The development and evaluation of a novel measure of discretionary food intake
==============================================================================
* Amara Channell Doig
* Leah M. Lipsky
* Allison Choe
* Tonja R. Nansel
## Abstract
**Background** While dietary guidelines recommend limiting foods high in added sugars, saturated fat, refined grains, and sodium, and all alcoholic beverages, there are no available methods for classifying discretionary foods or quantifying discretionary food intake using dietary intake data.
**Objective** To develop and evaluate a nutrient-based method to classify discretionary foods and compare with two established methods for classifying foods to limit.
**Design** Foods in the 2017-2018 USDA Food and Nutrient Database for Dietary Studies (FNDDS, n=6909) were classified as discretionary, hyperpalatable, and ultra-processed, and their nutrient values were compared. Correlations of discretionary, hyperpalatable, and ultra-processed food intakes with overall diet quality were evaluated using 2017-2018 National Health and Nutrition Examination Survey (NHANES) data.
**Participants** All non-pregnant NHANES participants age 2+ years with two valid 24-hour recalls (n=6136) were included.
**Main outcome measures** Outcomes included food-level nutrient density (Nutrient Rich Foods 9.3 index) and person-level Healthy Eating Index-2020 scores (HEI total, HEI-adequacy, and HEI-moderation).
**Statistical Analyses** Differences in nutrient density of FNDDS foods classified as discretionary, hyperpalatable, and ultra-processed vs. not were compared using t-tests and ANOVA. Fisher’s z-transformation compared associations of HEI scores with intake (percent kcal) from discretionary, hyperpalatable, and ultra-processed foods.
**Results** The difference in nutrient density (p<0.001) of discretionary vs. non-discretionary foods (mean diff=74.8, 95%CI:70.6-78.9) was larger than that of ultra-processed vs. non-ultra-processed (mean diff=29.7, 95%CI:25.1-34.2) and hyperpalatable vs. non-hyperpalatable (mean diff=53.2, 95%CI:49.3-57.1). Discretionary foods contributed 77% of energy intake in U.S. children and adults, while hyperpalatable foods and ultra-processed foods contributed 71% and 58%, respectively. Inverse associations of HEI-2020 total (r=-0.72) and HEI-adequacy scores (r=-0.67) with discretionary food intake were stronger than those with hyperpalatable (r=-0.40; - 0.23) and ultra-processed food intake (r=-0.49;-0.43); associations with HEI-moderation scores were similar across all classifications.
**Conclusions** The discretionary food classification method effectively distinguishes between low-and high-nutrient dense foods and is strongly associated with adherence to dietary guidelines. The method may improve diet quality assessment and inform public health interventions.
## Introduction
The 2020 – 2025 Dietary Guidelines for Americans (DGA) recommend limiting intake of food and beverages high in added sugars, saturated fat, refined grains, and sodium, along with limiting alcoholic beverages1 due to associations with adverse health outcomes.2–12 While the DGA does not provide a term for these foods, the Australian Dietary Guidelines describes these foods as “discretionary.”13 These foods are often energy-dense, contain high amounts of nutrients to limit, and displace intake of nutrient-rich foods needed to meet nutritional requirements (i.e., “core foods”).13 However, the DGA do not provide specific nutrient-based standards for identifying discretionary foods. Given that people select and consume foods, not nutrients, establishing nutrient-based criteria for classifying discretionary foods would facilitate accessible, food-based dietary guidance, policy, and intervention targets to improve diet quality.14
Current methods for evaluating diet or food quality do not enable classifying individual foods as discretionary. For example, the Healthy Eating Index (HEI) 2020, a measure of alignment of overall diet quality with the DGA, includes component scores for total intakes of refined grains, sodium, added sugars, saturated fat relative to total intake, which lower the total HEI score.15 The Nutrient Rich Food Index and the Food Compass methods score individual foods on continuous scales based on nutrient content, but do not provide guidance on specific cut-off values for discriminating between discretionary and non-discretionary foods.16–18 While the Australian Guide to Healthy Eating discretionary food definition has been used to assess intake in multiple studies,19–23 it does not provide a quantitative definition that can be applied across all food groups.24
Methods have been developed to designate foods as ultra-processed or hyperpalatable, although these attributes are distinct from discretionary foods. The NOVA method classifies foods according to their level of industrial processing.25 Foods are classified as either minimally processed, processed, culinary ingredients, or ultra-processed; the proportion of energy intake from ultra-processed foods has been used as the primary dietary exposure in hundreds of analyses.26 While there is overlap between foods that are ultra-processed and nutrient-poor,27,28 numerous ultra-processed foods are nutrient-dense (e.g., packaged whole grain breads, low-sugar cereal), and several foods not classified as ultra-processed are nutrient-poor (e.g., homemade pastries).27,28 Additionally, although intake of ultra-processed foods is typically associated with lower diet quality,29,30 diets high in ultra-processed foods can also be of high diet quality.31
The recently-developed Hyperpalatable Food Score classifies foods based on thresholds of nutrient combinations thought to increase palatability and activate the brain’s reward system (i.e., fat plus sodium, fat plus simple sugars, carbohydrates plus sodium),32 as determined from a overview of food reward survey measures and experimental studies that examined a small number of foods thought to be hyperpalatable (e.g. fast foods or sweets).32 However, the measure classifies several nutrient-dense foods as hyperpalatable (e.g. chives, arugula), while classifying several nutrient-poor foods as non-hyperpalatable (e.g. fried vegetables, gummy bears, high-sugar fruit snacks), which may lead to misclassification if used to quantify intake of foods to limit.32
To address current limitations in assessing discretionary food intake, this study develops and evaluates a nutrient-based method for classifying discretionary foods in food databases and dietary intake data. Identifying discretionary foods at the food-level, rather than the nutrient-level, allows for direct translation to dietary guidance, intervention targets, and public health initiatives.
## Methods
### Data sources
#### FNDDS
The 2017-2018 USDA Food and Nutrient Database for Dietary Studies (FNDDS)33 was used to evaluate nutrient thresholds for classifying individual foods as discretionary. This dataset contains nutrient values for 7083 individual foods and beverages consumed in the United States. The accompanying Food Patterns Equivalents Database (FPED)34 converts the foods and beverages to their corresponding USDA food patterns components (i.e., teaspoon equivalents of added sugars, gram equivalents of saturated fat, and cup equivalents of grains).
#### NHANES
The 2017-2018 National Health and Nutrition Examination Survey (NHANES)35,36 was used to evaluate the discretionary food score (percent energy intake from discretionary foods) with respect to person-level diet quality. NHANES is a cross-sectional, nationally representative survey of health and nutrition in the United States. The National Center for Health Statistics Ethics Review Board approved the study,37 and all participants provide written consent.36 The dietary intake portion of the survey, What We Eat in America (WWEIA), collects 24-hour dietary recalls from individuals of all ages using the interviewer-administered five-step USDA Automated Multiple-Pass Method.36,38 Dietary recalls are proxy-reported for children 5 years and younger and proxy-assisted for children ages 6-11 years. The sample used for this analysis includes all participants aged two years and older who had two reliable dietary recalls from the 2017-2018 NHANES cycle. Of 6502 participants with two reliable dietary recalls, pregnant individuals (n=44) and participants under two years of age (n=322) were excluded. The final sample included 6136 participants.
