The translational implications of applying multiple measures to evaluate the nutrient quality of the food supply: a case study of two food pantries in Montana

Abstract

Lower-income populations experience food insecurity, are less likely to meet dietary recommendations, and develop noncommunicable diseases at higher rates than the general U.S. population. Food pantries, which provide food to individuals in need, present an opportunity to decrease these disparities. The purpose of this study was to assess the nutrient quality of the food supply using multiple measures in two food pantry food environments and examine the methodological impactions for translation from research to practice. Nutrient quality of the food supply at two food pantries located in southwest Montana was evaluated using the Healthy Eating Index (HEI) 2015, NOVA classification system, and UnProcessed Pantry Project (UP3) framework every other month during 2018 and 2019. From a total of 63,429 pounds, 291,070 servings, and 32,818 calories of food, processed and ultraprocessed food (UPF) accounted for 57% of servings, 59% of pounds, and 67% of calories. UPF accounts for the highest proportion of food calories compared to its weight. Simutaneously, the food pantries’ food supply had total HEI scores of 77.55 and 79.45 out of a total possible score of 100. NOVA, UP3, and HEI measured multiple aspects that increased the understanding of the nutrient quality of the food supply in two food pantries. A multifaceted approach should be applied, or an all-inclusive tool should be developed, to speed the translation of evidence to practice when assessing and promoting a food supply that limits UPF, increasing the availability of nutritious food and decreasing health disparities for low-income populations.

Implications.

Practice: Multifaceted nutrient assessments of food environments can be used to identify food items that will improve access to healthy food for low-income populations that experience diet and health disparities.

Policy: Policy makers and advocates should rely on multifaceted nutrient assessments of the food supply in order to develop strategies that increase the availability of nutritious food, limit ultraprocessed food, and decrease health disparities for low-income populations in food environments, such as food pantries.

Research: Future research should apply multifaceted nutrient assessments to larger, more diverse sample sizes and in food environments that serve low-income populations in order to provide baseline data to use to improve the food supply.

Introduction

Poor diets are a main contributor to the development of noncommunicable diseases (NCDs), and incidents of NCD have steadily increased globally over the past half century [1]. NCD are now responsible for 41 million deaths annually [1]. Dietary patterns are highly complex and shaped by the interaction of unlimited socioecological factors [2].

The food environment is a prevailing factor that influences the availability, affordability, convenience, and desirability of foods within a food supply [3]. Research suggests that food environments in the USA frequently do not encourage healthy dietary behaviors, but an opportunity does exist for individual and sustained dietary changes through food environment improvements [4]. For example, increasing access to affordable and fresh foods in the food environment may temper the influence that income has on dietary behaviors [2, 3].

The increasing dominance of ultraprocessed food (UPF) in the global food supply promotes unhealthy food environments [4–9] and an escalation in diet-related NCD [6, 8–15]. UPF are foods and beverages high in added sugars, saturated fats, sodium, and additives and low in fiber, vitamins, and minerals [4]. An analysis of the U.S. National Health and Nutrition Survey from 2007 to 2012 demonstrated that approximately 60% of the calories consumed by Americans are UPF, with a significant 1% increase in UPF consumption observed during each 2 year survey cycle [9]. Substantial evidence exists about the relationship between UPF consumption, dietary quality, and health to warrant UPF as a key consideration in the measurement of food patterns [15].

It is known that lower socioeconomic populations experience food insecurity, are more likely to consume UPF, are less likely to meet dietary recommendations, and develop NCD at higher rates than the general U.S. population [9, 16–18]. The toxic combination of low access to healthy, affordable, and nutrient-dense foods and low cost of hyperpalatable UPF across food environments contributes to these disparities [19–21], yet scant studies directly address UPF among lower socioeconomic populations [22].

The food environment of food pantries within the emergency food system provides a promising opportunity to intervene to decrease dietary disparities among lower socioeconomic groups [23]. The emergency food system operates as a sophisticated network of food banks, food pantries, and other programs to provide food for needy households (Fig. 1) [22]. Commonly, food banks store foods that are distributed to local food programs, such as food pantries, soup kitchens, and shelters. Some food pantries do not work with food banks and both store and distribute foods to clients. The food pantry decides strategies to distribute food to clients through a prepacked box or a self-choice shopping model. Typically, the client is given (i.e., prepacked box) or allowed to select (i.e., self-choice shopping) a certain number of foods with food groups based upon household size.

Fig 1.

Food acquisition and distribution in the emergency food system. Original source: [22]. Food banks and food pantries acquire food and beverages through donations and purchases from individuals, food retailers, farmers and processors, national companies and organizations, and federal commodities. Food rescue, or redirecting food that would be wasted, from any of these locations is common. Food banks typically collect, store, and distribute food and beverages for distribution through food pantries to individuals and families in need. Food pantries typically distribute food to clients. Food pantries range in type of organization, capacity, and size from independent nonprofits to affiliations with local organizations, such as faith-based organizations.

Accessed annually by over 40 million Americans and increasing following the onset of the novel coronavirus (COVID-19) pandemic, the emergency food system provides an entry point to offer nutritious food and limit UPF for low-income and food-insecure populations [24, 25]. The nutrient quality of the food supply in the emergency food system varies [22, 26, 27]. Findings across studies suggest that food banks and food pantries tend to rely primarily on nonperishable food to meet the immediate food deprivation needs of clients, and the capacity to provide perishable, nutrient-dense food varies [22, 26–30]. For example, a food retailer may donate excess frozen carrots and, then, at another time, donate a layer cake with frosting. Since donations are typically not nutritionally equivalent, this creates wide variability in clients’ access to food day-to-day.

