Sage Journals: Discover world-class research

Abstract

The persistence of food insecurity leads millions of Americans to seek assistance from hunger relief programs including charitable organizations such as food banks. A greater understanding of the relationships between diet and disease has led to discussions on the role of the food banking industry to source more nutritious foods. However, an empirical process for how nutritional quality can be applied in a food bank setting has not been well defined. The aim of this study is to use the Nutrient-Rich Food Index (NRFI) to establish a set of cut-point values that score and categorize the nutrient density of foods, while developing an application with visual appeal and function. NRFI scores were calculated using nutrition information available for 8,751 foods recorded in the U.S. Department of Agriculture (USDA) Standard Reference Database. Tertile cut points (>26; 5.5-26.0; <5.5) were established and visualized using a three-tier color-coded approach (green, yellow, and red). The intention of this approach was to help food sourcing managers at food banks make evidence-based, informed decisions when considering nutrient density in food procurement. Furthermore, scoring and categorization can help track inventory overtime and create a data mechanism for strategic planning.

Keywords

food insecurity hunger and health nutrient density food bank

Introduction

Healthy People 2020 objectives are focused on reducing rates of household food insecurity by 6% and eliminating childhood hunger (U.S. Department of Health and Human Services [USDHHS], Office of Disease Prevention and Health Promotion, 2014). Despite the abatement of household food insecurity levels in recent years, the rate of reduction may not be robust enough to meet the 2020 target. Today, nearly 42 million Americans report living in a household that is food insecure (Coleman-Jensen, Rabbitt, Gregory, & Singh, 2017).

Food insecure households may access a number of federally funded nutrition assistance programs including the Supplemental Nutrition Assistance Program (SNAP); Women, Infants, and Children’s (WIC) program; among others. However, these programs have strict eligibility guidelines. For families who are food insecure and do not meet eligibility criteria, charitable food assistance programs including food banks and food pantries provide food to those in need. Forty-six and a half million people each year depend on the Feeding America network of 200 food banks and 46,000 partner agencies for hunger relief assistance (Weinfield et al., 2014). Individuals who access network resources are not only hungry, but many are chronically sick. A recent study conducted by Feeding America identified that nearly 30% of network households reported someone in the home with diabetes and nearly 60% reported a family member with hypertension (Weinfield et al., 2014).

Food insecurity and limited food access are directly linked to poor dietary patterns (Leung, Epel, Ritchie, Crawford, & Laraia, 2014), which can perpetuate chronic diseases. With a greater understanding of the connections between diet, disease, and the impact on clients in the food banking network, food banks have prioritized the acquisition of foods that are of higher nutritional quality (Campbell, Ross, & Webb, 2013). Yet, organizational capacities can limit the ability to source these foods. One gap in organizational capacity which has been identified is the absence of an empirical approach to define nutritional quality. As a result, food banks use common sense when determining the nutritional quality of foods (Campbell et al., 2013).

Community-Based Participatory Research (CBPR) provides a framework to draw on the extensive knowledge and expertise from food banks, food systems, and their client needs in partnership with university faculty. CBPR includes ongoing partnerships which allow for multiple iterations and refinement of assessment, research, dissemination, and action to address a local problem (Israel et al., 2003). Within this framework, the Blue Ridge Area Food Bank (BRAFB) and faculty engaged in a long-term commitment to improve the nutritional quality of foods provided to those facing food insecurity. Today, the BRAFB serves 106,000 people each month through an expansive network of 200 food pantries across a 25-county service spanning central and western Virginia. Our partnership was galvanized by an established set of guiding principles built upon the notion that solutions must be grounded with empirical evidence and operationalized within the organizational structure and culture. In the development and maintenance of the partnership, the BRAFB and faculty iterative process have involved developing an evidence-based tool to inform the procurement of high-quality, nutritious foods.

The 2015 Dietary Guidelines for Americans (2015) highlights nutritional density when defining dietary quality. Nutritionally dense foods and beverages include “vitamins, minerals, and other substances that contribute to adequate nutrient intakes or may have positive health effects, with little or no solid fats and added sugars, refined starches, and sodium” (USDHHS & U.S. Department of Agriculture [USDA], Office of Disease Prevention and Health Promotion, 2015). Validated approaches to quantify the nutritional density of foods have been defined in the literature (Fulgoni, Keast, & Drewnoswski, 2009; Nanney et al., 2016). The Nutrient-Rich Food Index (NRFI; Fulgoni et al., 2009) is based on the nutritional composition per calorie of nine nutrients to encourage and three nutrients to limit (Figure 1). Additional variants of the NRFI9.3, referred to as the family of NRFI, have also been validated with fewer nutrients to encourage. In a previous publication, Seidel, Laquatra, Woods, and Sharrard (2015) published an approach that utilized the NRFI6.3 to rank the nutritional quality of foods in the inventory of The Greater Pittsburgh Area Food Bank (GPAFB).