### Development of the Discretionary Food Score
Thresholds for defining discretionary foods were based on scoring standards for the Healthy Eating Index – 2020 (HEI),39 the 2020-2025 Dietary Guidelines for Americans (DGA),1 the Food and Drug Administration (FDA) consumer guidance,40,41 and the 2021 American Heart Association (AHA) Guidance for Cardiovascular Health Promotion and Disease Reduction.42 The primary dietary components considered included added sugars, saturated and total fats, sodium, refined grains, and alcohol, based on guidance from the DGA and AHA to limit intakes of these components. Initial cut-points were evaluated for face validity with respect to food categorizations using the FNDDS, and modifications were applied as appropriate. Since the discretionary food criteria are based on dietary recommendations for people over age two years, infant formula and human milk were excluded from these analyses.
***Face validity.*** Face validity of initial and revised criteria was assessed by examining how well they classified foods consistently referred to in the literature as discretionary or non-discretionary. For example, foods referred to as discretionary, junk, or extra in the literature include sweets, salty snacks, sugar sweetened beverages, high sugar cereal, processed meats, fried potatoes, and high-fat fast food (e.g. pizza, chicken nuggets).19,43–52 Foods that are referred to as nutrient-dense in the DGA include “vegetables, fruits, whole grains, seafood, beans, peas, and lentils, unsalted nuts and seeds, eggs, fat-free and low-fat dairy products, and lean meats and poultry” that are prepared with minimal amounts of added sugars, saturated fats, and sodium.1 Criteria that classified these key foods accordingly were considered to have strong face validity. ***Added Sugars.*** Since individual foods that modestly exceed the DGA limit for individual nutrients may still contribute to a healthful diet in balance with overall intake, the threshold for added sugars used to classify discretionary foods was set to ≥20%, which is twice the maximum recommended daily intake of added sugars (10% of total energy intake).1
***Fats.*** Using the same rationale as for added sugars, the initial criterion for saturated fat was set to ≥20% of kcal, which is twice the DGA maximum recommended intake of saturated fat in the overall diet.1 Implementation of this criterion resulted in misclassifications in the vegetables and snacks and sweets food groups; several vegetables cooked with a small amount of butter (e.g. broccoli, collards, or carrots cooked with butter) met the criterion due to the inherently low energy content of vegetables, whereas fried vegetables (e.g., fries, onion rings, vegetable fritters) and several high-fat snacks and sweets (e.g., potato chips, doughnuts) were classified as non-discretionary since they contained primarily unsaturated fat. To address these weaknesses, the saturated fat criterion was not applied to the vegetable food group, and several thresholds of total fat were evaluated for classifying fried vegetables, high-fat snacks, and sweets as discretionary. A threshold of ≥9% total fat by weight aligned most closely with expected classifications, and was, therefore, applied to these food groups. The threshold of ≥20% kcal from saturated fat was applied to all other food groups.
***Sodium.*** An initial threshold of ≥2.0 grams of sodium per 1,000 kcal was evaluated, based on the value corresponding to the minimum score of zero for the HEI-2020 sodium component.39 However, several vegetables cooked with a small amount of salt met this criterion due to their low energy content. To allow for a small amount of sodium in low-calorie foods, the threshold was revised to align with the FDA definition of a high-sodium food (460 mg per serving), which corresponds to 20% of the recommended maximum daily intake (2300 mg per 2000 kcal).40,41 The food weight for “quantity not specified” in the FNDDS Portions and Weights33 dataset was used to indicate portion sizes of individual foods. For foods with multiple subcodes, the mean of the portion weights for all subcodes was applied.
***Refined Grains.*** The initial threshold of >50% of total grains from refined grains, based on the DGA criterion that half of grains consumed should be whole grains,1 was selected for classifying foods as discretionary. However, this criterion resulted in misclassification of several grain-based foods because FNDDS classifies sprouted grain bread, corn tortillas, and other foods made from masa as refined grains, which resulted in their classification as discretionary. This conflicts with guidance by WIC,53 National School Lunch Program,54 and Child and Adult Care Food Program,55 which defines these foods as whole grains. Therefore, an additional threshold based on the American Heart Association whole grain definition42 was added, such that foods with a carbohydrate to fiber ratio of ≥10:1 and >50% of total grains from refined grains were classified as discretionary. Additionally, to avoid misclassifying mixed dishes containing only a small amount of refined grains (e.g., soup thickened with cornstarch), the grain criteria was applied only to foods containing at least 0.5 ounce-equivalents of total grains.
***Alcohol.*** Based on the DGA guidance to limit alcohol intake,1 all alcoholic beverages were classified as discretionary.
***Final Criteria.*** The final criteria for classifying discretionary foods are presented in **Figure 1**. Any food meeting at least one of the thresholds is classified as discretionary. In addition to the dichotomous discretionary classification system, an ordinal classification was created by summing the number of thresholds met.
[Figure 1.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/F1)
Figure 1.
Criteria for classification of foods as discretionary
Finally, based on the DGA core elements of a healthy diet1 and HEI-2020 adequacy components, it was determined that no whole plant foods (i.e. whole fruit, vegetables, whole grains, legumes, nuts, or seeds) should be classified as discretionary. Only three food items
(Brazil nuts, fresh coconut, and avocado [raw or for use on a sandwich]) were misclassified (due to their naturally occurring fat); these were recoded manually as non-discretionary.
### Evaluation of the discretionary food classification and score
The discretionary food classification was evaluated at the food level by examining nutrient density in discretionary versus non-discretionary foods. At both the food and person level, the discretionary food classification method was compared with that of two alternative classification methods used to identify foods to limit – ultra-processed foods and hyperpalatable foods. These differ conceptually from, but have substantial overlap with, discretionary foods. Since the discretionary food criteria were specifically intended to identify nutrient-poor foods, it should demonstrate greater precision in identifying discretionary foods and stronger associations with diet quality and nutrient density than the other two classification methods. The discretionary food score was also evaluated at the person level by examining associations of discretionary food intake with diet quality.35,36
***Ultra-processed Foods***. Ultra-processed foods were classified using a linkage database created by researchers at the University de Sao Paulo and the NCI,56,57 which assigns one of the four NOVA groups to the disaggregated (i.e. ingredient) food codes and calculates the corresponding weight and energy for each group. To enable direct comparisons with the discretionary food classification method, foods classified at the ingredient level rather than the food level were considered ultra-processed if ultra-processed ingredients comprised >50% of the total food weight. Person-level intake of ultra-processed foods (NOVA group 4) was calculated as the percent of daily energy intake from ultra-processed food for each participant in NHANES. ***Hyperpalatable foods.*** The hyperpalatable food measure defines foods meeting any of the following criteria as hyperpalatable: >25% kcal from fat and ≥0.30% sodium by weight, >20% kcal from fat and >20% sugar, or >40% kcal from carbohydrates and ≥0.20% sodium by weight.32 These criteria were based on nutrient compositions that were thought to increase palatability, selected by evaluating the nutrient content of foods described as hyper-palatable in the literature.32 The person-level percent of daily energy intake from hyperpalatable foods was calculated for each participant in NHANES.
