods in U.S. supermarkets, grocery stores, and convenience stores.
Methods Study Design and Sample This study utilized 2 years of cross-sectional pooled data (2011– 2012) from the Bridging the Gap Community Obesity Measures Project (BTG-COMP), one of the only nationwide data sets of the directly observed food environment in the U.S. (www.bridgingth egapresearch.org/). As part of the BTG-COMP, aspects of the food environment were measured via direct observation in 317 communities spanning 42 states. The observed communities were school enrollment zones (or catchment areas from which the school draws its students) for nationally representative samples of eighth-, tenth-, and 12th-grade public school students in the continental U.S. from the Monitoring the Future study.37 A preliminary list of stores was generated annually by combining and de-duplicating separate business lists from InfoUSA and Dun & Bradstreet. After initially screening each store via telephone for eligibility (sold at a minimum snacks and drinks) and store type criteria (described in Measures section), a stratified probability sample of stores was selected in each community, with three store types (supermarkets, grocery stores, and “limited service” stores such as convenience stores, liquor stores, drug stores, dollar
stores) as the strata. Given limitations previously identified in the validity of commercial databases,38–40 a half-open interval procedure was implemented to help address error stemming from an incomplete list of businesses.41,42 This procedure involved observing stores sampled from the business list and stores discovered in the field.43 Field staff conducted data collection at 6,002 eligible food stores in 2011–2012 (completion rate, 495%). Approximately 2% of stores were missing data on one or more of the variables of interest for this analysis. Because few limited service stores other than convenience stores carry prepared foods (10.2% in this sample), other limited service stores were excluded, and results are based on the remaining 4,361 stores. This includes 620 supermarkets, 620 grocery stores, and 3,121 convenience stores. More-detailed descriptions of the sample design and data collection procedures are provided elsewhere.33,43
Measures Food and beverage availability at each store was measured using the Bridging the Gap Food Store Observation Form.44 This instrument assesses the availability of five prepared, ready-to-eat food items: cold sandwich (wrapped/ready-to-eat or made-toorder), prepared vegetable salad (excluding salads premixed with mayonnaise or dressing like coleslaw or potato salad) or salad bar, pizza, hot dog/hamburger, and taco/burrito/taquito. These items were based on a previously existing instrument and feedback on common prepared food items from field staff members who visited stores during pilot work in 2010.44 In addition, pizza, hot dogs, hamburgers, and Mexican mixed dishes are major contributors to energy and solid fat intake.45–47 Using these items, three dichotomous (none or at least one) outcome measures were derived: any prepared foods (at least one of the five items); availability of prepared vegetable salad or salad bar (hereafter referred to as “prepared salads”); and availability of less-healthy prepared food items (at least one of the following three items: pizza, hot dog/ hamburger, and taco/burrito/taquito). Because of wide variations in their ingredients, the cold sandwich item was included in the “any prepared foods” measure but was considered neither healthier nor less healthy. Food store type was operationalized using Food Marketing Institute descriptions (http://www.fmi.org/research-resources/ supermarket-facts) and prior studies.48–50 Supermarkets had fresh (unprocessed) meat, four or more cash registers, and at least two of the following: butcher, bakery, and deli. Grocery stores had fresh meat but did not meet the other supermarket criteria. Convenience stores had no fresh meat and sold a limited selection of staple groceries or other convenience items. American Community Survey (2007–2011) 5-year estimate data,51 aggregated based on census block groups intersecting the communities, were used to measure racial/ethnic and socioeconomic characteristics. Racial/ethnic composition was classified as majority white (Z50% residents non-Hispanic white) and nonwhite (not majority non-Hispanic white). Socioeconomic characteristics were measured by household poverty rate (tertiles). National Center for Education Statistics urban-centric locale codes were used to categorize community urbanicity as urban (small, midsize, and large cities); suburban (small, midsize, and large suburbs plus distant and fringe towns); and rural (distant, fringe, and remote rural areas plus remote towns).52 www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562 Several covariates helped to account for potential confounders, including census division/region (New England, Middle Atlantic, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, and Pacific); year; and store size as proxied by number of cash registers (one, two to three, four to seven, and eight or more).
555
Table 1. Weighted Descriptive Statistics for Communitya Characteristics All stores (n¼4,361) % Census division/region
Statistical Analysis
New England
4.1
Descriptive statistics for community characteristics were calculated for all stores. Percentages of stores that carried any of the assessed prepared foods, prepared salads, and less-healthy prepared food items were also calculated for all stores and by store type and community characteristics. Multivariable generalized linear models (GLM) with Poisson distribution were used to regress the two outcomes of primary interest—availability of prepared salads and availability of less-healthy prepared food items—on store type (in models based on all stores only); community racial/ethnic composition; poverty level; urbanicity; the other prepared food type; and covariates (region, year, and store size [stratified models only]). Each model was estimated for all stores and stratified by store type: supermarkets, grocery stores, and convenience stores. Generalized Hausman specification tests were used to test whether prevalence ratios (PRs) differed significantly across the stratified models. Multiplicative interaction terms between racial/ethnic composition and poverty level were added to each model in order to determine whether racial/ethnic composition–prepared food associations differed by poverty level. When interaction terms were statistically significant, estimated predicted probabilities for availability were calculated for contrasting communities, with all other covariates held at their original distributions. For example, the predicted probability of prepared salad availability was estimated for (1) communities set as being majority white and low-poverty and (2) communities set as being non-white and high-poverty, with all other covariates (e.g., region, urbanicity, year, store type) at their original distributions. All analyses were conducted using Stata, version 11.2, and completed in 2014. Sample weights were applied and estimation was undertaken (using svy commands) to account for the complex sample design for selection of communities and for stores clustered within communities, to obtain results for communities where nationally representative samples of eighth-, tenth-, and 12th-grade (traditional) public school students reside in the continental U.S. GLM with Poisson distribution was selected in order to estimate PRs for the relatively common outcomes.53
Middle Atlantic
10.6
West North Central
4.9
East North Central
10.4
South Atlantic
25.6
East South Central
6.0
West South Central
18.3
Mountain
4.7
Pacific
15.4
Results Table 1 shows characteristics of the store sample. Table 2 reports the prevalence of prepared, ready-to-eat foods and bivariate associations of the three prepared food measures with store type and community characteristics. Overall, 63.6% of stores sold prepared foods of the types we assessed (cold sandwich, salad, pizza, hot dog/hamburger, and taco/burrito/taquito). Thirty-six percent of stores offered at least one of the less-healthy items and considerably fewer stores (20.0%) offered prepared salads. October 2015
Racial/ethnic composition Majority white
69.9
Non-whiteb
30.1
Poverty levelc Low poverty
25.3
Moderate poverty
35.5
High poverty
39.3
d
Urbanicity Urban
37.0
Suburban
42.5
Rural
20.6
Store type Supermarket
11.3
Grocery store
11.0
Convenience store
77.8
a
Median area (minus water and military bases) of communities: 47.0 square miles (SD¼217.6; minimum¼0.1, maximum¼1833.0). b Not majority white (450% non-Hispanic black [5.4%], 450% Hispanic [7.3%], or Other [17.5%; not majority non-Hispanic white, non-Hispanic black, or Hispanic]). c The median household poverty rate was 14.0% in 2011 and 12.3% in 2012. In 2011, low poverty ranged from 1.8% to 8.7%, moderate poverty ranged from 8.8% to 16.2%, and high poverty ranged from 16.2% to 37.2%. In 2012, low poverty ranged from 1.9% to 7.9%, moderate poverty ranged from 8.1% to 12.7%, and high poverty ranged from 12.7% to 40.6%. d Urban was defined as small, midsize, and large cities; suburban was small, midsize, and large suburbs plus distant and fringe towns; and rural was distant, fringe, and remote rural areas plus remote towns, based on National Center for Education Statistics urban-centric locale codes.
Table 3 presents multivariable associations among community characteristics and availability of prepared salads, across all stores and stratified by store type.