One recent systematic review assessed the nutritional quality of food bags distributed by food pantries across nine cross-sectional studies and found large variability in adequacy of energy, macronutrients, micronutrients, food groups, and perishability of food items [26]. These studies used dietary analysis software and compared nutrients with the respective country’s dietary guidelines [31–39]. A majority assessed food bags at one to two food pantries over three to four continuous or noncontinuous days. Researchers and practitioners have assessed the food supply available to clients at food pantries by focusing on compliance with food groups established by the U.S. Dietary Guidelines for Americans, including the application of the Healthy Eating Index (HEI), Food Assortment Scoring Tool (FAST), a Food Inventory Log, and Supporting Wellness at Pantries [40–46]. Inventories were conducted at one to six food pantries or food banks for 1 day across several months and up to daily for 1 month. These studies resoundingly found that the food supply is out of compliance with dietary guidance and are limited in fresh food. However, no research assesses the availability of UPF specifically in this vulnerable food environment, although they are anecdotally widely purchased by or donated to food pantries and food banks in the emergency food system because they are shelf stable, easy to store, and require less infrastructure.

With large numbers of households accessing food pantries within the emergency food system annually and health disparities documented, a comprehensive approach to measuring nutrient quality that includes UPF availability is needed. Therefore, this research leverages a case study approach to evaluate the nutrient quality of the food supply using multiple measures at two food pantry sites across a 1 year period in order to examine the methodological implications for translation from research to practice.

METHODS

Measuring the food supply of food pantries

Together, the HEI 2015, NOVA food classification system, and the UnProcessed Pantry Project (UP3) framework were applied to assess the nutrient quality of the food supply at two food pantry sites in the emergency food system across a 1 year period [22, 47–50].

The HEI 2015 uses a scoring system with a scale 0–100, where a score of 100 represents a diet in complete agreement with Dietary Guidelines for Americans [47, 48]. The score is totaled from 13 components that fall into two groups, namely adequacy or moderation. Adequacy is composed of foods that the Dietary Guidelines encourage Americans to consume, including total fruits, whole fruits, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and fatty acids. Higher intake of adequacy foods results in a higher score. Moderation is composed of foods Americans are encouraged to limit, including refined grains, sodium, added sugars, and saturated fats [48]. Lower intake of moderation foods results in a higher score.

The NOVA (not an acronym) classification system divides food into groups based on the level of processing rather than nutrients [21, 49]. This system divides foods into unprocessed or minimally processed, processed culinary ingredients, processed, and ultraprocessed [50, 51].

The UP3 framework expands upon the NOVA classification system to consider processing and key nutrients related to the development on NCDs, including sodium, added sugar, and saturated fat. Food is defined as fresh, pantry staples, lightly prepared (<200 mg sodium, <9 g added sugar, and <1.5 g saturated fat), heavily prepared (≥200 mg sodium, ≥9 g added sugar, and ≥1.5 g saturated fat), and ultraprocessed [46].

Data source

Data were collected from the food supply of two food pantries located in Montana that each offered a self-select shopping model for customers. The food supply was defined as the food inventory on the customer-facing shelves. This unit of analysis was selected as previous research has established that the food available influences the healthfulness of what is selected within the food pantry by a client of a given day [52].

Both food pantries accepted food donations and made purchases of bulk food items from food banks and food retailers. Food Pantry 1 operated within the same location as a food bank. The food bank stored and distributed food to food pantries in the region, including one food pantry in this study. This food bank also operated a food rescue program that collected food items from farmers, bakeries, groceries, and distributors, which increased the pantry’s ability to offer fresh foods. Food Pantry 2 was not affiliated with a food bank. This food pantry purchased approximately 85% of its food with donated funds, which allowed it to procure specific items within each food group, including fresh items, for households and limit the inflow of items higher in sodium, saturated fat, additives, and added sugar. These food pantries were selected because of their diverse storage and distribution models within the emergency food system, engagement in a community advisory board with the research team, involvement in the development and application of the UP3 Framework [22], and agreement to participate in the case study.

Food pantry staff and the research team engaged in an iterative process to develop a protocol that would capture foods typically represented in the inventory of the food supply. To detect seasonal changes in the food supply, each pantry was visited every other month at six time points over a 1 year period between September 2018 and August 2019. With food deliveries occurring on a schedule almost daily, food pantry directors agreed that conducting the inventory during the first 2 weeks of the month on Tuesday through Friday would ensure a consistent representation of the food supply. Each of the food pantries restocked the customer-facing shelves daily to be able to offer households a variety of food items within defined food groups based upon household size during shopping hours. To capture an accurate representation of the variability in the food supply, dates to conduct the food inventory within the agreed upon time frame were randomly selected and agreed upon with food pantry staff. The food pantry staff would ensure that the customer-facing shelves were stocked as usual by the time the food inventory began and were instructed not to make any changes from usual protocol in the food supply. The food inventory was conducted before clients began shopping.

Research staff were trained by the lead author and a research dietitian using a protocol and two practice sessions to assess and record food items in the pantry. The first training session occurred outside of the food pantry and included reviewing the protocol, rating a random selection of food, recording required information, and discussing discrepancies with researchers. The second training session occurred inside of the food pantry and included reviewing the protocol again, rating a random selection of food, recording required information, and establishing interrater reliability with 10 food items and another researcher.

On average, five research staff would attend each inventory and each session lasted 4 hours. For each food inventory item, information was recorded about item name (e.g., reduced fat peanut butter), brand, number of units, servings per unit, total number of servings, and food group (e.g., fruit, vegetable, and grain). NOVA and UP3 classifications were made for each item by reading the item’s list of ingredients and referring to the appropriate classification criteria.