***Diet quality.*** HEI-202039,58 total scores (range 0-100), HEI-2020 adequacy scores (range 0-60), and HEI-2020 moderation scores (range 0-40) were calculated for each participant in NHANES using the simple scoring algorithm.59 The adequacy score is calculated as the sum of 9 adequacy components (total fruit, whole fruits, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafoods and plant proteins, and fatty acids), and the moderation score is calculated as the sum of the 4 moderation components (refined grains, sodium, added sugars, and saturated fats). The HEI-2020 total score is calculated as the sum of 13 components (the 9 adequacy components and 4 moderation components). Higher scores represent closer alignment with the DGA.39
***Nutrient density.*** Nutrient density for food items in the FNDSS was calculated using the Nutrient Rich Foods Index (NRF 9.3), which is scored as a sum of the percent of the daily recommended values of nutrients to encourage (protein, fiber, vitamin A, vitamin C, Vitamin E, calcium, iron, magnesium, and potassium) minus the percent of the maximum recommended daily value for nutrients to limit (saturated fat, added sugars, and sodium). Higher scores indicate higher nutrient density.60
## Statistical Analysis
The statistical programming code used to create the discretionary food classification and score is provided in supplemental materials. All analyses were completed in SAS 9.4,61 and statistical significance was set at p<0.05.
***FNDDS.*** To compare food classification methods at the food level, descriptive statistics for the number of discretionary, hyperpalatable, and ultra-processed foods, and the energy, added sugars, refined grains, saturated fat, and sodium were calculated using FNDDS. The percent of foods in each FNDDS food group classified as discretionary, hyperpalatable, and ultra-processed were plotted on a radar chart to visualize differences in classification. Differences in nutrient density of discretionary vs. non-discretionary, hyperpalatable vs. non-hyperpalatable, and ultra-processed vs. non-ultra-processed foods were compared using t-tests; additionally, ANOVA was used to compare mean differences in food-level nutrient density by food classification method, using Bonferroni correction to account for multiple comparisons.
***NHANES.*** The discretionary food score was evaluated at the person level using the NHANES dietary data. Descriptive statistics were used to summarize sociodemographic characteristics and the percent of daily energy (kcal), added sugars (g), and sodium (mg) from discretionary, hyperpalatable, and ultra-processed foods. Additionally, the correlations of percent of daily energy intake (kcal) from discretionary, hyperpalatable, and ultra-processed foods with HEI-2020, HEI-2020 moderation, and HEI-2020 adequacy scores were examined using Pearson’s ρ.
Differences in the magnitude of the correlations were assessed using the fisher’s z-transformation. Design-based statistical methods for all NHANES analyses accounted for the complex, multistage sampling design62 and were weighted using the dietary recall day 2 sample weights.
## Results
### Face validity of the Discretionary Food Classification
Of 6909 foods and beverages in the FNDDS (excluding formula and human milk), 4555 (66%) were classified as discretionary based on the selected nutrient thresholds. Of these, 60% met one threshold, 33% met two thresholds, and 7% met three or more thresholds. Sodium was the most common threshold classifying discretionary foods (47%), followed by refined grains (41%), added sugars (24%), saturated fat (20%), and percent fat by weight (16%). The number (%) of foods classified as discretionary for each FNDDS food group and WWEIA food category description are presented in **Table 1**. Vegetables contained the smallest proportion of discretionary foods (18%), while snacks contained the largest proportion of discretionary foods (97%). The fruit and beverage food groups had the smallest proportion of foods (<1%) that met two nutrient thresholds, while sweets (50%) had the highest. Only snacks (4%), mixed dishes (6%), and sweets (28%) had more than 2% of foods that met three or more thresholds. Within the WWEIA categories, all milkshakes, frankfurters, pizza, nachos, potato chips, cookies and brownies, doughnuts, and soft drinks were classified as discretionary. Conversely, none of the low-fat or non-fat milk, grapes, melon, broccoli, spinach, string beans, or tomatoes were classified as discretionary.
View this table:
[Table 1.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/T1)
Table 1. WWEIA category description and n (%) of foods classified as discretionary and meeting each discretionary criterion by food group
### Proportion of discretionary foods, ultra-processed foods, and hyperpalatable foods in FNDDS
The proportion of foods classified as >50% ultra-processed (32%) and hyperpalatable (62%) in the FNDDS were smaller than those classified as discretionary (66%). Twenty percent of foods met the criteria for all three classification methods, 6% were classified as both discretionary and >50% ultra-processed, 29% were classified as both discretionary and hyperpalatable, and 11% were classified only as discretionary (**Figure 2A**).
[Figure 2.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/F2)
Figure 2.
**A**) Venn diagram of the overlap between foods that meet the criteria to be classified as discretionary, hyperpalatable, and >50% ultra-processed. **B**) Radar plot showing the percent of foods classified as discretionary, >50% ultra-processed, and hyperpalatable by FNDDS food group.
The percent of FNDDS foods classified as discretionary, hyperpalatable, and >50% ultra-processed by food group are displayed in **Figure 2B**. In each food group, more foods were classified as discretionary than ultra-processed. There was a smaller proportion of discretionary than hyperpalatable foods within the protein, grains, and vegetable groups, but a larger proportion of discretionary than hyperpalatable foods within the sugars, fats, snacks, and sweets groups.
The percent of FNDDS foods in each food group and category that were classified as discretionary, hyperpalatable, and ultra-processed are presented in **Table 2**. In the milk group, a substantially greater proportion of foods were classified as discretionary than ultra-processed, including most cheese, flavored milk, and milkshakes. Fewer differences were observed between discretionary and hyperpalatable classifications; however, a greater proportion of low-fat (53% vs 7%) and nonfat flavored milks (53% vs 0%) were classified as discretionary than hyperpalatable. Conversely, a greater proportion of reduced fat milks (not flavored) were classified as hyperpalatable than as discretionary (83% vs 17%). The discretionary classification method identified a greater proportion of protein foods than the ultra-processed classification, but a smaller proportion than the hyperpalatable classification. For example, more bacon (85% vs 8%), cold cuts (91% vs 51%), and ground beef (80% vs 0%) were classified as discretionary than ultra-processed. Conversely, a smaller proportion of legumes (13% vs 38%), beef (excluding ground) (68% vs 91%), chicken (whole pieces) (36% vs 94%), and nuts and seeds (5% vs 44%) were classified as discretionary than as hyperpalatable. Within the vegetable food group, a greater proportion of French fries (95% vs 75% or 64% respectively), fried vegetables (97% vs 73% or 20%), and coleslaw (58% vs 42% or 0%) were classified as discretionary than hyperpalatable or ultra-processed. Conversely, in nearly all the other vegetable subgroups, a greater proportion of foods were classified as hyperpalatable than the other two methods. For example, vegetables that met only the hyperpalatable criterion included baked sweet potato, canned lima beans, and cooked cabbage, carrots, chard, pumpkin, and kohlrabi. A greater proportion of fruits were classified as discretionary than the other two classification methods, including high sugar canned fruits, fried fruits (e.g. banana chips), and chocolate covered fruits. In snacks, sweets, and sugars food groups, a greater proportion of foods were classified as discretionary than as ultra-processed or hyperpalatable.
