Diabetes and Health-Friendly Food Pantry Shelf Design and Implementation
A thesis submitted to the Graduate School of the University of Cincinnati in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
Department of Rehabilitation, Exercise, and Nutrition Sciences College of Allied Health Sciences
Madison Kelly Busch, RD, LD
B.S., Miami University 2016 Oxford, Ohio
Committee Chairs: Seung-Yeon Lee, Ph.D. Francoise Knox Kazimierczuk PhD, RDN, CSSD, LD, ATC, CSCS, NSCA-CPT, FAND
ABSTRACT
Objectives: The objective of this pilot study was to develop and implement the “Most Diabetes and Health Friendly” Shelf Initiative in an urban choice food pantry.
Methods: The “Most Diabetes & Health Friendly” shelf initiative was developed using grocery store marketing techniques and community based participatory research (CBPR). The program aimed to assist food pantry clients in identifying healthier food options available in a choice-food pantry. The criteria for healthier food options were established based on a literature review and current dietary recommendations. The healthiest options within each pantry section were highlighted using grocery store marketing techniques: priming, colored labeling, framing, and placement. The initiative was pilot tested in a choice pantry in Cincinnati, OH and usage was tracked for 6 months. The number of items on the intervention shelf versus general shelf was counted before and after pantry sessions. The total number items available, number items taken, and percentage items taken was tracked, calculated, then compared by section and shelf.
Results: A higher total number of items was taken from the intervention shelf (1987 items) compared to the general shelf (1863 items). The intervention shelf had a higher percentage of items taken per month (32-47%) compared to the general shelf (25-45%); however, the percentage items taken from the general shelf increased overtime. There was no clear pattern of choices shown across food sections.
Conclusions and Implications: Findings suggested the initiative was feasible to implement within a choice food pantry. A higher percentage of food items were taken from the intervention shelves compared to regular shelves. Further qualitative research is needed to determine whether the initiative is able to impact food pantry client’s dietary intake, as well as which outside factors most influence food choices within this population.
ii
iii ACKNOWLEDGEMENTS
There have been many individuals who have dedicated time and energy to assist with this project, in which this would have not been possible without. I would first like to thank my advisor, Seung-Yeon-Lee, PhD, for her guidance and mentorship throughout this study as well as graduate school. I would also like to thank Francoise Knox Kazimierczuk PhD, RDN for her willingness to serve on my thesis committee. I am also grateful for Claire Moorman, who volunteered to assist in data collection and was a pleasure to work with. Additionally, I would like to thank the Jewish Family Service Heldman Family Food Pantry, especially Sandee Golden and Ali Ulanski, who welcomed us in with a strong passion to better serve their community. I would also like to express my gratitude for the Community Based Participatory Research team who volunteered their time and energy to help create an effective and realistic intervention for their peers. Finally, I would like to thank my family and friends for their unwavering support and encouragement while furthering my education.
iv TABLE OF CONTENTS
Introduction………………………………………………………………………………………..1
Literature Review………………………………………………………………………………….2
Significance………………………………………………………………………………………14
Methods…………………………………………………………………………………………..15
Setting…………………………………………………………………………………....15
Development of the Diabetes and Health-Friendly Food Pantry Shelf Initiative………..17
Finalizing the Most Diabetes and Health-Friendly Food Pantry Shelf Initiative………..22
Implementation & Monitoring…………………………………………………………...25
Results………...………………………………………………………………………………….28
Discussion………………………………………………………………………………………..40
Conclusion……………………………………………………………………………………….46
References………………………………………………………………………………………..47
Appendix 1……………………………………………………………………………………….51
v LIST OF TABLES
Table 1. JFS food pantry users…………………………………………………………………...28
Table 2. Combined number of all items taken by month………………………………………...29
Table 3. Percentage of total items taken per month……………………………………………...30
Table 4. Total number of items taken by food category per month……………………………...31
Table 5. Percentage of items taken by food category per month………………………………...31
vi LIST OF FIGURES
Figure 1. Procedures for the development of the intervention…………………………………...18
Figure 2. Combined percentages of all items taken per month…………………………………..30
Figure 3. Percentage of applesauce taken………………………………………………………..32
Figure 4. Percentage of beans/lentils taken………………………………………………………33
Figure 5. Percentage of bread taken……………………………………………………………...34
Figure 6. Percentage of canned fruit taken………………………………………………………35
Figure 7. Percentage of canned vegetables taken………………………………………………..36
Figure 8. Percentage of cereal/oatmeal taken……………………………………………………37
Figure 9. Percentage of pasta taken……………………………………………………………...38
Figure 10. Percentage of rice taken………………………………………………………………39
vii INTRODUCTION
Food insecurity—defined as a lack of dependable access to adequate, safe, and nutritious food for all household members to live a healthy life—is a risk factor for developing type 2 diabetes and other diet-related chronic diseases.1-15 This association is thought to be related to the poor nutritional quality of easily-accessed, cheap foods and a lack of nutrition education.7-15
Despite the immense problem, a literature search revealed a lack of substantial interventions to target the serious, yet preventable public health burden. Choice food panties, which are a medium for food insecure individuals to choose and obtain food, are an ideal site for nutrition and health-related intervention.5 Applying consumer marketing techniques within a choice food pantry may promote healthier food selections. Consumer marketing techniques have historically been used by large companies to increase item sales, however have more recently been utilized in nutrition research to study the impact of utilizing the techniques on the promotion of healthy food options.16-18 The literature review revealed several successful techniques to improve selection of healthy foods including priming, colored labeling, framing, and product placement.18-31 It is hypothesized that the combination of these techniques used in a choice food pantry may improve the clients’ choices in the food pantry. If successful, this would provide a feasible and affordable solution for food pantries to adopt in order to improve client awareness to choose the most healthful options available for his or her condition regardless of income status or nutrition knowledge. The purpose of this study was to develop a diabetes-friendly food shelf intervention to target the food insecure population, then to track the number of food items taken from the intervention shelves compared to the general shelves.
1 LITERATURE REVIEW
Food Insecurity
Food insecurity is the lack of dependable access to adequate, safe, and nutritious food necessary for a healthy life. There are three categories to describe the severity of food insecurity including marginal food security, low food security, and very low food security. Marginal food security consists of those who may not always be defined as food insecure, however are households that have problems, at times, or anxiety about acquiring adequate food, but the quality, variety, and quantity of their food intake is not substantially reduced. Households with low food security report frequent reduced quality, variety, or desirability of the typical diet.
Households with very low food security are the most insecure, having multiple indications of disrupted eating patterns and reduced food intake.1 The common worries of food insecure individuals include running out of food, inability to afford balanced meals, and having to eat less or skip meals.2 Within the United States, 12.3% of households suffer from food insecurity with an additional 4.9% suffering from very low food security.2 In 2017, Cincinnati, Ohio identified a higher prevalence of food insecurity than the national average, with 19-31% of the adult population as food insecure, depending on county.3
Food security has been associated with multiple negative health outcomes for adults including physical impairments, obesity, psychological suffering, socio-familial disturbances, and increased risk for diet-related chronic diseases.4 Many of the associated chronic conditions are diet-related and could be prevented with access to adequate nutrition and basic health care including hypertension, chronic heart disease, chronic kidney disease, stroke, and type 2 diabetes.1
Food insecurity is considered an independent risk factor for the development of type 2
2 diabetes, which is problematic considering the already high national diabetes prevalence.5
According to the 2017 National Diabetes Statistic Report, an estimated 23 million people in the
United States have diagnosed diabetes. An additional 7.2 million people are estimated to have undiagnosed diabetes, with an added 84.1 million people with prediabetes.6 Moreover, an estimated 13% of adults in the Greater Cincinnati area had diagnosed diabetes, which is above the national average. This percentage increased by race (14% of African American adults), poverty level (17% at or below the 100% federal poverty level), and age (26% ages 65 years and older).3 A study of the impact of the different levels of food security on the development of chronic disease found that there is 60% increased risk of developing type 2 diabetes for those who suffer from marginal food security compared to those who are food secure. Worsened food insecurity status is then associated with a further increased risk of developing type 2 diabetes.
Individuals with very low food security are associated with an increase of over 100% prevalence of diabetes compared to high-security households.1
Nationally, food insecurity status is associated with increased risk of developing diet- related chronic diseases including prediabetes and type 2 diabetes.1,2,4,5 The same significant association between diabetes and food insecurity was consistently found amongst recent research findings.4,5,7-12 Type 2 diabetes is a preventable and manageable condition through a healthy diet and adequate exercise, therefore the increasing prevalence has potential to be reversed.6 Despite the evident need, a literature review revealed very few programs and interventions available nationwide to target the needs of at-risk food insecure individuals. There is an obvious nutrition and basic health care gap, including lack of intervention and education, in the food insecure population to combat development of chronic disease.5
3 Furthermore, food insecurity was also found to be associated with additional negative health outcomes such as decreased fruit and vegetable intake4,5,9,10, increased risk of overweight and obesity4,7, poor diet quality4,5,7, and poor quality of life.11,13 Food insecure individuals reported frequent intake of highly processed carbohydrates, sugar, and fat-laden food sources, which are typically the cheapest and/or most convenient options available.7,8,12 These foods are unfortunately also part of an unhealthy dietary pattern associated with the development of obesity, type 2 diabetes, and other chronic diseases.7
Furthermore, research has demonstrated a socioeconomic inequality within diabetes care.