Zenk et al / Am J Prev Med 2015;49(4):553–562
556
Table 2. Percentage of Stores That Sold Prepared Foods by Store Type and Community Characteristics Sold any prepared fooda % (95% CI)
Offered prepared vegetable salads or salad bar % (95% CI)
Offered any less-healthy prepared food itemsb % (95% CI)
63.6 (61.1, 66.1)
20.0 (18.0, 22.1)
36.4 (33.9, 39.0)f
Supermarkets
90.1 (87.0, 92.5)
74.5 (70.3, 78.2)
25.3 (21.3, 29.7)
Grocery stores
35.7 (31.2, 40.5)
16.2 (12.6, 20.6)
17.0 (13.6, 21.0)
63.7 (60.9, 66.5)
12.7 (10.9, 14.6)
40.8 (37.8, 43.9)
67.6 (64.9, 70.1)
22.6 (20.5, 24.9)
39.3 (36.3, 42.4)
54.5 (48.4, 60.4)
13.9 (10.8, 17.7)
29.7 (25.1, 34.7)
Low-poverty
68.4 (63.5, 72.9)
29.2 (26.3, 32.3)
39.6 (34.9, 44.5)
Moderate poverty
66.5 (63.0, 69.8)
20.8 (16.7, 25.5)
37.2 (32.6, 42.0)
57.9 (52.9, 62.8)
13.4 (11.2, 16.0)
33.7 (29.5, 38.1)
Suburban
60.6 (55.7, 65.4)
16.6 (13.6, 20.0)
31.8 (27.1, 37.0)
Urban
64.4 (60.9, 67.7)
24.7 (21.8, 27.8)
36.3 (33.3, 39.5)
Rural
67.5 (61.0, 73.3)
16.6 (13.6, 20.1)
44.9 (39.1, 50.8)
All stores (n¼4,361) Store typec,d,e
Convenience stores Community racial/ethnic composition Majority white Non-white Community poverty level
High-poverty Community urbanicity
c,d,e
c,d
d,e
At least one of the following five items: cold sandwich, vegetable-based salad/salad bar, pizza, hot dog/hamburger, Mexican (e.g., taco, burrito, taquito). b At least one of three items: pizza, hot dog/hamburger, Mexican (e.g., taco, burrito, taquito) c 2 χ test po0.05 for any prepared food. d 2 χ test po0.05 for prepared salads. e 2 χ test po0.05 for any less-healthy prepared food item. f 63.6% of stores carried none, 17.2% carried one item, 14.9% carried two items, and 4.3% carried three items. a
Overall, rural stores were 26% less likely to carry prepared salads than suburban stores (PR¼0.74, 95% CI¼0.62, 0.88). Stores in high-poverty communities were 29% less likely to carry prepared salads than those in lowpoverty communities (PR¼0.71, 95% CI¼0.59, 0.86). The prevalence of prepared salads in high-poverty communities was particularly low in convenience stores (PR¼0.64, 95% CI¼0.47, 0.87). Among supermarkets, there were statistically significant interactions between community racial/ethnic composition and community poverty level for prepared salad availability (not shown; non-white, moderate poverty [PR¼0.71, 95% CI¼0.53, 0.96, p¼0.024]; non-white, high poverty [PR¼0.73, 95% CI¼0.53, 1.02, p¼0.064]). Figure 1 shows estimated predicted probabilities of prepared salad availability at supermarkets for communities with different racial/ethnic compositions and poverty levels. Among supermarkets, the predicted probability of offering prepared salad was 0.79 (95% CI¼0.73, 0.85) in majority-white, low-poverty
communities compared with 0.71 (95% CI¼0.62, 0.79) in majority-white, high-poverty communities (p¼0.099) and 0.65 (95% CI¼0.55, 0.75) in non-white, high-poverty communities (p¼0.033). Table 4 shows multivariable associations among community characteristics and availability of the less-healthy prepared food items, across all stores and stratified by store type. Overall, rural stores were 14% more likely to carry at least one less-healthy prepared food item than suburban stores (PR¼1.14, 95% CI¼1.001, 1.30). Although overall, there were no statistically significant associations, when stratified by store type, grocery stores in high-poverty communities were less likely to offer at least one less-healthy prepared food item than those in low-poverty communities (PR¼0.45, 95% CI¼0.26, 0.78). The likelihood of stores carrying at least one less-healthy prepared food item was 24% lower in non-white communities than majority white communities (PR¼0.76, 95% CI¼0.66, 0.87). Stratified models revealed that this association was confined to www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562
557
Table 3. Multivariable Associations Between Community Characteristics and Availability of Prepared Vegetable Salads or Salad Bara All storesb (n¼4,361) PR (95% CI)
Supermarketsb,c (n¼620) PR (95% CI)
Grocery storesb,d (n¼620) PR (95% CI)
Convenience storesb,e (n¼3,121) PR (95% CI)
0.93 (0.75, 1.16)
1.03 (0.90, 1.17)
1.20 (0.75, 1.92)
0.87 (0.62, 1.22)
0.83 (0.72–0.97)
1.04 (0.56–1.92)
0.77 (0.59 - 1.01)
0.82 (0.62–1.06)
0.92 (0.79–1.08)
0.74 (0.47–1.17)
0.82 (0.55–1.24)
0.98 (0.81–1.18)
0.96 (0.85–1.07)
0.84 (0.52–1.37)
1.01 (0.76–1.34)
0.71** (0.59–0.86)
0.90 (0.78–1.02)
0.88 (0.51–1.51)
0.64**f (0.47–0.87)
Variables Urbanicity Ref: Suburban Urban Rural
**
0.74
(0.62–0.88)
*
Racial/ethnic composition Ref: Majority white Non-white Poverty level Ref: Low Moderate High Store type Ref: Supermarket Grocery store
0.27*** (0.22–0.34)
Convenience store
0.16*** (0.14–0.18)
Note: Boldface indicates statistical significance *po0.05; **po0.01; ***po0.001. a Each data column provides results from a single generalized linear model (GLM) with Poisson distribution. Thus, all independent variables are mutually adjusted for one another. Although the outcome variable is binary, prevalence ratios using GLM estimation with Poisson distribution, and not binomial distribution, were used because of convergence issues experienced in the estimation of GLM with binomial distribution. As a result, SEs may be larger.53 b Each of the four models shown in the table included the following covariates: census division entered as eight dummy variables (Northeast, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, Pacific) with Middle Atlantic as the reference category, a dummy for year (2011) with 2012 as the reference category, and a dummy variable for presence of at least one unhealthy prepared food item. c Covariates also included one dummy variable for number of cash registers (8þ), with 4–7 registers as the reference category. d Covariates also included three dummy variables for number of cash registers (2 or 3, 4–7, 8þ), with 1 register as the reference category. e Covariates also included two dummy variables for number of cash registers (2 or 3, 4–7), with 1 register as the reference category. f Generalized Hausman specification test comparing convenience stores and supermarket: 4.06, p¼0.04. PR, prevalence ratio.
convenience stores (PR¼0.72, 95% CI¼0.63, 0.82). There were no significant interactions between community racial/ethnic composition and poverty level (not shown).
Discussion In the first nationwide examination of prepared, readyto-eat food availability in food stores, this study found that well over half of stores carried at least one of the prepared foods we assessed, with 36% offering at least one unhealthy prepared food item and 20% offering prepared salads. Controlling for racial/ethnic composition, poverty level, and other factors, rural food stores were less likely to carry prepared salads and more likely to carry unhealthy prepared food items as compared with stores in suburban communities. In addition, October 2015
convenience stores in high-poverty communities were less likely to offer prepared salads. Among supermarkets, the likelihood of prepared salad availability was higher in majority-white, low-poverty communities than nonwhite, high-poverty communities. Results of this study suggest that there is substantial room to increase the healthfulness of prepared food offerings within stores. Interventions to increase availability of fresh fruits and vegetables and healthier packaged food options in stores have had some success in improving psychosocial factors (e.g., intentions to purchase/consume) and dietary behaviors.54 School and workplace studies have shown that the addition of salad bars improves dietary behaviors.55,56 However, further research is needed to better understand the extent to which interventions that attempt to change the supply side of the prepared food
558
Zenk et al / Am J Prev Med 2015;49(4):553–562
Figure 1. Predicted probabilities of prepared salad availability at supermarkets (n¼620) by community racial/ethnic composition and poverty level.a,b a
Difference between majority-white, low-poverty community and majority-white, high-poverty community is marginally statistically significant (p¼0.099). b Difference between majority-white, low-poverty community and nonwhite, high-poverty community is statistically significant (p¼0.033).
market may need to be accompanied by demand-side interventions to change community residents’ food preferences through nutrition education, alteration of social norms, or other approaches. As pointed out by Sharkey and colleagues,11,34 studies focused solely on traditional fast food restaurants likely underestimate communities’ exposure to prepared, ready-to-eat foods or fast food, particularly in rural areas. A conservative 36.4% of food stores in this nationwide sample offered at least one less-healthy prepared food item, and stores in rural areas—including supermarkets—were more likely to carry at least one lesshealthy prepared food item than suburban stores. Thus, efforts to reduce exposure to unhealthy prepared foods, such as fast food restaurant moratoriums and fast food industry reformulation of offerings, may have limited success if prepared foods within stores are ignored.57,58 The inclusion of prepared food items intended to feed one person from several types of chain food stores in the U.S. Food and Drug Administration final menu labeling requirements is a positive development for increasing consumer awareness of the caloric contribution and nutritional quality of prepared foods (http://www.fda. gov/Food/IngredientsPackagingLabeling/LabelingNutri tion/ucm248732.htm).59,60 This study provides additional evidence of environmental barriers faced by rural residents in making healthy food choices. Rural populations are at increased risk for obesity and adverse nutrition-related health outcomes.61–65 They live farther from and have fewer available supermarkets and other healthy food sources
(e.g., stores selling fruits and vegetables) than urban residents.26,66,67 The relatively few stores located in rural areas tend to predominately sell energy-dense, nutrientpoor foods and have fewer and more-expensive healthier food options.68,69 This study suggests that rural stores provide a particularly unhealthy mix of prepared foods, with more stores carrying unhealthy prepared foods and fewer stores carrying prepared salads as a healthy alternative. As such, improving prepared food selections in rural stores should be among the prioritized interventions and policies for addressing rural food access.70 A unique contribution of our study is the ability to examine prepared food availability at a relatively large sample of supermarkets. The fact that supermarkets located in rural communities (versus suburban communities) and in non-white, high-poverty communities (versus majority-white, low-poverty communities) were less likely to offer prepared salads provides additional evidence of disparities in healthy food availability within supermarkets.33,71 Thus, although introducing supermarkets into underserved communities through private and public partnerships and policies such as the Healthy Food Financing Initiative is an important step in improving access to healthy foods, these efforts should be accompanied by attention to the foods that are for sale within these new stores. Although they were equally as likely to offer lesshealthy prepared food items, convenience stores in highpoverty communities were 36% less likely to carry prepared salads than those in low-poverty communities. On the other hand, grocery stores in high-poverty (versus low-poverty) communities and convenience stores in non-white (versus majority-white) communities were less likely to offer less-healthy prepared food items. It is possible that small stores in these communities are less likely to have infrastructure for prepared foods.54,72,73
Limitations Strengths of this study are its focus on healthier and lesshealthy prepared food availability, inclusion of communities nationwide, and ability to make comparisons across a continuum of community characteristics. Nonetheless, the study has limitations. First, prepared food availability was measured based on a small number of items and only prepared foods sold by the store itself, not through separate businesses located within a store (e.g., separate counter for a chain fast food restaurant), were included. Therefore, prevalence estimates may underestimate prepared food availability. It is also possible that findings on community differences would change with the inclusion of additional less-healthy prepared foods (e.g., fried chicken, egg rolls) as well as other healthier items beyond www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562
559
Table 4. Multivariable Associations Between Community Characteristics and Availability of Less-Healthy Prepared Food Itemsa All storesb (n¼4,361) PR (95% CI)
Supermarketsb,c (n¼620) PR (95% CI)
Grocery storesb,d (n¼620) PR (95% CI)
Convenience storesb,e (n¼3,121) PR (95% CI)
0.94 (0.81, 1.09)
0.86 (0.58, 1.27)
0.90 (0.54, 1.48)
0.99 (0.87, 1.13)
1.14 (1.00–1.30)
1.38 (0.93–2.04)
1.08 (0.70–1.67)
1.14 (1.00–1.30)
0.76*** (0.66–0.87)
1.04 (0.71–1.53)
1.19 (0.72–1.97)
0.72***f (0.63–0.82)
Moderate
0.89 (0.78–1.01)
0.78 (0.54–1.14)
0.68 (0.42–1.08)
0.97 (0.85–1.10)
High
0.96 (0.66–0.87)
1.15 (0.76–1.72)
0.45**g (0.26–0.78)
1.02 (0.89–1.17)
Variables Urbanicity Ref: Suburban Urban Rural
*
Racial/ethnic composition Ref: Majority white Non-white Poverty level Ref: Low
Store type Ref: Supermarket Grocery store
1.14 (0.91–1.43)
Convenience store
2.73*** (2.30–3.25)
Note: Boldface indicates statistical significance *po0.05; **po0.01; ***po0.001. a Each data column provides results from a single generalized linear model (GLM) with Poisson distribution. Thus, all independent variables are mutually adjusted for one another. Although the outcome variable is binary, prevalence ratios using GLM estimation with Poisson distribution, and not binomial distribution, was used due to convergence issues experienced in the estimation of GLM with binomial distribution. As a result, SEs may be larger.53 b Each of the four models shown in the table included the following covariates: census division entered as eight dummy variables (Northeast, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, Pacific) with Middle Atlantic as the reference category, a dummy for year (2011) with 2012 as the reference category, and a dummy variable for presence of prepared salads. c Covariates also included one dummy variable for number of cash registers (8þ), with 4–7 registers as the reference category. d Covariates also included three dummy variables for number of cash registers (2 or 3, 4–7, 8þ), with 1 register as the reference category. e Covariates also included two dummy variables for number of cash registers (2 or 3, 4–7), with 1 register as the reference category. f Generalized Hausman specification test comparing convenience stores to grocery stores: 4.24, p¼0.04. g Generalized Hausman specification test comparing grocery stores to supermarkets: 9.46, p¼0.002. Generalized Hausman test comparing grocery stores to convenience stores: 8.49, p¼0.004. PR, prevalence ratio.
prepared salads (e.g., rotisserie chicken, cut fresh fruit), which may be more common in some types of communities because of local food preferences or other factors. Second, assumptions were made about which prepared foods are healthier and less healthy. Prepared salads can be less healthy when they include additional fat and sodium from cheeses, processed meats, or full-fat dressings, for example, though salads premixed with dressings were excluded. Similarly, tacos and pizza can be healthy choices, depending on their ingredients and preparation. Third, prepared food prices were not measured. Prepared salads may be relatively expensive compared to lesshealthy foods, and might not be an economically feasible choice for some people. Fourth, the findings are October 2015
generalizable to communities where eighth-, tenth-, and 12th-grade public school students reside and may not apply to all U.S. communities. Fifth, because this is a cross-sectional study, causal relationships between community characteristics and prepared food availability cannot be established. It is possible that findings reflect differences in demand (e.g., prepared salads are offered in fewer stores and are less available in rural communities because there is less demand).