View this table:
[Table 2.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/T2)
Table 2. Heatmap showing percent of foods classified as discretionary, ultra-processed, and hyperpalatable by food category. The colors range from green (low percentage) to red (high percentage).
### Nutrient Density
Differences in nutrient density by classification method are depicted in **Figure 3A**. Nutrient density was significantly different between foods meeting criteria for discretionary, ultra-processed, and hyperpalatable as compared with those that did not. The mean difference in nutrient density by discretionary foods classification (mean diff: 74.8, 95% CI: 70.6-78.9) was significantly greater than by either hyperpalatable (mean diff: 53.2, 95% CI:49.3-57.1) or ultra-processed (mean diff: 29.7, 95% CI: 25.1-34.2) classification (p<0.001). Mean nutrient density decreased linearly with increasing number of discretionary thresholds met, although there was no significant difference between those meeting two versus three thresholds (**Figure 3B**). Foods classified as 25-50% ultra-processed had a lower nutrient density than those that were 0-25% ultra-processed, but did not differ from those that were greater than 50% ultra-processed (**Figure 3C**). A comparison of the nutrient and energy content for discretionary, hyperpalatable, and ultra-processed foods is presented in **Table 3** in the supplemental materials.
[Figure 3.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/F3)
Figure 3.
**A**) Nutrient density according to food classification as discretionary, hyperpalatable, and >50% ultra-processed versus not. **B**) Nutrient density according to number of discretionary thresholds met. **C)** Nutrient density according to percent of food that is ultra-processed.
### Evaluation of the association with diet quality in U.S adults and children ages 2 years and older
Sample characteristics are presented in **Table 4**. Intake of discretionary foods contributed on average 77% of total energy, whereas intake of hyperpalatable foods contributed 71% and ultra-processed foods contributed 58%. Foods meeting only one discretionary food criterion contributed 42% of total energy intake, while foods meeting two criteria contributed 26% of energy intake, and foods meeting three or more criteria contributed 9% of energy intake.
View this table:
[Table 4.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/T3)
Table 4. Sample Characteristics
Discretionary foods contributed 94% of total added sugars, while hyperpalatable foods contributed 49%, and ultra-processed foods contributed 82%. Discretionary foods contributed 78% of total sodium intake, whereas hyperpalatable foods contributed 81% and ultra-processed foods contributed 56%.
Intake (% kcal) from discretionary, hyperpalatable, and ultra-processed foods was negatively correlated with HEI-2020 total, HEI-moderation, and HEI-adequacy scores (**Table 5**).
View this table:
[Table 5.](http://medrxiv.org/content/early/2025/03/20/2025.03.20.25324317/T4)
Table 5. Pearson Correlation of intake (% kcal) of Discretionary, Hyperpalatable, and Ultra-processed foods with Healthy Eating Index scores.
The magnitudes of the negative correlations of HEI-total and HEI-adequacy scores with discretionary foods were larger than those with hyperpalatable and ultra-processed foods (p-difference<0.0001). The magnitude of the negative correlation of HEI-moderation scores with discretionary food intake was larger than that with hyperpalatable and ultra-processed intake (p-difference <0.0001).
## Discussion
In this study, a new method for classifying discretionary foods demonstrated high face validity, as evidenced by the high proportion of discretionary foods in the snacks, sweets, and mixed dishes food groups (sensitivity), and the low proportion of discretionary foods in the vegetable and fruit food groups (specificity). Additionally, more nutrient-poor foods, including a greater proportion of sweets, snacks, and fried vegetables, were classified as discretionary than as either hyperpalatable or ultra-processed, suggesting the discretionary food classification method may have greater precision with respect to capturing unhealthy foods. In contrast, the difference in nutrient density between non-discretionary versus discretionary foods was larger than the difference between non-ultra-processed versus ultra-processed and non-hyperpalatable versus hyperpalatable, suggesting the discretionary food classification method may also have greater precision with respect to excluding healthy foods. The inverse associations of the discretionary food score (reflecting percent energy intake from discretionary foods) with HEI-2020 total, moderation, and adequacy components in U.S. children and adults further support the construct alignment with respect to overall adherence to U.S. dietary guidelines.
Because foods may meet one or more of the discretionary food classification thresholds, the method can be used categorically (discretionary versus non-discretionary) or ordinally (number of thresholds met). Nutrient density was lower in discretionary versus non-discretionary foods and decreased linearly with increasing number of thresholds met. Future research could examine whether associations of discretionary food intake with health outcomes differ by the number of discretionary thresholds met.
The moderate association of the discretionary food score with the HEI-2020 moderation score suggests alignment with the DGA, but also indicates considerable non-overlapping variance between the two scores. The discretionary food score may more precisely reflect intake of nutrient-poor foods than the HEI-2020 moderation score,39 as it accounts for nutrients not fully captured in the HEI-2020 moderation score (e.g. unsaturated fat from chips and many sweets), does not truncate intake, and does not capture nutrients from core foods (e.g., saturated fat from low-fat milk or lean meat63). Future research elucidating whether these measures are differentially related to health outcomes such as cardiometabolic risk factors would be informative.
The discretionary food classification method was compared to food processing and hyperpalatable food classifications given expected overlap in these constructs. However, the three classifications are conceptually distinct. Although the hyperpalatable food definition is nutrient based, it specifies combinations of nutrients thought to increase palatability, rather than thresholds for single nutrients or food groups to limit.32 While the hyperpalatable food measure would be expected to effectively classify foods that are high in nutrients to limit, findings herein suggest that the thresholds do not discriminate effectively between discretionary and non-discretionary foods, as demonstrated by the classification of some non-discretionary foods as hyperpalatable (e.g. starchy vegetables and legumes cooked with salt) and the classification of some nutrient poor foods as non-hyperpalatable (e.g. ices, sorbets, and candy). In contrast to the hyperpalatable food measure, ultra-processed foods are classified based on processing without respect to nutrient value.25 Since not all ultra-processed foods are nutrient poor (e.g., industrially-produced whole grain bread),64 and not all discretionary foods are ultra-processed (e.g., homemade desserts), the discretionary food score may help inform the extent to which associations of ultra-processed food intake with health outcomes are attributable to their nutritional properties or their industrial processing.65,66
The discretionary food classification method has utility for research, practice, and policy.