Prepackaged items were organized by brand and subvarieties of brands, then counted and recorded separately. All information on prepackaged food items (e.g., canned vegetables, dry pasta, and yogurt) was recorded. For example, “S&W fresh cut green beans, low sodium” was recorded separately from “S&W French cut green beans, no salt added.” In rare instances when freshly packaged items lacked a nutrition label (e.g., fresh bakery or deli items) or the food item did not list weight or volume, a researcher used a digital scale to record weight and wrote as much detail as possible to match the item to its corresponding food code, described next.

Each food item was matched with its corresponding description and food code in the Food and Nutrient Database for Dietary Studies (FNDDS) and Food Patterns Equivalent Database (FPED). The FNDDS provides nutrient values for foods and beverages reported in What We Eat in America, the dietary intake component of the National Health and Nutrition Examination Survey, while the FPED converts the foods and beverages in the FNDDS to U.S. Department of Agriculture food pattern equivalents [53]. Together, the databases provide the information needed to compute HEI 2015 scores.

Analysis

FNDDS and FPED values are reported per 100 g serving. Therefore, the total number of servings of each food pantry inventory item (as provided in the food pantry inventory data set) was multiplied by the item’s average weight in grams per serving, listed in the FNDDS Portions and Weights data set under “quantity not specified.” These values were then divided by 100 to determine the number of standard 100 g servings per food pantry inventory item. SAS macros provided by the National Cancer Institute were used to compute HEI 2015 scores for each month of data collection using the Simple HEI Scoring Algorithm [54]. Mean scores were assessed by food pantry, month, and stratified by UP3 and NOVA classifications. A two-tailed t-test was used to assess the difference in means between Food Pantry 1 and Food Pantry 2.

Nutrient composition was assessed overall and by month to detect seasonal variation in the percentage of total calories, pounds, and servings of food distributed in each UP3 and NOVA classification category. HEI scores were also assessed for seasonal variation, with mean scores reported overall and by month. The most frequent foods (referred to as top foods) overall and within each UP3, NOVA, and HEI category were computed. Top foods overall were determined by the highest percentage of total calories, pounds, and servings, while top foods within each UP3 and NOVA category were determined by the highest percentage of calories, pounds, and servings within the respective classification. Top foods within each HEI component were reported in cups, ounces, or grams, corresponding to standards used to compute component scores.

RESULTS

In total, the two food pantries provided 63,429 pounds of food, amounting to 291,070 servings and 32,818 calories bimonthly across a 12 month period of data collection (Table 1), of which the majority were processed (25% of servings, 24% of pounds, and 25% of calories) or ultraprocessed (32% of servings, 35% of pounds, and 42% of calories) according to NOVA classifications. Fresh or minimally processed foods accounted for 36% of food servings, 38% of food pounds, and 31% of food calories as categorized by NOVA and UP3. According to the UP3 classification, 18% of servings, 18% of pounds, and 19% of calories were “lightly prepared” and 7% of servings, 6% of pounds, and 5% of calories were “heavily prepared” processed foods. Food pantries distributed 10,307 pounds of food in August (summer), 45% of which was classified as fresh, according to UP3 and NOVA classifications. Comparatively, fresh food only accounted for 28% of the 13,583 pounds of food distributed in April (spring).

Table 1.

Quantity and distribution of food supply at two Montana food pantries, by NOVA and UnProcessed Pantry Project (UP3) classification, September 2018–August 2019

			Distribution by NOVA classificationb				Distribution by UP3 classificationc
Food pantry	Unit	n	1	2	3	4	1	2	3	4	5
Total, overall	Servings	291,070	36%	7%	25%	32%	36%	7%	18%	7%	32%
	Pounds	63,429	38%	2%	24%	35%	38%	2%	18%	6%	35%
	Caloriesa	32,818	31%	2%	25%	42%	31%	2%	19%	5%	42%
Food Pantry 1	Servings	188,588	35%	5%	27%	34%	35%	5%	19%	8%	33%
	Pounds	41,781	36%	2%	25%	36%	36%	2%	19%	6%	36%
	Caloriesa	20,819	28%	3%	27%	42%	28%	3%	22%	6%	42%
Food Pantry 2	Servings	102,359	39%	10%	24%	28%	33%	10%	17%	7%	28%
	Pounds	21,566	42%	1%	22%	34%	36%	1%	16%	6%	34%
	Caloriesa	11,984	37%	2%	20%	40%	42%	2%	15%	5%	40%

^aNumber of calories, expressed in thousands.

^b1: unprocessed/minimally processed; 2: processed culinary ingredients; 3: processed; 4: ultraprocessed.

^c1: fresh, 2: pantry staples; 3: lightly prepared; 4: heavily prepared; 5: ultraprocessed.

The two Montana food pantries had mean total HEI scores of 77.55 and 79.45, respectively, out of a total possible score of 100 across the six data collection points (Table 2). The food pantries received the highest scores in total vegetables, greens and beans, total protein, and seafood and plant protein, achieving maximum possible scores in these components. Scores were lowest in sodium, achieving 1.78 and 2.91, respectively, out of 10 points, and dairy, which had scores of 5.09 and 4.55, respectively, out of 10. Food Pantry 1 had a larger food inventory, distributing significantly more calories than food Pantry 2 (p < .01). While mean total scores were similar between the two pantries (p = .57), component scores differed in refined grains and saturated fat, with Food Pantry 1 receiving a lower score in refined grains and a higher score in saturated fat, on average, compared to Food Pantry 2 (p = .01).

Table 2.