Low socioeconomic individuals have been found to have poor disease management and worse intermediate health outcomes resulting in higher risk complications compared to higher socioeconomic classes.14 A study with Ippolito et al (2017) found that food insecure individuals with diagnosed diabetes suffer from worsened overall diabetes management compared to food secure individuals, supported by higher hemoglobin A1c levels and increased prevalence of diabetes distress and depression. Greater food insecurity further worsened these outcomes.11,13
The study also reported increased frequency of hypoglycemic episodes and poor adherence to imperative diabetes medications.11 Food insecure individuals with diabetes face many added challenges including increased frequency of life-threatening hyperglycemic and hypoglycemic episodes that puts them at a serious health risk.15 Inconsistent blood glucose levels are caused by inconsistent carbohydrate intake and poor medication adherence as well as an unhealthy diet pattern.2 Food insecurity has been shown to be associated with poor medication adherence due to having to choose between paying for medications or paying for food. Many food insecure individuals with diabetes report having to skip or reduce diabetes medications because of cost.
Individuals with diabetes have also been found to substitute healthy foods for inexpensive, high
4 caloric foods and to have frequent cycles of binging and fasting contributing to poor blood glucose control.15 Unfortunately, these unhealthy food sources are often what is most affordable and most easily accessed for the food insecure population.8
Choice Food Pantries
Food pantries are one important response to the food insecurity issue by helping households improve access to food, and therefore are a great medium to provide a food assistance to the food insecure population.7 Choice pantries are a type of food pantry that allows clients to choose for themselves or their household members what food items they receive.16 The architecture resembles a grocery store in which pantry users shop with a volunteer to pick out a specific allotment of food items from a variety of categories. Food pantry users are able to utilize the food pantry monthly. There different types of choice pantries including limited-client choice and full-client choice. Limited-client choice pantries are those in which clients are able to choose a predetermined amount of product from each category (i.e. canned fruits, bread, pasta, etc.). A full-client choice pantry is a pantry that most closely resembles a grocery store in which clients choose between categories and get various amounts of products within each category and make trade-offs between categories.16 Regardless of the specific type, choice pantries have benefits over a traditional food pantry to both the pantry and the clients. Benefits of a choice food panty include allowing the pantry to accept a wide arrange of food donations, to reduce food waste, and the ability to provide a variety of nutritious options. This type of pantry also gives clients a gained sense of dignity and control while allowing them to choose foods that are the best possible fit for their own situation and needs.17 Choice food pantries may also provide a medium for health care professionals to access and meet the basic nutrition and health care needs of food insecure individuals, therefore presenting itself to be a plausible setting for a nutrition
5 intervention.5 Many food pantries are able to offer heathy food options; however, individuals may lack the knowledge needed to choose the most appropriate items.
The current architecture and environment of a food pantry may unintentionally exacerbate client selection of less healthy items by not realizing the implications of environmental and architectural factors on consumer choice. Research has shown that unplanned food decisions are highly influenced by in-store marketing techniques and physical environment.18 For example, by including a dessert and pastries category to pick items from, this unintentionally encourages the selection of two sugar-laden items. By placing peaches canned in heavy syrup on an eye level shelf, while the peaches canned in water on the bottom shelf of the fruit category, this unintentionally urges the consumer towards choosing the peaches canned in heavy syrup over the healthier alternative. In a population that is already at an increased health risk, an environment that is not supportive of a nutritious eating pattern may be detrimental.18
People make an average of over 200 decisions about food each day. Some of these food decisions are rational and deliberate; however, some are also made subconsciously as a result of quick, instinctive actions.19 These subconscious food decisions can be largely influenced by the environment and marketing techniques.20
Very few interventions have been implemented in a food pantry setting to specifically target the high risk of developing type 2 diabetes in the food insecure population. A study by
Seligman et al (2015) provided a diabetes management intervention to food pantry clients, including screening and monitoring of glycemic control, access to diabetes-appropriate food, primary care referrals, and diabetes education. Diabetes-appropriate food was prepackaged into boxes and distributed every two weeks. Foods included whole grains, lean meats, beans, low- sodium vegetables, no sugar-added fruit, and shelf stable dairy products as well as supplemented
6 fresh produce, milk, yogurt, cheese, bread, and frozen lean meet. Recipes and cooking tips were also included. The cost of the boxes averaged $16 each. After the six-month intervention, participants reported an increased intake of fruits and vegetables as well as high satisfaction with the diabetes box. 88% of participants preferred the diabetes box foods over the regular food pantry options, while 10% reported giving away or throwing out the contents. Results showed small improvements in hemoglobin A1c and confidence in diabetes self-management.5
Although the intervention showed promising results, providing this level of care may not feasible within the average food pantry. The majority of food pantries have a limited budget that could not afford purchasing $16 worth of diabetes-appropriate foods for each client.
Additionally, providing a prepared allotment of food takes away the clients’ sense of control and dignity that a choice food pantry provides.17 Nevertheless, the highly reported satisfaction with the diabetes-appropriate food provided and the improved health outcomes continues to support the need for an easily-implemented intervention to help clients identify diabetes-appropriate items in a food pantry setting. Interventions should be simple and cost-effective in order to be widely disseminated within a variety of diverse food pantries settings regardless of budget and access to materials.
Health Marketing Techniques
Grocery stores and the food industry utilize a variety of marketing techniques to get consumers to purchase their products. To consumers, these techniques may appear as bright colors and signs, discounts, tasting events, shelf/store placement, and media.22 Based on the premise of consumer marketing, prototypes could also be manipulated to impact health-related behavior.23, 24 A marketing research survey of U.S. adults revealed that consumers are motivated by marketing techniques to purchase more healthful foods. These techniques included in-store
7 coupons and specials, availability of convenient healthy foods, product labels or advertising on packages, and labels or signs on shelves that highlighted the more healthful options. In-store tastings and recipes were also a motivating factor.21 Moreover, a study by Moore et al (2016) revealed that consumers may be influenced by external marketing cues and factors regardless of whether they are aware or not.21 Nutrition research has begun to study the impact of utilizing grocery store marketing techniques, including priming, colored labeling, framing, and product placement, to promote healthy food options.21-31 Studies have utilized these techniques with the intention to enhance the visibility and overall appeal of healthy options to consumers. Improving the marketing of health foods may improve nutrition choices and have the potential to improve overall dietary intake, and therefore nutrition-related health outcomes.21-31
Priming
The literature review revealed two studies that focused on priming individuals with messages before shopping for foods, then evaluated the types of foods purchased.23,26 Priming refers to activating the mental representation of a goal or other meaningful construct by means of subtle external cues that can affect subsequent cognitive professes and behavior.25 The first study by Papies et al (2014) used healthy primes to reduce unhealthy snack purchases among overweight consumers in a grocery store. Researchers tested the effectiveness of a simple health prime in hopes of combating the food temptations in the grocery store. The prime consisted of a recipe flyer that either contained a health and diet prime or a flyer that did not. As a result, the healthy prime group reduced the amount of unhealthy snack purchases by 75% in overweight and obese individuals. Furthermore, the healthy prime only worked if the consumer paid initial attention to the flyer; however, no conscious awareness during grocery shopping was necessary
8 to see an effect.26 These findings suggest that priming shoppers with healthy messages may be a highly viable and easily implemented intervention tool to promote more healthy food choices.
Another study that used the priming technique utilized terror management and mortality reminders instead of focusing on the benefits of healthy foods. In this study, grocery shoppers in the experimental group were exposed to a reminder of mortality and then asked to visualize a healthy meal prior to shopping. The mortality reminder consisted of a series of true or false questions with a scale measuring fear of death and then positive and negative self-reported affect. As a result, shoppers exposed to subtle mortality reminders purchased more nutritious foods compared to the control group. This study further supports that priming individuals before grocery shopping may be a feasible and successful way to positively impact health-related behaviors.23
Colored Labeling
Welcoming colors and engaging displays are also able to attract consumers to products.16,21 The literature review revealed studies that used colors to draw attention to particular goods in order to increase their attractiveness. Thorndike et al (2012) conducted a two-phase cafeteria intervention to improve the purchasing of healthy items over the unhealthy options. All food items were color coded to resemble a traffic light according to the nutrition information: green (choose often), yellow (choose less often), and red (there is a better choice in green or yellow). The categories were based on the current USDA healthy eating recommendations.
Possible positive criteria included being a fruit or vegetable, being a whole grain, and having lean or low-fat dairy or protein as one of the first three ingredients. Possible negative criteria included having a saturated fat content of over 5 grams per entrée or over 2 grams per item and/or having a caloric content of over 500 calories per entrée, over 200 calories per item, or
9 over 100 calories per beverage or condiment. Educational handouts and posted signs were available to explain the color coding. After 3 months of data collection, results showed a significant increase in green item sales with a significant decrease in red item sales.27 Findings from this study supports the color coding of healthy items to improve consumption.