The ability to classify individual foods as discretionary or not may facilitate clearer consumer guidance and could improve product labeling and clinical dietary education.14 For example, in 2016, the Chilean government mandated warning labels on foods using similar nutrient thresholds to the discretionary food classification. Evaluation of the policy showed that consumers decreased overall intake of sugar, sodium, and saturated fat,67,68 suggesting that clearly labeled foods led to eating behavior changes. In addition, the discretionary food classification method allows for monitoring and evaluating consumer purchasing behaviors and the availability of discretionary foods in the food supply. For example, the discretionary food classification can be used for surveilling the quality of restaurant foods. Finally, the discretionary food score can be applied across widely varying diets using objective thresholds based on dietary guidance and allows for comparison of findings across studies of diverse populations.
### Strengths and limitations
Strengths of this study include the use of objective criteria from established dietary guidelines to define discretionary foods and the application across the full U.S. food database. The use of the FNDDS for classifying discretionary foods is easily implemented in any study using NHANES WWEIA data or the National Cancer Institute Automated Self-Administered 24-hour (ASA24®).69 However, as the criteria were developed for FNDSS they will likely require adaptation for use in studies based on other food and nutrient databases. Lastly, while self-reported dietary intake data are susceptible to response bias, this may not impact comparisons of classifications of foods as reported (since classification is applied after food reporting), unless misreporting is differential by processing level or nutrient thresholds used by the classification methods. For example, if individuals misreport foods high in sugars, saturated fat, and sodium, then all classification methods would be similarly affected by misreporting.
## Conclusions
These findings support the utility of the discretionary food classification method for effectively identifying non-recommended foods with lower nutrient density and for quantifying discretionary food intake, which is inversely related to adherence to dietary guidelines in the U.S. population. By establishing nutrient-based criteria that effectively identify foods to limit, the discretionary food measure has utility for diet quality assessment, behavioral nutrition interventions, and food policy.
## Data Availability
All data produced in the present study are available upon reasonable request.
* Received March 20, 2025.
* Revision received March 20, 2025.
* Accepted March 20, 2025.
* © 2025, Posted by Cold Spring Harbor Laboratory
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license
## References
1. 1.U.S. Department of Agriculture and U.S. Department of Health and Human Services UaH. Dietary Guidelines for Americans, 2020-2025. 2020. December 2020. [DietaryGuidelines.gov](https://DietaryGuidelines.gov).
2. 2.Hu H, Zhao Y, Feng Y, et al. Consumption of whole grains and refined grains and associated risk of cardiovascular disease events and all-cause mortality: a systematic review and dose-response meta-analysis of prospective cohort studies. The American Journal of Clinical Nutrition. 2023/01/01/ 2023;117(1):149-159. doi:10.1016/j.ajcnut.2022.10.010
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajcnut.2022.10.010&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=36789934&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
3. 3.Graudal N, Jürgens G, Baslund B, Alderman MH. Compared With Usual Sodium Intake, Low-and Excessive-Sodium Diets Are Associated With Increased Mortality: A Meta-Analysis. American Journal of Hypertension. 2014;27(9):1129–1137. doi:10.1093/ajh/hpu028
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/ajh/hpu028&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=24651634&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
4. 4.Schulze MB, Manson JE, Ludwig DS, et al. Sugar-Sweetened Beverages, Weight Gain, and Incidence of Type 2 Diabetes in Young and Middle-Aged Women. JAMA. 2004;292(8):927–934. doi:10.1001/jama.292.8.927
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/jama.292.8.927&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=15328324&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
[Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000223429200017&link_type=ISI)
5. 5.Imamura F, O’Connor L, Ye Z, et al. Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes: systematic review, metaanalysis, and estimation of population attributable fraction. BMJ: British Medical Journal. 2015;351:h3576. doi:10.1136/bmj.h3576
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjE4OiIzNTEvanVsMjFfMTEvaDM1NzYiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyNS8wMy8yMC8yMDI1LjAzLjIwLjI1MzI0MzE3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==)
6. 6.Bhupathiraju SN, Tobias DK, Malik VS, et al. Glycemic index, glycemic load, and risk of type 2 diabetes: results from 3 large US cohorts and an updated meta-analysis123. The American Journal of Clinical Nutrition. 2014/07/01/ 2014;100(1):218-232. doi:10.3945/ajcn.113.079533
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiYWpjbiI7czo1OiJyZXNpZCI7czo5OiIxMDAvMS8yMTgiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyNS8wMy8yMC8yMDI1LjAzLjIwLjI1MzI0MzE3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==)
7. 7.Hunter RW, Dhaun N, Bailey MA. The impact of excessive salt intake on human health. Nature Reviews Nephrology. 2022/05/01 2022;18(5):321-335. doi:10.1038/s41581-021-00533-0
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41581-021-00533-0&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=35058650&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
8. 8.Breeze P, Sworn K, McGrane E, Abraham S, Cantrell A. Relationships between sodium, fats and carbohydrates on blood pressure, cholesterol and HbA1c: an umbrella review of systematic reviews. *BMJ Nutrition*, Prevention & Health. 2024:e000666. doi:10.1136/bmjnph-2023-000666
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NjoiYm1qbnBoIjtzOjU6InJlc2lkIjtzOjc6IjcvMS8xOTEiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyNS8wMy8yMC8yMDI1LjAzLjIwLjI1MzI0MzE3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==)
9. 9.Kris-Etherton PM, Petersen K, Van Horn L. Convincing evidence supports reducing saturated fat to decrease cardiovascular disease risk. BMJ Nutr Prev Health. 2018;1(1):23–26. doi:10.1136/bmjnph-2018-000009
[FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6NjoiYm1qbnBoIjtzOjU6InJlc2lkIjtzOjY6IjEvMS8yMyI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI1LzAzLzIwLzIwMjUuMDMuMjAuMjUzMjQzMTcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)
10. 10.Yang Q, Zhang Z, Gregg EW, Flanders WD, Merritt R, Hu FB. Added Sugar Intake and Cardiovascular Diseases Mortality Among US Adults. JAMA Internal Medicine. 2014;174(4):516–524. doi:10.1001/jamainternmed.2013.13563
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1001/jamainternmed.2013.13563&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=24493081&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
[Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000336841900008&link_type=ISI)
11. 11.DiNicolantonio JJ, Lucan SC, O’Keefe JH. The Evidence for Saturated Fat and for Sugar Related to Coronary Heart Disease. Progress in Cardiovascular Diseases. 2016/03/01/ 2016;58(5):464-472. doi:10.1016/j.pcad.2015.11.006
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.pcad.2015.11.006&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=2676342&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
[Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1989AT16800002&link_type=ISI)
12. 12.Sawicki CM, Jacques PF, Lichtenstein AH, et al. Whole-and Refined-Grain Consumption and Longitudinal Changes in Cardiometabolic Risk Factors in the Framingham Offspring Cohort. The Journal of Nutrition. 2021/09/01/ 2021;151(9):2790-2799. doi:10.1093/jn/nxab177
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/jn/nxab177&link_type=DOI)
13. 13.National Health and Medical Research Council. Australian Dietary Guidelines. 2013.