Conclusions Despite these limitations, based on the foods assessed in this study, less-healthy prepared foods are commonly available in stores, and customers’ ability to opt for
560
Zenk et al / Am J Prev Med 2015;49(4):553–562
prepared salads is limited, especially in rural stores and convenience stores in high-poverty communities. Continued growth in and demand for prepared, ready-to-eat foods within stores highlight the need for public health professionals to work with storeowners and managers to improve the healthfulness of prepared food offerings in addition to packaged foods and fresh produce, the focus of most research and interventions to date. Data collection and analysis were supported by the Robert Wood Johnson Foundation through grants (Grant ID 64702 and 70157) to the Bridging the Gap program at the University of Illinois at Chicago. The Foundation had no role in study design; data collection, analysis, or interpretation; writing the report; or the decision to submit for publication. We thank Christopher Quinn for assistance with data management. No financial disclosures were reported by the authors of this paper.
References 1. Powell LM, Nguyen BT, Han E. Energy intake from restaurants: demographics and socioeconomics, 2003–2008. Am J Prev Med. 2012;43(5):498–504. http://dx.doi.org/10.1016/j.amepre.2012.07.041. 2. Fryer CD, Ervin RB. Caloric intake from fast food among adults: United States, 2007–2010. NCHS Data Brief. 2013;114:1–8. 3. Duffey KJ, Gordon-Larsen P, Jacobs Jr DR, Williams OD, Popkin BM. Differential associations of fast food and restaurant food consumption with 3-y change in body mass index: the Coronary Artery Risk Development in Young Adults study. Am J Clin Nutr. 2007;85(1): 201–208. 4. Pereira MA, Kartashov AI, Ebbeling CB, et al. Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis. Lancet. 2005;365(9453):36–42. http://dx.doi.org/10.1016/ S0140-6736(04)17663-0. 5. Rosenheck R. Fast food consumption and increased caloric intake: a systematic review of a trajectory towards weight gain and obesity risk. Obes Rev. 2008;9(6):535–547. http://dx.doi.org/10.1111/j.1467-789X. 2008.00477.x. 6. Boone-Heinonen J, Gordon-Larsen P, Kiefe CI, Shikany JM, Lewis CE, Popkin BM. Fast food restaurants and food stores: longitudinal associations with diet in young to middle-aged adults: the CARDIA study. Arch Intern Med. 2011;171(13):1162–1170. http://dx.doi.org/10. 1001/archinternmed.2011.283. 7. Nguyen BT, Powell LM. The impact of restaurant consumption among U.S. adults: effects on energy and nutrient intakes. Public Health Nutr. 2014;17:2445–2452. 8. Bowman SA, Vinyard BT. Fast food consumption of US adults: impact on energy and nutrient intakes and overweight status. J Am Coll Nutr. 2004;23(2):163–168. http://dx.doi.org/10.1080/07315724.2004. 10719357. 9. Powell LM, Nguyen BT. Fast-food and full-service restaurant consumption among children and adolescents: effect on energy, beverage, and nutrient intake. JAMA Pediatr. 2013;167(1):14–20. http://dx.doi.org/10. 1001/jamapediatrics.2013.417. 10. U.S. Department of Agriculture, U.S. DHHS, eds. Dietary Guidelines for Americans 2010. 7th ed. Washington, DC: U.S. Government Printing Office; 2010.
11. Sharkey JR, Johnson CM, Dean WR, Horel SA. Focusing on fast food restaurants alone underestimates the relationship between neighborhood deprivation and exposure to fast food in a large rural area. Nutr J. 2011;10:10. http://dx.doi.org/10.1186/1475-2891-10-10. 12. Zenk SN, Powell LM. U.S. secondary schools and food outlets. Health Place. 2008;14(2):336–346. http://dx.doi.org/10.1016/j.healthplace. 2007.08.003. 13. Fleischhacker S, Evenson K, Rodriguez D, Ammerman A. A systematic review of fast food access studies. Obes Rev. 2011;12(5): e460–e471. http://dx.doi.org/10.1111/j.1467-789X.2010.00715.x. 14. Jeffery RW, Baxter JE, McGuire MT, Linde JA. Are fast food restaurants an environmental risk factor for obesity? Int J Behav Nutr Phys Act. 2006;3(2):1–6. 15. Hickson DA, Roux AVD, Smith AE, et al. Associations of fast food restaurant availability with dietary intake and weight among African Americans in the Jackson heart study, 2000–2004. Am J Public Health. 2011;101(suppl 1):S301–S309. http://dx.doi.org/10.2105/AJPH.2010. 300006. 16. Moore LV, Roux D, Ana V, Nettleton JA, Jacobs DR, Franco M. Fastfood consumption, diet quality, and neighborhood exposure to fast food: the Multi-ethnic Study of Atherosclerosis. Am J Epidemiol. 2009;170(1):29–36. http://dx.doi.org/10.1093/aje/kwp090. 17. Reitzel LR, Regan SD, Nguyen N, et al. Density and proximity of fast food restaurants and body mass index among African Americans. Am J Public Health. 2014;104(1):110–116. http://dx.doi.org/10.2105/AJPH. 2012.301140. 18. Khan T, Powell LM, Wada R. Fast food consumption and food prices: evidence from panel data on 5th and 8th grade children. J Obes. 2012;2012:857697. http://dx.doi.org/10.1155/2012/857697. 19. Forsyth A, Wall M, Larson N, Story M, Neumark-Sztainer D. Do adolescents who live or go to school near fast-food restaurants eat more frequently from fast-food restaurants? Health Place. 2012;18(6): 1261–1269. http://dx.doi.org/10.1016/j.healthplace.2012.09.005. 20. He M, Tucker P, Gilliland J, Irwin JD, Larsen K, Hess P. The influence of local food environments on adolescents’ food purchasing behaviors. Int J Environ Res Public Health. 2012;9(4): 1458–1471. http://dx.doi.org/10.3390/ijerph9041458. 21. Powell LM, Chaloupka FJ, Bao Y. The availability of fast-food and fullservice restaurants in the United States: associations with neighborhood characteristics. Am J Prev Med. 2007;33(4 suppl): S240–S245. http://dx.doi.org/10.1016/j.amepre.2007.07.005. 22. Block JP, Scribner RA, DeSalvo KB. Fast food, race/ethnicity, and income: a geographic analysis. Am J Prev Med. 2004;27(3):211–217. 23. Kwate NO, Yau CY, Loh JM, Williams D. Inequality in obesigenic environments: fast food density in New York City. Health Place. 2009; 15(1):364–373. http://dx.doi.org/10.1016/j.healthplace.2008.07.003. 24. Morland K, Wing S, Diez Roux A, Poole C. Neighborhood characteristics associated with the location of food stores and food service places. Am J Prev Med. 2002;22(1):23–29. http://dx.doi.org/10.1016/S07493797(01)00403-2. 25. James P, Arcaya MC, Parker DM, Tucker-Seeley RD, Subramanian SV. Do minority and poor neighborhoods have higher access to fast-food restaurants in the United States? Health Place. 2014;29: 10–17. http://dx.doi.org/10.1016/j.healthplace.2014.04.011. 26. Powell LM, Slater S, Mirtcheva D, Bao Y, Chaloupka FJ. Food store availability and neighborhood characteristics in the United States. Prev Med. 2007;44(3):189–195. http://dx.doi.org/10.1016/j.ypmed.2006. 08.008. 27. Fleischhacker SE, Evenson KR, Rodriguez DA, Ammerman AS. A systematic review of fast food access studies. Obes Rev. 2011;12(5): e460–e471. http://dx.doi.org/10.1111/j.1467-789X.2010.00715.x. 28. Horowitz CR, Colson KA, Hebert PL, Lancaster K. Barriers to buying healthy foods for people with diabetes: evidence of environmental disparities. Am J Public Health. 2004;94(9):1549–1554. http://dx.doi.org/ 10.2105/AJPH.94.9.1549.