A similar study was later completed with a sample of minority, low-income adults in which food items were color-coded and labeled with green indicating healthy choices or red indicating unhealthy choices. Items were categorized the same as the Thorndike et al (2012) study, with positive and negative criteria based on the USDA recommendations.27 Like the previous findings, colored labeling decreased red item purchases while simultaneously increasing green item purchases.28
An assessment of labeling influence on customers’ awareness of health and healthy purchases was evaluated via survey before and after implementation of the traffic light labeling intervention.29 The survey assessed basic demographics as well as if the consumer noticed any nutrition information in the cafeteria and how important nutrition information is when making decisions about foods and beverages. Traffic light labels were assessed by asking if the consumer noticed the red, yellow, and green labels, then asking if the labels influenced the items purchased. As a result, the number of people that reported health and nutrition as an important factor in food decisions increased from baseline to after the labeling intervention. A higher proportion of clients also reported noticing and looking at nutrition information after intervention compared to baseline. The same trend was seen for all sex and racial subgroups.29 These findings further support a color-coded labeling intervention to promote consumer awareness and increased importance of making healthy food choices while shopping. Benefits of the intervention may be even more significant than the survey was able to show since research
10 supports that clients may not always be aware of the external marketing factors influencing their food decisions.21
Moreover, at baseline, lower socioeconomic status individuals were more likely to purchase unhealthy foods compared to individuals with a higher socioeconomic status. However, after implementation of the color-coded intervention, similar improvements in the number of healthy items purchased was seen across all socioeconomic classes and race/ethnicities despite the differences at baseline.28 Color coding healthy items in the food pantry may be a feasible way to help all clients more easily identify the items without need for additional education. Simplified labeling techniques, such as traffic light color coding, allow for complex nutrition information to be conveyed quickly in a simple way.28
Framing
Another possible solution to help clients more easily identify the healthiest items available is to physically to frame the healthy items to separate them from unhealthy selections.
Products that are easiest to see, comprehend, and stand out are the products that are most likely to be selected. Targeting options by framing them makes the cognitive selection of these items easier.16 A study by Payne et al (2014) placed colorful lines of tape in shopping carts that segregated specific sections for fruits and vegetables with intention to observe changes in consumer food selection. As a result, the study found that the consumers with segregated shopping carts bought more fruits and vegetables compared to the control group.30 These findings support framing foods with colored tape to highlight healthy items and to encourage their selection.
11 Placement
Items that are seen first and are more accessible are more likely to be selected.16 The second phase of the studies by Thorndike et al (2012) and Levy et al (2012) tested the effect of rearranging the architecture of a cafeteria.27,28 This was completed after the significant findings from the three-month traffic light color-coding phase. Green items were rearranged to be at eye level (between 5-6 feet), while red items were placed on the bottom shelf. Healthy items were also more available throughout the store and purposefully placed to be seen first and more often.27 After another three months of data collection, significant results for both studies showed that changing the architecture further increased green item sales and further decreased red item sales.27, 28 These changes were seen consistently across race/ethnicity and socioeconomic status.28
In addition to making products eye level and more available, products that are placed first in a presentation of goods have been shown to encourage selection of that product.31 A study by
Wilson et al (2016) manipulated the display of products in a food pantry to test the effectiveness of environmental changes on consumer product choices. An assortment of snack bars was added to the dessert table and considered the healthier dessert option due to lower calories, increased fiber, and set portion size compared to the typical cake and cookie sections. Placing the bars at the beginning of the aisle increased consumer selection compared to when placed at the end of the aisle.31 This is thought to be because products in the first position may make the cognitive food decision easier.16, 20, 31 The literature supports simple changes in architecture, such as making products eye level, seen first, and more available, as effective ways to improve the selection of targeted products.16, 27, 28, 31 Changing the architecture of a food pantry to promote healthy items may be a feasible, yet subtle intervention that places no cognitive demand on the clients.
12 Implementation is suggested to be simple and inexpensive while still showing potential for wide dissemination and effect.28
13 SIGNIFICANCE
The comprehensive literature review reveals that there seems to be a lack of successful interventions to promote nutritious food options in the choice food pantry setting targeting food insecure populations, including adults with type 2 diabetes and other diet-related chronic disease1,4,5 Since type 2 diabetes is a preventable and manageable condition with proper diet and other lifestyle modifications, an intervention to combat the serious, yet preventable public health burden is necessary. 2 The architecture and environment within a choice food pantry may subconsciously influence food pantry clients’ selection of food items.18 Utilizing consumer marketing techniques to make changes to the food pantry architecture and environment may be a feasible and cost-effective way to encourage food panty users to choose healthier options in hopes of minimizing the public health burden. Previous studies have found making changes such as priming23, 26, colored labeling27, 28, 29, framing30, and product placement16, 27, 28, 31 as successful techniques.
14 METHODS
This study was composed of two parts: 1) developing the “Most Diabetes & Health
Friendly” pantry shelves initiative using a community based participatory research approach; and
2) implementing the initiative and tracking the numbers of food items taken from the shelves for
6 months at an urban choice pantry.
Setting
The most diabetes and health friendly pantry initiative was implemented at the Jewish
Family Service Heldman Family Food Pantry (JFS) in Cincinnati, Ohio between February 2018 and July 2018. The food pantry is affiliated with the larger non-profit Jewish Family Service
Center organization in Cincinnati, Ohio that provides a variety of care services to individuals in times of need without regard to religion, race, age, disability, sexual orientation, national origin, or ability to pay. Services include, but are not limited to, Medicare support, caregiver services, emergency financial assistance, homelessness support, adoption connection, and case management. The mission of the Jewish Family Services Center is to strengthen lives and the community by providing professional social services to families and individuals in times of need.
The food pantry is a part of the Vital Services Support in which provides a source for free food, personal care, and household care items.36
The food pantry is classified as a limited-client choice food pantry. Choice pantries are a type of food pantry that allows clients to choose for themselves what products they receive within varying degrees of healthiness and brands. The architecture resembles a grocery store in which pantry users shop with a volunteer to pick out a specific allotment of food items from a variety of categories for the month. Limited-client choice pantries are those in which clients are able to choose a predetermined amount of product from each category (i.e. canned fruits, bread,
15 pasta, etc.) based on household size. This is in comparison to a full-client choice pantry in which clients choose between categories, get various amounts of products, and may make trade-offs between categories.17 The majority of choice food banks are considered limited-client choice due to the available funds and resources.17
The pantry is unique in the way it offers two sides of the choice food pantry, Kosher and
Non-Kosher, to meet the needs of both Jewish and non-Jewish households. The Kosher sections were established related to the federal poverty statistics and studies of poverty within Jewish communities, estimating that as many as 9% of Jewish households in the Cincinnati area could be in need of food assistance.36 Upon entering the facility, clients sign into the front desk and are given a number. Individuals are called up in numerical order to meet with a social worker to discuss demographic and personal data including household size and whether they prefer to shop
Kosher or Non-Kosher. Clients are then assigned to a volunteer personal shopper who assists the client in choosing the correct amount of items depending on household size.
The food pantry services an average of 130-150 households per month. The majority of clients report themselves to be Caucasian (67%) or African American (26%) adults. The monthly household income is mostly between 0-99% of the Federal Poverty Guidelines (58%) or between
100-149% of Federal Poverty Guidelines (31%). The food panty is open to clients four days each month divided into five sessions for two hours at a time. Households are able to utilize the pantry one time per month. Individuals are able to select a pre-determined allotment of foods based on household size from each category with the assistance of a volunteer personal shopper. The amounts of items allotted per category may vary month to month depending on donations, food costs, and special events.36 All client food choices from the JFS food pantry between February
2018 and July 2018 were included in this study.
16 Development of the Diabetes and Health-Friendly Food Pantry Shelf Initiative
Community Based Participatory Research Approach was used for this study. CBPR is a partnership approach to research that involves community members, organization representatives, and researchers in all aspects of the research process, in which all partners provide expertise, decision making, and share ownership.32 As the first step in our procedures, a steering committee was formed with academic partners including a faculty member and three nutrition graduate students in Nutrition program at the University of Cincinnati, as well as the
Food Pantry Volunteer Manager, a Public Ally Americorps Food Pantry Assistant, three volunteers, and two clients from JFS. Through monthly meetings, the pantry initiative was developed and implemented following the procedures as shown in Figure 1.