Figure 1.

U.S. FDA (USDHHS, FDA, 2018) Nutrition Facts Label (NFL) with NRFI nutrients highlighted, nutrients to limit in red and nutrients to encourage in green.

Although the approach was well characterized, cut-points to establish parameters for their rating system were not provided (Seidel et al., 2015). The aim of this study was to establish a set of tertile cut-points that were used to develop parameters for a food scoring system. As the NRFI produces a numeric score for each food item, BRAFB desired cut-points for more intuitive decision making and to establish a metric for benchmarking and to set goals. Furthermore, this article describes a visual, color-coded approach for a data application that is based on established food labeling structures to increase accessibility and allow for a wider audience. The intention is to utilize these tools to assist food sourcing managers in making evidence-based, informed decisions during food procurement at BRAFB.

Method

Two data sets were analyzed separately. First, a database of food and beverage items (n = 464) acquired and distributed by BRAFB during the 2015-2016 fiscal year was developed (Sample 1). Items purchased through the commodity food program and the food bank purchasing program (n = 255) and a random sample of donated items (n = 209) were combined to create a representative sample of food bank items. Nutrition information was captured through a combination of data entry using product food labels and queries from the U.S. Department of Agriculture (USDA) Food Composition Databases (USDA, Agricultural Research Service, 2018). Second, food items from the USDA National Nutrition Database for Standard Reference Release 28 (2016; N = 8,751) were used to evaluate a more expansive, generalizable set of food items (Sample 2).

For each food item, the NRFI6.3 was used and included six nutrients to encourage (protein, fiber, vitamin A, vitamin C, calcium, and iron) and three nutrients to limit (saturated fat, total sugar, and sodium). These nutrients represent the information currently available on food labels. For each nutrient to encourage, a ratio of 100 g to the Reference Daily Value was calculated; the maximum ratio must not exceed 1 for any single nutrient. The sum of these six ratios was multiplied by 100, then divided by the energy density of the item, and then multiplied by 100 again. Energy density is defined as kcal/100 g. A similar algorithm was used for the nutrients to limit. A ratio of the nutrient content per 100 g to the maximum recommended value was calculated; no upper value was established for these ratios. The sum of these three ratios was multiplied by 100, then divided by the energy density of the item, and then multiplied by 100 again. To create the NRFI6.3, the nutrient to limit score was subtracted from the nutrients to encourage score (Figure 2).

Figure 2.

A visual representation of the NRFI6.3 Algorithm, validated by Fulgoni et al. (2009).

After excluding items with one or more missing values for these nine nutrients and items with 0 kcal/100 g from further analysis, 6,098 items remained (69.4%) in Sample 2. Although the NRFI6.3 may exaggerate the nutritional value of items with very low energy density, the exclusion of items with less than 5 kcal/100 g did not change the tertiles. In each sample, three categories of scores were established based upon tertiles.

Results

Within Sample 1 from the food bank, NRFI6.3 scores ranged from −91.20 to 719.17. The cut-point values for the categories were 9.50 and 30.0 (Table 1). The median of scores was 13.9.

Table 1.

NRFI6.3 Tertile Values of Food Bank Inventory and USDA Foods (Expressed With a Color-Coded Traffic Light System as a Categorical Approach: Red, Yellow, and Green).

Value	Red	Yellow	Green
Food bank (Sample 1)	<9.50	9.50-30.0	>30.0
N = 464	N = 154	N = 156	N = 154
USDA (Sample 2)	< 5.5	5.5-26.0	>26.0
N = 6,098	N = 2,033	N = 2,032	N = 2,033

Note. NRFI = Nutrient-Rich Food Index; USDA = U.S. Department of Agriculture.

Within Sample 2 from the USDA, NRFI6.3 scores ranged from –1,081.2 to 6,451.7. The median of scores was 15.6. Just over one-quarter of items, 25.1%, had nutrients to limit that exceeded nutrients to encourage. The cut-points for categories were 5.5 and 26 (Table 1).