www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562 29. Baker EA, Schootman M, Barnidge E, Kelly C. The role of race and poverty in access to foods that enable individuals to adhere to dietary guidelines. Prev Chronic Dis. 2006;3(3):A76. 30. Morland K, Filomena S. Disparities in the availability of fruits and vegetables between racially segregated urban neighbourhoods. Public Health Nutr. 2007;10(12):1481–1489. http://dx.doi.org/10.1017/ S1368980007000079. 31. Franco M, ez-Roux AV, Nettleton JA, et al. Availability of healthy foods and dietary patterns: the Multi-ethnic Study of Atherosclerosis. Am J Clin Nutr. 2009;89(3):897–904. http://dx.doi.org/10.3945/ajcn.2008.26434. 32. Gustafson A, Hankins S, Jilcott S. Measures of the consumer food store environment: a systematic review of the evidence 2000-2011. J Community Health. 2012;37(4):897–911. http://dx.doi.org/10.1007/ s10900-011-9524-x. 33. Zenk SN, Powell LM, Rimkus L, et al. Relative and absolute availability of healthier food and beverage alternatives across communities in the United States. Am J Public Health. 2014;104(11): 2170–2178. http://dx.doi.org/10.2105/AJPH.2014.302113. 34. Creel JS, Sharkey JR, McIntosh A, Anding J, Huber JC. Availability of healthier options in traditional and nontraditional rural fast-food outlets. BMC Public Health. 2008;8(1):395. http://dx.doi.org/ 10.1186/1471-2458-8-395. 35. Lee SH, Rowan MT, Powell LM, et al. Characteristics of prepared food sources in low-income neighborhoods of Baltimore City. Ecol Food Nutr. 2010;49(6):409–430. http://dx.doi.org/10.1080/03670244.2010.524102. 36. Neckerman KM, Lovasi L, Yousefzadeh P, et al. Comparing nutrition environments in bodegas and fast-food restaurants. J Acad Nutr Diet. 2014;114(4):595–602. http://dx.doi.org/10.1016/j.jand.2013.07.007. 37. Bachman JG, Johnston LD, O’Malley PM, Schulenberg JE, eds. The Monitoring the Future Project After Thirty-Seven Years: Design and Procedures (Monitoring the Future Occasional Paper No. 76). Ann Arbor, MI: Institute for Social Research, 2011. 38. Powell LM, Han E, Zenk SN, et al. Field validation of secondary commercial data sources on the retail food environment in the U.S. Health Place. 2011;17(5):1122–1131. http://dx.doi.org/10.1016/ j.healthplace.2011.05.010. 39. Liese AD, Colabianchi N, Lamichhane AP, et al. Validation of 3 food outlet databases: completeness and geospatial accuracy in rural and urban food environments. Am J Epidemiol. 2010;172(11): 1324–1333. http://dx.doi.org/10.1093/aje/kwq292. 40. Fleischhacker SE, Evenson KR, Sharkey J, Pitts SBJ, Rodriguez DA. Validity of secondary retail food outlet data: a systematic review. Am J Prev Med. 2013;45(4):462–473. http://dx.doi.org/10.1016/j.amepre. 2013.06.009. 41. Iannacchione VG, Staab JM, Redden DT. Evaluating the use of residential mailing addresses in a metropolitan household survey. Public Opin Q. 2003;67(2):202–210. http://dx.doi.org/10.1086/374398. 42. Lavrakas PJ. Encyclopedia of Survey Research Methods. Los Angeles, CA: Sage; 2008. http://dx.doi.org/10.4135/9781412963947. 43. Barker D, Chaloupka F, Chriqui J, et al., eds. Bridging the Gap Community Obesity Project: Methodology. Bridging the Gap Research Paper Series, Paper #1. Chicago: Bridging the Gap, 2014. 44. Rimkus L, Powell LM, Zenk SN, et al. Development and reliability testing of a food store observation form. J Nutr Educ Behav. 2013;45(6): 540–548. http://dx.doi.org/10.1016/j.jneb.2013.02.005. 45. Bachman JL, Reedy J, Subar AF, Krebs-Smith SM. Sources of food group intakes among the U.S. population, 2001-2002. J Am Diet Assoc. 2008;108(5):804–814. http://dx.doi.org/10.1016/j.jada.2008.02.026. 46. Drewnowski A, Rehm CD. Energy intakes of U.S. children and adults by food purchase location and by specific food source. Nutr J. 2013;12: 59. http://dx.doi.org/10.1186/1475-2891-12-59. 47. Reedy J, Krebs-Smith SM. Dietary sources of energy, solid fats, and added sugars among children and adolescents in the United States. J Am Diet Assoc. 2010;110(10):1477–1484. http://dx.doi.org/10.1016/j. jada.2010.07.010.
October 2015
561
48. Connell CL, Yadrick MK, Simpson P, Gossett J, McGee BB, Bogle ML. Food supply adequacy in the lower Mississippi Delta. J Nutr Educ Behav. 2007;39(2):77–83. http://dx.doi.org/10.1016/j.jneb.2006.10.007. 49. Farley TA, Rice J, Bodor JN, Cohen DA, Bluthenthal RN, Rose D. Measuring the food environment: Shelf space of fruits, vegetables, and snack foods in stores. J Urban Health. 2009;86(5): 672–682. http://dx.doi.org/10.1007/s11524-009-9390-3. 50. Galvez MP, Morland K, Raines C, et al. Race and food store availability in an inner-city neighbourhood. Public Health Nutr. 2008;11(06): 624–631. http://dx.doi.org/10.1017/S1368980007001097. 51. American Community Survey Office. The 2006-2010 ACS 5-year summary file technical documentation. 2011. www2.census.gov/ acs2010_5yr/summaryfile/ACS_2006-2010_SF_Tech_Doc.pdf. 52. National Center for Education Statistics. NCES locale codes. www.nces. ed.gov/programs/handbook/data/pdf/appendix_d.pdf. 53. Cummings P. Methods for estimating adjusted risk ratios. Stata J. 2009;9(2):175. 54. Gittelsohn J, Rowan M, Gadhoke P. Interventions in small food stores to change the food environment, improve diet, and reduce risk of chronic disease. Prev Chronic Dis. 2012;9:E59. 55. Jeffery RW, French SA, Raether C, Baxter JE. An environmental intervention to increase fruit and salad purchases in a cafeteria. Prev Med. 1994;23(6):788–792. http://dx.doi.org/10.1006/pmed.1994.1135. 56. Slusser WM, Cumberland WG, Browdy BL, Lange L, Neumann C. A school salad bar increases frequency of fruit and vegetable consumption among children living in low-income households. Public Health Nutr. 2007;10(12):1490–1496. http://dx.doi.org/10.1017/S1368980007000444. 57. Sturm R, Cohen DA. Zoning for health? The year-old ban on new fastfood restaurants in south LA. Health Aff. 2009;28(6): w1088–w1097. http://dx.doi.org/10.1377/hlthaff.28.6.w1088. 58. Kirkpatrick SI, Reedy J, Kahle LL, Harris JL, Ohri-Vachaspati P, KrebsSmith SM. Fast-food menu offerings vary in dietary quality, but are consistently poor. Public Health Nutr. 2014;17(4): 924–931. http://dx.doi.org/10.1017/S1368980012005563. 59. Elbel B, Kersh R, Brescoll VL, Dixon LB. Calorie labeling and food choices: a first look at the effects on low-income people in New York City. Health Aff. 2009;28(6):w1110–w1121. http://dx.doi.org/10.1377/hlthaff.28.6.w1110. 60. Waterlander WE, Zenk SN. Food labelling, food retail availability and food pricing—moving from research to action? Public Health Nutr. 2015;18(1):2–7. http://dx.doi.org/10.1017/S1368980014002638. 61. Befort CA, Nazir N, Perri MG. Prevalence of obesity among adults from rural and urban areas of the United States: findings from NHANES (2005‐2008). J Rural Health. 2012;28(4): 392–397. http://dx.doi.org/10.1111/j.1748-0361.2012.00411.x. 62. Davis AM, Bennett KJ, Befort C, Nollen N. Obesity and related health behaviors among urban and rural children in the United States: data from the National Health and Nutrition Examination Survey 2003-2004 and 2005-2006. J Pediatr Psychol. 2011;36(6): 669–676. http://dx.doi.org/10.1093/jpepsy/jsq117. 63. James WL. All rural places are not created equal: revisiting the rural mortality penalty in the United States. Am J Public Health. 2014;104 (11):2122–2129. http://dx.doi.org/10.2105/AJPH.2014.301989. 64. Singh GK, Siahpush M. Widening rural–urban disparities in life expectancy, U.S., 1969–2009. Am J Prev Med. 2014;46(2): e19–e29. http://dx.doi.org/10.1016/j.amepre.2013.10.017. 65. Kulshreshtha A, Goyal A, Dabhadkar K, Veledar E, Vaccarino V. Urban-rural differences in coronary heart disease mortality in the United States: 1999-2009. Public Health Rep. 2014;129(1):19–29. 66. Sharkey JR, Horel S. Neighborhood socioeconomic deprivation and minority composition are associated with better potential spatial access to the groundtruthed food environment in a large rural area. J Nutr. 2008;138(3):620–627. 67. Gantner LA, Olson CM, Frongillo EA, Wells NM. Prevalence of nontraditional food stores and distance to healthy foods in a rural food environment. J Hunger Environ Nutr. 2011;6(3): 279–293. http://dx.doi.org/10.1080/19320248.2011.597829.
562
Zenk et al / Am J Prev Med 2015;49(4):553–562
68. Liese AD, Weis KE, Pluto D, Smith E, Lawson A. Food store types, availability, and cost of foods in a rural environment. J Am Diet Assoc. 2007;107(11):1916–1923. http://dx.doi.org/10.1016/j.jada.2007. 08.012. 69. Fleischhacker S, Johnson D, Quinn E, Pitts SBJ, Byker C, Sharkey JR. Advancing rural food access policy research priorities: process and potential of a transdisciplinary working group. J Agric Food Syst Community Dev. 2013;3(4):201–212. 70. Johnson DB, Quinn E, Sitaker M, et al. Developing an agenda for research about policies to improve access to healthy foods in rural communities: a concept mapping study. BMC Public Health. 2014;14: 592. http://dx.doi.org/10.1186/1471-2458-14-592.
71. Franco M, Diez Roux AV, Glass TA, Caballero B, Brancati FL. Neighborhood characteristics and availability of healthy foods in Baltimore. Am J Prev Med. 2008;35(6):561–567. http://dx.doi.org/10. 1016/j.amepre.2008.07.003. 72. Dannefer R, Williams DA, Baronberg S, Silver L. Healthy bodegas: increasing and promoting healthy foods at corner stores in New York City. Am J Public Health. 2012;102(10):e27–e31. http://dx.doi.org/ 10.2105/AJPH.2011.300615. 73. Cavanaugh E, Green S, Mallya G, Tierney A, Brensinger C, Glanz K. Changes in food and beverage environments after an urban corner store intervention. Prev Med. 2014;65:7–12. http://dx.doi.org/ 10.1016/j.ypmed.2014.04.009.
www.ajpmonline.org
Results: Overall, 63.6% of stores sold prepared foods, with 20.0% offering prepared salads and 36.4% offering at least one less-healthy item. Rural stores were 26% less likely to carry prepared salads (prevalence ratio [PR]¼0.74, 95% CI¼0.62, 0.88) and 14% more likely to carry at least one less-healthy prepared food item (PR¼1.14, 95% CI¼1.00, 1.30). Convenience stores in high-poverty communities were less likely to carry prepared salads than those in low-poverty communities (PR¼0.64, 95% CI¼0.47, 0.87). Among supermarkets, prepared salads were more likely to be carried in majority-white, low-poverty communities than in non-white, high-poverty communities.
Conclusions: Increasing the healthfulness of prepared foods within stores may offer an important opportunity to improve the food environment. (Am J Prev Med 2015;49(4):553–562) & 2015 American Journal of Preventive Medicine
Introduction
I
n 2007–2008, more than one third of American children, adolescents, and adults reported consuming fast food on a given day.1 Fast food comprised an average of 11.3% of American adults’ total daily calories in 2007–2010.2 Fast food intake is associated with higher intake of calories, fat, sodium, and sugar; lower intake of vegetables and micronutrients; and higher body weight From the College of Nursing (Zenk), School of Public Health (Powell), Institute for Health Research and Policy (Isgor, Rimkus), and Department of Economics (Chaloupka), University of Illinois at Chicago, Chicago, Illinois; and Barker Bi-Coastal Health Consultants Inc. (Barker), Calabasas, California Address correspondence to: Shannon N. Zenk, PhD, University of Illinois at Chicago, 845 S. Damen Ave., 9th Floor, Chicago IL 60612. E-mail: [email protected]. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2015.02.025
& 2015 American Journal of Preventive Medicine
and weight gain.1,3–9 Increasingly, sources of fast food, or food prepared and served quickly (www.merriam-web ster.com/dictionary/fast-food),10 extend well beyond traditional fast food restaurants (where customers order and pay at the counter and food is consumed either on premises or off premises as carryout11) to food stores. According to a 2010 report, almost two thirds of Americans indicated that they purchased prepared ready-to-eat/heat food from a grocery store or supermarket in the past month (www.packagedfacts.com/ Prepared-Foods-Ready-2694891/). In another survey, 82% reported that they purchased a prepared food or drink from a convenience store at least once a month (http://nrn.com/archive/c-store-growth-poses-threat-quickservice). Close location of sources of fast food and other prepared, ready-to-eat food to where individuals live, work, and play may contribute to their convenience and
Published by Elsevier Inc.