17
1. Formation of a steering committee composed of 11 academic and community partners
2. Discussion on major health concerns in food pantry clients and potential intervention
strategies
3. Discussion on potential ideas for the Most Diabetes and Health Friendly Food Pantry
Shelf Initiative and volunteer training
o Initial idea: Traffic light shelf design
o Second idea: Two color shelf design using Green and Blue
o Final idea: one color shelf design using Green
4. Discussion on potential names of shelf and selection of the name -
5. Creation and discussion on magnets, poster, and broacher
6. Implementation of the initiative at the pantry
7. Six-month monitoring on the number of food items taken from the shelves by food
categories
Figure 1. Procedures for the development of the “Diabetes & Health Friendly” food pantry shelf initiative
The major health concern discussed by the steering committee included the ability to provide adequate components of a nutritious diet to promote long, healthy lives for all food pantry clients. The main condition in concern was type 2 diabetes and pre-diabetes. The food pantry needed an affordable, yet beneficial and feasible intervention. These concerns aligned with findings from the literature review suggesting that food insecurity is correlated with type 2 diabetes.1,3,5 The purpose of developing the Most Diabetes & Health Friendly shelf was to create a way for food pantry users to most easily identify the healthiest foods available regardless of nutrition knowledge or education.
18 The initial idea for the intervention shelf was to create a traffic light-coded system using stickers on individual food items as shown to be successful in previous studies.27,28,29 However, after discussion with several steering committee members, the Food Pantry and Volunteer
Manager and Public Ally Americorp Food Pantry Assistant decided this idea was not feasible to be implemented at the food pantry. Labeling every food item with a green, yellow, or red sticker was thought to be too time demanding and too challenging for the volunteer stockers. Unlike the cafeterias in previous studies, 27-28 food donations to the food pantry were not consistent nor delivered at one specific time, therefore would require continuous evaluation, sorting based on criteria, and marking. This method did not seem feasible without the addition of trained staff.
Furthermore, committee members had concern that labeling some foods as red, “poor choice” could be seen judgmental and against JFS’s value of creating a nonjudgmental environment within the organization. Choosing red food items over green or yellow could provide a basis for judgement, therefore the idea would not be an acceptable. Additionally, the clients using the food pantry are in need of any food to sustain life, therefore discouraging any available food choices is not recommendable.
At the first steering committee meeting, two new simplified color-coded systems for labeling were proposed. The committee discussed that labeling entire pantry shelf sections with colors would be more feasible and equally as successful compared to color coding individual items with a sticker. The first color system proposed used the two colors green and blue; however, the second color system used only the color green. The steering committee found the single color system more simple to use and understand compared to the two color system. The color green was selected because it was perceived to represent a healthy choice by committee members. The steering committee members associated the color green with ‘Go,’ ‘Healthy,’ and
19 ‘Good Choice.’ We then chose the name “Diabetes Friendly” as the green label to indicate that these items were recommended for a healthy diet to prevent development of diabetes or manage the disease. The steering team decided to focus the dietary components of the intervention on whole grains, sodium, and added sugar based on the recommendations and dietary guidelines from the American Diabetes Association, United States Department of Agriculture, and Feeding
America’s Food to Encourage.33-35 However, we did not want to discourage consumption of canned fruits and vegetables by not labeling them green based on the added sodium content or syrup concentration. The steering committee believed that even though the canned fruits and vegetables may contain a high sodium or added sugar content, they may still a good source of nutrients that could benefit the individual, over not eating fruits and vegetables are all, and therefore should not be discouraged. This was especially a concern if the food pantry were to run out of green color-coded canned fruits and vegetables in the green shelf sections. To prevent this problem, the addition of the blue shelf section to highlight the best options available without discouraging any fruit or vegetable consumption was suggested.
Within the two-color system proposal, the blue shelf sections were considered to be even healthier than the green shelf. This shelf would have even more strict nutrient guidelines, including the following: 100% whole grain for rice, pasta, and bread, <140mg sodium for canned vegetables and beans, less than or equal to 5g added sugar for cereal/oatmeal products, and fruit canned in light syrup, extra light syrup, fruit concentrate, or water. The green shelf requirements included the following: all whole grain rice, pasta, bread, and cereal products and all fruit, vegetable, and bean products regardless of nutrient content. All blue items would also qualify for the green shelf, but the green shelf did not qualify for the blue shelf. The idea was for the blue shelf to be a smaller subsection within the green shelf section to highlight the best options
20 available for the category. The green shelf was originally to be labeled “Diabetes Friendly” with the blue shelf labeled “Better Choice!”.
Naming the shelves proved to be difficult. It was challenging to find a way to describe the blue shelf as being healthier than the green shelf without using judgmental words. “Better
Choice!”, “Most Friendly”, and “Healthier Choice!” were all deemed inappropriate to represent the organization’s values. We also came up with the names “Health Special!” and “Blue Light
Special!” which were not considered too judgmental, however had already been used in the pantry’s Blue Light Special program. After more in-depth discussion, the steering committee came up with the names “Diabetes Friendly” for the green shelf and “Even More Friendly” for the blue shelf. Using the word “Friendly” in both labels helped tie the two shelves together as well.
Next, we provided two volunteer training sessions to introduce the food pantry initiative.
The volunteers who attended the training sessions found the names “Diabetes Friendly” and
“Even More Friendly” to be confusing. When having the volunteer shoppers practice role playing by explaining the shelves and answering basic questions, we noticed that the majority of them had a hard time distinguishing the two shelves. The majority of volunteers were not clear how the green shelves differed from the blue shelves. Volunteers commonly thought that the
“Diabetes Friendly” shelf was only for those with diabetes, while the “Even More Friendly” shelf was for those without, when in reality they were both for clients with and without diabetes.
Both shelves were considered healthy options, however the blue “Even More Friendly” shelf was the overall healthiest available. Furthermore, it took volunteers twice the amount of time to stock food items on shelves when they practiced sorting food items into sections based on the criteria for green shelves and blue shelves. Considering volunteers’ confusion and the increased stocking
21 time, the committee members concluded that this pantry shelf design may not feasible nor sustainable to implement over time.
Based on the discussion, the shelf designed was modified to the single-color system to make it simpler for volunteers to implement. In addition, implementing only one colored shelf was expected to make the clients’ decisions quicker and simpler by eliminating confusion. The color green was selected to represent the intervention. The criteria for food items for the green shelves included the following: 100% whole grain for rice, pasta, and bread, <140mg sodium for canned vegetables and beans, less than or equal to 5g added sugar for cereal/oatmeal products, and fruit canned in light syrup, extra light syrup, fruit concentrate, or water. Finally, the name of the green shelf was finalized as “Most Diabetes & Health Friendly”. The word “diabetes” was used to target clients or their household members with diabetes and “health” was intended to be inclusive and encouraging for clients and their household members without interested in a healthy eating, who may or may not have diabetes. A poster, handout, and volunteer training sessions to fit the changes were created.
Finalizing the Most Diabetes and Health-Friendly Food Pantry Shelf Initiative
Criteria for food items categorized to the “Most Diabetes & Health Friendly” pantry shelves
The criteria for food items categorized to the green shelves were established considering feasibility as well as the 2015-2020 USDA MyPlate Recommendations, 2014 American Diabetes
Association Nutrition Therapy Guidelines, and 2015 Feeding America Detailed Foods to
Encourage.33-35 The criteria was approved by a registered dietitian. Food pantry categories included applesauce, beans/lentils, bread, canned fruit, canned vegetables, cereals/oatmeal, pasta, and rice. Other food groups and items were not included due to difficulty simplifying sorting criteria for volunteers. A Volunteer Stocking Protocol document for reference was developed
22 and reviewed and finalized by the steering committee. The simplified criteria for the “Most
Diabetes & Health Friendly” shelves are as followed: Applesauce were required to be unsweetened and low sugar. Beans/lentils were required to be <140 mg sodium or no salt added.
Bread was required to have whole grain or whole wheat listed as the first ingredient. Canned fruit was required to be canned in water, 100% juice, juice from concentrate, or extra light syrup.
Canned vegetables were required to be <140 mg sodium or no salt added. Cereals/oatmeal were required to have whole grain or whole wheat listed as the first ingredient and have less than or equal to 5 grams sugar. Pastas were required to have whole grain or whole wheat listed as the first ingredient. Rice was required to have whole grain or whole wheat listed as the first ingredient.
Poster and Handout
A poster explaining the “Most Diabetes & Health Friendly” shelves was created. We used a positive, yet clear message to urge clients towards selecting the healthier options, careful not to discourage selection of general items in a negative way. A handout which explained the “Most
Diabetes & Health Friendly” shelf in detail was developed. The handout explained the benefits of consuming whole grains and reducing sodium intake and added sugar intake. It also encouraged clients to rinse canned vegetables and canned fruits to potentially reduce added sodium or added sugar consumption. Contents of the information, wording, font style and size, color scheme, photos, and final design were finalized by the steering committee members.
Volunteer and Staff Training
Training sessions were held for both volunteer stockers and volunteer shoppers. The role of a stocker was to help sort foods in the beginning of each month and place (stock) the items on the correct shelf. Shoppers’ role was to assist food pantry clients in the selection of items during
23 open pantry hours. They are responsible for walking the clients through the aisles and explaining the allotted amount from each category based on household size. Multiple training sessions were held to accommodate for different schedules. Sessions began by giving a background of the current food insecurity and diabetes health burden in Cincinnati, the association between intake of sodium, added sugar, and simple carbohydrates and risk for type 2 diabetes, and the importance of having healthy diet for disease management and prevention.