Discussion

This study compared two sets (food bank vs. USDA) of cut-point values to establish a three-tier categorical system using a previously validated metric of nutrient density, the NRFI6.3 (Fulgoni et al., 2009). Although values and ranges for each category were similar between the groups, the USDA National Nutrition Database for Standard Reference Release 28 (USDA, Agricultural Research Service, 2016) offered a larger data set to establish cut-point values. The sequential use of the separate samples in the analysis illustrates the iterative processes and capacity building that characterize CBPR (Israel, Eng, Schulz, & Parker, 2005). The initial sample of items represented only a small subset of potential foods. Expansion to the second sample addressed the need to have cut-points that could be generalized beyond the food bank inventory and capture a wider range of food items and types.

To place greater contextualization around the cut-points established by the NRFI6.3, we also adapted a color-coded, traffic light system as a categorical approach: green, yellow, and red. The traffic light design translates the research into a solution that is fully operational and preferred by BRAFB; this balance of research with functional capacity demonstrates another characteristic of CBPR (Israel et al., 2005). Cut-points expressed in a traffic light design allows BRAFB’s food sourcing manager to make time-sensitive decisions and establish a metric for benchmarking and to set goals. The visual analogy of the traffic light is widely adopted as a mechanism to communicate the nutritional quality of food. A recent meta-analysis found that traffic light approaches are most effective with helping consumers make healthy choices compared to other methods evaluated (Cecchini & Warin, 2016). Furthermore, the USDHHS, National Institutes of Health’s Eat Play Grow curriculum utilizes the stoplight approach to reinforce messages on healthy eating (USDHHS, National Heart, Lung and Blood Institute, 2013).

Relative to their calories, foods in the green category (foods to encourage) have a greater proportion of nutrients to encourage; the yellow category (foods to advise) includes foods that contain some nutrients to encourage and nutrients to limit; and the red category (foods to limit) includes foods that have a greater proportion of nutrients to limit. With the NRFI, food scores and traffic light categories can be compared across and within food groups or food category. Figure 3 demonstrates the evaluation of food items across food categories. In this example, corn flakes rank the highest, green, as this food has a greater proportion of total nutrients to encourage (fiber, protein, iron, calcium, and vitamin A), while remaining low in nutrients to limit. In this example, a canned peach product ranks in the yellow category. Although fruits and vegetables are considered nutritionally dense foods, the composition of canned items can vary, and lower scores are typically driven by higher amounts of sodium and sugar added during food processing. Figure 4 further illustrates this concept as we show a comparison of three different green bean products. In these examples, nutrients to encourage have little variability across product, so the difference in traffic light category is largely driven by the difference in sodium content.

Figure 3.

Visual hierarchy samples incorporating NRFI Scoring for foods with color coding for quick reference to reinforce decision making (based on the established visual patterns of FDA NFL labeling).

Figure 4.

Visual hierarchy samples incorporating NRFI Scoring for foods within a food category with color-coding for quick reference to reinforce decision making (based on the established visual patterns of FDA NFL labeling).

Finally, the visual presentation of nutrition information was adapted to conform to a contextual hierarchy, based on the U.S. Food and Drug Administration’s (FDA) Nutrition Facts Labeling (NFL; USDHHS, FDA, 2018) system. This allows users to easily recognize and prioritize individual nutrients when a direct comparison is required (Figure 2). For BRAFB, the product labeling and NFL information is often the easiest way to identify food items and verify if the item has been previously purchased.

Implications

In recent years, conversations regarding the responsibility of food banks in providing food to the nation’s most vulnerable have shifted to emphasize nutritional quality. Food banks recognize the importance of this shift (Campbell et al., 2012) and are responding. To provide guidance to their network of food banks, Feeding America has proposed an evolving set of recommendations, called “Foods to Encourage” (F2E), and has previously published, to the network of food banks, a second revision (Feeding America, 2015). This approach requires foods to be sorted into categories and then “scored” using parameters set around saturated fat, sodium and sugar, and for some foods, fiber. This approach overlooks a paramount principal of the Dietary Guidelines for Americans, nutrient density, which emphasizes the consumption of foods with healthful nutrients, including vitamins, minerals, and fiber (USDHHS & USDA, Office of Disease Prevention and Health Promotion, 2015).