14. 14.Tapsell LC, Neale EP, Satija A, Hu FB. Foods, Nutrients, and Dietary Patterns: Interconnections and Implications for Dietary Guidelines. Advances in Nutrition. 2016/05/01/ 2016;7(3):445-454. doi:10.3945/an.115.011718
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoiYWR2YW5udXQiO3M6NToicmVzaWQiO3M6NzoiNy8zLzQ0NSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI1LzAzLzIwLzIwMjUuMDMuMjAuMjUzMjQzMTcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)
15. 15.Shams-White MM, Pannucci TE, Lerman JL, et al. Healthy Eating Index-2020: Review and Update Process to Reflect the Dietary Guidelines for Americans,2020-2025. Journal of the Academy of Nutrition and Dietetics. Sep 2023;123(9):1280-1288. doi:10.1016/j.jand.2023.05.015
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2023.05.015&link_type=DOI)
16. 16.Drewnowski A. The Nutrient Rich Foods Index helps to identify healthy, affordable foods1234. The American Journal of Clinical Nutrition. 2010/04/01/ 2010;91(4):1095S-1101S. doi:10.3945/ajcn.2010.28450D
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiYWpjbiI7czo1OiJyZXNpZCI7czoxMDoiOTEvNC8xMDk1UyI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI1LzAzLzIwLzIwMjUuMDMuMjAuMjUzMjQzMTcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)
17. 17.Mozaffarian D, El-Abbadi NH, O’Hearn M, et al. Food Compass is a nutrient profiling system using expanded characteristics for assessing healthfulness of foods. Nature Food. 2021/10/01 2021;2(10):809-818. doi:10.1038/s43016-021-00381-y
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s43016-021-00381-y&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=37117986&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
18. 18.Ortenzi F, Kolby M, Lawrence M, et al. Limitations of the Food Compass Nutrient Profiling System. The Journal of Nutrition. 2023/03/01/ 2023;153(3):610-614. doi:10.1016/j.tjnut.2023.01.027
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.tjnut.2023.01.027&link_type=DOI)
19. 19.Mauch CE, Golley RK, Hendrie GA. Variety Predicts Discretionary Food and Beverage Intake of Australian Adults: A Cross-Sectional Analysis of an Online Food Intake Survey. Journal of the Academy of Nutrition and Dietetics. 2024;124(4):509–520. doi:10.1016/j.jand.2023.07.019
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2023.07.019&link_type=DOI)
20. 20.James-Martin G, Baird DL, Hendrie GA. Strategies to Reduce Consumption of Unhealthy Foods and Beverages: Scenario Modeling to Estimate the Impact on the Australian Population’s Energy and Nutrient Intakes. Journal of the Academy of Nutrition and Dietetics. 2021/08/01/ 2021;121(8):1463-1483. doi:10.1016/j.jand.2020.12.003
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2020.12.003&link_type=DOI)
21. 21.Fayet-Moore F, McConnell A, Cassettari T, Tuck K, Petocz P, Kim J. Discretionary intake among Australian adults: prevalence of intake, top food groups, time of consumption and its association with sociodemographic, lifestyle and adiposity measures. Public Health Nutrition. 2019;22(9):1576–1589. doi:10.1017/S1368980018003361
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1017/S1368980018003361&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=30681049&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
22. 22.Spence AC, Campbell KJ, Lioret S, McNaughton SA. Early Childhood Vegetable, Fruit, and Discretionary Food Intakes Do Not Meet Dietary Guidelines, but Do Show Socioeconomic Differences and Tracking over Time. Journal of the Academy of Nutrition and Dietetics. 2018/09/01/ 2018;118(9):1634-1643.e1. doi:10.1016/j.jand.2017.12.009
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2017.12.009&link_type=DOI)
23. 23.Manohar N, Hayen A, Do L, Scott J, Bhole S, Arora A. Early life and socio-economic determinants of dietary trajectories in infancy and early childhood – results from the HSHK birth cohort study. Nutrition Journal. 2021/09/07 2021;20(1):76. doi:10.1186/s12937-021-00731-3
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1186/s12937-021-00731-3&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=34493286&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
24. 24.Australian Bureau of Statistics. Australian Health Survey: Users’ Guide, 2011–13— Discretionary Food List. 2014.
25. 25.Monteiro CA, Cannon G, Levy RB, et al. Ultra-processed foods: what they are and how to identify them. Public Health Nutrition. 2019;22(5):936–941. doi:10.1017/S1368980018003762
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1017/S1368980018003762&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=30744710&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
26. 26.Lane MM, Gamage E, Du S, et al. Ultra-processed food exposure and adverse health outcomes: umbrella review of epidemiological meta-analyses. BMJ. Feb 28 2024;384:e077310. doi:10.1136/bmj-2023-077310
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1136/bmj-2023-077310&link_type=DOI)
27. 27.Braesco V, Souchon I, Sauvant P, et al. Ultra-processed foods: how functional is the NOVA system? European Journal of Clinical Nutrition. 2022/09/01 2022;76(9):1245-1253. doi:10.1038/s41430-022-01099-1
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41430-022-01099-1&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=35314769&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
28. 28.Dicken SJ, Batterham RL, Brown A. Nutrients or processing? An analysis of food and drink items from the UK National Diet and Nutrition Survey based on nutrient content, the NOVA classification and front of package traffic light labelling. British Journal of Nutrition. 2024;131(9):1619–1632. doi:10.1017/S0007114524000096
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1017/S0007114524000096&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=38220223&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
29. 29.Liu J, Steele E, Karageorgou D, Micha R, Monteiro C, Mozaffarian D. Consumption of Ultra-Processed Foods and Diet Quality Among U.S. Adults and Children. Current Developments in Nutrition. 2020/06/01/ 2020;4:nzaa046_043. doi:10.1093/cdn/nzaa046_043
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/cdn/nzaa046_043&link_type=DOI)
30. 30.Rossato SL, Khandpur N, Lo C-H, et al. Intakes of Unprocessed and Minimally Processed and Ultraprocessed Food Are Associated with Diet Quality in Female and Male Health Professionals in the United States: A Prospective Analysis. Journal of the Academy of Nutrition and Dietetics. 2023;123(8):1140–1151.e2. doi:10.1016/j.jand.2023.03.011
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2023.03.011&link_type=DOI)
31. 31.Hess JM, Comeau ME, Casperson S, et al. Dietary Guidelines Meet NOVA: Developing a Menu for A Healthy Dietary Pattern Using Ultra-Processed Foods. The Journal of Nutrition. 2023/08/01/ 2023;153(8):2472-2481. doi:10.1016/j.tjnut.2023.06.028
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.tjnut.2023.06.028&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=37356502&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
32. 32.Fazzino TL, Rohde K, Sullivan DK. Hyper-Palatable Foods: Development of a Quantitative Definition and Application to the US Food System Database. Obesity. 2019/11/01 2019;27(11):1761-1768. doi:10.1002/oby.22639
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/oby.22639&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=31689013&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
33. 33.U.S. Department of Agriculture ARS. USDA Food and Nutrient Database for Dietary Studies 2017-2018. *Food Surveys Research Group Home Page*. 2020;
34. 34.Bowman SC, JC.; Friday, JE.; Moshfegh, AJ. Food Patterns Equivalents Database 2017-2018: Methodology and User Guide. 2020. [http://www.ars.usda.gov/nea/bhnrc/fsrg](http://www.ars.usda.gov/nea/bhnrc/fsrg)
35. 35.Centers for Disease Control and Prevention and National Center for Health Statistics. Data from: National Health and Nutrition Examination Survey Data 2017-2018. Hyattsville, MD.