Am J Prev Med 2015;49(4):553–562
553
554
Zenk et al / Am J Prev Med 2015;49(4):553–562
utilization, and consequently poorer diet and weight outcomes. A 2005 study in the 20 largest U.S. cities found, for example, that 46% of neighborhoods had at least one fast food restaurant within walking distance.12 This likely underestimates the accessibility of fast food and other prepared food because it does not take into account availability in stores. A study of six rural Texas counties showed that the median distance to the nearest outlet at which fast food or other prepared food could be purchased decreased from 4.5 miles to 2.7 miles when all opportunities (traditional fast food restaurants and non-traditional outlets such as supermarkets, grocery stores, and convenience stores) were included.11 Although some research shows no association,13,14 other studies have found that living in neighborhoods with greater accessibility of fast food restaurants is associated with greater utilization, lower dietary quality, and higher body weight.6,13,15–20 In general, research suggests that low-income, racial/ ethnic minority, and some central-city neighborhoods in the U.S. may be disproportionately exposed to fast food restaurants as compared with more-advantaged neighborhoods.12,21–27 These communities may also have fewer healthy food options at food stores, even in supermarkets.28–33 However, limited empirical evidence exists on the healthfulness of prepared food offerings at food stores and how it may differ by community characteristics.34–36 As the first nationwide examination of prepared food availability in food stores, this study examines associations among community characteristics and the availability of both healthier and less-healthy prepared, ready-to-eat foods in U.S. supermarkets, grocery stores, and convenience stores.
Methods Study Design and Sample This study utilized 2 years of cross-sectional pooled data (2011– 2012) from the Bridging the Gap Community Obesity Measures Project (BTG-COMP), one of the only nationwide data sets of the directly observed food environment in the U.S. (www.bridgingth egapresearch.org/). As part of the BTG-COMP, aspects of the food environment were measured via direct observation in 317 communities spanning 42 states. The observed communities were school enrollment zones (or catchment areas from which the school draws its students) for nationally representative samples of eighth-, tenth-, and 12th-grade public school students in the continental U.S. from the Monitoring the Future study.37 A preliminary list of stores was generated annually by combining and de-duplicating separate business lists from InfoUSA and Dun & Bradstreet. After initially screening each store via telephone for eligibility (sold at a minimum snacks and drinks) and store type criteria (described in Measures section), a stratified probability sample of stores was selected in each community, with three store types (supermarkets, grocery stores, and “limited service” stores such as convenience stores, liquor stores, drug stores, dollar
stores) as the strata. Given limitations previously identified in the validity of commercial databases,38–40 a half-open interval procedure was implemented to help address error stemming from an incomplete list of businesses.41,42 This procedure involved observing stores sampled from the business list and stores discovered in the field.43 Field staff conducted data collection at 6,002 eligible food stores in 2011–2012 (completion rate, 495%). Approximately 2% of stores were missing data on one or more of the variables of interest for this analysis. Because few limited service stores other than convenience stores carry prepared foods (10.2% in this sample), other limited service stores were excluded, and results are based on the remaining 4,361 stores. This includes 620 supermarkets, 620 grocery stores, and 3,121 convenience stores. More-detailed descriptions of the sample design and data collection procedures are provided elsewhere.33,43
Measures Food and beverage availability at each store was measured using the Bridging the Gap Food Store Observation Form.44 This instrument assesses the availability of five prepared, ready-to-eat food items: cold sandwich (wrapped/ready-to-eat or made-toorder), prepared vegetable salad (excluding salads premixed with mayonnaise or dressing like coleslaw or potato salad) or salad bar, pizza, hot dog/hamburger, and taco/burrito/taquito. These items were based on a previously existing instrument and feedback on common prepared food items from field staff members who visited stores during pilot work in 2010.44 In addition, pizza, hot dogs, hamburgers, and Mexican mixed dishes are major contributors to energy and solid fat intake.45–47 Using these items, three dichotomous (none or at least one) outcome measures were derived: any prepared foods (at least one of the five items); availability of prepared vegetable salad or salad bar (hereafter referred to as “prepared salads”); and availability of less-healthy prepared food items (at least one of the following three items: pizza, hot dog/ hamburger, and taco/burrito/taquito). Because of wide variations in their ingredients, the cold sandwich item was included in the “any prepared foods” measure but was considered neither healthier nor less healthy. Food store type was operationalized using Food Marketing Institute descriptions (http://www.fmi.org/research-resources/ supermarket-facts) and prior studies.48–50 Supermarkets had fresh (unprocessed) meat, four or more cash registers, and at least two of the following: butcher, bakery, and deli. Grocery stores had fresh meat but did not meet the other supermarket criteria. Convenience stores had no fresh meat and sold a limited selection of staple groceries or other convenience items. American Community Survey (2007–2011) 5-year estimate data,51 aggregated based on census block groups intersecting the communities, were used to measure racial/ethnic and socioeconomic characteristics. Racial/ethnic composition was classified as majority white (Z50% residents non-Hispanic white) and nonwhite (not majority non-Hispanic white). Socioeconomic characteristics were measured by household poverty rate (tertiles). National Center for Education Statistics urban-centric locale codes were used to categorize community urbanicity as urban (small, midsize, and large cities); suburban (small, midsize, and large suburbs plus distant and fringe towns); and rural (distant, fringe, and remote rural areas plus remote towns).52 www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562 Several covariates helped to account for potential confounders, including census division/region (New England, Middle Atlantic, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, and Pacific); year; and store size as proxied by number of cash registers (one, two to three, four to seven, and eight or more).
555
Table 1. Weighted Descriptive Statistics for Communitya Characteristics All stores (n¼4,361) % Census division/region
Statistical Analysis
New England
4.1
Descriptive statistics for community characteristics were calculated for all stores. Percentages of stores that carried any of the assessed prepared foods, prepared salads, and less-healthy prepared food items were also calculated for all stores and by store type and community characteristics. Multivariable generalized linear models (GLM) with Poisson distribution were used to regress the two outcomes of primary interest—availability of prepared salads and availability of less-healthy prepared food items—on store type (in models based on all stores only); community racial/ethnic composition; poverty level; urbanicity; the other prepared food type; and covariates (region, year, and store size [stratified models only]). Each model was estimated for all stores and stratified by store type: supermarkets, grocery stores, and convenience stores. Generalized Hausman specification tests were used to test whether prevalence ratios (PRs) differed significantly across the stratified models. Multiplicative interaction terms between racial/ethnic composition and poverty level were added to each model in order to determine whether racial/ethnic composition–prepared food associations differed by poverty level. When interaction terms were statistically significant, estimated predicted probabilities for availability were calculated for contrasting communities, with all other covariates held at their original distributions. For example, the predicted probability of prepared salad availability was estimated for (1) communities set as being majority white and low-poverty and (2) communities set as being non-white and high-poverty, with all other covariates (e.g., region, urbanicity, year, store type) at their original distributions. All analyses were conducted using Stata, version 11.2, and completed in 2014. Sample weights were applied and estimation was undertaken (using svy commands) to account for the complex sample design for selection of communities and for stores clustered within communities, to obtain results for communities where nationally representative samples of eighth-, tenth-, and 12th-grade (traditional) public school students reside in the continental U.S. GLM with Poisson distribution was selected in order to estimate PRs for the relatively common outcomes.53
Middle Atlantic
10.6
West North Central
4.9
East North Central
10.4
South Atlantic
25.6
East South Central
6.0
West South Central
18.3
Mountain
4.7
Pacific
15.4
Results Table 1 shows characteristics of the store sample. Table 2 reports the prevalence of prepared, ready-to-eat foods and bivariate associations of the three prepared food measures with store type and community characteristics. Overall, 63.6% of stores sold prepared foods of the types we assessed (cold sandwich, salad, pizza, hot dog/hamburger, and taco/burrito/taquito). Thirty-six percent of stores offered at least one of the less-healthy items and considerably fewer stores (20.0%) offered prepared salads. October 2015
Racial/ethnic composition Majority white
69.9
Non-whiteb
30.1
Poverty levelc Low poverty
25.3
Moderate poverty
35.5
High poverty
39.3
d
Urbanicity Urban
37.0
Suburban
42.5
Rural
20.6
Store type Supermarket
11.3
Grocery store
11.0
Convenience store
77.8
a
Median area (minus water and military bases) of communities: 47.0 square miles (SD¼217.6; minimum¼0.1, maximum¼1833.0). b Not majority white (450% non-Hispanic black [5.4%], 450% Hispanic [7.3%], or Other [17.5%; not majority non-Hispanic white, non-Hispanic black, or Hispanic]). c The median household poverty rate was 14.0% in 2011 and 12.3% in 2012. In 2011, low poverty ranged from 1.8% to 8.7%, moderate poverty ranged from 8.8% to 16.2%, and high poverty ranged from 16.2% to 37.2%. In 2012, low poverty ranged from 1.9% to 7.9%, moderate poverty ranged from 8.1% to 12.7%, and high poverty ranged from 12.7% to 40.6%. d Urban was defined as small, midsize, and large cities; suburban was small, midsize, and large suburbs plus distant and fringe towns; and rural was distant, fringe, and remote rural areas plus remote towns, based on National Center for Education Statistics urban-centric locale codes.
Table 3 presents multivariable associations among community characteristics and availability of prepared salads, across all stores and stratified by store type.