The training for volunteer stockers focused on the criteria for food items to be placed on the “Most Diabetes & Health Friendly” shelves. Volunteers were taught how to read Nutrition
Facts labels to determine the sodium content per serving in canned vegetables and beans/lentils, as well as the sugar content per serving in cereal/oatmeal. They were taught to identify whole grain and whole wheat products by looking for the Whole Grains Counsel’s Whole Grain Stamp and/or by looking for “whole grain” or “whole wheat” listed first on the ingredient list for bread, pasta, and rice. They were also taught how to read the item labels to find what kind of liquid canned fruits were packaged in (i.e. heavy syrup, light syrup, water, etc.). The Volunteer Sorting
Protocol document that summarized stocking guidelines was distributed for reference. Stockers then physically practiced reading the Nutrition Facts labels and other nutrition information on packages and stocking items to the appropriate shelf in the food pantry. Guidance from a registered dietitian and nutrition graduate students was available during these sessions.
The primary purpose of the shopper training session was to familiarize volunteers with the “Most Diabetes & Health Friendly” shelves initiative. Volunteers were also taught how to introduce the green shelves initiative to clients and answer basic questions. The training session began by introducing the intervention shelf and teaching volunteers to recognize the green “Most
Diabetes & Health Friendly” shelf, shelf labels, posters, and handouts. Shoppers were instructed
24 to make food pantry clients aware of the new “Most Diabetes & Health Friendly” shelf and express that it was designed to help identify the healthiest options available. They were taught not to urge the client to make one food choice over another and/or shame a client for choosing an item from the general shelf over the intervention shelf. Shoppers were then educated on the benefits of following a diet pattern rich with whole grains, that meets sodium recommendations, and limits added sugars. They were taught that items from the “Most Diabetes & Health
Friendly” shelf are part of a healthy diet and may protect against the development of chronic diseases, such as diabetes and heart disease. Shoppers had the opportunity to role play the interaction between the personal shopper and the food pantry client by answering possible questions that may come up. Shoppers were given a “Shopper Cheat Sheet” to summarize the green shelves, how to introduce them, and possible benefits of choosing food items on the green shelves. “Shopper Cheat Sheets” were also available in each shopping cart for quick reference during open pantry hours. Furthermore, volunteer shoppers were given a brief review of their roles by a nutrition graduate student or the food pantry manager in a short meeting prior to each pantry session.
Implementation and Monitoring
The Most Diabetes and Health Friendly Food Pantry Shelf initiative was implemented in the food pantry and the number of items taken from the shelves by food category were tracked for a total of 24 sessions between February 2018 and July 2018 (Appendix 1). Data from Kosher and Non-Kosher pantry sections were collected separately. The poster explaining the “Most
Diabetes & Health Friendly” shelves was placed in five locations around the food pantry visible to clients during check-in, in the waiting line, and while shopping around the pantry. The handout was handed to clients during the check-in process in February and then displayed and
25 readily available within the following months. During the first month of intervention, a registered dietitian was available in the pantry to answer any questions about the green shelves and the handout.
A nutrition graduate student was available for questions during stocking hours for the first two months of the study. Proper placement of items was checked after stocking and corrected as needed. The number of food pantry clients who shopped at the pantry and their household sizes per session were obtained from the food pantry manager.
The number of items taken from each food category within the “Most Diabetes & Health
Friendly” shelves versus the general shelves at each session were tracked for six consecutive months. A nutrition graduate student and undergraduate student hand-counted and recorded all food items before and after each pantry session on physical documents. The number of food items in each category were counted twice to ensure accuracy. The graduate or undergraduate student corrected shelf placement of food items prior to data collection as needed. Data was collected for both the “Most Diabetes & Healthy Friendly” shelves and the general shelves as well as for both Kosher and Non-Kosher sections. Data was manually entered into Microsoft
Excel Software and double checked for input accuracy.
Data Analysis
All descriptive statistics calculations, tables, and graphs were conducted in Microsoft
Excel Software Version 16.17. The number of clients who shopped at the food pantry and their household size was obtained from the food pantry manger and JFS records. The number of items taken from each food category within the “Most Diabetes & Health Friendly” shelves versus the general shelves at each session were calculated per month. The number of items available was determined by hand counting the total items available at the beginning of the month for each
26 item in each category, and on each shelf. We also considered additions or subtractions to this number between pantry sessions. Items may have been added due to restocking and new donations. Items may have been subtracted due to discarding expired or spoiled items.
The number of items taken per session was calculated by subtracting the number of items remaining on a shelf after the pantry session from the number of items prior to the session on each shelf, in each category. The numbers taken per session were added to calculate the total number of items taken per month by food category. Finally, the sum of the total number taken in each food category gave the total items taken from green shelves versus general shelf.
The percentage of items taken from the green shelf and general shelf by each food group per month was also calculated due to the inconsistent number of food items available in each food category prior to pantry sessions. The percentage of items taken per month was calculated by dividing the total number taken by the number of items available for each shelf and food category. The overall percent of items taken per month from the green shelves versus the general shelves was calculated by dividing the total number items taken per month by the total number of items available per month, for each shelf.
Among the data collection from a total of 24 sessions, data from the two pantry sessions on June 19 and June 20 in 2018 were excluded from data analysis because of inaccurate data collection. Between these dates, volunteers and employees of the food pantry began to restock food items onto shelves prior to data collection causing the number of items remaining on the shelves to be inaccurate. Because the number of items taken from the green shelves and general shelves in the Kosher section versus Non-Kosher section were not distinctively different, data from the Kosher and non-Kosher sections were combined, then summarized and reported together in this study.
27 RESULTS
The number of clients who shop in the food pantry
The number of households who used on-site shopping at JFS, not food bag delivery service, during the study ranged from 61 to 89 households (106 to 146 individuals) per month.
The usage per pantry sessions also greatly varied, ranging from 1 to 34 households as shown in
Table 1. The greatest number of individuals utilized the food pantry in June. The least number of individuals utilized the food pantry in February (Table 1).
Table 1. Number of JFS food pantry clients per session and per month
Total # households that Total # individuals who Date used on-site shopping benefited from on-site at JFS shopping at JFS 2/7/18 13 34 2/8/18 16 23 2/21/18 31 45 2/22/18 1 4 February Total 61 106 3/14/18 13 19 3/15/18 27 40 3/27/18 14 27 3/28/18 34 49 March Total 88 135 4/4/18 7 10 4/5/18 10 14 4/24/18 28 60 4/25/18 33 60 April Total 78 144 5/9/18 13 15 5/10/18 15 34 5/22/18 21 32 5/23/18 27 50 May Total 76 131 6/6/18 18 23 6/7/18 24 41 6/19/18 20 28 6/20/18 27 54 June Total 89 146 7/11/18 18 35 7/12/08 19 26 7/24/18 22 32 7/25/18 29 51 July Total 88 144
28 The number of food items taken from the “Most Diabetes & Health Friendly” shelves
A higher total number of items was taken from the intervention shelves compared to the general shelves throughout the 6-month study. A greater number of food items were taken from general shelves compared to green, intervention shelves in May and July (Table 2). The number of available items greatly varied by month. For example, of the food categories monitored, there were 2,088 total food items available at the food pantry in March, while only 1,605 total items available in July.
Table 2. Combined number of all items taken from the intervention shelf vs. general shelf by month
Feb March Apr May June* July Total Intervention Shelf 342 404 360 313 211 357 1987 General Shelf 260 349 333 369 167 385 1863
*June only includes data for two of the four possible pantry sessions
When comparing the overall percentage of available items taken per month, a higher percentage of items from the “Most Diabetes & Health Friendly” intervention shelf compared to general shelf items for all six months of the study, as shown in Figure 2 and Table 3. The intervention shelf also had a higher total percentage of items taken. The percentage of items taken from the general shelf slightly increased over time. The difference in percentages between the intervention shelf and general shelf lessened overtime with clients choosing an increasing number of general shelf items. The greatest difference in percentage of items taken per shelf took place in March, while the smallest difference took place in July (Table 3) (Figure 2).