36. 36.Ahluwalia N, Dwyer J, Terry A, Moshfegh A, Johnson C. Update on NHANES Dietary Data: Focus on Collection, Release, Analytical Considerations, and Uses to Inform Public Policy. Advances in Nutrition. 2016/01/01/ 2016;7(1):121-134. doi:10.3945/an.115.009258
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoiYWR2YW5udXQiO3M6NToicmVzaWQiO3M6NzoiNy8xLzEyMSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI1LzAzLzIwLzIwMjUuMDMuMjAuMjUzMjQzMTcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)
37. 37.National Center for Health Statistics. Ethics Review Board Approval. Accessed 3/6, 2025. [https://www.cdc.gov/nchs/nhanes/about/erb.html](https://www.cdc.gov/nchs/nhanes/about/erb.html)
38. 38.Raper N, Perloff B, Ingwersen L, Steinfeldt L, Anand J. An overview of USDA’s Dietary Intake Data System. Journal of food composition and analysis. 2004;
39. 39.Shams-White MM, Pannucci TE, Lerman JL, et al. Healthy Eating Index-2020: Review and Update Process to Reflect the Dietary Guidelines for Americans, 2020-2025. J Acad Nutr Diet. 2023/09/01/ 2023;123(9):1280-1288. doi:10.1016/j.jand.2023.05.015
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2023.05.015&link_type=DOI)
40. 40.U.S. Food and Drug Association F. Sodium in Your Diet. U.S. Food & Drug Administration Updated 03/05/2024. Accessed 7/20, 2024. [https://www.fda.gov/food/nutrition-education-resources-materials/sodium-your-diet](https://www.fda.gov/food/nutrition-education-resources-materials/sodium-your-diet)
41. 41.U.S. Food and Drug Association F. Daily Value and Percent Daily Value on the Nutrition and Supplement Facts Labels Updated 2024. Accessed 8/7, 2024. [https://www.fda.gov/food/nutrition-facts-label/daily-value-nutrition-and-supplement-facts-labels](https://www.fda.gov/food/nutrition-facts-label/daily-value-nutrition-and-supplement-facts-labels)
42. 42.Lloyd-Jones DM, Hong Y, Labarthe D, et al. Defining and Setting National Goals for Cardiovascular Health Promotion and Disease Reduction. Circulation. 2010;121(4):586–613. doi:doi:10.1161/CIRCULATIONAHA.109.192703
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTQ6ImNpcmN1bGF0aW9uYWhhIjtzOjU6InJlc2lkIjtzOjk6IjEyMS80LzU4NiI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI1LzAzLzIwLzIwMjUuMDMuMjAuMjUzMjQzMTcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)
43. 43.Sui Z, Wong WK, Louie JCY, Rangan A. Discretionary food and beverage consumption and its association with demographic characteristics, weight status, and fruit and vegetable intakes in Australian adults. Public Health Nutrition. 2017;20(2):274–281. doi:10.1017/S1368980016002305
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1017/S1368980016002305&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=27572276&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
44. 44.Aburto TC, Pedraza LS, Sánchez-Pimienta TG, Batis C, Rivera JA. Discretionary Foods Have a High Contribution and Fruit, Vegetables, and Legumes Have a Low Contribution to the Total Energy Intake of the Mexican Population. J Nutr. Sep 2016;146(9):1881s–7s. doi:10.3945/jn.115.219121
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6OToibnV0cml0aW9uIjtzOjU6InJlc2lkIjtzOjExOiIxNDYvOS8xODgxUyI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI1LzAzLzIwLzIwMjUuMDMuMjAuMjUzMjQzMTcuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)
45. 45.Batis C, Barrientos-Gutierrez T, Basto-Abreu A. Associated substitution and complementation patterns of processed discretionary foods and drinks on total energy and added sugar intake. Journal of Human Nutrition and Dietetics. 2023/10/01 2023;36(5):1942-1950. doi:10.1111/jhn.13175
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/jhn.13175&link_type=DOI)
46. 46.An R. Beverage Consumption in Relation to Discretionary Food Intake and Diet Quality among US Adults, 2003 to 2012. Journal of the Academy of Nutrition and Dietetics. 2016/01/01/ 2016;116(1):28-37. doi:10.1016/j.jand.2015.08.009
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2015.08.009&link_type=DOI)
47. 47.Liu Q, Wang L, Guo S, Allman-Farinelli M, Rangan A. Development and validation of an online tool to assess perceived portion size norms of discretionary foods. European Journal of Clinical Nutrition. 2023/08/01 2023;77(8):815-822. doi:10.1038/s41430-023-01290-y
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41430-023-01290-y&link_type=DOI)
48. 48.Liu J, Lee Y, Micha R, Li Y, Mozaffarian D. Trends in junk food consumption among US children and adults, 2001–2018. *The American Journal of Clinical Nutrition*. 2021/09/01/ 2021;114(3):1039-1048. doi:10.1093/ajcn/nqab129
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/ajcn/nqab129&link_type=DOI)
49. 49.Rangan A, Hector D, Randall D, Gill T, Webb K. Monitoring consumption of ‘extra’ foods in the Australian diet: Comparing two sets of criteria for classifying foods as ‘extras’. Nutrition & Dietetics. 2007;64(4):261–267. doi:10.1111/j.1747-0080.2007.00161.x
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/j.1747-0080.2007.00161.x&link_type=DOI)
50. 50.Rangan AM, Randall D, Hector DJ, Gill TP, Webb KL. Consumption of ‘extra’ foods by Australian children: types, quantities and contribution to energy and nutrient intakes. European Journal of Clinical Nutrition. 2008/03/01 2008;62(3):356-364. doi:10.1038/sj.ejcn.1602720
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/sj.ejcn.1602720&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=17356553&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
[Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000254289400008&link_type=ISI)
51. 51.Johnson BJ, Grieger JA, Wycherley TP, Golley RK. Theoretical Reductions in Discretionary Choices Intake via Moderation, Substitution, and Reformulation Dietary Strategies Show Improvements in Nutritional Profile: A Simulation Study in Australian 2-to 18-Year-Olds. Journal of the Academy of Nutrition and Dietetics. 2019/05/01/ 2019;119(5):782-798.e6. doi:10.1016/j.jand.2018.10.016
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2018.10.016&link_type=DOI)