Healthy Eating Index (HEI) scores of food supply at two Montana food pantries

HEI component	Maximum possible score	Food Pantry 1Mean (SD)	Food Pantry 2Mean (SD)	p-valuea
Energy (kcal)		3,469,807.11 (269,996.41)	1,997,410.17 (588,769.92)	<.01
Total vegetables	5	5.00 (0.00)	5.00 (0.00)	n/ab
Greens and beans	5	5.00 (0.00)	5.00 (0.00)	n/ab
Total fruit	5	4.27 (1.14)	3.90 (1.22)	.6
Whole fruit	5	5.00 (0.00)	4.62 (0.92)	.34
Whole grains	10	8.26 (1.84)	7.62 (3.77)	.72
Dairy	10	5.09 (0.68)	4.55 (1.51)	.43
Total protein	5	5.00 (0.00)	5.00 (0.00)	n/ab
Seafood and plant protein	5	5.00 (0.00)	5.00 (0.00)	n/ab
Fatty acid ratio	10	6.92 (2.30)	8.20 (1.98)	.33
Sodium	10	1.78 (2.01)	2.91 (1.78)	.33
Refined grains	10	7.06 (2.34)	9.91 (0.21)	.01
Saturated fat	10	9.22 (0.78)	7.98 (0.61)	.01
Added sugar	10	9.96 (0.11)	9.78 (0.41)	.32
Total HEI Score	100	77.55 (6.55)	79.45 (4.52)	.57

Using number of servings × average serving weight to calculate quantity.

SD standard deviation.

^a  p-values were calculated using a two-tailed t-test.

^b  p-value could not be calculated due to insufficient variation in the data.

Total HEI scores were highest in fresh, unprocessed foods (90.01), while processed and ultraprocessed foods received progressively lower scores of 72.06 and 60.14, respectively, when stratified by NOVA classification. When stratified by UP3 classification, heavily prepared foods received a higher total HEI score (74.35) than lightly prepared processed foods (69.11), while scores in fresh and ultraprocessed foods remained similar to NOVA values (Table 3). Heavily prepared foods received higher HEI component scores than lightly prepared foods in refined grain (10 vs. 2.10 out of 10), total fruit (3.43 vs. 2.06 out of 5) and whole fruit (5 vs. 3.10 out of 5), contributing to the higher total HEI score. UPF received the lowest HEI scores in sodium (0 out of 10), total fruit (1.39 out of 5), and saturated fat (5.05 out of 10). Notably, UPF still received maximum scores in total vegetables (5 out of 5), greens and beans (5 out of 5), total protein (5 out of 5), and seafood and plant protein (5 out of 5), suggesting limitations in the use of HEI as an exclusive indicator of nutrition quality. The food pantries received the lowest total HEI score in January (72.42) and the highest total HEI score in August (82.59).

Table 3.

Healthy Eating Index (HEI) scores of food pantry food supply, stratified by NOVA and UnProcessed Pantry Project (UP3) classifications

	NOVA classificationa				UP3 classificationb
HEI component	1	2	3	4	1	2	3	4	5
Energy (kcal)	10,246,124	788,353	8,128,344	13,655,261	10,246,124	788,353	6,313,790	1,786,302	13,636,272
Total vegetables	5.00	2.62	5.00	5.00	5.00	2.62	5.00	5.00	5.00
Greens and beans	5.00	0.00	5.00	5.00	5.00	0.00	5.00	5.00	5.00
Total fruit	5.00	0.00	2.35	1.39	5.00	0.00	2.06	3.43	1.39
Whole fruit	5.00	0.00	3.64	2.40	5.00	0.00	3.10	5.00	2.41
Whole grains	9.24	0.83	10.00	3.55	9.24	0.83	10.00	5.92	3.55
Dairy	8.15	0.94	2.43	4.22	8.15	0.94	2.54	2.03	4.22
Total protein	5.00	0.56	5.00	5.00	5.00	0.56	5.00	5.00	5.00
Seafood and plant protein	5.00	1.75	5.00	5.00	5.00	1.75	5.00	5.00	5.00
Fatty acid ratio	5.41	7.83	8.62	8.15	5.41	7.83	8.89	7.97	8.15
Sodium	7.29	0.00	0.00	0.00	7.29	0.00	0.42	0.00	0.00
Refined grains	10.00	10.00	5.02	7.04	10.00	10.00	2.10	10.00	7.03
Saturated fat	10.00	0.00	10.00	5.05	10.00	0.00	10.00	10.00	5.03
Added sugar	10.00	0.00	10.00	8.35	10.00	0.00	10.00	10.00	8.34
Total HEI Score	90.09	24.53	72.06	60.14	90.09	24.53	69.11	74.35	60.12

^a1: unprocessed/minimally processed; 2: processed culinary ingredients; 3: processed; 4: ultraprocessed.

^b1: fresh, 2: pantry staples; 3: lightly prepared; 4: heavily prepared; 5: ultraprocessed.

Food pantries were limited in the variety of inventory items (Table 4). The top five distributed food items, as measured in servings, pounds, and calories, contributed to 18.80% of all servings, 22.40% of all pounds, and 27.15% of all calories distributed. When measured in servings, peanut butter (6.42%), onions (3.98%), pasta (3.52%), whole wheat bread (2.56%), and multigrain bread (2.31%) were distributed in the highest quantity. By weight, top foods included canned tomato soup (5.72%), pasta (4.99%), canned chicken noodle soup (4.32%), apples (4.03%), and low-fat milk (3.34%). The majority of calories came from peanut butter (10.90%), pasta (6.87%), macaroni and cheese (4.05%), dry lentils (2.77%), and potatoes (2.56%).

Table 4.