The family of NRFI algorithms is built around the concept of nutrient density and as a validated metric to score the nutritional quality of foods, and this approach has been applied in other food banking organizations. Previously, colleagues from the GPAFB (Seidel et al., 2015) described the application of the NRFI6.3 for food sourcing. They concluded that food sourcing managers can use the NFRI6.3 to be proactive with the acquisition of nutritious foods, while creating a mechanism to track inventory, in an evidence-based way (Seidel et al., 2015). Despite these valuable insights, cut-point parameters in which GPAFB is used to establish food categories were not published. This inspired our efforts to generate and publish a series of cut-point parameters. Subsequently, a three-tier categorization, green, yellow, and red, was applied to establish greater contextualization around the cut-points to assist with more intuitive decision making. This coupled with the hierarchical presentation of the information mimicking the NFL provides a consistent “look and feel,” so nutrition information is easily recognized.

At this time, quantitative metrics to distinguish nutritional quality has limited use across the network of America’s food banks (Campbell et al., 2012) and presents an untapped opportunity to develop systemic approaches that would streamline the ability to keep nutrition in the forefront of decision making during food procurement (Shimada, Ross, Campbell, & Webb, 2013). In fact, the application of nutritional quality can be critical as food banks consider their strategic goals, consider how to assess their impact, (Shimada et al., 2013) and generate data that may be helpful in establishing organizational policies.

Considering the empirical nature of our approach, the BRAFB and faculty are in the process of piloting a user-friendly solution called Nourish. The solution was designed to remedy the ambiguous use of common sense to decide nutritional quality by operationalizing the NRFI6.3 and apply our categorization and visual approach. The system utilizes the USDA Food Composition Databases (USDA, Agricultural Research Service, 2018) to obtain nutritional information on procured foods, assists in real-time decision making, and creates connections with the inventory management system to track progress and create mechanisms for reporting. Our system aligns with recent recommendations published by Feldman and Schwartz (2018) indicating that Food Banks should consider implementing a nutrition-tracking system to monitor and assess the nutritional quality of foods.

One limitation to our work is the lack of available data for added sugars on nutrition facts labels and in the USDA Food Composition Databases (USDA, Agricultural Research Service, 2018). The use of total sugar in the NRFI6.3 rather than added sugar may overestimate the negative component of this nutrient compared to using only added sugar. It is anticipated that updates to the NFL, to include information on added sugar, should occur in 2020. When this information becomes ubiquitously available, these data point should replace total sugars in the NRFI algorithm.

Conclusion

The ongoing partnership between BRAFB and university faculty have facilitated a collaborative and equitable process to problem solving. The partnership was guided by two principles: solutions developed had to be grounded in evidence and have the capacity to be operational. Multiple iterations of the project were developed with constant input and assessment. The resulting work delivers an empirically based scoring and categorizing system that easily evaluates the dietary quality of foods in food banks. This project also fits within the larger context of what it means to engage our society with supporting healthy eating patterns for all, a key recommendation of the 2015-2020 Dietary Guidelines for Americans (USDHHS & USDA, Office of Disease Prevention and Health Promotion, 2015). Food Banks, as community-based advocates, can act as one sector to influence the dietary choices of Americans. University faculty can work in partnership with food banks to provide solutions that help guide the creation of strong policies and strategic visions grounded in scientific evidence which may strengthen that influence.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Michelle Battista Hesse graduated with a PhD in Human Nutrition from The Ohio State University in 2010 and completed her dietetic internship and passed the Registered Dietitian’s examination in 2011. Before joining the Food Bank in 2018, Hesse was an assistant professor at James Madison University where she worked collaboratively with academic and community colleagues to develop operational solutions for food banks to procure more nutritious foods. She also taught several courses for the dietetics program.

Andrew Peachey holds a PH in Community Health Education and MA in Applied Geography from University of North Carolina at Greensboro. Peachey is the Academic Unit Head (acting) for the department of Health Sciences at James Madison University. Peachey’s research interests include exploring the built enviornment and its relationship to health.

David Wang is an assistant professor in the School of Media Arts and Design at James Madison University. Wang received a masters in Fine Arts and Communications from Louisiana Tech University. With more than 20 years of industry experience, Wang has worked as an art director, photographer and interactive developer in advertising, communications and higher education.

References

Campbell

E. C.

Ross

Webb

K. L.

(2013). Improving the nutritional quality of emergency food: A study of food bank organizational culture, capacity and practices. Journal of Hunger and Environmental Nutrition, 8, 261-280.

Cecchini

Warin

(2016). Impact of food labelling systems on food choices and eating behaviours: A systematic review and meta-analysis of randomized studies. Obesity Reviews, 17, 201-210.