29 Table 3. Percentage of total food items taken from the intervention shelves vs. general, non-green shelves per month (%)
Feb March April May June* July Total Intervention Shelf 38.7% 44.3% 45.7% 38.6% 32.3% 47.2% 40.7% General Shelf 24.5% 29.7% 34.5% 36.7% 23.5% 45.4% 32.3%
Proportion = (Total food items taken from each shelf / Total food items available on shelf) x100 *June only includes data for two of the four possible pantry sessions
100.0% 90.0% 80.0% 70.0% 60.0% 47.2% 50.0% 44.3% 45.7% 45.4% 38.7% 38.6% 34.5% 36.7% 40.0% 29.7% 32.3% 30.0% 24.5% 23.5% 20.0% 10.0% 0.0% February March April May June July
Intervention shelf General shelf
Figure 2. Combined percentage of all items taken from the intervention shelves vs. general shelves per month
Proportion = (Total food items taken from each shelf / Total food items available on shelf) x 100 * June only includes data for two of the four possible pantry sessions
30 Table 4. Total number of items taken from the intervention shelves vs. general shelves by food category per month
February March April May June July (n=61) (n=88) (n=78) (n=76) (n=89) (n=88) # # # # # # # # # # # # Available Taken Available Taken Available Taken Available Taken Available Taken Available Taken I-Ap 77 31 78 41 59 29 89 33 64 24 79 43 G-Ap ------I-BL 69 28 107 28 98 26 86 19 67 21 73 36 G-BL 169 35 259 47 235 53 226 69 165 22 189 74 I-Br 16 15 39 39 1 1 1 1 - - 6 5 G-Br 42 15 22 22 20 20 20 20 6 6 37 32 I-CFr 259 118 268 154 229 174 210 127 182 93 191 132 G-CFr 34 7 40 11 33 23 25 20 4 2 15 13 I-Cveg 364 107 299 93 296 86 298 73 238 41 282 80 G-Cveg 359 81 361 92 288 85 311 102 253 57 249 89 I-CO 65 20 86 25 55 21 69 25 70 12 90 37 G-CO 100 45 95 46 87 44 124 57 62 23 116 57 I-Pas 18 10 7 3 19 11 22 10 11 5 12 6 G-Pas 267 46 315 88 239 68 223 69 171 42 156 77 I-Ri 19 13 28 21 30 12 36 25 22 15 24 18 G-Ri 88 31 84 43 62 40 76 32 51 15 86 43 n = number of clients who shopped at the food pantry I = Intervention, green “Most Diabetes & Health Friendly” shelf; G = General shelf Ap = Applesauce; BL = Beans/lentils; Br = Bread; CFr = Canned fruit; Cveg = Canned vegetables; CO = Cereal/oatmeal;
Table 5. Percentage of items taken from the taken from the intervention shelf vs. general shelf by food category per month
February March April May June July I-Applesauce 40.3% 52.6% 49.2% 37.1% 37.5% 54.4% G-Applesauce ------I-Beans/Lentils 40.6% 26.2% 26.5% 22.1% 31.3% 49.3% G-Beans/Lentils 20.7% 18.1% 22.6% 30.5% 13.3% 39.2% I-Bread 93.8% 100.0% 100.0% 100.0% N/A 83.3% G-Bread 35.7% 100.0% 100.0% 100.0% 100.0% 86.5% I-Canned Fruit 45.9% 57.5% 76.0% 60.5% 51.1% 69.1% G-Canned Fruit 17.6% 27.5% 69.7% 80.0% 50.0% 86.7% I-Canned Vegetables 29.4% 31.1% 29.1% 24.5% 17.2% 28.4% G-Canned Vegetables 22.6% 25.5% 29.5% 32.8% 22.5% 35.7% I-Cereal/Oatmeal 30.8% 29.1% 38.2% 36.2% 17.1% 41.1% G-Cereal/Oatmeal 45.0% 48.4% 50.6% 46.0% 37.1% 49.1% I-Pasta 55.6% 42.9% 57.9% 45.5% 45.5% 50.0% G-Pasta 17.2% 27.9% 28.5% 30.9% 24.6% 49.4% I-Rice 68.4% 75.0% 40.0% 69.4% 68.2% 75.0% G-Rice 35.2% 51.2% 64.5% 42.1% 29.4% 50.0% I = Intervention, green “Most Diabetes & Health Friendly” shelf; G = General shelf
31 Availability of Food Items in Green and General Shelves by Food Category
Throughout the study, there was an inconsistent supply of donations resulting in an inconsistent number of options. For example, there was not a supply of applesauce fit for the general shelf throughout the study, while there was ample supply of applesauce categorized to the intervention shelf. Furthermore, food pantry clients had the right to accept or decline food categories depending on preference and needs. Because of this, food groups have different choice patterns and trends. Availability of items within the same food category also varied per month. As shown in Table 4, there were 39 bread items available on the intervention shelf in
March compared to only 1 bread item available on the same shelf in April. Further results within each food category are described in the following.
Applesauce 100.0%
80.0%
54.4% 60.0% 52.6% 49.2% 40.3% 37.1% 37.5% 40.0%
20.0%
0.0% February March April May June July Intervention shelf General shelf
Figure 3. Percentage of applesauce taken from the intervention shelf vs. general shelf by month
Applesauce
Throughout the study, there was no supply of general shelf applesauce available to pantry clients. Only green, intervention shelf applesauce was available. The number of items available on the “Most Diabetes & Health Friendly” shelf ranged from 59-89 items per month. Between
24-43 items per month were taken from the “Most Diabetes & Health Friendly” shelf (Table 4).
32 The percentage taken from the “Most Diabetes & Health Friendly” shelf ranged from 37.1% to
54.4% (Figure 3).
Beans & Lentils 100.0%
80.0%
60.0% 49.3% 40.6% 39.2% 40.0% 26.5% 30.5% 31.3% 26.2% 22.1% 20.7% 18.1% 22.6% 20.0% 13.3%
0.0% February March April May June July Intervention shelf General shelf
Figure 4. Percentage of beans/lentils taken from intervention shelf vs. general shelf by month
Beans and Lentils
The number of items available on the intervention shelf ranged from 67-107 items, which was less than the number of items available on the general shelf, ranging from 165-235 items.
Among those, 19-36 items from the intervention shelf were taken per month compared to 22-74 items from the general shelf were taken per month (Table 4). The percentage of beans/lentils taken from the intervention shelf decreased from February to May, then increased again in June and July. The percentage of general shelf beans and lentils increased over time. The percentage taken from the intervention shelf ranged from 22.1% to 49.3%. The percentage taken from the general shelf items ranged from 13.3% to 39.2% (Figure 4).
33 Bread 100.0% 100.0% 100.0% 93.8% 100.0% 100.0% 86.5% 83.3% 80.0%
60.0%
40.0% 35.7%
20.0%
0.0% February March April May June July Intervention shelf General shelf
Figure 5. Percentage of bread taken from the intervention shelf vs. general shelf by month
No bread available on the intervention shelf in June.
Bread
The availability of bread donation varied considerably across months. It was more common to have adequate supply of general shelf bread items compared to intervention shelf bread items. The number of items available on the intervention shelf ranged from 0-39 items, which was less than the number of items available per month on the general shelf ranging from
6-42 items per month. In June, there were bread items available on the intervention shelf.
Between 0-39 items per month were taken from the intervention shelf compare to 6-32 items per month from the general shelf (Table 4). The percentage taken from the intervention shelf ranged from 83.3% to 100% compared to 35.7% to 100% taken from the general month (Figure 5).
34 Canned Fruit 100.0% 80.0% 86.7% 80.0% 76.0% 69.7% 69.1% 57.5% 60.5% 60.0% 45.9% 51.1% 50.0% 40.0% 27.5% 20.0% 17.6%
0.0% February March April May June July Intervention shelf General shelf
Figure 6. Percentage of canned fruit taken from the intervention shelf vs. general shelf by month
Canned Fruit
The number of items available per month on the intervention shelf ranged from 182-268 items, which was greater than the number of items available on the general shelf ranging from 4-
10 items per month. Food pantry clients chose between 93-174 items per month from the intervention shelf compared to 2-23 items per month from the general shelf (Table 4). The percentage of canned fruit taken from the intervention shelf increased from February to April, dipped in May and June, then increased again in July. The percentage of canned fruit taken from the general shelf increased over time, except June. The percentage taken from the intervention shelf ranged from 45.9% to 76.0% and the percentage taken from the general shelf items ranged from 17.6% to 80.0%. The percentage of canned fruit taken from the green shelf was higher than those from general shelf during the first 3 months, however a higher percentage was taken from general shelf than the green shelf in May and July (Figure 6).
35 Canned Vegetables 100.0%
80.0%
60.0%
35.7% 40.0% 29.4% 31.1% 29.5% 32.8% 25.5% 29.1% 28.4% 22.6% 24.5% 22.5% 20.0% 17.2%
0.0% February March April May June July Intervention shelf General shelf
Figure 7. Percentage of canned vegetables taken from the intervention shelf vs. general shelf by month
Canned Vegetables
The number of canned vegetable items available per month on the intervention shelf ranged from 238-364 items, which was similar to number of items available on the general shelf ranging from 249-359 items per month. Between 41-107 items were taken per month from the intervention shelf compared to between 57-102 items taken per month from the general shelf
(Table 4). The percentage of canned vegetables taken from the intervention shelf remained relatively consistent over 6 months, while the percentage of general shelf canned vegetables gradually increased over time. The percentage taken from the intervention shelf ranged from
17.2% to 31.1%, while the percentage of items taken from the general shelf ranged from 22.5% to 35.7% (Figure 7).
36 Cereal & Oatmeal 100.0%
80.0%
60.0% 48.4% 50.6% 45.0% 46.0% 49.1% 41.1% 38.2% 36.2% 37.1% 40.0% 30.8% 29.1% 20.0% 17.1%
0.0% February March April May June July Intervention shelf General shelf
Figure 8. Percentage of cereal/oatmeal taken from the intervention shelf vs. general shelf by month
Cereal & Oatmeal
The number of cereal/oatmeal items available per month on the intervention shelf ranged from 55-90 items, which was less than the number of items available on the general shelf ranging from 57-124 items per month. Between 12-37 items per month were taken from the intervention shelf compared to between 23-57 items taken per month from the general shelf (Table 4). The percentage of cereal/oatmeal taken from both shelves remained relatively consistent overtime.