Top food items distributed by two food pantries in Montana, in servings, pounds, and calories

Food item	Quantity	Unit	Percentage of total
Total	291,070	Servings
Peanut butter	18,697	Servings	6.42
Onions, raw	11,594	Servings	3.98
Pasta	10,252	Servings	3.52
Bread, whole wheat	7,460	Servings	2.56
Bread, multigrain	6,726	Servings	2.31
Total	63,429	Pounds
Tomato soup, canned, or ready-to-serve	3,627	Pounds	5.72
Pasta	3,164	Pounds	4.99
Chicken or turkey noodle soup, canned or ready-to-serve	2,741	Pounds	4.32
Apple, raw	2,558	Pounds	4.03
Milk, low fat (1%)	2,117	Pounds	3.34
Total	32,818	Calories (in thousands)
Peanut butter	3,578	Calories (in thousands)	10.90
Pasta	2,253	Calories (in thousands)	6.87
Macaroni or noodles with cheese	1,330	Calories (in thousands)	4.05
Lentils, dry	910	Calories (in thousands)	2.77
Potato, whole, with peel	840	Calories (in thousands)	2.56

This lack of variety becomes more striking when comparing top food items contributing to each HEI scoring component. For example, 70% of total vegetables and greens and beans scoring components come from dry lentils, canned beans (not sure as to type), and canned kidney beans; apples, bananas, and canned cranberry sauce account for 62% of total fruit and 63% of whole fruit; brown rice, whole wheat bread, and multigrain bread account for 68% of all ounces of whole grains; low-fat milk, macaroni and cheese, and fat-free milk account for 54% of dairy; and dry lentils, peanut butter, and canned beans (not sure as to type) account for 72% of both total protein and seafood and plant protein scoring components (Table 5).

Table 5.

Top food items distributed by two food pantries in Montana, by Healthy Eating Index (HEI) score component

HEI score component	Quantity	Unit	%
Total vegetables	5,358,821	Cups
Lentils, dry	2,830,871	Cups	53
Beans, canned, NS as to type	467,232	Cups	9
Red kidney beans, canned	445,094	Cups	8
Greens and beans	5,326,381	Cups
Lentils, dry	2,830,871	Cups	53
Beans, canned	467,232	Cups	9
Red kidney beans, canned	445,094	Cups	8
Total fruit	23,891	Cups
Apple, raw	10,561	Cups	44
Banana, raw	3,054	Cups	13
Cranberries, cooked or canned	1,167	Cups	5
Whole fruit	22,530	Cups
Apple, raw	10,561	Cups	47
Banana, raw	3,054	Cups	14
Cranberries, cooked or canned	1,167	Cups	5
Whole grains	42,327	Ounces
Rice, brown	12,313	Ounces	29
Bread, whole wheat	9,749	Ounces	23
Bread, multigrain	6,610	Ounces	16
Dairy	21,002	Cups
Milk, low fat (1%)	3,937	Cups	19
Macaroni or noodles with cheese	3,722	Cups	18
Milk, fat free (skim)	3,597	Cups	17
Total protein	30,260,411	Ounces
Lentils, dry	11,373,150	Ounces	38
Peanut butter	8,605,599	Ounces	28
Beans, canned, NS as to type	1,877,126	Ounces	6
Seafood and plant protein	30,195,769	Ounces
Lentils, dry	11,373,150	Ounces	38
Peanut butter	8,605,599	Ounces	28
Beans, canned, NS as to type	1,877,126	Ounces	6
Fatty acid ratio (mono and poly unsaturated fat/saturated fat)
Mono and poly unsaturated fat	694,260	Grams
Peanut butter	230,203	Grams	33
Lentils, dry	30,597	Grams	4
Macaroni or noodles with cheese	23,646	Grams	3
Saturated fat	319,517	Grams
Peanut butter	61,775	Grams	19
Macaroni or noodles with cheese	31,687	Grams	10
Ground beef, raw	10,659	Grams	3
Sodium	60,723,547
Chicken or turkey noodle soup, canned or ready-to-serve	4,301,230	Grams	7
Pasta	3,329,850	Grams	5
Tomato soup, canned, or ready-to-serve	3,175,236	Grams	5
Refined grains	64,239
Pasta	20,094	Ounces	31
Macaroni or noodles with cheese	6,741	Ounces	10
Cornbread muffin	3,152	Ounces	5
Added sugar	113,497
Cranberries, cooked or canned	16,336	Grams	14
Peanut butter	8,197	Grams	7
Tomato soup, canned, or ready-to-serve	6,745	Grams	6

When assessing food variety by UP3 and NOVA, onions, apples, and dry lentils contribute to 20% of all fresh and minimally processed food servings; apples, potatoes (whole, with peel), and fat-free milk contribute to 26% of all fresh and minimally processed food pounds; and brown rice, dry lentils, and potatoes (whole, with peel) contribute to 29% of all fresh and minimally processed food calories. According to NOVA classification, pasta, multigrain bread, and whole wheat bread accounted for 26% of all processed food servings; pasta, whole grain pasta, and tuna (canned, water pack) accounted for 31% of total processed food pounds; and pasta, wholegrain bread, and multigrain bread accounted for 41% of all processed food calories (Table 6). When assessing lightly and heavily prepared processed foods by UP3 classification, pasta, multigrain bread, and whole wheat bread account for 36% of all lightly prepared food servings; pasta, wholegrain pasta, and canned tomatoes account for 40% of all lightly prepared food pounds; and pasta, whole grain pasta, and multigrain bread account for 52% of all lightly prepared food calories. In contrast, canned tuna, canned green peas, and canned green beans account for 38% of heavily prepared food servings; canned tuna, canned pear, and canned green beans account for 37% of all heavily prepared food pounds; and canned tuna, peanut butter, and pork (not sure as to type) account for 32% of all heavily prepared food calories. Peanut butter, canned tomato soup, and canned chicken noodle soup accounted for 29% of UPF calories; canned tomato soup, canned chicken noodle soup, and macaroni and cheese accounted for 34% of all UPF pounds; and peanut butter, macaroni and cheese, and canned tomato soup accounted for 39% of all UPF calories according to both UP3 and NOVA classifications (Table 7). Peanut butter was a top food item in both heavily prepared and UPF categories, depending on the specific brand’s ingredients and degree of processing.