Coleman-Jensen

Rabbitt

M. P.

Gregory

C. A.

Singh

(2017). Household food security in the United States in 2016. U.S. Department of Agriculture. Retrieved from https://www.ers.usda.gov/publications/pub-details/?pubid=84972

Feeding America. (2015). Feeding America’s foods to encourage background. Retrieved from http://hungerandhealth.feedingamerica.org/wp-content/uploads/legacy/mp/files/tool_and_resources/files/f2e-background-detail.v1.pdf

Feldman

Schwartz

M. B.

(2018). A tipping point: Leveraging opportunities to improve the nutritional quality of food bank inventory. Mazon. Retrieved from https://mazon.org/assets/download-files/MAZON-TippingPointReport-FINAL.pdf

Fulgoni

V. L.

III Keast

D. R.

Drewnoswski

(2009). Development and validation of the Nutrient-Rich Foods Index: A tool to measure nutritional quality of foods. Journal of Nutrition, 8, 1549-1554.

Israel

B. A.

Eng

Schulz

A. J.

Parker

E. A.

(2005). Introduction to methods in community-based participatory research for health. In Israel

B. A.

Eng

Schulz

A. J.

Parker

E. A.

(Eds.), Methods in community-based participatory research for health (pp. 3-29). San Francisco, CA: Jossey-Bass.

Israel

B. A.

Schulz

A. J.

Parker

E. A.

Becker

A. B.

Allen

A. J.

Guzman

J. R.

(2003). Critical issues in developing and following community based participatory research principles. In Minkler

Wallerstein

(Eds.), Community based participatory research for health (pp. 53-76). San Francisco, CA: Jossey-Bass.

Leung

C. W.

Epel

E. S.

Ritchie

L. D.

Crawford

P. B.

Laraia

B. A.

(2014). Food insecurity is inversely associated with diet quality of lower-income adults. Journal of the Academy of Nutrition and Dietetics, 114, 1943-1953.

10.

Nanney

M. S.

Grannon

K. Y.

Cureton

Hoolihan

Janowiec

Wang

King

R. P.

(2016). Application of the Healthy Eating Index-2010 to the hunger relief system. Public Health Nutrition, 19, 2906-2914.

11.

Seidel

Laquatra

Woods

Sharrard

(2015). Applying a Nutrient-Rich Foods Index algorithm to address nutrient content of food bank food. Journal of the Academy of Nutrition and Dietetics, 115, 695-700.

12.

Shimada

Ross

Campbell

E. C.

Webb

K. L.

(2013). A model to drive research-based policy change: Improving the nutritional quality of emergency food. Journal of Hunger and Environmental Nutrition, 8, 281-293.

13.

U.S. Department of Agriculture, Agricultural Research Service. (2016). USDA national nutrient database for standard reference release 28. Retrieved from https://www.ars.usda.gov/northeast-area/beltsville-md-bhnrc/beltsville-human-nutrition-research-center/nutrient-data-laboratory/docs/usda-national-nutrient-database-for-standard-reference/

14.

U.S. Department of Agriculture, Agricultural Research Service. (2018). United States Department of Agriculture food composition databases. Retrieved from https://ndb.nal.usda.gov/ndb/

15.

U.S. Department of Health and Human Services, Food and Drug Administration. (2018). Labeling and nutrition. Retrieved from https://www.fda.gov/Food/LabelingNutrition/ucm274593.htm#overview

16.

U.S. Department of Health and Human Services, National Heart, Lung and Blood Institute. (2013). GO, SLOW, WHOA early childhood health lesson. Retrieved from https://www.nhlbi.nih.gov/health/educational/wecan/tools-resources/eatplaygrow-gsw.htm

17.

U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion. (2014). Healthy people 2020. Washington, DC: Author.

18.

U.S. Department of Health and Human Services, & U.S. Department of Agriculture, Office of Disease Prevention and Health Promotion. (2015). 2015–2020 dietary guidelines for Americans (8th ed.). Retrieved from https://health.gov/dietaryguidelines/2015/

19.

Weinfield

N. S.

Mills

Borger

Gearing

Macaluso

Montaquila

Zedlewski

(2014). Hunger in America 2014: National report. Feeding America. Retrieved from http://help.feedingamerica.org/HungerInAmerica/hunger-in-america-2014-full-report.pdf

Establishing a Three-Tier Color-Coded Approach to Categorize the Nutrient Density of Food Bank Foods

Abstract

Keywords

Introduction

Method

Results

Discussion

Implications

Conclusion

Footnotes

Declaration of Conflicting Interests

Funding

Author Biographies

References