Zenk et al / Am J Prev Med 2015;49(4):553–562
556
Table 2. Percentage of Stores That Sold Prepared Foods by Store Type and Community Characteristics Sold any prepared fooda % (95% CI)
Offered prepared vegetable salads or salad bar % (95% CI)
Offered any less-healthy prepared food itemsb % (95% CI)
63.6 (61.1, 66.1)
20.0 (18.0, 22.1)
36.4 (33.9, 39.0)f
Supermarkets
90.1 (87.0, 92.5)
74.5 (70.3, 78.2)
25.3 (21.3, 29.7)
Grocery stores
35.7 (31.2, 40.5)
16.2 (12.6, 20.6)
17.0 (13.6, 21.0)
63.7 (60.9, 66.5)
12.7 (10.9, 14.6)
40.8 (37.8, 43.9)
67.6 (64.9, 70.1)
22.6 (20.5, 24.9)
39.3 (36.3, 42.4)
54.5 (48.4, 60.4)
13.9 (10.8, 17.7)
29.7 (25.1, 34.7)
Low-poverty
68.4 (63.5, 72.9)
29.2 (26.3, 32.3)
39.6 (34.9, 44.5)
Moderate poverty
66.5 (63.0, 69.8)
20.8 (16.7, 25.5)
37.2 (32.6, 42.0)
57.9 (52.9, 62.8)
13.4 (11.2, 16.0)
33.7 (29.5, 38.1)
Suburban
60.6 (55.7, 65.4)
16.6 (13.6, 20.0)
31.8 (27.1, 37.0)
Urban
64.4 (60.9, 67.7)
24.7 (21.8, 27.8)
36.3 (33.3, 39.5)
Rural
67.5 (61.0, 73.3)
16.6 (13.6, 20.1)
44.9 (39.1, 50.8)
All stores (n¼4,361) Store typec,d,e
Convenience stores Community racial/ethnic composition Majority white Non-white Community poverty level
High-poverty Community urbanicity
c,d,e
c,d
d,e
At least one of the following five items: cold sandwich, vegetable-based salad/salad bar, pizza, hot dog/hamburger, Mexican (e.g., taco, burrito, taquito). b At least one of three items: pizza, hot dog/hamburger, Mexican (e.g., taco, burrito, taquito) c 2 χ test po0.05 for any prepared food. d 2 χ test po0.05 for prepared salads. e 2 χ test po0.05 for any less-healthy prepared food item. f 63.6% of stores carried none, 17.2% carried one item, 14.9% carried two items, and 4.3% carried three items. a
Overall, rural stores were 26% less likely to carry prepared salads than suburban stores (PR¼0.74, 95% CI¼0.62, 0.88). Stores in high-poverty communities were 29% less likely to carry prepared salads than those in lowpoverty communities (PR¼0.71, 95% CI¼0.59, 0.86). The prevalence of prepared salads in high-poverty communities was particularly low in convenience stores (PR¼0.64, 95% CI¼0.47, 0.87). Among supermarkets, there were statistically significant interactions between community racial/ethnic composition and community poverty level for prepared salad availability (not shown; non-white, moderate poverty [PR¼0.71, 95% CI¼0.53, 0.96, p¼0.024]; non-white, high poverty [PR¼0.73, 95% CI¼0.53, 1.02, p¼0.064]). Figure 1 shows estimated predicted probabilities of prepared salad availability at supermarkets for communities with different racial/ethnic compositions and poverty levels. Among supermarkets, the predicted probability of offering prepared salad was 0.79 (95% CI¼0.73, 0.85) in majority-white, low-poverty
communities compared with 0.71 (95% CI¼0.62, 0.79) in majority-white, high-poverty communities (p¼0.099) and 0.65 (95% CI¼0.55, 0.75) in non-white, high-poverty communities (p¼0.033). Table 4 shows multivariable associations among community characteristics and availability of the less-healthy prepared food items, across all stores and stratified by store type. Overall, rural stores were 14% more likely to carry at least one less-healthy prepared food item than suburban stores (PR¼1.14, 95% CI¼1.001, 1.30). Although overall, there were no statistically significant associations, when stratified by store type, grocery stores in high-poverty communities were less likely to offer at least one less-healthy prepared food item than those in low-poverty communities (PR¼0.45, 95% CI¼0.26, 0.78). The likelihood of stores carrying at least one less-healthy prepared food item was 24% lower in non-white communities than majority white communities (PR¼0.76, 95% CI¼0.66, 0.87). Stratified models revealed that this association was confined to www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562
557
Table 3. Multivariable Associations Between Community Characteristics and Availability of Prepared Vegetable Salads or Salad Bara All storesb (n¼4,361) PR (95% CI)
Supermarketsb,c (n¼620) PR (95% CI)
Grocery storesb,d (n¼620) PR (95% CI)
Convenience storesb,e (n¼3,121) PR (95% CI)
0.93 (0.75, 1.16)
1.03 (0.90, 1.17)
1.20 (0.75, 1.92)
0.87 (0.62, 1.22)
0.83 (0.72–0.97)
1.04 (0.56–1.92)
0.77 (0.59 - 1.01)
0.82 (0.62–1.06)
0.92 (0.79–1.08)
0.74 (0.47–1.17)
0.82 (0.55–1.24)
0.98 (0.81–1.18)
0.96 (0.85–1.07)
0.84 (0.52–1.37)
1.01 (0.76–1.34)
0.71** (0.59–0.86)
0.90 (0.78–1.02)
0.88 (0.51–1.51)
0.64**f (0.47–0.87)
Variables Urbanicity Ref: Suburban Urban Rural
**
0.74
(0.62–0.88)
*
Racial/ethnic composition Ref: Majority white Non-white Poverty level Ref: Low Moderate High Store type Ref: Supermarket Grocery store
0.27*** (0.22–0.34)
Convenience store
0.16*** (0.14–0.18)
Note: Boldface indicates statistical significance *po0.05; **po0.01; ***po0.001. a Each data column provides results from a single generalized linear model (GLM) with Poisson distribution. Thus, all independent variables are mutually adjusted for one another. Although the outcome variable is binary, prevalence ratios using GLM estimation with Poisson distribution, and not binomial distribution, were used because of convergence issues experienced in the estimation of GLM with binomial distribution. As a result, SEs may be larger.53 b Each of the four models shown in the table included the following covariates: census division entered as eight dummy variables (Northeast, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, Pacific) with Middle Atlantic as the reference category, a dummy for year (2011) with 2012 as the reference category, and a dummy variable for presence of at least one unhealthy prepared food item. c Covariates also included one dummy variable for number of cash registers (8þ), with 4–7 registers as the reference category. d Covariates also included three dummy variables for number of cash registers (2 or 3, 4–7, 8þ), with 1 register as the reference category. e Covariates also included two dummy variables for number of cash registers (2 or 3, 4–7), with 1 register as the reference category. f Generalized Hausman specification test comparing convenience stores and supermarket: 4.06, p¼0.04. PR, prevalence ratio.
convenience stores (PR¼0.72, 95% CI¼0.63, 0.82). There were no significant interactions between community racial/ethnic composition and poverty level (not shown).
Discussion In the first nationwide examination of prepared, readyto-eat food availability in food stores, this study found that well over half of stores carried at least one of the prepared foods we assessed, with 36% offering at least one unhealthy prepared food item and 20% offering prepared salads. Controlling for racial/ethnic composition, poverty level, and other factors, rural food stores were less likely to carry prepared salads and more likely to carry unhealthy prepared food items as compared with stores in suburban communities. In addition, October 2015
convenience stores in high-poverty communities were less likely to offer prepared salads. Among supermarkets, the likelihood of prepared salad availability was higher in majority-white, low-poverty communities than nonwhite, high-poverty communities. Results of this study suggest that there is substantial room to increase the healthfulness of prepared food offerings within stores. Interventions to increase availability of fresh fruits and vegetables and healthier packaged food options in stores have had some success in improving psychosocial factors (e.g., intentions to purchase/consume) and dietary behaviors.54 School and workplace studies have shown that the addition of salad bars improves dietary behaviors.55,56 However, further research is needed to better understand the extent to which interventions that attempt to change the supply side of the prepared food
558
Zenk et al / Am J Prev Med 2015;49(4):553–562
Figure 1. Predicted probabilities of prepared salad availability at supermarkets (n¼620) by community racial/ethnic composition and poverty level.a,b a
Difference between majority-white, low-poverty community and majority-white, high-poverty community is marginally statistically significant (p¼0.099). b Difference between majority-white, low-poverty community and nonwhite, high-poverty community is statistically significant (p¼0.033).
market may need to be accompanied by demand-side interventions to change community residents’ food preferences through nutrition education, alteration of social norms, or other approaches. As pointed out by Sharkey and colleagues,11,34 studies focused solely on traditional fast food restaurants likely underestimate communities’ exposure to prepared, ready-to-eat foods or fast food, particularly in rural areas. A conservative 36.4% of food stores in this nationwide sample offered at least one less-healthy prepared food item, and stores in rural areas—including supermarkets—were more likely to carry at least one lesshealthy prepared food item than suburban stores. Thus, efforts to reduce exposure to unhealthy prepared foods, such as fast food restaurant moratoriums and fast food industry reformulation of offerings, may have limited success if prepared foods within stores are ignored.57,58 The inclusion of prepared food items intended to feed one person from several types of chain food stores in the U.S. Food and Drug Administration final menu labeling requirements is a positive development for increasing consumer awareness of the caloric contribution and nutritional quality of prepared foods (http://www.fda. gov/Food/IngredientsPackagingLabeling/LabelingNutri tion/ucm248732.htm).59,60 This study provides additional evidence of environmental barriers faced by rural residents in making healthy food choices. Rural populations are at increased risk for obesity and adverse nutrition-related health outcomes.61–65 They live farther from and have fewer available supermarkets and other healthy food sources
(e.g., stores selling fruits and vegetables) than urban residents.26,66,67 The relatively few stores located in rural areas tend to predominately sell energy-dense, nutrientpoor foods and have fewer and more-expensive healthier food options.68,69 This study suggests that rural stores provide a particularly unhealthy mix of prepared foods, with more stores carrying unhealthy prepared foods and fewer stores carrying prepared salads as a healthy alternative. As such, improving prepared food selections in rural stores should be among the prioritized interventions and policies for addressing rural food access.70 A unique contribution of our study is the ability to examine prepared food availability at a relatively large sample of supermarkets. The fact that supermarkets located in rural communities (versus suburban communities) and in non-white, high-poverty communities (versus majority-white, low-poverty communities) were less likely to offer prepared salads provides additional evidence of disparities in healthy food availability within supermarkets.33,71 Thus, although introducing supermarkets into underserved communities through private and public partnerships and policies such as the Healthy Food Financing Initiative is an important step in improving access to healthy foods, these efforts should be accompanied by attention to the foods that are for sale within these new stores. Although they were equally as likely to offer lesshealthy prepared food items, convenience stores in highpoverty communities were 36% less likely to carry prepared salads than those in low-poverty communities. On the other hand, grocery stores in high-poverty (versus low-poverty) communities and convenience stores in non-white (versus majority-white) communities were less likely to offer less-healthy prepared food items. It is possible that small stores in these communities are less likely to have infrastructure for prepared foods.54,72,73
Limitations Strengths of this study are its focus on healthier and lesshealthy prepared food availability, inclusion of communities nationwide, and ability to make comparisons across a continuum of community characteristics. Nonetheless, the study has limitations. First, prepared food availability was measured based on a small number of items and only prepared foods sold by the store itself, not through separate businesses located within a store (e.g., separate counter for a chain fast food restaurant), were included. Therefore, prevalence estimates may underestimate prepared food availability. It is also possible that findings on community differences would change with the inclusion of additional less-healthy prepared foods (e.g., fried chicken, egg rolls) as well as other healthier items beyond www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562
559
Table 4. Multivariable Associations Between Community Characteristics and Availability of Less-Healthy Prepared Food Itemsa All storesb (n¼4,361) PR (95% CI)
Supermarketsb,c (n¼620) PR (95% CI)
Grocery storesb,d (n¼620) PR (95% CI)
Convenience storesb,e (n¼3,121) PR (95% CI)
0.94 (0.81, 1.09)
0.86 (0.58, 1.27)
0.90 (0.54, 1.48)
0.99 (0.87, 1.13)
1.14 (1.00–1.30)
1.38 (0.93–2.04)
1.08 (0.70–1.67)
1.14 (1.00–1.30)
0.76*** (0.66–0.87)
1.04 (0.71–1.53)
1.19 (0.72–1.97)
0.72***f (0.63–0.82)
Moderate
0.89 (0.78–1.01)
0.78 (0.54–1.14)
0.68 (0.42–1.08)
0.97 (0.85–1.10)
High
0.96 (0.66–0.87)
1.15 (0.76–1.72)
0.45**g (0.26–0.78)
1.02 (0.89–1.17)
Variables Urbanicity Ref: Suburban Urban Rural
*
Racial/ethnic composition Ref: Majority white Non-white Poverty level Ref: Low
Store type Ref: Supermarket Grocery store
1.14 (0.91–1.43)
Convenience store
2.73*** (2.30–3.25)
Note: Boldface indicates statistical significance *po0.05; **po0.01; ***po0.001. a Each data column provides results from a single generalized linear model (GLM) with Poisson distribution. Thus, all independent variables are mutually adjusted for one another. Although the outcome variable is binary, prevalence ratios using GLM estimation with Poisson distribution, and not binomial distribution, was used due to convergence issues experienced in the estimation of GLM with binomial distribution. As a result, SEs may be larger.53 b Each of the four models shown in the table included the following covariates: census division entered as eight dummy variables (Northeast, South Atlantic, East North Central, East South Central, West North Central, West South Central, Mountain, Pacific) with Middle Atlantic as the reference category, a dummy for year (2011) with 2012 as the reference category, and a dummy variable for presence of prepared salads. c Covariates also included one dummy variable for number of cash registers (8þ), with 4–7 registers as the reference category. d Covariates also included three dummy variables for number of cash registers (2 or 3, 4–7, 8þ), with 1 register as the reference category. e Covariates also included two dummy variables for number of cash registers (2 or 3, 4–7), with 1 register as the reference category. f Generalized Hausman specification test comparing convenience stores to grocery stores: 4.24, p¼0.04. g Generalized Hausman specification test comparing grocery stores to supermarkets: 9.46, p¼0.002. Generalized Hausman test comparing grocery stores to convenience stores: 8.49, p¼0.004. PR, prevalence ratio.