Table 6.

Top food items distributed by two food pantries in Montana, in servings, pounds, and calories, stratified by NOVA classification

Minimally processed				Culinary ingredients				Processed				Ultraprocessed
Food item	Quantity	Unit	%	Food item	Quantity	Unit	%	Food item	Quantity	Unit	%	Food item	Quantity	Unit	%
Total	105,657	Servings		Total	18,996	Servings		Total	74,200	Servings		Total	92,217	Servings
Onions, raw	8,066	Servings	8	Sugar, cinnamon	3,528	Servings	19	Pasta	9,556	Servings	13	Peanut butter	17,738	Servings	19
Apple, raw	6,374	Servings	6	Sugar, NFS	2,847	Servings	15	Bread, multigrain	5,556	Servings	7	Tomato soup, canned, or ready-to-serve	5,283	Servings	6
Lentils, dry	6,128	Servings	6	Vegetable oil, NFS	2,691	Servings	14	Bread, whole wheat	4,600	Servings	6	Chicken or turkey noodle soup, canned or ready-to-serve	4,120	Servings	4
Potato, whole, with peel	5,310	Servings	5	Butter, stick, salted	2,072	Servings	11	Tuna, canned, water pack	4,523	Servings	6	Bread, whole wheat	2,785	Servings	3
Egg, whole, raw	4,734	Servings	4	Biscuit mix	1,260	Servings	7	Pasta, whole grain	3,252	Servings	4	Macaroni or noodles with cheese	2,599	Servings	3
Total	24,371	Pounds		Total	1,296	Pounds		Total	15,442	Pounds		Total	22,320	Pounds
Apple, raw	2,558	Pounds	10	Chicken or turkey broth, bouillon, or consommé	790	Pounds	61	Pasta	2,949	Pounds	19	Tomato soup, canned, or ready-to-serve	3,494	Pounds	16
Rice, brown	2,233	Pounds	9	Biscuit mix	83	Pounds	6	Pasta, whole grain	1,004	Pounds	7	Chicken or turkey noodle soup, canned or ready-to-serve	2,734	Pounds	12
Potato, whole, with peel	1,990	Pounds	8	Sugar, cinnamon	62	Pounds	5	Tuna, canned, water pack	848	Pounds	5	Macaroni or noodles with cheese	1,318	Pounds	6
Milk, fat free (skim)	1,926	Pounds	8	Vegetable oil, NFS	53	Pounds	4	Tomatoes, from canned, stewed	628	Pounds	4	Peanut butter	1,251	Pounds	6
Milk, low fat (1%)	1,736	Pounds	7	Sugar, NFS	50	Pounds	4	Peas, green, canned	511	Pounds	3	Chicken or turkey rice soup, canned, or ready-to-serve	689	Pounds	3
Total	10,246	Caloriesa		Total	788	Caloriesa		Total	8,128	Caloriesa		Total	13,655	Caloriesa
Rice, brown	1,236	Caloriesa	12	Vegetable oil, NFS	215	Caloriesa	27	Pasta	2,100	Caloriesa	26	Peanut butter	3,394	Caloriesa	25
Lentils, dry	910	Caloriesa	9	Biscuit mix	122	Caloriesa	16	Pasta, whole grain	674	Caloriesa	8	Macaroni or noodles with cheese	1,321	Caloriesa	10
Potato, whole, with peel	840	Caloriesa	8	Sugar, cinnamon	106	Caloriesa	14	Bread, multigrain	530	Caloriesa	7	Tomato soup, canned, or ready-to-serve	523	Caloriesa	4
Apple, raw	603	Caloriesa	6	Butter, stick, salted	104	Caloriesa	13	Bread, whole wheat	417	Caloriesa	5	Cranberries, cooked or canned	407	Caloriesa	3
Banana, raw	406	Caloriesa	4	Sugar, NFS	88	Caloriesa	11	Tuna, canned, water pack	331	Caloriesa	4	Cornbread muffin	315	Caloriesa	2

^aCalories expressed in thousands

Table 7.

Top food items distributed by two food pantries in Montana, in servings, pounds, and calories, stratified by UnProcessed Pantry Project (UP3) classification