52. 52.Nansel TR, Schwedhelm C, Lipsky LM, Faith MS, Siega-Riz AM. Socioeconomic Characteristics and the Home Food Environment Are Associated With Feeding Healthful and Discretionary Foods During the First Year of Life in the Pregnancy Eating Attributes Study. Journal of the Academy of Nutrition and Dietetics. 2025;125(2):228–238.e1. doi:10.1016/j.jand.2024.05.011
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2024.05.011&link_type=DOI)
53. 53.Food and Nutrition Service. WIC Food Packages - Regulatory Requirements for WIC-Eligible Foods. Accessed 01/15, 2024. [https://www.fns.usda.gov/wic/food-packages/regulatory-requirements](https://www.fns.usda.gov/wic/food-packages/regulatory-requirements)
54. 54.Food and Nutrition Service. *Grain Requirements for the National School Lunch Program and School Breakfast Program*. Vol. 2024. Policy Memo. 2012. [https://www.fns.usda.gov/cn/grain-requirements-national-school-lunch-program-and-school-breakfast-program](https://www.fns.usda.gov/cn/grain-requirements-national-school-lunch-program-and-school-breakfast-program)
55. 55.Food and Nutrition Service. Grain Requirements in the CACFP: Questions and Answers. United States Department of Agriculture (USDA). Accessed 01/14, 2024. [https://www.fns.usda.gov/cacfp/grain-requirements-cacfp-questions-and-answers](https://www.fns.usda.gov/cacfp/grain-requirements-cacfp-questions-and-answers)
56. 56.Steele EM, O’Connor LE, Juul F, et al. Identifying and Estimating Ultraprocessed Food Intake in the US NHANES According to the Nova Classification System of Food Processing. The Journal of Nutrition. 2023/01/01/ 2023;153(1):225-241. doi:10.1016/j.tjnut.2022.09.001
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.tjnut.2022.09.001&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=36913457&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
57. 57.National Cancer Institute Division of Cancer Control & Population Sciences. Linking ASA24 and NHANES Data with the Nova Classification System for Industrial Food Processing. [https://epi.grants.cancer.gov/nova/](https://epi.grants.cancer.gov/nova/)
58. 58.Kennedy ET, Ohls J, Carlson S, Fleming K. The Healthy Eating Index: Design and Applications. Journal of the American Dietetic Association. 1995/10/01/ 1995;95(10):1103-1108. doi:10.1016/S0002-8223(95)00300-2
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0002-8223(95)00300-2&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=7560680&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
[Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1995RY28200006&link_type=ISI)
59. 59.Program. NCIDoCCPSEaGR. Healthy Eating Index SAS Code. Accessed 4/10, 2024. [https://epi.grants.cancer.gov/hei/sas-code.html](https://epi.grants.cancer.gov/hei/sas-code.html)
60. 60.Fulgoni VL, Keast DR, Drewnowski A. Development and Validation of the Nutrient-Rich Foods Index: A Tool to Measure Nutritional Quality of Foods. The Journal of Nutrition. 2009/08/01/ 2009;139(8):1549-1554. doi:10.3945/jn.108.101360
[Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6OToibnV0cml0aW9uIjtzOjU6InJlc2lkIjtzOjEwOiIxMzkvOC8xNTQ5IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjUvMDMvMjAvMjAyNS4wMy4yMC4yNTMyNDMxNy5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=)
61. 61.*SAS Enterprise Miner 13.1*. SAS Institute Inc.;
62. 62.Akinbami LJ CT, Davy O, Ogden CL, Fink S, Clark J, et al. *National Health and Nutrition Examination Survey*, *2017–March 2020 prepandemic file: Sample design, estimation, and analytic guidelines*. Vol. 2(190). 2022. Vital Health Stat.
63. 63.Täger M, Peltner J, Thiele S. Evaluation of diet quality by means of the Healthy Eating Index and its modified variants. Ernährungs Umschau. 2016;63:110–118.
64. 64.Derbyshire E. Are all ‘ultra-processed’ foods nutritional demons? A commentary and nutritional profiling analysis. Trends in Food Science & Technology. 2019/12/01/ 2019;94:98-104. doi:10.1016/j.tifs.2019.08.023
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.tifs.2019.08.023&link_type=DOI)
65. 65.Astrup A, Monteiro CA, Ludwig DS. Does the concept of “ultra-processed foods” help inform dietary guidelines, beyond conventional classification systems? NO. The American Journal of Clinical Nutrition. 2022/12/01/ 2022;116(6):1482-1488. doi:10.1093/ajcn/nqac123
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/ajcn/nqac123&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=35670128&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
66. 66.Visioli F, Del Rio D, Fogliano V, Marangoni F, Ricci C, Poli A. Ultra-processed foods and health: are we correctly interpreting the available evidence? European Journal of Clinical Nutrition. 2024/09/26 2024;doi:10.1038/s41430-024-01515-8
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s41430-024-01515-8&link_type=DOI)
67. 67.Taillie LS, Bercholz M, Popkin B, Rebolledo N, Reyes M, Corvalán C. Decreases in purchases of energy, sodium, sugar, and saturated fat 3 years after implementation of the Chilean food labeling and marketing law: An interrupted time series analysis. PLOS Medicine. 2024;21(9):e1004463. doi:10.1371/journal.pmed.1004463
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pmed.1004463&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=39331649&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)
68. 68.Taillie LS, Bercholz M, Popkin B, Reyes M, Colchero MA, Corvalán C. Changes in food purchases after the Chilean policies on food labelling, marketing, and sales in schools: a before and after study. The Lancet Planetary Health. 2021;5(8):e526–e533. doi:10.1016/S2542-5196(21)00172-8
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S2542-5196(21)00172-8&link_type=DOI)
69. 69.Subar AF, Kirkpatrick SI, Mittl B, et al. The Automated Self-Administered 24-hour dietary recall (ASA24): a resource for researchers, clinicians, and educators from the National Cancer Institute. J Acad Nutr Diet. Aug 2012;112(8):1134–7. doi:10.1016/j.jand.2012.04.016
[CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jand.2012.04.016&link_type=DOI)
[PubMed](http://medrxiv.org/lookup/external-ref?access_num=22704899&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2025%2F03%2F20%2F2025.03.20.25324317.atom)