The percentage taken from the intervention shelf ranged from 17.1% to 41.1%, while the percentage taken from the general shelf items ranged from 37.1% to 50.6% (Figure 8).
37 Pasta 100.0%
80.0% 57.9% 60.0% 55.6% 50.0% 49.4% 42.9% 45.5% 45.5% 40.0% 28.5% 30.9% 27.9% 24.6% 17.2% 20.0%
0.0% February March April May June July Intervention shelf General shelf
Figure 9. Percentage of pasta taken from the intervention shelf vs. general shelf by month
Pasta
The number of items available per month on the intervention shelf ranged from 7-22 items, which was less than the number of items available on the general shelf, ranging from 156-
315 items per month. Between 3-10 items were taken per month from the intervention shelf, compared to between 42-88 items taken per month from the general shelf (Table 4). The percentage of pasta taken from the intervention shelf was highest in February and April. The percentage of general shelf pasta slightly increased over time. The percentage taken from the intervention shelf ranged from 42.9% to 57.9%, while the percentage of items taken from the general shelf ranged from 17.2% to 49.4% (Figure 9).
38 Rice 100.0%
75.0% 75.0% 80.0% 68.4% 69.4% 68.2% 64.5% 60.0% 51.2% 50.0% 42.1% 35.2% 40.0% 40.0% 29.4% 20.0%
0.0% February March April May June July Intervention shelf General shelf
Figure 10. Percentage of rice taken from the intervention shelf vs. general shelf by month
Rice
The number of items available per month from the intervention shelf ranged from 15-36 items, which was less than the number of items available on the general shelf ranging from 51-88 items per month. Between 12-25 items were taken per month from the intervention shelf compared to between 15-43 items taken per month from the general shelf (Table 4). The percentage of rice taken from the intervention shelf remained relatively consistent over the study, except for a dip in April as shown in Figure 10. The percentage of rice taken from the general increased from February to April, then dipped in May and June. The percentage taken from the intervention shelf ranged from 40.0% to 75.0%, while the percentage taken from the general shelf items ranged from 29.4% to 64.5% (Figure 10).
39 DISCUSSION
Our results showed that food pantry users chose a higher total number of items, as well as a higher percentage of combined items from the “Most Diabetes & Health Friendly” intervention shelf compared to the general shelf. There was not a clear trend in the percentage of food items taken from the intervention shelves during the 6 months of data collection, however the percentage of items taken from the general shelves showed a gradual increase over time. The data trends for the percentage of items taken per month varied between food categories and did not show a consistent 6 month pattern, suggesting that factors other than health and nutrition likely impacted food decisions.
Outside factors that may have influenced consumer choices include donation quality and quantity and the types of foods available on each shelf category.16,18, 22 The number of items available was not consistent throughout the study. JFS did not consistently have optimal supply of foods for each shelf and category throughout the study, especially in applesauce, bread, and pasta sections. According to research by Wright et al, consumers often attribute limited availability to product demand, which implies a higher quality item and therefore may have impacted food pantry users’ choices.24 Research also suggests that consumer choices are influenced by the appeal and familiarity of brands and packaging.18,20-21,24 The availability of a preferred item or brand may have influenced food selection over the location of the food shelves.
Finally, food choices within the food pantry population may have been influenced by package size and the type of cooking method required.28 These outside factors may have impacted study results by influencing consumer choices over the nutritional quality and shelf placement. Factors that may have contributed to each food category are as followed:
40 Only applesauce categorized to the “Most Diabetes & Health Friendly” shelf was available during this study. The majority of applesauce supply was unsweetened and canned.
Food pantry users did not have the opportunity to choose from different shelves or products.
Beans and lentils were available on both shelves throughout the study, however the brand, size, and types of items varied. The intervention shelf contained both canned and dry bag forms. The general shelf only contained canned beans/lentils, however had an increased flavor variety including baked beans and pre-seasoned beans. The types of items available, difference in brands, items size, and access to a kitchen may have impacted consumer choices. For example, the dry bags of beans and lentils contained more product, however required access to kitchen equipment to cook. Because the dry bags were mostly categorized as “Most Diabetes &
Health Friendly” options, this may have been an outside factor affecting results. The percentage of beans/lentils taken showed a U-shaped pattern for items on the “Most Diabetes & Health
Friendly” shelf. The initial interest in beans/lentils from the intervention shelf may have been related to a photo of black beans in the educational handout introduced in February. The decrease in the intervention shelf paired with the increase in the general shelf may have been related to disliking the low sodium taste and/or to decreased marketing and interest in the study as time progressed. The increase of all bean/lentil products in July may have been influenced by the increase in total food pantry clients.
The stock of bread items was typically lower than other food categories, however the majority of bread was taken each month regardless of shelf label and nutrition quality. Note, the lower percentage of items taken from the general shelf in February was likely related to a large bread donation at the end of the month after many of the clients had already received their monthly portions. The availability of bread categorized to the general shelf was consistently
41 higher than the intervention shelf. There was only one intervention shelf bread item available for the entire month in April and May. Taking one of the “Most Diabetes & Health Friendly” shelf items made a substantial change in the percentage of items taken compared to taking one general shelf item due to the number of items available.
The majority of canned fruit donations were classified as intervention shelf items. The general shelf was notably smaller holding only 4-40 items compared to the green, intervention shelf holding between 182-259 items Taking one of the general shelf items made a substantial change in the percentage of items taken compared to taking one green shelf items, therefore the total number of items taken better reflects consumer choice patterns for canned fruit. The smallest supply of canned fruit on the general shelf was available in May-June, which reflects the increase in percentage of items taken over the “Most Diabetes & Health Friendly” shelf.
Furthermore, the variety and types of items available on the shelves greatly differed. The general shelf mostly consisted of sweetened cranberry sauce compared to the wide variety of fruits found on the “Most Diabetes & Health Friendly” shelf. The differences in availability and variety between shelves may have been an outside factor effecting results.
JFS had ample supply and availability of canned vegetables for each shelf throughout the study, however the same types of vegetables were not typically found on both the intervention and general shelf. For example, carrots may have been an available option on the general shelf, however not an option on the “Most Diabetes & Health Friendly” shelf. If a client only liked carrots, he or she may have chosen the item based on availability rather than nutritional quality or shelf placement. Therefore the types of vegetables available may have influenced consumer choices. The percentage of items taken from the “Most Diabetes & Health Friendly” shelf
42 decreased while the percentage taken from the general shelf increased throughout the study. This may have been related to decreased marketing or interest in the study as time progressed.
Cereal/oatmeal items were available for both shelves through the study, however the types of items and brands varied. The results show that pantry users consistently chose a higher percentage of available items from the general shelf compared to the “Most Diabetes & Health
Friendly” shelf, suggesting that outside factors beyond nutritional quality impacted consumer choices. The majority of items on the general shelf were sweetened and flavored compared to items available on the “Most Diabetes & Health Friendly” shelf. Additionally, general shelf items typically contained cartoons on the front and were marketed to children. It was also uncommon to have the same brands and products available on each shelf. For example, the general shelf may have contained Frosted Cheerios, while the intervention shelf did not contain supply of a similar option, such as plain Cheerios. Therefore, if a client preferred Cheerios, he or she may choose the item based on brand availability and familiarity rather than nutritional quality or shelf placement. The limited amount of items and brands available may have influenced consumer choices.
The majority of available pasta items were categorized to the general shelf. There were only between 7-22 items stocked on the intervention shelf at one time compared to 171-315 items stocked on the general shelf. Because of this, taking one of the intervention shelf items made a substantial change in the percentage of items taken compared to taking one general shelf item. The smallest supply of pasta on the general shelf was available in July, which reflects the increase in percentage taken during that month. Results indicate that food pantry users took a higher percentage of available items from the “Most Diabetes & Health Friendly” shelf, however chose a higher total number of general, non-green shelf items. Furthermore, the types/form of
43 pasta available on the shelves differed. For example, the general shelf commonly contained six forms of pastas while the intervention shelf would only contain one or two different forms.
Therefore, if a client preferred one form, he or she may choose the item based on availability rather than nutritional quality or shelf placement. The limited amount of items and types available may have influenced client choices.
Throughout the study, rice donations differed by availability and product type which may have impacted consumer choices. The “Most Diabetes & Health Friendly” shelf consistently contained less than half of the number of items and compared to the amount of items available on the general shelf. The general shelf was notably larger holding 51-88 items compared to the intervention shelf holding between 19-36 items. Taking one of the green, intervention shelf items made a substantial change in the percentage of items taken compared to taking one general, non- green shelf item. The types of rice products, size, and cooking method needed also differed greatly. This may be a reason for the percentage taken pattern shift seen in Figure 10. The intervention shelf included only dry whole grain rice and occasionally quinoa, while the general shelf contained a wider variety of products including dry rice, flavored/seasoned rice, rice mixes, and microwave rice. The types of items available, item size, and access to kitchen equipment may have impacted consumer choices. For example, the dry bags of rice contained more product, however required access to a kitchen to cook. Because the microwave rice was categorized as general shelf options, this may have influenced results.