(1) Fresh				(2) Pantry staples				(3) Lightly prepared
Food item	Quantity	Unit	%	Food item	Quantity	Unit	%	Food item	Quantity	Unit	%
Total	1,05,657	Servings		Total	18,996	Servings		Total	52,367	Servings
Onions, raw	8,066	Servings	8	Sugar, cinnamon	3,528	Servings	19	Pasta	9,556	Servings	18
Apple, raw	6,374	Servings	6	Sugar, NFS	2,847	Servings	15	Bread, multigrain	5,396	Servings	10
Lentils, dry	6,128	Servings	6	Vegetable oil, NFS	2,691	Servings	14	Bread, whole wheat	3,976	Servings	8
Potato, whole, with peel	5,310	Servings	5	Butter, stick, salted	2,072	Servings	11	Tomatoes, canned	1,928	Servings	4
Egg, whole, raw	4,734	Servings	4	Biscuit mix	1,260	Servings	7	Plum, rock salt, dried	1,674	Servings	3
Total	24,371	Pounds		Total	1,296	Pounds		Total	11,514	Pounds
Apple, raw	2,558	Pounds	10	Chicken or turkey broth, bouillon, or consommé	790	Pounds	61	Pasta	2,949	Pounds	26
Potato, whole, with peel	1990	Pounds	8	Biscuit mix	83	Pounds	6	Pasta, whole grain	1,004	Pounds	9
Milk, fat free (skim)	1,926	Pounds	8	Sugar, cinnamon	62	Pounds	5	Tomatoes, canned	544	Pounds	5
Milk, low fat (1%)	1,736	Pounds	7	Vegetable oil, NFS	53	Pounds	4	Milk, dry, reconstituted, low fat (1%)	508	Pounds	4
Rice, brown	1,432	Pounds	6	Sugar, NFS	50	Pounds	4	Bread, multigrain	428	Pounds	4
Total	10,246	Caloriesa		Total	788	Caloriesa		Total	6,314	Caloriesa
Rice, brown	1,236	Caloriesa	12	Vegetable oil, NFS	215	Caloriesa	27	Pasta	2,100	Caloriesa	33
Lentils, dry	910	Caloriesa	9	Biscuit mix	122	Caloriesa	16	Pasta, whole grain	674	Caloriesa	11
Potato, whole, with peel	840	Caloriesa	8	Sugar, cinnamon	106	Caloriesa	14	Bread, multigrain	515	Caloriesa	8
Apple, raw	603	Caloriesa	6	Butter, stick, salted	104	Caloriesa	13	Bread, whole wheat	361	Caloriesa	6
Banana, raw	406	Caloriesa	4	Sugar, NFS	88	Caloriesa	11	Red kidney beans, canned	98	Caloriesa	2
(4) Heavily prepared				(5) Ultraprocessed
Food item	Quantity	Unit	%	Food item	Quantity	Unit	%
Total	21,765	Servings		Total	91,966	Servings
Tuna, canned, water pack	4,730	Servings	22	Peanut butter	17,738	Servings	19
Peas, green, canned	1,434	Servings	7	Tomato soup, canned, or ready-to-serve	5,283	Servings	6
Beans, string, green, canned	1,999	Servings	9	Chicken or turkey noodle soup, canned or ready-to-serve	4,120	Servings	4
Pear, cooked or canned, in light syrup	1,017	Servings	5	Bread, whole wheat	2,785	Servings	3
Corn, yellow, canned	853	Servings	4	Macaroni or noodles with cheese	2,599	Servings	3
Total	3,923	Pounds		Total	22,273	Pounds
Tuna, canned, water pack	886	Pounds	23	Tomato soup, canned, or ready-to-serve	3,494	Pounds	16
Pear, cooked or canned, in light syrup	280	Pounds	7	Chicken or turkey noodle soup, canned or ready-to-serve	2,734	Pounds	12
(4) Heavily prepared				(5) Ultraprocessed
Food item	Quantity	Unit	%	Food item	Quantity	Unit	%
Peas, green, canned	269	Pounds	7	Macaroni or noodles with cheese	1,318	Pounds	6
Beans, string, green, canned	339	Pounds	9	Peanut butter	1,251	Pounds	6
Corn, yellow, canned	154	Pounds	4	Chicken or turkey rice soup, canned, or ready-to-serve	689	Pounds	3
Total	1,786	Caloriesa		Total	13,636	Caloriesa
Tuna, canned, water pack	346	Caloriesa	19	Peanut butter	3,394	Caloriesa	25
Peanut butter	147	Caloriesa	8	Macaroni or noodles with cheese	1,321	Caloriesa	10
Pork, NFS	97	Caloriesa	5	Tomato soup, canned, or ready-to-serve	523	Caloriesa	4
Peas, green, canned	83	Caloriesa	5	Cranberries, cooked or canned	407	Caloriesa	3
Pear, cooked or canned, in light syrup	72	Caloriesa	4	Cornbread muffin	315	Caloriesa	2

^aCalories expressed in thousands.

Discussion

The results underscore the need to gain a comprehensive understanding of foods available in order to improve nutrient quality for the populations served in a specific food environment. Our study found the majority of the food supply in the two food pantries to be processed food or UPF. Two thirds of all calories came from processed food or UPF compared to just a third coming from fresh food or pantry staple items. This finding is critical because food pantries and their clients are particularly vulnerable to UPF because they are shelf stable and inexpensive, but they are also typically high in added sugars, saturated fats, sodium, and additives and low in fiber, vitamins, and minerals [4, 21].

Using the UP3 classification system allowed us to further tease the difference between lightly prepared and heavily prepared food, which showed that most processed foods were lightly prepared. This distinction is important since processed foods vary greatly in their nutrition and healthfulness. For example, a lightly prepared natural peanut butter with no salt or sugar added, while no longer in its original unprocessed nut form, offers significant health benefits compared to peanut butter that is heavily prepared with salt, added sugar, and added oils.

We expected heavily prepared food as classified by the UP3 framework to also receive lower HEI scores than lightly prepared foods. In our study, both lightly and heavily prepared foods received close to 0 (0.42 and 0.00, respectively) out of the maximum 10 points for sodium content, and both received the maximum 10 points for low levels of saturated fat and added sugar. UP3 classifies foods as lightly or heavily prepared using specific sodium, saturated fat, and added sugar cutoff points. This contradiction suggests a limitation in the HEI scoring system as it lacks the sensitivity to detect nutrition differences in processed ingredients, which can make a large difference in the healthfulness of an individual’s diet. Overall, when classified by NOVA standards, processed food and UPF had lower HEI scores than fresh or minimally processed food. Still, the relatively high total HEI scores among UPF and inconsistencies in HEI component scores suggest a need for a broader and more comprehensive measurement of nutrient quality across food environment settings, especially those that serve populations vulnerable to UPF.