prepared salads (e.g., rotisserie chicken, cut fresh fruit), which may be more common in some types of communities because of local food preferences or other factors. Second, assumptions were made about which prepared foods are healthier and less healthy. Prepared salads can be less healthy when they include additional fat and sodium from cheeses, processed meats, or full-fat dressings, for example, though salads premixed with dressings were excluded. Similarly, tacos and pizza can be healthy choices, depending on their ingredients and preparation. Third, prepared food prices were not measured. Prepared salads may be relatively expensive compared to lesshealthy foods, and might not be an economically feasible choice for some people. Fourth, the findings are October 2015
generalizable to communities where eighth-, tenth-, and 12th-grade public school students reside and may not apply to all U.S. communities. Fifth, because this is a cross-sectional study, causal relationships between community characteristics and prepared food availability cannot be established. It is possible that findings reflect differences in demand (e.g., prepared salads are offered in fewer stores and are less available in rural communities because there is less demand).
Conclusions Despite these limitations, based on the foods assessed in this study, less-healthy prepared foods are commonly available in stores, and customers’ ability to opt for
560
Zenk et al / Am J Prev Med 2015;49(4):553–562
prepared salads is limited, especially in rural stores and convenience stores in high-poverty communities. Continued growth in and demand for prepared, ready-to-eat foods within stores highlight the need for public health professionals to work with storeowners and managers to improve the healthfulness of prepared food offerings in addition to packaged foods and fresh produce, the focus of most research and interventions to date. Data collection and analysis were supported by the Robert Wood Johnson Foundation through grants (Grant ID 64702 and 70157) to the Bridging the Gap program at the University of Illinois at Chicago. The Foundation had no role in study design; data collection, analysis, or interpretation; writing the report; or the decision to submit for publication. We thank Christopher Quinn for assistance with data management. No financial disclosures were reported by the authors of this paper.
References 1. Powell LM, Nguyen BT, Han E. Energy intake from restaurants: demographics and socioeconomics, 2003–2008. Am J Prev Med. 2012;43(5):498–504. http://dx.doi.org/10.1016/j.amepre.2012.07.041. 2. Fryer CD, Ervin RB. Caloric intake from fast food among adults: United States, 2007–2010. NCHS Data Brief. 2013;114:1–8. 3. Duffey KJ, Gordon-Larsen P, Jacobs Jr DR, Williams OD, Popkin BM. Differential associations of fast food and restaurant food consumption with 3-y change in body mass index: the Coronary Artery Risk Development in Young Adults study. Am J Clin Nutr. 2007;85(1): 201–208. 4. Pereira MA, Kartashov AI, Ebbeling CB, et al. Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis. Lancet. 2005;365(9453):36–42. http://dx.doi.org/10.1016/ S0140-6736(04)17663-0. 5. Rosenheck R. Fast food consumption and increased caloric intake: a systematic review of a trajectory towards weight gain and obesity risk. Obes Rev. 2008;9(6):535–547. http://dx.doi.org/10.1111/j.1467-789X. 2008.00477.x. 6. Boone-Heinonen J, Gordon-Larsen P, Kiefe CI, Shikany JM, Lewis CE, Popkin BM. Fast food restaurants and food stores: longitudinal associations with diet in young to middle-aged adults: the CARDIA study. Arch Intern Med. 2011;171(13):1162–1170. http://dx.doi.org/10. 1001/archinternmed.2011.283. 7. Nguyen BT, Powell LM. The impact of restaurant consumption among U.S. adults: effects on energy and nutrient intakes. Public Health Nutr. 2014;17:2445–2452. 8. Bowman SA, Vinyard BT. Fast food consumption of US adults: impact on energy and nutrient intakes and overweight status. J Am Coll Nutr. 2004;23(2):163–168. http://dx.doi.org/10.1080/07315724.2004. 10719357. 9. Powell LM, Nguyen BT. Fast-food and full-service restaurant consumption among children and adolescents: effect on energy, beverage, and nutrient intake. JAMA Pediatr. 2013;167(1):14–20. http://dx.doi.org/10. 1001/jamapediatrics.2013.417. 10. U.S. Department of Agriculture, U.S. DHHS, eds. Dietary Guidelines for Americans 2010. 7th ed. Washington, DC: U.S. Government Printing Office; 2010.
11. Sharkey JR, Johnson CM, Dean WR, Horel SA. Focusing on fast food restaurants alone underestimates the relationship between neighborhood deprivation and exposure to fast food in a large rural area. Nutr J. 2011;10:10. http://dx.doi.org/10.1186/1475-2891-10-10. 12. Zenk SN, Powell LM. U.S. secondary schools and food outlets. Health Place. 2008;14(2):336–346. http://dx.doi.org/10.1016/j.healthplace. 2007.08.003. 13. Fleischhacker S, Evenson K, Rodriguez D, Ammerman A. A systematic review of fast food access studies. Obes Rev. 2011;12(5): e460–e471. http://dx.doi.org/10.1111/j.1467-789X.2010.00715.x. 14. Jeffery RW, Baxter JE, McGuire MT, Linde JA. Are fast food restaurants an environmental risk factor for obesity? Int J Behav Nutr Phys Act. 2006;3(2):1–6. 15. Hickson DA, Roux AVD, Smith AE, et al. Associations of fast food restaurant availability with dietary intake and weight among African Americans in the Jackson heart study, 2000–2004. Am J Public Health. 2011;101(suppl 1):S301–S309. http://dx.doi.org/10.2105/AJPH.2010. 300006. 16. Moore LV, Roux D, Ana V, Nettleton JA, Jacobs DR, Franco M. Fastfood consumption, diet quality, and neighborhood exposure to fast food: the Multi-ethnic Study of Atherosclerosis. Am J Epidemiol. 2009;170(1):29–36. http://dx.doi.org/10.1093/aje/kwp090. 17. Reitzel LR, Regan SD, Nguyen N, et al. Density and proximity of fast food restaurants and body mass index among African Americans. Am J Public Health. 2014;104(1):110–116. http://dx.doi.org/10.2105/AJPH. 2012.301140. 18. Khan T, Powell LM, Wada R. Fast food consumption and food prices: evidence from panel data on 5th and 8th grade children. J Obes. 2012;2012:857697. http://dx.doi.org/10.1155/2012/857697. 19. Forsyth A, Wall M, Larson N, Story M, Neumark-Sztainer D. Do adolescents who live or go to school near fast-food restaurants eat more frequently from fast-food restaurants? Health Place. 2012;18(6): 1261–1269. http://dx.doi.org/10.1016/j.healthplace.2012.09.005. 20. He M, Tucker P, Gilliland J, Irwin JD, Larsen K, Hess P. The influence of local food environments on adolescents’ food purchasing behaviors. Int J Environ Res Public Health. 2012;9(4): 1458–1471. http://dx.doi.org/10.3390/ijerph9041458. 21. Powell LM, Chaloupka FJ, Bao Y. The availability of fast-food and fullservice restaurants in the United States: associations with neighborhood characteristics. Am J Prev Med. 2007;33(4 suppl): S240–S245. http://dx.doi.org/10.1016/j.amepre.2007.07.005. 22. Block JP, Scribner RA, DeSalvo KB. Fast food, race/ethnicity, and income: a geographic analysis. Am J Prev Med. 2004;27(3):211–217. 23. Kwate NO, Yau CY, Loh JM, Williams D. Inequality in obesigenic environments: fast food density in New York City. Health Place. 2009; 15(1):364–373. http://dx.doi.org/10.1016/j.healthplace.2008.07.003. 24. Morland K, Wing S, Diez Roux A, Poole C. Neighborhood characteristics associated with the location of food stores and food service places. Am J Prev Med. 2002;22(1):23–29. http://dx.doi.org/10.1016/S07493797(01)00403-2. 25. James P, Arcaya MC, Parker DM, Tucker-Seeley RD, Subramanian SV. Do minority and poor neighborhoods have higher access to fast-food restaurants in the United States? Health Place. 2014;29: 10–17. http://dx.doi.org/10.1016/j.healthplace.2014.04.011. 26. Powell LM, Slater S, Mirtcheva D, Bao Y, Chaloupka FJ. Food store availability and neighborhood characteristics in the United States. Prev Med. 2007;44(3):189–195. http://dx.doi.org/10.1016/j.ypmed.2006. 08.008. 27. Fleischhacker SE, Evenson KR, Rodriguez DA, Ammerman AS. A systematic review of fast food access studies. Obes Rev. 2011;12(5): e460–e471. http://dx.doi.org/10.1111/j.1467-789X.2010.00715.x. 28. Horowitz CR, Colson KA, Hebert PL, Lancaster K. Barriers to buying healthy foods for people with diabetes: evidence of environmental disparities. Am J Public Health. 2004;94(9):1549–1554. http://dx.doi.org/ 10.2105/AJPH.94.9.1549.