One major strength of this study was the use of CBPR approach to develop a more relevant, feasible, and sustainable intervention. Additionally, the intervention was implemented for a total of six months to track the long-term number of food items taken from the intervention shelves versus the general shelves. This time frame demonstrated feasibility and sustainability of
44 the “Most Diabetes & Health Friendly” Food Shelves Initiative in a choice food pantry setting over time. A limitation of this study was that we were unable to control for the variance in availability of food items, which may have influenced the number of items taken from the intervention shelves versus the general shelves. Second, client choices were only assessed by counting the number of items taken from the intervention and general shelves after open pantry sessions. We did not track individual client food choices and patterns overtime via pre- and post- tests.
45 CONCLUSION
Findings suggested that implementing a diabetes and health-focused food shelf intervention in a choice food pantry is feasible long term. Considering inconsistent availability of items on the intervention shelves due to variability in donations, providing education for donors by using donation guidelines to request healthier items may need to be implemented concurrently with this initiative. An assessment of changes in the selection of foods categorized to the intervention shelves using pre- and post-tests may be needed to determine the effectiveness of behavior changes from this initiative. In addition, a qualitative study can be conducted to explore clients’ perception of the “Most Diabetes & Health Friendly” shelf initiative including how it has influenced food selection behaviors and how the intervention may be improved and better promoted.
46 REFERENCES
1. Gregory CA., & Coleman-Jensen A. Food insecurity, chronic disease, and health among working-age adults, ERR-225. USDA. July 2017.
2. Food security in the U.S. United States Department of Agriculture Economic Research Service. https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the- us/. Updated September 6, 2017. Accessed April 15, 2018.
3. City of Cincinnati community health assessment. City of Cincinnati Health Department. https://www.cincinnatioh.gov/health/assets/File/EDIT%20THIS%20CHA_12_21_17%20 FINAL.pdf. Updated December 21, 2017. Accessed April 15, 2018.
4. Robaina KA, Martin, KS. Food insecurity, poor diet quality, and obesity among food pantry participants in Hartford, CT. J Nutr Educ Behav. 2013;45(2): 159-164.
5. Seligman HK, Lyles C, Marshall MB, Prendergast K et al. A pilot food bank intervention featuring diabetes-appropriate food improved glycemic control among clients in three states. Health Aff. 2015;34(11): 1956-1963.
6. National diabetes statistics report 2017. Centers for Disease Control and Prevention, United States Department of Health and Human Services. https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics- report.pdf2017. Published 2017. Accessed April 15, 2018.
7. Seligman HK, Davis TC, Schillinger D, Wolf MS. Food insecurity is associated with hypoglycemia and poor diabetes self-management in a low-income sample with diabetes. J Health Care Poor Underserved. 2010;21(4): 1227.
8. Murillo R, Reesor LM, Scott CW, Hernandez DC. Food Insecurity and Pre-diabetes in Adults: Race/Ethnic and Sex Differences. Am J Health Behav. 2017;41(4): 428-436.
9. Mayer VL, McDonough K, Seligman H, Mitra M, Long JA. Food insecurity, coping strategies and glucose control in low-income patients with diabetes. Public Health Nutr. 2016;19(6): 1103-1111.
10. Lyles CR, Wolf MS, Schillinger D et al. Food insecurity in relation to changes in hemoglobin A1c, self-efficacy, and fruit/vegetable intake during a diabetes educational intervention. Diabetes Care. 2013;36(6): 1448-1453.
11. Ippolito MM, Lyles CR, Prendergast K, Marshall MB, Waxman E, Seligman HK. Food insecurity and diabetes self-management among food pantry clients. Public Health Nutr. 2017;20(1): 183-189.
47 12. Berkowitz SA, Baggett TP, Wexler DJ, Huskey KW, Wee CC. Food insecurity and metabolic control among US adults with diabetes. Diabetes Care. 2013; 36(10): 3093- 3099.
13. Gholami A, Khazaee-Pool M, Rezaee N et al. Household food insecurity is associated with health-related quality of life in rural type 2 diabetic patients. Arch Iran Med. 2017; 20(6).
14. Grintsova O, Maier W, Mielck A. Inequalities in health care among patients with type 2 diabetes by individual socio-economic status and regional deprivation: a systematic literature review. J Equity Health. 2014;13(1): 43.
15. Vivian EM, Ejebe, IH. Identifying knowledge deficits of food insecure patients with diabetes. Curr Diabetes Rev. 2014;10(6): 402-406.
16. Wilson N. When the cupboards are bare: Nudging food pantry clients to healthier foods. J Consum Res. 2016;1(1): 125-133.
17. Client choice food pantries. End Hunger in America. https://www.endhungerinamerica.org/getting-started/client-choice-food-pantries/. Updated 2018. Accessed April 15, 2018.
18. Inman JJ, Winer RS, & Ferraro R. The interplay among category characteristics, customer characteristics, and customer activities on in-store decision making. J Mark. 2009; 73(5): 19-29.
19. Kahneman D. Thinking, fast and slow. New York, NY: Straus and Giroux; 2011.
20. Wansink, B. Change their choice! Changing behavior using the CAN approach and activism research. P&M. 2015;32(5): 486-500.
21. Moore LV, Pinard CA, Yaroch AL. Features in grocery stores that motivate shoppers to buy healthier foods. J Community Health. 2016;41(4): 812-817.
22. Lancaster G, Massingham L. Markets and customers: consumer and organizational buyer behavior and marketing strategy. Ess Mark Man. 2017; 43-69.
23. McCabe S, Arndt J, Goldenberg JL et al. The effect of visualizing healthy eaters and mortality reminders on nutritious grocery purchases: An integrative terror management and prototype willingness analysis. Health psychol. 2015;34(3): 279.
24. Wright SA, da Costa Hernandez JM, Sundar A, Dinsmore J, Kardes FR. If it tastes bad it must be good: Consumer naïve theories and the marketing placebo effect. Int J Res Mark. 2013;30(2): 197-198.
48 25. Bargh JA, Chartrand TL. The mind in the middle: a practical guide to priming and automaticity research. Handbook of Research Methods in Social and Personality Psychology. New York: Cambridge University Press. 2000; 253–285.
26. Papies EK, Potjes I, Keesman M, Schwinghammer S, Van Koningsbruggen GM. Using health primes to reduce unhealthy snack purchases among overweight consumers in a grocery store. Int J Obes. 2014;38(4): 597.
27. Thorndike AN, Sonnenberg L, Riis J, Barraclough S, Levy DE. A 2-phase labeling and choice architecture intervention to improve healthy food and beverage choices. Am J Public Health. 2012;102(3): 527-533.
28. Levy DE, Riis J, Sonnenberg LM, Barraclough SJ, Thorndike AN. Food Choices of Minority and Low- Income Employees: A Cafeteria Intervention. Am J Prev Med. 2012;43 (3), 240–48.
29. Sonnenberg L, Gelsomin E, Levy DE, Riis J, Barraclough S, Thorndike AN. A traffic light food labeling intervention increases consumer awareness of health and healthy choices at the point-of-purchase. Prev Med. 2013;57(4): 253-257.
30. Payne CR, Niculescu M, Just DR, Kelly MP. Shopper marketing nutrition interventions. Physiol Behav. 2014;136: 111-120.
31. Wilson NL, Just DR, Swigert J, Wansink B. Food pantry selection solutions: a randomized controlled trial in client-choice food pantries to nudge clients to targeted foods. J Public Health. 2016;39(2): 366-372.
32. Hartwig K, Calleson D, Wallace M. Unit 1: Community-Based Participatory Research: Getting Grounded. In: Developing and Sustaining Community-Based Participatory Research Partnerships: A Skill-Building Curriculum. 2006.
33. Evert AL, Boucher JL, Cypress M, Dunbar SA, Franz MJ, Mayer-Davis EJ, Neumiller JJ, Nwankwo R, Verdi CL, Urbanski P, Yancy WS. Nutrition Therapy Recommendations for the Management of Adults with Diabetes. American Diabetes Association. Diabets Care: Jan 2014: 37(1) S120-S143. http://care.diabetesjournals.org/content/37/Supplement_1/S120
34. U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015–2020 Dietary Guidelines for Americans. 8th Edition. December 2015. Available at http://health.gov/dietaryguidelines/2015/guidelines/.
35. Feeding America. Feeding America’s Detailed Foods to Encourage. July 2015. Available at http://hungerandhealth.feedingamerica.org/wpcontent/uploads/legacy/mp/files/tool_ and_resources/files/f2e-background-detail.v1.pdf
49
36. Jewish Family Service of the Cincinnati Area. https://www.jfscinti.org/. Published 2015. Accessed April 23, 2018.
50 Appendix 1. Photos of the “Most Diabetes & Health Friendly” food shelf intervention
51