Overall, HEI scores reported in our study are higher than the national average score of 59 for Americans, although this score represents food consumed by Americans and not the food available within the food supply as this study investigated [55]. The food pantries in this study also had higher HEI scores, on average, than those reported in a 2015 study on nutrition quality among food pantries in Minnesota, which reported a mean score of 62.7 [43, 56]. The HEI scores were relatively high, even with a large percentage of UPF in the food supply. The food pantries in this study make bulk purchases of less-processed food items, such as rice, canned tomatoes, and canned beans to ensure that food pantry shelves were stocked with healthful staple foods. Food Pantry 1 benefited from a food rescue program that collected 1,187,777 pounds of food from groceries, bakeries, farmers, and distributors between July 2018 and June 2019 [56], which allows for increased availability and variety of fresh foods, like fruits and vegetables, to pantry customers. Food Pantry 2 benefited from a majority of their food supply being purchased using monetary donations, which provided the ability to procure specific items within desired food groups and nutrient guidelines. The higher HEI scores highlight that these food pantries’ focused procurement strategies paid off in ensuring access to nutritious foods that align with national dietary guidance but that the consideration of UPF ingredients is warranted. For example, whole wheat bread with high fructose corn syrup contributes to a positive whole grain component score for HEI, but this positive score does not provide discernible guidance toward whole grain breads without added sugar.

The food pantries lacked diversity in their inventory. The top three food items contributing to HEI total vegetable scoring component were all various forms of legumes, accounting for 70% of all total vegetable servings. The food pantries also displayed inconsistencies in fresh food availability throughout the year. For example, access to fresh foods was more of a challenge for clients in April than in August. Healthy diets should ideally be diverse and varied, consisting of a wide range of colors and nutrients [57]. Clients with food sensitivities may be particularly impacted by this lack of diversity when trying to meet their daily nutrition needs. An estimated 8.0% of children and 10.0% of adults are reported to have food allergies in the USA, with milk and peanuts cited as top sources of allergy [58–60]. Yet, peanut butter was the top source of total food servings and calories distributed by the food pantries, suggesting a need for more variety and options, especially in alternatives to common allergy-inducing foods.

By using diverse units of measurement captured by three assessment tools, our study offers a wholistic and nuanced examination of the food pantry food environment. Quantifying food in pounds, servings, and calories allowed us to show how UPF make up a higher proportion of the total calories than the total weight of food pantry distribution. These findings follow the logic that fresh, unprocessed foods often contain higher water content and are less calorically dense than UPF. Similarly, using multiple measurement units allowed us to more meaningfully assess the proportion of pantry staples. Pantry staples, such as spices and cooking oils, are typically used in small amounts and, therefore, contain many servings for their calories and weight. Pantry staples are an important component for cooking food, which contributes to a healthy diet [61].

Future research should focus on developing a simple tool that assesses alignment with dietary guidance and is sensitive to detect slight differences in the nutrient composition, and UPF in particular, between similar products. For example, such a measure could be applied to understand the implications of COVID-19 on the food supply of food pantries. Furthermore, the translation of such research data should be carefully considered in the context of the time, skill, and behaviors required for the end user. For example, in a food pantry, many individuals face time and skill barriers to preparing minimally processed food, and a diversity of lightly prepared food should be available to also fulfill the guidelines for a healthy diet.

Two food pantries were recruited to participate in a case study of the food supply over a 1 year period. These two food pantries may be uncharacteristic of all food pantries given their geographic location, desire to participate in the case study, and focus on procuring a nutrient-dense food supply. Even with a small sample size, the research achieved its purpose to demonstrate that assessing the nutrient quality of the food supply using multiple measures has significant implications for translating research data to practical changes in the food available within food pantries. Future work should focus on applying these multifaceted measures to a larger number of food pantries and food banks in diverse locations to develop best practices for procuring a nutritious food supply across food pantries. Expanding this case study work to a larger sample size will demonstrate if programmatic and policy decisions can be made to improve the food supply in food environments for health-disparate populations based upon such multifaceted tools.

Conclusions

Using NOVA, UP3, and HEI allows us to compare the strengths and weaknesses of each assessment tool while providing three unique measurements of the food environment. Assessing food item quantities in pounds, servings and calories, and food quality by NOVA, UP3, and HEI standards, therefore, adds to the body of literature by providing a better understanding of the nutrient composition of this unique food environment, which is vulnerable to self-stable and easy to distribute UPF. The evaluation herein emphasizes distinctions of nutrient quality and can be applied by researchers, administrators, and policy makers and advocates of food pantry programs and in other food environments to assess and promote a food supply that increases the availability of nutritious food, limits UPF, and decreases health disparities for low-income populations.

Funding:

This study was funded by the National Institute of General Medical Sciences of the National Institutes of Health (NIH) under P20GM104417. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Compliance with Ethical Standards

Conflicts of Interest: The authors declare that they have no conflicts of interest.

Author Contributions: CBS conceived the study and developed the method; CBS, LL, BW collected the data; CBS and EW wrote the first draft of the manuscript; EW conducted the data analysis; all authors contributed to draft iterations and revisions; all authors approved the final manuscript.

Ethical Approval: This article does not contain any studies with human participants performed by any of the authors.

Informed Consent: No informed consent was required as human participants were not engaged in the research.