www.ajpmonline.org
Zenk et al / Am J Prev Med 2015;49(4):553–562 29. Baker EA, Schootman M, Barnidge E, Kelly C. The role of race and poverty in access to foods that enable individuals to adhere to dietary guidelines. Prev Chronic Dis. 2006;3(3):A76. 30. Morland K, Filomena S. Disparities in the availability of fruits and vegetables between racially segregated urban neighbourhoods. Public Health Nutr. 2007;10(12):1481–1489. http://dx.doi.org/10.1017/ S1368980007000079. 31. Franco M, ez-Roux AV, Nettleton JA, et al. Availability of healthy foods and dietary patterns: the Multi-ethnic Study of Atherosclerosis. Am J Clin Nutr. 2009;89(3):897–904. http://dx.doi.org/10.3945/ajcn.2008.26434. 32. Gustafson A, Hankins S, Jilcott S. Measures of the consumer food store environment: a systematic review of the evidence 2000-2011. J Community Health. 2012;37(4):897–911. http://dx.doi.org/10.1007/ s10900-011-9524-x. 33. Zenk SN, Powell LM, Rimkus L, et al. Relative and absolute availability of healthier food and beverage alternatives across communities in the United States. Am J Public Health. 2014;104(11): 2170–2178. http://dx.doi.org/10.2105/AJPH.2014.302113. 34. Creel JS, Sharkey JR, McIntosh A, Anding J, Huber JC. Availability of healthier options in traditional and nontraditional rural fast-food outlets. BMC Public Health. 2008;8(1):395. http://dx.doi.org/ 10.1186/1471-2458-8-395. 35. Lee SH, Rowan MT, Powell LM, et al. Characteristics of prepared food sources in low-income neighborhoods of Baltimore City. Ecol Food Nutr. 2010;49(6):409–430. http://dx.doi.org/10.1080/03670244.2010.524102. 36. Neckerman KM, Lovasi L, Yousefzadeh P, et al. Comparing nutrition environments in bodegas and fast-food restaurants. J Acad Nutr Diet. 2014;114(4):595–602. http://dx.doi.org/10.1016/j.jand.2013.07.007. 37. Bachman JG, Johnston LD, O’Malley PM, Schulenberg JE, eds. The Monitoring the Future Project After Thirty-Seven Years: Design and Procedures (Monitoring the Future Occasional Paper No. 76). Ann Arbor, MI: Institute for Social Research, 2011. 38. Powell LM, Han E, Zenk SN, et al. Field validation of secondary commercial data sources on the retail food environment in the U.S. Health Place. 2011;17(5):1122–1131. http://dx.doi.org/10.1016/ j.healthplace.2011.05.010. 39. Liese AD, Colabianchi N, Lamichhane AP, et al. Validation of 3 food outlet databases: completeness and geospatial accuracy in rural and urban food environments. Am J Epidemiol. 2010;172(11): 1324–1333. http://dx.doi.org/10.1093/aje/kwq292. 40. Fleischhacker SE, Evenson KR, Sharkey J, Pitts SBJ, Rodriguez DA. Validity of secondary retail food outlet data: a systematic review. Am J Prev Med. 2013;45(4):462–473. http://dx.doi.org/10.1016/j.amepre. 2013.06.009. 41. Iannacchione VG, Staab JM, Redden DT. Evaluating the use of residential mailing addresses in a metropolitan household survey. Public Opin Q. 2003;67(2):202–210. http://dx.doi.org/10.1086/374398. 42. Lavrakas PJ. Encyclopedia of Survey Research Methods. Los Angeles, CA: Sage; 2008. http://dx.doi.org/10.4135/9781412963947. 43. Barker D, Chaloupka F, Chriqui J, et al., eds. Bridging the Gap Community Obesity Project: Methodology. Bridging the Gap Research Paper Series, Paper #1. Chicago: Bridging the Gap, 2014. 44. Rimkus L, Powell LM, Zenk SN, et al. Development and reliability testing of a food store observation form. J Nutr Educ Behav. 2013;45(6): 540–548. http://dx.doi.org/10.1016/j.jneb.2013.02.005. 45. Bachman JL, Reedy J, Subar AF, Krebs-Smith SM. Sources of food group intakes among the U.S. population, 2001-2002. J Am Diet Assoc. 2008;108(5):804–814. http://dx.doi.org/10.1016/j.jada.2008.02.026. 46. Drewnowski A, Rehm CD. Energy intakes of U.S. children and adults by food purchase location and by specific food source. Nutr J. 2013;12: 59. http://dx.doi.org/10.1186/1475-2891-12-59. 47. Reedy J, Krebs-Smith SM. Dietary sources of energy, solid fats, and added sugars among children and adolescents in the United States. J Am Diet Assoc. 2010;110(10):1477–1484. http://dx.doi.org/10.1016/j. jada.2010.07.010.
October 2015
561
48. Connell CL, Yadrick MK, Simpson P, Gossett J, McGee BB, Bogle ML. Food supply adequacy in the lower Mississippi Delta. J Nutr Educ Behav. 2007;39(2):77–83. http://dx.doi.org/10.1016/j.jneb.2006.10.007. 49. Farley TA, Rice J, Bodor JN, Cohen DA, Bluthenthal RN, Rose D. Measuring the food environment: Shelf space of fruits, vegetables, and snack foods in stores. J Urban Health. 2009;86(5): 672–682. http://dx.doi.org/10.1007/s11524-009-9390-3. 50. Galvez MP, Morland K, Raines C, et al. Race and food store availability in an inner-city neighbourhood. Public Health Nutr. 2008;11(06): 624–631. http://dx.doi.org/10.1017/S1368980007001097. 51. American Community Survey Office. The 2006-2010 ACS 5-year summary file technical documentation. 2011. www2.census.gov/ acs2010_5yr/summaryfile/ACS_2006-2010_SF_Tech_Doc.pdf. 52. National Center for Education Statistics. NCES locale codes. www.nces. ed.gov/programs/handbook/data/pdf/appendix_d.pdf. 53. Cummings P. Methods for estimating adjusted risk ratios. Stata J. 2009;9(2):175. 54. Gittelsohn J, Rowan M, Gadhoke P. Interventions in small food stores to change the food environment, improve diet, and reduce risk of chronic disease. Prev Chronic Dis. 2012;9:E59. 55. Jeffery RW, French SA, Raether C, Baxter JE. An environmental intervention to increase fruit and salad purchases in a cafeteria. Prev Med. 1994;23(6):788–792. http://dx.doi.org/10.1006/pmed.1994.1135. 56. Slusser WM, Cumberland WG, Browdy BL, Lange L, Neumann C. A school salad bar increases frequency of fruit and vegetable consumption among children living in low-income households. Public Health Nutr. 2007;10(12):1490–1496. http://dx.doi.org/10.1017/S1368980007000444. 57. Sturm R, Cohen DA. Zoning for health? The year-old ban on new fastfood restaurants in south LA. Health Aff. 2009;28(6): w1088–w1097. http://dx.doi.org/10.1377/hlthaff.28.6.w1088. 58. Kirkpatrick SI, Reedy J, Kahle LL, Harris JL, Ohri-Vachaspati P, KrebsSmith SM. Fast-food menu offerings vary in dietary quality, but are consistently poor. Public Health Nutr. 2014;17(4): 924–931. http://dx.doi.org/10.1017/S1368980012005563. 59. Elbel B, Kersh R, Brescoll VL, Dixon LB. Calorie labeling and food choices: a first look at the effects on low-income people in New York City. Health Aff. 2009;28(6):w1110–w1121. http://dx.doi.org/10.1377/hlthaff.28.6.w1110. 60. Waterlander WE, Zenk SN. Food labelling, food retail availability and food pricing—moving from research to action? Public Health Nutr. 2015;18(1):2–7. http://dx.doi.org/10.1017/S1368980014002638. 61. Befort CA, Nazir N, Perri MG. Prevalence of obesity among adults from rural and urban areas of the United States: findings from NHANES (2005‐2008). J Rural Health. 2012;28(4): 392–397. http://dx.doi.org/10.1111/j.1748-0361.2012.00411.x. 62. Davis AM, Bennett KJ, Befort C, Nollen N. Obesity and related health behaviors among urban and rural children in the United States: data from the National Health and Nutrition Examination Survey 2003-2004 and 2005-2006. J Pediatr Psychol. 2011;36(6): 669–676. http://dx.doi.org/10.1093/jpepsy/jsq117. 63. James WL. All rural places are not created equal: revisiting the rural mortality penalty in the United States. Am J Public Health. 2014;104 (11):2122–2129. http://dx.doi.org/10.2105/AJPH.2014.301989. 64. Singh GK, Siahpush M. Widening rural–urban disparities in life expectancy, U.S., 1969–2009. Am J Prev Med. 2014;46(2): e19–e29. http://dx.doi.org/10.1016/j.amepre.2013.10.017. 65. Kulshreshtha A, Goyal A, Dabhadkar K, Veledar E, Vaccarino V. Urban-rural differences in coronary heart disease mortality in the United States: 1999-2009. Public Health Rep. 2014;129(1):19–29. 66. Sharkey JR, Horel S. Neighborhood socioeconomic deprivation and minority composition are associated with better potential spatial access to the groundtruthed food environment in a large rural area. J Nutr. 2008;138(3):620–627. 67. Gantner LA, Olson CM, Frongillo EA, Wells NM. Prevalence of nontraditional food stores and distance to healthy foods in a rural food environment. J Hunger Environ Nutr. 2011;6(3): 279–293. http://dx.doi.org/10.1080/19320248.2011.597829.
562
Zenk et al / Am J Prev Med 2015;49(4):553–562
68. Liese AD, Weis KE, Pluto D, Smith E, Lawson A. Food store types, availability, and cost of foods in a rural environment. J Am Diet Assoc. 2007;107(11):1916–1923. http://dx.doi.org/10.1016/j.jada.2007. 08.012. 69. Fleischhacker S, Johnson D, Quinn E, Pitts SBJ, Byker C, Sharkey JR. Advancing rural food access policy research priorities: process and potential of a transdisciplinary working group. J Agric Food Syst Community Dev. 2013;3(4):201–212. 70. Johnson DB, Quinn E, Sitaker M, et al. Developing an agenda for research about policies to improve access to healthy foods in rural communities: a concept mapping study. BMC Public Health. 2014;14: 592. http://dx.doi.org/10.1186/1471-2458-14-592.
71. Franco M, Diez Roux AV, Glass TA, Caballero B, Brancati FL. Neighborhood characteristics and availability of healthy foods in Baltimore. Am J Prev Med. 2008;35(6):561–567. http://dx.doi.org/10. 1016/j.amepre.2008.07.003. 72. Dannefer R, Williams DA, Baronberg S, Silver L. Healthy bodegas: increasing and promoting healthy foods at corner stores in New York City. Am J Public Health. 2012;102(10):e27–e31. http://dx.doi.org/ 10.2105/AJPH.2011.300615. 73. Cavanaugh E, Green S, Mallya G, Tierney A, Brensinger C, Glanz K. Changes in food and beverage environments after an urban corner store intervention. Prev Med. 2014;65:7–12. http://dx.doi.org/ 10.1016/j.ypmed.2014.04.009.
www.ajpmonline.org
